Prominent intraspecific genetic divergence within Anopheles gambiae sibling species triggered by habitat discontinuities across a riverine landscape.

The Anopheles gambiae complex of mosquitoes includes malaria vectors at different stages of speciation, whose study enables a better understanding of how adaptation to divergent environmental conditions leads to evolution of reproductive isolation. We investigated the population genetic structure of closely related sympatric taxa that have recently been proposed as separate species (An. coluzzii and An. gambiae), sampled from diverse habitats along the Gambia river in West Africa. We characterized putatively neutral microsatellite loci as well as chromosomal inversion polymorphisms known to be associated with ecological adaptation. The results revealed strong ecologically associated population subdivisions within both species. Microsatellite loci on chromosome-3L revealed clear differentiation between coastal and inland populations, which in An. coluzzii is reinforced by a unusual inversion polymorphism pattern, supporting the hypothesis of genetic divergence driven by adaptation to the coastal habitat. A strong reduction of gene flow was observed between An. gambiae populations west and east of an extensively rice-cultivated region apparently colonized exclusively by An. coluzzii. Notably, this 'intraspecific' differentiation is higher than that observed between the two species and involves also the centromeric region of chromosome-X which has previously been considered a marker of speciation within this complex, possibly suggesting that the two populations may be at an advanced stage of differentiation triggered by human-made habitat fragmentation. These results confirm ongoing ecological speciation within these most important Afro-tropical malaria vectors and raise new questions on the possible effect of this process in malaria transmission.

Preliminary studies developing methods for the control of Chrysomya putoria, the African latrine fly, in pit latrines in The Gambia.

OBJECTIVE: To explore ways of controlling Chrysomya putoria, the African latrine fly, in pit latrines. As pit latrines are a major source of these flies, eliminating these important breeding sites is likely to reduce village fly populations, and may reduce the spread of diarrhoeal pathogens. METHODS: We treated 24 latrines in a Gambian village: six each with (i) pyriproxyfen, an insect juvenile hormone mimic formulated as Sumilarv(®) 0.5 G, a 0.5% pyriproxyfen granule, (ii) expanded polystyrene beads (EPB), (iii) local soap or (iv) no treatment as controls. Flies were collected using exit traps placed over the drop holes, weekly for five weeks. In a separate study, we tested whether latrines also function as efficient flytraps using the faecal odours as attractants. We constructed six pit latrines each with a built-in flytrap and tested their catching efficiency compared to six fish-baited box traps positioned 10 m from the latrine. Focus group discussions conducted afterwards assessed the acceptability of the flytrap latrines. RESULTS: Numbers of emerging C. putoria were reduced by 96.0% (95% CIs: 94.5-97.2%) 4-5 weeks after treatment with pyriproxyfen; by 64.2% (95% CIs: 51.8-73.5%) after treatment with local soap; by 41.3% (95% CIs = 24.0-54.7%) after treatment with EPB 3-5 weeks after treatment. Flytraps placed on latrines collected C. putoria and were deemed acceptable to local communities. CONCLUSIONS: Sumilarv 0.5 G shows promise as a chemical control agent, whilst odour-baited latrine traps may prove a useful method of non-chemical fly control. Both methods warrant further development to reduce fly production from pit latrines. A combination of interventions may prove effective for the control of latrine flies and the diseases they transmit.

Localized breeding of the Anopheles gambiae complex (Diptera: Culicidae) along the River Gambia, West Africa.

A study was undertaken to identify the major larval habitats of the Anopheles gambiae (Giles) complex in rural Gambia. Mosquito larvae and pupae were sampled along transects and in specific habitats in the central region of the country during the rainy seasons of 1996 and 1997. The sampling showed that the major breeding sites were located on the flooded alluvial soils bordering the river. The largest numbers of larvae were found during September, one month after the peak rains. Polymerase chain reaction analysis of specimens showed that Anopheles melas (Theobald) was the dominant species in the flooded areas (81.5%), followed by A. gambiae sensu stricto (Giles) (18.0%) and A. arabiensis (Patton) (0.5%). By sampling in specific habitats it was evident that A. arabiensis was mainly breeding in rain-fed rice fields along the edge of the alluvial soils. Anopheles melas and A. gambiae s.s. often coexisted but whereas A. melas were found in water with a salinity of up to 72% sea water (25.2 g NaCl l(-1)), A. gambiae s.s. only occurred in water with up to 30% sea water (10.5 g NaCl l(-1)). Anopheles melas larvae were found in association with plant communities dominated by sedges and grasses (Eleocharis sp., Paspalum sp., Sporobolus sp.) and sea-purslane Sesuvium portulacastrum (L.) and the presence of cattle hoof prints, whereas A. gambiae s.s. larvae mainly occurred in association with Paspalum sp. and Eleocharis sp. The study showed that even during the peak rainy season, breeding of the A. gambiae complex is almost entirely restricted to the extensive alluvial areas along the river.

Changes in house design reduce exposure to malaria mosquitoes.

House design may affect an individual's exposure to malaria parasites, and hence to disease. We conducted a randomized-controlled study using experimental huts in rural Gambia, to determine whether installing a ceiling or closing the eaves could protect people from malaria mosquitoes. Five treatments were tested against a control hut: plywood ceiling; synthetic-netting ceiling; insecticide-treated synthetic-netting ceiling (deltamethrin 12.5 mg/m2); plastic insect-screen ceiling; or the eaves closed with mud. The acceptability of such interventions was investigated by discussions with local communities. House entry by Anopheles gambiae, the principal African malaria vector, was reduced by the presence of a ceiling: plywood (59% reduction), synthetic-netting (79%), insecticide-treated synthetic-netting (78%), plastic insect-screen (80%, P < 0.001 in all cases) and closed eaves (37%, ns). Similar reductions were also seen with Mansonia spp., vectors of lymphatic filariasis and numerous arboviruses. Netting and insect-screen ceilings probably work as decoy traps attracting mosquitoes into the roof space, but not the room. Ceilings are likely to be well accepted and may be of greatest benefit in areas of low to moderate transmission and when used in combination with other malaria control strategies.

Comparison of deltamethrin tablet formulation with liquid deltamethrin and permethrin for bednet treatment in The Gambia.

The study aim was to compare three formulations, tablet deltamethrin, liquid deltamethrin and liquid permethrin, for their impact on vector behaviour and persistence. Product acceptance, perceived side-effects and user's perceptions of effectiveness were also investigated. At the beginning of the 1998 rainy season, 255 nets in a Gambian village were dipped in one of the three insecticides. Chemical residue analysis immediately after dipping showed that the target doses were reached for the liquid insecticides, but tablet deltamethrin deposited significantly less. Insecticide persistence at 5 months, however, was highest for the tablet formulation. Susceptibility tests established that Anophelines in this area were sensitive to both insecticides. All three formulations appeared effective as very few live Anophelines, or other mosquitoes, were caught under the treated nets. This conclusion was supported by the bioassay data with both deltamethrin formulations giving over 90% mortality soon after dipping and at 3 months, and at 5 months 70.8 and 79.6% were obtained for deltametrin liquid and tablet, respectively. Permethrin appeared less effective at all times (72.4, 86.8, 59.0%). There were no serious side-effects reported by the villagers following dipping. All three treatments were perceived as effective by the majority (92%) of users and most (93%) wanted to use the insecticide again. Deltamethrin tablets thus appear as good as permethrin for treating bednets in The Gambia. In addition, a tablet formulation is considerable easier to pack and distribute.

Parasitaemia and gametocytaemia after treatment with chloroquine, pyrimethamine/sulfadoxine, and pyrimethamine/sulfadoxine combined with artesunate in young Gambians with uncomplicated malaria.

As part of a study to assess the infectivity of gametocytes after treatment with four antimalarial regimens, the efficacy of each treatment was also determined. From September to December 1998, 598 children with uncomplicated malaria were treated; 135 received chloroquine (CQ) alone, 276 received pyrimethamine/sulfadoxine (Fansidar, PSD) alone, 113 received PSD with a single dose of artesunate (PSD + 1ART) and 74 received PSD combined with three doses of artesunate (PSD + 3ART). On day 28 19/63 (30.2%; 95% C.I. 19.2% to 43.1%) of children treated with CQ alone, 5/134 (3.7%; 95% C.I. 1.2% to 8.5%) treated with PSD alone, 1/71 (1.4%, 95% C.I. 0.0% to 7.9%) treated with PSD + 1ART and 0/45 (0.0%; 95% C.I. 0.0% to 7.9%) treated with PSD + 3ART were parasitaemic. The proportion of children with gametocytes on day 7 after treatment with CQ alone was 16/89 (18.0%; 95% C.I. 10.6% to 27.6%), 98/174 (56.3%; 95% C.I. 48.6% to 63.8%) after treatment with PSD alone, 8/70 (11.4%; 95% C.I. 5.1% to 21.3%) after treatment with PSD + 1ART and 4/46 (8.7%; 95% C.I., 2.4% to 20.8%) after treatment with PSD + 3ART. CQ thus has a lower efficacy than PSD or either of the PSD and artesunate combinations. Use of PSD alone as an alternative first line treatment results in a very high post-treatment gametocyte prevalence that is likely to enhance transmission. There would be greater and more sustainable benefits from using PSD and artesunate combinations.

Artesunate reduces but does not prevent posttreatment transmission of Plasmodium falciparum to Anopheles gambiae.

Combination therapy that includes artemisinin derivatives cures most falciparum malaria infections. Lowering transmission by reducing gametocyte infectivity would be an additional benefit. To examine the effect of such therapy on transmission, Gambian children with Plasmodium falciparum malaria were treated with standard regimens of chloroquine or pyrimethamine-sulfadoxine alone or in combination with 1 or 3 doses of artesunate. The infectivity to mosquitoes of gametocytes in peripheral blood was determined 4 or 7 days after treatment. Infection of mosquitoes was observed in all treatment groups and was positively associated with gametocyte density. The probability of transmission was lowest in those who received pyrimethamine-sulfadoxine and 3 doses of artesunate, and it was 8-fold higher in the group that received pyrimethamine-sulfadoxine alone. Artesunate reduced posttreatment infectivity dramatically but did not abolish it completely. The study raises questions about any policy to use pyrimethamine-sulfadoxine alone as the first-line treatment for malaria.

Permethrin-treated bed nets do not have a 'mass-killing effect' on village populations of Anopheles gambiae s.l. in The Gambia.

In The Gambia, the use of permethrin-treated bed nets has led to a reduction in morbidity and mortality from malaria in children. However, no clear evidence has been found for a 'mass-killing effect' on the mosquito vectors as a result of this intervention. Two further entomological studies to investigate this phenomenon have been carried out. In one study, 20 villages were paired so that bed nets in one member of each pair were treated with permethrin. In the other, a cross-over design was used in which treated and untreated bed nets were exchanged between 2 villages. Longevity, biting rate and resting density of the malaria vector population and sporozoite rates were assessed in both studies. Malaria vectors were equally abundant and long-lived, and as likely to be infective, in villages with treated bed nets as in those with untreated nets. However, a clear reduction in the density of the indoor-resting population of mosquitoes in rooms with treated bed nets was found, probably reflecting the excito-repellency of the insecticide. This study confirmed that, in The Gambia, the protection against death and morbidity from malaria seen in children using treated bed nets must be due primarily to personal protection rather than to a 'mass-killing effect' on the mosquito vector population at a village level.

Comparison of bednets treated with alphacypermethrin, permethrin or lambdacyhalothrin against Anopheles gambiae in the Gambia.

In the Gambian village of Saruja, where malaria is transmitted mainly by mosquitoes of the Anopheles gambiae complex, a trial was undertaken of the acceptability and efficacy of bednets treated with one of three pyrethroid insecticides--alphacypermethrin 40 mg/m2, permethrin 500 mg/m2 and lambdacyhalothrin 10 mg/ m2. Fewer mosquitoes were found alive under nets treated with insecticide than under control nets. Significantly more dead mosquitoes were found under nets treated with alphacypermethrin than under nets treated with permethrin or lambdacyhalothrin. Side-effects were reported by a proportion of the users of nets treated with each of the insecticides, but none were severe and their prevalence was similar between treatment groups. Unwashed nets treated with alphacypermethrin were more effective at killing anopheline mosquitoes in bioassays than nets treated with permethrin or lambdacyhalothrin. Killing activity was reduced when nets were washed, irrespective of which insecticide was used. Bednets treated with alphacypermethrin are well accepted, effectively killed anopheline mosquitoes and should therefore be evaluated for personal protection against malaria transmission.

Anopheles gambiae gonotrophic cycle duration, biting and exiting behaviour unaffected by permethrin-impregnated bednets in The Gambia.

Permethrin-impregnated bednets protect children against malaria in The Gambia, where Anopheles gambiae complex mosquitoes are the main vectors of malaria. However, no effect has been found on mosquito density, parous rates or sporozoite rates in An.gambiae sensu lato populations; only a reduction in the numbers of mosquitoes resting indoors in rooms with treated bednets. A possible explanation for this paradox is that exposure to treated bednets leads to changed vector behaviour such as a shift in biting time, a diversion to biting outdoors instead of indoors, to biting animals instead of humans, or to increased duration of the gonotrophic cycle. To investigate these possibilities, we observed the biting and existing behaviour of An.gambiae in ten pairs of villages, in half of which the residents used permethrin-treated bednets. The possible influence of treated bednets on the gonotrophic cycle length was evaluated by mark-release-recapture experiment. No significant difference was found between villages with treated and untreated bednets in the indoor/outdoor ratio of human biting, in mean biting times or in human blood indices of An.gambiae females found resting indoors in the mornings. The proportions of unfed, fed or gravid An.gambiae females collected in exit traps, and the number of females exiting showed no significant differences between rooms with treated and untreated bednets. Indications for a gonotrophic cycle length of 2 days were found. No evidence for any change in duration of the gonotrophic cycle in relation to exposure to treated bednets was found, although the number of recaptures was low in the villages with treated bednets. Since equal numbers of infective An.gambiae were found in villages with treated or untreated bednets, and no changes in mosquito behaviour were detected, we cannot account for how children are protected against malaria by treated bednets. One possibility is that mosquitoes divert to bite other hosts, including adults.

Geographical perspectives on bednet use and malaria transmission in The Gambia, West Africa.

Insecticide-impregnated bednets are now widely accepted as an important tool in reducing malaria-related deaths in children in Africa. Defining the circumstances in which net treatment programmes are likely to be effective is essential to a rational development of this control strategy. In The Gambia a National Impregnated Bednet Programme was introduced into the primary health care system in 1992. Prior to its introduction baseline epidemiological and entomological studies were conducted throughout the country. These studies showed that in areas where mosquito biting nuisance was high, people protected themselves with bednets and that where mosquito densities (and therefore bednet usage) was low malaria prevalence rates were relatively high. Since the national programme is designed to assist only those people who already own a bednet (by providing the insecticide) an understanding of the factors which determine bednet ownership is needed to help evaluate the programme's effectiveness and provide guidelines for increasing bednet usage. Village scale bednet usage rates and malaria prevalence rates obtained from the baseline survey were correlated with certain geographical variables: dominant ethnic group, area, habitat, distance from the River Gambia and distance from the 'bluffline' (the interface between the sandstone soils and alluvial soils which border the river system). In a multiple regression analysis, bednet usage was independently associated with area (P < 0.001), ethnic group (P = 0.010), habitat (P = 0.006) and distance from the river (P = 0.013). A negative association of bednet usage with malaria prevalence persisted after allowing for the other variables. Malaria prevalence was not independently associated with area, ethnic group, habitat or distance from the river. Our analysis showed that the impregnated bednet programme is likely to be most effective in villages which are sited near to or on the alluvial soils in the middle and lower river zones. These villages, which were originally settled for easy access to the river (for transport) and its swampy margins (for rice production) are within the flight distance of mosquitoes that have their breeding sites on the poorly drained alluvial soils. Variation in malaria prevalence rates (after bednet usage has been taken into account) may be related to factors such as poverty and access to health care, and/or to localized differences in the ecology of The Gambia, which determine the duration and intensity of transmission. If the National Bednet Programme is to be effective throughout The Gambia it is vital to develop promotional activities which will encourage bednet usage in areas where nuisance biting by mosquitoes is low.

Movement of Anopheles gambiae s.l. malaria vectors between villages in The Gambia.

Movement of mosquitoes belonging to the Anopheles gambiae complex (mixed wild populations of An.arabiensis, An.gambiae and An.melas) between three neighbouring rural villages in The Gambia was investigated by mark-release-recapture. A total of 12,872 mosquitoes were collected in bednets, marked with a magenta fluorescent powder and released over a 15-day period in one of the villages. A further 15,507 mosquitoes were collected in exit traps, marked with a yellow powder and released over the same period. Mosquitoes were captured daily in all three villages using pyrethrum spray catches, as well as bednet and exit trap catches. The catching period extended for 6 days after the last day of release. Of the mosquitoes released, 372 (1.3%) were recaptured 2-21 days later. Of these recaptures, 272 were caught in the release village, and 98 were caught in other villages situated 1-1.4 km away. The 'movement index' between villages was calculated as 17.2% (12.2-22.4% confidence limits) for mosquitoes released after feeding and 20.1% (14.7-25.3%) for those released unfed. These results suggest that movement of mosquitoes between neighbouring villages in The Gambia seriously affects the entomological evaluation of pyrethroid-impregnated bednet programmes in areas where treated and untreated villages are interspersed.

Efficacy trial of malaria vaccine SPf66 in Gambian infants.

SPf66 malaria vaccine is a synthetic protein with aminoacid sequences derived from pre-erythrocytic and asexual blood-stage proteins of Plasmodium falciparum. SPf66 was found to have a 31% protective efficacy in an area of intensive malaria transmission in Tanzanian children, 1-5 years old. We report a randomised, double-blind, placebo-controlled trial of SPf66 against clinical P falciparum malaria in Gambian infants. 630 children, aged 6-11 months at time of the first dose, received three doses of SPf66 or injected polio vaccine (IPV). Morbidity was monitored during the following rainy season by means of active and passive case detection. Cross-sectional surveys were carried out at the beginning and at the end of the rainy season. An episode of clinical malaria was defined as fever (> or = 37.5 degrees C) and a parasite density of 6000/microL or more. Analysis of efficacy was done on 547 children (316 SPf66/231 IPV). No differences in mortality or in health centre admissions were found between the two groups of children. 347 clinical episodes of malaria were detected during the three and a half months of surveillance. SPf66 vaccine was associated with a protective efficacy against the first or only clinical episode of 8% (95% CI -18 to 29, p = 0.50) and against the overall incidence of clinical episodes of malaria of 3% (95% CI -24 to 24, p = 0.81). No significant differences in parasite rates or in any other index of malaria were found between the two groups of children. The findings of this study differ from previous reports on SPf66 efficacy from South America and from Tanzania. In The Gambia, protection against clinical attacks of malaria during the rainy season after immunisation in children 6-11 months old at time of the first dose was not achieved.

PIP: During December 1993 to November 1994, in the Upper River Division of The Gambia, 630 infants aged 6-11 months at time of first dose received three doses of malaria vaccine SPf66 or injected polio vaccine (IPV) and were followed up at home during the rainy season using active and passive case detection methods to determine the protective efficacy of SPf66 against clinical episodes of malaria due to Plasmodium falciparum. The researchers were able to use data on only 547 children (316 SPf66/231 IPV) to determine efficacy. The definition of clinical malaria was fever (37.5 degrees Celsius or higher) and a parasite density of at least 6000/mcl. The two groups were essentially the same in terms of mortality, health center admissions, parasite rates, or any other index of malaria. Health workers identified 347 clinical episodes of malaria during the three months of surveillance. SPf66 vaccine had a protective efficacy against first or only clinical episode of malaria of 8% (p = 0.5). Its protective efficacy was 3% against all clinical episodes of malaria. The results of this trial were different than those from earlier reports on SPf66 efficacy from South America and Tanzania. Possible reasons accounting for the different findings were a mistake in coding syringes for the third dose, substantially less intensity of malaria infection in The Gambia than South America, younger children in The Gambia than in Tanzania, genetic differences in the populations, and difference in length of follow-up. In conclusion, protection against malaria during the rainy season after immunization with SPf66 vaccine in infants aged 6-11 months did not occur.

Entomological evaluation of the Gambia's National Impregnated Bednet Programme.

Entomological studies were conducted in paired study villages in three of the five study areas used for the epidemiological assessment of the Gambia's National Impregnated Bednet Programme. Baseline data collected in 1991 were compared with post-intervention data from 1992, when one of each village pair (from areas II, III and V) was included in the treatment programme in which villagers' nets were dipped in permethrin. In a longitudinal study, indoor-resting densities of Anopheles gambiae s.l. in the treated villages were significantly reduced, when compared with their paired untreated village, in areas II (t = 3.32, 13 degrees of freedom, P = 0.006) and III (t = 3.71, 13 degrees of freedom, P = 0.003). However, this was not associated with higher outdoor-biting rates in the evenings in the treated villages than in the controls. The reduction in vector population was most evident in area II, where the treated village was relatively isolated and 74% of the population slept under a treated net. Anopheles gambiae s.l. were also collected during two fortnightly periods in area V, where the sporozoite rates in 1991 had been relatively high (7.65% and 6.07%). There was no significant decrease in the sporozoite rate in the treated village in area V in 1992, despite the fact that this village was over 5 km from its nearest neighbouring village. However, the proportion of villagers sleeping under a treated net in this village was < 50%. No clear evidence for an epidemiologically significant 'mass killing effect', resulting from the village-wide use of treated nets, emerges in this or any other study undertaken in The Gambia. However, such an effect is indicated by results from Salikene, where the village mosquito population was relatively isolated and where the majority of the population slept under a treated net. This situation is unlikely to be repeated in most Gambian villages in the primary-health-care scheme, since treated and untreated villages are generally interspersed and the mosquito populations move freely between them. We must therefore conclude that, on a national scale, the nets provide an exceptionally high level of personal rather than community protection. The mechanisms underlying the personal protection are, however, still unclear.

Village trial of bednets impregnated with wash-resistant permethrin compared with other pyrethroid formulations.

A village-scale field trial of pyrethroid-impregnated mosquito nets was undertaken in The Gambia, West Africa, in the Mandinka village of Saruja (13 degrees 13'N, 14 degrees 55'W) during July-November 1989. Nearly all the villagers possessed and used their own bednets. Anopheles gambiae is the main vector of human malaria in the area. An experimental wash-resistant formulation of permethrin was compared with standard emulsifiable concentrate (EC) formulations of permethrin and lambda-cyhalothrin, versus placebo-treated bednets. Target concentrations of pyrethroids on bednets were permethrin 500 mg/m2 and lambda-cyhalothrin 25 mg/m2. The experimental design involved random allocation of a treatment to one net per family. Whereas 68% of people questioned said they washed their nets fortnightly, observations during the 16-week trial period showed that only 4/130 (3%) of nets involved in the trial had been washed as frequently as once per month. Early morning searches for mosquitoes under bednets (1 day/week for 16 weeks) found significantly more mosquitoes (60% An.gambiae) in placebo-treated nets than in pyrethroid-treated nets. The numbers found with each of the three pyrethroid treatments did not differ significantly from each other. Insecticidal efficacy of the treatments was tested by bioassays using wild-caught unfed mosquitoes exposed to netting for 3 min. Linear regression analysis of bioassay mortality against number of times that a net had been washed by villagers showed that nets impregnated with the wash-resistant permethrin retained their insecticidal properties better than nets impregnated with lambda-cyhalothrin or with the standard permethrin formulation.

Malaria prevalence is inversely related to vector density in The Gambia, West Africa.

Baseline epidemiological and entomological studies were conducted in 5 different areas of The Gambia before the introduction of a national malaria control programme, the objective of which was to treat all the bed nets belonging to people living in primary health care villages with insecticide. All malariometric indices used (parasite density, parasite rates, splenomegaly, and packed cell volume) indicated that malaria transmission was more intense in the east of the country than elsewhere. High transmission in the east was associated with a high sporozoite rate but not with the greatest vector abundance; the lowest malaria prevalence rates were found in villages which were close to very productive breeding sites of Anopheles gambiae s.l. Bed net usage was strongly correlated with vector density and the highest malaria rates were found in villages where bed net usage was relatively low. These results suggest that in The Gambia malaria prevalence rates are reduced where nuisance biting by mosquitoes is sufficient to encourage the population to protect themselves with bed nets.

A malaria control trial using insecticide-treated bed nets and targeted chemoprophylaxis in a rural area of The Gambia, west Africa. 7. Impact of permethrin-impregnated bed nets on malaria vectors.

The impact of permethrin-impregnated bed nets on malaria vectors was studied in 6 pairs of villages during the rainy season in 1989. In each pair, the residents of one village had their nets treated whilst those of the other remained untreated. Routine collections of mosquitoes were made outdoors in the early evening using human-biting collections, and indoors with insecticide sprays, light traps and by searches under bed nets. Mosquitoes of the Anopheles gambiae complex, An. gambiae sensu stricto, An. arabiensis and An. melas, were present in large numbers for 5 months of the study period. These mosquitoes were susceptible to permethrin as judged by bioassay results. Outdoor human-biting rates in the early evening in communities with treated bed nets were similar to those in communities with untreated nets. In villages with treated bed nets most biting occurred outdoors in the early evening with little taking place under impregnated nets. The insecticidal activity of permethrin-impregnated bed nets, dipped by the local population, provided good individual protection against mosquitoes throughout the rainy season and bed nets remained effective even when washed up to 3 times. There was little to suggest that the use of insecticide-treated nets reduced the survival of mosquito populations in villages with impregnated nets. The absence of the expected village-wide effects of net impregnation may have resulted from the circulation of mosquitoes between villages with treated and untreated nets. The proportion of mosquitoes which fed on humans did not differ significantly between villages with treated and untreated nets.(ABSTRACT TRUNCATED AT 250 WORDS)