Breeding Sites of Culex quinquefasciatus (Say) during the Rainy Season in Rural Lowland Rainforest, Rivers State, Nigeria

Breeding sites of Culex quinquefasciatus (Say) were investigated during a World Bank Assisted Project on Integrated Vector Management (IVM ) for malaria control at 5 v illages (Ipo, Ozuaha, Oman wa, Omademe, Ub ima) in the Ikwerre Local Govern ment Area (LGA ). The study was in the rainy season, June-September, 2009, in ru ral lowland rain- forest, Rivers State, Nigeria. Samp ling was undertaken daily in a randomly selected (100 m x 100 m) grid in each village. Nu mber of Cx. quinquefasciatus (Say) immatures was appro ximately 6-fold that of An. gambiae s.l. Nearly 80% of Cx. quinquefasciatus immatures were fro m container-type breeding sites (metal, p lastic containers, "calabashes", tyres) and least fro m phytotelmata. The differences were significant (FCal=35695636; FTab=39.86; d f 1, 1; p 0.01). In the phytotelmata breeding site-types, about 67% of Cx. quinquefasciatus immatures were fro m depressions on trees; the difference in total numbers among site-types was significant (FCal=2.47x10 8 ; FTab=39.86; df 1,1; p<0.01). The container-type breeding sites yielded approximately 90% of immatures at Omanwa, but decreased to 60% at Omademe. In contrast, nearly 70% of immatures were fro m pools at Ozuaha and 80% fro m phytotelmata at Ub ima. These variat ions were significant (FCal=4305.94; F Tab=39.86; df 1,1; p<0.01). The results indicate that Cx. quinquefasciatus had invaded rural areas and their breed ing sites ubiquitous and variable among villages, highlighting the needs for studies locally, prior to larviciding.


Introduction
It is estimated that globally, 1.2 billion people are at risk to ly mphatic filariasis (LF), major cause of acute and chronic mo rb id ity affecting hu mans in trop ical and subtrop ical count ries [1,2]. The absen ce o f a no nhu man reservo ir, availability of safe, single-dose, two-drug treatment regimes, capable o f reducing micro filariae to very lo w levels and remarkable imp rovement in techniques for diagnosing the disease, resulted in advocacy for a global strategy to eliminate filariasis through mass drug administration (M DA) [3]. This culminated in a global alliance by the World Health Organ izat ion and other agencies to eliminate ly mph at ic filariasis by 2020 [4]. The main goal is to break the cycle of transmission of the parasites, between mosquitoes and humans mainly through MDA with Albendazole in co mbination with either Ivermect in or Diethylcarbamazine citrate [5]. Although there had been p rogress since the init iat ion of MDA Programmes, challenges have emerged. These include: the inability of some countries to sustain MDA, non-availability of data to indicate level and duration of treat ments to eliminate LF, and a shift towards linking MDA for LF control and other neglected diseases [6]. These challenges have led to gro wing concerns on the effect iveness of MDA alone to eliminate LF, without vector control as a co mplement [7]. Thus an integrated strategy involving vector control is now thought to have great potential to beco me an important supplementary component of the filariasis elimination campaign.
Culex quinquefasciatus (Say) is an urban vector of nocturnally periodic Wuchereria bancrofti in West Africa [8]. The distribution of the urban vector of bancroftian filariasis, Cx. quinquefasciatus is expanding with urbanizat ion; many rural areas that were free of this vector are now being colonized [9]. Reliable data on breeding sites of species ensure the effectiveness of new control methods. Nicolas et al. [10] controlled this species, by applying Bacillus sphaericus against the immatures at breeding sites; they estimated that this approach would be more cost effective than the use of chemical insecticides. There had been reports of cases of bancroftian filariasis in rural Local Govern ment Areas (LGAs) adjacent to the Ikwerre LGA of Rivers State, Nige-ria [11,12]. Records indicate that patients had been treated for filariasis in the General hospital at Isiokpo, Ikwerre LGA headquarters. Studies were therefore undertaken, June-September, 2009 (Rainy season) in 5 villages (Ozuaha, Ipo, Omanwa, Omademe, Ub ima) in Ikwerre LGA , Rivers State, Nigeria. These villages were selected by the State Government for a World Bank-assisted project on Integrated Vector Management (IVM ) for malaria control. Studies on Cx. quinquefaciatus were undertaken simultaneously with those on Anopheles spp. to determine whether Cx. quinquefasciatus described as primarily urban (service 2008) had invaded rural areas and to identify their breed ing sites.

Methods
A 100 x 100m grid was randomly selected in each village. Potential breed ing sites were samp led with a 100ml-laddle and a Pasteur pipette for sites with large and small volu mes of water respectively, daily for 4 months, June-September, 2009. These sites included Phytotelmata (water-containing holes in plants), Pools (puddles, gutters) and containers (water receptacles in villages, such as the calabash, fro m the fruit of Cresenta cujeta Linn, Bignonaceae), plastic and metallic containers. Culex larvae were identified by the descriptions in Service [13]. Immatures were reared to adults to confirm that they were those of Cx. quinquefasciatus as per the keys of Gilles and de Meillon [14] and Gillett [15]. Anopheles gambiae s.l. immatures were also collected and reared to adults during the study. Adults of An.gambiae s.l. were identified by the keys of Holstein [16], Gillett [15] and Louis [17]. Nu mbers of immatures were recorded as per breeding site-types and villages. Analysis of Variance with one observation per cell was used for statistical analyses. Calculated F values were co mpared to tabulated F values to determine significance.

Di versity and Yiel d of Breeding Site-Types
The total number of Cx. quinquefasciatus immatures was approximately 6-fold that of An. gambiae s.l. immatures.
Nearly 80% of all immatures were co llected fro m container-type breeding sites (metal and plastic containers, "calabash", and tyres) and least (about 5%) fro m phytotelmata (Table1). The differences were significant (F Cal =35695636; F Tab =39.86; df 1,1; p<0.01). The highest number of immatures per breeding site was also in the container category (Table 2); the differences were significant (F Cal =16.5; df 1,3; p<0.05). Approximately 80% were fro m plastic containers and least 1% fro m tyres in the container-type breeding sites. The d ifferences were significant (F Cal =4.7 x 10.8; F Tab =8.53; df 1,2; p<0.01). The highest number of immatures per breeding site was fro m calabash in the container-type ( In the phytotelmata breed ing site-type, about 67% of immatures were co llected fro m depressions on trees; the difference in total nu mbers among site-types was significant (F Cal =2.47 x 10 8 ; F Tab =8.53; df 1,1; p<0.01).

Variati on in Site-Type Utilizati on across Villages
At Ipo and Omanwa, about 90% of immatures were fro m container-type breeding sites; this figure decreased to 60% at Omademe. In contrast, at Ozuaha, nearly 70% o f immatures were fro m pools, while at Ubima mo re than 80% were fro m phytotelmata. The variat ions in breeding site preference patterns among villages were significant (F Cal =4305.94; F Tab =39.86; d f 1,1; p<0.01).

Discussion
The high numbers of total Cx. quinquefasciatus larvae collected in the villages, indicated that the species, once considered an urban species [18,13] is increasing in distribution and colonizing rural pockets that were once free of this mosquito, an aspect of Cx. quinquefasciatus distribution that was also observed by Chavasse et al. [9]. In urban areas, the typical breeding sites were described as stagnant polluted water and in rural areas, mainly privies (pit latrines) [18,13]. The extensive distribution of breeding sites (tyres, plastic and metal containers, calabashes, gutters, puddles, leaf axils, depressions on trees, etc) is an indication of changes in the oviposition behaviour of Cx. quinquefasciatus; pit latrines, typical of polluted habitats did not occur in the randomly selected grids. The dominance of containers as breeding sites of Cx. quinquefasciatus had been reported by Hardine et al. [19] and Bockarie et al. [8].
Potential capacity to enhance offspring survivorship and fecundity contributes by definit ion, to parental care or juvenile dispersal; offspring survival and growth may depend strongly on the quality of the habitat in which they are deposited. Thus, when potential habitats vary in their suitability for juveniles, females are expected to choose habitats that maximize fitness [20]. In mosquitoes, such oviposition habitat selection was demonstrated in response to physical and chemical suitability for larval develop ment [21]. This may account for the variation in preference among metallic, plastic containers and, calabashes. The low numbers of immatures in phytotelmata site-types may be associated with habitat size and limited resource availability [22][23][24] .
Although container-type sites were preferred at Omanwa, Ipo and Omademe, pools and phytotelmata were the selected sites at Ozuaha and Ubima respectively. The pronounced preference for containers at Ipo and Omanwa might be attributed to the extreme rural conditions in the villages, lim-ited access to running water and therefore ut ilization of containers for collecting and storing rainwater. Un fortunately, these containers served as mosquito breeding sites. The dominance of phytotelmata at Ubima might be attributed to the extensive occurrence of plantain/banana groves in the randomly selected grids. These observations indicate that variations are co mmon and therefore breeding site preference studies should be undertaken in each locality before the commencement of larviciding. The ubiquitous nature of Cx. quinquefasciatus breeding sites is highlighted by its high percent occurrence in all site-type categories.

Conclusions
Cx. quinquefasciatus, long considered an urban species has invaded the rural lowland rainforest. Most immatures were collected fro m container-type breeding sites (metal and plastic containers, "calabash", and tyres) and least from phytotelmata. The h ighest numbers of immatures were collected fro m plastic containers and calabashes. Degree of breeding-site utilizat ion varied across villages.