Malaria transmission dynamics in central Côte d'Ivoire: the influence of changing patterns of irrigated rice agriculture

Malaria transmission dynamics in central Cote d’Ivoire:the influence of changing patterns of irrigated riceagriculture

B. G. KOUDOU1,2 , 3 , Y. TANO2, M. DOUMBIA4, C. NSANZABANA5,

G. CISSE1 , O. GIRARDIN1, D. DAO1, 2 , E . K. N’GORAN1,2 , P . VOUNATSOU3,

G. BORDMANN3, J . KEISER3, M. TANNER3 and J. UTZINGER3

1Centre Suisse de Recherches Scientifiques, Abidjan, Cote d’Ivoire, 2Universite d’Abidjan-Cocody, Abidjan, Cote d’Ivoire,3Swiss Tropical Institute, Basel, Switzerland, 4Universite d’Abobo-Adjame, Abidjan, Cote d’Ivoire and 5Universite de

Neuchatel, Neuchatel, Switzerland

Abstract. The dynamics of malaria transmission was studied comparatively in thevillages of Zatta and Tiemelekro, central Cote d’Ivoire, from February 2002 toAugust 2003. Prominent agroecosystems in these villages are irrigated rice grow-ing and vegetable farming, respectively. Mosquitoes (Diptera: Culicidae) werecollected on human bait at night and by pyrethrum knock-down spray sheetcollections at four randomly selected sentinel sites in each village. In 2002, for atotal of 96 man-nights per village, 7716 mosquitoes were collected in Zatta and3308 in Tiemelekro. In 2003, with half the sampling effort, 859 and 2056 mosqui-toes were collected in Zatta and Tiemelekro, respectively. Anopheles gambiaeGiless.l. was the predominant mosquito and the key malaria vector throughout, fol-lowed by An. funestus Giles. Anthropophily among adult female Anophelesexceeded 95% in both villages. Comparison between years revealed that the bitingrate of An. gambiae s.l. in Zatta decreased several-fold from 49.3 bites per personper night (b/p/n) in 2002 to 7.9 b/p/n in 2003 (likelihood ratio test(LRT)¼ 1072.66; P< 0.001). Although the biting rate remained fairly constantin Tiemelekro, the difference between years was significant (16.1 vs. 18.2 b/p/n;LRT¼ 148.06; P< 0.001). These observations were paralleled by a markeddecrease in the infective rate of An. gambiae s.l. in Zatta (4.6–1.2%), and anincrease in Tiemelekro (3.1–7.6%). Meanwhile, the entomological inoculationrate of An. gambiae s.l. decreased 21-fold in Zatta, from 789 to 38 infective bitesper person per year (ib/p/y), whereas it remained high in Tiemelekro (233 vs.342 ib/p/y). The interruption of irrigated rice growing in Zatta in 2003, conse-quential to a farmers’ conflict over land, might be the underlying cause for thesignificant reduction in malaria transmission, whereas more stable conditionsoccurred in Tiemelekro.

Key words. Anopheles gambiae s.l., An. funestus, biting rate, entomologicalinoculation rate, irrigated rice growing, malaria, parity ratio, sporozoite rate,vegetable farming, Cote d’Ivoire.

Introduction

Over the past 30 years, a reduction in the mean annualprecipitation occurred in sub-Saharan Africa, particularly

Correspondence: Jurg Utzinger, Department of Public Health

and Epidemiology, Swiss Tropical Institute, P. O. Box, CH-4002

Basel, Switzerland. Tel.: þ 41 61 225 2666; fax: þ41 61 225 2678;

e-mail: [email protected]

Medical and Veterinary Entomology (2005) 19, 27–37

# 2005 The Royal Entomological Society 27

in the Sahel (Hulme, 2001). In turn, this has importantimplications for meeting current and future needs of foodsecurity of the predominantly agricultural-based macroeco-

nomies. Despite significant imports of cereal, the numberof malnourished children living in sub-Saharan Africaincreased by more than 75% in the last three decades; in

1997 it was estimated that 32.7 million children under5 years of age were malnourished (Rosegrant & Meijer,2002). This must be juxtaposed with rapid population

growth that outpaces production increases in agriculturein most countries of sub-Saharan Africa (Ijumba et al.,2002a; De Plaen et al., 2003). To remedy this situation,thousands of small irrigation and multi-purpose dams

have been constructed in Africa, and the irrigated land sur-face has been extended concurrently (Amerasinghe, 2003;Keiser et al., 2004). In Cote d’Ivoire, for example, an esti-

mated 500 small dams have been built in the last threedecades (Aka et al., 2000).It is currently estimated that irrigated rice agroecosys-

tems represent 55% of the world’s harvested rice area andcontribute to approximately three-quarters of the world’srice production (Keiser et al., 2002). Although irrigated ricecultivations usually result in higher yields, these often come

along with the creation of additional and more permanentbreeding sites, which lead to higher densities of mosquitoes.Such observations have been made in Kenya (Chandler &

Highton, 1975), Burkina Faso (Robert et al., 1988), TheGambia (Lindsay et al., 1991) and Mali (Dolo et al., 2004).However, the impact on the transmission dynamics of mos-

quito-borne diseases in general, and malaria in particular;hence on incidence and prevalence rates in human popula-tions, is not straightforward. It is governed by several con-

textual determinants. For example, some studies suggestthat very high mosquito densities can lead to reduced long-evity of mosquitoes, and hence reduce transmission (Robertet al., 1992; Klinkenberg et al., 2002; Dolo et al., 2004).

Other important underlying risk factors that are drivers ofthe changing patterns of malaria transmission are the over-all level of malaria endemicity, common agriculture prac-

tices in a given agroecosystem, proximity of humansettlements to productive breeding sites, means of andcoverage with personal protective measures, and access to

health care (Ijumba & Lindsay, 2001; Doannio et al., 2002;Staedke et al., 2003; van der Hoek et al., 2003). Recentstudies came to the conclusion that for most parts of sub-

Saharan Africa that are characterized by stable malariatransmission, the introduction of irrigated agriculture hadno or only little impact on malaria transmission (Ijumba &Lindsay, 2001; Ijumba et al., 2002b; Keiser et al., 2004;

Sissoko et al., 2004). These observations have beenexplained, at least partially, by high proportions of zoophilyamong potential malaria vectors (Robert et al., 1985; Faye

et al., 1993), good access to effective antimalarial drugs(Faye et al., 1992; Ijumba & Lindsay, 2001), and personalprotection, e.g. sleeping under insecticide-treated nets

(ITNs) (Ijumba & Lindsay, 2001; De Plaen et al., 2003;Henry et al., 2003). On the other hand, irrigated rice grow-ing, particularly in the semi-arid savannah zone of Africa

can alter the malaria transmission pattern from seasonal toperennial (Dolo et al., 2004; Keiser et al., 2004; Sissoko et al.,2004).

We have recently noticed that location of human habita-tion to irrigated ricefields, the practice of double rice culti-vation (regardless of whether or not it is synchronized

among plots) and intensive vegetable farming influence thetransmission dynamics of malaria in rural parts of centralCote d’Ivoire (Girardin et al., 2004). The purpose of the

current study is to advance our understanding of theserelationships; hence we carried forward a comparativeappraisal of key malaria transmission parameters in twovillages with different agroecosystems. One village is located

in close proximity to irrigated ricefields, whereas vegetablefarming is a prominent agricultural activity in the othervillage. In addition, irrigated rice cultivation was inter-

rupted during part of the study, providing a uniqueopportunity to examine its effect on malaria transmissionpatterns. Our entomological work presented here was com-

plemented by repeated cross-sectional malaria parasite ratesurveys and analysis of routinely collected health data,which will be published elsewhere.

Materials and methods

Study area

The study was carried out in the villages of Tiemelekro

(geographical coordinates: 6�500 N, �4�170 W) and Zatta(6�880 N, �5�390 W), located in central Cote d’Ivoire. Itformed part of a project that aimed at agricultural intensi-

fication through off-season vegetable farming in the formervillage, so that local farmers could improve their livelihoodsand socio-economic well-being. A detailed description ofTiemelekro, including climatic conditions, current health-

care delivery structures and key demographic and socio-economic indicators, has been presented recently (Girardinet al., 2004).

Zatta is located 7 km north-west of Yamoussoukro, thecapital city of Cote d’Ivoire. The mean annual temperaturein this village is 26.5�C and the mean annual precipitation is

1280mm. There is a long rainy season between April andJuly and a shorter one in October/November. A dispensary,run by two local nurses, is located in Zatta and also covers

nearby settlements. Two small dams were constructed inthis village in the mid-1970s. Since 1997, irrigated rice hasbeen cultivated on an estimated surface area of 36 ha, inclose proximity to human habitations. However, due to

unstable socio-political conditions and a farmers’ conflictover land, rice irrigation was interrupted in 2000 and againin 2003.

Cross-sectional household surveys

Preceding our entomological work, between December

2001 and February 2002, a cross-sectional survey was

28 B. G. Koudou et al.

# 2005 The Royal Entomological Society, Medical and Veterinary Entomology, 19, 27–37

carried out in 110 randomly selected households in eachvillage. Data collection included basic demographic indica-tors, access to clean water and improved sanitation and

common protective measures to prevent mosquito bites. Inaddition, geographical coordinates of all households andpotential mosquito breeding sites were mapped, using a

hand-held Magellan 320 global positioning system (GPS,Thales Navigation, Santa Clara, CA, U.S.A.).

Mosquito collections

A total of 10 entomological surveys were carried out

between February 2002 and August 2003. Adult mosquitoeswere collected with two different methods at four randomlyselected sentinel sites in each village (Fig. 1). Firstly, night

catches were done on human bait, employing aspirators andflashlights. The collectors were protected against malariaby appropriate chemoprophylaxis and they had been im-munized against yellow fever. The first four surveys, done

every other month beginning in February 2002, were con-ducted from 18.00 to 06.00 hours. A total of eight collectorsper village (one person stationed inside and one person

outside each of the sentinel sites) performed the surveys,each lasting for three consecutive nights. The samplingeffort per survey therefore was 24 man-nights per village.

Due to socio-political unrest in Cote d’Ivoire, commencingin September 2002, the next survey could only be done inFebruary 2003. Furthermore, the onset of mosquito collec-tion had to be delayed until 22.00 hours, and only indoor

night catches were allowed (one person per sentinel site).We adhered to an additional precautionary measure,namely mosquito collections were carried out only over

two consecutive nights per survey. As a result, samplingefforts for all six surveys done in 2003 were eight man-

nights per village, and the human night bite catches were4 h shorter than during the surveys in the preceding year.However, analysis of the 2002 surveys showed that only

very few mosquitoes were collected prior to 22.00 hours(< 5%). Consequently, we considered all nights sampled ineach of the entomological surveys as complete man-nights.

Overall, 96 and 48 human night bait catches were carriedout in 2002 and 2003, respectively.

The second method consisted of pyrethrum knock-down

spray sheet collections performed inside randomly selectedsleeping rooms at the sentinel sites (Service, 1976). Tensleeping rooms were sprayed in each of the four surveys

carried out in 2002, yielding a total sampling effort of 40sleeping rooms sprayed per village. In 2003, eight sleepingrooms were sprayed per survey; hence the overall samplingeffort was 48 sleeping rooms sprayed per village.

Mosquitoes were brought to the laboratory and identifiedto genus or species level. The physiological age of adultfemale Anopheles, and hence parity ratio (proportion of

female mosquitoes having laid eggs at least once), wasdetermined by dissection of ovaries and examination oftracheoles (Detinova, 1962). The identification of sporo-

zoites and oocysts obtained from dissection specimens wascarried out by enzyme-linked immunosorbent assay(ELISA), according to standard procedures (Beier et al.,1988). Bloodmeals from randomly selected female adult

Anopheles were selected and blotted onto filter paper foridentification of the source host. Antibodies were chosenwith respect to the abundance of potential hosts in the study

area; namely, human, bovine, chicken and porcine,previously marked with peroxidase and kept at 4�C. Theentomological inoculation rate (EIR), i.e. the number of

infective bites per person per unit of time, was estimatedby multiplication of the number of bites per person and themosquitoes’ infective rate.

Kilometers

0 1 2

Kilometers

0 1

Figure legend

N

A B

Household plots(Irrigated) rice growingVegetable farmingPotential mosquito breeding sitesSentinel mosquito collection sites

1.50.5

Fig. 1. Household locations in relation to (irrigated) rice growing and vegetable farming, potential mosquito breeding sites and sentinel sites

for adult mosquito collection in the villages of Tiemelekro (A) and Zatta (B), central Cote d’Ivoire.

Malaria transmission in central Cote d’Ivoire 29


Statistical analysis

All data were analysed with version 8.0 of the STATA

software package (STATA Corporation, College Station,TX, U.S.A.). We calculated 95% confidence intervals (CI)for sporozoite rates and parity ratios based on an exact

binomial distribution. For biting rates, 95% CI were calcu-lated on the basis of a Poisson distribution. w2-test andFisher’s exact test when appropriate, were used to compare

overall sporozoite rates between villages and between years.Finally, Poisson regression models were fitted to comparethe biting rates and EIRs between villages and years.The likelihood ratio test (LRT) was employed to assess

statistical significance.

Results

Village characteristics

According to the 1998 census, the estimated populationsizes of Tiemelekro and Zatta were 14 138 and 3315,

respectively. Tiemelekro therefore can be considered as asmall town. This is justified on significantly different live-lihoods; for example, a smaller proportion of people in

Tiemelekro are engaged in subsistence agriculture whencompared to Zatta. However, we employ the term ‘village’for both study locations throughout. In Tiemelekro there

are three small, non-irrigated ricefields (surface areas: 0.2,0.5 and 2.0 ha), as well as numerous temporary breedingsites located outside the village in the eastern and southernparts (Fig. 1A).

In Zatta, as already mentioned above, an estimated 36 haare utilized for irrigated rice growing. The ricefields arelocated as close as 250m from the nearest household. In

addition, several temporary breeding sites were discoveredwithin and outside the village (Fig. 1B).

Household characteristics

Living conditions and several of the investigated

household characteristics are comparable between thetwo study villages (Table 1). For example, similar pro-portions of houses utilized iron-corrugated sheets as

roofing material (93.8% in Tiemelekro vs. 92.9% inZatta), and had running water at home (74.1% vs.65.4%). On the other hand, improved sanitation facilities

were less prominent in Tiemelekro than in Zatta (17.0%vs. 47.6%). With regard to personal protective measuresagainst mosquito bites, the proportion of people sleepingunder a bednet was similarly low in both villages

(8.4–11.2%), whereas use of fumigating coils was muchmore pronounced in Zatta (47.3%) when compared toTiemelekro (9.1%).

Composition and abundance of mosquito fauna

Table 2 summarizes the number and percentage of the

mosquito fauna collected in the two study villages eitherby human night bait catches or pyrethrum knock-downspray sheet collections. In 2002, for a total of 96 man-

nights, 3308 and 7716 mosquitoes were collected onhuman bait inside and outside the sentinel sites in Tieme-lekro and Zatta, respectively. Another 225 and 703 mosqui-

toes, respectively, were caught by early morning spraying of40 randomly selected sleeping rooms per village. In thesubsequent year, for a total sampling effort of 48 man-nights in each village, 2056 mosquitoes were collected in

Tiemelekro and 859 in Zatta. Spray catches in 48 sleepingrooms per village revealed 622 and 78 mosquitoes, respec-tively.

Anopheles gambiae s.l. and An. funestus were the twomalaria vectors encountered in both villages. Anopheles

Table 1. Comparative descriptive statistics for the two study villages in early 2002, as assessed by a cross-sectional household survey

(all numbers are percentages except estimated population sizes).

Study village

Indicator Tiemelekro Zatta

Estimated population (according to 1998 census) 14 138 3315

Population at working age engaged in agriculture 42.8 74.4

House construction materials (walls and roofs)

Cement bricks and iron-corrugated sheets 35.5 82.6

Roughcast mud and iron-corrugated sheets 58.3 10.3

Roughcast mud and thatched grass 6.2 7.1

Educational attainment

Total illiteracy rate among people� 15 years 50.3 57.3

Secondary school 11.2 8.0

Running water at home 74.1 65.4

Improved sanitation facilities at home 17.0 47.6

Protective measures against mosquito bites

Modern drugs 2.0 1.3

Bednets 11.2 8.4

Fumigating coils 9.1 47.3

Insecticide sprays 5.9 9.5



gambiae s.l. was the predominant anopheline speciesthroughout; it represented 53.0–98.9% in each of the

two study villages in 2002 or 2003, depending on thecollection method. The percentage of mosquitoes belongingto Mansonia spp. was between 14.1% and 39.1% of all

mosquitoes caught either by human bait or spray catch in2002 or 2003. Only few Culex spp. and Aedes spp. were col-lected in Tiemelekro, whereas the number ofCulex spp. caughtin Zatta ranged between 3.8% and 11.2% in 2002–2003.

Human night biting pattern and biting rate

In Zatta, the human biting rate of An. gambiae s.l. duringthe night showed two peaks; the first one occurred at23.00–24.00 hours and the second one at 04.00–05.00 hours.

High biting rates were recorded in all four surveys carried outin 2002. There was considerable variation ranging from 26.2bites per person per night (b/p/n) in August to 67.7 b/p/n in

April (Table 3). The estimated overall biting rate for this yearwas 49.3 b/p/n. In 2003, a several-fold lower biting rate wasobserved, namely 7.9 b/p/n, which was a highly significant

reduction (LRT¼ 1072.66;P< 0.001). Higher proportions ofAn. gambiae s.l. were collected inside houses rather than out-side, hence this species tended to be endophagic in this setting.As already shown in Table 2, noAn. funestuswere collected in

Zatta in 2002, but in the following year, a low overall bitingrate of 1.2 b/p/n was recorded.In Tiemelekro, the night biting activity of An. gambiae s.l.

showed only one peak; it occurred at 01.00–02.00 hours.

The overall biting rate of An. gambiae s.l. in 2002 was 16.1b/p/n. It was slightly higher in 2003, namely 18.2 b/p/n. This

difference was statistically significant (LRT¼ 148.06;P< 0.001). The proportions of An. gambiae s.l. collectedduring the 2002 surveys indicated that this species

was equally endophagic than exophagic. Considerablylower overall biting rates were observed for An. funestus;2.8 b/p/n and 5.4 b/p/n for 2002 and 2003, respectively.The difference between years was statistically significant

(LRT¼ 143.02; P< 0.001).

Sporozoite rate of malaria vectors

Table 4 shows the numbers of mosquitoes tested and

infected; hence the estimated sporozoite rates for An. gam-biae s.l. and An. funestus separately for each survey. Theoverall annual values for each village are also given. In

Tiemelekro, a significantly higher overall sporozoite ratewas observed in 2003 when compared to 2002 (7.6% vs.3.1%; w2¼ 13.13; P< 0.001). A significant difference in the

overall sporozoite rate was also observed in Zatta betweenyears, with a four-fold lower rate observed in 2003 (4.6% vs.1.2%; w2¼ 5.92; P¼ 0.015). In the latter year, sporozoiteswere only found during the rainy season (surveys conducted

in June and July 2003). Comparison of the overall sporo-zoite rates of An. gambiae s.l. between villages revealed nosignificant difference in 2002 (w2¼ 3.03; P¼ 0.082), but a

highly significant difference in 2003 (w2¼ 12.46; P< 0.001).

Table 2. Number (%) of mosquito fauna collected by human night bait catches and pyrethrum knock-down spray sheet collections (spray

catch) in the two study villages in the years 2002 and 2003.

Tiemelekro Zatta

Mosquito fauna

Human night

bait catch Spray catch

Human night

bait catch Spray catch

Year 2002*

Anopheles gambiae s.l. 1545 (46.7) 96 (42.7) 4733 (61.3) 521 (74.1)

Anopheles funestus 270 (8.2) 85 (37.8) 0 (0.0) 0 (0.0)

Anopheles pharoensis 151 (4.6) 0 (0) 395 (5.1) 6 (0.9)

Mansonia africana 937 (28.3) 29 (12.9) 1260 (16.3) 91 (12.9)

Mansonia uniformis 352 (10.6) 15 (6.7) 470 (6.1) 32 (4.6)

Culex spp. 40 (1.2) 0 (0.0) 819 (10.6) 53 (7.5)

Aedes spp. 13 (0.4) 0 (0.0) 36 (0.5) 0 (0.0)

Other species 0 (0.0) 0 (0.0) 3 (0.04) 0 (0.0)

Total 3308 (100) 225 (100) 7716 (100) 703 (100)

Year 2003yAnopheles gambiae s.l. 875 (42.6) 478 (76.8) 379 (44.1) 63 (80.8)

Anopheles funestus 261 (12.7) 14 (2.3) 58 (6.7) 1 (1.3)

Anopheles pharoensis 28 (1.4) 0 (0.0) 47 (5.5) 0 (0.0)

Mansonia africana 632 (30.7) 83 (13.3) 191 (22.2) 8 (10.3)

Mansonia uniformis 156 (7.6) 34 (5.5) 83 (9.7) 3 (3.8)

Culex spp. 88 (4.3) 13 (2.1) 96 (11.2) 3 (3.8)

Aedes spp. 16 (0.8) 0 (0.0) 5 (0.6) 0 (0.0)

Other species 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)

Total 2056 (100) 622 (100) 859 (100) 78 (100)

*Total sampling effort per village: 96 human night bite catches and 40 randomly selected sleeping rooms sprayed.

yTotal sampling effort per village: 48 human night bite catches and 48 randomly selected sleeping rooms sprayed.



The overall sporozoite rate for An. funestus in Tiemelekroshowed an increase from 4.1% in 2002 to 6.6% in 2003,

which, however, was not statistically significant (Fisher’sexact test; P¼ 0.447).

Parity ratio of malaria vectors

Table 5 shows the numbers of Anopheles tested and thosehaving laid eggs at least once, hence the estimated parity

ratios during the different surveys. In Tiemelekro, the over-all parity ratios of An. gambiae s.l. in 2002 and 2003 were66.1% and 58.6%, respectively, which was a statistically

significant difference (w2¼ 7.04; P¼ 0.008). Lower parityratios were measured for this malaria vector in Zatta;

34.5% and 44.6% in 2002 and 2003, respectively. The dif-ference between years was also statistically significant(w2¼ 7.97; P¼ 0.005). Finally, estimated parity ratios ofAn. gambiae s.l. were significantly higher in Tiemelekro

when compared to Zatta both in 2002 (w2¼ 212.16;P< 0.001) and in 2003 (w2¼ 11.38; P< 0.001).In Tiemelekro, the overall parity ratio of An. funestus was

higher in 2003 (72.5%) when compared to the precedingyear (63.8%), but there was no statistically significant dif-ference between years (w2¼ 2.37; P¼ 0.124).

Table 3. Number of mosquitoes caught by human night bait catches and variation of biting rate (i.e. bites per person per night) of Anopheles

gambiae s.l. and An. funestus in the two study villages between February 2002 and August 2003 (values in brackets are 95% confidence

interval).

Tiemelekro Zatta

An. gambiae s.l. An. funestus An. gambiae s.l. An. funestus

Survey date No. caught Biting rate No. caught Biting rate No. caught Biting rate No. caught Biting rate

2002

February 29 1.2 (0.8–1.7) 69 2.9 (2.2–3.6) 1226 51.1 (48.3–54.0) 0 0.0 (0.0–0.2)*

April 355 14.8 (13.3–16.4) 12 0.5 (0.3–0.9) 1625 67.7 (64.5–71.1) 0 0.0 (0.0–0.2)*

June 250 10.4 (9.2–11.8) 19 0.8 (0.5–1.2) 1253 52.2 (49.4–55.2) 0 0.0 (0.0–0.2)*

August 911 38.0 (35.5–40.5) 170 7.1 (6.1–8.2) 629 26.2 (24.2–28.3) 0 0.0 (0.0–0.2)*

Overall in 2002 1545 16.1 (15.3–16.9) 270 2.8 (2.5–3.2) 4733 49.3 (47.9–50.7) 0 0.0 (0.0–0.04)*

2003

February 69 8.6 (6.7–10.9) 108 13.5 (11.1–16.3) 54 6.8 (5.1–8.8) 6 0.8 (0.3–1.6)

April 101 12.6 (10.3–15.3) 101 12.6 (10.3–15.3) 66 8.3 (6.4–10.5) 0 0.0 (0.0–0.5)*

May 58 7.3 (5.5–9.4) 12 1.5 (0.8–2.6) 48 6.0 (4.4–8.0) 16 2.0 (1.1–3.2)

June 355 44.4 (39.9–49.2) 4 0.5 (0.1–1.3) 84 10.5 (8.4–13.0) 10 1.3 (0.6–2.3)

July 277 34.6 (30.7–39.0) 10 1.3 (0.6–2.3) 67 8.4 (6.5–10.6) 10 1.3 (0.6–2.3)

August 15 1.9 (1.0–3.1) 26 3.3 (2.1–4.8) 60 7.5 (5.7–9.7) 16 2.0 (1.1–3.2)

Overall in 2003 875 18.2 (17.0–19.5) 261 5.4 (4.8–6.1) 379 7.9 (7.1–8.7) 58 1.2 (0.9–1.6)

*One-sided 97.5% CI.

Table 4. Number of mosquitoes tested, number of infected mosquitoes, and variation of sporozoite rates (in percentage, including 95%confidence interval) of Anopheles gambiae s.l. and An. funestus in the two study villages between February 2002 and August 2003.

Tiemelekro Zatta


Survey date

No. tested

(infected)

Sporozite

rate

No. tested

(infected)

Sporozite

rate

No. tested

(infected)

Sporozite

rate

No. tested

(infected)

Sporozite

rate

2002

February 29 (0) 0.0 (0.0–11.9) 30 (0) 0.0 (0.0–11.6) 620 (42) 6.8 (4.9–9.0) 0 (0) 0

April 281 (6) 2.1 (0.8–4.6) 12 (1) 8.3 (0.2–38.5) 424 (10) 2.4 (1.1–4.3) 0 (0) 0

June 250 (8) 3.2 (1.4–6.2) 14 (1) 7.1 (0.2–33.9) 554 (32) 5.8 (4.0–8.1) 0 (0) 0

August 239 (11) 4.6 (2.3–8.1) 67 (3) 4.5 (0.9–12.5) 500 (12) 2.4 (1.2–4.2) 0 (0) 0

Overall in 2002 799 (25) 3.1 (2.0–4.6) 123 (5) 4.1 (1.3–9.2) 2098 (96) 4.6 (3.7–5.6) 0 (0) 0

2003

February 20 (0) 0.0 (0.0–16.8) 29 (1) 3.4 (0.1–17.8) 35 (0) 0.0 (0.0–10.0) 3 (0) 0.0 (0.0–70.8)

April 58 (8) 13.8 (6.1–25.4) 50 (4) 8.0 (2.2–19.2) 36 (0) 0.0 (0.0–9.7) 0 (0) 0

May 52 (10) 19.2 (9.6–32.5) 8 (2) 25.0 (3.2–65.1) 35 (0) 0.0 (0.0–10.0) 13 (0) 0.0 (0.0–24.7)

June 151 (7) 4.6 (1.9–9.3) 2 (0) 0.0 (0.0–84.2) 69 (1) 1.4 (0.0–7.8) 8 (0) 0.0 (0.0–36.9)

July 206 (12) 5.8 (3.0–10.0) 6 (0) 0.0 (0.0–45.9) 36 (2) 5.6 (0.7–18.7) 8 (1) 12.5 (0.3–52.6)

August 15 (1) 6.7 (0.2–31.9) 26 (1) 3.8 (0.1–19.6) 30 (0) 0.0 (0.0–11.6) 15 (0) 0.0 (0.0–21.8)

Overall in 2003 503 (38) 7.6 (5.4–10.2) 122 (8) 6.6 (2.9–12.5) 241 (3) 1.2 (0.3–3.6) 47 (1) 2.1 (0.1–11.3)



Entomological inoculation rate

Malaria transmission in both villages was primarily dueto An. gambiae s.l. and, to a lesser extent, due to An. funes-tus, particularly in Tiemelekro. Table 6 shows the estimated

EIRs for each of the two malaria vectors during the 10entomological surveys carried out in the two villages. InZatta, the estimated annual EIR of An. gambiae s.l. in

2002 was very high, namely 789 infective bites. A 21-foldlower EIR was estimated in the subsequent year (38 infec-tive bites), which was a highly significant reduction(LRT¼ 838.14; P< 0.001). Estimated annual EIRs of

An. gambiae s.l. in Tiemelekro were high in both years; 233infective bites in 2002 and 342 infective bites in 2003. Thedifference between years was statistically significant

(LRT¼ 20.79; P< 0.001). Comparison of annual EIRs

between villages revealed highly significant differences bothin 2002 (LRT¼ 319.41; P< 0.001) and 2003 (LRT¼ 279.73;

P< 0.001). In Tiemelekro, the highest EIRs were observedin June–August. In Zatta, high EIRs were estimated ineach of the four surveys conducted in 2002, but in the

following year, only those two surveys conducted duringthe main rainy season (June and July) revealed positiveEIRs. The remaining four surveys, as well as the February

2002 and February 2003 surveys done in Tiemelekro,revealed EIRs of zero. However, care is needed in theinterpretation of these data, as the number of mosquitoesexamined was small.

With regard to An. funestus in Tiemelekro, low EIRs wereestimated for three of the four surveys carried out in 2002.Higher EIRs were measured in 2003, with the highest one

observed in April. In Zatta, the only survey that revealed a

Table 5. Number of mosquitoes tested, number of mosquitoes that laid eggs at least once, and variation of parity ratios (in percentage,

including 95% confidence interval) of Anopheles gambiae s.l. and An. funestus in the two study villages between February 2002 and August

2003.

Tiemelekro Zatta


Survey date

No. tested

(laid eggs) Parity ratio

No. tested


No. tested


No. tested


2002

February 26 (16) 61.5 (40.6–79.8) 68 (51) 75.0 (63.0–84.7) 386 (130) 33.7 (29.0–38.6) 0 (0) 0

April 256 (127) 49.6 (43.2–55.9) 11 (3) 27.3 (6.0–61.0) 483 (85) 17.6 (14.3–21.3) 0 (0) 0

June 241 (186) 77.2 (71.4–82.3) 11 (8) 72.7 (39.0–94.0) 674 (275) 40.8 (37.1–44.6) 0 (0) 0

August 214 (158) 73.8 (67.4–79.6) 117 (70) 59.8 (50.4–68.8) 287 (142) 49.5 (43.6–55.4) 0 (0) 0

Overall in 2002 737 (487) 66.1 (62.5–69.5) 207 (132) 63.8 (56.8–70.3) 1830 (632) 34.5 (32.4–36.8) 0 (0) 0

2003

February 25 (14) 56.0 (34.9–75.6) 34 (27) 79.4 (62.1–91.3) 27 (23) 85.2 (66.3–95.8) 0 (0) 0

April 61 (27) 44.3 (31.5–57.6) 26 (19) 73.1 (52.2–88.4) 27 (6) 22.2 (8.6–42.3) 0 (0) 0

May 41 (36) 87.8 (73.8–95.9) 8 (8) 100.0 (53.1–100.0) 33 (17) 51.5 (33.5–69.2) 3 (2) 66.7 (9.4–99.2)

June 183 (104) 56.8 (49.3–64.1) 2 (0) 0.0 (0.0–84.2) 32 (13) 40.6 (23.7–59.4) 1 (1) 100.0 (2.5–100.0)

July 164 (96) 58.5 (50.6–66.2) 6 (2) 33.3 (4.3–77.7) 47 (15) 31.9 (19.1–47.1) 8 (4) 50.0 (15.7–84.3)

August 14 (9) 64.3 (35.1–87.2) 26 (18) 69.2 (48.2–85.7) 36 (16) 44.4 (27.9–61.9) 9 (0) 0 (0.0–33.6)

Overall in 2003 488 (286) 58.6 (54.1–63.0) 102 (74) 72.5 (62.8–80.5) 202 (90) 44.6 (37.6–51.7) 21 (7) 33.3 (14.6–57.0)

Table 6. Variation of nightly entomological inoculation rate (EIR) of Anopheles gambiae s.l. and An. funestus in the two study villages

estimated at 10 entomological surveys carried out between February 2002 and August 2003, and annual EIRs.

Survey date

Tiemelekro

An. gambiae s.l. An. funestus

Zatta

An. gambiae s.l. An. funestus

2002

February 0 0 3.46 0

April 0.32 0.04 1.60 0

June 0.33 0.06 3.02 0

August 1.75 0.32 0.63 0

Annual EIR for 2002 233 16 789 0

2003

February 0 0.47 0 0

April 1.74 1.01 0 0

May 1.39 0.38 0 0

June 2.06 0 0.15 0

July 2.02 0 0.47 0.16

August 0.13 0.13 0 0

Annual EIR for 2003 342 119 38 10



positive EIR for An. funestus was the one conducted inJuly 2003.

Anthropophily rate of malaria vectors

In Zatta, a random sample of 377 bloodmeals derivedfrom adult female An. gambiae s.l. was investigated. The

anthropophily rate, as measured by the proportion ofbloodmeals of human origin, was 98.4%. A similarly highanthropophily rate was found in Tiemelekro (95.6%),

following the examination of 389 random bloodmealsfrom the same vector. The remaining bloodmeals wereeither of bovine or porcine origins (Tiemelekro), or from

bovines (Zatta). Only three bloodmeals were tested amongadult female An. funestus specimens collected in Zatta; twoof them were of human origin. In Tiemelekro, a total of 19

bloodmeals among adult female An. funestus wereexamined; 17 of which were of human and two of porcinorigin. These high anthropophily rates observed in bothvillages emphasize that An. gambiae s.l. and An. funestus

predominantly feed on humans.

Discussion

The study presented here was designed to comparativelyassess malaria transmission dynamics in two differentagroecosystems, namely irrigated rice growing and vege-

table farming, in the wet savannah zone of central Coted’Ivoire. Due to socio-political unrest in the country and afarmers’ conflict over land, irrigated rice cultivation was

interrupted in the second half of the investigation. Theseevents provided a unique opportunity to not only comparemalaria transmission patterns between agroecosystems, but

also to further our understanding of the effect of changingpatterns of irrigation in a ricefield proximal to humanhabitations. Our data confirm that An. gambiae s.l. andAn. funestus are the principal malaria vectors in central

Cote d’Ivoire (Dossou-Yovo et al., 1995; Doannio et al.,1999, 2002; N’Guessan et al., 2001). Further north in thedrier savannah zone of Burkina Faso, malaria transmission

is also mainly due to An. gambiae s.l. and An. funestus(Robert et al., 1985), whereas in the irrigated Sahel ofMali,An. gambiae s.l. is often the only malaria vector species

encountered (Klinkenberg et al., 2002; Dolo et al., 2004).Next, our results confirm that the density of An. gambiaes.l. was several-fold higher in irrigated rice agroecosystems

when compared to traditional crop cultivation, includingvegetable growing (Dolo et al., 2004). Indeed, the numberof An. gambiae s.l. caught by far outnumbered An. funestusin the irrigated rice agroecosystem (Robert et al., 1985).

Finally, our data confirm that the anthropophilic mosquitofauna in the two study villages largely consisted ofAn. gambiae s.l., followed byMansonia spp. Similar findings

had been reported three decades ago from rural parts ofKenya, where irrigated rice agriculture was performed.Hence it was concluded there that members of the

An. gambiae complex andMansonia spp. were the predomin-

ant ricefield-breeding mosquitoes (Surtees, 1970; Chandler& Highton, 1975). More recent studies carried out inBurkina Faso and Cote d’Ivoire came to the same conclu-

sions (Robert et al., 1988; Doannio et al., 2002).We found highest biting rates of An. gambiae s.l. towards

the end of the main rainy season, which is typical for the

rural savannah of West Africa (Faye et al., 1993). However,in Zatta, high biting rates of An. gambiae s.l. were observedthroughout 2002. A likely explanation of this observation is

that mosquito breeding took place to a large extend in theirrigated ricefield; hence was independent to that year’srainfall pattern. As can be seen from the map of this village(Fig. 1B), the irrigation scheme is as close as 250m from the

nearest households. Support for this explanation is derivedfrom two sources. First, Anopheles larvae were found in thetwo breeding sites connected to the ricefield. Second, inter-

ruption of rice irrigation in 2003 coincided with lower den-sities of adult Anopheles and, concurrently, the biting ratesdecreased substantially. Similar findings have been reported

before; consistently higher biting rates were observedamong villagers performing smallholder rice irrigationschemes than among those living in different agro-ecosys-tems, except during the rainy season (Marrama et al., 1995;

Ijumba et al., 2002a; Dolo et al., 2004). For example,in Senegal, the biting rate in a village near a ricefield was17-fold higher than that observed in a village located more

than 5 km away from a ricefield (Faye et al., 1993). Signifi-cantly higher biting rates and an increase in malaria trans-mission has recently been documented in an irrigated

subarid ecosystem of Madagascar (Marrama et al., 2004).Peak biting rates of An. funestus were observed in the dry

spell between the long and the short rainy season (August).

These findings are consistent with previous entomologicalinvestigations in the humid African savannah; the highfrequencies of An. funestus were usually observed betweenAugust and October (Robert et al., 1985; Dossou-Yovo

et al., 1995). Interestingly, in Tiemelekro, we also observedhigh biting rates of An. funestus during the dry season of2003 (surveys conducted in February and April).

Our results also revealed that in Zatta, particularly in2002, An. gambiae s.l. showed a significant tendency toendophagy, which is in agreement with previous findings

from a traditional wet savannah village of central Coted’Ivoire (Dossou-Yovo et al., 1995). In Tiemelekro, on theother hand, An gambiae s.l. and An. funestus were equally

endophageous than exophageous. The underlying reasonsare not well understood, but previous research has shownthat it depends on ecological factors and, in the case ofmembers of the An. gambiae complex, on the species itself

(Robert et al., 1985).With regard to the infective rate of the malaria vectors,

our data from Zatta in 2002 show that the highest rates

occurred concurrently with a high biting rate. A differentpattern was observed in Tiemelekro, namely high infectiverates coincided with periods of low biting rates. Import-

antly, our results showed that malaria vectors’ infectiverates in rural areas are significantly elevated when asso-ciated with irrigated rice agriculture. These observations



are in contrast to those made by others, reporting lowerinfective rates in irrigated rice cultivations when comparedto non-irrigated rural areas (Dossou-Yovo, 1999). When

rice irrigation was interrupted in Zatta in 2003, malariatransmission was restricted to the second half of the mainrainy season. These findings are in agreement with previous

observations made elsewhere in the West African savannah(Robert et al., 1985).In Zatta, the estimated EIR was very high during the

period of rice cultivation, but it decreased significantlytowards the end of the rice growing cycle. Clearly, theproximity of the irrigated ricefields to the households is a

key risk factor. Studies carried out in other malaria endemicsettings also reported enhanced transmission in relation tointensification of rice cultivation (Gratz, 1988). For exam-ple, a study in Burundi found that the vectorial capacity of

An. arabiensis Patton is 150 times higher in rice cultivatingareas when compared to cotton plantations (Coosemans,1985). Recent surveys carried out in northern Cote d’Ivoire

indicate that intensification of shallow rice cultivations wasaccompanied by higher malaria incidence. In addition, itweakened the households’ capacities to invest in protective

measures against mosquito bites, and curtailed women’scapacity to promptly and effectively manage disease epi-sodes (De Plaen et al., 2003). However, no impact of irri-gated rice cultivation on malaria was found in another

epidemiological setting in Cote d’Ivoire (Henry et al.,2003), as well as most other settings in sub-Saharan Africa,where malaria transmission is stable (Ijumba & Lindsay,

2001; Keiser et al., 2004).Previous research in irrigated rice agroecosystems of

neighbouring Burkina Faso and Mali found low parity

ratios of An. gambiae s.l. accompanied by lowanthropophily of these mosquitoes (Robert et al., 1985;Dolo et al., 2004). Here, we confirm low parity ratios of

these malaria vectors in an area of irrigated rice cultivationof central Cote d’Ivoire, but we found a very high anthro-pophily exceeding 95%. The subject of our current researchis to identify the different sibling species of the An. gambiae

complex. This in turn will enhance our understanding of thedifferent levels of anthropophily and zoophily observed in theregion, and will facilitate comparison with previous studies.

We summarize that several factors are likely to explainthe high malaria transmission rates observed in Zatta in theyear 2002. Prominent is the close proximity of the irrigated

ricefield to the households. Previous studies carried out inurban settings of Senegal and Uganda demonstrated thatclinical malaria cases were strongly associated with close

proximity to breeding places, acting as the main malariatransmission sites (Trape et al., 1992; Staedke et al., 2003).Other contributing factors include the current cultivationmethods, characterized by overlaps between several agricul-

tural cycles; the malaria vectors’ high anthropophily; andthe low percentage of people currently sleeping under ITNs.These factors are coupled with precarious living conditions.

There is, for example, no mosquito screening of windows ordoors, and eaves are left open. Implementation of suchsimple changes in the house design can significantly reduce

exposure to mosquito bites (Lindsay et al., 2003). Interrup-tion of irrigated rice cultivation brought about a significantreduction in all transmission parameters investigated here,

and also resulted in lower malaria prevalence rates (thesedata will be presented elsewhere). By contrast, the transmis-sion dynamics of malaria in Tiemelekro, where vegetable

farming is performed and traditional subsistence crops aregrown, showed only small variation from one year to another.

In view of these findings we advance three recommenda-

tions and briefly discuss their feasibility. First, the locationof the irrigated ricefield in Zatta might be relocated at least1–2 km away from the village boundaries (Carter et al.,

2000). Obviously, such a radical step would require accep-tance among rice farmers and the larger village community,and is compounded by increasing land pressure. Further-more, 1–2 km might not suffice, as wider flight ranges have

been reported for Anopheles (Briet et al., 2003; Killeen et al.,2003). Second, in the case of continued irrigated rice agri-culture in close proximity to households, alternative irriga-

tion practices should be considered, such as different waterregimens with regard to water depth and frequency of irri-gation application. In Asia, for example, intermittent irriga-

tion, when properly designed and implemented on suitablesoil types, brought about significant reduction in mosquitoproductivity. Moreover, this strategy was often coupledwith water savings, increased rice yields and reduced

methane emissions (Keiser et al., 2002). However, only fewexperimental field trials have so far been conducted inAfrica (for a recent example see Mutero et al., 2000), thus

more research is required to rigorously evaluate the suit-ability of this proposed strategy in different African settingswhere malaria is endemic and irrigated rice growing is

performed. Finally, personal protective measures (i.e. sleep-ing under ITNs) should be vigorously promoted, as thisstrategy has a proven track record of reducing malaria-

related morbidity and mortality (Lengeler, 2004). Socialmarketing approaches offer an innovative and cost-effectiveavenue for rapid scaling-up of ITN coverage rates(Armstrong Schellenberg et al., 2001; Hanson et al., 2003).

The evidence-base is compelling that ITNs significantlyreduce child mortality and morbidity, as well as transmissionin any given setting.

Acknowledgements

Our thanks are addressed to the village authorities and

villagers of Tiemelekro and Zatta and the district healthofficers of Dimbokro and Yamoussoukro for their commit-ment in the present study. A series of excellent comments

made by an anonymous referee are duly acknowledged.This investigation received financial support from the‘Fonds Ivoiro Suisse de Developpement Economique etSociale’ (FISDES), the Swiss Academy of Natural Sciences

(SANW) and the Swiss Federal Commission for Fellow-ships for Foreign Students (CFBEE). It was partiallysupported by the National Centre of Competence in

Research (NCCR) North–South, ‘Research Partnerships



for Mitigating Syndromes of Global Change’, IndividualProject #4 (IP4) entitled ‘Health and Well-being’. B.G.K. isgrateful to the Universite d’Abidjan-Cocody and the Swiss

Tropical Institute, and J.K. and J.U. to the Swiss NationalScience Foundation (Project Nos PMPDB�106221 andPPOOB�102883, respectively).

References

Aka, M., Pagano, M., Saint-Jean, L. et al. (2000) Zooplankton

variability in 49 shallow tropical reservoirs of Ivory Coast (West

Africa). International Review of Hydrobiology, 85, 491–504.

Amerasinghe, F.P. (2003) Irrigation and mosquito-borne diseases.

Journal of Parasitology, 89 (Suppl.), S14–S22.

Armstrong Schellenberg, J.R.M., Abdulla, S., Nathan, R. et al.

(2001) Effect of large-scale social marketing of insecticide-

treated nets on child survival in rural Tanzania. Lancet, 357,

1241–1247.

Beier, M.S., Schwartz, I.K., Beier, J.C. et al. (1988) Identification

of malaria species by ELISA in sporozoite and oocyst infected

Anopheles from western Kenya. American Journal of Tropical

Medicine and Hygiene, 39, 323–327.

Briet, O.J.T., Dossou-Yovo, J., Akodo, E., van de Giesen, N. &

Teuscher, T.M. (2003) The relationship between Anopheles

gambiae density and rice cultivation in the savannah zone and

forest zone of Cote d’Ivoire. Tropical Medicine and International

Health, 8, 439–448.

Carter, R., Mendis, K.N. & Roberts, D. (2000) Spatial targeting of

interventions against malaria. Bulletin of the World Health

Organization, 78, 1401–1411.

Chandler, J.A. & Highton, R.B. (1975) The succession of mosquito

species (Diptera, Culicidae) in rice fields in the Kisumu area of

Kenya, and their possible control. Bulletin of Entomological

Research, 65, 295–302.

Coosemans, M.H. (1985) Comparaison de l’endemie malarienne

dans une zone de riziculture et dans une zone de culture de coton

dans la Plaine de la Rusizi, Burundi. Annales de la Societe belge

de Medicine Tropicale, 65 (Suppl. 2), 187–200.

De Plaen, R., Geneau, R., Teuscher, T., Koutoua, A. & 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. Tropical Medicine and International Health,

8, 459–470.

Detinova, T.S. (1962) Age-Grouping Methods in Diptera of Medical

Importance, with Special Reference to Some Vectors of Malaria.

Monograph no. 47, World Health Organization, Geneva,

Switzerland.

Doannio, J.M.C., Dossou-Yovo, J., Diarrassouba, S. et al. (1999)

Efficacite des moustiquaires pre-impregnees de permethrine

Olyset Net1 en zone de resistance des vecteurs aux pyrethri-

noides. I – Evaluation entomologique. Medecine Tropicale, 59,

349–354.

Doannio, J.M.C., Dossou-Yovo, J., Diarrassouba, S. et al. (2002)

La dynamique de la transmission du paludism a Kafine, un

village rizicole en zone de savane humide de Cote d’Ivoire.

Bulletin de la Societe de Pathologie Exotique, 95, 11–16.

Dolo, G., Briet, O.J.T., Dao, A. et al. (2004) Malaria transmission

in relation to rice cultivation in the irrigated Sahel of Mali. Acta

Tropica, 89, 147–159.

Dossou-Yovo, J. (1999) Etude ethiologique des moustiques

vecteurs du paludisme en rapport avec les aspects parasitologi-

ques de la transmission du plasmodium dans la Region de

Bouake. These de Doctorat d’Etat en Entomologie Medicale,

Universite de Cocody, Abidjan.

Dossou-Yovo, J., Doannio, J.M.C., Riviere, F. & Chauvancy, G.

(1995) Malaria in Cote d’Ivoire wet savannah region: the

entomological input. Tropical Medicine and Parasitology, 46,

263–269.

Faye, O., Diallo, S., Gaye, O., N’Dri, O. & Faye, O. (1992)

Efficacite comparee de l’utilisation des pieges lumineux du type

CDC et des sujets humains pour l’echantillonnage des popula-

tions anopheliennes. Bulletin de la Societe de Pathologie

Exotique, 85, 185–189.

Faye, O., Fontenille, D., Herve, J.P., Diack, P.A., Diallo, S. &

Mouchet, J. (1993) Le paludisme en zone sahelienne du Senegal.

1. Donnees entomologiques sur la transmission. Annales de la

Societe belge de Medicine Tropicale, 73, 21–30.

Girardin, O., Dao, D., Koudou, B.G. et al. (2004) Opportunities

and limiting factors of intensive vegetable farming in malaria

endemic Cote d’Ivoire. Acta Tropica, 89, 109–123.

Gratz, N.G. (1988) The impact of rice production on vector-

borne disease problems in developing countries. Vector-Borne

Disease Control in Humans through Rice Agroecosystem

Management. Proceedings of the Workshop on Research and

Training Needs in the Field of Integrated Vector-Borne

Disease Control in Riceland Agroecosystems of Developing

Countries, 9–14 March 1987. International Rice Research

Institute, WHO/FAO/UNEP Panel of Experts on Environ-

mental Management for Vector Control, Manila, The

Philippines.

Hanson, K., Kikumbih, N., Armstrong Schellenberg, J. et al.

(2003) Cost-effectiveness of social marketing of insecticide-

treatednets formalaria control in theUnitedRepublic ofTanzania.

Bulletin of the World Health Organization, 81, 269–276.

Henry, M.-C., Rogier, C., Nzeyimana, I. et al. (2003) Inland valley

rice production systems and malaria infection and disease in the

savannah of Cote d’Ivoire. Tropical Medicine and International

Health, 8, 449–458.

van der Hoek, W., Konradsen, F., Amerasinghe, P.H., Perera, D.,

Piyaratne, M.K. & Amerasinghe, F.P. (2003) Towards a risk

map of malaria for Sri Lanka: the importance of house location

relative to vector breeding sites. International Journal of

Epidemiology, 32, 280–285.

Hulme, M. (2001) Climatic perspectives on Sahelian desiccation:

1973–98. Global Environmental Change, 11, 19–29.

Ijumba, J.N. & Lindsay, S.W. (2001) Impact of irrigation on

malaria in Africa: paddies paradox. Medical and Veterinary

Entomology, 15, 1–11.

Ijumba, J.N., Mosha, F.W. & Lindsay, S.W. (2002a) Malaria

transmission risk variations derived from different agricultural

practices in an irrigated area of northern Tanzania. Medical and

Veterinary Entomology, 16, 28–38.

Ijumba, J.N., Shenton, F.C., Clarke, S.E., Mosha, F.W. &

Lindsay, S.W. (2002b) Irrigated crop production is associated

with less malaria than traditional agricultural practices in

Tanzania. Transactions of the Royal Society of Tropical Medicine

and Hygiene, 96, 476–480.

Keiser, J., Utzinger, J. & Singer, B.H. (2002) The potential of

intermittent irrigation for increasing rice yields, lowering water

consumption, reducing methane emissions, and controlling

malaria in African rice fields. Journal of the American Mosquito

Control Association, 18, 329–340.

Keiser, J., Caldas de Castro, M.,Maltese,M.F., Bos, R., Tanner,M.,

Singer, B.H. & Utzinger, J. (2004) The effect of irrigation and



large dams on the burden of malaria on global and regional

scale. American Journal of Tropical Medicine and Hygiene, in

press.

Killeen, G.F., Knols, B.G.J. & Gu, W. (2003) Taking malaria

transmission out of the bottle: implications of mosquito

dispersal for vector-control interventions. Lancet Infectious

Diseases, 3, 297–303.

Klinkenberg, E., Huibers, F., Takken, W. & Toure, Y.T. (2002)

Water management as a tool for malaria mosquito control?

Irrigation and Drainage Systems, 16, 201–212.

Lengeler, C. (2004) Insecticide-treated bed nets and curtains for

preventing malaria. Cochrane Database of Systematic Reviews,

CD000363.

Lindsay, S.W., Wilkins, H.A., Zieler, H.A., Daly, R.J., Petrarca, V.

& Byass, P. (1991) Ability of Anophes gambiae mosquitoes to

transmit malaria during the dry and wet seasons in an area of

irrigated rice cultivation in The Gambia. Journal of Tropical


Lindsay, S.W., Jawara, M., Paine, K., Pinder, M., Walraven, G.E.

& Emerson, P.M. (2003) Changes in house design reduce

exposure to malaria mosquitoes. Tropical Medicine and Interna-

tional Health, 8, 512–517.

Marrama, L., Jambou, R., Rakotoarivony, I. et al. (2004) Malaria

transmission in southern Madagascar: influence of the environ-

ment and hydro-agricultural works in sub-arid and humid

regions. Part 1. Entomological investigations. Acta tropica, 89,

193–203.

Marrama, L., Rajaonarivelo, E., Laventure, S. & Rabarison, P.

(1995) Anopheles funestus et la riziculture sur les plateaux de

Madagascar. Sante, 5, 415–419.

Mutero, C.M., Blank, H., Konradsen, F. & van der Hoek, W.

(2000) Water management for controlling the breeding of

Anopheles mosquitoes in rice irrigation schemes in Kenya. Acta

Tropica, 76, 253–263.

N’Guessan, R., Darriet, F., Doannio, J.M.C., Chandre, F. &

Carnevale, P. (2001) Olyset Net1 efficacy against pyrethroid-

resistant Anopheles gambiae and Culex quinquefasciatus after 3

years’ field use in Cote d’Ivoire. Medical and Veterinary

Entomology, 15, 97–104.

Robert, V., Gazin, P., Boudin, C., Molez, J.F., Ouedraogo, V. &

Carnevale, P. (1985) La transmission du paludisme en zone de

savane arboree et en zone rizicole des environs de Bobo

Dioulasso (Burkina Faso). Annales de la Societe belge de

Medicine Tropicale, 65 (Suppl. 2), 201–214.

Robert, V., Ouari, B., Ouedraogo, V. & Carnevale, P. (1988) Etude

ecologique des Culicidae adultes et larvaires dans une riziere en

Vallee du Kou, Burkina Faso. Acta tropica, 45, 351–359.

Robert, V., van den Broek, A., Stevens, P. et al. (1992) Mosquitoes

and malaria transmission in irrigated rice-fields in the Benoue

valley of northern Cameroon. Acta Tropica, 52, 201–204.

Rosegrant, M.W. & Meijer, S. (2002) Appropriate food policies

and investments could reduce child malnutrition by 43% in 2020.

Journal of Nutrition, 132, 3437S–3440S.

Service, M.W. (1976) Mosquito Ecology: Field Sampling Methods.

Applied Science Publishers, London, U.K.

Sissoko, M.S., Dicko, A., Briet, O.J.T. et al. (2004) Malaria

incidence in relation to rice cultivation in the irrigated Sahel of

Mali. Acta Tropica, 89, 161–170.

Staedke, S.G., Nottingham, E.W., Cox, J., Kamya, M.R.,

Rosenthal, P.J. & Dorsey, G. (2003) Proximity to mosquito

breeding sites as a risk factor for clinical malaria episodes in an

urban cohort of Ugandan children. American Journal of Tropical


Surtees, G. (1970) Effects of irrigation on mosquito populations

and mosquito-borne diseases in man, with particular reference to

ricefield extension. International Journal of Environmental

Studies, 1, 35–42.

Trape, J.-F., Lefebvre-Zante, E., Legros, F., Ndiaye, G.,

Bouganali, H., Druilhe, P. & Salem, G. (1992) Vector density

gradients and the epidemiology of urban malaria in Dakar,

Senegal. American Journal of Tropical Medicine and Hygiene,

47, 181–189.

Accepted 21 September 2004



Documents

Malaria transmission dynamics in central Côte d'Ivoire: the influence of changing patterns of irrigated rice agriculture