Spatiotemporal distribution of insecticide resistance in Anopheles culicifacies and Anopheles subpictus in Sri Lanka

Transcription

1 Transactions of the Royal Society of Tropical Medicine and Hygiene (2005) 99, Spatiotemporal distribution of insecticide resistance in Anopheles culicifacies and Anopheles subpictus in Sri Lanka L.A. Kelly-Hope a, A.M.G.M. Yapabandara b, M.B. Wickramasinghe b, M.D.B. Perera b, S.H.P.P. Karunaratne c, W.P. Fernando b, R.R. Abeyasinghe b, R.R.M.L.R. Siyambalagoda b, P.R.J. Herath b, G.N.L. Galappaththy b, J. Hemingway a, a Vector Research Group, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA, UK b Ministry of Health, Anti-Malarial Campaign, P.O. Box 1472, Colombo 5, Sri Lanka c Department of Zoology, University of Peradeniya, Peradeniya, 20400, Sri Lanka Received 20 September 2004; received in revised form 26 April 2005; accepted 6 May 2005 KEYWORDS Malaria; Anopheles culicifacies; Anopheles subpictus; Insecticide resistance; Sri Lanka; Geographical information systems Summary The malaria situation in Sri Lanka worsened during the 1990s with the emergence and spread of resistance to the drugs and insecticides used for control. Chloroquine resistance has increased rapidly over this period, but adverse changes in malaria transmission are more closely associated with insecticide use rather than drug resistance. Insecticide susceptibility tests were routinely carried out in key anopheline vectors across the country for more than a decade. These sentinel data were combined with data collected by other research programmes and used to map the spatial and temporal trends of insecticide resistance in the main vectors, Anopheles culicifacies and A. subpictus, and to examine the relationship between insecticide resistance, changes in national spraying regimens and malaria prevalence. Both species had widespread resistance to malathion, the insecticide of choice in the early 1990s. Both species were initially susceptible to the organophosphate and pyrethroid insecticides used operationally from 1993, but some resistance has now been selected. The levels of malathion and fenitrothion resistance in A. subpictus were higher in some ecological regions than others, which may be related to the distribution of sibling species, agricultural pesticide exposure and/or environmental factors. The study highlights that the emergence and spread of insecticide resistance is a constant threat and that active surveillance systems are vital in identifying key vectors and evidence of resistance Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved. Corresponding author. Tel.: ; fax: address: hemingway@liverpool.ac.uk (J. Hemingway) /$ see front matter 2005 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved. doi: /j.trstmh

2 752 L.A. Kelly-Hope et al. 1. Introduction Malaria is a major public health problem in Sri Lanka (Briet et al., 2003, 2005; Roll Back Malaria, 2005). Approximately half the population live in malarious areas and control is dependent upon antimalarial chemotherapy and indoor residual insecticide spraying for vector control. The situation has worsened over the past two decades as successful interventions have been threatened by the emergence and spread both of drug and insecticide resistance. Malaria cases have increased four-fold from 1980, to confirmed cases in 2000 (Annual Parasite Index, 20.7%). Plasmodium falciparum prevalence has gradually increased over many years and now represents approximately 25% of cases, with P. vivax making up the remainder. Anopheles culicifacies is the main vector of malaria, and A. subpictus is a significant secondary vector in Sri Lanka (Amerasinghe et al., 1992, 1999; Carter, 1930; Carter and Jacocks, 1929), although natural plasmodium infections occur in 11 other anopheline species (Amerasinghe and Ariyasena, 1991; Amerasinghe and Indrajith, 1994; Amerasinghe et al., 1991, 1992, 1999; Herath et al., 1983; Mendis et al., 1990, 1992; Ramasamy et al., 1992; Yapabandara, 1997; Yapabandara and Curtis, 2004). Resistance to chloroquine, the first-line antimalarial of choice in Sri Lanka, was first recorded in 1984 in the Matale District (Ratnapala et al., 1984). This resistance has now spread to most districts, with parasite resistance rates ranging from 30% to 65%. Despite this resistance, chloroquine remains the first-line treatment in Sri Lanka (Galappaththy, 2002). Historically, malaria control in Sri Lanka has focused on controlling the vector with effective indoor residual insecticide spraying, which targets endophagic and/or endophilic adults. The national Malaria Eradication Program used DDT in all malarious areas from 1946, but this was replaced owing to the development of widespread vector resistance in 1976 (Wickramasinghe, 1981). Malathion was the insecticide of choice from 1977, and resistance to malathion in A. culicifacies was first detected in 1982 (Herath et al., 1987). Over the next decade, malathion resistance spread throughout the country, which, together with the coincident escalation of malaria, prompted the national Anti- Malaria Campaign (AMC) to revise its programme according to the WHO Global Malaria Control Strategy (WHO, 1993). Malaria control activities were devolved to the provinces in 1989, and in 1994 the pyrethroid lambdacyhalothrin and the organophosphate fenitrothion were first introduced and incorporated into rotational indoor residual spray regimens. The use of permethrin-impregnated bed nets also became more common, particularly in the North-East and North-Central Provinces in the late 1990s. AMC staff carried out routine insecticide susceptibility testing on a wide range of insecticides and Anopheles spp., and comprehensive sentinel data are available from Additional entomological projects were carried out by staff at Peradeniya University. The extent of this data collection is unique, but the data have yet to be collated into an accessible format that can be used by those outside the AMC. Here we map the spatial and temporal trends of insecticide resistance in A. culicifacies and A. subpictus throughout Sri Lanka using these data. A secondary aim was to examine the relationship between insecticide resistance, insecticide regimens, drug resistance and malaria incidence in Sri Lanka. 2. Methods 2.1. Mapping trends of insecticide resistance Annual insecticide susceptibility testing data were collected for the malaria-endemic regions of the country as part of AMC activities between 1991 and 2003 and for a project funded by the National Science Foundation between 1999 and These data were mapped using geographical information system (GIS) software (ArcGIS 8.2; Environmental Systems Research Institute, Redlands, CA, USA). Mosquito collection sites were georeferenced using longitude and latitude coordinates obtained from the National Imagery and Mapping Agency database ( Levels of insecticide resistance were calculated for A. culicifacies and A. subpictus based on the percentage mortality rates following exposure to WHO discriminating dosages of insecticideimpregnated papers in a standard WHO test kit for 1 h (except for fenitrothion, where the WHO recommend a 2 h exposure period), followed by a 24 h recovery period. Insecticides tested during included: the organophosphates malathion (5%) and fenitrothion (1%), the pyrethroids lambdacyhalothrin (0.1%), deltamethrin (0.025% and 0.05%), permethrin (0.25% and 0.75%), cyfluthrin (0.15%) and cypermethrin 0.1%, the carbamates propoxur (0.1%) and bendiocarb (0.1%), the pseudopyrethroid etofenprox (0.1%), and the organochlorine DDT (4%); where two discriminating concentrations are quoted, these reflect the recommended increases in the WHO discriminating dosage during the monitoring period.

3 Insecticide resistance in malaria vectors in Sri Lanka 753 Susceptibility tests were carried out in situ on samples of wild-caught, blood-fed female mosquitoes collected in cattle-baited net traps and huts during the night. For data collected between 1991 and 1996, insecticide resistance was tested in sample sizes 40, as no control groups were tested during this period, while for data from 1997, sample sizes that had 20 in both the exposure and control group were examined. Sites where small numbers of mosquitoes were collected were also mapped to identify areas of low vector abundance. In addition, overall percentage mortality rates of A. culicifacies and A. subpictus were examined in coastal (within 10 km) and inland areas, as well as in different geographical regions based on zonal maps of rainfall, temperature, elevation, paddy field density and population density. Maps were imported from the Sri Lankan Department of Census and Statistics (2004) and NOAA/NGDC ( Internet sources. Maps were digitised on screen in ArcGIS 8.2, and each variable was divided into two distinct regions based on the zones provided. For each mosquito species, mortality rates in each region were pooled, and the overall means were compared using the Mann Whitney U test with Bonferroni correction for multiple comparison. Resistance data were examined only for the insecticides that were commonly used between 1991 and These included malathion, fenitrothion, lambdacyhalothrin and permethrin Population case studies Three examples are presented as population case studies, which examined the relationship between insecticide resistance, insecticide treatment regimens, malaria incidence and chloroquine resistance over time. Case study locations involved data sets with more than 5 years of mortality data between 1991 and 2003 for one insecticide, tested against either A. culicifacies or A. subpictus. Insecticide resistance data were compared with information on insecticide treatment regimens and the number of mosquitoes caught at each location. Until 1999, AMC entomological teams recorded and tested all Anopheles mosquitoes caught at each collection site, thus providing basic information on vector abundance. Malaria incidence data from passive surveillance carried out between 1991 and 2002 were available from the AMC offices and based on blood slide positivity (%). Information on the level of chloroquine resistance by district was obtained from Galappaththy (2002). In addition, mortality rates of other Anopheles spp. at each location were compared with those of A. culicifacies or A. subpictus to determine whether the vector species had similar patterns of resistance to the non-vectors and whether the latter could be used as surrogates for testing the vectors when their abundance was low. 3. Results 3.1. Mapping trends of insecticide resistance Mosquitoes were collected from 79 sites in Sri Lanka between 1991 and 2003 (Figure 1). Resistance testing was carried out in most regions of the country, however the location and number of collection sites varied annually, with no sites in the northern districts because of civil war. National data for A. subpictus in 1993 as well as for A. culicifacies and A. subpictus in 1998 and 1999 were also unavailable for analysis. Malathion and fenitrothion were the most frequently tested insecticides for both mosquito species. Resistance levels varied between locations and over time (Figure 1). Overall, levels of malathion resistance were high in all regions (mortality 50%), except for A. subpictus in most coastal areas (mortality 75%). High levels of resistance to DDT (mortality 50%) occurred in both species. From 1995, low levels of resistance (mortality 75%) to lambdacyhalothrin, permethrin, deltamethrin, cyfluthrin, cypermethrin, propoxur and etofenprox were detected. However, higher levels of resistance in A. culicifacies to fenitrothion (in 2000 and 2001) and permethrin (2001), and in A. subpictus to fenitrothion (1997), permethrin (1995 and 2001), lambdacyhalothrin (2001) and cyfluthrin (2003) were evident in some localities (Figure 1). Collection sites and regional mortality rates for A. culicifacies and A. subpictus are shown in Table 1. For A. culicifacies, minimal differences in resistance were found between regions, however tests with permethrin were limited due to the small number of collection sites. For A. subpictus, higher levels of resistance to malathion (P < 0.001) and fenitrothion occurred in inland regions (compared with coastal), to malathion where annual average temperatures were less than 30 C, and to fenitrothion in areas 100 m a.s.l. (compared with <100 m). Some resistance was also evident in high paddy field density and sparsely populated areas (compared with ha paddy fields and >100 inhabitants/km 2 ).

4 754 L.A. Kelly-Hope et al. Figure 1 Locations and levels of resistance in Anopheles culicifacies and A. subpictus between 1991 and 2003.

5 Insecticide resistance in malaria vectors in Sri Lanka 755 Figure 1 (Continued ) Mapping trends of insecticide resistance Population case study 1 The Rambukkana Health Area in the northern region of the Kegalle District is close (4 6 km) to the borders of the Kurunegala and Kandy Districts (Figure 2). Malathion susceptibility tests were carried out in A. culicifacies in , and There was a significant decrease in mosquito mortality between 1993 and 1996 (Figure 2). The numbers of A. culicifacies collected for testing during this period were 4703 in 1992, 40 in 1993, 29 in 1994, 52 in 1995, 1031 in 1996 and 34 in Anopheles aconitus, A. annularis, A. barbirostris, A. jamesi, A. pallidus, A. tessellatus, A. vagus and A. varuna showed high mortalities ( 80%) with malathion. For malaria control, malathion was replaced in the Kegalle District with deltamethrin in 2002, in Kandy with fenitrothion in 2002, and in Kurunegala with lambdacyhalothrin in 1994 (used until 2000), etofenprox in 1999 (until 2001) and deltamethrin in 2002 (Figure 2). Malaria data from two health areas in this region were examined together owing to their close proximity and similar prevalence patterns. The average proportion of blood slides testing positive for P. vivax was 10.7% and for P. falciparum was 1.9% in 1991, which decreased to 2.0% and 0.3%, respectively, by 2002 (Figure 2). Overall, P. falciparum prevalence remained low and constant ( 2%), whilst P. vivax fluctuated interannually, with peaks recorded in 1993 (23.2%) and 1996 (9.3%). The greatest decrease in P. vivax prevalence occurred between 1993 and In 1999, 42.0% of P. falciparum-positive patients (RI late 42.0%, RI early 21.4%, RII 4.8%, RIII 31.0%) had parasites that were resistant to chloroquine in Kurunegala (n = 100). No data were available for Kegalle Population case study 2 The Galewela Health Area lies in the western region of the Matale District, close (5 9 km) to the borders of the Kurunegala and Anuradhapura Districts (Figure 2). Malathion susceptibility tests were carried out in A. culicifacies in and , with no significant difference in mortality found between years. The numbers of A. culicifacies decreased during this period, with 514 recorded in 1991, 337 in 1993, 190 in 1994, 0 in 1995, 198 in 1996 and 48 in Anopheles aconitus, A. jamesi, A. maculates, A. subpictus and A. varuna had high mortalities ( 80%) with malathion, whilst A. vagus was resistant. Malathion antimalarial spraying in the Matale District ceased in early 1997 and was replaced in 1998 with deltamethrin (until 2003) and lambdacyhalothrin (until 1999), whilst in Anuradhapura, fenitrothion ( ), lambdacyhalothrin ( ) and cyfluthrin (2000) were used as alternatives (for Kurunegala see previous case study). In this region, the proportion of blood slides testing positive for P. vivax was 23.8% and for P. falciparum was 12.1% in 1991, which gradually decreased to 0.8% and 0.04%, respectively, by 2002 (Figure 2). The greatest decrease for P. vivax occurred between 1994 and In 1999, 36% of P. falciparum-positive patients (RI late 52.8%, RI early 22.2%, RII 2.8%, RIII 22.2%) had parasites that were resistant to chloroquine in Matale (n = 100) Population case study 3 The Buttala Health Area lies in the central western region of the Monaragala District, km from the border of the Badulla District (Figure 2).

6 Figure 2 Population case studies. Shown in each location are: (top) percent mortality of Anopheles culicifacies after exposure to malathion (white bars) or fenitrothion (shaded bars) (including 95% binomial CIs), with a number of small sample sizes included for general trend analysis in Rambukkana 1994 (n = 13), 1995 (n = 30) and 1997 (n = 11), in Galewela 1997 (n = 32) and in Buttala 1992 (n = 31); (middle) annual Plasmodium vivax and P. falciparum malaria positivity (and 95% binomial CIs) based on an average of blood slides in Rambukkana, in Galewela and in Buttala; and (bottom) year in which a particular insecticide was used in the region surrounding the health area, with the vertical arrows indicating the introduction of a new insecticide. Mal: malathion; Lam: lambdacyhalothrin; Del: deltamethrin; Fen: fenitrothion; Eto: etofenprox; Cyf: cyfluthrin. 756 L.A. Kelly-Hope et al.

7 Insecticide resistance in malaria vectors in Sri Lanka 757 Table 1 Number of collection sites and overall mortality rates of Anopheles culicifacies and A. subpictus in different geographical zones Species and insecticide Zone Coastal and inland Rainfall Temperature Elevation Paddy fields Population >100/km 2 100/km 2 > ha ha >100 m m >30 C 30 C >2000 mm 2000 mm Inland Coastal A. culicifacies Malathion 63.0 (8) 59.4 (44) 62.7 (36) 54.5 (16) 60.4 (21) 60.1 (31) 64.8 (23) 56.6 (29) 62.2 (30) 57.5 (22) 62.6 (28) 57.4 (24) Fenitrothion 89.3 (7) 94.6 (17) 82.6 (24) (0) 92.0 (7) 93.5 (17) 93.4 (15) 92.4 (9) 92.8 (17) 93.6 (7) 90.5 (11) 95.2 (13) Lambdacyhalothrin (4) 99.9 (17) 99.8 (12) (9) (10) 99.8 (11) 99.7 (8) 100 (13) 99.6 (6) (15) 99.7 (8) (13) Permethrin 61.5 (2) 82.6 (7) 77.9 (9) (0) 85.6 (3) 74.0 (6) 74.0 (6) 85.6 (3) 90.2 (6) 53.3 (3) 71.5 (6) 90.7 (3) A. subpictus Malathion 87.4 (19) 43.0 (13) * 69.2 (32) (0) 45.4 (7) 76.1 (25) 74.0 (26) 49.1 (6) 72.9 (10) 67.7 (22) 63.4 (16) 75.3 (16) Fenitrothion 89.7 (13) 72.5 (12) 81.1 (24) 88.0 (1) 61.0 (3) 84.2 (22) 85.3 (21) 60.8 (4) 82.6 (17) 78.8 (8) 69.5 (8) 87.0 (17) Lambdacyhalothrin (9) 91.8 (13) 94.9 (21) (1) (5) 93.7 (17) 94.9 (14) 95.6 (8) 96.5 (12) 93.5 (10) 95.5 (11) 94.8 (11) Permethrin 75.4 (5) 86.9 (18) 83.8 (21) 90.8 (2) 86.6 (5) 83.8 (18) 87.0 (17) 77.1 (6) 89.5 (15) 75.0 (8) 72.0 (7) 89.9 (16) * P < Malathion susceptibility tests were carried out in A. culicifacies in , 2000 and In addition, fenitrothion susceptibility tests were carried out in A. culicifacies in 1991, and (n = 535, 97.9%). The numbers of A. culicifacies collected for testing during this period were 145 in 1991, 31 in 1992, 187 in 1993, 102 in 1994, 147 in 1995, 0 in 1996 and 1997, and 483 in Anopheles annularis, A. jamesi, A. tessellates, A. vagus and A. varuna had high mortalities with malathion and fenitrothion ( 80%), in contrast to A. nigerrimus, A. subpictus and A. peditaeniatus. Malathion was replaced in the Monaragala District with fenitrothion in 1999 (until 2001), lambdacyhalothrin in 2000 (until 2001) and cyfluthrin in The proportion of blood slides testing positive for P. vivax was 17.8% and for P. falciparum was 2.0% in 1995, which decreased to 1.0% and 0.2%, respectively, by 2002 (Figure 2). Overall, P. falciparum prevalence remained relatively low (<4%), whilst P. vivax increased dramatically in 1998 (28.6%) and 1999 (26%), and then decreased to 1% by In 1999, 62.0% of P. falciparum-positive patients (RI late 46.8%, RI early 25.8%, RII 17.7%, RIII 9.7%) had parasites that were resistant to chloroquine in Monaragala (n = 100). 4. Discussion Mapping the spatial and temporal trends of insecticide resistance in A. culicifacies and A. subpictus in Sri Lanka over a decade has illustrated the heterogeneity of resistance present on the island. Malathion was primarily used for indoor residual spraying until the mid-1990s, when increases in malaria and malathion resistance led to its gradual replacement. The high levels of malathion resistance in many locations for both Sri Lankan vectors are probably a direct result of malathion used for malaria control for 20 years, as the major resistance mechanism in A. culicifacies is malathion-specific and this insecticide has never been used for agricultural treatments on the island. There is now evidence of emerging resistance to the newly introduced insecticides after only a few years of usage. In contrast to malathion, this may be the result of extensive agricultural use of the same insecticides over many years. Cross-resistance, whereby a mechanism responsible for resistance to one insecticide confers resistance to other chemically or functionally related insecticides (Brogdon and McAllister, 1998; Hemingway et al., 2002, 2004; Karunaratne, 1999) may also be playing a role. Major mechanisms

8 758 L.A. Kelly-Hope et al. of resistance involve either an alteration within the target site of the insecticide, which prevents it from binding to its target, or an alteration in the rate of insecticide metabolism, i.e. detoxification or sequestration, which occurs when increased levels or modified activities of esterases, monooxygenases or glutathione S-transferases (GST) prevent it from reaching its site of action. Selection by insecticides with shared target sites or detoxification pathways may therefore produce crossresistance to insecticides such as fenitrothion, even though this insecticide has had very restricted agricultural use in Sri Lanka. The main mechanisms associated with DDT and malathion resistance in Sri Lankan A. culicifacies and A. subpictus are primarily metabolic and involve carboxylesterases (malathion) or monooxygenases and GSTs (DDT) (Hemingway et al., 1991; Herath et al., 1987, 1988; Karunaratne and Hemingway, 2001). More recently, an altered acetylcholinesterase conferring organophosphate resistance has been detected in both species (Karunaratne, 1999). These mechanisms can confer broad-spectrum cross-resistance and may now be maintained by selection with the new insecticides used for malaria control. The selection of malathion resistance was almost certainly a direct result of antimalarial activities, because malathion use was restricted to public health treatments from the early 1980s and the resistance selected in A. culicifacies was malathion-specific (Karunaratne and Hemingway, 2001). However, identifying sources of selection pressure with the insecticides currently employed in vector control is more complex, as most are used both in agriculture and public health (Department of Agriculture, 2004). Anopheles subpictus is subjected to more agricultural insecticide pressure than A. culicifacies, as the former breeds in paddy fields whereas the latter mainly breeds in riverbed pools (Hemingway et al., 1987; Herath and Joshi, 1989). However, as both species rest and feed indoors, adult female mosquitoes will continue to experience pressure from permethrin-impregnated bed nets and AMC residual spray activities. The alternation of chemically unrelated insecticides was introduced in the mid-1990s to reduce malaria transmission and to delay the development of insecticide resistance. Our analyses, despite their limited sentinel nature, suggest that the changes in spray regimens have stopped resistance to the new insecticides from becoming widespread, and potentially reduced malaria transmission. This is in contrast to the earlier protracted use of DDT, which eventually selected higher levels of resistance in both vectors which subsequently remained in the populations despite the withdrawal of DDT in 1977, and the malathion use that rapidly selected for resistance which also subsequently stayed at a high level despite the reduced use of malathion since The challenge now is to maintain or improve the efficacy of control without exacerbating the insecticide resistance problems. In Mexico, studies have shown that annual planned rotations or a fine-scale mosaic system of two unrelated insecticides are effective in slowing the rate of selection for pyrethroid resistance compared with long-term use of a pyrethroid. The baseline studies for this work are published (Hemingway et al., 1997; Penilla et al., 1996, 1998) and the full results are detailed in a series of 10 papers that have been accepted for publication by Medical and Veterinary Entomology. Spatial differences in the resistance patterns within Sri Lanka may be due to differences in vector composition, distribution, biology, genetic structure or behaviour. The higher resistance to malathion and fenitrothion in A. subpictus from inland regions compared with coastal regions, and the opposite trend with permethrin, correlate well with the observations of Abhayawardana et al. (1996), who found A. subpictus species A to be abundant predominantly in inland regions and moderately resistant to malathion and fenitrothion, whereas species B was confined to the coast and resistant to permethrin. Although in routine surveillance it is not possible to distinguish A. subpictus A and B, this suggests that different resistance mechanisms may prevail in the two siblings, with species A more resistant to organophosphates and species B more resistant to pyrethroids, with little evidence of introgression between the two sibling species. A similar phenomenon between sibling species is evident in the A. gambiae s.s. molecular (chromosomal) forms in West Africa, where pyrethroid resistance (due to a kdr mutation) is more frequent in the A. gambiae molecular S-form (Savanna) than in the M-form (Mopti) (Awolola et al., 2003; Chandre et al., 1999; Fanello et al., 2003a, 2003b), although movement of genes between sibling species occurs in some areas (Diabate et al., 2003). Anopheles subpictus was more resistant to malathion and fenitrothion in more elevated, cooler regions of the country, which may be an extension of the inland coastal trend, but it could be related to temperature or to changes in landscape/vegetation that occur with altitude. Some resistance in A. subpictus was also evident in rural rice-growing regions. This is probably related to agricultural organophosphate pesticide use (Department of Agriculture, 2004) in the paddy field breeding sites of A. subpictus,

9 Insecticide resistance in malaria vectors in Sri Lanka 759 and reflects similar organophosphate resistance trends as the Japanese encephalitis vector, Culex tritaeniorhynchus, which is a paddy field breeder in Sri Lanka (Karunaratne and Hemingway, 2000). Anopheles subpictus species A is also more endophilic and seasonally abundant than species B (Abhayawardana et al., 1996), and thus should be more exposed to indoor residual insecticide treatments. Our three population case studies suggest that regular residual spraying of a single insecticide increases resistance, thereby sustaining or increasing the risk of malaria if effective alternatives are not introduced. For instance, in the Rambukkana Health Area, malathion was used until 2002 and resistance patterns changed during this time, with the highest levels found in 1996 and Interestingly, the additional use of lambdacyhalothrin in 1994 and etofenprox in 1999 in this region corresponded with a reduction in malaria. Similarly, in the Galewela Health Area, the introduction of lambdacyhalothrin in 1994 and 1998, fenitrothion in 1996 and etofenprox in 1999 probably played a key role in the decline of malarial disease. In contrast, in the Buttala Health Area, malathion resistance (>50%) increased between 1991 and 2001, although not significantly. There was a rapid increase in malaria in 1998 and 1999, which prompted a switch from malathion to fenitrothion. This delay in changing to a more effective insecticide may account for the significantly higher malaria prevalence found in Buttala compared with Rambukkana and Galewela, where alternatives to malathion were introduced 5 years earlier. It appears that alternating different insecticides for malaria control in Sri Lanka has the potential to reduce the incidence of disease and to extend the effective lifespan of insecticides. In the three population case studies, there was a 2 6- fold reduction in malaria prevalence within 2 years of introducing additional or alternative insecticides to malathion. Whilst P. vivax, which remains fully chloroquine susceptible, may have been affected by increased drug treatment, these reductions, particularly in P. falciparum prevalence, are not likely to be a result of better drug treatment, as all districts had moderate to high levels (36 62%) of chloroquine resistance and no new antimalarials were introduced as first-line drugs during the study period (Galappaththy, 2002). In a similar situation in southern Africa, sulfadoxine pyramethamine drug resistance and pyrethroid insecticide resistance were selected coincidently in the same region (Hargreaves et al., 2000; Roper et al., 2003). In this instance, as both drug and insecticide treatments were changed simultaneously, it was not possible to attribute accurately the contribution of each change to the resultant reduction in malaria transmission. The success of malaria control programmes is dependent on knowledge of the local epidemiology and associated risk factors, as these vary between populations. In Sri Lanka, human migration, housing type, age, gender, socio-economic status, proximity to water sources, rainfall, vegetation and cultivation type are all risk factors for malaria (Gamage- Mendis et al., 1991; Gunawardena et al., 1998; Klinkenberg et al., 2004; Konradsen et al., 2003; Mendis et al., 1990; van der Hoek et al., 1998, 2003). The populations presented in our case studies vary environmentally, socio-economically and demographically, and therefore will have different prevailing risk factors that will impact on control programmes. For example, Rambukkana and Galewela are more urbanised than Buttala, which has a large rural population, with many living in poorly constructed houses, which is an important risk factor. This region also receives an annual influx of visitors and religious pilgrims to the Kataragama Shrine (Harrigan, 1998), which may contribute to the region s high malaria incidence. Anopheles peditaeniatus, A. nigerrimus and A. vagus had varying degrees of insecticide resistance. These species are primarily exophagic paddy field breeders (Yapabandara, 1997), hence their resistance patterns should reflect selection by agricultural insecticides. Many other non-vectors remained fully insecticide susceptible. Hence, nonvectors cannot be used as surrogates to extrapolate to the resistance status of the vectors. Although the malaria situation in Sri Lanka initially worsened over the past two decades, more recently there has been a dramatic drop in malaria case numbers (Briet et al., 2005). This is presumably predominantly because of renewed access by control personnel to the northern districts of the country following the ceasefire in the civil war. From our data, we consider that in the districts less affected by war, the switching of spraying from malathion to alternatives has also contributed importantly to the decline in national totals of malaria cases. The emergence and spread of insecticide resistance is a constant threat to these control programmes as the vector populations respond to increasing insecticide selection pressures, and active surveillance systems are vital in identifying key vectors and evidence of resistance. Conflicts of interest statement The authors have no conflicts of interest concerning the work reported in this paper.

10 760 L.A. Kelly-Hope et al. Acknowledgements We are grateful to Dr R.R. Abeyasinghe for comments on the manuscript, and to all the Regional Malaria/Medical Officers and AMC entomological staff who were involved in the collection of this national sentinel data. We thank Dr Ian Hastings for statistical advice. This study was partly funded by the Gates Malaria Partnership and the National Science Foundation, Sri Lanka. References Abhayawardana, T.A., Wijesuriya, S.R., Dilrukshi, R.K., Anopheles subpictus complex: distribution of sibling species in Sri Lanka. Indian J. Malariol. 33, Amerasinghe, F.P., Ariyasena, T.G., Survey of adult mosquitoes (Diptera: Culicidae) during irrigation development in the Mahaweli Project, Sri Lanka. J. Med. Entomol. 28, Amerasinghe, F.P., Indrajith, N.G., Postirrigation breeding patterns of surface water mosquitoes in the Mahaweli Project, Sri Lanka, and comparisons with preceding developmental phases. J. Med. Entomol. 31, Amerasinghe, F.P., Amerasinghe, P.H., Peiris, J.S., Wirtz, R.A., Anopheline ecology and malaria infection during the irrigation development of an area of the Mahaweli Project, Sri Lanka. Am. J. Trop. Med. Hyg. 45, Amerasinghe, P.H., Amerasinghe, F.P., Wirtz, R.A., Indrajith, N.G., Somapala, W., Pereira, L.R., Rathnayake, A.M.S., Malaria transmission by Anopheles subpictus (Diptera: Culicidae) in a new irrigation project in Sri Lanka. J. Med. Entomol. 29, Amerasinghe, P.H., Amerasinghe, F.P., Konradsen, F., Fonseka, K.T., Wirtz, R.A., Malaria vectors in a traditional dry zone village in Sri Lanka. Am. J. Trop. Med. Hyg. 60, Awolola, T.S., Brooke, B.D., Koekemoer, L.L., Coetzee, M., Absence of the kdr mutation in the molecular M form suggests different pyrethroid resistance mechanisms in the malaria vector mosquito Anopheles gambiae s.s. Trop. Med. Int. Health 8, Briet, O.J., Gunawardena, D.M., van der Hoek, W., Amerasinghe, F.P., Sri Lanka malaria maps. Malar. J. 2, 22. Briet, O.J., Galappaththy, G.N., Konradsen, F., Amerasinghe, P.H., Amerasinghe, F.P., Maps of the Sri Lanka malaria situation preceding the tsunami and key aspects to be considered in the emergency phase and beyond. Malar. J. 4, 8. Brogdon, W.G., McAllister, J.C., Insecticide resistance and vector control. Emerg. Infect. Dis. 4, Carter, H.F., Further observations on the transmission of malaria by anopheline mosquitoes in Ceylon. Ceylon J. Sci. 2, Carter, H.F., Jacocks, W.P., Observations on the transmission of malaria by anopheline mosquitoes in Ceylon. Ceylon J. Sci. 2, Chandre, F., Manguin, S., Brengues, C., Dossou Yovo, J., Darriet, F., Diabate, A., Carnevale, P., Guillet, P., Current distribution of a pyrethroid resistance gene (kdr) in Anopheles gambiae complex from west Africa and further evidence for reproductive isolation of the Mopti form. Parassitologia 41, Department of Agriculture, Data obtained from website. Agriculture in Sri Lanka. Pesticides. Colombo, Sri Lanka. [accessed 20 April 2005]. Department of Census and Statistics, Data obtained from website. Colombo, Sri Lanka. gov.lk/index.asp [accessed 20 April 2005]. Diabate, A., Baldet, T., Chandre, C., Dabire, K.R., Kengne, P., Guiguemde, T.R., Simard, F., Guillet, P., Hemingway, J., Hougard, J.M., KDR mutation, a genetic marker to assess events of introgression between the molecular M and S forms of Anopheles gambiae (Diptera: Culicidae) in the tropical savannah area of West Africa. J. Med. Entomol. 40, Fanello, C., Petrarca, V., della Torre, A., Santolamazza, F., Dolo, G., Coulibaly, M., Alloueche, A., Curtis, C.F., Toure, Y.T., Coluzzi, M., 2003a. The pyrethroid knock-down resistance gene in the Anopheles gambiae complex in Mali and further indication of incipient speciation within An. gambiae s.s. Insect. Mol. Biol. 12, Fanello, C., Carneiro, I., Ilboudo-Sanogo, E., Cuzin-Ouattara, N., Badolo, A., Curtis, C.F., 2003b. Comparative evaluation of carbosulfan- and permethrin-impregnated curtains for preventing house-entry by the malaria vector Anopheles gambiae in Burkina Faso. Med. Vet. Entomol. 17, Galappaththy, G.N.L., A study of chloroquine resistant Plasmodium falciparum malaria in Sri Lanka. MD Thesis, Postgraduate Institute of Medicine, University of Colombo, Sri Lanka. Gamage-Mendis, A.C., Carter, R., Mendis, C., De Zoysa, A.P., Herath, P.R., Mendis, K.N., Clustering of malaria infections within an endemic population: risk of malaria associated with the type of housing construction. Am. J. Trop. Med. Hyg. 45, Gunawardena, D.M., Wickremasinghe, A.R., Muthuwatta, L., Weerasingha, S., Rajakaruna, J., Senanayaka, T., Kotta, P.K., Attanayake, N., Carter, R., Mendis, K.N., Malaria risk factors in an endemic region of Sri Lanka, and the impact and cost implications of risk factor-based interventions. Am. J. Trop. Med. Hyg. 58, Hargreaves, K., Koekemoer, L.L., Brooke, B.D., Hunt, R.H., Mthembu, J., Coetzee, M., Anopheles funestus resistant to pyrethroid insecticides in South Africa. Med. Vet. Entomol. 14, Harrigan, P., From Kailasa to Kataragama: sacred geography in the cult of Skanda-Murukan, in: Kataragama: The Mystery Shrine. Institute of Asian Studies, Chennai, India. [accessed 20 April 2005]. Hemingway, J., Jayawardena, K.G.I., Weerasinghe, I., Herath, P.R.J., The use of biochemical tests to identify multiple insecticide resistance mechanisms in field-selected populations of Anopheles subpictus Grassi (Diptera: Culicidae). Bull. Entomol. Res. 77, Hemingway, J., Miyamoto, J., Herath, P.R.J., A possible novel link between organophosphorous and DDT insecticide resistance genes in Anopheles: supporting evidence from fenitrothion metabolism studies. Pestic. Biochem. Physiol. 39, Hemingway, J., Penilla, R.P., Rodriguez, A.D., James, B.M., Edge, W., Rogers, H., Rodriguez, M.H., Resistance management strategies in malaria vector mosquito control. A large-scale field trial in Southern Mexico. Pestic. Sci. 51, Hemingway, J., Field, L., Vontas, J., An overview of insecticide resistance. Science 298, Hemingway, J., Hawkes, N.J., McCarroll, L., Ranson, H., The molecular basis of insecticide resistance in mosquitoes. Insect. Biochem. Mol. Biol. 34, Herath, P.R., Joshi, G.P., Pesticide selection pressure on Anopheles subpictus in Sri Lanka: comparison with two other

11 Insecticide resistance in malaria vectors in Sri Lanka 761 Sri Lankan anophelines. Trans. R. Soc. Trop. Med. Hyg. 83, Herath, P.R.J., Abeyawardene, T.A., Padmalal, U.K.G.K., A study of the role of indigenous anopheline species in the transmission of human malaria in Sri Lanka. Proc. Sri Lanka Assoc. Adv. Sci. 39, 6. Herath, P.R.J., Hemingway, J., Weerasinghe, I.S., Jayawardena, K.G.I., The detection and characterization of malathion resistance in field populations of Anopheles culicifacies B in Sri Lanka. Pestic. Biochem. Physiol. 29, Herath, P.R.J., Jayawardena, K.G.I., Hemingway, J., Harris, J., DDT resistance in Anopheles culicifacies Giles and A. subpictus Grassi (Diptera: Culicidae) from Sri Lanka: a field study on the mechanisms and changes in gene frequency after cessation of DDT spraying. Bull. Entomol. Res. 78, Karunaratne, S.H.P., Insecticide cross-resistance spectra and underlying resistance mechanisms of Sri Lankan anopheline vectors of malaria. Southeast Asian J. Trop. Med. Public Health 30, Karunaratne, S.H.P.P., Hemingway, J., Insecticide resistance spectra and resistance mechanisms in populations of Japanese encephalitis vector mosquitoes, Culex tritaeniorhynchus and Cx. gelidus, in Sri Lanka. Med. Vet. Entomol. 14, Karunaratne, S.H., Hemingway, J., Malathion resistance and prevalence of the malathion carboxylesterase mechanism in populations of mosquito vectors of disease in Sri Lanka. Bull. World Health Organ. 79, Klinkenberg, E., van der Hoek, W., Amerasinghe, F.P., A malaria risk analysis in an irrigated area in Sri Lanka. Acta Trop. 89, Konradsen, F., Amerasinghe, P., van der Hoek, W., Amerasinghe, F., Perera, D., Piyaratne, M., Strong association between house characteristics and malaria vectors in Sri Lanka. Am. J. Trop. Med. Hyg. 68, Mendis, C., Gamage-Mendis, A.C., De Zoysa, A.P., Abhayawardena, T.A., Carter, R., Herath, P.R., Mendis, K.N., Characteristics of malaria transmission in Kataragama, Sri Lanka: a focus for immuno-epidemiological studies. Am. J. Trop. Med. Hyg. 42, Mendis, C., Herath, P.R., Rajakaruna, J., Weerasinghe, S., Gamage-Mendis, A.C., Mendis, K.N., De Zoysa, A.P., Method to estimate relative transmission efficiencies of Anopheles species (Diptera: Culicidae) in human malaria transmission. J. Med. Entomol. 29, Penilla, R.P., Rodriguez, A.D., Hemingway, J., Estrada, J.L.T., Arredondo-Jimenez, J.I., Rodriguez, M.H., Rotational and mosaic strategies for delaying the development of insecticide resistance in mosquitoes baseline data for a large scale field trial in Southern Mexico, in: Proceedings of the 2nd International Conference on Insect Pests in the Urban Environment, BPCC, pp Penilla, R.P., Rodriguez, A.D., Hemingway, J., Torres, J.L., Arredondo-Jimenez, J.I., Rodriguez, M.H., Resistance management strategies in malaria vector mosquito control. Baseline data for a large-scale field trial against Anopheles albimanus in Mexico. Med. Vet. Entomol. 12, Ramasamy, R., De Alwis, R., Wijesundere, A., Ramasamy, M.S., Malaria transmission at a new irrigation project in Sri Lanka: the emergence of Anopheles annularis as a major vector. Am. J. Trop. Med. Hyg. 47, Ratnapala, R., Subramaniam, K., Yapabandara, M.G., Fernando, W.P., Chloroquine resistant Plasmodium falciparum in Sri Lanka. Ceylon Med. J. 29, Roll Back Malaria, World Malaria Report [accessed 20 April 2005]. Roper, C., Pearce, R., Bredenkamp, B., Gumede, J., Drakeley, C., Mosha, F., Chandramohan, D., Sharp, B., Antifolate antimalarial resistance in southeast Africa: a populationbased analysis. Lancet 361, van der Hoek, W., Konradsen, F., Dijkstra, D.S., Amerasinghe, P.H., Amerasinghe, F.P., Risk factors for malaria: a microepidemiological study in a village in Sri Lanka. Trans. R. Soc. Trop. Med. Hyg. 92, van der Hoek, W., Konradsen, F., Amerasinghe, P.H., Perera, D., Piyaratne, M.K., Amerasinghe, F.P., Towards a risk map of malaria for Sri Lanka: the importance of house location relative to vector breeding sites. Int. J. Epidemiol. 32, WHO, Implementation of the Global Malaria Control Strategy. Report of a WHO Study Group on the Implementation of the Global Plan of Action for Malaria Control World Health Organization, Geneva, Technical Report Series, No Wickramasinghe, M.B., Malaria and its control in Sri Lanka. Ceylon Med. J. 26, Yapabandara, A.M.G.M., Control of anopheline vectors in a gem mining area in Sri Lanka. PhD thesis, University of London, UK. Yapabandara, A.M., Curtis, C.F., Vectors and malaria transmission in a gem mining area in Sri Lanka. J. Vector Ecol. 29,