Title page Title: Malaria incidence in relation to rice cultivation in the irrigated Sahel of Mali Authors Olivier J.T. Briët, IWMI, P.O. Box 2075, Colombo, Sri Lanka. Phone +94 12 787 404, Fax: +94 12 786 854, E-mail: [email protected] (corresponding author) Keita H. Dembélé, Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de Pharmacie et d’Odonto-Stomatologie, BP 1805, Mali. Fax: +223 228 109, Email: [email protected] Alassane Dicko, Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de Pharmacie et d’Odonto-Stomatologie, BP 1805, Mali. Fax: +223 228 109, Email: [email protected] Ogobara K. Doumbo, Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de Pharmacie et d’Odonto-Stomatologie, BP 1805, Mali. Fax: +223 228 109, Email: [email protected] (honorary corresponding author) Christophe Rogier Institut de Médecine Tropicale du Service de Santé des Armées, Le Pharo, Marseilles. B P 46, 13998 Marseille-armées. France. E-mail: [email protected] Issaka Sagara, Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de Pharmacie et d’Odonto-Stomatologie, BP 1805, Mali. Fax: +223 228 109, E-mail: [email protected] Mady Sissoko, Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de Pharmacie et d’Odonto-Stomatologie, BP 1805, Mali. Fax: +223 228 109, E-mail: [email protected] Mahamadou S. Sissoko, Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de Pharmacie et d’Odonto-Stomatologie, BP 1805, Mali. Fax: +223 228 109, Email: [email protected] Moussa Sogoba, Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de Pharmacie et d’Odonto-Stomatologie, BP 1805, Mali. Fax: +223 228 109, Email: [email protected] Yeya T. Touré, Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de Pharmacie et d’Odonto-Stomatologie, BP 1805, Mali. Fax: +223 224 987, Email: [email protected]

Malaria incidence in relation to rice cultivation in the irrigated Sahel of Mali Mahamadou Soumana Sissoko a, Alassane Dicko a; Olivier Johan Tavai Briët b,*, Mady Sissoko a

, Issaka Sagara a, Hawa Dembélé Keita a, Moussa Sogoba a, Christophe Rogier c, Yeya

Tiémoko Touré a, Ogobara K. Doumbo a, ** a

Département d’Epidémiologie des Affections Parasitaires, Faculté de Médecine, de

Pharmacie et d’Odonto-Stomatologie, BP 1805, Bamako, Mali. b

West Africa Rice Development Association, 01 BP 2551 Bouaké 01, Bouaké, Côte d’Ivoire.

c

Institut de Médecine Tropicale du Service de Santé des Armées, Le Pharo, B P 46, 13998

Marseille-armées, France. Summary Seven repeated cross-sectional parasitological surveys, collecting a total of 13912 blood samples, were carried out from September 1995 to February 1998 in three irrigated rice growing villages and three villages without irrigated agriculture in the area surrounding Niono, Mali. Parasite prevalence varied according to season and agricultural zone, but showed similar patterns for villages within the same zone. Overall, malaria prevalence was 47% in the villages without irrigated agriculture and 34% in the irrigated rice growing villages. In a village in the irrigated zone, and a village in the non-irrigated zone, 1067 and 608 children up to the age of 14 years, respectively, were followed in a passive malariological study for the period of 13 months. Fevers were attributed to malaria using a statistical method, taking into account the parasitaemia in afebrile controls from the cross-sectional surveys. The incidence of malaria fevers differed markedly between the two zones and over time. In the village in the irrigated zone, the incidence of malaria fevers was fairly constant over the year at 0.7 per 1000 children per day. In the village without irrigated agriculture, incidence was low during the dry season (at 0.6 per 1000 children per day), whereas it was high during the rainy season (at 3.2 per 1000 children per day). These results correspond well to the malaria transmission observed in a concurrent entomological survey. Rice cultivation in the semi-arid sub-Saharan environment altered the transmission pattern from seasonal to perennial, but reduced annual incidence more than two-fold. Key words: Plasmodium falciparum, malaria, incidence, rice, irrigation, semi-arid, Mali *Corresponding author. Present address: IWMI, P.O. Box 2075, Colombo, Sri Lanka. Tel.: +94 12 787 404; fax: +94 12 786 854.

E-mail address: [email protected] (O.J.T. Briët) **Honorary corresponding author

Introduction In West Africa, rice is becoming increasingly popular as a staple food for the growing population. To meet the growing demand, expansion of arable land by irrigation is perhaps the most effective, especially in the arid and semi-arid areas. Large scale rice irrigation projects have been established to meet this demand and to assure economic growth. As a consequence, the rice production in sub-Saharan Africa has more than doubled during the last tree decades. Although the irrigated Sahel represents only 7% of the area cultivated with rice in West Africa, it is responsible for 18% of the region’s rice production (WARDA 1999). Irrigation generally favours the development of large populations of disease vectors such as anopheline mosquitoes responsible of transmission of malaria. There is, therefore, a concern that irrigation leads to increased transmission and morbidity. However, the relationship between rice irrigation and malaria is not straightforward and varies according to endemicity and seasonality. Several studies in stable malaria endemic areas have shown that malaria prevalence or incidence is equal or less in irrigated-rice growing areas compared with neighbouring areas without irrigated rice cultivation (Lindsay et al. 1991; Faye et al. 1993; Faye et al. 1995; Boudin et al. 1992; Couprié et al. 1985; Josse et al. 1987; Henry et al. 2003). One study, however, did show a positive association of rice cultivation with parasitaemia prevalence, but the villages with rice cultivation were far remote from health care facilities and were isolated during the rainy season (Gonçalves et al. 1996). Yet another study found both high and low prevalences in villages close to undeveloped or developed swamps for rice cultivation (Gbakima 1994). In situations where malaria is absent or unstable, introduction of irrigated rice cultivation may enhance malaria morbidity and mortality. This has often been described in studies outside Africa (Lacey and Lacey 1990; Service 1989; Tyagi and Chaudhary 1997). In Africa, negative effects of irrigated rice cultivation or irrigation structures such as microdams were observed in areas at moderate altitude and in highlands (Coosemans 1991; Laventure et al. 1996; Ghebreyesus et al. 1999). Irrigation might push malaria transmission over a threshold (Bradley 1988) in areas where transmission would otherwise be very low or non-existent, such as desert fringes and highlands, due to unfavourable conditions for vector and parasite development (Shililu et al. 1998). In the southern Sahel of Mali, with an annual rainfall above 400 mm, climatic conditions are suitable for stable, yet seasonal transmission of malaria (Craig et al. 1999). Irrigation, especially during the dry season, might alter the transmission pattern from seasonal to annual. This study examined whether irrigated rice farming in the Sahel alters the dynamic and magnitude of malaria infection and morbidity. Material and methods

Study area The district of Niono is located in the Ségou Region, a Sahelian area of Mali, 350 km from Bamako. The district is accessible throughout the year. The rainy season lasts 5 months

(June to October). The irrigation scheme ‘Office du Niger’ divides the area into two zones: an irrigated zone and a typical Sahelian environment (non-irrigated zone) outside the irrigated perimeter. For a further description of the study area, see Dolo et al. in this issue (2003). After a preliminary study in September 1995, six typical villages, three in the irrigated perimeter that practice double rice cropping (Ténégué, Tissana, Niessoumana) and three from outside the irrigated perimeter (the non-irrigated zone) (Dokobougou, Toumakoro and Kalanampala), with mainly millet cultivation, were selected.

Data collection: Cross-sectional surveys Seven cross-sectional malariological surveys were performed in each village studying children of 0-14 years of age. We expected this age range to be the most vulnerable to malaria morbidity in the endemic setting. Two surveys were carried out in the cold dry season: in January / February 1997 and February / March 1998; two in the hot dry season when the offseason rice crop was cultivated in the irrigated zone: in April 1996 and April / May 1997; and three were made in the rainy season: in September 1995, September 1996 and October / November 1997. In the surveys of September 1996 and January / February 1997, one out of two households was selected at random and studied. For the other surveys, the sample size was exhaustive. A census was done in September 1995 in each village. In each village, a temporary medical post was set up. A team consisting of a medical doctor, a medical student, a pharmacist, a laboratory technician and two locally-recruited assistants spent about one week in each village for each survey. For each survey, all villages were visited within the period of 25 days. During the first survey, a census was taken of the whole population. All children up to the age of 14 years were interviewed. Age, sex, weight, medical history and drug use during the preceding 15 days were recorded. All children were clinically examined (including palpation for splenomegaly) and those found ill were treated. Axillary temperature was taken using electronic thermometers. In the field laboratory, thick blood smears were made from all children. The blood smears were stored in WHO collection boxes and stained with Giemsa 3% for 45 minutes, 24 hours after collection. Blood slides were read in the main laboratory in Bamako (DEAP) with ocular 10, objective 100, and the species and number of malaria parasites per 300 leukocytes was recorded. This was based on the assumption that the mean concentration of the blood was 7500 leukocytes/µl. The accuracy of the readings was checked on 10% of randomly selected blood smears.

Data collection: Longitudinal survey From 1 October 1996 to 31 October 1997, a longitudinal survey was conducted in Kalanampala and Ténégue. A medical student was permanently based in each village, assisted by two locally-recruited and instructed health workers. All parents were asked to present their children to the medical student in case of perceived illness. Upon presentation, children were interviewed, their temperature recorded, and abdomens palpated. For those

children with an axillary temperature of 37.5°C or above, a thick smear was made, which was prepared as in the cross-sectional surveys. Those children with an axillary temperature superior to 38.5°C received paracetamol at 15−30mg/kg body weight per day. On the basis of the clinical examination, children diagnosed with malaria were treated following the guidelines of the the National Malaria Control Program. Mild malaria cases were treated with chloroquine 25 mg/kg per treatment (10 mg/kg first and second day, 5 mg/kg on the third day). Drugs were taken in front of the medical staff. Patients with severe and complicated malaria (convulsions, coma, severe anaemia, etc.) were transfered to the district hospital of Niono and treated with quinine 25mg/kg per day for three days.

Data management and analysis: Prevalence Data were recorded on standard case report forms and and checked in the field. Data were entered and cleaned in the computer program Epi Info. The data of the cross-sectional surveys were analyzed independently to determine how the prevalence of asexual forms varied over the surveys, villages and age groups, using multiple binomial regression (R statistical package).

Data management and analysis: Incidence Estimates of the number of illnesses attributable to malaria infection were calculated from the patients presented at the longitudinal surveys with an axillary temperature equal or superior to 37.5°C. Blood smears taken from asymptomatic children (with no illness on the day of examination or during the preceding 15 days) from those cross-sectional surveys which fell just before or during the longitudinal survey (September 1996, January 1997, April 1997, October 1997) served as controls. Of those children that had two blood slides with different scores taken on the same day, both samples were excluded when it was not clear which one was erroneous. Those readings of children that became ill within 3 days (as observed in the longitudinal survey) were excluded from the controls. The fraction of fever cases attributable to malaria was calculated following the binomial regression method described by Smith et al. (1994a; 1994b). The odds ratio (OR) for the risk of fever at each level of parasite density was estimated in a binomial regression model taking into account the effects of age, season and zone and their interaction terms with parasite density. The binomial regression model was built using the statistical package R. The basic binomial regression model used was: Log (pi/(1-pi))=β0+β1xi where pi is the fitted probability τ

that individual i is a case rather than a control, xi is the parasite density to the power of τ, τ

and β0, β1 and τ are parameters to be estimated. For the finding of the exponent τ, age classes (see Table 1) were fitted as an additional linear term in the model. Several models with different values of τ were fitted. The value of τ kept for the following analysis was the τ value of the model that had the lowest deviance, i.e. gave the best fit to the data set. The additional linear terms ‘village and ‘survey period’ were added to the model consisting of ‘age’

and ‘parasite density to the power of τ’. Statistical significance of the effect of these variables and their interaction terms was tested using likelihood ratio tests. Only significant (P < 0.05) terms remained in the model. For each case of fever, the OR for the risk of fever corresponding to the parasite density observed in that case was calculated according to the estimates of the parameters of the final model, in comparison with the risk of fever for the same case if the parasite density was nill. Then, for each case, the attributable fraction was calculated as AF=(OR-1)/OR. That is the probability that a fever case exposed to its parasite density can be attributed to malaria. The clinical malaria incidence densities were calculated by dividing the number of malaria attacks by the number of person-days under survey in each group and the incidence density ratios were tested using a Poisson regression model (R statistical package).

Ethical consideration The study was approved by the FMPOS IRB based on the understanding of local cultural behaviour. Community oral informed consent was obtained from the villages before each cross-sectional survey and oral informed consent was obtained from each family. Results

Population Table 1 shows the distribution of children in age groups in the studied villages, as counted in the census at the first cross-sectional survey (September 1995). Ténégué was the largest of the three villages studied in the irrigated zone. It was 1.7 times larger than Kalanampala, which was the largest of the three villages studied in the non-irrigated zone. There was no significant difference in age distributions among villages or of sex distribution within age classes (Chi-square test, P < 0.05). However, in a comparison of the malaria incidence between zones for all age groups pooled, the incidence figures were standardised using the average population distribution of the compared villages. In the irrigated zone, the Mianka are the most important ethnic group, making up 68%, 45%, and 75% of the population in Ténégué, Tissana and Niessoumana, respectively. The Bamana are also important in Tissana (24%). In the Non irrigated zone, in Toumakoro and Kalanampala, the Bamana are the most important ethnic groups with 81% and 95% respectively. In Dokobougou, the Peulh are more important than the Bamana (73% versus 21%). In both zones, the Islam is the most important religion (> 95%).

Malaria parasitology In both zones, Plasmodium falciparum was the most important malaria species (Table 2).

Plasmodium malariae was also present but only few children carried this infection. Overall, in the non-irrigated zone, the prevalence of P. falciparum and of P. malariae asexual forms was significantly higher than in the irrigated zone (Mantel-Haenszel Chi square, P < 0.05).

Prevalence of P. falciparum varied strongly with age group, generally increasing with age. Figure 1 shows the parasite prevalences for the different cross-sectional surveys in the villages. Within each village, there was a large variation over surveys; even for surveys that were performed in the same season but in different years, parasite rates were significantly different (Mantel-Haenszel Chi square, P < 0.05). Although the prevalence of P. falciparum infection showed somewhat similar patterns for villages within each zone, in multiple binomial regression, the interaction term of ‘villages with surveys’ was significant within each zone (P

< 0.05). The prevalence of P. falciparum infection was significantly higher in the non-irrigated zone than in the irrigated zone for the rainy season surveys of September (multiple binomial regression with zone and age group as explanantory variables, OR = 3.9; 95% Confidence Interval: 3.2−4.7, OR = 2.9; 95% CI: 2.3−3.8, and OR = 4.1; 95% CI: 3.4−4.9 for 1995, 1996 and 1997, respectively), and the hot dry season survey of April 1996 (OR = 1.4; 95% CI: 1.2−1.7). However, for the April / May 1997 survey, it was higher in the irrigated zone (OR = 1.9; 95% CI: 1.6−2.3), and during the cold dry season surveys in January / February 1997 and February / March 1998, there was no significant difference. In the irrigated zone, Niessoumana had consistently higher parasitological indices than Ténégué and Tissana, which were very similar.

Binomial regression models and incidence Figure 2 shows the proportional distribution over parasite classes of asymptomatic children from the villages Kalanampala (non-irrigated zone) and Ténégué (irrigated zone), subject to the cross-sectional surveys held shortly before, during and shortly after the longitudinal study, by age group. The distribution over the parasite classes differed markedly between age groups, villages and seasons. In the village without irrigated agriculture, during the rainy season, the parasite load in asymptomatic children was much higher than during the dry season. In the irrigated rice growing village, however, the opposite was observed: during the rainy season, the parasite load was lower than during the dry season. Terms significant in the binomial regression model were: ‘parasite density’, ‘age’, ‘village’, ‘survey period’, and interaction terms ‘village with period’, ‘village with age’, ‘age with period’, ‘parasite density with age’, ‘parasite density with village’, ‘parasite density with period’, and ‘parasite density with village and with period’. The risk of becoming a fever case due to malaria is estimated by the parasite density and the interaction terms of the parasite density with the other variables. The best-fitting value of the exponent τ was 0.49. Figure 3 shows that in Kalanampala (non-irrigated zone), the total fever incidence had a large amplitude, with a marked seasonal peak in September, whereas in Ténégué (irrigated zone) the incidence fluctuated much less over the seasons, with small peaks in December, April and October. The seasonality of the malaria-attributable fevers was more marked in both villages. Figure 4 shows the malaria-attributable fever incidence over the months studied, as well as the malaria transmission (EIR). In Kalanampala (non-irrigated zone), the incidence of malaria

was significantly higher in the rainy season (3.2 per 1000 children per day) than in the dry season (0.6 per 1000 children per day; Incidence Density Ratio = 5.4; 95% CI: 4.2−6.9). For Ténégué (irrigated zone), this difference was also significant, yet less pronounced (0.8 vs 0.6 cases per 1000 children per day; IDR = 1.5; 95% CI: 1.2−1.8). During the rainy season, the incidence of malaria in Kalanampala (non-irrigated zone) was significantly higher than in Ténégué (irrigated zone) ( IDR = 3.4; 95% CI: 2.3−5.0). At the beginning of the dry season (November and December), the incidence in Kalanampala diminished, whereas that in Ténégué increased. The incidence in Ténégué was at a higher level than in Kalanampala during the hot dry season (January−April), the period when the offseason rice crop is grown in the irrigated zone. Over the period of 13 months, the incidence was more than two-fold higher in Kalanampala (non-irrigated zone) than in Ténégué (irrigated zone) (1.8 vs 0.8 cases per 1000 children per day; IDR = 2.4; 95% CI: 2.1−2.7). The seasonal pattern of malaria incidence in the village without irrigated agriculture corresponded well to the transmission observed in the concurrent entomological study (Dolo et al. 2003): very low incidence when transmission was not detectable during the dry season, and a very high incidence and transmission at the end of the rainy season. In the irrigated rice growing village, both incidence and transmission were at a low level, but the fluctuations in incidence were not explained by the fluctuations in transmission. Discussion In the district of Niono, malaria was meso-endemic, with a prevalence in 0−14 year olds between 11 and 77%, depending on the season and year. The variability of the levels of parasitaemia in asymptomatic children (which varied with age, survey and zone) illustrates the need for the use of an attributable fraction model. With a fixed parasitaemia threshold for the definition of clinical cases, the incidence density ratios between zones and seasons were larger (except for the irrigated zone, rainy season vs dry season). Annual incidence density of malaria episodes in 0−9 year old children was about 0.7 in the non-irrigated zone, and about 0.3 in the village in the irrigated zone. This was a factor 3 to 7 lower than the annual incidences found by Henry et al. (2003) in the savannah zone of Côte d'Ivoire, where malaria is hyper-endemic (prevalence in the 0−9 year olds >70%). At the relatively low level of transmission in the semi-arid Sahel, there was a strong relationship between the number of infectious bites and the number of malaria episodes, which is illustrated by the strong seasonal variation in both malaria transmission and incidence in the non-irrigated zone. This seasonal pattern of malaria incidence in the village without irrigated agriculture is very similar to that of southern Niger (with a similar climate), which generally shows a peak in October (Julvez et al. 1992). The results from our study are comparable to those of a recent study in northern Tanzania, which showed 2.4 to 2.7 times higher incidence of malaria in a non–irrigated area

compared to a rice irrigated area, at low malaria prevalence levels of 12.5% and 29.4% (in 1–4 year olds) for the irrigated and non irrigated areas, respectively (Ijumba et al. 2002). At higher levels of annual transmission, such as in the savannah zone of Côte d'Ivoire, malaria incidence tends to show a dampened response to seasonal variation in transmission, up to one infectious bite per night (Charlwood et al. 1998), and the marginal impact of transmission reducing strategies is lower (Smith et al. 1998; Smith et al. 1993; McElroy et al. 1994; Trape and Rogier 1996). The more than two-fold lower annual incidence in the irrigated zone (as compared to the non-irrigated zone) is most likely directly related to lower transmission (Dolo et al. 2003). Transmission was still sufficient to warrant immunity development in the population; the fact that it was perennial may have improved the immune status of the population in the irrigated zone as compared to the non irrigated zone. The lower overall transmission can be ascribed to wide bed net use in the presence of alternative hosts and reduced vector longevity (Dolo et al. 2003). Access to health care and antimalarials may have played a role as well. Irrigation can contribute to improvement of socio-economic status and therefore access to health care. However, irrigation does not necessarily improve health status. Sometimes, irrigated agriculture may lead to a reduction in the capacity of households to manage disease episodes (De Plaen et al. 2003). Aside from socio-economic status, differences in knowledge may also have an impact on behaviour. In our study area, this was unlikely, as an anthropological study showed a similar level of knowledge on protective measures and auto-medication in both the irrigated and non irrigated zone (Sagara

et al., unpublished data). In conclusion, irrigated rice cultivation in the Sahelian environment altered the malaria incidence patterns from seasonal to perennial, but reduced annual incidence more than two-fold. Acknowledgements This study was undertaken within the framework of the WARDA/ WHO-FAO-UNEP PEEM/ IDRC/ DANIDA/ Norway Health Research Consortium on the Association between Irrigated Rice Ecosystems and Vector-borne Diseases in West Africa. The Consortium received financial support from the International Development Research Center (IDRC), Ottawa, Canada, the Danish International Development Agency (DANIDA) and the Royal Government of Norway. References Boudin, C., Robert, V., Carnevale, P. and Ambroise, T. P.; 1992. Epidemiology of Plasmodium

falciparum in a rice field and a savanna area in Burkina Faso. Comparative study on the acquired immunoprotection in native populations. Acta Trop. 51, 103-111. Bradley, D. J.; 1988. The epidemiology of ricefield-associated diseases. In: Vector borne disease control in humans through rice agroecosystem management. Proceedings of a workshop jointly sponsored by Int. Rice Res. Inst. and the WHO/FAO/UNEP Panel

of Experts on Environmental Management for Vector Control, International Rice Research Institute, Manila, pp. 29-39. Charlwood, J. D., Smith, T., Lyimo, E., Kitua, A. Y., Masanja, H., Booth, M., Alonso, P. L. and Tanner, M.; 1998. Incidence of Plasmodium falciparum infection in infants in relation to exposure to sporozoite-infected anophelines. Am.J.Trop.Med.Hyg. 59, 243-251. Coosemans, M.; 1991. Development of a campaign strategy against malaria in a rice-growing region of Burundi. Bull.Mem.Acad.R.Med.Belg. 146, 157-165. Couprié, B., Claudot, Y., Same-Ekobo, A., Issoufa, H., Léger-Debruyne, M., Tribouly, J. and Ripert, C.; 1985. Etude épidemiologique du paludisme dans les régions rizicoles de Yagoua et de Maga (Nord-Cameroon). Bull.Soc.Pathol.Exot.Filiales 78, 191-204. Craig, M. H., Snow, R. W. and Le Sueur, D.; 1999. A climate based distribution model of malaria transmission in sub-Saharan Africa. Parasitol.Today 15, 105-111. De Plaen, R., Geneau, R., Teuscher, T., Koutoua, A. and Seka, M. L.; 2003. Living in the paddies: a social science perspective on how inland valley irrigated rice cultivation affects malaria in Northern Cote d'Ivoire. Trop.Med.Int.Health 8, 459-470. Dolo, G., Briët, O. J. T., Dao, A., Traoré, S. F., Bouaré, M., Sogoba, M., Niaré, O., Bagayogo, M., Sangare, D., Doumbo, O. and Touré, Y. T.; 2003. Malaria transmission in relation to rice cultivation in the irrigated Sahel of Mali. this journal ?, 0000-0000. Faye, O., Fontenille, D., Gaye, O., Sy, N., Molez, J. F., Konate, L., Hebrard, G., Herve, J. P., Trouillet, J., Diallo, S. and Mouchet, J.; 1995. Paludisme et riziculture dans le delta du fleuve Sénégal. Ann.Soc.Belg.Med.Trop. 75, 179-189. Faye, O., Gaye, O., Herve, J. P., Diack, P. A. and Diallo, S.; 1993. Malaria in the sahelian zone of Senegal. 2. Parasitic indices. Ann.Soc.Belg.Med.Trop. 73, 31-36. Gbakima, A. A.; 1994. Inland valley swamp rice development: malaria, schistosomiasis, onchocerciasis in South Central Sierra Leone. Public Health 108, 149-157. Ghebreyesus, T. A., Haile, M., Witten, K. H., Getachew, A., Yohannes, A. M., Yohannes, M., Teklehaimanot, H. D., Lindsay, S. and Byass, P.; 1999. Incidence of malaria among children living near dams in northern Ethiopia: community based incidence survey. Brit.Med.J. 319, 663-666. Gonçalves, A., Ferrinho, P. and Dias, F.; 1996. The epidemiology of malaria in Prabis, GuineaBissau. Mem.Inst Oswaldo Cruz 91, 11-17. Henry, M. C., Rogier, C., Nzeyimana, I., Assi, S. B., Dossou-Yovo, J., Audibert, M., Mathonnat, J., Keundjian, A., Akodo, E., Teuscher, T. and Carnevale, P.; 2003. Inland valley rice production systems and malaria infection and disease in the savannah of Cote d'Ivoire. Trop.Med.Int.Health 8, 449-458. Ijumba, J. N., Shenton, F. C., Clarke, S. E., Mosha, F. W. and Lindsay, S. W.; 2002. Irrigated crop production is associated with less malaria than traditional agricultural practices in Tanzania. Trans.R.Soc.Trop.Med.Hyg. 96, 476-480.

Josse, R., Josseran, R. and Audibert, M.; 1987. Malariometry and seasonal variations of malaria indices in the rice fields of Maga (north Cameroon) and in the adjacent region. Cah.ORSTOM, Ser.Entomol.Med.Parasitol. printed 1989, 25, 63-71. Julvez, J., Develoux, M., Mounkaila, A. and Mouchet, J.; 1992. Diversity of malaria in the sahelo-saharian zone. A review of the situation in Niger, West Africa. Ann.Soc.Belg.Med.Trop. 72, 163-177. Lacey, L. A. and Lacey, C. M.; 1990. The medical importance of riceland mosquitoes and their control using alternatives to chemical insecticides. J.Am.Mosq.Control Assoc. 6, 1-93. Laventure, S., Mouchet, J., Blanchy, S., Marrama, L., Rabarison, P., Andrianaivolambo, L., Rajaonarivelo, E., Rakotoarivony, I. and Roux, J.; 1996. Rice, a source of life and death on the Madagascar plateaux. Sante 6, 79-86. Lindsay, S. W., Wilkins, H. A., Zieler, H. A., Daly, R. J., Petrarca, V. and Byass, P.; 1991. Ability of Anopheles gambiae mosquitoes to transmit malaria during the dry and wet seasons in an area of irrigated rice cultivation in The Gambia. J.Trop.Med.Hyg. 94, 313-324. McElroy, P. D., Beier, J. C., Oster, C. N., Beadle, C., Sherwood, J. A., Oloo, A. J. and Hoffman, S. L.; 1994. Predicting outcome in malaria: correlation between rate of exposure to infected mosquitoes and level of Plasmodium falciparum parasitemia. Am.J.Trop.Med.Hyg. 51, 523-532. Service, M. W.; 1989. Rice, a challenge to health. Parasitol.Today 5, 162-164. Shililu, J. I., Maier, W. A., Seitz, H. M. and Orago, A. S.; 1998. Seasonal density, sporozoite rates and entomological inoculation rates of Anopheles gambiae and Anopheles

funestus in a high-altitude sugarcane growing zone in Western Kenya. Trop.Med.Int.Health 3, 706-710. Smith, T., Armstrong-Schellenberg, J. R. M. and Hayes, R.; 1994a. Attributable fraction estimates and case definitions for malaria in endemic areas. Stat.Med. 13, 23452358. Smith, T., Charlwood, J. D., Kihonda, J., Mwankusye, S., Billingsley, P., Meuwissen, J., Lyimo, E., Takken, W., Teuscher, T. and Tanner, M.; 1993. Absence of seasonal variation in malaria parasitaemia in an area of intense seasonal transmission. Acta Trop. 54, 5572. Smith, T., Charlwood, J. D., Kitua, A. Y., Masanja, H., Mwankuyse, S., Alonso, P. L. and Tanner, M.; 1998. Relationships of malaria morbidity with exposure to Plasmodium

falciparum in young children in a highly endemic area. Am.J.Trop.Med.Hyg. 59, 252257. Smith, T., Genton, B., Baea, K., Gibson, N., Taime, J., Narara, A., Al-Yaman, F., Beck, H. P., Hii, J. and Alpers, M.; 1994b. Relationships between Plasmodium falciparum infection and morbidity in a highly endemic area. Parasitol. 109, 539-549.

Trape, J. F. and Rogier, C.; 1996. Combating malaria morbidity and mortality by reducing transmission. Parasitol.Today 12, 236-240. Tyagi, B. K. and Chaudhary, R. C.; 1997. Outbreak of falciparum malaria in the Thar Desert (India), with particular emphasis on physiographic changes brought about by extensive canalization and their impact on vector density and dissemination. J.Arid.Env. 36, 541-555. WARDA; 1999. Medium Term Plan 2000-2002. West Africa Rice Development Association, M'be, Ivory Coast, 1-117. Figure legends Fig. 1 The prevalence of asexual forms of P. falciparum parasites in children upto 14 years of age during seven cross-sectional surveys in three villages in the non-irrigated zone (filled squares = Kalanampala; filled circles = Toumakoro; filled triangles = Dokobougou) and three villages in the irrigated zone (open squares = Ténégué; open circles = Tissana; open triangles = Niessoumana). Fig. 2 The proportional distribution of asymptomatic children over parasite classes (filled bars ( ) ≥ 5000; gray bars ( ) 2500−4999; densely speckled bars ( ) 1000−2499; medium densely speckled bars ( ) 500−999; lightly speckled bars ( ) 100−499; open bars ( ) 1−99 trophozoites/µl) from (a) Kalanampala in the non-irrigated zone, and (b) Ténégué in the irrigated zone, presented by survey and age group. Fig. 3 Incidence of fevers (complete bars) and malaria-attributable fevers (filled bars) per month in (a) Kalanampala (non-irrigated zone) and (b) Ténégué (irrigated zone), standardized for the average age distribution of the villages Kalanampala and Ténégué. Fig. 4 Malaria-attributable fever incidence in Kalanampala (filled squares and solid lines) and Ténégué (open squares and solid lines) over the months studied, and the entomological inoculation rates from Dolo et al. (2003)(filled and open circles for Kalanampala and Ténégué, respectively). Tables Table 1

The distribution of the study population by village and age group from

census data in September 1995. Age group

Irrigated

Non-irrigated zone

zone Ténégué

Tissana

Niessoumana Dokobougou Toumakoro

Kalanampala

0-1

286

259

111

66

86

146

2-4

239

191

95

66

76

135

5-9

314

277

125

106

100

187

10-14

228

221

72

70

73

140

Table 2

The prevalence of malaria infection in asymptomatic children and in children

that were ill during the 15 preceding days, by species, in the non-irrigated zone and in the irrigated zone. Data from the 7 cross-sectional surveys were pooled. Asymptomatic children Non-irrigated

Symptomatic children

Irrigated zone Non-irrigated zone Irrigated zone

zone number of slides

3308

5826

1443

3335

46.5

31.3

42.2

35.8

% P. malariae

1.1

0.6

1.2

0.7

% P. falciparum +

0.9

0.4

1.5

0.8

48.5

32.3

44.9

37.3

Malaria species % P. falciparum

P. malariae Total prevalence