Utilization of ‘swamp’ rice fields by members of the anopheles Gambiae complex in Gambia

Abstract

Background Whilst it is well known that rice production in Africa increases the production of Anopheles gambiae mosquitoes, most studies have investigated this in irrigation schemes. Here I examine the colonisation of 'swamp' rice fields by An. gambiae mosquitoes in The Gambia and examine some of the factors responsible for the presence and absence of these vectors in field and semi-field conditions. This work is of relevance to large-scale larval control programmes that have identified rice fields as a major source of malaria vectors in The Gambia.Methods Larval and adult mosquito surveys were carried out in rice fields near Tamba Koto village in rural Gambia from June to January, 2006, a period that included the wet season, the period of most malaria transmission. Three transects each 500m long and 200m apart situated on the edge of the River Gambia floodplains were routinely surveyed. Larval sampling using area sampler and dippers was done at regular intervals along each transect. Adult sampling with emergence traps placed over water. I compared three different water treatments that were commonly found under field conditions in Gambian rice fields: the presence of algae, cow dung and urea. The number of larvae and pupae were counted daily for 14 days in 16 artificial breeding sites of plastic bowls filled with tap water, each with a different treatment and an untreated control bowl. This trial was repeated four times over 12 weeks period.Results Three hundred and seventy-five (375) anopheline larvae were caught during the larval survey in the field, with 349 larvae (93%) collected within the first 350m from the landward edge of the paddy fields. Out of the An. gambiae complex collected, 36 (45%) were An. arabiensis, 23 (29%) were An. gambiae sensu stricto and 21 (26%) were An. melas. A total of 263 adult mosquitoes were collected during the adult emergence study. Sixty - eight belonged to the An. gambiae complex, 139 were culicines, 30 were Aedes, and 26 were anophelines which could not be identified by Polymerase Chain Reaction. Out of the 68 An. gambiae sensu lato, 61 were An. arabiensis, 4 were An. gambiae s.s. and 1 was An. melas. Of the 68 An. gambiae s.l. caught, 62 (91.1%) were caught along the edges of rice fields. Although fish are common and evident animals in rice fields, there have been relatively few studies on the ecology of indigenous species. We did not find fish in any of the sampling points within 350m of the village, but did find them far away close to the river.A total of 17,467 mosquito larvae were collected from 16 bowls during four months of the semi field trial. Of these 75% were early instars and 25% late instars. Of the total number of mosquito larvae sampled, 6,233 (36.7%) were identified as anopheline and 11,234 (63.3%) culicine. Of the anophehnes sampled, 5164 (83%) were early instars whilst 1069 (17%) were late instars. Field surveys showed that paddies closest to the land had no fish and were rich in cow dung.More anopheline larvae were produced from cow dung treatments (n=l,718; 33.3%) and, bowls treated with algae (n=l,453; 28.1%) compared to water alone (n=l,064; 21%).Similar numbers were found in bowls with urea (n=929; 18%) compared with water alone. Similar trends were observed for culicine larvae. Out of 8,021 early stage larvae,2,675 (33%) were found in bowls with cow dung, 2,554 (32%) from algae, 1,532 (19%)) from water alone and 1,260 from urea (16%).One hundred and thirty-eight adult mosquitoes of An. gambiae complex were identified. Fifty-nine were An. gambiae s.s., 74 were An. arabiensis and 2 An. melas. Out of 59 An. gambiae s.s., 24 (41%) were recorded from cow dung treatments and 21 (36%) from algae treatments. Of the 74 An. arabiensis, 35 (47%)) were recorded form cow dung whilst 28 (38%) were from algae. This result indicated that both cow dung and filamentous algae supported production and development of An. gambiae complex. Discussion Anopheles arabiensis was the dominant member of the An. gambiae complex in the study area. Nearly all breeding of both An. gambiae complex species and culicines took place in rice paddies within 350m zones from the edge of the paddies closest to human settlement. Both types of mosquitoes were collected together and were more likely to be found on the edges of paddies compared to the centres. Most aquatic invertebrates were also more frequently found in rice fields close to villages. In contrast, fish were more common in rice fields close to the river. It is impossible to be certain that fish are important predators of mosquitoes in this setting since mosquitoes may simply colonise different habitats to fishes. Here the paddies containing highest numbers of mosquitoes were situated closest to the village and were demarcated by raised embankments to prevent salt water encroachment, which could have also prevented fish from getting in them. Both cow dung (Organic manure) and algae (Biofertilizer) favoured larval breeding of both anopheline and culicine species. These results suggest that since most cow dung was found in paddies close to the settlement, this may also have contributed to higher numbers of larvae in the same areas. Larval control measures should therefore target rice fields close to human settlements and along the edges of rice paddies for successful source reduction.