BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofitpublishers, academic institutions, research libraries, and research funders in the common goal of maximizing access tocritical research.

Organophosphate Resistance in Trinidad and Tobago Strains ofAedes aegyptiAuthor(s): Karen A. Polson, Samuel C. Rawlins, William G. Brogdon, and DaveD. ChadeeSource: Journal of the American Mosquito Control Association, 26(4):403-410.2010.Published By: The American Mosquito Control AssociationDOI: http://dx.doi.org/10.2987/10-6019.1URL: http://www.bioone.org/doi/full/10.2987/10-6019.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in thebiological, ecological, and environmental sciences. BioOne provides a sustainable onlineplatform for over 170 journals and books published by nonprofit societies, associations,museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated contentindicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercialuse. Commercial inquiries or rights and permissions requests should be directed to theindividual publisher as copyright holder.

ORGANOPHOSPHATE RESISTANCE IN TRINIDAD AND TOBAGOSTRAINS OF AEDES AEGYPTI

KAREN A. POLSON,1 SAMUEL C. RAWLINS,2 WILLIAM G. BROGDON3AND DAVE D. CHADEE1,4

ABSTRACT. Aedes aegypti larvae from 8 sites in Trinidad and 1 in Tobago were assayed againsttemephos, malathion, and fenthion using the Centers for Disease Control and Prevention time-mortality–based bioassay method. Resistance ratios (RRs) and resistance thresholds (RTs) for each insecticide werecalculated in relation to the Caribbean Epidemiology Center reference susceptible strain. Results showed thatthe Haleland Park and Tobago strains were susceptible to fenthion and malathion, respectively (RRs , 1),while the San Fernando strain had a high RR (33.92) to malathion. All other strains had low-level resistanceto fenthion and malathion. Resistance to temephos was more intense with 4 strains showing high-levelresistance. The established RT was 60 min for fenthion, 75 min for bendiocarb, and 120 min for temephosand malathion. At the RTs, all Trinidad strains were resistant to temephos (11.50–74.50% mortality), 7resistant to fenthion (21.25–78.75% mortality), and 5 resistant to malathion (56.25–77.50% mortality). Theother strains were incipiently resistant (80–97% mortality). Despite the discrepancies between the RR levelsand RT status, it is evident that the organophosphate insecticide resistance is prevalent in Trinidad andTobago populations of Ae. aegypti. These results suggest that operational failure could soon occur andalternative strategies should be developed and implemented to reduce the probability of further selectionpressure on resistant Ae. aegypti populations in Trinidad and Tobago.

KEY WORDS Aedes aegypti, insecticide resistance, organophosphates, Trinidad and Tobago

INTRODUCTION

Aedes aegypti (L.), the main vector of dengueand yellow fever, is of significant public healthimportance in the tropics. The global incidence ofdengue has increased dramatically in the pastdecade and now there are approximately 2.5billion people at risk with an estimated 50–100million cases of dengue fever and 250,000–500,000 cases of dengue hemorrhagic feverworldwide (WHO 2008). At present, there is notreatment or vaccine available for dengue and,therefore, the only available means of preventionis through the control of the mosquito vector.

Insecticides play a key role in Ae. aegypticontrol programs and a survey conducted in 1990revealed that the majority of Caribbean countriesrelied heavily upon the use of insecticides forlarval control (Nathan 1993). Organophosphate(OP) insecticides such as fenthion, fenitrothion,and malathion are mainly used for residual orspace spraying, while temephos is used forlarviciding with 1% sand-core granules appliedto domestic water containers (Chadee and Raha-man 2000). These insecticides have been widelyused in the Caribbean region to control Ae.aegypti for over the last 30–40 years (Rawlins1998). However, with the continued use andreliance on insecticides, resistance to these insec-

ticides has been reported (Georghiou et al. 1987,Rawlins and Ou Hing Wan 1995, Rawlins 1998).

Generally, the reported levels of Ae. aegyptiresistance to OPs in the Caribbean area havevaried from low to moderate resistance levels (5-to 10-fold), to high resistance (.10-fold) (Geor-ghiou et al. 1987, Rawlins and Ragoonanansingh1990, Mekuria et al. 1991, Rawlins and Ou HingWan 1995, Rawlins 1998). Studies conducted byRawlins and Ou Hing Wan (1995) showedresistance to temephos, malathion, and fenthionin some larval populations of Ae. aegypti in theCaribbean. In 1998, Rawlins reviewed the prev-alence of temephos and malathion resistance inlarval and adult Ae. aegypti populations, respec-tively, from several Caribbean islands and notedthat there was general widespread prevalence ofresistance to temephos with only limited occur-rence of varying levels of resistance to malathion(Rawlins 1998).

Aedes aegypti was reintroduced into Trinidadbetween 1961 and 1962 and into Tobago in 1983following the earlier ‘‘successful’’ eradicationconducted by the Pan American Health Organi-zation (Chadee 1984, Chadee et al. 1984). Sincethen, Ae. aegypti has become widespread inTrinidad and Tobago although vector controlefforts have been ongoing mainly through the useof OP insecticides (Yan et al. 1998). Strongselection pressure on Ae. aegypti populations hasinevitably led to the development of insecticideresistance: for example, high levels of resistanceto temephos and fenthion and low levels tomalathion (Rawlins and Ragoonanansingh 1990);low levels of resistance to temephos, malathion,and fenthion (Rawlins and Ou Hing Wan 1995);4 To whom correspondence should be addressed.

1 The University of the West Indies, St. Augustine,Trinidad and Tobago.

2 Caribbean Epidemiology Center (CAREC), Port ofSpain, Trinidad and Tobago.

3 Centers for Disease Control and Prevention (CDC),Atlanta, GA 30333.

Journal of the American Mosquito Control Association, 26(4):403–410, 2010Copyright E 2010 by The American Mosquito Control Association, Inc.

403

and low to moderate levels of resistance totemephos (Rawlins 1998).

The purpose of this study was to assess thecurrent status of insecticide resistance in someTrinidad and Tobago larval populations of Aeaegypti and to compare these results with datafrom the last study conducted .10 years ago(Rawlins 1998).

MATERIALS AND METHODS

Mosquitoes

Aedes aegypti eggs were collected from 8geographically distinct areas in Trinidad and 1in Tobago (Fig. 1) in 2006, using enhancedovitraps (Reiter et al. 1991). The Port of Spainstrains (St. Clair, St. James, and Sea Lots) arelocated in the northwestern part of Trinidad. TheHaleland Park strain is located on the outskirts ofPort of Spain, the Curepe and Valencia strains inthe northeastern part of the island, the SpringVale strain in the central part, and San Fernandoin the southern part of the island. Colonies wereestablished for all field strains and bioassays wereconducted on late 3rd- to early 4th-stage larvae ofthe F2–F4 generation. The Caribbean Epidemiol-ogy Center (CAREC) strain of Ae. aegyptimosquitoes (Rawlins and Ou Hing Wan 1995,Rawlins 1998), a known susceptible laboratorystrain from Trinidad, and the Rockefeller(ROCK) strain, a susceptible population ob-tained from the Centers for Disease Controland Prevention (CDC) laboratory, were used asthe reference susceptible strains.

Bioassays

Determination of diagnostic dosages: Bioassayswere conducted using the method developed byCDC (Ocampo et al. 2000, CDC 2004). Thismethod is a variation of the WHO bioassay(WHO 1981) and follows the same principles asthe CDC bottle bioassay for adult mosquitoes(Brogdon and McAllister 1998). This assay is atime-mortality assay that measures the intoxica-tion rate of the insecticide. Diagnostic dosages oftemephos, malathion, and fenthion were firstdetermined by exposing 20 6 5 larvae of theCAREC strain, in replicates of 4, to 6 concen-trations of each insecticide (1, 10, 20, 100, 200,and 250 mg/100 ml) and mortality was recorded at15-min intervals until all larvae died or up to 2 hafter the start of the experiment. The diagnosticdosage was established as the concentration ofeach insecticide beyond which further increasesdid not result in a faster kill of mosquitoes(saturation point).

Bioassay of field strains: Field strains of Ae.aegypti larvae were subjected to the knowndiagnostic dosage of each insecticide and mortal-

ity counts recorded as stated earlier. The totalnumber of mosquito larvae in each cup wasrecounted and the percentage mortality calculat-ed for each 15-min interval.

Controls: As an internal control, the CARECstrain larvae were exposed to the diagnosticdosage, for the respective insecticide being tested,simultaneously with each assay. Also, for eachassay, 1 ml of acetone was added to a cupcontaining 20 6 5 larvae of the field strain beingtested.

Data analyses

Lethal times and resistance ratios: Lethal time(LT) values, LT50 and LT90 with 95% fiduciallimits, were calculated for each mosquito strainand insecticide by log-probit analyses of time-mortality–correlated data (Throne et al. 1995).The probit analysis was done with PROBITHwritten in Mathematica version 7 (Wolfram,Champaign, IL). The Pearson goodness-of-fitchi-square test was used to determine thegoodness of fit between the regression lines andobserved data. The LT values with overlappingfiducial limits were not considered to be signifi-cantly different.

Field strains were compared with the CARECreference strain by calculation of a resistanceratio (RR), using the following formula:

Resistance ratio~LT50 or LT90 field strain

LT50 or LT90 CAREC strain

The criteria proposed by Mazarri and Georghiou

(1995) were used to classify resistance ratios as

high (.10-fold), medium (between 5 and 10), and

low (,5). The degree of resistance compared to

Fig. 1. Map of Trinidad and Tobago showinglocalities from which Aedes aegypti eggs were collected.

404 JOURNAL OF THE AMERICAN MOSQUITO CONTROL ASSOCIATION VOL. 26, NO. 4

the susceptible CAREC strain was assessed at the

LT90 level.Resistance thresholds and susceptibility status:

Graphs of time versus percentage mortality wereconstructed for the CAREC and ROCK refer-ence susceptible strains and the resistance thresh-old (RT), the time in which 100% mortality wasobserved, was ascertained for each insecticide.Mean percentage mortalities of mosquitoes fromfield strains were compared to the CAREC and/or ROCK strain and mosquitoes which survivedthe respective RT were considered to be resistant(Brogdon and McAllister 1998). The experimentswere allowed to run until the cutoff time of120 min and mean percentage mortalities wererecorded again at the end of the experiments.Statistical analyses were carried out using Krus-kal–Wallis and Dunn’s nonparametric tests,comparing mean percentage mortalities of theCAREC susceptible reference strain with the fieldstrains. Field strains were determined to besignificantly different from the CAREC referencestrain if P , 0.05. At both the RTs and end ofexperiments, mosquito strains were also assessedas being resistant, using the criteria defined byDavidson and Zahar (1973) and modified byWHO (1998). Mosquitoes showing 98–100%mortality were classified as susceptible, ,80%mortality as resistant, and 80–97% mortality asincipiently resistant.

RESULTS

Diagnostic dosages for each insecticide weredetermined as 100 mg/100 ml for fenthion and200 mg/100 ml for temephos and malathion.Different resistance levels to these insecticideswere observed among the strains (Tables 1–4).Control mortalities were negligible, ,5%; there-fore, no correction was necessary.

Lethal times and resistance ratios

Overall results showed that there was resistancein most populations to fenthion, malathion, andtemephos.

Fenthion: The slopes of the regression lines forthe field strains were lower than that of theCAREC strain in all cases, suggesting a high levelof heterogeneity and the deviations in linearity ofthe regression lines were significant in 4 strains (P, 0.05) (Table 1). Except for the Haleland Parkstrain which was susceptible (RR 0.92), all fieldstrains showed low-level resistance, with RRsranging from 1.33 (San Fernando) to 3.82 (St.James) (Table 1).

Malathion: Except for the Sea Lots strain, theslope values of all the field strains were lowerthan that of the CAREC strain, suggesting highheterogeneity. The differences between the ob-

Tab

le1.

Let

hal

tim

esan

dre

sist

an

cera

tio

so

fA

edes

aegypti

stra

ins

inre

spo

nse

tofe

nth

ion

(100

mg/1

00

ml)

.1

Str

ain

Nu

mb

er2

LT

50

(FL

)3L

T90

(FL

)3S

lop

e6

SE

x2

P-v

alu

e

Res

ista

nce

rati

o4

LT

50

LT

90

CA

RE

C72

39.6

2(3

7.8

3–41.3

5)

48.5

3(4

6.1

6–51.8

9)

14.5

36

1.6

10.3

30.9

91.0

01.0

0R

ock

efel

ler

94

41.0

7(3

9.4

6–42.6

)50.3

9(4

8.2

3–53.3

8)

14.4

26

1.4

61.3

40.9

71.0

41.0

4H

ale

lan

dP

ark

66

29.1

3(2

4.4

6–33.2

)44.8

2(3

9.0

8–55.1

6)

6.8

46

0.6

712.2

10.0

60.7

40.9

2S

eaL

ots

88

85.8

4(8

1.7

3–90.3

8)

146.0

8(1

33.3

1–164.9

)5.5

56

0.4

51.0

50.9

82.1

73.0

1S

t.C

lair

78

54.1

2(5

1.5

3–56.6

1)

76.8

2(7

2.7

6–82.1

)8.4

36

0.6

51.5

20.9

61.3

71.5

8S

t.Ja

mes

84

79.7

1(6

9.8

3–92.6

2)

185.2

4(1

44.6

8–287.6

7)

3.4

96

0.3

012.8

40.0

52.0

13.8

2C

ure

pe

89

56.6

8(5

0.2

–62.8

1)

97.8

6(8

6.3

–117.5

)5.4

06

0.3

614.9

20.0

21.4

32.0

2V

ale

nci

a87

64.3

5(6

0.7

5–67.9

2)

116.0

1(1

07.1

7–128.1

)5.0

06

0.3

66.0

20.4

21.6

22.3

9S

pri

ng

Vale

80

53.1

9(4

6.3

4–59.2

9)

81.6

5(7

2.4

–97.8

2)

6.8

86

0.4

918.7

50.0

11.3

41.6

8S

an

Fer

nan

do

87

47.0

464.7

79.2

36

0.7

04,9

60.8

20.0

01.1

91.3

3T

ob

ago

76

72.5

6(6

1.7

2–84.7

6)

134.8

6(1

09.3

8–204.2

8)

4.7

66

0.3

823.5

20.0

01.8

32.7

8

1D

iag

no

stic

do

sag

efo

rfe

nth

ion

.2

Nu

mb

ero

fla

rva

ea

ssa

yed

.3

LT

50,

LT

90

5ti

me

(min

ute

s)re

qu

ired

tok

ill

50%

,9

0%

larv

al

sam

ple

;9

5%

fid

uci

al

lim

its

(FL

)in

pa

ren

thes

es.

4R

esis

tan

cera

tio

5L

T50(9

0)

fiel

dst

rain

/LT

50(9

0)

Cari

bb

ean

Ep

idem

iolo

gy

Cen

ter

(CA

RE

C)

stra

in.

DECEMBER 2010 ORGANOPHOSPHATE RESISTANCE IN TRINIDAD AND TOBAGO 405

Tab

le2.

Let

hal

tim

esan

dre

sist

an

cera

tio

so

fA

edes

aeg

ypti

stra

ins

inre

spo

nse

tom

ala

thio

n(2

00

mg/1

00

ml)

.1

Str

ain

Nu

mb

er2

LT

50

(FL

)3L

T90

(FL

)3S

lop

e6

SE

x2

P-v

alu

e

Res

ista

nce

rati

o4

LT

50

LT

90

CA

RE

C83

60.3

8(5

7.2

2–63.4

7)

98.6

7(9

2.3

8–106.9

3)

6.0

06

0.4

11.3

90.9

71.0

01.0

0R

ock

efel

ler

83

72.3

4(6

6.8

3–77.5

9)

93.3

5(8

5.9

9–106.7

6)

11.5

76

0.8

917.9

20.0

11.2

00.9

5H

ale

lan

dP

ark

68

102.1

(92.4

–116.2

8)

259.9

4(2

04.9

5–374.7

2)

3.1

66

0.3

68.2

30.2

21.6

92.6

3S

eaL

ots

89

80.6

9(7

7.1

7–84.3

4)

128.3

4(1

19.4

8–140.6

3)

6.3

66

0.4

88.0

50.2

31.3

41.3

0S

t.C

lair

78

92.6

6(8

5.4

2–102.0

8)

220.4

8(1

83.0

9–287.8

4)

3.4

06

0.3

31.8

10.9

41.5

32.2

3S

t.Ja

mes

78

91.9

9(8

4.0

4–102.4

9)

243.8

4(1

98.0

8–328.1

4)

3.0

36

0.2

97.2

90.2

91.5

22.4

7C

ure

pe

87

65.6

9(6

0.7

–71.0

6)

170.5

2(1

47.2

5–206.9

5)

3.0

96

0.2

51.8

90.9

31.0

91.7

3V

ale

nci

a85

91.5

1(8

6.8

4–96.9

6)

158.5

(142.8

6–182.4

4)

5.3

76

0.4

67.6

20.2

71.5

21.6

1S

pri

ng

Vale

77

52.0

7(4

1.8

6–61.8

4)

119.7

3(9

5.5

9–177.6

3)

3.5

46

0.2

719.9

90.0

00.8

61.2

1S

an

Fer

nan

do

91

450.3

4(2

63.1

8–1,4

69.8

6)

3,3

47.0

2(1

,132.6

8–38,3

27.7

5)

1.4

76

0.2

91.1

20.9

87.4

633.9

2T

ob

ago

68

46.4

4(3

6.7

6–54.9

3)

85.1

0(7

0.8

6–115.0

7)

4.8

76

0.3

623.6

50.0

00.7

70.8

6

1D

iag

no

stic

do

sag

efo

rm

ala

thio

n.

2N

um

ber

of

larv

ae

ass

ay

ed.

3L

T50,

LT

90

5ti

me

(min

ute

s)re

qu

ired

tok

ill

50%

,9

0%

larv

al

sam

ple

;9

5%

fid

uci

al

lim

its

(FL

)in

pa

ren

thes

es.

4R

esis

tan

cera

tio

5L

T50(9

0)

fiel

dst

rain

/LT

50(9

0)

Cari

bb

ean

Ep

idem

iolo

gy

Cen

ter

(CA

RE

C)

stra

in.

Tab

le3.

Let

hal

tim

esan

dre

sist

an

cera

tio

so

fA

edes

aegypti

stra

ins

inre

spo

nse

tote

mep

ho

s(2

00

mg/1

00

ml)

.1

Str

ain

Nu

mb

er2

LT

50

(FL

)3L

T90

(FL

)3S

lop

e6

SE

x2

P-v

alu

e

Res

ista

nce

rati

o4

LT

50

LT

90

CA

RE

C77

30.1

2(1

8.0

8–39.8

9)

67.6

5(5

1.0

9–112.4

4)

3.6

56

0.2

739.4

80.0

01.0

01.0

0R

ock

efel

ler

77

48.7

8(3

6.2

5–60.7

9)

95.6

4(7

4.7

–156)

4.3

86

0.3

41.6

30.0

01.6

21.4

1H

ale

lan

dP

ark

81

98.9

6(9

3.1

–106.4

2)

181.1

2(1

58.6

8–218.7

)4.8

86

0.4

73.6

60.7

23.2

92.6

8S

eaL

ots

88

1,2

15.0

1(4

28.0

2–54,4

42.2

4)

11,7

61.2

7(1

,828.6

8–1.1

3E

+07)

1.2

96

0.3

81.5

20.9

640.3

4173.8

5S

t.C

lair

86

202.0

8(1

59.6

4–356.1

9)

408.1

2(2

63–1,2

09.2

4)

4.1

96

0.9

25.3

50.5

06.7

16.0

3S

t.Ja

mes

66

720.6

9(2

80.5

1–472,2

96.1

4)

3,4

24.1

9(6

89.9

3–2.4

2E

+08)

1.8

96

0.7

33.8

30.7

023.9

350.6

2C

ure

pe

84

156.9

2(1

37.0

6–202.3

6)

279.0

7(2

12.9

2–475.5

5)

5.1

26

0.8

71.2

50.9

75.2

14.1

3V

ale

nci

a89

290.6

1(2

01.3

8–633.8

5)

1,0

81.0

5(5

29.4

9–5,1

17.2

9)

2.2

56

0.4

32.7

50.8

415.9

815.9

8S

pri

ng

Vale

70

296.1

1(1

97.4

–776.7

6)

1,0

93.0

5(5

00.2

4–7,3

78.0

9)

2.2

66

0.4

90.4

30.9

99.8

316.1

6S

an

Fer

nan

do

79

87.6

2(7

5.6

–106.1

2)

181.2

4(1

38.6

4–316.2

4)

4.0

66

0.3

618.7

20.0

52.6

82.6

8T

ob

ago

72

51.4

3(4

7.2

5–55.5

5)

114.4

9(1

02.6

5–131.4

4)

3.6

96

0.2

81.9

50.9

31.7

11.6

9

1D

iagn

ost

icd

osa

ge

for

tem

eph

os.

2N

um

ber

of

larv

ae

ass

ayed

.3

LT

50,

LT

90

5ti

me

(min

ute

s)re

qu

ired

tok

ill

50%

,9

0%

larv

al

sam

ple

;9

5%

fid

uci

al

lim

its

(FL

)in

pa

ren

thes

es.

4R

esis

tan

cera

tio

5L

T50(9

0)

fiel

dst

rain

/LT

50(9

0)

Cari

bb

ean

Ep

idem

iolo

gy

Cen

ter

(CA

RE

C)

stra

in.

406 JOURNAL OF THE AMERICAN MOSQUITO CONTROL ASSOCIATION VOL. 26, NO. 4

served and expected regression lines were signif-icant in the ROCK, Tobago, and Spring Valestrains (P , 0.05) (Table 2). Low-level resistanceto malathion was also evident in 7 strains, withRRs ranging from 1.61 (Valencia) to 2.47 (St.James). The Tobago strain was susceptible (RR0.86), but the San Fernando strain was highlyresistant (RR 33.92) to malathion (Table 2).

Temephos: Higher heterogeneity than theCAREC strain was ascertained in 4 strains asevidenced by lower slope values; and slopedifferences were significant in the CAREC,ROCK, Anguilla, and San Fernando strains (P, 0.05) (Table 3). Although no field strain wassusceptible to temephos, low resistance levelswere seen in 4 strains with RRs ranging from 1.69(Tobago) to 4.13 (Curepe). The St. Clair strainshowed medium-level resistance to temephos (RR6.03) while the Valencia (RR 15.98), Spring Vale(RR 16.16), St. James (RR 50.62), and Sea Lots(RR 173.85) strains exhibited high levels ofresistance (Table 3). Resistance to temephos wasmore significant than to fenthion and malathion.

Percentage mortalities—classification ofresistance status

The RT value was established as 60 min forfenthion and 120 min for malathion and tempe-hos. The mean percentage larval mortalities ofdifferent strains at for each insecticide at 120 minare presented in Table 4. The ROCK strain wasas susceptible as the CAREC strain to temephosand fenthion; however, it was slightly moresusceptible to malathion with 100% mortality

compared to 97.50% mortality in the CARECstrain. The RT for malathion and temephos wasthe same as the 120-min end of experiment cutofftime.

Based on the WHO classification of resistance,none of the field strains were susceptible to any ofthe 3 insecticides (,98% mortality).

Fenthion: The Haleland Park (94% mortality)and San Fernando (85% mortality) strains wereincipiently resistant to fenthion while all otherstrains were resistant, with percentage mortalitiesranging from 21.25% (Sea Lots) to 78.75%(Spring Vale) (Table 4). The responses of 4 fieldstrains at the RT were significantly different fromthat of the CAREC reference strain, while at120 min, the responses of only the Sea Lots andSt. James strains were significantly different fromthe CAREC strain (P , 0.05).

Malathion: Four strains were incipiently resis-tant to malathion, with mortality levels rangingfrom 80% (Sea Lots) to 95.25% (Tobago). Theother 5 strains were resistant to malathion, withmortalities ranging from 56.25% (Haleland Park)to 77.50% (Curepe) (Table 4). The Haleland Parkand St. Clair strains were the only field strainsthat were significantly different from the CARECreference strain, in response to malathion (P ,0.05).

Temephos: The resistance to temephos wasevident in all strains (11.50–74.50% mortality),except Tobago which was incipiently resistant(90% mortality) (Table 4). At the end of theexperiments, the Haleland Park, San Fernando,and St. Clair strains were the only ones that had100% mortality to fenthion. Six strains were

Table 4. Percentage mortalities of Aedes aegypti larval strains to organophosphate insecticides.

Strain

Percentage mortality1,2

Fenthion (100 mg/100 ml)3

Malathion(200 mg/100 ml)3

Temephos(200 mg/100 ml)3

60 min4 120 min5 120 min4 120 min4

CAREC 100.00 6 0.00 100.00 6 0.00 97.50 6 2.89 100.00 6 0.00ROCK 100.00 6 0.00 100.00 6 0.00 100.00 6 0.00 100.00 6 0.00Haleland Park 94.00 6 4.24 100.00 6 0.00 56.25 6 5.97* 59.25 6 7.40Sea Lots 21.25 6 7.25* 76.75 6 6.70* 80.00 6 3.72 11.50 6 1.75*St. Clair 67.00 6 3.34 100.00 6 0.00 64.00 6 8.80* 13.25 6 3.86*St. James 39.50 6 3.59* 65.50 6 10.72* 71.75 6 4.53 18.75 6 4.21*Curepe 59.00 6 6.72 94.25 6 6.65 77.50 6 8.22 26.00 6 4.30*Valencia 48.50 6 6.01* 87.25 6 7.37 76.75 6 1.31 17.75 6 4.19*Spring Vale 78.75 6 3.15 97.50 6 5.00 83.25 6 2.28 18.75 6 2.69*San Fernando 85.00 6 3.24 100.00 6 0.00 93.50 6 2.22 74.50 6 4.70Tobago 35.50 6 5.81* 85.50 6 6.56 95.25 6 1.60 90.00 6 4.49

1 Percentage mortality 6 SEM.2 Asterisks indicate percentage mortalities significantly different from the Caribbean Epidemiology Center (CAREC) reference

susceptible strain (P , 0.05). Susceptibility status: susceptible, $98% mortality; incipiently resistant, 80–97% mortality; resistant,.80% mortality.

3 Diagnostic dosages of insecticides.4 RT, resistance threshold: time (minutes) in which 100% mortality was observed in the CAREC and/or Rockefeller (ROCK)

reference strains.5 Time at which experiment ended—same as RT for malathion and temephos.

DECEMBER 2010 ORGANOPHOSPHATE RESISTANCE IN TRINIDAD AND TOBAGO 407

significantly different to the CAREC referencestrain in response to temephos (P , 0.05)(Table 4).

DISCUSSION

The diagnostic dosages for each insecticideused in these experiments were established usingthe CDC protocol (Ocampo et al. 2000, CDC2004), with 200 mg/100 ml (2 mg/liter) fortemephos and malathion and 100 mg/100 ml(1 mg/liter) for fenthion. Results of this studyconfirm the continued prevalence of OP insecti-cide resistance in Trinidad and Tobago larvalstrains of Ae. aegypti. Earlier studies showed theSt. James strain to exhibit low resistance totemephos, with RR , 5, while the Curepe strainwas highly resistant to all 3 insecticides (RRs .10) (Rawlins and Ou Hing Wan 1995, Rawlins1998). In contrast, 10 years later, the presentstudy showed that the St. James strain is veryresistant to temephos, with RR (LT90) 50.62, andthe Curepe strain low RRs (LT90) to the 3insecticides. It is noteworthy that caution must betaken in comparing the results of the previousstudies with these present results as the bioassaymethods used were different. Rawlins (1998)studies used the WHO dose-mortality assayswhile the present study used the CDC larvalassay method. The method used in the presentstudy was based on a time-mortality assay thatmeasured the intoxication rate, that is, the time ittook for the insecticide to effectively reach itstarget (Brogdon et al. 1992). The WHO estab-lished diagnostic dosages of 0.02 mg, 1.0 mg, and0.05 mg/liter for temephos, malathion, andfenthion, respectively, after a 24-h exposureperiod (WHO 1980), but as confirmed by ourresults, these diagnostic dosages are not applica-ble to all mosquito species or populations (WHO1986, Brogdon and McAllister 1998, Ocampo etal. 2000).

In this assay, the slope of the regression line isreflective of the level of heterogeneity in themosquito populations. Susceptible strains wouldhave a steeper slope, indicative of the short timein which larval mortality is observed. A shallowslope on the other hand, could be caused byresistant individuals, which may have taken alonger time to die. Probit analyses that wereconducted generated high LT values, which, insome instances, exceeded the 120-min experimen-tal period. This is a result of the extrapolation ofthe probit analysis program.

Although it is important to know the dosage ofinsecticides required to kill a population, it isequally important to ascertain the time requiredfor a particular dosage to kill all the susceptiblesin a population, hence the practical usefulness ofthe CDC bioassay method. This method has theadvantages of using fewer test subjects and a

smaller amount of insecticides with resultsproduced in a short time. It is usually difficultto obtain enough mosquitoes from the field forexperiments and so the fact that results can beobtained from a small number of insects is ofpractical significance. In one study, Brogdon andMcAllister (1998) assayed several species ofmosquitoes to various insecticides, comparingthe WHO and CDC bioassays and a biochemicalassay. They found that the resistance levels by theWHO method were 50% lower for OPs and 90%lower for pyrethroids than in the other methodsand that the CDC assay method was moresensitive in detecting resistance levels at lowfrequencies.

Discrepancies were observed between theresistance status based on the mean percentagemortalities and WHO-based classification at theRTs and the levels of resistance based on theRRs. Strains that were susceptible by WHOcriteria ($98% mortality) were not necessarilysusceptible by RRs (RR , 1). However, therewas some agreement in that 25% of strains whichwere resistant (,80% mortality) also exhibitedhigh-level RRs (.10), and 73% of strains whichwere incipiently resistant (80–97% mortality) alsohad low-level RRs (1–5). The inconsistenciescould be due to differences in the experimentalmethods and also to the fact that probit analysesassume homogeneity in the populations beingtested, although field populations are usuallyheterogeneous (Brogdon and McAllister 1998).

The levels of resistance to the insecticides testedvaried among strains that are from differentgeographical locations within Trinidad and To-bago. The San Fernando strain was similar to theTobago strain in being incipiently resistant tofenthion, while the other strains were similar inresistance (,80% mortality). With respect tomalathion, the Spring Vale, San Fernando, andTobago strains were similar to the Haleland Parkstrain in showing incipient resistance; whereas,the other strains were resistant. All Trinidadianstrains were resistant to temephos, but theTobago strain was incipiently resistant. SinceTobago became reinfested with Ae. aegypti afterTrinidad (Chadee 1984, Chadee et al. 1984),Tobago mosquito populations have a shorterhistory of insecticide selection pressure than theTrinidad strains (Yan et al. 1998). This couldexplain the lower levels of resistance in theTobago strains to 2 insecticides. Conversely, theselection pressure exerted for vector control onthe island of Tobago may have been less intensethan on the bigger sister-island. The variations inresistance status among mosquito populationsfrom localities within the same or differentgeographical areas may be due to the intensityand frequency of vector control activities and thepassive movement of Ae. aegypti from differentareas, which may have resulted in a mixing of the

408 JOURNAL OF THE AMERICAN MOSQUITO CONTROL ASSOCIATION VOL. 26, NO. 4

gene pool with the introduction of foreign strains(Wallis et al. 1984).

Generally, selection pressure exerted by adulti-cides is lower than that exerted by larvicides.Rawlins (1998) showed evidence of low-levelresistance to malathion in 3 Trinidad populationsof Ae. aegypti. This, coupled with the fact thatboth malathion and fenthion are not routinelyused and the possibility of cross-resistance, couldexplain the generally lower levels of resistance tothese insecticides than to temephos. Organophos-phates act by inhibiting acetylcholinesterase,causing impairment of the nerve impulses acrossthe synapses and mutations in this target site canresult in resistance. The action of this mechanismcould be responsible for the development ofcross-resistance between fenthion, malathion,and temephos.

The emergence of resistance to temephos isimportant, as temephos is the only insecticide thatis certified by WHO as safe for use in potablewater, hence its widespread use in dengue controlprograms. It is not uncommon for field operatorsto over- or under-treat containers with insecti-cides because of their inability to accuratelyestimate the water volume and hence dosagerequired for each container type (Georghiou et al.1987, Chadee and Rahaman 2000). This practicemay have contributed to the development ofresistance as inaccurate dosages may exertselection pressure by allowing the survival ofresistant heterozygotes. If resistant populationsare continuously subjected to temephos, one canpredict that, with emerging higher levels ofresistance, temephos may become completelyineffective.

Results in the present study showed thatstatistically, mortality levels in some field strainswere not significantly different from the CARECreference strain in response to fenthion, malathi-on, and temephos (Table 4). The CAREC andROCK reference strains, with $98% mortalityobserved to all 3 insecticides, fit into the WHOclassification of a ‘‘susceptible’’ population.Therefore, we have confidence in our results thusvindicating the methodology used in classifyingresistance levels.

The fight against dengue fever in Trinidad andTobago is ongoing, but based on the results of thepresent study, the evidence of insecticide resis-tance is now available. It is clear that vectorcontrol programs must adopt alternative controlor management strategies in order to suppress theAe. aegypti populations and thus prevent dengueoutbreaks or epidemics. In addition, it is impor-tant to continually monitor the insecticide resis-tance status of the mosquito populations in orderto be able to make informed decisions on the useof insecticides and possibly suggest alternativeinsecticides or vector management strategies thatmay be more efficacious. For example, the

Integrated Vector Management approach canextend the life of insecticides by implementingsource reduction programs through community-based environmental sanitation, which can reducethe need for insecticides and/or reduce theselection pressure on resistant mosquito popula-tions and continued reliance on insecticides.

ACKNOWLEDGMENTS

The authors would like to sincerely thank thestaff of the Insect Vector Control Division andthe Port of Spain Health Corporation VectorControl Unit for their invaluable assistance in thecollection of Aedes aegypti eggs from the field.The authors are also indebted to the Centers forDisease Control and Prevention (CDC) for use ofits Entomology Laboratory and resources forconduct of the experiments and to ShannonMcClintock, also of the CDC, for statisticalassistance. Thanks also to Indira Omah-Maharajfor her useful comments on the manuscript and tothe Caribbean Epidemiology Center (CAREC)for graciously providing the CAREC referencesusceptible Ae. aegypti strain.

REFERENCES CITED

Brogdon WG, Beach RF, Barber AM, Cordon-RosalesC. 1992. A generalized approach to detection oforganophosphate resistance in mosquitoes. Med VetEntomol 6:110–114.

Brogdon WG, McAllister JC. 1998. Simplification ofadult mosquito bioassays through use of time-mortality determinations in glass bottles. J Am MosqControl Assoc 14:159–164.

CDC [Centers for Disease Control and Prevention].2004. Evaluating mosquitoes for insecticide resistance:a web based instruction [Internet]. Atlanta, GA:Centers for Disease Control and Prevention [accessedSeptember 5, 2004]. Available from: http://www.cdc.gov/ncidod/wbt/resistance/assay/larval/index.html.

Chadee DD. 1984. Aedes aegypti aboard boats at Portof Spain, Trinidad, West Indies (1972–82). MosqNews 44:1–3.

Chadee DD, Cornell NK, Le Maitre A, Ferreira SB.1984. Surveillance for Aedes aegypti in Tobago, W.I.Mosq News 44:490–492.

Chadee DD, Rahaman A. 2000. Use of water drums byhumans and Aedes aegypti in Trinidad. J Vector Ecol25:28–35.

Davidson G, Zahar AR. 1973. The practical implica-tions of resistance of malaria vectors to insecticides.Bull WHO 49:465–483.

Georghiou GP, Wirth M, Tran H, Saume F, KnudsenAB. 1987. Potential for organophosphate resistancein Aedes aegypti (Diptera:Culicidae) in the Caribbeanarea and neighboring countries. J Med Entomol24:290–294.

Mazarri MB, Georghiou GP. 1995. Characterization ofresistance to organophosphate, carbamate and pyre-throid insecticides in field populations of Aedesaegypti from Venezuela. J Am Mosq Control Assoc11:315–322.

DECEMBER 2010 ORGANOPHOSPHATE RESISTANCE IN TRINIDAD AND TOBAGO 409

Mekuria Y, Gwinn TA, Williams DC, Tidwell MA.1991. Insecticide susceptibility of Aedes aegypti fromSanto Domingo, Dominican Republic. J Am MosqControl Assoc 7:69–72.

Nathan MB. 1993. Critical review of Aedes aegypti controlprograms in the Caribbean and selected neighbouringcountries. J Am Mosq Control Assoc 9:1–7.

Ocampo CB, Brogdon WG, Orrego CM, Toro G,Montoya-Lerma J. 2000. Insecticide susceptibility inAnopheles pseudopunctipennis from Colombia: com-parison between bioassays and biochemical assays.J Am Mosq Control Assoc 16:331–338.

Rawlins SC. 1998. Spatial distribution of insecticideresistance in Caribbean populations of Aedes aegyptiand its significance. Pan Am J Pub Health 4:243–251.

Rawlins SC, Ou Hing Wan J. 1995. Resistance in someCaribbean populations of Aedes aegypti to severalinsecticides. J Am Mosq Control Assoc 11:59–65.

Rawlins SC, Ragoonanansingh R. 1990. Comparativeorganophosphorous insecticide susceptibility in Carib-bean populations of Aedes aegypti and Toxorhynchitesmoctezuma. J Am Mosq Control Assoc 6:315–319.

Reiter P, Amador MA, Conlon N. 1991. Enhancementof the CDC ovitrap with hay infusion for enhancedmonitoring of Aedes aegypti populations. J Am MosqControl Assoc 7:52–55.

Throne JE, Weaver DK, Chew V, Baker JE. 1995.Probit analysis of correlated data: multiple observa-tions over time at once concentration. J EconEntomol 88:1510–1512.

Wallis GP, Tabachnick WJ, Powell JR. 1984. Geneticheterogeneity among Caribbean populations of Aedesaegypti. Am J Trop Med Hyg 33:492–498.

WHO [World Health Organization]. 1980. Resistance ofvectors of disease to pesticides. Geneva, Switzerland:World Health Organization. Fifth report of theWHO Expert Committee on Vector Biology andControl. WHO Technical Report Series 1980; 655.82 p. Available from World Health Organization,Avenue Appia 20, 1211 Geneva 27, Switzerland.

WHO [World Health Organization]. 1981. Instructionsfor determining the susceptibility or resistance ofmosquito larvae to insecticides. WHO/VBC/81.807.Geneva, Switzerland: World Health Organization.

WHO [World Health Organization]. 1986. Resistance ofvectors and reservoirs of disease to pesticides. Geneva,Switzerland: World Health Organization. Tenthreport of the WHO Expert Committee on VectorBiology and Control. World Health OrganizationTechnical Report Series 737. 87 p. Available fromWorld Health Organization, Avenue Appia 20, 1211Geneva 27, Switzerland.

WHO [World Health Organization]. 1998. Test proce-dures for insecticide resistance monitoring in malariavectors, bioefficacy and persistence of insecticides ontreated surfaces. WHO/CDS/CPC/MAL/98.12. Gene-va, Switzerland: World Health Organization.

WHO [World Health Organization]. 2008. Dengue anddengue haemorrhagic fever [Internet]. FactsheetNo. 117. Geneva, Switzerland: World Health Orga-nization [accessed February 4, 2010]. Available from:http://www.who.int/mediacentre/factsheets/fs117/en/.

Yan G, Chadee DD, Severson DW. 1998. Evidence forgenetic hitchhiking effect associated with insecti-cide resistance in Aedes aegypti. Genetics 148:793–800.

410 JOURNAL OF THE AMERICAN MOSQUITO CONTROL ASSOCIATION VOL. 26, NO. 4