[Anopheles and malaria transmission in Ambohimena, a village in the Occidental fringe of Madagascar Highlands].

The anopheline vectors and malaria transmission were investigated in the Middle West of Madagascar, in the village of Ambohimena (at the altitude of 940 meters) during two years (August 1996 to July 1998). This village is located outside the vector control area, where yearly DDT house spraying campaigns have been conducted between 1993 and 1998. Collection of mosquitoes was mainly based on all night man billing collections (650 man-nights), pyrethrum spray catches (224 bedrooms) and direct collections in outdoor resting places (140 toilets, 61 pigsties, 33 holes, 19 sheds, 79 sisal hedges, 70 cart shelters). Blood fed anophelines allowed analysis of the origin of blood with an ELISA method. Presence of circum-sporozoite protein was assessed with another ELISA method. The total number of collected anophelines was 14,280. Two malaria vectors were identified: Anopheles funestus Giles, 1900 and An. arabiensis Patton, 1902. An. funestus was the most abundant mosquito, especially during the hot rainy season. Two peaks of abundance were observed (in December and April). Endophagic rate (for mosquitoes aggressive for man) of 35.3%, an endophilic rate (for resting mosquitoes) of 78.0% and an anthropophilic rate (for indoor resting mosquitoes) of 64.0% were calculated. The average parity rate was relatively low (61.2%). The Plasmodium falciparum immunological sporozoite rate was 0.20%. An. funestus presented a higher vectorial capacity during the first round of rice cultivation (January) than during the second round (April-May). An. arabiensis was mostly abundant in December and January at the beginning of the rainy season. This species was exophagic (endophagic rate = 27.5%) and zoophilic (anthropophilic rate = 7.8%). The sporozoitic index was determined as zero (number of examined mosquitoes = 871). In this village, An. arabiensis presented only marginal importance for malaria transmission. Malaria transmission occurred from December to April. Annual entomological inoculation rate, only due to An. funestus, was 8.96 during the first year, and 3.17 during the second year. In this area where transmission is moderately stable, we suggest an extension of vector imagocidal control activities up to the western fringes of the Highlands.