BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, researchlibraries, and research funders in the common goal of maximizing access to critical research.

Assessment of Microencapsulated Formulations for Improved Residual Activity ofBacillus thuringiensisAuthor(s): Patricia Tamez-Guerra, Michael R. McGuire, Robert W. Behle, Baruch S. Shasha, and Luis J.Galń WongSource: Journal of Economic Entomology, 93(2):219-225. 2000.Published By: Entomological Society of AmericaDOI: http://dx.doi.org/10.1603/0022-0493-93.2.219URL: http://www.bioone.org/doi/full/10.1603/0022-0493-93.2.219

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, andenvironmental sciences. BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’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-commercial use. Commercial inquiriesor rights and permissions requests should be directed to the individual publisher as copyright holder.

BIOLOGICAL AND MICROBIAL CONTROL

Assessment of Microencapsulated Formulations for Improved ResidualActivity of Bacillus thuringiensis

PATRICIA TAMEZ-GUERRA,1, 2 MICHAEL R. MCGUIRE,1, 3 ROBERT W. BEHLE,1

BARUCH S. SHASHA,4 AND LUIS J. GALAN WONG2

J. Econ. Entomol. 93(2): 219Ð225 (2000)

ABSTRACT Bacillus thuringiensis Berliner is a highly efÞcacious bioinsecticide used to controllepidopteran pests in the Þeld. Unfortunately, it has limited residual activity on plants becausesunlight inactivates spores and crystals and they can be washed off by rain. To minimize loss ofactivity, formulations must contain UV protectants, stickers, or both. We tested '80 formulationsand determined optimal combinations of ingredients and spray drying conditions for improving B.thuringiensis residual activity after simulated rain and simulated sunlight. B. thuringiensis stability,after simulated sunlight (xenon light/8 h) and rain (5 cm/50 min), was improved using formulationsbased on lignin, corn ßours, or both, with up to 20% of the active ingredient, when compared withtechnical powder or Dipel 23 in laboratory assays. Two formulations, made with corn ßours orlignin 1 pregelatinized corn ßour (PCF), killed 51.6 and 75.3% of Ostrinia nubilalis (Hubner)neonates after rain, respectively, versus 27% for technical powder.When the insecticidal activitywastested after simulated sunlight, corn ßour-based formulations killed 78.5% of test larvae, and thelignin 1 PCF formulation killed 70.4%, in contrast to technical powder which caused an average of29% mortality. Formulations made with Dipel 23 rather than technical powder, caused 62.5%mortality (corn ßour-based formulations), and 72.3% mortality (lignin 1 PCF), versus 53.4% forDipel 23 after rain. When tested after simulated sunlight, formulations killed 95% of the larvae(average of both formulations) versus 82% for Dipel 23. In a Þeld test, formulations were appliedto cabbage and insecticidal activity was determined against Trichoplusia ni (Hubner) neonatesexposed to treated leaves. Insecticidal activity of the corn ßour-based formulations was comparableto Dipel 23 for 4 d after treatment, but was signiÞcantly better than Dipel 23 7 d after application.A lignin and PCF-based formulation showed signiÞcantly higher residual activity than Dipel 23, 4and 7 d after application.

KEY WORDS Bacillus thuringiensis, lepidoptera, biopesticide, spray dried formulations, lignin,pregelatinized corn ßour

AMONG THE MICROBIAL pesticides used to control insectpests, Bacillus thuringiensis Berliner has been the Þrstviable alternative to the use of chemical pesticides(Roush and Shelton 1997). This bacterium producesan insecticidal protein crystal during the sporulationprocess. Unfortunately, both spores and crystals areadversely affected by natural environmental condi-tions (Ignoffo 1992). Rain physically removes themfromthe foliage (McGuire andShasha 1990), and solarirradiation inactivates the insecticidal protein (Pozs-gay et al. 1987, Pusztai et al. 1991).

One way to improve the residual activity of B. thu-ringiensis is bydeveloping formulationswith solar pro-tective materials and rain resistant stickers. Naturalingredients such as pregelatinized cornstarch and

ßour, gluten, casein, and lignin have been studied asadjuvants to improve the residual activity of B. thu-ringiensis in response to rain and solar exposure(McGuire et al. 1996; Behle et al. 1997a, 1997b; Shashaet al. 1998). When natural ingredients were used asadjuvants, they improved the residual activity of B.thuringiensis in the Þeld. They worked as Þlm-formingmaterials while drying on the leaves, giving protectionto the active agent. The disadvantage of these mate-rials is that their efÞcacy depends on the concentra-tion of solids in the tank mix (e.g., a concentration of2% wt:vol for ßour/sucrose, 1% for gluten, or 0.5% forcasein or lignin;McGuire et al. 1996; Behle et al. 1997a,1997b; Shasha et al. 1998). Because the amount ofsolids is based on the spray volume, and the sprayvolume needed is different in each crop or sprayingsystem (from 50 to 500 liter/ha), these adjuvants havenot been readily accepted by industry.

Flour-basedmicrogranular formulationsweremadeby spray drying to overcome the limitations associatedwith the use of adjuvants as protective materials (Ta-mez-Guerra et al. 1996). The spray drying process tiedthe protective material and B. thuringiensis togetherinto a microgranular product that could be used in-

This article reports the results of research only. Mention of aproprietary product does not constitute an endorsement or recom-mendation by the USDA for its use.

1 Bioactive Agents Research Unit. USDAÐARS-NCAUR, 1815 N.University Street Peoria, IL 61604Ð3339.

2 Departamento de Microbiologõa e Inmunologõa. Fac. de CienciasBiologicas, UANL. A. P. 414. Sn Nicolas de los Garza, NL, Mexico66450.

3 To whom reprint request should be addressed.4 Retired.

dependently of the spray volume. The microgranuleswere easy to spray with conventional farm sprayingsystems (Tamez-Guerra et al. 1999). When micro-granular formulations were tested in laboratory andÞeld conditions, they showed good solar protectionand shelf life, but failed to avoid wash-off by rain(Tamez-Guerra et al. 1996, Tamez-Guerra et al. 1999).In addition, spray dried ßour formulations providedsolar protection when up to 10% of the active ingre-dient was added, but started losing protection whenthe amount of active ingredient was increased (Ta-mez-Guerra et al. 1996).

The current study was conducted to improve theresidual insecticidal activity of B. thuringiensis afterrain and solar exposure, with formulations based onnatural ingredients and use of the spray dry process.

Materials and Methods

Bacillus thuringiensis. Technical powder of B. thu-ringiensis variety kurstaki provided by Abbott Labo-ratories (North Chicago, IL) was used in formulationstested in laboratory bioassays and Þeld experiments.Thespore-crystalpowdercontained64,000 IU/mgandwas derived from the strain HD-1 used commercially.In addition, Dipel 23 WP, containing 32,000 IU/mg(Abbott Laboratories), was used to make selectedformulations and as a commercial bioinsecticide com-parison in Þeld tests.

Formulations. To improve B. thuringiensis residualactivity, .80 formulations were tested. Formulationswere prepared varying the pHs, ingredients, B. thu-ringiensis amount, and spray drying conditions. Allformulations were tested under simulated rain andsolar exposure. Matrix forming ingredients includedxanthan gum, gelatin, wheat-gluten, pregelatinizedcornstarch, pregelatinized corn ßour, nixtamalizedcorn ßour, and potassium lignate. Some formulationsincluded ingredients such as sugar, oil, and acid (lacticor citric). Many test formulations contained .1 poly-

mer. In addition, we tested formulations containing10Ð50% B. thuringiensis technical powder. After pre-liminary tests under simulated rain and simulated sun-light, 6 formulations were selected for more extensivetesting under laboratory andÞeld conditions. Selectedformulation ingredients included pregelatinized cornßour(PCF)(Flour965, IllinoisCerealMills, Paris, IL),nixtamalized corn ßour (NCF) (Maseca, Molinos Az-teca, S. A. de C. V., Guadalupe NL, Mexico), andpotassium lignate (made from kraft lignin, Westvaco,Charleston, SC) (Table 1). Formulations made with-out lignin were mixed following the procedures re-ported byTamez-Guerra et al. (1996). In formulationswith 2 corn ßours or a combination of corn ßour andlignin, potassium lignate was dissolved in water-iso-propanol, from 17.3 to 30% vol:vol (Table 1), followedby mixing with B. thuringiensis (technical or Dipel23), and calcium chloride. Formulations with ligninand PCF had the highest viscosity, and the watervolume was increased to avoid feed problems duringthe drying process.

Spray Drying Conditions. A Niro atomizer (Niro,Columbia,MD)wasused to spraydry all formulations.Spray drying conditions were inlet temperature of1308C, outlet temperature of 80 6 58C, air pressure of5.6 kg/cm2, and ßow rate of 18 ml/min. The formu-lation residence time for all formulations varied froma few seconds to 30 min.

Insect Cultures. Ostrinia nubilalis (Hubner) wasreared using procedures reported by Bartelt et al.(1990). Neonates were used in laboratory bioassays.Trichoplusia ni (Hubner)was reared according toGuyet al. (1985). ArtiÞcial diet contained wheat germ(37.5 g), casein (30.0 g), sucrose (30.0 g), Vander-zantÕs vitaminmix (12.5),Wesson salt (9.0 g), ascorbicacid (6.0 g), cellulose (4.5 g), methyl paraben (3.0 g),cholesterol (1.25), choline chloride (1.0 g), chlortet-racycline (0.25 g), agar (17.0), linoleic acid (1.7 ml),20%acetic acid (18ml), anddeionizedwater (800ml).Neonate T. ni were used in Þeld test bioassays.

Table 1. List of ingredients to microencapsulate B. thuringiensis

Ingredient, g

Formulation

NCF PCFNCF 1PCF

LigninNCF 1lignin

PCF 1lignin

NCF 75.0 75.0 35.0 Ñ 25.0 ÑPCF Ñ Ñ 35.0 Ñ Ñ 25.0Potassium lignatea Ñ Ñ Ñ 65.0 50.0 50.02-propanol, mlb Ñ Ñ 130.0 400.0 300.0 300.0B. thuringiensisc,d 21.5, 43 21.5, 43 20, 40 19.5, 39 21.5, 43 21.5, 43Water, ml 750.0 850.0 750.0 1,300.0 1,000.0 1,000.0PHd 4.5, 4.8 4.3, 4.6 4.5, 4.7 7.9, 8.2 7.6, 7.9 7.6, 8.0Moisture contentd,e 5.8, 5.4 7.5, 6.7 5.0, 5.4 6.2, 7.5 4.5, 5.5 9.1, 8.9Yield, %e,f 93, 91 99, 91 92, 89 86, 86 91, 87 88, 90

All formulations containedCaCl2 (10.0 g) at 10%wt: vol.NCF, nixtamalized cornßour,Maseca;GuadalupeN.L.,Mexico. PCF, pregelatinizedcorn ßour, Flour 965; Illinois Central Mills, Paris, IL.

a Potasium lignate; made by mixing 90.0g of kraft lignin (Westvaco, Charleston Heights, SC) in 200 ml of deionized water and 11.0 g ofpotassium hydroxide. The mixture was dried under the hood and sieved through 30-mesh sieve.

b 2-propanol; Fisher, Pittsburgh, PA.c B. thuringiensis technical powder (64,000 IU/mg) supplied by Abbott; Dipel 2X (32,000 UI/mg, Abbott, North Chicago, IL).d First value for Bt technical powder and 2nd value for Dipel 2X.e MARK 2 Moisture Analyzer; Denver Instruments, Denver, CO.f Yield calculated based on ingredients and collected product weights.

220 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 93, no. 2

Laboratory Bioassays. Bioassays on cotton leaveswere conducted with O. nubilalis following methodsreported by Behle et al. (1997a). Cotton plants, Gos-sypium hirsutum L. ÔDES 119Õ, were grown in a green-house in plastic pots (15 cm diameter), 2 plants perpot. When each plant had 5 or more fully expandedtrue leaves, the cotyledon and immature true leaveswere trimmed from the plant, leaving 10 large trueleaves. These remaining leaves were washed with tapwater, to remove soil or dirt that could interfere withlarval survival during the assay, and left to dry. Treat-ments, consisting of spray dried formulations added towater were applied to plants by spraying in a tracktype booth sprayer (DeVries Research Track SprayBooth, Hollandale, MN), as described by Behle et al.(1997a). Laboratory bioassay treatments consisted ofspray dried formulations, unformulated active agent(either technical powder or Dipel 23, depending onthe source of B. thuringiensis used for the spray driedformulations), and deionized water (control). All for-mulations were applied at a dose of 6,400 IU of B.thuringiensis/ml, '1/10th Þeld rate. Rainfastness andsolar protection assayswereperformedusing the samemethods reported by Tamez-Guerra et al. (1996).Rainfastnesswas tested by subjecting treated plants to5 cm of simulated rain ('50 min of spraying). Solarstability was tested by exposing plants for 8 h to sim-ulated sunlight from a Suntest CPS (Atlas, ) xenon-based light source (McGuire et al., Chicago, IL, 1996).After treated plants were exposed, 10 leaf disks (33mm diameter) were cut from the plants and placed inplastic petri plates (50 by 9 mm) containing a Þlterpaper disk to absorb excess moisture. Ten O. nubilalisneonates were placed in the dish and the dishes werecapped with a sealing lid. The dishes (10 per treat-ment) were incubated for 3 d in the dark at 288C andthe percentage mortality was determined for eachdish. Each experiment was done 3 times (replicates)and the total percentage mortality based on 100 larvaeper treatment per replicatewas used as the dependentvariable. Data were analyzed with analysis of variance(ANOVA), and means were separated using a pro-tected least signiÞcant difference (LSD) test (Statistix1994).

Field Test. Tests were conducted in August andOctober 1997 at Peoria, IL. These tests were designedto measure the effect of formulation ingredients onefÞcacy and residual activity of B. thuringiensis. Cab-bage plants, Brassica oleracea L., ÔBravoÕ, were trans-planted in Þeld plots '1 mo before the 1st Þeld test.Each plot was 1 row 5 m long. Cabbage plants werespaced 61 cm apart within rows and rows were 1 mapart. Treatments consisted of 6 selected formulations(PCF, NCF, lignin, PCF 1 NCF, PCF 1 lignin, andNCF 1 lignin, Table 1) made with B. thuringiensistechnical powder as active ingredient; Dipel 23 WP(used as a commercial B. thuringiensis standard tocompare with the formulations), and the untreatedcontrol. Rates of B. thuringiensis were constant for alltreatments at 43 1010 IU/ha applied in 160 literswaterper hectare. Each formulation was applied once to 40cabbage plants with a hand-held CO2 sprayer with 3

hollow cone nozzles directed at the row (Conejet 8X,Spraying Systems, Wheaton, IL), using a spray pres-sure of 3.5 kg/cm2. After each treatment was applied,the spray system was ßushed with water before ap-plication of the next formulation. Applications weredone between 0600 and 0700 hours. After application,formulations were allowed to dry before collectingleaf samples for assay. Forty leaves (replicates) wereremoved from different plants in each treatment at 0,1, 4, and 7 d after application, usually between 0800and 0900 hours. Leaf samples were taken to the lab-oratory for bioassay to determine insecticidal activity.Bioassays were performed in the laboratory as de-scribed before with cotton plants, but insecticidal ac-tivity was tested against T. ni neonates instead of O.nubilalis. Untransformed percentage insect mortalitywas analyzedusingANOVA.Meanswere separatedbyusing LSD tests, and contrasts were used for compar-isons among groups of treatment means.

Results

LaboratoryBioassays: FormulationswithTechnicalB. thuringiensis. PCF, NCF, and potassium lignatewere the polymers selected for further testing afterthe preliminary screening, because formulationsbased on these natural ingredients showed the highestresidual activity after simulated rain and simulatedsunlight. When PCF, NCF, and lignin-based formula-tions were assayed against O. nubilalis, laboratory ex-periments showed no differences in initial insecticidalactivity among formulated and unformulated B. thu-ringiensis. Bioassays performed to evaluate rainfast-ness and solar stability of the formulations made withtechnical B. thuringiensis showed that formulationscontaining PCF had signiÞcantly better residual ac-tivity than unformulated technical B. thuringiensis(Table 2). Formulations with the highest residual in-secticidal activity after rain were PCF and PCF 1NCF.Also, PCFandPCF1NCFformulations showedthe highest activity after exposure to simulated sun-light, causing 83.4 and 78.5% mortality, respectively,versus 33%mortality for unformulatedB. thuringiensis.

Table 2. O. nubilalis percentage mortality after feeding onflour-based, spray dried formulations of technical B. thuringiensisapplied to cotton after simulated rain (5 cm/50 min) and simulatedsunlight (xenon light, 8 h)

Formulation

% mortality

Postapplication treatment

None Rain Sunlight

Technical onlya 90.0a 27.0f 33.0efNCF 92.0a 35.6ef 63.0cPCF 87.0a 47.3de 83.4abNCF1PCF 95.0a 51.6cd 78.5bControl 5.0g 5.6g 3.3g

NCF, nixtamalized corn ßour; PCF, pregelatinized corn ßour.Means (n 5 3) followed by the same letter are not signiÞcantlydifferent. F 5 59.3; df 5 14, 42; P , 0.001. LSD (P 5 0.05), criticalvalue for comparison 5 15.3; standard error for comparison 5 4.8.

a Technical B. thuringiensis only (not spray dried).

April 2000 TAMEZ-GUERRA ET AL.: RESIDUAL ACTIVITY OF B. thuringiensis 221

Lignin 1 PCF formulation was the most rainfast,causing 75.3% mortality versus 27% mortality for theunformulated B. thuringiensis after rain exposure (Ta-ble 3). All lignin formulations provided higher solarprotection of B. thuringiensis (63Ð70.4% mortality)when compared with the unformulated technicalpowder (26% mortality).

Formulations with Dipel 23 WP. Spray dried for-mulations were made with Dipel 23 to determine ifPCF, NCF, or lignin could protect a commercial prod-uct from solar and rain degradation. Laboratory ex-periments showed no differences in initial insecticidalactivity among formulations and Dipel 23 (Table 4).All formulations lost activity during exposure to sim-ulated rain. Among formulations without lignin, noformulation had signiÞcantly better residual activityafter rain than Dipel 23 (Table 4). Among lignin-based formulations, only the formulation containinglignin 1 PCFhad signiÞcantly better rainfastness thanthe other treatments (72.3% mortality) (Table 4).None of the ßour or lignin-based formulations lost

activity during exposure to simulated sunlight,whereas Dipel 23 did lose activity (99% mortalityinitial activity, 82% mortality after sunlight).

Field Experiments. Two Þeld experiments, per-formed in August and October 1997, were used tocompare residual activity of the formulations madewith technical B. thuringiensis versus a commercialproduct (Dipel 23). Because a signiÞcant experimentdate 3 treatment interaction occurred (F 5 15.67;df 5 1, 2,472; P , 0.001), each experiment was ana-lyzed separately. Similarly, because signiÞcant sampleday 3 treatment interactions occurred (August: F 515.46; df521, 1,224;P,0.001;October:F566.27; df521, 1,224; P , 0.001), each sampling day was analyzedseparately. The sample day 3 treatment interactionalso suggests that formulations lost activity at differentrates.

Results showed signiÞcant differences in initial in-secticidal activity againstT. ni among formulations andDipel 23. In August, no rainfall occurred during theresidual activity test, whereas during the October test,4.2 cm of rain occurred between day 2 and 3 samples.Untreated controls averaged 13% mortality across thesamplingdates in the 1st test (Table 5), and3.4% in the2nd (Table 6). Lignin 1 PCF formulation showed thebest residual insecticidal activity in both experiments.The most dramatic differences were observed in the4thdafter application.Contrast analysisdemonstratedthat lignin-based formulations generally had signiÞ-cantlybetter(P,0.05) insecticidal activity than treat-ments of ßour-based formulations [August: t 5 6.46(day 1), t 5 11.92 (day 4), t 5 0.67 not signiÞcant (day8); October: t 5 9.61 (day 1), t 5 14.27 (day 4), t 514.27 (day 8)] or Dipel 23 formulations [August: t 53.34 (day 1), t 5 4.65 (day 4), t 5 3.03 (day 8);October: t 5 0.87 not signiÞcant (day 1), t 5 9.64 (day4), t 5 6.63 (day 8)]. Overall, results showed that B.thuringiensis formulations containing NCF, PCF, or

Table 3. O. nubilalis percentage mortality after feeding onlignin 1 flour-based spray dried formulations of technical B. thu-ringiensis applied to cotton after simulated rain (5 cm/50 min) andsimulated sunlight (xenon light, 8 h)

Formulation

% mortality

Postapplication treatment

None Rain Sunlight

Technical only 92.0a 27.0e 26.0eLignin 89.0ab 52.0cd 68.5bLignin1NCF 94.0a 35.3de 63.0bcLignin1PCF 92.0a 75.3b 70.4bControl 7.0f 4.6f 6.6f

NCF, nixtamalized corn ßour; PCF, pregelatinized corn ßour.Means (n 5 3) followed by the same letter are not signiÞcantlydifferent. F 5 60.6; df 5 14, 42; P , 0.001. LSD (P 5 0.05), criticalvalue for comparison 5 16.2; standard error for comparison 5 5.1.

a Technical B. thuringiensis only (nor spray dried).

Table 4. O. nubilalis percentage mortality after feeding onflour and lignin-based spray dried formulations of Dipel 2X WPapplied on cotton after simulated rain (5 cm/50 min) and simulatedsunlight (xenon light, 8 h)

Dipel 2X-based

formulations

% mortality

Postapplication treatment

No treatment Rain Sunlight

Dipel 2X only 99.0a 53.4ef 82.0bcNCF 96.3ab 48.4efg 98.5aPCF 99.0a 58.1de 95.8abNCF1PCF 99.3a 62.5de 95.3abLignin 98.0ab 52.8ef 97.9abLignin1NCF 99.0a 41.3fgh 97.0abLignin1PCF 98.0ab 72.3cd 94.4abControl 34.0ghi 22.0hi 26.2hi

NCF, nixtamalized corn ßour; PCF, pregelatinized corn ßour. Datarepresent 3 replications average. Means followed by the same letterare not signiÞcantly different. F 5 21.4; df 5 23, 48; P , 0.001. LSD(P 5 0.05), critical value for comparison 5 16.57; standard error forcomparison 5 8.24.

Table 5. T. ni percentage mortality after feeding on flour andlignin-based spray dried B. thuringiensis formulations or Dipel 2XWP applied on cabbage in the field in August 1997

Formulation

% mortality

Days after application

0 1 4 8

Dipel 2X 98.7ab 77.7bc 42.4c 23.5bNCF 88.4bc 68.5c 32.4de 25.2bPCF 84.4c 70.6c 26.9e 38.8aNCF1PCF 92.5b 81.7b 30.4de 42.3aLignin 99.8a 97.7a 57.5b 37.2aLignin1NCF 83.7c 75.4bc 36.7cd 28.1bLignin1PCF 99.5a 92.7a 79.5a 35.5abControl 15.4d 26.9d 5.4f 5.9c

F 138.0 42.8 55.8 13.9SEC 3.4 4.6 4.1 4.4CVC 6.6 9.1 8.1 8.6

NCF, nixtamalized corn ßour; PCF, pregelatinized corn ßour; SEC,standard error for comparison; CVC, critical value for comparison(P , 0.05). Data represent 40 replications average. df 5 7, 312 foreach day. Means followed by the same letter in the same column arenot signiÞcantly different, LSD (P 5 0.05).

222 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 93, no. 2

lignin 1 NCF had signiÞcantly lower residual insec-ticidal activity than lignin and lignin 1 PCF formula-tions 4 d after application. However, PCF and NCF 1PCF formulations showed no loss of activity betweenday 4 and 8 in the August Þeld test (no rain, Table 5).In the Þeld test performed in October, all formulationskept losing activity across days (Table 6).Theaverageofthe results among the 2 experiments is shown in Fig. 1.

Discussion

Most sprayable formulations of microbial pesti-cides have short residual activity in the Þeld and canbe improved with the addition of sunscreens orrainfast agents. Previous work in our laboratoriesidentiÞed several natural products that, when usedas spray tank adjuvants, signiÞcantly extended theactivity of B. thuringiensis. Unfortunately, the suc-cess of natural ingredients as adjuvants depends ona speciÞc amount of solids per volume when appliedin the Þeld (McGuire et al. 1996; Behle et al. 1997a,1997b; Shasha et al. 1998). One method of producingsprayable formulations with natural ingredients,which do not depend on the amount of solids pervolume, is by forming uniform microencapsulatedparticles using spray-drying techniques (Thies1994). The efÞcacy of B. thuringiensis microcapsulesmade of natural ingredients was 1st tested in a for-mulation containing cornstarch, ßour, and sugar(Tamez-Guerra et al. 1996). Improved residual in-secticidal activity after simulated solar irradiationwas demonstrated when spray dried formulationswere compared with the same levels of ingredientsused as adjuvants (Tamez-Guerra et al. 1996). Inaddition, microencapsulation of B. thuringiensis in anixtamalized corn ßour-based formulation wasshown to control Epilachna varivestis Mulsant pop-ulations better than unformulated powder as well as

the chemical insecticide carbaryl in the Þeld (Ta-mez-Guerra et al. 1999). Flour-based microencap-sulated formulations successfully protected B. thu-ringiensis from solar degradation and improved thepalatability and stability of B. thuringiensis but failedto avoid wash-off by rain. In addition, active ingre-dient rates had to be ,10% of total solids to maintainactivity after solar exposure (Tamez-Guerra et al.1996). The formulations presented herein weremade with '20% technical B. thuringiensis and ex-hibited improved solar stability and rainfastness.

This study evaluated the optimal combination ofnatural ingredients and spray drying conditions thatwould improve rainfastness and allow an increase inthe concentration of the B. thuringiensis. Approxi-mately 80 formulations were screened, testing naturalingredients (which showed rainfastness and solar pro-tection abilities when used as adjuvants) and varyingpH and spray-drying conditions. Spray drying param-eters included inlet and outlet temperatures and feedrates. Inlet temperature is not as critical as outlettemperature because the center of the microparticlewill never exceed the outlet temperature. A relativelyheat stable pathogen such as B. thuringiensis may sur-

Fig. 1. T. ni percentage mortality after feeding on cab-bage leaves treated with spray dried formulations of B. thu-ringiensis or Dipel 23 WP. Data represent the average from2 Þeld experiments. (A) Flour-based formulations. (B) Lig-nin-based formulations.

Table 6. T. ni percentage mortality after feeding on flour andlignin-based spray dried B. thuringiensis formulations or Dipel 2XWP applied on cabbage in the field in October 1997

Formulation

% mortality

Days after application

0 1 4 8

Dipel 2X 98.0ab 94.0a 22.3c 5.0dNCF 97.1ab 87.0b 22.0c 13.8cPCF 94.2b 75.2c 22.0c 3.8dNCF1PCF 95.2b 81.9b 20.0c 10.9cLignin 97.9ab 95.4a 31.5b 8.3dLignin1NCF 97.4ab 95.7a 38.1b 21.7bLignin1PCF 99.8a 96.6a 59.4a 27.7aControl 6.8c 5.2d 1.6b 0.5e

F 672.1 190.2 32.1 27.7SEC 1.7 3.1 4.2 2.5CVC 3.4 6.2 8.2 4.9

NCF, nixtamalized corn ßour; PCF, pregelatinized corn ßour; SEC,standard error for comparison; CVC, critical value for comparison(P , 0.05). Data represent 40 replications average. df 5 7, 312 foreach day. Means followed by the same letter in the same column arenot signiÞcantly different, LSD (P 5 0.05).

April 2000 TAMEZ-GUERRA ET AL.: RESIDUAL ACTIVITY OF B. thuringiensis 223

vive a higher outlet temperature than other ento-mopathogens. However, conditions across facilitieswill vary and it is important to determine optimalconditions based on pathogen type, spray dryer type,and formulation ingredients.

Natural ingredients tested included pregelatinizedstarch, casein, xanthangum, gelatin, gluten, andothers(McGuire and Shasha 1990; Behle et al. 1997a, 1997b;Oh et al. 1998). None of these ingredients, whichshowed wash-off resistance while used as adjuvants,provided rainfastness under simulated conditions asspray dried formulations. We found that PCF andlignin improved B. thuringiensis residual activity afterrain.

In previous tests of natural ingredients as adju-vants, a mixture of ßour/sucrose improved the lon-gevity of B. thuringiensis compared with a commer-cial product (McGuire et al. 1996). Flour/sucrose-based microgranule formulations reported beforewere made with equal parts of sucrose and corn-starch or ßour as the matrix forming material (Ta-mez-Guerra et al. 1996). Because sucrose dissolvesrapidly in water, microcapsules made with cornßour and sucrose did not maintain their integrityonce mixed in water for application to plants. Thus,after the active agent was released, the matrix nolonger provided solar protection. For this reason,the formulations selected in this study were madewithout sugar. Potassium lignate is soluble in water,and spray dried formulations quickly redissolve.However, CaCl2 or other divalent cation, crosslinkspotassium lignate during the spray dry process andthe resulting microgranule is not water soluble(Shasha et al. 1998). When we substituted acrosslinking agent (CaCl2) for the sugar in the lig-nin-based mixture before spray drying, the resultingmicrocapsules held their shapes in the spray tankand showed an improvement in rainfastness.

Because the activity ofB. thuringiensis is affected byexposure to sunlight, commercial products containingredients to increase B. thuringiensis solar stability.In laboratory tests we observed that formulationsmade with Dipel 23 had relatively higher residualactivity than formulations made with B. thuringiensistechnical power (Tables 2 and 3 versus 4). To deter-mine the relative effectiveness of the selected formu-lations in the Þeld, we tested formulations made withtechnical powder against Dipel 23. Results showedthat after 4 d, the formulation that maintained thehighest residual activity in both Þeld testswas lignin 1PCF, demonstrating its potential as an improved for-mulation.

However, although photoinactivation is consid-ered the major environmental factor affecting theresidual activity of B. thuringiensis among ento-mopathogenic microorganisms (Ignoffo 1992), rain-fall is an important factor to consider (McGuire etal. 1996). In our Þeld test, we expected to lose moreactivity in August because the light-hour per day islonger than in October for the Þeld location (Peoria,IL). We observed lower residual activity in theOctober test after the 4th d, because of rain (4.3 cm)

between day 2 and 3. Also, this rain can explain whyPCF and NCF 1 PCF based formulations kept losingactivity in the October Þeld test. These formulationsprovided solar protection as well as the lignin basedformulations in laboratory tests, and did not loseactivity between 4 and 8 d after application in theÞeld test in August.

In summary, the formulations presented in thismanuscript are prepared with natural ingredients,which are relatively inexpensive and commonly avail-able. In addition, the spray drying process and equip-ment are well known and broadly available. Finally,formulations can be made with up to 20% B. thurin-giensis.

Acknowledgments

We thank Monica Wetzel and Erica Bailey for their ex-cellent technical assistance, and Ricardo Gomez-Flores(FCB-UANL, Monterrey, NL, Mexico) and Leslie C. Lewis(USDA-ARS,Ames, IA) for their critical reviewof themanu-script. We also thank Randall L. Pingel (USDA-ARS, Fargo,ND) for support of a portion of this research.

References Cited

Bartelt, R. J., M. R. McGuire, and D. A. Black. 1990. Feedingstimulants for the European corn borer (Lepidoptera:Pyralidae): additives to a starch-based formulation for Ba-cillus thuringiensis Berliner. Environ. Entomol. 19: 182Ð189.

Behle,R.W.,M.R.McGuire,R.L.Gillespie, andB. S. Shasha.1997a. Effects of alkaline gluten on the insecticidal ac-tivity of Bacillus thuringiensis. J. Econ. Entomol. 90: 354Ð360.

Behle, R. W., M. R. McGuire, and B. S. Shasha. 1997b. Ef-fects of sunlight and simulated rain on residual activity ofBacillus thuringiensis formulations. J. Econ. Entomol. 90:1560Ð1566.

Guy,R.H., N.C.Leppla, and J. R.Rye. 1985. Trichoplusia ni,pp. 487Ð494. In P. Singh and R. F. Moore [eds.], Hand-book of insect rearing, vol. II. Elsevier, New York.

Ignoffo, C. M. 1992. Environmental factors affecting per-sistence of entomopathogens. Fla. Entomol. 75: 516Ð525.

McGuire, M. R., and B. S. Shasha. 1990. Sprayable self-en-capsulating starch formulations for Bacillus thuringiensis.J. Econ. Entomol. 83: 1813Ð1817.

McGuire, M. R., B. S. Shasha, C. E. Eastman, and H.Oloumi-Sadeghi. 1996. Starch and ßour based spray-able formulations: effect on rainfastness and solar sta-bility of Bacillus thuringiensis. J. Econ. Entomol. 89:863Ð869.

Oh, K. S., B. Y. Oh, S. S. Park, and J. K. Lee. 1998. Improve-ment of Bacillus thuringiensis wettable powder to en-hance adherence. RDA. J. Crop. Prot. 40: 57Ð62.

Pozsgay, M., P. Fast, H. Kaplan, and P. R. Carey. 1987. Theeffect of sunlight on the protein crystals from Bacillusthuringiensis var. kurstaki HD 1 and NRD 12: a Ramanspectroscopic study. J. Invertebr. Pathol. 50: 246Ð253.

Pusztai,M., P. Fast, L.Gringorten,H.Kaplan, T.Lessard, andP. Carey. 1991. The mechanism of sunlight-mediated in-activation of Bacillus thuringiensis crystals. BiochemistryJ. 273: 43Ð47.

Roush R. T., and A. M. Shelton. 1997. Assessing the odds:the emergence of resistance to Bt transgenic plants. Nat.Biotechnol. 15: 816Ð817.

224 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 93, no. 2

Shasha, B. S., M. R. McGuire, and R. W. Behle. 1998. Lignin-based pest control formulations. U. S. Patent No. 5,750,467.

Analytical Software. 1994. Statistix analytical software, ver-sion 4.0. Analytical Software, St. Paul, MN.

Tamez-Guerra, P.,C.Garcıa-Gutierrez,H.Medrano-Roldan,L. J. Galan-Wong, and C. F. Sandoval-Coronado. 1999.Spray dried microencapsulated Bacillus thuringiensis for-mulations for the control of Epilachna varivestis Mulsant.Southwest. Entomol. 24: 37Ð48.

Tamez-Guerra, P., M. R. McGuire, H. Medrano-Roldan, L. J.Galan-Wong, B. S. Shasha, and F. Vega. 1996. Sprayablegranule formulations for Bacillus thuringiensis. J. Econ.Entomol. 89: 1424Ð1430.

Thies, C. 1994. Microencapsulation: mini answers to majorproblem. TodayÕs Chemist at Work. November: 40Ð45.

Received for publication 4 June 1999; accepted 21 October1999.

April 2000 TAMEZ-GUERRA ET AL.: RESIDUAL ACTIVITY OF B. thuringiensis 225