This study was conducted in the south-western region of Nigeria. Samples were collected from five states in the region namely Ogun, Oyo, Osun, Ondo and Ekiti states (Figure 1). Standing water points found in each locality were systematically scrutinized for mosquito larvae. Anopheles larvae were recognized by their morphology and their horizontal position on water surface 13. Larvae were sieved, concentrated in small buckets, washed with clean water and taken to the insectary of the Department of Zoology, University of Ibadan, Nigeria, where they were reared to adult stage for WHO bio-assays 14. A Global Positioning System (GPS) device was used to record the latitude and longitude of each identified breeding site. These parameters were used for accurate positioning of surveyed sites on a map (Figure 2).

Water samples collected from each identified breeding site were macroscopically described, and their pH and dissolved oxygen analyzed in situ using the multi-meter (Consort C535, Multi-variates). Macroscopical parameters considered in this study include the presence of petroleum particles in the water, the thickness of the water and the presence of oil layer at the surface of the breeding sites.

Two to five-day old Anopheles emerging from the identified breeding sites, were bio-assayed using discriminating doses of insecticides. Bio-assays were carried out following WHO protocols 14. Females of An. gambiae were exposed for one hour to papers impregnated with diagnostic dosages of permethrin (0.75%). For each test, five test tubes were used: one for the control and four for exposed mosquitoes. Control tubes contained filter papers impregnated with silicon oil (insecticide carrier), whereas treated papers were impregnated with diagnostic doses of insecticide plus carrier. An average of twenty mosquitoes was introduced in each tube. After one hour, mosquitoes were transferred to holding containers and provided with cotton pads with 10% sucrose solution. Mortalities were recorded at 24 hours and the susceptibility status of the population was graded according to the WHO protocol 14.

In each locality, 20–30 females Anopheles from WHO Bio-assays were analysed at molecular level. PCR-species 15 was conducted in all Anopheles to identify the various members of An. gambiae complex encountered in each site. The next set of PCR focused on molecular forms 16 and involved only An. gambiae sensu stricto (s.s.) from the previous screening. The PCR forms sub-grouped the An. gambiae s.s. into two molecular forms: An. gambiae s.s. (M) and An. gambiae s.s. (S). The last series of PCRs was to assess the presence of the kdr mutations in An. gambiae ss. populations 17

Following the exposure of females An. gambiae to permethrin impregnated papers, a good spread of resistance was recorded in all the five states surveyed. Out of the 15 tested populations of An. gambiae s.l., 12 populations were clearly resistant to permethrin, whereas three were within the range of susceptibility as defined by WHO standards 14.

In Oyo state, out of the five localities studied (Ojoo, Basorun, UI, Challenge and Oja-tuntun), resistance was clearly recorded at Basorun (70% mortality), Ojoo (80% mortality), Challenge (81% mortality) and Oja-tuntun (81% mortality). No case of resistance was recorded among Anopheles spp. from UI (100% mortality). In Osun state, susceptibility was recorded at Modakeke (97% mortality) and resistance at Lagere (94% mortality) localities.

In Ondo state, the Anopheles populations from the two localities surveyed: Ilesha garage (Akure) and Owena (Owena) were all resistant to permethrin (89% and 75%, respectively).

In Ekiti state, resistance was recorded with Anopheles from the two localities sampled; 85% mortality rate was recorded at Ati-kankan1 (Ado-ekiti) and 88% at Ati-kankan2 (Ado-ekiti).

In Ogun state, resistance was recorded at Ogere1 (75% mortality) and Ogere2 (88% mortality), situated on the Lagos-Ibadan express way and at Mobalufon (80% mortality) in Ijebu-ode. A susceptibility pattern was noticed at Imowo (96% mortality), locality of Ijebu-ode.

In most screened breeding sites, cadavers of larvae were recorded: probably due to lethal activities of petroleum products. The summary of the results of bio-assays is shown in (Table 1).

Using geographical information recorded with the GPS, localities screened were projected on a map of Nigeria and areas of permethrin resistance demarcated. The map generated from this study showed a spread of resistance in most states of South Western Nigeria. The levels of resistance registered in the localities of Basorun, Owena, and Ogere were relatively consistent compare to other sites. On the other hand, no resistance was found in UI. A reduced susceptibility to permethrin was recorded in Modakeke and Imowo (Figure 2).

Cross-analyses were made with data on the quality of the breeding sites (physico-chemical parameters, presence of petroleum particles) and the pyrethroid susceptibility status of emerging adults of An. gambiae. Very slight differences were obtained between the pH of breeding sites producing resistant populations of An. gambiae and those producing susceptible populations. Contrary to data on the pH, breeding sites having petroleum particles and therefore low levels of oxygen seemed to produce resistant strains of mosquito. It is possible that the survival rate of eggs laid by susceptible mosquitoes in oily breeding sites is low. This inhibition of eggs or larvae development (P = 0.0015) could explain the low presence of susceptible populations of Anopheles in oily breeding sites. In the three breeding sites with little or no oil particles (breeding sites No.13, 5, 15), a consistent number of pyrethroid susceptible populations of An. gambiae was recorded (Table 2 and 3).

In the laboratory, both pyrethroid susceptible and resistant strains of Anopheles exhibited a high reluctance for laying eggs in breeding sites containing petroleum products. Out of a mean number of 1,900 eggs laid in each cage by the susceptible strain Anopheles UI, 4.5% (86 eggs) of these eggs were deposited in oily breeding sites and 95.5% (1814 eggs) in the clean breeding site (control site). With the resistant strain Anopheles Ojoo, Out of a mean number of 2,425 eggs laid in cages, 77% (1875 eggs) of these eggs were deposited in breeding sites with clean water (control site) whereas 23% (550 eggs) of these eggs were laid in breeding site with petroleum debris (Figure 3). This partially reveals an acceptable level of tolerance of oily breeding sites by resistant females of An. gambiae compare to susceptible ones (UI/Ojoo, P = 0.00000). The very low tolerance to oily breeding sites by pyrethroid susceptible populations of Anopheles during oviposition translates the high selection pressure exercised on this mosquito species in localities where most breeding sites are covered with petroleum products.

Of a total of 400 eggs of Anopheles UI monitored during this experiment, 200 eggs were introduced in breeding sites containing oil particles (test) and the remaining 200 in the breeding site with no oil (control). Results generated showed that 61% (122 eggs) of the 200 eggs hatched in oily breeding sites and 87% (174 eggs) of 200 eggs hatched in breeding site with no petroleum products (control). With the Ojoo strain, 84% (168 out of 200 eggs) hatching rate was recorded in the control breeding site whereas 63% (126 out of 200 eggs) hatching rate was found in the breeding site with oil products. These results point to a significant impact of oil products on the hatching of eggs of both tested strains but with a more pronounced hatching inhibition on the susceptible strain Anopheles UI (UI, P = 0.00000 and Ojoo, P = 0.0002) (Figure 4).

A catastrophic effect of oil was recorded during the development of the susceptible larvae of Anopheles UI. Very few larvae of this strain (1%) could reach the pupae stage in breeding sites polluted with petroleum products (only 1 pupa from the 122 L1-larvae emerging from hatched eggs). With Anopheles Ojoo 27% (34 out of 126 hatched individuals) of larvae were able to reach the pupae stage when reared in water with oily debris. When the rearing was conducted in clean breeding sites (control breeding sites), we recorded a relatively high number of pupae from both strains (64% and 66% of L1-larvae emerging from hatched eggs got to the pupae stage in non oily breeding sites with the susceptible and resistant strain respectively) (Figure 5). Although a lethal effect of petroleum products was globally observed with both strains, the destructive impact of these products was more prominent on larvae of UI compare to those of Ojoo (UI/Ojoo, P = 0.00000). These results suggest a high selection pressure at larval stage of resistant populations of Anopheles by spilled petroleum products (Figure 5).