Anopheles funestus is one of the major vectors of malaria in Africa. Cytogenetic studies conducted on populations from West Africa have shown variable degrees of polymorphism with a genetic structure leading to the description of two chromosomal forms called “Folonzo” and “Kiribina” that exhibit limited gene flow. Because studies on allopatric populations showed bionomical heterogeneities, the present study was undertaken during three consecutive years (2006, 2007, and 2008) in an area of sympatry in Senegal, in order to assess their bionomical characteristics and compare their epidemiologic role in malaria transmission. Overall, the two forms coexisted in the study area; the Kiribina form being more abundant and exhibiting higher biting rates. Based on an enzyme-linked immunosorbent assay, the anthropophilic rates were statistically comparable and were, respectively, 30.7% and 28.6% for Kiribina and Folonzo. Plasmodium falciparum circumsporozoite rates were also comparable and were 2.7% for Kiribina and 3.1% for Folonzo. Both forms were involved in malaria transmission; Kiribina being responsible for 68% of transmission. Thus, due to the limited gene flow between the two forms, the introduction of transgene for Plasmodium resistance in one of the two forms could be a disadvantage for the implementation of control strategies based on the use of Plasmodium-refractory genetically modified individuals. Nevertheless, it could represent an advantage limiting the insurgence of insecticide resistance gene spread between forms and should be taken into account for the implementation of control strategies.
Anopheles funestus is among the major vectors of malaria in sub-Saharan Africa. It has a continent-wide distribution that extends from northern Sudan to South Africa and from the west to the east cost and is very common on the islands of Madagascar. It is a very efficient vector of human plasmodia throughout its distribution area because of its close association with human environments (Gillies and de Meillon 1968). In some localities in tropical Africa, the vectorial capacity of this species can often exceed that of An. gambiae (Fontenille et al. 1997, Fontenille and Simard 2004, Rajaonarivelo et al. 2004). The re-emergence of An. funestus has been stressed in areas where its populations have disappeared (Labbo et al. 2004, Elissa and Karch 2005, Dia et al. 2008). Thus, any comprehensive control method against malaria vectors cannot ignore this species. Its polytene chromosomes were first described by Green and Hunt (1980) from material collected mainly in southern and East Africa. Several other cytogenetic studies conducted thereafter, mainly in West Africa, showed variable degrees of polymorphism in Burkina Faso (Boccolini et al. 1994), in Mali (Boccolini et al. 1998), and in Senegal (Dia et al. 2000), with a genetic structure among this species leading to the description of two distinct chromosomal forms, “Folonzo” and “Kiribina,” with limited gene flow and contrasting degrees of chromosomal polymorphism (Costantini et al. 1999, Guelbeogo et al. 2005). Further evidence for slight but significant genetic differentiation between the two chromosomal forms was found using 16 microsatellite loci dispersed over the whole chromosomal complement and the ND5 region of mtDNA (Michel et al. 2005). Despite the obvious genetic differentiation of the two forms, conflicting results were observed in allopatric areas where these forms are present. In fact, while Costantini et al. (1999) showed apparent association of inversions on 2R and 3R arms with behavioral heterogeneities of epidemiological relevance, extensive recent studies carried out by Guelbeogo et al. (2005) in other sites in Burkina Faso did not find statistical differences in the feeding habits among carriers of different chromosomal arrangements. The behavioral aspects of the two forms seem to be related to the availability of alternative hosts. In addition, allopatric observations of the two forms in Senegal showed that the Kiribina form is more anthropophilic and has higher infection rates than the Folonzo form (Fontenille et al. 1997, Dia et al. 2003), whereas in Burkina Faso, the Kiribina form appear to be less anthropophilic and endophilic than the Folonzo form. The present study was undertaken during three consecutive years in an area where the two forms are sympatric and synchronous in order to assess their bionomical characteristics and to compare their roles in malaria transmission.
MATERIALS AND METHODS
The present study was carried out in the village of Kouvar (13°23'N, 13°37'W) situated in eastern Senegal, in an area of sudanian domain, 3 km from The Gambia River. The rainy season lasts from June to October. Kouvar, with 1,100 inhabitants, is situated 500 m from a permanent pool that permits the development of An. funestus aquatic stages during the last part of the rainy season. The vegetation is composed of shrubs and open wooded savanna replacing the climatic forest that formerly covered the area and whose vestiges are left in relic areas. Degradation of the natural vegetation is due to the combined action of persistent rainfall deficit and the extension of cultivated areas.
Mosquito collections and field processing
Adult mosquitoes were collected in October and November during three consecutive years in 2006, 2007, and 2008 using the pyrethrum spray catch (PSC) method in eight selected bedrooms from 15:00 to 17:00 and aspiration in two different pit shelters. After mosquito collections, anopheline vectors were identified morphologically and the ovaries from the half-gravid females of An. funestus were dissected and immediately preserved in Carnoy's fixative (three parts pure ethanol: one part glacial acetic acid). They were then held at ambient temperature in the field for 12 to 36 h and stored at -20° C in the laboratory until processing. The blood meals were blotted onto filter paper to determine the host source. The rest of the mosquitoes were stored individually in numbered vials with desiccant for laboratory processing.
The origin of the blood meals was identified as human, bovine, ovine, and horse using an enzyme-immunosorbent assay (ELISA). The crushed head-thoraces were tested by ELISA for Plasmodium falciparum circumsporozoite (CS) antigen detection. Polytene chromosomes preparations from the ovarian nurse cells were carried out according to the standard method of Hunt (1973). The paracentric chromosomal inversions were then identified by microscopic examination under a phase-contrast microscope and scored following the nomenclature of Sharakhov et al. (2004). Each individual was subsequently assigned to the corresponding chromosomal forms according to the algorithm developed by Guelbeogo et al. (2005).
The frequency of each chromosomal form was determined after cytogenetic identification. The anthropophilic rate (AR) of each form was calculated as the proportion of human blood to the total meals determined for all specimens belonging to each form. The circumsporozoite protein rate (CSR) was calculated the same way for all cytogenetically identified specimens as the proportion of mosquitoes found to contain the CS protein of P. falciparum. The man biting rate (MBR) was inferred from the PSC by dividing the number of mosquitoes fed on humans by the mean number of sleepers in bedrooms where these mosquitoes were collected. The entomological inoculation rate (EIR) for each form was calculated as the product of the man biting rate (MBR) and the CSR. For statistical analysis, the anthropophilic and circumsporozoite rates between Kiribina and Folonzo forms were compared using the chi-square test in Epi Info version 6.04 (CDC, Atlanta GA, U.S.A.) with P values <0.05 considered as significant.
A total of 772 anopheline mosquitoes, represented by An. gambiae s.l. and An. funestus, was collected in 2006 (260 specimens), in 2007 (180) specimens, and in 2008 (332 specimens), by the PSC method. An. funestus represented, respectively, 43.5%, 60.6%, and 66.3% in 2006, 2007, and 2008. A significant difference was observed in the abundance of An. funestus among the three years (c2=32.03, df=2, p<0.0001).
Observations on polytene chromosomes were carried out on 276 An. funestus half-gravid females. The whole polytene complement was read in 213 specimens that were all assigned to one of the two An. funestus chromosomal forms. During the three years of study, the Kiribina form was more abundant (69.5%) than the Folonzo form. The respective frequencies of the former were 59% in 2006, 79.6% in 2007, and 68.3% in 2008. The frequencies of the two forms between the three years were not statistically different (c2=4.73, df=2, p=0.09). Only five An. funestus were collected in two pit shelters in 2006. Three karyotyped specimens were assigned to the Kiribina form.
The indoor resting densities of endophilic females for each form were estimated. The mean number of females per room per day (FRD) was 9.2 during the three-year study. For 2006, 2007, and 2008, the mean number of FRD was higher in the Kiribina form than in the Folonzo form (Figure 1). The corresponding values were 4.2, 5.4, and 9.4 FRD for the Kiribina form and 2.9, 1.4, and 4.4 FRD for the Folonzo form respectively, in 2006, 2007, and 2008.
Trophic preferences and anthropophilic rate
A total of 183 blood meals was tested by ELISA to determine the host source, 127 from Kiribina and 56 from Folonzo. The two forms had taken their blood meals from human, bovine, ovine, and equine hosts in 2006, 2007 and in 2008 (Table 1). Globally, the mean anthropophilic rate was 28.6% (CI95%=18.4–41.5) and 30.7% (CI95%=23.4–39.2), respectively, for the Folonzo and Kiribina forms during the three-year study. These frequencies were comparable (c2=0.08, df=1, p=0.77). No significant differences were observed between the anthropophilic rates of the two forms in 2006 (c2=0.02, df=1, p=0.90), in 2007 (Fisher exact test p=0.51), and in 2008 (c2=0.27, df=1, p=0.60). The anthropophilic rates for each form were comparable among the three years (c2=3.7, df=2, p=0.16 for Kiribina) and (c2=1.73, df=2, p=0.42 for Folonzo).
Table 1. Number (No.) and percentage (%) of Kiribina and Folonzo forms fed on each vertebrate host among resting specimens.
Circumsporozoite protein and entomological inoculation rates
The crushed heads and thoraces of 212 karyotyped specimens (65 Folonzo and 147 Kiribina) were processed by ELISA for P. falciparum circumsporozoite CS antigen detection. Overall, 3.1% (CI95%=0.8–11) and 2.7% (CI95%=1.1–6.8) of Folonzo and Kiribina forms were positive, respectively (Table 2). No significant difference was observed between the two forms (Fisher exact test p=0.60).
Table 2. Circumsporozoite protein and entomological inoculation rate calculated by enzyme-linked immunosorbent assay for P. falciparum for each chromosomal forms.
Nt: number of specimens tested by ELISA, CSR: circumsporozoite rate, EIR: entomological inoculation rate.
The mean entomological inoculation rate EIR due to An. funestus was estimated to be 4.12 infected bites (Tables 2 and 3). The Kiribina form was responsible for 68% and Folonzo for 32% of this transmission.
Table 3. Human-biting rate of the Kiribina and Folonzo forms during the rainy seasons 2006–2008.
No. houses sampled
MBR*: man-biting rate adjusted with the anthropophilic rate of each chromosomal form.
The chromosomal analysis of Anopheles funestus populations from Kouvar village during three consecutive years confirmed the synchronous and sympatric presence of the two chromosomal forms already described in this species (Lochouarn et al. 1998, Dia et al. 2000). This study also established that the presence of these two forms is not a transitory phenomenon (Guelbeogo et al. 2005), as each form was collected during each of the three years of study. Our study is the first to study the bionomical characteristics of the two chromosomal forms in an area of sympatry. It was conducted in October and November during the three years targeted because previous studies made in the village had shown that the highest densities of An. funestus are observed during this period that corresponds to the end of the rainy season (Lochouarn et al. 1998, Dia et al. 2000).
Globally, the Kiribina form was predominant during the three years of study. Indeed, the predominance of Kiribina form is very common in An. funestus populations from savanna areas (Lochouarn et al. 1998, Costantini et al. 1999, Dia et al. 2000) and may reflect a stability of climatic conditions in the Kouvar village, as the same trend was observed during each of the three years. A recent study carried out in sympatric areas of Burkina Faso has shown that the frequencies of the two forms can be different over time (Guelbeogo et al. 2009). A longitudinal survey during the rainy season and the early- to mid-dry season could shed light on this phenomenon.
The bionomic analysis showed that the two chromosomal forms exhibit comparable anthropophilic and infection rates and are both involved in malaria transmission; Kiribina was more involved due to its higher biting rate. However, significant bionomical differences were observed between allopatric populations of the two chromosomal forms. Indeed, Costantini et al. (1999) showed that Kiribina form has a lower anthropophilic and a less marked endophilic infection rate, whereas the Folonzo form was characterized by a higher vectorial capacity. In Senegal, conflicting observations have been observed, with a higher anthropophilic rate in Kiribina form (Dia et al. 2001).
Taking into account the low number of specimens collected in pit shelters, it is probable that the two forms are endophilic; Folonzo being more endophilic than Kiribina. If this hypothesis is verified, such observations have important epidemiological implications. In fact, vector control strategies using IRS as well as ITNs should be useful to significantly reduce malaria transmission by these forms. However, from an epidemiological standpoint, our findings have not been able to demonstrate a clear difference of the two forms in their ability to transmit malaria parasites. The relative abundance of the two forms may reflect different aquatic breeding conditions (Guelbeogo et al. 2009), as well as adult population dynamics that could affect the transmission of malaria as already observed in An. gambiae by Toure et al. (1998). Thus, for genetic control purposes, the introduction of transgenes for Plasmodium resistance in one of the two forms could be a disadvantage for gene spread due to the limited gene flow between the two forms and, on the other hand, a limit for insecticide resistant gene spread between forms. Hence, any strategy for controlling malaria transmission by An. funestus chromosomal forms should take these observations into account.
We thank the inhabitants of Kouvar for their collaboration. This research was financially supported by WHO MIM/TDR grant 50090 and the Institut Pasteur de Dakar.