The results presented here show that in areas with low leishmaniasis incidence (<15 per 1000), indoor residual spraying with alphacypermethrin can lead to a significant reduction in leishmaniasis transmission. In both localities under intervention, leishmaniasis incidence and gravidity rate have been significantly reduced. However, sand fly abundance was not significantly affected. In fact, sand flies were collected from animal shelters, which are also reproduction, resting and blood feeding sites. As IRS is directed exclusively against adult stages, larvae will not be affected until the emergence of adults and their contact with treated surfaces. Once on the walls, young blood fed adults are killed and do not have time to rest inside to digest their blood meal. It is known that sand fly larval development is slow and uneven mainly under temperate conditions [10, 11]. The emergence of young flies that belonged to the same batch of eggs may extend over a long period (up to two months). The continuous emergence of adults continually replenished the population.
Similar results were observed in Brazil by Passerat de Silans et al.  who showed that residual insecticide applications of cypermethrin had no effect on the overall density of peridomiciliar Lutzomyia longipalpis populations. Alexander et al.  also found that residual spraying of walls with delthamethrin in a Colombian village surrounded by forest had no perceptible effect on the number of sand flies entering houses; although the insecticidal activity of the treated surfaces was undiminished during the study period. While other authors reported that residual pyrethroid spraying reduces sand fly indoor density in various localities in South America [14–16] and China  probably, because of their repulsive effect.
The residual effects of pyrethroid insecticides reported in the literature are variable according to formulations, doses used, types of surfaces tested and species of sand flies exposed. In this program, the residual effect of alphacypermethrin, at a dose of 30 mg/m2, remained for less than three months (about 10 weeks). A shorter duration was reported by Morsy et al.  in Egypt where the exposition of Ph. papatasi, during 30 min, on cement walls treated with permethrin, 75 days after treatment involved only 51.9% of mortality. A similar duration was reported by Passerat De Silans et al.  who indicated that residual activity of cypermethrin against Lu. longipalpis in Brazil was limited to two months.
Longer periods were also reported in rural areas of the new World tropics. Le Pont et al.  reported that one deltamethrin spray application in Bolivia was enough to eliminate Lu. longipalpis from the dwelling during 9 months. Falcao et al.  demonstrated that the residual action of deltamethrin remained for more than one year on walls of concrete and lime painted in a Brazilian focus of American cutaneous leishmaniasis. Davies et al.  confirmed that, with a concentration of 25 mg/m2 of lambdacyhalothrin, the insecticide remained 100% lethal for the sand fly Lu verrucarum for up to 6 months in endemic villages of Peru.
Hence, because of the low residual activity of alphacypermethrin (less than three months) and the long sand flies development period in Morocco (more than 6 months), one spray round per year is not sufficiently powerful to control sand fly populations in many regions of Morocco. Indeed, sand fly populations peaked twice during the season, at the end of spring: May-June and again at the end of summer: August-September . This bi-modal population density pattern is characteristic of Ph. sergenti in semi arid habitats and has to be taken into account when determining the best application schedule of insecticide. Ambient conditions (temperature and humidity) were more favourable to transmission during periods corresponding to the second peak. Insecticide persistence should largely cover this period.
Application of insecticide during the two years of the study in the two localities has taken place after the first peak (end of June). Insecticide remained effective during the following months (July-August). The insecticide has no more efficacy at the beginning of September, when conditions become favourable to transmission. This suggests that two spray rounds are needed to maintain good effectiveness. The first should begin early in the season before the first peak of sand fly populations early in May (in order to kill the maximum number of sand flies throughout the whole season) and the second round early in August. However, such a strategy implicates an increase in the cost of leishmaniasis prevention and cannot be implemented everywhere. It must be reserved only for areas with the highest leishmaniasis transmission.