Prior to assessing the larvicidal activities of PP, focus group discussions (FGD), in-depth interviews and questionnaires were conducted in some traditional communities were PP are still used for larval control. The rationale of FGD and interviews was to generate enough information on: i) the empirical use of PP, ii) the mode of application, iii) the periods and eventually iv) the frequencies of utilization of PP.

Larvae of Anopheles Ladji were exposed to four petroleum products, petrol, kerosene, engine oil and waste oil from mechanics, and their larvicidal activities were recorded and quantified. Lethal concentrations of the various petroleum products were determined after treatment of Anopheles breeding sites with increasing doses of PP in the laboratory. For each petroleum product, the lowest concentration (LoC₁₀₀) capable of inhibiting the development of larvae to the adult stage (100% mortality of exposed larvae) was determined as well as the highest concentration (HiC) not having any observable impact on the growth of larvae (0% mortality of exposed larvae). In between both concentrations, the LC₅₀ (concentration leading to the mortality of 50 % of exposed larval population) was also determined for each PP.

Operationally, the importance of determining the LoC₁₀₀ resides in the fact that this is the concentration which produces the best lethal larvicidal activity with fewer products. At this concentration, the treatment process is financially economical. As for the HiC, also referred to as the NOEL (no effect level) this corresponds to the quantity of petroleum products "wasted" in the environment during treatment of breeding sites. The HiC is the concentration that has no observable toxicity on Anopheles larvae and is, therefore, the concentration at which mutant or resistant populations of Anopheles could be gradually selected. Both operational concentrations were the main focuses of this field and laboratory investigations on malaria control tools and the selection of resistance in malaria vectors.

Increasing quantities of each PP were used to treat surfaces of breeding sites constituted in the laboratory. Known volumes of PP were introduced in 255 cm² bowls each containing well water and 25 larvae of Anopheles Ladji (second to third stage larvae). Larvae were fed throughout the experiment with biscuit and yeast. Each dilution was replicated four times, making a total of 100 larvae tested at each concentration. Similar breeding sites with no traces of petroleum products (controls) were constituted and monitored in tandem.

Mortality rates as well as the number of adults emerging from each breeding site were determined. Data generated were pooled and used for plotting the curve of activity of each PP on the resistant Anopheles Ladji strain. Thus, the HiC and the LoC₁₀₀ corresponding to petrol, kerosene, engine oil and waste oil was determined by doses response analysis during this laboratory assessment. In between the HiC and the LoC₁₀₀, the LC₅₀ was determined for each PP.

Two potential modes of action of petroleum products were screened during this study: i) the killing of larvae by "suffocation" through the oil film/layer produced by PP at the surface of breeding sites, and ii) the direct lethal activity through dissolved particles of PP in breeding sites, referred here as "contact toxicity". Water samples with visible residues of petroleum products were collected in the field and more precisely in areas treated with petroleum products. These samples (crude samples from the field) were used in the laboratory for the simulation of two types of breeding sites:

i) The first set of breeding sites known as "unsieved or crude" was directly reconstituted by putting two litres of the water from the field into laboratory bowls.

ii) The second set of breeding sites known as "sieved or clean" was reconstituted by sieving to clean two litres of the crude sample from the field.

One hundred larvae of Anopheles Ladji were introduced in each bowl and reared to adults. Four replicates were made for each simulation, making a total of 400 larvae monitored in "un-sieved/crude" and "sieved/clean" breeding sites. Control bowls were constituted alongside using well water with no traces of petroleum products and containing the same number of Anopheles larvae. For each type of breeding site (sieved and un-sieved), mortality rates were determined and statistical tests of comparison used to associate larval lethality with either sieved or un-sieved simulations.

Sociological information from focus group discussions, in-depth interviews and questionnaires conducted in studied communities were compiled and analysed using Excel and Text base-beta computer software.

Larvicidal activities of each PP on the resistant strains Ladji were analysed and plotted with Excel and, key values tested with Stat-calc-Epi-info software. The potential mode of action of PP on Anopheles Ladji was determined through statistical comparison using Stat-calc-Epi-info software.

Results from qualitative and quantitative surveys conducted in some traditional communities (Gbodjo, Ladji, and Ketonou) in the Republic of Benin and more specifically in localities where Anopheles populations are resistant to standard insecticides (mortality rates with permethrin < 60%,) confirmed the empirical use of PP against several insects of great nuisance like: flies, mosquito larvae and cockroaches. The PP mainly used in these communities are kerosene and waste engine oil from mechanics. Others, such as petrol and engine oil, are said to be used occasionally. These products are either sprayed on the ground, on table surfaces, on standing water points and in latrines. Out of a total number of 65 key respondents interviewed in the three communities, 73% spray kerosene in standing water points and latrines to control mosquito growth and nuisance, whereas 9% use waste engine oil, 5% engine oil and 2% petrol. Information extracted from the interviewed populations also showed that 11% do not know much about PP. This sociological investigation revealed that this technique of controlling vectors using PP is transferred in the community from parents to offspring without any scientific explanation to support the efficacy of this strategy (Figure 1).

Although abandoned long ago by National Programmes of Malaria Control in most African countries, PP are still used in several traditional communities in the Republic of Benin. This is probably due to the availability of these products, which are sold on the streets of many cities in this country and, are often freely available from mechanical workshops and industries (case of waste engine oil). The low income of communities in traditional settings also accounts immensely in the slow adoption of synthetic insecticides and, therefore, solidifies their attachment to PP as the main malaria control tool. The use of PP for vector control in the studied communities is transferred from generation to generation without any scientific explanation. The availability of PP in several localities in the Republic of Benin and their relative cost-effectiveness suggests this method of mosquito control may have some benefit to the communities.

This study has clearly established the larvicidal properties of petrol, kerosene, engine oil and waste oil against Anopheles Ladji larvae, the most resistant strain currently identified in the Republic of Benin 13. The relatively low efficacy of petrol as recorded during this experiment could be explained by its high volatility compared to kerosene, engine oil and waste oil. The high volatility of petrol does not allow its persistency in breeding sites. This low persistency results in a low residual effect of this product in treated breeding sites. The relatively high efficacy of kerosene, engine oil and waste oil is likely to be due to their elevated persistency in breeding sites after treatment.

The HiC and the LoC₁₀₀ defined and determined in this research are key operational values for this assessment which focused on the analysis of malaria control tools and the selection of resistance in A. gambiae. The importance of the LoC₁₀₀ resides in the fact that this is the concentration which produces the best lethal activity with fewer products. It, therefore, corresponds to the cost of product during the treatment of breeding sites. On the other hand, the HiC corresponds to the quantity of petroleum products "wasted" in the environment during treatments of breeding sites. This value defines the threshold at which resistant populations of Anopheles could be gradually selected. The HiC values recorded in this study reflects the high probability of theses products to contribute in the emergence of resistant populations of Anopheles in studied localities. It is possible that these products might have contributed through a form of cross-resistance to the elevated cases of pyrethroid resistance currently recorded in various localities in the Republic of Benin. The empirical use of PP and the absence of knowledge on ideal concentrations to use during the treatment of breeding sites have over the years contributed to the spread in the environment of none active quantities of PP which in return would have contributed to the selection of resistant populations of mosquitoes. These 'mosquito-friendly' concentrations of PP found in standing water bodies (HiC) have probably induced over the years a cross resistance to pyrethroids in Anopheles populations. However this hypothesis needs to be investigated thoroughly using advanced molecular techniques.

The basic information revealed by this study on the mode of action of PP on mosquito larvae highlights an active lethality of PP by "contact-toxicity" rather than suffocation. It is possible that some heavy metals from petroleum products dissolve, diffuse into the water and are ingested by Anopheles larvae after treatment of the breeding sites. The elevated mortality in sieved breeding sites probably occurs following the ingestion of these metals by larvae. In-depth analyses have to be conducted to identify and quantify these metals and to monitor their progression in the digestive system of Anopheles larvae after ingestion.