Journal of the American Mosquito Control Association, 18(4):329-340,2OO2Copyright @ 2OO2 by the American Mosquito Control Association, Inc.

THE POTENTIAL OF INTERMITTENT IRRIGATION FOR INCREASINGRICE YIELDS, LOWERING WATER CONSUMPTION, REDUCING

METHANE EMISSIONS. AND CONTROLLING MALARIA IN AFRICANRICE FIELDS

JENNIFER KEISER, JURG UTZINGER AND BURTON H. SINGER

Office of Population Research, Princeton University, Princeton, NJ 08544

ABSTRACT, Rice production in sub-Saharan Africa has more than doubled in the last 3 decades and thepotential to further develop rice-harvested areas is considerable. Several studies have demonstrated that thetransformation of arable land into rice irrigation might create suitable habitats for large populations of diseasevectors. Prominent among those are anopheline mosquitoes responsible for transmission of malaria. The methodof irrigation on an intermittent basis during the rice-cropping calendar has gained renewed interest as a potentiallyeffective malaria control strategy since the early 1980s. We review the experiences of the past 80 years withintermittent irrigation in the cultivation of rice. This method has been shown to reduce significantly the densityof malaria vectors by curtailing their larval development. Furthermore, reduced methane emissions and watersavings with at least equal yields were achieved in intermittently irrigated rice fields. We explore and discussunder what conditions intermittent irrigation might be beneficial in new rice-growing areas and identify stepsthat have to be taken to expand such programs in the future.

KEY WORDS Malaria, intermittent irrigation, rice fields, Africa, vector control, environmental management

INTRODUCTION

The world's population is projected to grow fromthe current 6 billion people to about 9.4 billion by2050. The last 4 decades have seen unprecedenteddemographic events, as dramatic changes in fertil-ity, mortality, and population growth rates oc-curred, with more people added to the world's pop-ulation over the past 50 years than in the precedingmillion years (Raleigh 1999). Unfavorable climaticconditions and inappropriate systems for access anddistribution of agricultural products contribute toserious threats of food security. By 2O20, the globaldemand for maize, rice, and wheat is estimated toincrease by approximately 4OVo. This translates toan average annual increase of about l.3Vo (Mann1999). At present, more than 800 million people,predominately women and young children, arechronically malnourished. A recent report pub-lished by the United Nations Food and AgricultureOrganization (FAO 2001) emphasizes that 17 coun-tries in sub-Saharan Africa currently are facing ex-ceptional food emergencies as a result of persis-tently difficult weather conditions, and ongoingcivil strife or war. Food imports and aid to Africahave risen to unprecedented levels in the past fewyears. Every year, African countries buy approxi-mately 3 million tons of rice, but food aid does notoffer a permanent solution.

Rice is the staple crop for about one half of theworld's population (Fischer et al. 2000). Rice is byfar the predominant crop on the Asian continent.The importance of rice to national food security isfurther increasing in Asia and elsewhere. In the1950s and 1960s, the Green Revolution and its un-derlying technological advances led to major in-creases in grain production, mainly through the de-

velopment of high-yield varieties. However, we arecurrently confronted with an evident decline in therate of crop yields (Conway and Toenniessen1999). An increase in the production of food ingeneral, and rice in particular, is urgently required.

Irrigation has played an important role in in-creasing agricultural output since ancient times andwill be a part of future strategies to enhance foodproduction. At present, only 8.5Vo of Africa's ag-ricultural production systems are under irrigation(FAO 1997), but recent trends have shown that ir-rigation increases by about lVo per annum, and con-siderable potential exists to develop rice-harvestedareas further.

Concern is rising in public health circles that ag-ricultural gains may be associated with substantialnegative health consequences. A large body of lit-erature documents increases in vectorborne diseas-es that are consequential to the introduction of newirrigation schemes. Currently, malaria accounts for300-500 million clinical attacks and more than Imillion deaths every year, mainly of children underthe age of 5 in sub-Saharan Africa (WHO 1998).Almost 9OVo of the global burden of malaria cur-rently is concentrated in sub-Saharan Africa (WHO1999). ln addition, recognition is growing that therelease of methane from stably flooded rice fieldsplays an important role in climate modification.Pressure to conserve freshwater also is rising dra-matically on a worldwide scale.

The purpose of this paper is to review the liter-ature on intermittent irrigation in rice field ecosys-tems as a potential strategy for reducing malariavector densities, increasing rice yields, and lower-ing water consumption and methane emissions. Wefirst introduce the main rice ecosystems in theworld and place particular emphasis on rice agri-

330 JouRNAL oF THE AMERICIN Moseutro Corrtol AssocIATIoN Vor-. 18, No. 4

culture in Africa. We then summarize the experi-ences made over the course of the past 80 years,since the lst experimental trial with intermittent ir-rigation was carried out in a rice field in Bulgaria.We clarify the circumstances under which waterconservation, increased rice production, and re-duced methane emission might be simultaneouslyachievable. Finally, we identify concrete steps thatcan be taken to expand such programs in the future.Progress will depend on a subtle interplay betweeneconomic, political, and scientific issues.

RICE ECOSYSTEMS

Rice is 1 of the major food grains of the world.The development of high-yield rice varieties and,consequently, lower prices have enhanced the im-portance of this crop. In 1996, rice was the predom-inant crop for 2,89O million people in Asia, 40 mil-lion in Africa. and 1.3 million in the Americas(FAO 2000). Some 120,000 varieties of rice are es-timated to exist worldwide and research is ongoingto develop and promote new rice varieties. Rice isthe most common crop under irrigation, becausewater affects the physical character of the plant, aswell as the nutrient and physicochemical character-istics of the soil (Self and de Datta 1988). Fieldduration is usually between 90 and 120 days, butnew varieties often mature earlier.

The technological advance that led to greatachievements in rice production over the past 40years was the development of high-yield rice vari-eties. Especially in Asia, this resulted in an enor-mous increase in rice yields. Growth in grain har-vests even exceeded population growth between1960 and 1985, at least partially because of theGreen Revolution. Technical approaches to increasethe yield and yield stability included modificationof plant types, with an increase in number of grainsper panicle, rice hybrids, and the selection of de-sirable recombinants. Biotechnological approachessuch as cloned novel genes currently are under in-vestigation (Khush 2001). However, expansion ofthe Green Revolution has also incurred costs to theenvironment, as fertilizers and chemical pest andweed control have led to pollution of freshwaterbodies and groundwater through leaching. In addi-tion, from 1985 onward, the growth in grain harvestfell behind population growth because of a slowerincrease of irrigation and fertilizer use. Overall, theadaption of new technologies has been a necessarybut not sufficient condition leading to an increasein rice yield.

Current terminology distinguishes among 5 riceecosystems: upland, very deep-water, deep-water,rain-fed lowland, and irrigated systems. Elevation,rainfall pattern, flooding, and drainage are the char-acteristics used to define these ecosystems. Uplandrice, grown without surface wate4 is seeded onslopes. Yields are usually low because of the lackof moisture. Deep-water rice and very deep-water

rice ecosystems are common in Southeast Asia.Rice plants are transplanted or seeded in floodedfields, usually adjacent to rivers or oceans. There-fore, the flooding patterns depend on rainfall, riverflow, floodplain geomorphology, or tidal fluctua-tions. Soils in these ecosystems often suffer fromsalinity or toxicity. Rain-fed lowland ecosystemsare predominant in relatively densely populated,poor rural regions and are typically characterizedby considerable variation in flooding and yields de-pending on the pattern and the total amount ofrain-fall. Rain-fed rice often is grown in areas that aredifficult to irrigate. Irrigated rice ecosystems nowrepresent 557o of the world's harvested rice areaand they contribute to 757a of the world's rice pro-duction. These ecosystems are characterized bycontrol of the water level and high yields (IRRI2OO2).

RICE AGRICULTURE IN AFRICA

In 2000, the total rice-harvested area of Africawas 7.7 million ha, a 5Vo share of the world's total.Within 3 decades. a 947o increase occurred in thearea used for rice production. The rice-growingarea in West Africa expanded by l44Vo, whereasexpansion was considerably lower in eastern andnorthern Africa, at 59 and 36Vo, respectively (Fig.la). Over the same period, the total rice productionincreased by 135Vo, with the most dramatic increaseoccurring in West Africa (2477o), followed byNorth Africa (13O7o). A significantly lower andonly moderate increase was attained in East Africa(39Vo; Fig. 1b). The increase in irrigated land overthe last 30 years is depicted on Fig. lc. The irri-gated areas grew by 48Va. However, at present, ir-rigated agriculture is not practiced widely in Westand East Africa. Here, rice production systems aredominated by upland rice. In West Africa, rain-fedand irrigated ecosystems only account for 15 and17Vo, respectively. On the other hand, upland riceis grown in 54Vo of the area (Garrity 1988). In starkcontrast, in North Africa, rice is grown mainly un-der irrigation with I crop per year.

Overall, irrigated land in sub-Saharan Africa lagsfar behind the rest of the world with only 3.5Va ofthe land currently irrigated (FAO 2002). Irrigatedagriculture in Africa can never sufficiently enhancefood production on its own to keep up with therapid growth of population. However, irrigationmust be part of a broader strategy. Since 1970, thepopulation has more than doubled in Africa (Fig.ld). According to recent estimates, rice agricultureworldwide is expected to expand by TOVo over thenext 25 years to support increases in food demand(Schimel 2000).

Irrigation provides an opportunity for agriculturein arid areas and can stabilize yields in regions withunpredictable rainfall (Imevbore 1987). Africa hasbeen the focus of several studies carried out underthe auspices of the FAO, which attempted to assess

Dpcer\aapn 2002 INrsnt\4rrrsNr Innrc.a.rroN ron Mer-a.ntn CoNrnol n Apnrca -t-t I

6

o -9 c

a 4r l a

21

0

North Africa East Africa West Afri€

mation on soil conditions, water resources and po-tential for irrigation are displayed on several Websites (FAO 2002). Four drawbacks must be kept inmind. First, 60Vo of the irrigation potential is lo-cated in humid regions, almost 25Vo in the Congoarea alone. In these regions, rain-fed agriculture hastraditionally dominated over irrigation (FAO 1997).Second, a considerable shortage exists of trainedmanpower for implementation, operation, andmaintenance of irrigated systems. Third, fundamen-tal aspects of existing land use patterns and humancultures and traditions must be considered beforeselecting a strategy for agricultural development.For example, substantial differences exist betweenEast and West Africa (Ellis and Galvin 1994).Fourth, the introduction of large-scale rice irriga-tion projects might constrain biodiversity or changevector distributions. The possible ecological con-sequences must be carefully examined before im-plementation.

RICE PRODUCTION AND MALARIAVECTORS

With virtually no exception, the agricultural sur-veys to investigate the potential for irrigation in Af-rica do not consider that rice growing under irri-gated water might create suitable habitats formalaria vectors. On the other hand, comparison be-tween the maps of Africa's irrigation potential andmalaria vector suitability areas (FAO 1997, MARA2OO2) reveal almost identical distributions. Of the5 rice ecosystems discussed above, upland rice,grown without water accumulation, is the only onenot associated with malaria. However. the diseaseis linked to shallow and standing water bodies thatoften are created by traditional rice production.About one quarter of all Anopheles species that areknown vectors of malaria are able to breed in ricefields (Carnevale and Robert 1987).

The introduction of irrigation with a network ofcanals and dams or the extension of wet periods forcontinuous rice cultivation establishes new aquaticenvironments (Bang 1988). Furthermore, the lon-gevity of the mosquito is enhanced because ofgreater humidity during irrigation (Service 1989).In principle, rice fields may be habitable by differ-ent species of mosquito throughout all stages ofplant growth (Lacey and Lacey 1990). Mosquitodensities usually decrease as the plants develop, be-cause mosquitoes have reduced access to the waterfor oviposition (Rafatjah 1988). In general, waterdepth, duration of flooding, temperature, cropgrowth rates, and the area of the rice field deter-mine vector productivity (Bradley 1988). Further-more, malaria transmission is greatly influenced byenvironmental and climatic factors, because vecto-rial capacity is strongly driven by humidity, rain-fall, and temperature. Protective host genes, clinicalimmune responses, and parasite variation all con-tribute to malaria morbidity and mortality. Prepro-

ts

c

tq

1 8

1 6

1 4

1 0

6

2

0

t

8 1 0

co ^

! -q

'-t 2

800700

g 600g 500

.5 4oo6 rnn- "'-3 zooc

100

0

North Africa EastAfrie WestAfrica

North Africa East Africa West Africa

NorthAfri€ EastAfrica WestAfri€ Atri€

Fig. l. Development of (a) rice-growing area, (b) riceproduction, (c) irrigation agriculture, and (d) total humanpopulation in Africa between 1970 and 20OO (open bars,1970; light gray bars, 1980; dark gray bars, 1990; blackbars, 2000).

areas suitable for agricultural production. The con-clusion was made that Africa has an irrigation po-tential of 42.5 million ha. This is about 77o of thetotal area with soils and terrain potentially suitablefor irrigation (6fi) million ha). At present, less than3OVo of these 42.5 million ha are used for irrigationof any crops, and more than 5O7o of this area hasbeen estimated to need rehabilitation before it couldbe utilized for irigation. Country-specific infor-

L

J J L JounNnr- op rHe ANleRrc.lN Mosqurro CoNrnor- AssocrartoN Vor-. 18, No. 4

ject data often are absent. Therefore, quantificationof the relative contribution to overall malaria trans-mission of vectors breeding in rice field ecosystemsis extremely difficult.

On the other hand, the association of rice pro-duction and incidence of malaria has been recog-nized for centuries. For example, this associationresulted lst in the regulation, and 2nd in the out-lawing of rice cultivation, making rice growing acapital offense in Spain in 1489 (Lacey and Lacey1990). Although the larger the production area ofrice cultivation, the higher the expected density ofanophelines, as demonstrated in China (Baolin1988), this does not necessarily imply an increaseof malaria incidence or an increased risk of expo-sure. as demonstrated in several African countries(for review, see Ijumba and Lindsay [2001]). Trans-mission intensities in communities living in closeproximity to irrigation schemes can be higher, equalto, but also lower, when compared to villages out-side rice irrigation systems (Ijumba and Lindsay2001). The relationship between irrigated area un-der rice cultivation and the annual parasite inci-dence of malaria in India revealed a negative cor-relation in 23 states (Sharma et al. 1994). Ricecultivation in Burkina Faso even led to decreasedmalaria transmission, most likely because of thedifferent vector status of Anopheles g,ambiae Gllesin the rice-growing areas (Service 1989). In con-trast, malaria epidemics were reported from variouscountries all over the world after rice transplanta-tion (Surtees 1971). In Burundi, the vectorial ca-pacity of Anopheles arabiensis Patton was shownto be 150 times higher in the rice field area com-pared to the cotton-growing area (Coosemans1985). Consequently, it has been postulated that theimpact of establishing new rice irrigation schemeson malaria might be less problematic in areas ofhigh and stable malaria transmission. The socialand economic development, often induced by in-creased rice production might be, at least partially,invested in antimalarial rneasures (e.g., purchase ofinsecticide-treated bed-nets), and hence contributeto a decline in malaria prevalence (Ijumba andLindsay 2001). However, the creation of new ricefields in areas where malaria transmission is lowmay lead to epidemics and pose a great toll on pri-mary health care systems (Najera 1988), as ob-served, for instance, in newly developed swamprice areas in Sierra Leone (Gbakima 1994). In ad-dition, agricultural development resulting in bettersocioeconomic conditions usually attracts laborfrom different regions, including workers from ar-eas with low malaria transmission who are mostvulnerable to malaria attacks.

The 2 most efficient malaria vectors in Africa,An. gambiae sensu stricto and An. arabiensis, havesimilar requirements for their larval habitats, buthave distinct preferences for microhabitats. Impor-tantly, the larvae of both species are found abun-dantly in man-made habitats and are able to readily

adapt to anthropogenically induced environmentalalterations or even to develop new behavior pat-terns (Collins and Besansky 1994). An importantfeature of An. gambiae is its preference for layingeggs in standing water near houses. With a flightrange of about 1.5 km (Surtees l97l), indoor-rest-ing densities of this mosquito are directly related tothe distance of houses to rice fields. Climate suit-ability zones for An. gambiae show a large range,because these mosquitoes are found in areas withtotal annual precipitation of 330-3,224 mm. Dis-tribution of An. arabiensis is restricted to a.reas withlittle variation in precipitation, ranging from237 to415 mm (Lindsay et al. 1998). Thus, An. gambiaepredominates in saturated environments, with den-sities peaking during the rainy season, whereas An.arabiensis is more common in dry areas. Membersof the An. gambiae complex breed in open pools,exposed to sunlight often in centers of rice fields.These mosquitoes tolerate high water temperatures,because sunlit pools often reach 40'C (Minakawaet al. 1999).

Anopheles niliTheobald and An. mouchetiEvansare 2 malaia vectors of regional importance thatbreed at the edges of rivers and in the forest fringes,respectively (Fonteni l le and Lochouarn 1999).Anopheles funestus GTles is the main malaria vectorin swampy habitats normally overgrown with veg-etation and shaded breeding sites. Therefore, An.

funestus might occur at high densities later in thegrowth period of rice.

EXPERIMENTAL TRIALS OFINTERMITTENT IRRIGATION FOR

CONTROL OF MALARIA

Water regimens in controlled rice surface irriga-tion schemes can be categorized into 4 types: stableirrigation (constant water level with minimum sup-ply to compensate for loss of water), renewal irri-gation (constant water level with periodic renewalof water), fluctuating or flowing irrigation (constantwater level with continuous irrigation and drain-age), and intermittent irrigation (periodic irrigationand drainage) (Mogi 1988). Intermittent irrigationis the method of alternately irrigating and, passivelyor actively, drying the field for several days. Theprocess starts about 2 wk after rice seedlings aretransplanted and lasts for about 10-15 wk until theplants reach maturity. Intermittent irrigation is onlyfeasible where climate conditions favor rapid dry-ing. Draining rice fields during rainy seasons or inwetlands often is impossible.

It is important to note that intermittent irrigationwas introduced in Asia about 300 years ago, pri-marily to obtain higher rice yields. The need formidseason drying to obtain higher yields also wasrecognized in Japan more than a century ago. In-termittent irrigation followed in the beginning ofthe 20th century. In 1955, a Japanese rice produc-

DscsNaeBn 2002 INreRMrtrpNr InrrcarroN pop MaLl.ma CoNrnor- rN Arnrca J J J

tion contest was won by application of intermittentirrigation (Mogi 1988).

A series of experimental trials dating back to thel92os has been carried out in different ecologicaland epidemiological settings to study the effects ofintermittent irrigation as a potential tool for malariacontrol. The basic concept is to intemrpt the repro-ductive cycle of the mosquito by withholding wateron a periodic basis from the rice field. Because tra-ditional rice fields are kept flooded for an entirecropping season, in certain areas even for 2 or 3crops, multiple mosquito life cycles can be com-pleted. In contrast, intermittent irrigation, curtailsthe development of the adult anopheline mosqui-toes. Intermittent irrigation is particularly feasiblein places where control of the water supply anddrainage are possible. For best results, intermittentirrigation should be applied to all rice fields overlarge areas and during the entire cropping season.The method is less feasible in drought-prone areas,because water is too valuable to be drained. It isimportant to note that the rice fleld itself cannot beconsidered in isolation, and fallow fields, drainageditches, or irrigation canals also must be monitoredcarefully because they provide optimal breedingplaces. For example, Knipe and Russel (1942) de-scribed in detail the filling of borrow pits and wellsto eliminate vectors from channels and tanks.

The study designs and the outcomes of the trialsof intermittent irrigation conducted in small exper-imental fields are summarized in Table 1. The first3 studies (Konsuloff 1922, Ananyan 1930, Eniko-lopov 1931) were done in Bulgaria, Armenia, andDaghestan, respectively, 70-80 years ago. In the2nd half of the l93os, several experiments wereconducted in Portugal (Hill and Cambournac 1941),Bali (Smalt 1937), and Indo-China (Antoine 1936).These countries were endemic for malaria at thetime of trial implementations. In the early 1940s,experimental trials were extended to India (Knipeand Russell 1942, Russell and Knipe 1942) andKenya (Grainger 1947). A paucity of further ap-praisal of intermittent irrigation existed for morethan 3 decades, which can probably be explainedby the advent and widespread application of indoorhouse spraying with DDT Experiments with inter-mittent irrigation were continued from the mid-1970s onward in China (Luh 1984), India (Kris-hanasamy et al. 2000), Indonesia (Mather and That1984), Japan (Mogi 1993), Kenya (Mutero et al.2000), and Peru (Chang, personal communication).The method of intermittent irrigation is integratedinto several current rice culture methods withoutspecial mention. For example, the World Health Or-ganization reported intermittent irrigation to be suc-cessfully practiced in Korea (WHO 1983).

Although the trials were conducted in distinc-tively different ecosystems and the experimentaldesigns varied from I site to another, I commonfeature occurred. Intermittent irrigation resulted inhighly significant reductions in the density of lar-

vae, pupae, and adult mosquitoes, reaching levelsof up to 95Vo.It is now widely acknowledged thatthe wet period of the irrigation cycle must be con-trolled most carefully to reduce the density of ma-laria vectors. Konsuloff (1922) and Jettmar (1951)identified additional lethal effects such as sunlight,presence of fungi, or ants that feed on larvae indried rice fields. The time required from egg de-position by Anopheles through larval developmentto the adult stage is approximately 18 days, de-pending on the species and its aquatic requirements(Hill and Cambournac l94l). Somewhat shorter cy-cles occur in the warmer tropics (Russell and Knipe1942, Mather and That 1984). To prevent emer-gence of adult Anopheles, a marked moisture re-duction of the soil is necessary. At present, little isknown about the exact length of time that larvaeand pupae can survive on dried soils. Some anec-dotal evidence exists that soils with only 2OVo mois-ture are hostile environments and kill the majorityof the larvae. Thus, larvae are unable to survivesufficiently long to complete the development cyclewhen the ground loses its surface water. However,several authors have described the longevity ofeggs and larvae of Anopheles in dried soil (for re-view, see Jettmar tl95ll). Annual rainfall patternsand complete soil drying must be considered. In 3studies carried out in India, Kenya, and Tanzania,the soil did not dry completely and consequently,mosquito densities continued to be high (Grainger1947, Krishanasamy et al. 2000, Van der Hoek etaI. 2000). Because the physical characteristics ofthe soil influence the drying pattern, this must bekept in mind when designing the wet-dry cycles ofintermittent irrigation. Reviewing the trials revealsthat only scarce information is available about soilcharacteristics and water-holding capacities. Russelland Knipe (1942) designed a series of experimentswith fixed wet periods of 5 days and dry periodsvarying between 1 and 5 days. Larval developmentwas intemrpted when the dry spells were 3 days orlonger. Interestingly, a modified form of wet irri-gation carried out in India, where the soil nevercompletely dried out but instead left many pools ofwater, also showed a reduction in the density ofmosquito larvae and pupae. Enhanced predationpressure in crowded puddles has been suggested asa possible reason (Rajendran et al. 1995). lnfor-mation on the effects of malaria incidence and mor-bidity are missing in all but I of the studies. In theSenjayakollai area in India, Knipe and Russell(1942) observed a reduction of spleen rates in chil-dren from 48Vo to 4Vo and a reduction in parasiterates from 42Vo to OVo in the 7941 malaria seasonafter introduction of intermittent irrigation. Futurestudies should monitor incidence rates and vectorpopulation densities to evaluate whether or not acausal link exists between intermittent irrigationand reduction of malaria morbidity. Rice yields andwater-saving outcomes also should be quantifiedwith great care, because these 2 outcomes are crit-

334 JourulL on tge Avpnrc,q.N Moseuno CoNrnol AssocrarroN Vor-. 18, No. 4

E ! i c ' 5 c

5 : E E : E H7 , . = . : . O . E Ei 6

' ; _ L , i :! r - ^ v , ) E o$ J 9 ! I : r X ! i ; b 0: 9 g s . : d ? ; b - ES - l s - t H g i g 3 j9 S f i Z

. = F . E ' - o- . = - o t

* = E I >

E E3f vI F O V E

T: ierliEi iieii*iEii ii,r FE g:z

% Q' C l

IN.q *. 9 - *

R S i ' -s Q r . 9s J { q

.F

c!

cd

6 A Oa > r > r > l

s b € € €E E > , x > .- 6 i i i

€ ! c ' l o *o " " .* . 2 6 6 6

* 6 8 ' ;> i i r ) r ) \ ^

q o

9 9 ,' E nR 6 i

" 5 - 96 ;

e r 8 .' € = v

' F Y 9

I i H " 5.= 9? *F, t ^, - Y x - E

B B x u + E $ 3o o t r s ^ - - o

5 5 f E ; g F l2 2 6 z

B B 3

Ja ,i< y

6 6 6z z z9 r Ep i a ! s b . E E > ,; , F E [ , ] E P E =L ' ' h ! ' a . : i . S o b- . = : E - - � c 9 : E !I - . = > E 1 2 - = : ' ' E

s : t : H ; q i $ : E9 a a b 5 F € N . d . =+

B

Z

B

x

z

4

F :: ,(Ja ! ^

{ i s c ' a: o . = - : o. v - R u =: . H . VI I v l vz

q . 3 q

i i : i = ^X x $ I i 6S : S i ^ \ - B =: i \ s ; = s F! P l q F ! . "S e S s , i f : S1 3 s ' < Y s q -

B

x

Ezi o i io > :

E i 5 ,.P s t

h € adv ' ) xn d 9+

. o. - i oH F ! , o t r

x x ' d c !a v - c i . a

' c 6 ^ - - 5

F o F ! i [ i : a= , ^ I N . i I O

+ E E ' 6 . 8 ; * tg E ! E ! ; . E = st : n(t) z \o

g E3 ! E -s,->

u q . r e

; : o ^ :Y .=E 0 6 Y t r

. E X r d t rh - n o 9

x , =

@ -

O \ o ; i Ni O \ r : * +

6 . . ^ = *t r l l x - = ' e' F . = ; n - , * o =

U , 9 g E : . ? H. e i . ' 6 . d g ^ J! v ? v € v H

. i > n € Y

g . ! l E ^ € =o ' = a b 0 ; t r +

- C . i i X 5 ; n d . i .

6 s J i E + v r

! i oc d t r . Y> 9 + kX > C P ; 'd - : g ' i ^ E gE E E R E : i

o

>'

cA C.l

o o3 B

cn o\

BC N

v o

z o* S E r

i : - . x r . =" o - ^ Sb 3 i a €

! o ! . q Ii N v l >

. : q a ; Y 99 Y > , i ' " or E € 5 b s

\o

} B

z zN F "5 .9

, i E x u' t r . E i : E0 d o - >C t r a Q a 'x - > x - ao d a ! o S )

- ! e - - j

i : h > i : 9 -

O : L O l ( . . l> t r ' O > m io 6 )

U) C)

o c )

o o

6- o \X * f i -

. = r ) 'd o ^ € C

e oc.r h trb0tr4 o, i <

= v + F v

6 d

9 : !' i 6

V E

tr

o

. o

hi q

6 ea )

.= : l

= -

ho' r

oI

o

.9I>'O

g

s,g>, -o): . ! )J 9

U

-

C)

-

o

!

str

o

. F

a

Cd

a F

h -> : io r ix cd F

€ r

U =' t r o

- d

b0-

oI

I

o

o

q

0

xo

Fa

o

Jo

.G-

DeceMsrn 2002 INrenunreNr InnIcanoN pon MlLeml CoNrnol IN AFRTCA 335

B }

) a v

z z

o v

o e o ez z

! > 'a E a> ' ! 1 o

v ! e ^

' . F E g3 r E f ^

o \ - Fr e \ . = . E

€ \ o

*. ? i ?s * ' 6. : E qs o t #i , v %

. i >

i ! "{ r n

(h

; .1u- .E F"t r F o - . :

. A : E S E9 - 6 6 s

t r ! . s

P E i f ; E :; u : ; s n

c.)

JZ !: >,

H $ - F E .; c R A E RH F i i . r iE 3 u f f 5 r

x ] i) i r c

* a € 1 9 0

B E F q * ; ; E ", t 9 ' & : i 5 g 6 3c ! d ; : , d O , G '; } u a : t r o . g c z - :F S O E

- ) , _e ; $ t6

= i F J- > E . A = o ^v = : e s >r x . , l F g 6 ' t r I6 U . t . g : E B BE ; V

! d i i c i l . E ! . . 9 o

E i ; s 8 , e r a E S* n O N ; c t . F -o q \ ! . ; * - - o= o \ o : Z Z

' 3 ; B ;

Y . s I J Z J I &

z z z z z

. . v 1 S : 9 V i " l + N o " k

r t : ia ,E ' : is ;$E.{g E: -+ E5; !?6 ? ; € ? E b = * - g E E : E - E E _ i ; l i E f g

i H F : + g 3 E ; E s = - q . ̂ ^ E - . i N d E ; s ff l i o b i - s i - 6 E U d : E 5 F E . i E E g s i f r r * l eF X > i : C . : n . = Y X t iE c . r - .

t r ' 5 g 5 g . c d , € g E L . = - ' = t r 7 r . = C - ' f C >( E E A S f i

: E

! ; . E !c c E 9 E r; ' : t h h o2 Y a - a ' = = d

- ! . i = ; : = . P6 6 J ; ' * s !Z Z F . . +

'

z

.3

8 = S ! = . = . r 4 ss ; p q : e 3 ^ 3

' i . ! ^

s ! i t = b ! i ! ! s tY , F o o P l i . ; , e . i ; ; f l: ) : F : i : F ' . ' . : 0S S S v E v e - \ t R vd < {

sr f^,= . E ? =

: i € E . u sE E o g E F! ' 2 ^ ! . ' E h ;E 5 , E 6 i < ;Z Z X v )

B

)(

z

o

o

d

!

3, a * i ' $ $ rE: . . -e i sH $ i - E * € : A ; e i ; g gl s i € E 6 5 : ; ' E E t [ * i :F : i i e g f s E E * s 5 a s s s, n : . , o l

nN

, .S >

+o

> ? .

. o 9O L

; 9

E orr

6

a)

' l

0)

c-s

61 .

O\ ;o

6 vV

; ; d

v c

E

0

!

. o

qa

d

: h

E tj {

o

d

0)

II

()

do

: : ^:/ c)

h o

: , 95 9

O

q

E

.tr-

;tr-

io

Q

-

Jounr.rAr- op rue AvenrclN Mosqurro CoNrnol AssocranoN Vo l . 18 , No .4

o

U

F

I

,|-

o

.tr-

o ! !' l 6

V E

E

0)

. o

o

a ^! ? s

6 A. n =

6 {

aoa

O

U

: Y ^> ' u> , ! ): . o

o 5U

B }^ ^v &

Z Z

. i ! o o P , ,

i ; : p l s i € H is i ! e ; i r : r E' l H ' 5 : c - ' o - X f , - o

E X H e * : P r t e i $g b o E t r i . : i ; B t r o i o oc

!

{ ,- ! . , t : 6

. ; t r c o 0 ? o - oo Q

- ' i ; l o b \

o s r ' d ! - ' a c t

9 t t r ' - > \ _ Nb 6 . i . i u > ,b ! E E 6 ; 3 !

k : 9 ; o i . v i j

o h o t r o t . = > lZ F

sP ^

< s ! xn { \ ;F + N i

. t E l i f rS U i ' gt > I =S - * r

F

, F 8 - d -: e Y 5

i . F F i -. q F r . O "

H i o b o o @ 1; H..E .E .d ds- -.i -.i ;

o . 9 t r . 2 t r - Y t rd ) ! O + O a 0 a

E X " j - q i \ c l 6 q r ;

F ^ t r > .o l >F > \NI o o m l t l ' o i a- t r6 - t r - t r - t r= E N z \ c d > a d > \ d

. ! S E Y B i T Y Eb E 6 m N \ o - o \ NU

9 3

: ? Fx F 6 ^

r 9 r €@

s

f-

a-l

d

o - Y ; >E E - U i' - bo y :! . E E J ^: . : c ? uq : o ? 9; u h ! EH . : E > d

O

; l e

X =

$

: oX ^" O ^d f xh > 5I

ical to ensure farmer cooperation and ascertain theirsatisfaction.

A substantial literature exists that discusses theeffects of drought stress during early rice devel-opment. Insufficient quantities of water during theearliest development stages can cause severe dam-age to the seedlings and, in turn, result in reducedyield. At this stage, the plant consumes a largeamount of water, in opposition to the concept ofintermittent irrigation. The results regarding theyield in the summarized studies are inconclusive.Highea but also lower, yields were reported withintermittent irrigation and no association can bemade between yield, water cycle, and rice variety.Previously, experiments carried out in variouscountries that were designed to establish the rela-tionship between the quantity of water applied andrice yields also reported equal, higher, or lower riceyields. Several of these experiments are summa-rized by Van der Hoek et al. (2000). Use of ricestrains that are tolerant of soil drought might be anadvantage to simultaneously save water and obtainhigh yields.

INFLUENCE OF INTERMITTENTIRRIGATION ON SAVING WATER,

REDUCING METHANE EMISSION. SOILS.AND WEEDS

Water currently is the most limiting resource forcrop production and will be critical for the sustain-ability of rice production and the development ofnew irrigation schemes. In Asia, where water al-ways has been regarded as an abundant resource,per capita availability declined by 4O-6OVo between1955 and 1990. Extrapolations indicate that mostAsian countries will have severe water problems bythe year 2025 (Riceweb 2OO2). Twice as much wa-ter is needed for rice growing as for the productionof any other cereal. The traditional practice of con-tinuous flooding of rice fields causes large waterlosses through deep percolation, especially if riceagriculture is further extended to permeable sandysoils (Sandhu et al. 1980). Therefore, future riceproduction will depend heavily on water-savingmeasures. Projects are underway to develop newrice varieties that mature earlier and are less waterdependent. Other research focuses on the cultiva-tion of wet-seeded rice. However, the overall trendis to conserve water in rice agriculture with well-adapted irrigation regimens.

High water-use efficiencies were obtained by in-termittent irrigation as compared to continuousflooding in 2 studies in India (Jha and Chandra1981, Pant et al. 1990). The soil characteristicswere demonstrated to be of central importance.Sandy-type soils, with a low water-holding capac-ity, favor intermittent irrigation as a water-savingmethod. In contrast, prolonged drying of the ricefields, especially on clay soils, may result in crack-

DscsNaeen 2002 INTERMITTENT IRRIGATIoN FoR MALARIA CoNTRoL tN Apnrca -)-) I

ing of the soil and hence increase the demand forwater (Lu et al. 2000, Bouman and Tuong 2001).

Flooded rice fields are I of the major biogenicsources of atmospheric methane and therefore areimportant contributors to the greenhouse effect. Thepotential for methane release from rice flelds hasIong been noted, but comprehensive measurementsof methane fluxes in rice fields have been reportedonly since the early 1980s. Rice field methane isgenerated through microbial metabolic processes inthe soil. These processes seem to be fuelled by rootexudation and death, and methane is transportedthrough the rice plant into the atmosphere (Redekeret al. 2000). Water regime, temperature, and soilproperties, as well as rice variety, are the majorfactors determining the production and flux ofmethane in rice fields. Methane concentration hasmore than doubled during the last 200 years and asustained increase is believed to contribute signifi-cantly to climatic changes. Thus, an intensificationof rice agriculture is causing rising concern withregard to increased methane production. Watermanagement was shown to have an enormous im-pact on methane emission rates: removing flood-water decreases methane emission, as a result ofsoil aeration. Several experiments conducted in dif-ferent countries are summarized by Wassmann etal. (2001). Intermittent irrigation could be demon-strated to reduce methane emission by l5-88vo(Sass et al. 1992, Wassmann et al. 2001).

Soil degradation caused by flooding has great ef-fects on nutrition of higher plants. Products re-leased by flooding are nitrogen gas, ferrous iron,sulfide sulfur, and organic acids, which alter theavailability of soil nutrients (McKee and McKevlin1993). Although continuous flooding has been re-ported to result in higher uptake of nutrients andthus a higher yield (Ogunremi et al. 1986), soilsdiffer widely in their capacities to release nutrients.No general conclusion on the influence of the waterregimen on nutrient uptake can be drawn. Indeedin some instances, drainage might be helpful forimproving soil quality. Drainage results in controlof salinity or acidity in problematic soils (Mathewet al. 2OOl), or the leaching of organic and inor-ganic toxins, which might accumulate in the soilbecause of low soil temperature (Neue and Bloom1989) .

Weeds are universal companions of rice fields,with yield losses reaching very high levels if theweeds are not controlled. The primary factors thatencourage weed communities within rice field hab-itats are the water status of the field and the cropplanting method. Maintaining standing water is be-lieved to be an effective weed control practice be-cause aerobic weed populations decrease with anincreased water depth (de Datta and Baltazar 1996).This strategy is in direct opposition to intermittentirrigation. However, interactions among water andweed management practices are complex, and arefurther complicated by soil and climatic variability

and heterogeneity. Furthermore, draining will ex-pose aquatic weeds to dry conditions. Studies fromthe humid tropics on possible effects of tillage andwater control on weed emergence and growth in thepresence and absence of herbicides have yieldedconflicting results, mainly because of site specific-ities (Bhagat et al. 1996). In addition, research onintegrated weed management is ongoing, as scien-tists are developing environmentally sound weedmanagement systems (Riceweb 2002).

CONCLUDING REMARKS ANDPERSPECTIVES

Over the last 3 decades, rice production on theAfrican continent has expanded enormously, and thisgrowth most likely will continue in the near future.The development of high-yield rice varieties willfurther increase the significance of rice agricultureto secure global food production needs. The impactsof land and water development are diverse, and oftenare difficult to predict or are scarcely visible in theplanning stage. Negative outcomes comprise thespread of malaria and other parasitic diseases (e.g.,schistosomiasis) or further environmental pollutionor water exploitation. Therefore, deliberate measureson environmental management must be taken to im-plement irrigation schemes in a way such that ad-verse health effects to rice growers and residents ofthe area are minimized and social and economic im-provement prevails. We evaluated and found inter-mittent irrigation to be a potentially suitable strategy.Intermittent irrigation showed marked reductions inthe young developmental stages of the malaria vec-tors. However. fields must be drained of all surfacewater to hold back mosquito breeding in small tem-porary puddles. This is not only important to reducethe density of anopheline mosquitoes, but alsobreeding of Aedes or Culex species, which areknown to transmit other human disease pathogens,including filarial worms and arboviruses.

When integrated in a malaria control programconsisting of a combination of measures, intermit-tent irrigation might also contribute to saving waterand reducing methane emission with at least equalrice yields. Feasibility studies must be carried outto scrutinize the method under local conditions, be-cause each setting has its own characteristics,which largely are driven by soil type, climate, andrice ecosystem. Intermittent irrigation requires suf-ficient flexibility to sequentially modify the inter-vention to obtain a high level of performance. En-tirely missing, especially in those trials that failedto control mosquito breeding and displayed lowyields, is an appraisal of how to adapt over timethe wet-dry cycle to display the desired outcome.Adaptive tuning of this strategy to the local eco-logical settings might take several years until inter-mittent irrigation exhibits a high level of perfor-mance, as shown with other environmentalmanagement interventions for successful malaria

338 Jounrval oF THE AMERTcAN Moseurro CoNrnol AssocrarroN VoL . 18 , No .4

control (Utzinger et al. 2001). Ongoing surveillanceduring pilot testing will help to determine new pa-rameters for the protocol. Such a strategy is cur-rently under way in Peru (Chang, personal com-munication). The characteristics of the soil are ofcentral importance for sufficient drying withoutcracking, to save water and to curtail the develop-ment of adult mosquitoes. A rather high perme-ability and low water storage capacity is mandato-ry. Sandy soils, which display these characteristics,are common over large parts of Africa. Becauserice agriculture has been extended to these soiltypes, intermittent irrigation could contribute to anenormous saving of water. Water drainage cannotbe recommended on less permeable soils, and inregions where water is scarce. It has to be kept inmind that at present, rice in Africa is mainly grownunder rain-fed conditions or as upland rice. Theseecosystems are less suitable for intermittent irriga-tion. However, the area of irrigated rice agricultureis constantly growing in Africa. Nigeria and Sudanare 2 exarnples in which intermittent irrigationcould be introduced on a large scale after carefulevaluation. At present, Nigeria is the largest riceproducer in Africa, with 16%o of the rice agriculturecurrently being irrigated. Here, rice production hasgrown at an average of l4%o per annum over thepast 2 decades, which has resulted in an enormousexpansion of the rice-growing area. The currentrice-growing area in Sudan is relatively small butrapidly increasing. ln 1997, total area was 2,94O ha,which has almost doubled over the last 3 years toreach a total area of 5,46O ha. In the same period,rice production was amplified by a factor of 4. Pro-posals to expand existing irrigation projects are un-derway and the construction of new hydraulicstructures for new schemes has been launched. Thepotential risks of increased malaria incidence ratesboth in Sudan and Nigeria are high. Periodic ma-laria epidemics are reported each year, and thesenegatively influence social and economic develop-ment. Although the focus of the present paper wasmainly on Africa, huge areas of irrigated rice fieldsin Asia provide appropriate conditions for the ap-plication of intermittent irrigation.

The recently launched Systemwide Initiative onMalaria and Agriculture (SIMA), with 120 fieldsites in Africa and Asia, might provide a suitableframework for pilot trials. One of the main objec-tives of SIMA is to bring together malaria and ag-ricultural research. Interventions that have beenproposed include water management practices toreduce mosquito breeding. In addition, the Inter-national Water Management Institute currently isimplementing studies on intermittent irrigation inChina and India. The Environmental Health Projectin Madagascar currently is planning a 4-year pro-gram that will link and integrate activities in publichealth, population, and environment, including in-termittent irrigation. Within these pilot trials, em-phasis also must be given to education and training

to assure proper operation and maintenance. Cur-rently, the acceptance of farmers toward intermit-tent irrigation often is lacking, because the appre-ciation that intermittent irrigation not only reducesthe risk of malaria but also might increase rice yieldand save water is missing. Thus, effective infor-mation, education, and communication strategiesare needed to accompany the promotion and imple-mentation of intermittent irrigation. Intermittent ir-rigation might become more attractive for farmersas a water-saving method with at least equal riceyields, especially if water becomes more costly.

ACKNOWLEDGMENTS

We thank Robert Bos from the WHO for hisvaluable assistance in reviewing the literature. JtirgUtzinger acknowledges financial support from theSwiss National Science Foundation and the Centerfor Health and Wellbeing at Princeton University.

REFERENCES CITED

Ananyan SA. 1930. Intermittent irrigation of rice fields asa measure of fight against malaria in Armenia in 1928.Trop Med Vet 8:39-47.

Antoine M. 1936. Prophylaxie rurale du paludisme parirrigation altern6e des rizibres. Bull Soc Med Chir In-dochin 6:129.

Bang YH. 1988. Vector-borne diseases associated withrice cultivation and their control in Southeast Asia. In:Vector-borne disease control in humans through nceagroecosystem management Proceedings of the work-shop on research and training needs in the field of in-tegrated vector-borne disease control in riceland agro-ecosystems of developing countries, March 9-14, 1987.International Rice Research Institute in collaborationwith the wHoiFAoruNEP Panel of Experts on Envi-ronmental Management for Vector Control. Manila,Philippines: International Rice Research Institute. p 93-99.

Baolin L. 1988. Environmental management for the con-trol of rice-field breeding mosquitoes in China. ln: Vec-tor-borne disease control in humnns through nceagroecosystem management Proceedings of the work-shop on research and training needs in the field of in-tegrated vector-borne disease control in riceland agro-ecosystems of developing countries, March 9-14, 1987.International Rice Research Institute in collaborationwith the WHO/FAOruNEP Panel of Experts on Envi-ronmental Management for Vector Control. Manila,Philippines: Intemational Rice Research Institute. pl l t -121.

Bhagat RM, Bhuiyan SI, Moody K. 1996. Water, tillageand weed interaction in lowland tropical rice: a review.Agr ic Water Manage 3l :165-184.

Bouman BAM, Tuong TP 2001. Field water managementto save water and increase its productivity in irrigatedlowland ice. Agric Water Manage 49:11-30.

Bradley DJ. 1988. The epidemiology of rice-field-associ-ated diseases. ln: Vector-borne disease control in hu-mans throuqh rice agroecosystem management Pro-ceedings of the workshop on research and trainingneeds in the field of integrated vector-bome diseasecontrol in riceland agroecosystems of developing coun-

DecBveen 2002 INrsnvlrrsNr IpnrcartoN FoR MALARIA CoNrnol IN AFRTcA 339

tries, March 9-14, 1987. International Rice ResearchInstitute in collaboration with the WHO/FAOruNEPPanel of Experts on Environmental Management forVector Control. Manila, Philippines: International RiceResearch Institute. p 29-39.

Carnevale P, Robert V. 1987. Introduction of irrigation inBurkina Faso and its effect on malaria transmission. In:Effects of agricultural development on vector-bornediseases Working Papers of the 7th annual meeting ofthe joint WHO/FAOruNEP Panel of Experts on Envi-ronmental Management for Vector Control, September7-ll, 1987. AGL/MISC/12I87. Rome, Italy: FAO. p57-67.

Cates MD. 1968. The effect of improved rice farmingtechniques on mosquito population in central Taiwan.Mosq News 28:582-585.

Collins FH, Besansky NJ. 1994. Vector biology and con-trof of malaria in Africa. Science 284:1874-1875.

Conway G, Toenniessen G. 1999. Feeding the world inthe twenty-first century. Nature 4O2:55*58.

Coosemans MH. 1985. Comparison of malarial endemic-ity in a rice-growing area and a cotton-growing area ofthe Rusizi Plain, Burundi. Ann Soc Belg Med Trop65(Suppl 2):187-20O.

De Datta SK, Baltazar AM. 1996. Integrated weed man-agement in rice in Asia. In: Naylor R, ed. Herbicidesin Asian rice: transitions in weed management Manila,Philippines: International Rice Research Institute.

Ellis J, Galvin KA. 1994. Climate patterns and land-usepractices in the dry zones of Africa. BioScience 44:l-13.

Enikolopov SK. 1931. An experiment in interrupted irri-gation of rice plantations as a prophylactic measure inthe control of malaria in Daghestan. Trop Med Vet 9:58-63.

FAO [United Nations Food and Agriculture Organization].1997. Irrigation potential in Africa: a basin approachFAO Land and Water Bulletin 4. Rome, Italy: FAO.

FAO [United Nations Food and Agriculture Organization].2OOO. FAO rice information Rome, Italy: FAO.

FAO. [United Nations Food and Agriculture Organization].2001. Rome, Italy. http://www.fao.orgAVAICENT/OISiPRESS-NEiPRESSENG/200 1/prenO I 48.htm.

FAO. lunited Nations Food and Agriculture Organiza-tionl. 2002. Rome, Italy http://www.fao.org.

Fischer KS, Barton J, Khush GS, Leung H, Cantrell R.2000. Collaborations in rice. Science 29O:279-28O.

Fontenille D, Lochouarn L. 1999. The complexity of themalaria vectorial system in Afica. Parassitologia 4l:26'7-27 l.

Garrity DP 1988. Tiopical rice agroecosystems: charac-teristics, distribution, and future trends. ln: Vector-borne disease control in humans through rice agroeco-system management Proceedings of the workshop onresearch and training needs in the field of integratedvector-borne disease control in riceland agroecosystemsof developing countries, March 9-14, 1987. Interna-tional Rice Research Institute in collaboration with theWHO/FAOruNEP Panel of Experts on EnvironmentalManagement for Vector Control. Manila, Philippines:International Rice Research Institute. p 13-28.

Gbakima AA. 1994. Inland valley swamp rice develop-ment: malaria, schistosomiasis, onchocerciasis in southcentral Sierra Leone. Public Health 108:149-157.

Ge FX, Sun Z, Jiang QB, Sun YC. 1981. The study onthe control of mosquitoes in the paddy fields by wet,

irrigation in the alluvial plain of the yellow iver. ActaEcol Sin 1:168-175.

Grainger WE. 1947 . The experimental control of mosquitobreeding in rice fields in Nyanza Province, Kenya, byintermittent irrigation and other methods. East Afr MedJ 24:16-22.

Hill RB, Cambournac FJC. 1941. Intermittent irrigation inrice cultivation and its effect on yield, water consump-tion and Anopheles production. Am J Trop Med 2l:123-744.

Ijumba JN, Lindsay SW. 2001. Impact of irrigation onmalaria in Africa: paddies paradox. Med Vet Entomoll 5 : l - l l .

Imevbore AMA. 1987. Vector borne disease hazards inchanging agricultural practices resulting from overalldevelopment in Africa. ln: Effects of agricultural de-velopment on vector-borne diseases Working papers ofthe 7th annual meeting of the joint WHO/FAOruNEPPanel of Experts on Environmental Management forVector Control, September 7-11, 1987. AGL{MISC/I2/87. Rome, Italy: FAO. p 18-22.

IRRI [International Rice Research Institute]. 2OO2. Ma-nila, Philippines. http://www.irri.or gl.

Jettmar HM. 1951. Uber die Resistenz der aquatischenEntwicklungsformen der Stechmuecken gegen Trock-nttng. Arch Hyg Bakteriol 134:ll-26.

Jha KB Chandra D. 1981. Irrigation requirement of highyielding rice varieties grown on soils having shallowwater table. Indian J Agric Sci 51:732-'137.

Khush GS. 2001. Challenges for meeting the global foodand nutrient needs in the new millennium. Proc NutrSoc 6O:-15--26.

Knipe FW, Russell PE 1942. A demonstration project inthe control of rural irrigation malaria by antilarval mea-sures. -/ Malaria Inst India 4:615-631.

Konsuloff S. 1922. Die Bekiimpfung der Malariamiickenin den Reisfeldern. J Appl Entomol 8:283-294.

Krishanasamy S, Amerasinghe FP, Sakthivadivel R, RaviG, Tewari SC, Van der Hoek W. 2000. Strategies forconserving water and effecting mosquito vector controlin ice ecosystems: a case study from Tamil Nadu, IndiaColombo, Sri Lanka: International Water ManagementInstitute.

Lacey LA, Lacey CM. 1990. The medical importance ofriceland mosquitoes and their control using alternativesto chemical insecticides. J Am Mosq Control Assoc6(Suppl): l-93.

Lindsay SW, Parson L, Thomas CJ. 1998. Mapping theranges and relative abundance of the two principal Af-rican malaria vectors, Anopheles gambiae sensu strictoand An. arabiensis, using climate data. Proc R SocLond B Biol Sci 265:847-854.

Lu J, Ookawa T, Hirasawa T 20OO. The effects of irri-gation regimes on the water use, dry matter, productionand physiological responses of paddy ice. Plant Soil223:2O9-218.

Luh PL. 1984. The wet irrigation method of mosquitocontrol in rice fields: an experience in intermittent ir-rigation in China. In: Mather TH, That TT, eds. Envi-ronmental management for vector control in rice fieldsFAO lrrigation & Drainage Papers 41. Rome, Italy:FAO. p 133-136.

Mann CC. 1999. Crop scientists seek a new revolution.Science 283:-310-314.

MARA [Mapping Malaria Risk in Africa]. 2002. Durbin,South Africa. http://www.mara.org.za.

Mather TH, That TTI eds. 1984. Environmental ftKtnape-

340 JounNeL or rHr Arrlszucer Mosquno Coxrnol AssocIerIoN Vol . 18 , No.4

ment for vector control in rice .fields FAO Irrigation &Drainage Papers 41. Rome, Italy: FAO.

Mathew EK, Panda RK, Nair M. 2001. Influence of sub-surface draining on crop production and soil quality ina low-lying acid sulphate soil. Agric Water Manage 4'7:191-209.

McKee WHJ, McKevlin MR. 1993. Geochemical pro-cesses and nutrient uptake by plants in hydric soils.Environ Toxicol C hem 12:2197 -22Oi .

Minakawa N, Mutero CM, Githure JI, Beier JC, Yan G.1999. Spatial distribution and habitat characterizationof anopheline mosquito larvae in western Kenya. Am JTrop Med Flyg 61:1010-1016.

Mogi M. 1988. Water management in rice cultivation andits relation to mosquito production in Japan. ln: Vector-bome disease control in humans through rice agroeco-system management Proceedings of the workshop onresearch and training needs in the field of integratedvector-borne disease control in riceland agroecosystemsof.developing countries, March 9-14, 1987. Interna-tional Rice Research Institute in collaboration with theWHO/FAO/UNEP Panel of Experts on EnvironmentalManagement fbr Vector Control. p l0l-109.

Mogi M. 1993. Effect of intermittent irrigation on mos-quitoes (Diptera: Culicidae) and larvivorous predatorsin rice fields. J Med Entomol 30:309-319.

Mutero CM, Blank H, Konradsen E Van der Hoek W.2000. Water management fbr controlling the breedingof Anopheles mosquitoes in rice irrigation schemes inKenya. Acta Trop 76:253-263.

Najera JA. 1988. Malaria and rice: strategies and control.ln: Vector-borne disease control in humans throughrice agroecosystem managemenl Proceedings of theworkshop on research and training needs in the field ofintegrated vector-borne disease control in ricelandagroecosystems of developing countries, March 9-14,1987. International Rice Research Institute in collabo-ration with rhe WHO/FAO/UNEP Panel of Experts onEnvironmental Management for Vector Control. Ma-nila, Philippines: lnternational Rice Research Institute.p 123-132.

Neue HU, Bloom PR. 1989. Nutrient kinetics and avail-ability in flooded rice soils Manila, Philippines: IRRI.

Ogunremi LT, Lal R, Babola, O. 1986. Effects of tillagemethods and water regimes on soil properties and yieldof lowland rice from a sandy loam soil in southwestNigeria. Soil Tillage Res 6:223*234.

Pant CP. Rathore TR. Mishra HS. 1990. Effect of inter-mittent irrigation on groundwater table contribution, ir-rigation requirement and yield of rice in mollisols ofthe Tarai region. Agric Water Manage 18:231-241.

Rafatjah HA. 1988. Malaria vector control: environmentalmanagement. In: Wernsdorfer WH, McGregor I, eds.Malaria: principles and practice of malariology NewYork, NY: Churchill Livingstone. p 1135-1172.

Rajendran R, Reuben R, Purushothaman S, VeerapatranR. 1995. Prospects and problems of intermittent irri-gation for control of vector breeding in rice fields rnsouthem lndia. Ann Trop Med Parasitol 89:541-549.

Raleigh VS. 1999. Trends in world population: how will

the millenium compare with the past? Human ReprodUpdate 5:5OO-505.

Redeker KR, Wang N, Low JC, McMillan A, Tyler SC,Cicerone RJ. 2000. Emissions of methyl halides andmethane from rice paddies. Science 29O:966-969.

Riceweb. 2OO2. lnternational Rice Research Institute, Ma-nila, Philippines. http//:www.riceweb.org.

Russell PE Knipe FW. 1942. On the intermittent irrigationof ricefields to control malaria in South India. J MalariaInst India 4:321-34O.

Sandhu BS, Khera KL, Prihar SS, Singh B. 1980. Irriga-tion needs and rice yield on a sandy loam soil as af-fected by continuous and intermittent submergence. ln-dian J Agric Sci 50:494-496.

Sass RL, Fisher FM, Wang YB, Turner Fl Jund MF 1992.Methane emissions from rice fields: the effect of flood-water management. B i o g e o c he m Cy c I e s 6:249 -262.

Schimel J. 2000. Rice, microbes and methane. Nature4O3:375-377.

Self LS, de Datta SK. 1988. The impact of water man-agement practices in rice production on mosquito vectorpropagation. ln'. Vector-borne disease control in hu-mans through rice agroecosystem management Pro-ceedings of the workshop on research and trainingneeds in the field of integrated vector-borne diseaseconffol in riceland agroecosystems of developing coun-tries. March 9-14. 1987. International Rice ResearchInstitute in collaboration with the WHo/FAoTNEPPanel of Experts on Environmental Management forVector Control. Manila, Philippines: International RiceResearch Institute. p 67-83.

Service MW 1989. The importance of ecological studieson malaria vectors. Bull Soc Vector Ecol 14:26-38.

Sharma VP, Srivastava A, Nagpal BN. 1994. A study ofthe relationship of rice cultivation and annual parasiteincidence of malaria in India. Soc Sci Med 38:165-778.

Smalt FH. .1937.

Periodieke drooklegging van sahaw's terbestrijding van malaria. Meded Dienst Volkogezondh 4:258,299.

Surtees G. 1971. Control of mosquitoes breeding in rice-fields. "/ Trop Med Hyg 74:255-259.

Utzinger I,Tozan Y, Singer BH. 2001. Efficacy and cost-effectiveness of environmental management for malariacontrol. Trop Med Int Health 6:677-687.

Van der Hoek W, Sakthivadivel R, Renshaw R, Silver JB,Birley MH, Konradsen F 2000. Alternate wet/dry irri-gation in rice cultivation: a practical way to save waterand control malaria and Japanese encephalitis Colom-bo, Sri Lanka: International Water Management Insti-tute.

Wassmann R. Lantin RS. Neue HU.2001. Methane emis-sion.from major rice ecosystems Dordrecht, The Neth-erlands: Kluwver & Irri.

WHO [World Health Organization]. 1983. Integratedvec-tor control Technical Report Series No. 688. Geneva,Switzerland: World Health Organization.

WHO [World Health Organization]. 1998. Malaria Ge-neva, Switzerland: World Health Organization.

WHO [World Health Organization]. 1999. Report of theWHO informal consultation on schistosomiasis controlWHO/CDS/CPC/SIP/99.2. Geneva, Switzerland: WorldHealth Organization.