Pest Management Science Pest Manag Sci 60:685–690 (online: 2004)DOI: 10.1002/ps.855

Development of biodegradable aluminiumcarboxymethylcellulose matrices formosquito larvicidesNisha Mathew and Muthuswami Kalyanasundaram∗Vector Control Research Centre (ICMR) Pondicherry—605 006, India

Abstract: The use of ecofriendly biodegradable controlled-release formulations of mosquito larvicidescould reduce the frequency of application and losses due to degradation of the insecticide comparedwith conventional formulations. Among the 20 matrices developed by entrapping the organophosphorusmosquito larvicide, fenthion, in carboxymethylcellulose ionotropically cross-linked with aluminium ionswhich were studied for release profiles, two matrices, CRF3b and CRF5b, were found to be stable for16 and 14 weeks under simulated field conditions. The average concentration of fenthion released perweek ranged from 0.06 to 3.5 mg litre−1 for CRF3b and 0.09 to 2.72 mg litre−1 for CRF5b. Of these twoformulations, CRF3b was the more stable, maintaining the concentration of the active ingredient at thelevel required to effect mosquito control. The cumulative release of fenthion per pellet was 80% fromCRF3b and 72% from CRF5b. Based on the study with fenthion, two similar matrices for triflumuron, abenzoylphenylurea insect growth regulator, STAR3b and STAR5b were developed. These matrices werestable up to 16 weeks with the average concentration of triflumuron released per week ranging from 0.05 to3.44 mg litre−1 for STAR3b and 0.07 to 2.71 mg litre−1 for STAR5b. The cumulative release of triflumuronper pellet was 75% from STAR3b and 76% from STAR5b. From the results of this study under simulatedconditions, it is estimated that the application of four pellets of either fenthion or triflumuron per squaremetre of the breeding surface may play a useful role in controlling Culex quinquefasciatus Say in larvalhabitats for about 4 months. 2004 Society of Chemical Industry

Keywords: controlled-release formulation; mosquito larvicides; fenthion; triflumuron; sodium carboxymethylcel-lulose; crosslinking; aluminium sulfate

1 INTRODUCTIONInsect pests constitute a major detriment to thehealth and the economic well-being of much of thehuman population. Mosquitoes are known to transmitmany diseases such as malaria, filariasis, Japaneseencephalitis, dengue and dengue haemorrhagic fever.Management of mosquitoes and mosquito-bornediseases by following integrated control strategiestailored to local conditions will continue to requirethe selective use of insecticides. In order to achieve themaximum benefit, the active material must be madeavailable to reach the target organism over a sufficientduration.

Controlled-release systems have gained importancein the past three decades in the search for safer,more efficient and selective means of dispensing pest-control agents, as they can substantially reduce unde-sired environmental side-effects.1–3 Silicate capsuleshave been investigated as controlled-release devices of

chlorpyrifos for Aedes control.4 Slow-release formula-tions of temephos and Bacillus thuringiensis Berlinervar israelensis (Bti) on corncob and dried coconuthusk carriers have been developed for the control ofAedes aegypti L in automobile tyres.5 Nisha et al6 havedeveloped a controlled release formulation of fenthionusing powdered gummy parts of plant material knownas ‘Jiggieth’, cork powder and plaster of Paris as carriermaterials for the control of Culex quinquefasciatus Sayin cesspits, soak pits and septic tanks.

Among derivatives of cellulose, ethyl cellulose isused for microencapsulation of the herbicide norflu-razon to decrease its mobility through soil and protectit from photodegradation.7 Sodium carboxymethyl-cellulose is a water-soluble polysaccharide with excel-lent gel-forming properties which makes its solutionideal for many industrial applications. Earlier stud-ies in this laboratory resulted in the developmentof cupric8 and ferric9 ion cross-linked matrices of

∗ Correspondence to: Muthuswami Kalyanasundaram, Vector Control Research Centre (ICMR), Pondicherry—605 006, IndiaE-mail: kannananthi@yahoo.comContract/grant sponsor: BaytexContract/grant sponsor: Starycide(Received 3 July 2002; revised version received 22 May 2003; accepted 4 December 2003)Published online 19 February 2004

2004 Society of Chemical Industry. Pest Manag Sci 1526–498X/2004/$30.00 685

N Mathew, M Kalyanasundaram

carboxymethyl cellulose for the controlled releaseof fenthion. When the commercial grade of sodiumcarboxymethylcellulose used in the present study wascross-linked with cupric and ferric ions, the result-ing matrices were found to sink within two to threeweeks, leading to loss of effect against larvae at the sur-face. The present paper deals with the development ofcontrolled-release formulations of mosquito larvicides,fenthion (Baytex, an organophosphorus larvicide)and triflumuron (Starycide, a benzoylphenylureainsect growth regulator) entrapped in aluminium ioncross-linked carboxymethylcellulose matrices.

2 EXPERIMENTAL2.1 Preparation of controlled-release matricesof fenthion and triflumuronAn aqueous slurry was prepared by mixing therequired amount of sodium carboxymethylcellulose(commercial grade, as a 100 g litre−1 aqueous solution,viscosity 65 dPa s at 28 ◦C) with water (25 ml) followedby the addition of fenthion 825 g litre−1 EC (0.5 ml =0.413 g AI) and the required amount of filling agent,celite powder. The slurry was made homogeneousby stirring. The floating agent (expanded polystyrenebeads, 0.25 g) was then added and mixed well. Anaqueous solution of aluminium sulfate (LaboratoryReagent Grade; 0.05 M) with pH adjusted to 3.2 using500 g litre−1 sodium hydroxide solution was used asthe cross-linking medium. The slurry was transferredto special moulds for making the pellets. These pelletswere subjected to four different cross-linking periods,8, 16, 24 and 48 h, for ionotropic cross-linking with thegelling medium. After cross-linking for the requiredperiod, the pellets were washed with water to removethe remaining gelling solution on the surface of thematrices. The matrices were then dried either at roomtemperature to constant weight for one to two weeksor under controlled temperature in an oven at 50 ◦Cfor 1–2 days. Twenty different matrices were prepared(Table 1).

Two matrices of triflumuron, STAR3b andSTAR5b, similar to the matrices CRF3b and CRF5bfor fenthion, were developed, using 1 g of triflumuron480 g kg−1 SC formulation, equivalent to 0.48 g of theactive ingredient, for each matrix.

2.2 Determination of cross-linking density2.2.1 StandardisationThe aluminium ion concentration in the cross-linkedmatrices was analysed by a colorimetric method.10 Astandard aluminium solution (1 ml = 0.1 mg Al) wasprepared by dissolving 0.124 g of Al2(SO4)3 · 18H2Oin 100 ml of water. From the standard solution,0.5, 1.0, 1.5, 2.0, 2.5 and 3 ml portions weretransferred to six volumetric flasks (50 ml capacity).A blank solution without the aluminium sulfatewas also prepared. The contents of the flasks werediluted with water (25 ml) and then hydrochloric acid(5.0 M; 2 ml), gum arabic solution (5.0 M; 1 ml) and

Table 1. Composition of different formulations

MatrixCMC powder

(g)Celite powder

(g)Cross linking

time (h)

CRF1a 1.25 5 24CRF1b 1.25 5 48CRF1c 1.25 5 8CRF1d 1.25 5 16CRF2a 2.5 2.5 24CRF2b 2.5 2.5 48CRF2c 2.5 2.5 8CRF2d 2.5 2.5 16CRF3a 3.75 1.25 24CRF3b 3.75 1.25 48CRF3c 3.75 1.25 8CRF3d 3.75 1.25 16CRF4a 5 0.25 24CRF4b 5 0.25 48CRF4c 5 0.25 8CRF4d 5 0.25 16CRF5a 5 0 24CRF5b 5 0 48CRF5c 5 0 8CRF5d 5 0 16

ammonium acetate solution (3.5 M; 5 ml) were added.The solutions were diluted to 45 ml with distilledwater and ammonium auritricarboxylate (aluminon)solution (2 g litre−1; 2ml) was then added. All thesamples were mixed well and heated to 80 ◦C overa water bath for 10 min. The solutions were cooledto room temperature and the transmittance wasmeasured at 525 nm in a Spectronic-2000 UV-VISspectrophotometer (Bausch & Lomb).

2.2.2 Sample analysisThe matrices identified as having the maximumstability and uniform release of active ingredient inpreliminary screening, CRF3a–d and CRF5a–d, wereselected for the determination of extent of cross-linking. These matrices were prepared, powderedusing a pestle and mortar and 100 mg from eachsample was dispersed in 25 ml of water and treated asin Section 2.2.1 to estimate the percentage aluminiumcontent. The experiment was repeated twice.

2.4 Fenthion and triflumuron analysis by HPLCA LC-6A Shimadzu HPLC instrument equippedwith dual pump, DGU-6A online degasser, CTO-6A column oven, SPD-6AV UV detector, C-R6AChromatopac, SCL-6A system controller and aRheodyne injector was used for the analysis. Watersamples were analysed for fenthion concentrationwithout prior solvent extraction. The samples werefiltered through a 0.45-µm cellulose nitrate filterpaper. The mobile phase used was methanol +water (75 + 25 by volume) at 1.0 ml min−1 flowrate. A reverse phase Phenomenex Ultracarb 5 µm,C8 column (150 × 4.6 mm) was used. The columnoven temperature was kept at 40 ◦C. The analysiswas carried out at 254 nm. The injection volume was

686 Pest Manag Sci 60:685–690 (online: 2004)

Biodegradable aluminium carboxymethylcellulose matrices for mosquito larvicides

kept as 10 µl. The retention times for fenthion andtriflumuron were 5.0 and 4.8 min respectively.

2.5 Bioassay for larvicidal and insect growthregulating activity against Culexquinquefasciatus with water samples fromsimulated field trialTwenty-five early fourth-instar larvae of Cx quinquefas-ciatus were exposed to 250-ml water samples collectedfrom the test and control tanks in 500-ml beakers. Apinch of larval food consisting of yeast powder and dogbiscuit (1 + 1 by mass) was provided. Two replicateswere set up. Larval mortality was observed after 24 hof exposure. The experiments were conducted at 28(±2) ◦C and 70–80% RH.

Twenty-five early third-instar larvae of Cx quinque-fasciatus were exposed continuously to 250-ml watersamples collected from test and control tanks in 500-ml beakers. A pinch of larval food consisting of yeastpowder and dog biscuit (1 + 1 by mass) was pro-vided. Two replicates were set up. Larval mortality,pupal mortality and inhibition in adult emergencewere taken as the criteria of response and observationswere recorded daily until all the test larvae either diedor emerged. All experiments were conducted at 28(±2) ◦C and 70–80% RH.

2.6 Preliminary screening (Stage I)Forty beakers (2000 ml capacity) were filled withl000 ml of tap water. One pellet from each type ofmatrix was placed in each beaker. Two replicates foreach type of matrix were set up for monitoring thestability and release profile of the active ingredient.Water samples were taken from each beaker forfenthion analysis at daily intervals. Bioassays wereconducted daily in 500-ml beakers using 250 ml of thecollected water samples against fourth-instar larvae ofCx quinquefasciatus. The water was changed each dayafter taking the water sample for fenthion analysis andbioassay. The experiment was continued for ten weeksand the stability of the matrices assessed on the basisof visual observation of matrix erosion and the dailyrelease of fenthion.

2.7 Secondary screening (Stage II)Based on the observed stability of the ‘b’ series inthe preliminary screening, the five matrices CRF1b,CRF2b, CRF3b, CRF4b and CRF5b were preparedand studied for release profile and matrix stability for20 weeks using the same procedure as in Stage I.

2.8 Screening under simulated field conditions(Stage III)Two fenthion matrices, CRF3b and CRF5b, wereselected based on optimum stability and uniformrelease rate as shown in Stage II. A simulated fieldtrial was carried out with these matrices, each pelletcontained 1.0 ml of 825 g litre−1 EC (0.825 g AI). Sixcement tanks filled with 25 litres of water with a surface

area of 0.5 m2 were used for the study. Two pelletsof CRF3b or CRF5b were introduced into each oftanks 1 to 4. Tanks 5 and 6 were kept as controls.Water samples were collected at weekly intervals andfenthion analysis by HPLC and bioassay carried out.

A similar simulated field trial was also carriedout with STAR3b and STAR5b formulations oftriflumuron. Two pellets of STAR3b or STAR5b wereintroduced into each of tanks 1–4. Tanks 5 and 6were kept as controls. Water samples were collectedat weekly intervals and triflumuron analysis by HPLCand bioassay carried out.

3 RESULTS DISCUSSION3.1 Preliminary screening for selection ofsuitable matricesThe twenty matrices prepared were studied for stabilityand release of fenthion under laboratory conditions.In the preliminary screening, it was observed thatin all the formulations there was a higher releaseof fenthion in the first week, ranging from 2.08 to7.68 mg litre−1; this was the result of the burstingof the matrix due to swelling in water. Similarobservations had been made with matrices crosslinkedwith cupric8 and ferric9 ions. This is advantageous,as initial high release of fenthion could suppressthe immature population in the breeding habitatsas soon as the pellets are introduced for mosquitolarval control. The aluminium carboxymethylcellulosematrices with a cross-linking period of 48 h (‘b’series) were found to have optimal stability andfloating characteristics for more than 60 days andthe release of fenthion ranged between 0.98 and4.29 mg litre−1. On the basis of this information,further studies were conducted by selecting the fivematrices of ‘b’ series. The observed better stabilityof the ‘b’ series matrices compared with matrices a,c and d was confirmed by the maximum percentagecross-linking exhibited by these matrices in terms ofpercentage of aluminium content (Table 2). However,for the microbial pesticide Bacillus thuringiensis varisraelensis, of three different carboxymethylcelluloseconcentrations, 10, 20 and 30 g kg−1, the 10 g kg−1

treatment at a cross-linking period of 1 h at pH3.4 had been found sufficient to yield controlled-release beads with optimum stability and release

Table 2. Cross-linking density by CRF3 and CRF5 matrices

MatricesCross-linkingdensity (% Al)

CRF3a 0.169CRF3b 0.171CRF3c 0.145CRF3d 0.162CRF5a 0.249CRF5b 0.264CRF5c 0.224CRF5d 0.230

Pest Manag Sci 60:685–690 (online: 2004) 687

N Mathew, M Kalyanasundaram

properties.11 It is to be noted that aluminium cross-linked carboxymethylcellulose matrices had betterstability and floating characteristics than those cross-linked with cupric or ferric ions when using thecommercial grade of sodium carboxymethylcellulose.

3.2 Stage II evaluationFenthion release from the five matrices was monitoredfor 20 weeks under laboratory conditions. At the endof the study period, most of the matrices had disinte-grated to an extent of 98%, except the matrices CRF3band CRF5b, for which disintegration was only 75%.The average release of fenthion concentration fromthese matrices per day is shown in Table 3. The valuesof the initial high release of fenthion due to burstingare: CRF1b, 2.62 mg litre−1; CRF2b, 3.3 mg litre−1;CRF3b, 2.07 mg litre−1; CRF4b, 2.36 mg litre−1;CRF5b, 2.39 mg litre−1 and CRF6b, 5.12 mg litre−1.The average daily release of fenthion from the matri-ces showed sustained release in the case of CRF3b,where the concentration of fenthion released rangedfrom 0.5 to 1.3 mg litre−1 from week 2 to week14 of the study. In the case of CRF5b, it wasbetween 0.8 and 3 mg litre−1. In all other cases thefluctuation was high, due to the instability of thematrices. The cumulative release profile of fenthionfrom these six matrices (Table 3) showed the fol-lowing values; CRF1b, 214.2 mg; CRF2b, 252.1 mg;CRF3b, 188.8 mg; CRF4b, 246.9 mg and CRF5b,204.4 mg. CRF3b and CRF5b were thus considered toshow appropriate cumulative release and uniform dailyrelease of fenthion. Based on the observed matrix sta-bility and fenthion release profile, the matrices CRF3band CRF5b were selected for the simulated field trial.

3.3 Simulated field evaluation (Stage III)The cumulative release in terms of mg per pelletand average concentration of fenthion released (mglitre−1) per week (Figs 1 and 2) from the two matricesCRF3b and CRF5b under simulated field conditionsgave the following values: CRF3b, 657 mg (80%release); CRF5b, 593 mg (72% release). The matricesCRF3b and CRF5b were found to be stable for16 and 14 weeks under simulated field conditions.Similar monolithic matrices of fenthion preparedby cross-linking with copper had been found to bestable for 14 weeks in ponds in controlling Mansoniamosquito larvae.12 The average release of fenthionper week from CRF3b matrix ranged from 0.06to 3.5 mg litre−1 whereas that from CRF5b rangedfrom 0.09 to 2.72 mg litre−1. Between these twoformulations, CRF3b was comparatively more stableand could maintain the concentration of the activeingredient to a level to effect mosquito control.Maintenance of insecticide concentration is important,as rapid degradation of fenthion can occur in pollutedwater, with half-life ranging from 11.4 to 14.5 h.13

The cumulative release of fenthion per pellet in thecase of CRF3b was 80% and that of CRF5b was72%. This shows that the loss due to degradation byconventional spraying could be considerably reduced.The bioassay conducted with the water samplescollected at weekly intervals showed 100% mortalityagainst Cx quinquefasciatus during the entire period ofevaluation.

The study has thus identified two matrices, CRF3band CRF5b, which could be effectively used inmosquito larval control in various breeding habitatsat the rate of four pellets (each pellet containing

Table 3. Average daily release (ADR) and cumulative release per week (CR/W) of fenthion from selected matrices

CRF1b CRF2b CRF3b CRF4b CRF5b

WeeksADR

(mg litre−1)CR/W(mg)

ADR(mg litre−1)

CR/W(mg)

ADR(mg litre−1)

CR/W(mg)

ADR(mg litre−1)

CR/W(mg)

ADR(mg litre−1)

CR/W(mg)

1 2.62 18.35 3.30 23.13 2.07 14.50 2.36 16.52 2.39 16.712 1.18 26.64 1.45 33.29 0.76 19.82 1.14 24.48 0.86 22.743 1.20 35.03 1.22 41.80 0.67 24.55 0.96 31.21 0.87 28.824 1.19 43.30 1.37 51.39 0.49 27.95 1.07 38.67 0.84 34.695 1.15 51.37 1.82 64.12 0.57 31.94 0.82 44.42 0.99 41.596 1.30 60.61 1.50 74.61 0.69 36.73 1.14 52.38 1.01 48.637 1.27 69.42 2.33 90.95 1.07 44.19 1.59 63.54 1.82 61.368 1.43 79.41 2.96 111.7 0.91 50.56 1.49 73.98 1.24 70.019 1.38 89.05 3.24 134.4 0.97 57.33 1.64 85.45 1.56 80.92

10 1.15 97.09 2.83 154.2 0.86 63.34 1.64 96.92 1.46 91.1711 1.24 105.8 2.55 172.0 1.16 71.46 2.97 117.7 2.24 106.912 2.39 122.5 2.58 190.1 1.30 80.30 3.33 141.0 2.26 122.713 2.92 142.9 2.54 207.9 1.31 89.16 3.34 164.4 2.98 143.514 2.49 160.4 1.44 217.9 1.33 97.66 2.09 179.1 1.87 156.615 1.75 172.7 1.27 226.8 1.72 110.5 2.62 197.4 2.23 172.216 1.69 184.5 0.89 233.1 2.37 126.9 2.66 216.0 1.71 184.217 1.53 195.2 0.93 239.5 2.68 145.8 1.99 229.9 1.18 192.518 1.22 203.8 0.61 243.9 2.46 162.7 1.14 237.8 0.65 197.119 0.94 210.4 0.65 248.4 2.24 178.7 0.79 243.4 0.72 202.220 0.65 214.2 0.59 252.1 1.69 188.8 0.62 246.9 0.36 204.4

688 Pest Manag Sci 60:685–690 (online: 2004)

Biodegradable aluminium carboxymethylcellulose matrices for mosquito larvicides

0

200

400

600

800

1000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 160

2

4

6

8

10

Cum

ulat

ive

rele

ase

(mg

pelle

t-1)

Ave

rage

rel

ease

(m

g lit

re-1

) pe

r w

eek

Weeks

Figure 1. Release profile of fenthion from CRF3b: (°) cumulativerelease; (�) average release per week.

0

200

400

600

800

1000

Cum

ulat

ive

rele

ase

(mg

pelle

t-1)

0

2

4

6

8

10

Ave

rage

rel

ease

(m

g lit

re-1

) pe

r w

eek

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Weeks

Figure 2. Release profile of fenthion from CRF5b: (°) cumulativerelease; (�) average release per week.

0.825 g of fenthion) m−2 surface area of the breedinghabitat.

Based on studies with fenthion, two similar matrices,STAR3b and STAR5b, were prepared for triflumuronand studied for stability and release of active ingredientunder simulated field condition. The average weeklyrelease and cumulative release of triflumuron arepresented in Figs 3 and 4. In the case of triflumuron,both STAR3b and STAR5b were found to be stableup to 16 weeks with the average weekly release of theactive ingredient ranging from 0.05 to 3.44 mg litre−1

for STAR3b and 0.07 to 2.71 mg litre−1 for STAR3b.The cumulative release of triflumuron was 360 mg forSTAR3b and 366 mg for STAR5b, accounting for 75and 76% of that initially present.

The bioassay conducted with the water samplescollected at weekly intervals showed 100% larvalmortality during the entire period of evaluation. Undersimulated conditions, the release of four triflumuronpellets m−2, each pellet containing 0.48 g of the activeingredient, could prevent breeding of this speciesfor about 4 months. These formulations need to betested under field condition in different breedingsites to optimise the requirement of matrices perhectare of the breeding surface. The environment-friendly carboxymethylcellulose-based matrices areadvantageous because the ingredients are inexpensive,readily available and non-toxic.

0

100

200

300

400

500

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 160

2

4

6

8

10

Cum

ulat

ive

rele

ase

(mg

pelle

t-1)

Ave

rage

rel

ease

(m

g lit

re-1

) pe

r w

eek

Weeks

Figure 3. Release profile of triflumuron from STAR3b: (°) cumulativerelease; (�) average release per week.

0

100

200

300

400

500

Weeks

Cum

ulat

ive

rele

ase

(mg

pelle

t-1)

0

2

4

6

8

10

Ave

rage

rel

ease

(m

g lit

re-1

) pe

r w

eek

1 2 3 4 5 6 7 8 9 10 11 12 1613 14 15

Figure 4. Release profile of triflumuron from STAR5b: (°) cumulativerelease; (�) average release per week.

ACKNOWLEDGEMENTSThe authors are grateful to Dr PK Das, Director, Vec-tor Control Research Centre for valuable suggestionsand guidance. Authors wish to acknowledge M/s Bay-ers India Ltd, Mumbai for providing the insecticides,Baytex and Starycide for developing the controlledrelease matrices and financial support for carrying outthis work. Technical assistance offered by Shri (late)D Jayakumar, Shri S Srinivasan and Ms Anilaku-mari, staff of Product Development at the ChemistryDivision of the Centre is also gratefully acknowledged.

REFERENCES1 Cardarelli NF, Controlled release pesticide formulations, CRC

Press, Florida, USA p 210 (1976).2 Kydonieus AF, Fundamental concepts of controlled release, in

Controlled release technology: methods, theory and applications,Vol I, ed by Kydonieus AF, CRC Press, Florida, pp 1–19(1980).

3 Schreiber MM, White MD, Wing RE, Trimnell D andShasha BS, Bioactivity of controlled release formulations ofstarch encapsulated EPTC. J Controlled Release 7:237–242(1988).

4 Schandle VB, Sjogren RD and Thies C, Preliminary evaluationof silicate controlled release capsules for mosquito control,Proc 44th Annu Conf Calif Mosq Vector Control Assoc, pp95–97 (1976).

5 Novak RJ, Gubler DJ and Underwood D, Evaluation of slowrelease formulations of temephos and Bacillus thuringiensis var

Pest Manag Sci 60:685–690 (online: 2004) 689

N Mathew, M Kalyanasundaram

israelensis for the control of Aedes aegypti in Puerto Rico. J AmMosq Control Assoc 1:449–453 (1985).

6 Nisha G, Sujatha CH and Kalyanasundaram M, Controlledrelease formulation of mosquito larvicide with biodegradableingredients. Indian J Med Res 86:728–732 (1987).

7 Perez-Martinez JI, Morillo E, Maqueda C and Gines JM, Ethylcellulose polymer microspheres for controlled release ofnorfluazon. Pest Manag Sci 57:688–694 (2001).

8 Prasad MP and Kalyanasundaram M, Development of envi-ronmentally compatible controlled release formulations of amosquito larvicide. Indian J Med Res 93:51–54 (1991).

9 Prasad MP and Kalyanasundaram M, Iron(III) carboxymethyl-cellulose as swellable erodible matrix for the controlled releaseof a mosquito larvicide. J Controlled Release 22:167–172(1992).

10 McCoy JW, Chemical analysis of industrial water, ChemicalPublishing Co, New York, pp 51–53 (1969).

11 Cokmus C and Elcin YM, Stability and controlled releaseproperties of carboxymethylcellulose-encapsulated Bacil-lus thuringiensis var israelensis. Pestic Sci 45:351–355(1995).

12 Prasad MP, Rajendran G, Sabesan S and Kalyanasundaram M,Field evaluation of biodegradable controlled release formula-tion of fenthion against Mansonia mosquitoes. Southeast AsiaJ Trop Med Pub Health 28:208–211 (1997).

13 Harriths VR and Kalyanasundaram M, Studies on the stabilityof fenthion, an organophosphorus larvicide in relationto certain abiotic factors. Trop Biomedicine 12:23–29(1995).

690 Pest Manag Sci 60:685–690 (online: 2004)