PR

OO

F

Acta Tropica xxx (2005) xxx–xxx

Effect of irrigated rice agriculture on Japanese encephalitis,including challenges and opportunities for integrated

vector management

3

4

5

Jennifer Keisera,∗, Michael F. Malteseb, Tobias E. Erlangera, Robert Bosc,Marcel Tannera, Burton H. Singerd, Jurg Utzingera

6

7

a Swiss Tropical Institute, P.O. Box, CH-4002 Basel, Switzerland8b St. Antony’s College, Oxford University, Oxford OX2 6JF, UK9

c Water, Sanitation and Health, World Health Organization, Avenue Appia 20, CH-1211 Geneva 27, Switzerland10d Office of Population Research, Princeton University, Princeton, NJ 08544, USA11

Received 29 October 2004; received in revised form 31 March 2005; accepted 5 April 2005

12

Abstract13

Japanese encephalitis (JE) is a disease caused by an arbovirus that is spread by marsh birds, amplified by pigs, and mainlytransmitted by the bite of infectedCulex tritaeniorhynchusmosquitoes. The estimated annual incidence and mortality rates are

fe yearsbetweenigated1963 to

illionludingtalluated dryfungi

egratedlations,

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

vector30

31

32

1 02 d

CO

RR

EC

TED

ACTROP 1666 1–18

30,000–50,000 and 10,000, respectively, and the estimated global burden to JE in 2002 was 709,000 disability-adjusted lilost. Here, we discuss the contextual determinants of JE, and systematically examine studies assessing the relationshipirrigated rice agriculture and clinical parameters of JE. Estimates of the sizes of the rural population and population in irrareas are presented, and trends of the rural population, the rice-irrigated area, and the rice production are analyzed from2003. We find that approximately 1.9 billion people currently live in rural JE-prone areas of the world. Among them 220 mpeople live in proximity to rice-irrigation schemes. In 2003, the total rice harvested area of all JE-endemic countries (excthe Russian Federation and Australia) was 1,345,000 km2. This is an increase of 22% over the past 40 years. Meanwhile, the torice production in these countries has risen from 226 millions of tonnes to 529 millions of tonnes (+134%). Finally, we evathe effect of different vector control interventions in rice fields, including environmental measures (i.e. alternate wet anirrigation (AWDI)), and biological control approaches (i.e. bacteria, nematodes, invertebrate predators, larvivorous fish,and other natural products). We conclude that in JE-endemic rural settings, where vaccination rates are often low, an intvector management approach with AWDI and the use of larvivorous fish as its main components can reduce vector popuand hence has the potential to reduce the transmission level and the burden of JE.© 2005 Published by Elsevier B.V.

Keywords: Japanese encephalitis; Geographical distribution; Global burden; Vector control; Rice agriculture; Irrigation; Integratedmanagement; Environmental control; Biological control

∗ Corresponding author. Tel.: +41 61 284 8218; fax: +41 61 284 8105.E-mail address:jennifer.keiser@unibas.ch (J. Keiser).

001-706X/$ – see front matter © 2005 Published by Elsevier B.V.oi:10.1016/j.actatropica.2005.04.012

UN

T P

RO

OF

8

2 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

1. Introduction1

Japanese encephalitis (JE) is a mosquito-borne viral2

disease. The JE virus is a member of the familyFla-3

viviridae. Clinically apparent infection takes place in4

one out of 200–300 infected patients (Pugachev et al.,5

2003). The disease is characterized by a wide range of6

presentations, as both the symptoms and the clinical7

course can differ broadly among patients. They range8

from mild flu-like symptoms to considerable neuro-9

logic symptoms, such as rigors, convulsions, polio-like10

flaccid paralysis, seizures or encephalomyelitis. Severe11

clinical cases are likely to have life-long neurological12

sequelae. Mostly children and young adults are affected13

(Solomon et al., 2000; Tsai, 2000; Ding et al., 2003;14

Halstead and Jacobson, 2003). The annual incidence15

and mortality estimates for JE are 30,000–50,000 and16

10,000, respectively (Solomon, 2004). However, there17

is considered to be severe under-reporting of JE and18

one study estimated the annual incidence at 175,000 per19

year (Tsai, 2000). JE outbreaks occur in cycles that may20

be linked to climatic patterns and the immune status of21

the populations. The great majority of cases and death22

occur in World Health Organization (WHO) regions of23

South-East Asia and the Western Pacific. In 2002, the24

estimated global burden of JE was 709,000 disability-25

adjusted life years (DALYs) lost (WHO, 2004). At26

present there are no established antiviral treatments27

against JE. Interferon alpha was the most promising28

drug in small open-label trials, but it failed to affect the29

o30

iol-31

o sive32

v ated33

i sia34

a and35

T ly re-36

d d37

J ns-38

m -East39

A ex-40

p , as41

w vel-42

o pro-43

d auses44

f a,45

1 ce46

o past47

50 years (Solomon et al., 2000). At present, the geo- 48

graphical distribution of JE ranges from Japan, mar-49

itime Siberia and the Republic of Korea in the North,50

to most parts of China and the Philippines in the East,51

Papua New Guinea in the South, and India and Nepal to52

the West (Broom et al., 2003). Recent outbreaks of JE 53

have been reported Southward in Australia, and west-54

ward in Pakistan (Solomon et al., 2000). 55

Currently, approximately 90% of the world’s rice 56

is produced in Asia (Consultative Group on Interna- 57

tional Agricultural Research et al., 1998). In most of 58

the countries, where JE outbreaks have been reported,59

rice is not only a staple food, but rice growing also is60

a major economic activity and key source of employ-61

ment and income generation (Consultative Group on 62

International Agricultural Research et al., 1998). Ef- 63

forts to further enhance the high annual rice production64

in these areas are essential to maintain food security.65

It has been estimated that in the next 25 years the de-66

mand for rice will rise by 65% in the Philippines, 51% 67

in Bangladesh, 45% in Viet Nam and 38% in Indonesia68

(http://www.biotech-info.net/riceexpert.html). Thai- 69

land, for example, is already in the planning stages70

of designing new rice-irrigation schemes for year-71

round irrigation (Consultative Group on International 72

Agricultural Research et al., 1998). Hence, there is con- 73

siderable concern in public health circles, as the in-74

tensification of rice production systems as well as the75

extension of the flooded surface area, particularly in76

semi-arid areas, contributes greatly to increased fre-77

q 78

the79

e JE;80

a ns 81

a rated82

v per83

i ex-84

t ole85

o Sec-86

o be-87

t s of88

J ing89

i in 90

J ions91

o e 92

c and93

r oun-94

t re-95

UN

CO

RR

EC

utcome in children with JE (Solomon et al., 2003).There has been a changing pattern in the epidem

gy of JE. On the one hand, primarily due to extenaccination campaigns, JE has been almost eliminn many economically advanced countries of East And South-East Asia (i.e. Japan, Republic of Koreaaiwan) and the burden of JE has been substantialuced in many other endemic countries (Halstead anacobson, 2003). On the other hand, intensified traission has been observed in other parts of Southsia and the Western Pacific, most likely due to anansion of irrigated agriculture and pig husbandryell as changing climatic factors. Water resource depment and management, in particular flooded riceuction systems, are considered among the chief c

or several JE outbreaks (Amerasinghe and Ariyasen991; Akiba et al., 2001). Conversely, the occurrenf the disease has changed considerably over the

ED

ACTROP 1666 1–1

uencies and intensities of JE outbreaks.The objectives of this paper were (i) to evaluate

ffect of irrigated rice agriculture on the burden ofnd (ii) to review different vector control interventiond discuss challenges and opportunities for integector management (IVM). The remainder of this pas structured as follows. First, we put forward contual determinants of JE, highlighting the important rf rice agroecosystems on JE vector populations.nd, we review studies assessing the relationship

ween irrigated rice growing and clinical parameterE. Third, we present estimates of the population liv

n proximity to irrigation and rice-irrigation schemesE-prone areas, stratified by relevant WHO sub-regf the world (WHO, 2004). Fourth, we quantify thhanges of the rice-irrigated area, rice production,ural population sizes over the past 40 years, in cries, where JE is currently endemic. Finally, we

CT

PR

OO

F

8

J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx 3

viewed studies that employed different environmental96

and biological control interventions to reduce larval JE97

vector populations, which have been implemented in98

irrigated rice production systems.99

2. Contextual determinants100

Fig. 1 depicts the contextual determinants of JE.101

The most important epidemiological features that gov-102

ern the transmission of JE include (i) environmental103

factors, i.e. agricultural practice, altitude, climate and104

the presence of pigs and marsh birds; (ii) human im-105

munization rates and vector control measures; and (iii)106

socioeconomic parameters.107

The principal vector of JE isCulex tritae-108

niorhynchus. Female specimens are infective 9–10109

days after having taken the viraemic blood meal,110

having undergone three gonotropic cycles (Gajanana111

et al., 1997). Other culicine mosquitoes that can trans-112

mit JE includeCx. bitaeniorhyncus,Cx. epidesmus,Cx.113

fuscocephala, Cx. gelidus, Cx. pseudovishnui, Cx. si-114

tiens,Cx. vishnuiandCx. whitmorei(Sehgal and Dutta,115

2003). In Australia,Cx. annulirostriswas found to be116

the major JE vector species (Hanna et al., 1996). In117

a recent study in Kerala, South India, JE was isolated118

fromMansonia indiana(Arunachalam et al., 2004).119

Although JE vectors are able to breed in ground wa-120

ter habitats, sunlit pools, roadside ditches, tidal marshes121

of low salinity, or man-made containers, one of their122

major preferred larval habitats are rice fields (Mogi, 123

1984; Sucharit et al., 1989). The ecology ofCulex 124

spp. in rice fields has been studied and reviewed in125

great detail (Suzuki, 1967; Lacey and Lacey, 1990). 126

Since very high densities of JE vector species were127

found consistently in rice fields, it was concluded128

that the impact of these man-made breeding sites is129

much more important than that of natural breeding130

places. For example, a significant increase in the abun-131

dance ofCx. tritaeniorhynchusand increased human-132

vector contacts have been noted following completion133

of the large rice-irrigation scheme in the Mahaweli134

project, Sri Lanka (Amerasinghe and Ariyasena, 1991;135

Amerasinghe, 1995). 136

JE vector abundance is closely related to agro-137

climatic features (Peiris et al., 1993; Phukan et al.,138

2004), most notably temperature and monthly rain-139

fall (Suroso, 1989; Solomon et al., 2000; Bi et al.,140

2003). In addition, potential JE vectors were rarely141

found at altitudes above 1200 m (Peiris et al., 1993). 142

However, the most important causative factors of JE143

is the management of paddy water, and the peak peri-144

ods of mosquito abundance are associated with cycles145

in local agricultural practices. In Thailand, the highest146

rminan

UN

CO

RR

E

Fig. 1. Contextual dete

ED

ACTROP 1666 1–1

ts of Japanese encephalitis.

T P

RO

OF

8

4 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

numbers of larvae and pupae of JE vectors were col-147

lected when the rice fields were ploughed with the water148

in the fields (also termed puddling). The vector popula-149

tion decreased after transplanting when the fields were150

flooded, and stayed low until harvesting (Somboon151

et al., 1989). In Malaysia, small plots in the rice fields,152

which are common before planting and contain veg-153

etation, were found to be conductive factors to facili-154

tate enhanced breeding of JE vectors; up to 40 pupae155

were collected per m2 (Heathcote, 1970). Furthermore,156

the height of the rice plants, water temperature, dis-157

solved oxygen, ammonia nitrogen and nitrate nitrogen158

strongly influence the abundance of immatures (Sunish159

and Reuben, 2001). The practice of paddy cultivation,160

proximity of houses to water bodies and suitable cli-161

matic factors were the most important environmental162

factors associated with several recent JE outbreaks in163

Northeast India (Phukan et al., 2004).164

The presence of pigs and marsh birds is crucial in165

the etiology of JE, as the virus is carried by birds and166

amplified by pigs (Broom et al., 2003). The latter are167

the most important natural host for transmission of JE168

to humans. Pigs have high and prolonged viraemias,169

are often common in endemic countries, and are gen-170

erally reared in open and unroofed pigpens, which171

are located near houses (Mishra et al., 1984; Solomon172

et al., 2000). In 2004, the world’s pig population was173

estimated at 951 million, of which 60% are found in174

Asia (www.faostat.fao.org). In this part of the world175

pig farming has increased considerably over the past176

1 on177

i178

s es are179

c180

e ith181

e has182

b cat-183

t 0184

T ian185

v in186

r ion187

r and188

5 ith189

h 56%,190

r191

ants192

c tion193

s ects194

of different environmental and biological control in-195

terventions in rice fields. Though JE vectors tend to196

bite and rest outdoors, self-protection behaviour, such197

as sleeping under insecticide-treated nets (ITNs), us-198

ing insect repellants, and insect-proofing homes and199

work places, might also assist in reducing JE trans-200

mission (WHO, 1997). For example, a population- 201

based case-control study in China, which evaluated202

the protective effect of ITNs against JE, showed that203

the risk of infection among children below 10 years204

was greatly reduced (Luo et al., 1994). On the other 205

hand, all 187 serologically confirmed JE cases in re-206

cent outbreaks in Northeast India reported that they207

had slept under a bed-net (Phukan et al., 2004). Fur- 208

thermore, application of deet-permethrin soap (“Mos-209

bar”) let to an 89–100% reduction in man-vector con-210

tact, including vectors transmitting JE (Mani et al., 211

1991b). 212

The third broad category of determinants that gov-213

ern JE transmission is people’s socioeconomic status.214

In Central China, more JE cases were observed among215

children living in poor quality houses, and whose par-216

ents had lower income. However, the sample size of the217

study was small, which might explain why the associa-218

tions were not statistically significant (Luo et al., 1995). 219

Religion, exposure to domestic animals, and household220

crowding were also described as risk factors associated221

with JE (Halstead and Jacobson, 2003). 222

3 223

an-224

a ans-225

m ct a226

s ork227

w e ir-228

r vid229

T ap-230

p ali-231

t 232

r a- 233

p in-234

c Chi-235

n e pa-236

p oned237

d are238

s 239

UN

CO

RR

EC

0 years, from 490 million pigs in 1994 to 574 millin 2004 (+17.1%) (www.faostat.fao.org). In Asia pigstocks increased. Humans, goats, cattle and horsonsidered dead-end hosts (Reuben et al., 1992). Forxample, in the Thanjavur district, India, an area wxtensive rice agriculture, a very low JE incidenceeen reported, which has been explained by a high

le to pig ratio (400:1) (Vijayarani and Gajanana, 200).he important role of birds was demonstrated in Indillages with or without herons in close proximity;ice-growing villages without herons, seroconversates in children aged 0–5 and 6–15 years were 0%%, respectively. In ecologically-similar villages werons, the corresponding rates were 50% andespectively (Mani et al., 1991a).

The second category of contextual determinomprises vaccination and transmission interruptrategies. Later in this paper, we review the eff

ED

ACTROP 1666 1–1

. Rice irrigation and JE incidence

As summarized in the previous section, the mgement of paddy water strongly influences the trission of JE. Hence, we were motivated to condu

ystematic literature review to identify published with an emphasis on the relationship between ric

igation and JE. We searched Biosis previews, Oechnologies, Medline and the Web of Sciencelying the following keywords: “Japanese enceph

is” and “water”, or “rice”, or “irrigation”, or “rice ir-igation and agriculture”, or “paddy”, or “field”. Pers published in English, French or German wereluded. We also considered manuscripts written inese, Japanese or Korean, if an abstract of thesers was accessible in English on the above-mentiatabases. The key findings of our literature reviewummarized here.

CT

PR

OO

F

8

J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx 5

Four studies in India analyzed the presence of rice240

irrigation and vector abundance in relation to the inci-241

dence of JE. In the Gorakhpur district in Uttar Pradesh,242

and the Mandya district of Karnataka, areas extensively243

developed for irrigated rice agriculture, the occurrence244

of JE was closely associated with high vector densi-245

ties, breeding in the fields or the canal system. The246

highest numbers of JE cases were observed shortly af-247

ter the mosquito densities peaked (Mishra et al., 1984;248

Kanojia et al., 2003). In addition, in the Mandya dis-249

trict, a high incidence of JE was found in extensively ir-250

rigated areas, while few cases occurred in villages with251

less irrigation or no irrigation systems (Geevarghese252

et al., 1994). In Assam, 78.6% of the JE cases occurred253

in families practicing rice cultivation (Phukan et al.,254

2004).255

Based on an epidemiological, serological and clin-256

ical study of 54 JE patients in the Northern Philip-257

pines, 41 cases (76%) were associated with irrigated258

rice fields and only one patient (2%) could not be linked259

to rice irrigation (Barzaga, 1989). In Taiwan, JE was260

monitored over 4 years and the highest numbers of261

cases were observed in the counties with the highest262

number of rice paddies (e.g. in Ilan county in 1969 a263

morbidity rate of 5.5 per 100,000 inhabitants). Fewer264

cases occurred in areas where dry farming had been265

adopted (e.g. in Yunlin county in 1969 a morbidity rate266

of 0.64 per 100,000 inhabitants) and no cases were re-267

ported in non-rice cultivated provinces (Okuno et al.,268

1975).269

ka,270

a han271

4 d out-272

b aths.273

T us-274

p high-275

e iga-276

t rted277

f ,278

2279

4280

k of281

J JE282

e CDC,283

2 is/284

risk-table.htm). The countries were grouped into differ-285

ent epidemiological sub-regions of the world, accord-286

ing to recent classifications of WHO, which is based287

primarily on child and adult mortality rates (WHO, 288

2004). The JE-endemic countries are located in seven289

of the 14 WHO sub-regions (Table 1). 290

Though sporadic JE cases have been reported291

from peri-urban areas, disease occurrence is largely292

restricted to rural settings (Self et al., 1973; Solomon 293

et al., 2000). Hence, inTable 1 we show data on 294

the rural population for the individual JE-endemic295

countries only (United Nations, 2004). For those 296

countries, where JE is endemic only in certain parts297

of the rural areas (e.g. in the Russian Federation JE298

outbreaks have been observed only in far Eastern299

maritime areas South of Chabarovsk (CDC, 2004;300

http://www.cdc.gov/ncidod/dvbid/jencephalitis/risk- 301

table.htm), we determined the rural endemic popula-302

tion of these areas consulting Encarta Encyclopedia303

(Microsoft Corporation, 2004). 304

We find that approximately 1.9 billion people cur-305

rently live in rural JE-prone areas of the world, the306

majority of them in China (766 million) and India (646307

million) (Tables 1 and 2). This figure is in line with a 308

recent “population at risk estimate”, quoting a popu-309

lation of 2 billion, 700 million of these children under310

the age of 15 years living in JE-endemic areas (Tsai, 311

2000). 312

As discussed above surface irrigation is an impor-313

tant risk factor in the epidemiology of JE. Conse-314

q ir-315

r ar-316

e rom317

t the318

r ir-319

r ar-320

e nal321

i 322

b n’s323

p rea,324

h area325

a rone326

a emic327

i esti-328

m op-329

u 330

t ies,331

r 332

UN

CO

RR

E

In 1985–1986 in the Mahaweli System H, Sri Lann epidemic of JE occurred, resulting in more t00 cases and 76 deaths. In 1987–1988 a seconreak took place with >760 cases and 138 dehe promotion of smallholder pig husbandry was sected to be responsible for these outbreaks. Thest number of cases occurred in areas with irr

ion and pig husbandry, while no cases were reporom non-irrigated areas with few pigs (Amerasinghe003).

. Population at risk

In order to estimate the current population at risE, we first compiled a list of all countries wherepidemics or sporadic cases have been reported (004;http://www.cdc.gov/ncidod/dvbid/jencephalit

ED

ACTROP 1666 1–1

uently, inTable 2we present the size of the totaligated and the rice-irrigated land in JE-endemicas, together with data on populations at risk f

hese areas and estimated DALYs, stratified byelevant WHO sub-regions. The numbers on theigated and the rice-irrigated areas in JE-proneas were calculated by multiplying the total natio

rrigated and rice-irrigated area (http://www.fao.org)y the “endemic fraction” (e.g. 17.7% of Pakistaopulation is estimated to live in the endemic aence we assume that 17.7% of both the totalnd the rice-irrigated areas are located in JE-preas). The sizes of the populations in JE-end

rrigated areas and JE-endemic rice areas wereated by multiplying the average national rural plation densities (United Nations, 2004) by the to-

al endemic area under irrigation and rice paddespectively.

UN

CO

RR

EC

TED

PR

OO

F

ACTROP 1666 1–18

6 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

Table 1Rural population/population in irrigated areas in JE-endemic countries stratified by relevant WHO sub-regions of the world (all numbers inthousand; n.d.: no data currently available)

The AmericasWHO sub-region 3

United States of America: Guam (150/n.d.) and Saipan (40/n.d.)

Eastern MediterraneanWHO sub-region 7

Pakistana (16,750/4,001)

EuropeWHO sub-region 10

Russian Federationb (614/n.d.)

South-East AsiaWHO sub-region 11

Indonesia (119,589/3033), Sri Lanka (15,057/1461), Thailand (42,796/4114)

WHO sub-region 12Bangladeshc (111,165/28,958), Bhutanc (2,065/17), Democratic People’s Republic of Korea (8,818/1,066), Indiad

(646,100/100,873), Myanmar (34,927/809), Nepale (12,500/329), Timor-Leste (719/51)

Western-PacificWHO sub-region 13

Australiaf (15/n.d.), Brunei Darussalam (85/<1), Japang (25,113/2376), Singapore (<100/n.d.)

WHO sub-region 14Cambodia (11,514/172), Chinah (766,757/24,796), Lao People’s Democratic Republic (4489/30), Malaysia (8814/99), PapuaNew Guinea (4958/n.d.), Philipines (31,182/1612), Republic of Korea (9,395/1081), Viet Nam (60,441/5460)

a Areas around Karachi and in the Lower Indus Valley, province of Sind (CDC, 2004,http://www.cdc.gov/ncidod/dvbid/jencephalitis/risk-table.htm); population estimates taken from (United Nations, 2004) and (Microsoft Corporation, 2004).

b Far Eastern maritime areas South of Chabarovsk (administrative district of Primorskij Kraij) (CDC, 2004,http://www.cdc.gov/ncidod/dvbid/jencephalitis/risk-table.htm); population estimates taken from (United Nations, 2004) and (Microsoft Corporation, 2004).

c Bangladesh and Bhutan are potential JE-endemic countries, but due to lack of data the situation has to be clarified (CDC, 2004,http://www.cdc.gov/ncidod/dvbid/jencephalitis/risk-table.htm).

d Reported cases from all states except Arunachal, Dadra, Daman, Diu, Gujarat, Himachal, Jammu, Kashmir, Lakshadweep, Meghalaya, NagarHaveli, Orissa, Punjab, Rajasthan and Sikkim (CDC, 2004,http://www.cdc.gov/ncidod/dvbid/jencephalitis/risk-table.htm); population estimatestaken from (United Nations, 2004) and (Microsoft Corporation, 2004).

e Hyperendemic in Southern Terai lowlands (BBIN, 2004).f Outbreaks on the Islands of Torres Strait and on mainland Australia at Cape York Peninsula (Hanna et al., 1996); at-risk to JE assumed for

all inhabitants at Cape York Peninsula; population estimates from (United Nations, 2004) and (Microsoft Corporation, 2004).g Rare-sporadic cases on all islands except Hokkaido; population estimates taken from (United Nations, 2004) and (Microsoft Corporation,

2004).h Cases in all provinces except Xizang (Tibet), Xinjiang, Qinghai (CDC, 2004,http://www.cdc.gov/ncidod/dvbid/jencephalitis/risk-table.htm);

population estimates taken from (United Nations, 2004) and (Microsoft Corporation, 2004).

According to our estimates 1,025,000–1,080,000333

km2 land is irrigated in JE-prone areas. We find that334

currently 180–220 million people are living in prox-335

imity to irrigation or rice-irrigation schemes in the336

JE-endemic regions, and thus are at risk of con-337

tracting the disease (Table 2). Irrigated agriculture is338

most pronounced in WHO sub-region 12. Accord-339

ing to our estimates 132–167 million people live in340

JE-endemic irrigated areas in this WHO sub-region341

(Table 2). 342

The estimated global burden of JE in 2002 was343

709,000 DALYs. WHO sub-regions 12 and 14 currently344

bear 84% of this burden (597,000 DALYs). The remain-345

ing 16% are thought to occur in WHO sub-regions 7

UNCORRECTED PROOF

AC

TR

J.Keise

reta

l./Acta

Tropica

xxx(2005)xxx–

xxx

Table 2Burden of JE and population at risk in endemic areas and living in close proximity to irrigation, stratified by relevant WHO sub-regions

WHOsub-region

DALYs lostdue to JE in2002a

Total area(km2)b

TotalJE-endemicareas (km2)c

Irrigated land inJE-endemic areasin 2002 (km2)d

Rice paddies inJE-endemic areasin 2003 (km2)d

Total populationin 2003 (×103)e

Rural endemicpopulation(×103)f

Population inJE-endemic irrigatedareas (×103)g

Population inJE-endemic riceareas (×103)g

3 n.d. 663 663 (100%) n.d. n.d. 1905 190 (100%) n.d. n.d.7 83,000 796,095 140,914 (17.7%) 31,507 (4.0%) 3912 (0.5%) 153,578 16,750 (10.9%) 4,001 (2.6%) 497 (0.3%)

10 n.d. 17,075,400 165,760 (1.0%) n.d. n.d. 143,246 614 (0.4%) n.d. n.d.11 29,000 2,483,300 2,483,300 (100%) 104,100 (4.2%) 233,888 (9.4%) 301,782 177,442 (58.8%) 8,609 (2.9%) 18,44812 277,000 4,437,432 3,401,672 (76.7%) 506,195 (11.4%) 521,384 (11.7%) 1,312,548 816,294 (62.2%) 132,103 (10.1%) 167,647 (12.8%)13 n.d. 8,125,410 330,670 (4.1%) 20,317 (0.25%)h 12,971 (0.16%)h 151,996 25,313 (16.6%) 2,376 (1.6%)h 1,517 (1.0%)h

14 320,000 11,539,430 6,902,490 (59.8%) 363,520 (3.2%)i 308,844 (2.7%) 1,560,475 897,550 (57.5%) 33,249 (2.1%)i 32,749 (2.1%)

Total 709,000 44,457,730 13,425,469 (30.1%) 1,025,639 (2.3%)h,i 1,080,999 (2.4%)h 3,623,815 1,934,154 (53.3%) 180,338 (4.9%)h,i 220,859 (6.1%)h

n.d.: no data.a WHO (2004).b Food and Agriculture Organization 2004,http://www.fao.org/waicent/portal/statisticsen.asp, Rome.c Estimated with aid of CDC, 2004 (http://www.cdc.gov/ncidod/dvbid/jencephalitis/risk-table.htm) andMicrosoft Corporation (2004).d Data for the whole country obtained from the Food and Agriculture Organization 2004 and multiplication by the endemic fraction.e United Nations (2004).f Microsoft Corporation (2004).g The size of the endemic irrigated population and endemic population in rice areas was estimated by multiplying the average national rural populationdensities (United Nations,

2004) by the total area under irrigation/of rice paddies in JE-endemic areas.h Omitting Singapore and Australia.i Omitting Papua New Guinea.

O

P

16661–18 7

T P

RO

OF

8

8 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

and 11, whereas no information is currently available346

for WHO sub-regions 3, 10 and 13.347

5. Trend of rice agriculture in endemic WHO348

sub-regions349

We compiled data for the relevant WHO sub-regions350

on the rice harvested area and rice production between351

1963 and 2003, which is available athttp://faostat.352

fao . org / faostat / form?collection = Production .Crops.353

Primary&Domain=Production&servlet=1&hasbulk=0354

&version=ext&language=EN. No distinction could355

be made between rain-fed rice and irrigated rice and356

no information was available on the intensity of the357

irrigated rice production (single, double or triple358

cropping).359

No data was obtainable for the JE-endemic coun-360

tries of WHO sub-regions 3 (Guam and Saipan) and361

Singapore. We excluded data on the rice harvested area362

and rice production in the Russian Federation, because363

the current size of the JE-endemic area there are small364

compared to the rest of the JE-endemic countries. We365

furthermore excluded Australia as outbreaks have so366

far only be reported Torres Strait Islands and on main-367

land Australia at Cape York Peninsula (Hanna et al.,368

1996).369

As mentioned above rice agriculture accounted for370

up to 1,080,000 km2 in all JE-prone areas in 2003. In371

the same year the total rice harvested area of these372

c was373

1 er374

t rice-375

i in-376

v area377

i378

2 ted379

a apan,380

f lved381

(382

w over383

t sub-384

r ficant385

p 12,386

a ppar-387

e rew388

u ars.389

M d in390

WHO sub-region 13. And finally, in WHO sub-region391

14, a substantial expansion of the rice-irrigated area has392

occurred between 1963 and 1973, but has decreased393

thereafter. 394

The total rice production has risen from 226 mil-395

lions of tonnes in 1963 to 529 millions of tonnes in396

2003 (+134%). The most significant increases of rice397

production occurred in WHO sub-regions 7, 11, 12 and398

14, as shown inFig. 2. On the other hand, in WHO sub-399

region 13 (mainly Japan) a reduction from 16.6 to 9.7400

millions of tonnes (−43%) has taken place. 401

We also depict inFig. 2the change of the rural pop- 402

ulation of the five most relevant WHO sub-regions,403

which overall increased from 1325 million inhabitants404

in 1963 to 2197 million inhabitants in 2003 (+66%). In405

the past decade a decrease in the rural population has406

been observed in WHO sub-regions 11 and 14. 407

6. Intervention strategies in rice fields 408

Our aim was to systematically review the litera-409

ture to identify published work on biological control410

strategies againstCx. tritaeniorhynchuslarvae in rice 411

fields. We did not include studies on the application of412

synthetic larvicides and insecticides againstCx. tritae- 413

niorhynchusin rice fields, as the use of chemical control414

was found to be of no operational value due to the short415

activity of these products, high costs and resistance de-416

velopment (Wada, 1988). We searched the same elec-417

t e 418

f 419

t t 420

a 421

l , 422

o r- 423

v s”,424

o 425

6 426

6 427

es428

i s, in-429

c of ir-430

r and431

d can432

b ch-433

UN

CO

RR

EC

ountries, including the non-JE-endemic regions,,345,000 million km2. This is an increase of 22% ov

he past 40 years. In some of these countries therrigated area has almost doubled over the periodestigated. For example, in Pakistan the irrigatedncreased from 12,861 km2 in 1963 to 22,100 km2 in003 (+72%). Marked declines in the rice-irrigarea have been observed in other countries. In J

or example, the rice-irrigated area has been ha32,760 km2 in 1963 to 16,650 km2 in 2003). InFig. 2e present the changes of the rice-irrigated area

he past 4 decades for 5 of the 7 relevant WHOegions, where JE has been reported to be a signiublic health problem. In WHO sub-regions 11 andsubstantial increase in the rice-irrigated area is ant. In WHO sub-region 7 (Pakistan) the rice area gntil 1993, but remained stable over the past 10 yeeanwhile, the rice-irrigated area has decrease

ED

ACTROP 1666 1–1

ronic databases as mentioned in Section3and used thollowing keywords: “Japanese encephalitis” or “Culexritaeniorhynchus” in combination with “alternate wend dry irrigation”, or “Azolla”, or “Bacillus”, or “bio-

ogical control”, or “bacteria”, or “control”, or “fungi”r “intermittent irrigation”, or “invertebrates”, or “laivorous fish”, or “natural products”, or “nematoder “predator”, or “water management”.

.1. Environmental control strategies

.1.1. Alternate wet and dry irrigation (AWDI)Traditionally, rice fields are flooded, which provid

deal breeding places for several mosquito specieluding those that transmit JE. The managementigation water that leads to the alternate wettingrying of fields, including the canals and ditches,e active or passive. An important feature of this te

UN

CO

RR

EC

TED

PR

OO

F

ACTROP 1666 1–18

J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx 9

Fig. 2. Changes of rice growing area, rice production and rural population at risk of JE in 5 WHO sub-regions between 1963 and 2003.

TED

PR

OO

F

ACTROP 1666 1–18

10 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

nique is that the soil can dry out, which in turn curtails434

the life cycle development of the mosquito from lar-435

vae and pupae to adult. In order to achieve a significant436

reduction of mosquito larvae, AWDI (also termed inter-437

mittent irrigation) has to be applied during the entire438

cropping season and should cover all rice fields that439

are connected by irrigation canals over a large area.440

This method is particularly feasible in places where441

control of the water supply and drainage is possible,442

hence where land and climatic conditions are suitable443

(Mogi, 1988). Growing water shortages in many areas444

create an incentive to better control irrigation water445

management. AWDI is one such strategy. The poten-446

tial of AWDI is, however, limited in areas where there447

is a threat of insufficient resources to re-flood the fields448

and where farmers perceive a risk of reduced yields by449

letting their fields dry out.450

The effects of AWDI on the abundance ofAnophe-451

les, which are the vector species of malaria, rice yield,452

water consumption and methane emission have been453

reviewed recently (van der Hoek et al., 2001; Keiser454

et al., 2002). In brief, this method had been introduced455

in Asia about 300 years ago, primarily to obtain higher456

rice yields. The first study of AWDI as a potential tool457

to reduce mosquito vectors was conducted in Bulgaria458

on mosquitoes of theAn. maculipenniscomplex in the459

1920s (Keiser et al., 2002). AWDI was compulsory in460

Portugal in the 1930s, when malaria was still a problem461

there.462

We found four studies analyzing the effect of AWDI463

o um-464

m465

m ply-466

i DI467

o h468

d ined469

i ere470

o od471

h ted.472

6473

6474

c-475

t -476

l re-477

s es478

h inst479 Tabl

e3

Alte

rnat

ew

etan

ddr

yirr

igat

ion

(AW

DI)

agai

nstCulextrita

eniorhyn

chusla

rvae

inric

efie

lds

Stu

dysi

te;r

efer

ence

Spe

cific

ityof

AW

DI

Out

com

e

C.tritaeniorhyn

chus

imm

atur

esC.tritaeniorhyn

chus

adul

tsR

ice

yiel

dW

ater

cons

umpt

ion

Hen

an,C

hina

(Lu-B

aulin

,198

8)Ir

rigat

ion

inte

rval

:5da

ysD

ecre

ase:

81–9

1%D

ecre

ase:

55–7

0%In

crea

se:1

3%D

ecre

ase:

50%

Tam

ilN

adu,

Indi

a(R

ajen

dran

etal

.,19

95)Ir

rigat

ion

inte

rval

:3–5

days

Dec

reas

e:75

–88%

Not

know

nIn

crea

se:4

%N

otkn

own

Kin

ryu,

Japa

n(M

ogi,

1993

)Ir

rigat

ion

inte

rval

:Sev

eral

days

Dec

reas

eN

otkn

own

Not

know

nN

otkn

own

Tsu

City

,Jap

an(Ta

kagi

etal

.,19

95)M

idse

ason

dryi

ngD

ecre

ase

offo

urth

inst

ars:

14.3

–48.

2%N

otkn

own

Not

know

nN

otkn

own

UN

CO

RR

EC

n densities of JE vectors and the key findings are sarized inTable 3. Overall,Cx. tritaeniorhynchusim-atures were reduced by 14–91% in rice fields ap

ng AWDI. One study investigated the effect of AWn theCx. tritaeniorhynchusadult population, whicecreased by 55–70%. The crop yield was exam

n two trials and increases between 4 and 13% wbserved in AWDI rice fields. The effect this methas on the incidence of JE remains to be investiga

.2. Biological control strategies

.2.1. BacteriaThe larvicidal activity of the spore forming ba

eria Bacillus sphaericusandB. thuringiensis israeensis have been discovered in 1965 and 1976,pectively (Mittal, 2003). Subsequently, these biocidave been evaluated in various formulations aga

CO

RR

EC

TED

PR

OO

F

ACTROP 1666 1–18

J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx 11

Table 4Application ofBacillusspp. againstCulex tritaeniorhynchuslarvae in rice fields

Study site; period (if known);reference

Intervention strategy Outcome

India, 1983–1984 (Kramer, 1984) Application ofBacillus sphaericusandBacillus thuringiensisH-14 inrice fields

91–99% reduction with doses of 0.5–1.5 kg/haCulextritaeniorhynchusandAnopheles subpictus, howeveractivity did not subsist beyond a few days

India, 1982 (Balaraman et al., 1983) Application ofBacillus thuringiensisH-14 in rice fields

100% reduction with doses of 27× 105 spores/ml, but firstand second instars reappeared 3 days after application

India (Sundararaj and Reuben, 1991) Application of microgel dropletformulation ofBacillus sphaericusinrice fields

44–79% reduction of early instar and 82–100% reduction oflate instar ofCulex tritaeniorhynchus,Cx. vishnuiandCx.pseudovishnuifor at least 5 weeks applying a dose of4.3 kg/ha

Korea (Rhee et al., 1983) Application ofBacillus thuringiensisisraelensisH-14 in rice field

95–98% reduction ofCulex tritaeniorhynchus. The residualeffects lasted only 24 h at doses of 0.8–1.2 l/ha ofBacillusthuringiensis israelensisH-14, SAN 402 I suspensionconcentrate containing 600 IU/mg

Rourkela city, India, June 1993 toOctober 1994 (Yadav et al., 1997)

Application ofBacillus sphaericus(strain B-101, serotype H5a, 5b) inrice field

Application ofB. sphaericus, when sprayed at 1 g/m2

significantly reduced larval and pupal counts (P< 0.0001) inrice fields. Duration of effect could not be determined asfields received periodicallyBacillus sphaericustreatedwastewater

mosquito vectors worldwide (Lacey and Lacey, 1990).480

B. thuringiensisproduces four key insecticidal pro-481

teins, whileB. sphaericusproduces a single binary482

toxin (Federici et al., 2003). These protein toxins bind483

to cells of the gastric caecum and posterior midgut of484

the mosquitoes causing intoxication, which eventually485

leads to death (Mittal, 2003). The advantages of these486

bacterial insecticides in terms of efficacy, specificity487

and environmental safety are well documented (Zahiri488

et al., 2004). A major drawback of these products,489

however, is their high cost of production, the labor-490

intensive delivery, as well as first reports of resistance491

(Sundararaj and Reuben, 1991; Federici et al., 2003).492

The effect of the individual biolarvicides depends on493

water temperature, pH, aquatic vegetation, the formula-494

tion applied, type of habitat, and the target mosquitoes495

(Mittal, 2003).B. sphaericus, for example, is known to496

exhibit a high activity againstCulexmosquitoes, while497

certainAedesspecies are not affected (Mittal, 2003;498

Zahiri et al., 2004). Hence, no general conclusion499

can be drawn about whetherB. sphaericusand B.500

thuringiensis israelensisformulations are suitable for501

JE vector control in rice fields. The results on the502

application of these bacterial formulations in rice503

fields againstCx. tritaeniorhynchusare summarized504

in Table 4. In Tamil Nadu, India, the application505

of 4.3 kg/ha of a microgel droplet formulation of506

B. sphaericus1593 M resulted in a 44–79% reduc-507

tion of early instar and 82–100% reduction of late508

instar culicinae larvae (C. fuscanus, C. pseudovish- 509

nui and C. tritaeniorhynchus) for at least 5 weeks 510

(Sundararaj and Reuben, 1991). Similarly, up to 511

95–98% ofC. tritaeniorhynchuslarvae were reduced 512

in three other field sites evaluatingB. sphaericusor 513

B. thuringiensisformulations (Balaraman et al., 1983; 514

Rhee et al., 1983; Kramer, 1984). However, the larvi- 515

cidal activity did not persist in these rice fields beyond516

a couple of days; in the Republic of Korea the residual517

effect ofB. thuringiensisH-14 was found to last only 518

24 h (Rhee et al., 1983). 519

6.2.2. Nematodes 520

Romanomermis culicivoraxis probably the best- 521

studied nematode parasite of mosquitoes. The prepar-522

asitic nemas are applied to the fields, where they locate523

a host, penetrate the cuticle and develop within the524

mosquito larvae (Lacey and Lacey, 1990). Although 525

this method has not been widely studied, we identified526

three publications assessing the effect of nematodes on527

JE vectors, which are summarized inTable 5. In Tai- 528

wan, application of 7000 nematodes per m2 rice field 529

yielded a reduction of 11–18%C. tritaeniorhynchus 530

(Lacey and Lacey, 1990). In two studies in China the 531

nematodeR. yunanensishas been distributed in rice 532

UN

EC

TED

PR

OO

F

8

12 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

Table 5Application of nematodes againstCulex tritaeniorhynchuslarvae in rice fields

Study site; period (if known); reference Intervention strategy Outcome

China, 1986–1995 (Peng et al., 1998) Predation efficacy ofRomamomermisyunanensisin rice fields

2000–4000 nematodes/m2 yielded parasitism rates of52.2–96.7% forCulex tritaeniorhynchus

China (Song and Peng, 1996) Predation efficacy ofRomamomermisyunanensisand a secondRomamomermisspecies in rice fields

1000–3000 nematodes/m2 resulted in parasitism rates of60.8–95.5% ofCulex tritaeniorhynchusandCx.quinquefasciatusandAnophelesspp.

Taiwan, 1974 (Lacey and Lacey, 1990;Mitchell et al., 1974)

Predation efficacy ofRomamomermisculicivorax in rice fields

Larvae reduction of 11–18% ofCulex tritaeniorhynchus

fields (2000–4000 nematodes/m2), which resulted in533

parasitism rates of 52.2–96.7% and 60.8–95.5%, re-534

spectively, ofC. tritaeniorhynchuslarvae (Song and535

Peng, 1996; Peng et al., 1998).536

6.2.3. Invertebrate predators537

Invertebrate predators, i.e. Coleoptera, Hemiptera538

or Odonata, though less common than the use of fish,539

are also known to substantially reduce mosquito lar-540

val populations in rice fields (Lacey and Lacey, 1990).541

However, they are highly sensitive to temperature, pres-542

ence of vertebrates, growth of rice and chemical pollu-543

tants (Lacey and Lacey, 1990). In India, the presence of544

notonectids was negatively associated with larval abun-545

dance ofCx. pseudovishnui,Cx. tritaeniorhynchusand546

Cx. vishnui(Sunish and Reuben, 2002).547

6.2.4. Larvivorous fish548

The use of larvivorous fish to reduce mosquitoes has549

a history of over 100 years (Lacey and Lacey, 1990).550

The mosquitofish,Gambusia affinis,is the most widely551

used predator. Other fish species includeTilapia spp.,552

Poecilia reticulataor Cyprinidae (Lacey and Lacey,553

1990). A detailed summary of the most common preda-554

cious fish is given in (Lacey and Lacey, 1990).555

After stocking rice fields with 1–10 natural preda-556

tor fishs per m2, larval populations ofCx. tritae-557

niorhynchuswere reduced by 55.2–87.8% (Table 6).558

Larvivorous fish cannot be applied in rice fields, where559

i ted560

t by561

t ides562

o563

a iting564

m re-565

s fish566

( -567

duction of exotic predators such asGambusiamight 568

displace the native fish populations to reduce their nat-569

ural value, as being observed in Japanese rice fields570

(Wada, 1988). Therefore, the compatibility of the cho-571

sen fish with local fauna and flora is of high importance572

(Lacey and Lacey, 1990). 573

6.2.5. Fungi 574

Fungi that have been studied extensively for their575

potential as biological mosquito control agents include576

Coelomomycesspp. andLagenidium giganteum. The 577

former have been investigated how they impact the578

development of JE vectors in China. Field observa-579

tions there showed a strong effect of the fungusCoelo- 580

momyces indicaon Cx.tritaeniorhynchus, as infected 581

larvae were unable to develop into adults (Liu and Hsu, 582

1982). However, fungi have not been applied for bio-583

logical control of JE vectors on a large scale so far, as584

practical problems, for example their production, have585

yet to be solved (Lacey and Lacey, 1990). 586

6.2.6. Other natural products 587

Natural products might also have high potential for588

reducing the proliferation of culicine mosquitoes in589

rice fields. Our literature search on the control of JE590

vectors in rice fields yielded in-depth investigation of591

only two natural products. In Tamil Nadu, India ap-592

plication of the floating water fernAzolla microphylla 593

greatly reduced immature mosquito populations. How-594

e 595

d - 596

c tion,597

l rol598

a 599

d or600

s 601

i t a602

UN

CO

RRrregular irrigation is practiced. It should also be no

hat predator populations are strongly influencedemperature, rice growth, vegetation, use of pesticr chemical pollutants (Lacey and Lacey, 1990). Inddition, recent research has shown that ovipososquitoes may move to other breeding sites in

ponse to the stocking of rice fields with predatoryAngelon and Petranka, 2002). Furthermore, the intro

ACTROP 1666 1–1

ver, the infestation of the rice field withAzollawasifficult to achieve, 80% coverage byAzollawas acomplished only 13–14 days after rice transplantaimiting its wider use as a biological mosquito contgent (Rajendran and Reuben, 1991).

Neem cake powder, made from freshly harvestetored neem kernels, from the neem tree (Azadirachtandica) rich in the active principle azadirachtin. A

UN

CO

RR

EC

TED

PR

OO

F

ACTROP 1666 1–18

J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx 13

Table 6Application of fish predators againstCulex tritaeniorhynchuslarvae in rice fields

Study site; period (if known); reference Intervention strategy Outcome

Bulkyo, Bosong-gun, Chollanamdo,Korea 1995–1996 (Lee, 1998)

Presence ofMisgurnus mizolepisinrice fields

Coefficients of correlation betweenMisgurnusmizolepisand abundance of mosquito larvae showednegative correlations inAnopheles sinensis(−0.66) andCulex tritaeniorhynchus(−0.47)

Suwon, 1992–1993 (Kim et al., 1994) Moroco oxycephalusandMisgurnusanguillicaudatusin rice fields

4 fish/m2 gave a reduction of 55.2% of immatures ofAnopheles sinensis, Culex pipiens pallensandCulextritaeniorhynchuscompared to control field after 8weeks

South Delhi, May–June 1980 (Mathuret al., 1981)

Presence ofGambusia affinisin ricefields

10 fish/m2 gave a mean reduction of 63.9% ofimmatures ofCulex tritaeniorhynchusin 4 weeks

China (Liao et al., 1991) Predation efficacy ofCtenopharyngdon idella

78 days after treatment with 1.9 fish/m2 a reduction of80%Culex tritaeniorhynchuslarvae was observed

India, June to October 1991 (Prasadet al., 1993)

Predation efficacy ofGambusiaaffinis in rice fields

87.8% mosquito (6 Anopheline and 4 Culicine) larvalcontrol in well submerged rice fields, no effect ofGambusia affinisin rice fields applying AWDI

Quanzhou county, China, 1988–1989(Wu et al., 1991)

Predation efficacy ofCypriniuscarpio,Ctenopharyngdon idellaandTilopia spp.

Population size of larvae and adultCulextritaeniorhynchusdecreased

South Korea, 1979 and 1981 (Yu et al.,1982)

Predation efficacy ofAphyocyprischinensis

Fish stocking of 1.5/m2 in Wondang-Ni rice paddyresulted in mosquito larval reduction of 98.9% in thethird week after fish introduction againstAnophelessinensis, Culex pipiens, Culex crientalisandCulextritaeniorhynchus

Jindo Island, Chollanam-do Province,South Korea, 1989 (Yu and Kim,1993)

Predation efficacy ofAphyocyprischinensisin combination withBacillus thuringiensis(H-14)

Aphyocypris chinensis(1.5 fish/m2) achieved areduction of 60.8-77.0% against both,AnophelessinensisandCulex tritaeniorhynchusin the first 2month. In the third monthBacillus thuringiensis(H-14)treatment was made, which yielded a satisfactory degreeof control maintained above 93.1% for 2 weeks

dosage of 500 kg/ha it yielded an 81% reduction of late603

instar culicine larvae and a 43% reduction in the pupal604

production (Rao et al., 1992). Stable product neem cake605

coated urea and the combination of neem cake coated606

urea and water management practices (i.e. AWDI) led607

to a 70–95% reduction of culicine immatures. The ben-608

efits were also found to be cost-effective as neem treat-609

ment gave higher values for mean number of produc-610

tive tillers, plant height and mean number of grains per611

panicle (Rao et al., 1995).612

7. Discussion and conclusion613

Water resource development and management, in614

particular the construction and operation of small and615

large dams and irrigation schemes, occurred at an enor-616

mous pace over the past 50 years (Gujja and Perrin,617

1999). The effects of these water projects and accom-618

panying ecological transformations are manifold, and619

inherently difficult to assess and quantify. Negative ef-620

fects include increased frequencies and transmission621

dynamics of water-based (e.g. schistosomiasis) and622

water-related vector-borne diseases, including malaria623

(Keiser et al., in press), lymphatic filariasis (Erlanger 624

et al., in press) and JE. On the other hand, such water625

projects are key for hydroelectric power production and626

food security; hence they can stimulate social and eco-627

nomic development. In this review, we have focused628

on irrigation schemes, with particular emphasis on rice629

growing. Currently, there is a paucity on high quality630

data on the effects of large and small dams on JE. 631

Over the last four decades rice production has ex-632

panded considerably in most countries where JE is cur-633

rently endemic. This growth will most likely continue634

for food security reasons. We find that in 2003 approx-635

imately 180–220 million people live near irrigation636

schemes in JE-prone countries, hence have a poten-637

T P

RO

OF

8

14 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

tial risk of contracting JE. These figures are unavoid-638

ably subject to a level of uncertainty. The possibility639

that we have overvalued the population at risk cannot640

be ruled out, which is justified on four grounds. First,641

pigs, which are the amplifying hosts, are not present642

in all rural irrigated sites. Second, the disease affects643

mainly children under the age of 15 years. This segment644

of the rural population in Asia accounts for approx-645

imately one-third of the total rural Asian population646

(Tsai, 2000). Third, it is presently not possible to pre-647

cisely map the population of irrigated areas in partially648

JE-endemic countries. Forth, no data are available on649

vaccination coverage in rural irrigated areas. However,650

it is also conceivable that we might have underesti-651

mated the actual population at risk of JE due to irriga-652

tion, as we have applied a rather conservative proxy,653

namely the national rural population density. Irrigated654

areas, however, are often characterized by a large pop-655

ulation density as they attract a great number of people656

(Keiser et al., in press). In addition, due to the predicted657

expansion of irrigation schemes, the rural population in658

JE-endemic areas living in proximity to irrigated rice659

agro ecosystems will further rise in the years to come.660

Unfortunately and in contrast to malaria (Keiser661

et al., in press), very few studies are presently available662

that investigated the effect of an introduction or an ex-663

pansion of rice-irrigation schemes on the incidence of664

JE. The few studies reviewed here clearly link JE inci-665

dence with rice agriculture. There are important gaps666

in our understanding of the effects of different types of667

w mis-668

s ies669

t d JE670

i ater671

p ent672

s urce673

d674

ice675

f sia,676

a cam-677

p lin-678

i ama679

o .g.680

J by681

s ieve682

1 one683

b quent684

b685

side effects occur in 10–30% and allergic reactions oc-686

cur in 0.6% of vaccinated adults (Jones, 2004). Fur- 687

thermore, clinically significant neurological adverse688

events, such as a meningoencephalitis, have been re-689

ported. In addition, these vaccines are expensive and690

production can hardly keep up with demand (Jones, 691

2004). In China a live attenuated vaccine (SA14-14-692

2) was developed in 1988 and the efficacy of a single693

dose was estimated to be 99.3% (Bista et al., 2001). 694

However, its use in other affected countries has been695

restricted by regulatory concerns over manufacturing696

and control (Solomon et al., 2003). 697

There is ongoing research to develop improved JE698

vaccines. These efforts received a boost in December699

2003, when the Bill and Melinda Gates Foundation700

donated US$ 27 million to the Children’s Vaccine Pro-701

gram at the Program for Appropriate Technology in702

Health for this purpose (UNO, 2003). For example, a 703

newly developed live attenuated vaccine (ChimeriVax-704

JE) that uses yellow fever 17D as a live vector for the705

envelope genes of SA14-14-2 virus has been tested in706

phase 2 clinical trials and appears to be well tolerated707

(Monath et al., 2003). It is currently undergoing a 2- 708

year clinical study in Australia, to assess the duration709

of immunity and to gain further knowledge on safety710

and immunogenicity (Jones, 2004). 711

It is important to recognize that strategies other than712

vaccination may play an important role for prevention713

and control of JE, particular in rural areas, where vac-714

cination coverage are sometimes low or where there715

i re-716

c ., 717

2 te-718

g ses719

a body720

l be-721

h nets722

( on723

( of724

a they725

h 726

s nted727

a ts to728

r ized729

a 730

ental731

a lds:732

B is 733

UN

CO

RR

EC

ater-related projects on the frequency and transion dynamics of JE. In our view longitudinal studhat track both a change in vector abundance annfectivity rates in humans, for example at large wrojects in the framework of health impact assessmtudies, covering the entire course of a water resoevelopment project, are warranted.

Clearly, vaccination is the current strategy of choor prevention and control of JE outbreaks in And extensive government-supported vaccinationaigns will remain the mainstay of control. Forma

nactivated vaccines, based on wild type Nakayr Beijing-1 strains grown in adult mouse brain (eE-Vax), are licensed for immunization against JEeveral Asian countries. However, in order to ach00% seroconversion, 2–3 primary doses andooster dose after a year are required, with subseoosts every 3–4 years (Pugachev et al., 2003). Minor

ED

ACTROP 1666 1–1

s no history of immunization against JE at all, asently documented for Northeast India (Phukan et al004). Vaccination of the far rural population is straically difficult and costly, in particular as three dore required to achieve adequate neutralizing anti

evels. As demonstrated in China self-protectionaviour, such as sleeping under insecticide-treatedITNs) can significantly reduce the risk of infectiLuo et al., 1994). On the other hand, all JE patientsrecent outbreak in Northeast India reported thatad slept under a bed-net (Phukan et al., 2004). Con-equently, irrigation schemes should be implemend maintained in a way that adverse health effecice growers and residents of the area are minimnd social and economic development improved.

We evaluated and discussed several environmnd biological vector control measures in rice fieacillus sphaericusand B. thuringiensis israelens

CT

PR

OO

F

8

J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx 15

were found to greatly reduce JE mosquito larvae. How-734

ever, it is currently not economical to apply these bac-735

terial toxins in rice fields, as they are costly, labor-736

intensive and often only have a short duration of activ-737

ity. Few studies are available on invertebrates, fungi,738

nematodes or the neem cake powder in rice fields,739

which allows assessing and quantification of their po-740

tential as JE mosquito control agents. Reviewing the741

literature has shown that in settings where the irriga-742

tion water in the rice fields can be managed, AWDI743

has considerable potential to reduce JE vector densi-744

ties. A similar conclusion can be made for the use of745

larvivorous fish. An integrated vector management ap-746

proach with AWDI and the use of larvivorous fish as747

its main components can reduce JE vector populations,748

and hence has the potential to reduce the transmission749

level and the burden of JE. It should be emphasized750

that these intervention strategies must be tailored care-751

fully to a specific setting, which renders it difficult to752

generalize reported experiences and results.753

Consequently, there is a pressing need to implement754

and monitor the performance of well-designed stud-755

ies to further strengthen our understanding of the con-756

textual determinants of environmental and biological757

control methods on vector abundance and clinical out-758

comes of JE in different ecological, epidemiological759

and socio-cultural settings.760

Acknowledgements761

en762

o is of763

t re-764

s par-765

t J.766

K ger767

( wiss768

N ort.769

W is770

m771

R772

A Ku-773

e en-774

26,775

776

Amerasinghe, F.P., 1995. Mosquito vector ecology: a case study from777

Sri Lanka and some thoughts on research issues. In: Tropical778

Diseases, Society and Environment, Proceedings from a WHO-779

TDR/SAREC Research Seminar, SAREC Documentation: Con-780

ference Reports 1995, pp. 135–156. 781

Amerasinghe, F.P., 2003. Irrigation and mosquito-borne diseases. J.782

Parasitol. 89, 14–22. 783

Amerasinghe, F.P., Ariyasena, T.G., 1991. Survey of adult784

mosquitoes (Diptera: Culicidae) during irrigation development785

in the Mahaweli Project Sri Lanka. J. Med. Entomol. 28, 387–786

393. 787

Angelon, K.A., Petranka, J.W., 2002. Chemicals of predatory788

mosquitofish (Gambusia affinis) influence selection of oviposi- 789

tion site by Culex mosquitoes. J. Chem. Ecol. 28, 797–806. 790

Arunachalam, N., Samuel, P.P., Hiriyan, J., Thenmozhi, V., Gajanana,791

A., 2004. Japanese encephalitis in Kerala, South India: can Man-792

sonia (Diptera: Culicidae) play a supplemental role in transmis-793

sion? J. Med. Entomol. 41, 456–461. 794

Balaraman, K., Balasubramanian, M., Jambulingam, P., 1983. Field795

trial of Bacillus thuringiensisH-14 (VCRC B-17) against Culex 796

and Anopheles larvae. Ind. J. Med. Res. 77, 38–43. 797

Barzaga, N.G., 1972–1985. A review of Japanese encephalitis cases798

in the Philippines Southeast Asian. J. Trop. Med. Public Health799

20, 587–592. 800

Bi, P., Tong, S., Donald, K., Parton, K.A., Ni, J., 2003. Climate801

variability and transmission of Japanese encephalitis in Eastern802

China. Vector Borne Zoonotic Dis. 3, 111–115. 803

Bista, M.B., Banerjee, M.K., Shin, S.H., Tandan, J.B., Kim, M.H.,804

Sohn, Y.M., Ohrr, H.C., Tang, J.L., Halstead, S.B., 2001. Efficacy805

of single-dose SA 14-14-2 vaccine against Japanese encephalitis:806

a case control study. Lancet 358, 791–795. 807

Broom, A.K., Smith, D.W., Hall, R.A., Johansen, C.A., Macken-808

zie, J.S., 2003. Arbovirus infections. In: Cook, G., Zumla, A.809

(Eds.), Manson’s Tropical Diseases, 21st ed. Saunders, London,810

pp. 725–764. 811

C hni-812

port813

w of814

aly.815

D 03.816

nese817

81,818

819

E nner,820

ment821

opu-822

823

F J.J.,824

Biol.825

826

G i, T.F.,827

litis in828

di-829

Med.830

831

G tudies832

ndya

UN

CO

RR

E

This work was part of a project entitled “Burdf water-related vector-borne diseases: an analys

he fraction attributable to components of waterources development and management”, whichially funded by the World Health Organization.eiser (Project no. PMPDB-10622) and J. Utzin

Project no. PPOOB-102883) are grateful to the Sational Science Foundation for financial suppe thank Dr. B. Kay for carefully reviewing thanuscript.

eferences

kiba, T., Osaka, K., Tang, S., Nakayama, M., Yamamoto, A.,rane, I., Okabe, N., Umenai, T., 1997. Analysis of Japanescephalitis epidemic in Western Nepal in. Epidemiol. Infect. 181–88.

ED

ACTROP 1666 1–1

onsultative Group on International Agricultural Research, Teccal Advisory Committee, and CGIAR Secretariat, 1998. Reof the Fifth External Programme and Management RevieInternational Rice Research Institute (IRRI), FAO, Rome, It

ing, D., Kilgore, P.E., Clemens, J.D., Wei, L., Zhi-Yi, X., 20Cost-effectiveness of routine immunization to control Japaencephalitis in Shanghai. Chin. Bull. World Health Organ.334–342.

rlanger, T.E., Keiser, J., Castro, M., Bos, R., Singer, B.H., TaM., Utzinger, J., in press. Effect of water resource developand management on lymphatic filariasis, and estimates of plations at risk. Am. J. Trop. Med. Hyg.

ederici, B.A., Park, H.W., Bideshi, D.K., Wirth, M.C., Johnson,2003. Recombinant bacteria for mosquito control. J. Exp.206, 3877–3885.

ajanana, A., Rajendran, R., Samuel, P.P., Thenmozhi, V., TsaKimura-Kuroda, J., Reuben, R., 1997. Japanese encephaSouth Arcot district, Tamil Nadu India: a three-year longitunal study of vector abundance and infection frequency. J.Entomol. 34, 651–659.

eevarghese, G., Mishra, A.C., Jacob, P.G., Bhat, H.R., 1994. Son the mosquito vectors of Japanese encephalitis virus in Ma

T P

RO

OF

8

16 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

District, Karnataka India. Southeast Asian J. Trop. Med. Public833

Health 25, 378–382.834

Gujja, B., Perrin, M., 1999. A place for dams in the 21st Century.835

Discussion Paper, World Wildlife Fund, p. 108.836

Halstead, S.B., Jacobson, J., 2003. Japanese encephalitis. Adv. Virus837

Res. 61, 103–138.838

Hanna, J.N., Ritchie, S.A., Phillips, D.A., Shield, J., Bailey, M.C.,839

Mackenzie, J.S., Poidinger, M., McCall, B.J., Mills, P.J., 1995.840

An outbreak of Japanese encephalitis in the Torres Strait Aus-841

tralia. Med. J. Aust. 165, 256–260.842

Heathcote, O.H., 1970. Japanese encephalitis in Sarawak: studies on843

juvenile mosquito populations. Trans. R. Soc. Trop. Med. Hyg.844

64, 483–488.845

Jones, T., 2004. A chimeric live attenuated vaccine against Japanese846

encephalitis. Expert Rev. Vaccines 3, 243–248.847

Kanojia, P.C., Shetty, P.S., Geevarghese, G., 2003. A long-term study848

on vector abundance & seasonal prevalence in relation to the849

occurrence of Japanese encephalitis in Gorakhpur district Uttar850

Pradesh. Indian J. Med. Res. 117, 104–110.851

Keiser, J., Castro, M., Maltese, M., Bos, R., Tanner, M., Singer, B.H.,852

Utzinger, J., in press. The effect of irrigation and large dams on853

the burden of malaria on global and regional scale. Am. J. Trop.854

Med. Hyg.855

Keiser, J., Utzinger, J., Singer, B.H., 2002. The potential of inter-856

mittent irrigation for increasing rice yields, lowering water con-857

sumption, reducing methane emissions, and controlling malaria858

in African rice fields. J. Am. Mosq. Control Assoc. 18, 329–340.859

Kim, H.C., Kim, M.S., Yu, H.S., 1994. Biological control of vector860

mosquitoes by the use of fish predatorsMoroco oxycephalusand861

Misgurnus anguillicaudatusin the laboratory and semi-field rice862

paddy. Korean J. Entomol. 24, 269–284.863

Kramer, V., 1984. Evaluation ofBacillus sphaericus& B. thuringien-864

sisH-14 for mosquito control in rice fields. Indian J. Med. Res.865

80, 642–648.866

Lacey, L.A., Lacey, C.M., 1990. The medical importance of rice-867

ical868

869

L onal870

rice871

872

L eed-873

heng874

875

L a876

g877

878

L l of879

ease880

ent.881

the882

883

L mic884

k of885

sian J.886

887

L Y.,888

ated889

with pyrethroid insecticide and vaccination against Japanese en-890

cephalitis. Trans. R. Soc. Trop. Med. Hyg. 88, 632–634. 891

Mani, T.R., Rao, C.V., Rajendran, R., Devaputra, M., Prasanna, Y.,892

Hanumaiah, Gajanana, A., Reuben, R., 1991a. Surveillance for893

Japanese encephalitis in villages near Madurai, Tamil Nadu, In-894

dia. Trans. R. Soc. Trop. Med. Hyg. 85, 287–291. 895

Mani, T.R., Reuben, R., Akiyama, J., 1991b. Field efficacy of “Mos-896

bar” mosquito repellent soap against vectors ofbancroftian filar- 897

iasisand Japanese encephalitis in Southern India. J. Am. Mosq.898

Control Assoc. 7, 565–568. 899

Mathur, K.K., Rahman, S.J., Wattal, B.L., 1981. Integration of lar-900

vivorous fish and temephos for the control ofCulex tritae- 901

niorhynchusbreeding. J. Commun. Dis. 13, 58–63. 902

Microsoft Corporation, 2004. Encarta Encyclopedia, Redmond,903

USA. 904

Mishra, A.C., Jacob, P.G., Ramanujam, S., Bhat, H.R., Pavri, K.M.,905

1983. Mosquito vectors of Japanese encephalitis epidemic in906

Mandya district (India). Indian J. Med. Res. 80, 377–389. 907

Mitchell, C.J., Chen, P.S., Chapman, H.C., 1974. Exploratory tri-908

als utilizing a mermithid nematode as a control agent for Culex909

mosquitos in Taiwan. Taiwan Yi Xue Hui Za Zhi 73, 241–254. 910

Mittal, P.K., 2003. Biolarvicides in vector control: challenges and911

prospects. J. Vector Borne Dis. 40, 20–32. 912

Mogi, M., 1984. Mosquito problems and their solution in relation to913

paddy rice production. Protect. Ecol. 7, 219–240. 914

Mogi, M., 1988. Water management in rice cultivation and its rela-915

tion to mosquito production in Japan. In: Vector Borne Disease916

Control in Humans Through Rice Agroecosystem Management.917

International Rice Research Institute in collaboration with the918

WHO/FAO/UNEP Panel of Experts, pp. 101–109. 919

Mogi, M., 1993. Effect of intermittent irrigation on mosquitoes920

(Diptera: Culicidae) and larvivorous predators in rice fields. J.921

Med. Entomol. 30, 309–319. 922

Monath, T.P., Guirakhoo, F., Nichols, R., Yoksan, S., Schrader, R.,923

Murphy, C., Blum, P., Woodward, S., McCarthy, K., Mathis, D.,924

ccine925

clini-926

ose927

nacti-928

213–929

930

O 975.931

iwan)932

f case933

934

P L.P.,935

.A.,936

nka:937

atic938

939

P , Q.,940

941

ina.942

943

P halitis944

ublic945

946

UN

CO

RR

EC

land mosquitoes and their control using alternatives to cheminsecticides. J. Am. Mosq. Control Assoc. 2 (Suppl.), 1–93.

ee, D.K., 1998. Effect of two rice culture methods on the seasoccurrence of mosquito larvae and other aquatic animals infields of Southwestern Korea. J. Vector Ecol. 23, 161–170.

iao, S., Xu, B.Z., Chen, C.Y., Liang, Z.P., Zhang, H.F., 1991. Bring grass carp against mosquitoes in rice field. Zhongguo Ji SChong Xue Yu Ji Sheng Chong Bing Za Zhi 9, 219–222.

iu, S.L., Hsu, Y.C., 1982. Effect of the fungusCoelomomyces indicon the viability ofCulex tritaeniorhynchuslarvae. Kun ChonXue Bao 25, 409–412.

u-Baulin, 1988. Environmental management for the controricefield-breeding mosquitoes in China. In: Vector Borne DisControl in Humans Through Rice Agroecosystem ManagemInternational Rice Research Institute in collaboration withWHO/FAO/UNEP Panel of Experts, pp. 111–121.

uo, D., Ying, H., Yao, R., Song, J., Wang, Z., 1995. Socio-econostatus and micro-environmental factors in relation to the risJapanese encephalitis: a case-control study. Southeast ATrop. Med. Public Health 26, 276–279.

uo, D., Zhang, K., Song, J., Yao, R., Huo, H., Liu, B., Li,Wang, Z., 1994. The protective effect of bed nets impregn

ED

ACTROP 1666 1–1

Johnson, C., Bedford, P., 2003. Chimeric live, attenuated vaagainst Japanese encephalitis (ChimeriVax-JE): phase 2cal trials for safety and immunogenicity, effect of vaccine dand schedule, and memory response to challenge with ivated Japanese encephalitis antigen. J. Infect. Dis. 188, 11230.

kuno, T., Tseng, P.T., Hsu, S.T., Huang, C.T., Kuo, C.C., 1Japanese encephalitis surveillance in China (Province of Taduring 1968–1971. I. Geographical and seasonal features ooutbreaks. Jpn. J. Med. Sci. Biol. 28, 235–253.

eiris, J.S., Amerasinghe, F.P., Arunagiri, C.K., Perera,Karunaratne, S.H., Ratnayake, C.B., Kulatilaka, TAbeysinghe, M.R., 1993. Japanese encephalitis in Sri Lacomparison of vector and virus ecology in different agro-climareas. Trans. R. Soc. Trop. Med. Hyg. 87, 541–548.

eng, Y., Song, J., Tian, G., Xue, Q., Ge, F., Yang, J., Shi1998. Field evaluations ofRomamomermis yunanensis(Nema-toda: Mermithidae) for control of Culicinae mosquitoes in ChFundam. Appl. Nematol. 21, 227–232.

hukan, A.C., Borah, P.K., Mahanta, J., 2004. Japanese encepin Assam Northeast India. Southeast Asian. J. Trop. Med. PHealth 35, 618–622.

CT

PR

OO

F

8

J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx 17

Prasad, H., Prasad, R.N., Haq, S., 1993. Control of mosquito breeding947

throughGambusia affinisin rice fields. Indian J. Malariol. 30,948

57–65.949

Pugachev, K.V., Guirakhoo, F., Trent, D.W., Monath, T.P., 2003. Tra-950

ditional and novel approaches to flavivirus vaccines. Int. J. Para-951

sitol. 33, 567–582.952

Rajendran, R., Reuben, R., 1991. Evaluation of the water fernAzolla953

microphylla for mosquito population management in the rice-954

land agro-ecosystem of South India. Med. Vet. Entomol. 5,955

299–310.956

Rajendran, R., Reuben, R., Purushothaman, S., Veerapatran, R.,957

1995. Prospects and problems of intermittent irrigation for con-958

trol of vector breeding in rice fields in Southern India. Ann. Trop.959

Med. Parasitol. 89, 541–549.960

Rao, D.R., Reuben, R., Nagasampagi, B.A., 1995. Development of961

combined use of neem (Azadirachta indica) and water manage-962

ment for the control of culicine mosquitoes in rice fields. Med.963

Vet. Entomol. 9, 25–33.964

Rao, D.R., Reuben, R., Venugopal, M.S., Nagasampagi, B.A.,965

Schmutterer, H., 1992. Evaluation of neem,Azadirachta indica,966

with and without water management, for the control of culicine967

mosquito larvae in rice-fields. Med. Vet. Entomol. 6, 318–324.968

Reuben, R., Thenmozhi, V., Samuel, P.P., Gajanana, A., Mani, T.R.,969

1992. Mosquito blood feeding patterns as a factor in the epidemi-970

ology of Japanese encephalitis in Southern India. Am. J. Trop.971

Med. Hyg. 46, 654–663.972

Rhee, H.I., Shim, J.C., Kim, C.L., Lee, W.J., 1983. Small scale973

field trial withBacillus thuringiensis israelensisH-14 for control974

of the vector mosquito (Culex tritaeniorhynchus) larvae in rice975

fields. Korean J. Entomol. 13, 39–46.976

Sehgal, A., Dutta, A.K., 2003. Changing perspectives in Japanese977

encephalitis in India. Trop. Doct. 33, 131–134.978

Self, L.S., Shin, H.K., Kim, K.H., Lee, K.W., Chow, C.Y., Hong,979

H.K., 1973. Ecological studies onCulex tritaeniorhynchusas a980

vector of Japanese encephalitis. Bull. World Health Organ. 49,981

982

S 351,983

984

S ghn,985

Neu-986

987

S M.,988

.,989

e en-990

trial.991

992

S uwan-993

n the994

Mai995

alth996

997

S ies of998

ding999

1000

S anese1001

tors.1002

1003

Sundararaj, R., Reuben, R., 1991. Evaluation of a microgel droplet1004

formulation ofBacillus sphaericus1593 M (Biocide-S) for con- 1005

trol of mosquito larvae in rice fields in Southern India. J. Am.1006

Mosq. Control Assoc. 7, 556–559. 1007

Sunish, I.P., Reuben, R., 2001. Factors influencing the abundance1008

of Japanese encephalitis vectors in ricefields in India-I. Abiotic1009

Med. Vet. Entomol. 15, 381–392. 1010

Sunish, I.P., Reuben, R., 2002. Factors influencing the abundance1011

of Japanese encephalitis vectors in ricefields in India-II Biotic.1012

Med. Vet. Entomol. 16, 1–9. 1013

Suroso, T., 1989. Studies on Japanese encephalitis vectors in In-1014

donesia. Southeast Asian J. Trop. Med. Public Health 20, 627–1015

628. 1016

Suzuki, T., 1967. Bibliography on Culex tritaeniorhynchus in Japan1017

1945–1966. Jpn. Environ. Sanit Cent. 1018

Takagi, M., Sugiyama, A., Maruyama, K., 1995. Effect of rice cultur-1019

ing practices on seasonal occurrence ofCulex tritaeniorhynchus 1020

(Diptera: Culicidae) immatures in three different types of rice-1021

growing areas in Central Japan. J. Med. Entomol. 32, 112–1022

118. 1023

Tsai, T.F., 2000. New initiatives for the control of Japanese encephali-1024

tis by vaccination: minutes of a WHO/CVI meeting, Bangkok,1025

Thailand, 13–15 October 1998. Vaccine 18 (Suppl. 2), 1–25. 1026

United Nations, 2004. World Urbanization Prospects: The 2003 Re-1027

visions, Population Division Department of Economics and So-1028

cial Affair of the United Nations, New York. 1029

UNO, 2003. Gates foundation grant to help fight Japanese Encephali-1030

tis. United Nations Wire Service, Washington,http://www. 1031

unwire.org/UNWire/20031209/44911099.asp. 1032

van der Hoek, W., Sakthivadivel, R., Silver, J.B., Konradsen, F., 2001.1033

Alternate wet/dry irrigation in rice cultivation: a practical way1034

to save water and control malaria and Japanese encephalitis?1035

International Water Management Institute; Colombo, Sri Lanka.1036

Research Report 47. 1037

Vijayarani, H., Gajanana, A., 2000. Low rate of Japanese encephalitis1038

an1039

111,1040

1041

W in rice1042

orne1043

Man-1044

ation1045

1046

W als1047

eva,1048

ol 1049

1050

W Re-1051

1052

W he1053

h in1054

h 22,1055

1056

Y acil-1057

riasis1058

ontrol1059

1060

UN

CO

RR

E

41–47.olomon, T., 2004. Flavivirus encephalitis. N. Engl. J. Med.

370–378.olomon, T., Dung, N.M., Kneen, R., Gainsborough, M., Vau

D.W., Khanh, V.T., 2000. Japanese encephalitis. J. Neurol.rosurg. Psychiatry 68, 405–415.

olomon, T., Dung, N.M., Wills, B., Kneen, R., Gainsborough,Diet, T.V., Thuy, T.T., Loan, H.T., Khanh, V.C., Vaughn, D.WWhite, N.J., Farrar, J.J., 2003. Interferon alfa-2a in Japanescephalitis: a randomised double-blind placebo-controlledLancet 361, 821–826.

omboon, P., Choochote, W., Khamboonruang, C., Keha, P., Sphanit, P., Sukontasan, K., Chaivong, P., 1989. Studies oJapanese encephalitis vectors in Amphoe Muang, ChiangNorthern Thailand. Southeast Asian J. Trop. Med. Public He20, 9–17.

ong, J., Peng, Y., 1996. Field trials of combined use of two specmermithid nematodes to control Anopheles and Culex breein China. Indian J. Malariol. 33, 161–165.

ucharit, S., Surathin, K., Shrestha, S.R., 1989. Vectors of Japencephalitis virus (JEV): species complexes of the vecSoutheast Asian J. Trop. Med. Public Health 20, 611–621.

ED

ACTROP 1666 1–1

infection in rural children in Thanjavur district (Tamil Nadu),area with extensive paddy cultivation. Indian J. Med. Res.212–214.

ada, Y., 1988. Strategies for control of Japanese encephalitisproduction systems in developing countries. In: Vector BDisease Control in Humans Through Rice Agroecosystemagement. International Rice Research Institute in collaborwith the WHO/FAO/UNEP Panel of Experts, pp. 153–160.

HO, 1997. Vector Control—Methods for Use by Individuand Communities. World Health Organization, WHO, Genhttp://www.who.int/docstore/watersanitationhealth/vectcontr/ch08.htm.

HO, 2004. The World Health Report 2004, The World Healthport 2004, Geneva, World Health Organisation.

u, N., Liao, G.H., Li, D.F., Luo, Y.L., Zhong, G.M., 1991. Tadvantages of mosquito biocontrol by stocking edible fisrice paddies. Southeast Asian J. Trop. Med. Public Healt436–442.

adav, R.S., Sharma, V.P., Upadhyay, A.K., 1997. Field trial of Blus sphaericus strain B-101 (serotype H5a, 5b) against filaand Japanese encephalitis vectors in India. J. Am. Mosq. CAssoc. 13, 158–163.

T P

RO

OF

18 J. Keiser et al. / Acta Tropica xxx (2005) xxx–xxx

Yu, H.S., Kim, H.C., 1993. Integrated control of encephalitis vector1061

(Culex tritaeniorhynchus) with native fishes (Aplocheilus latipes1062

andAphyocypris chinensis) andBacillus thuringiensis(H-14)1063

in marshes in Jindo Island of Korea. Korean J. Entomol. 23,1064

221–230.1065

Yu, H.S., Lee, D.K., Lee, W.J., 1982. Mosquito control by the re-1066

lease of fish predatorAphyocypris chinensisin natural mosquito

breeding habitats of rice paddies and stream seepage in South1067

Korea. Korean J. Entomol. 12, 61–68. 1068

Zahiri, N.S., Federici, B.A., Mulla, M.S., 2004. Laboratory and1069

simulated field evaluation of a new recombinant ofBacillus 1070

thuringiensisssp. israelensis andBacillus sphaericusagainst 1071

Culex mosquito larvae (Diptera: Culicidae). J. Med. Entomol.1072

41, 423–429. 1073

U

NC

OR

RE

C

E

D

ACTROP 1666 1–18