United StatesDepartment ofAgriculture

AgriculturalResearchService

ARS–154

January 2001

Bt Cotton

Management of theTobacco Budworm-BollwormComplex

&

United StatesDepartment ofAgriculture

AgriculturalResearchService

ARS–154

January 2001

Bt Cotton

Management of theTobacco Budworm-BollwormComplex

&

D.D. Hardee, J.W. Van Duyn,M.B. Layton, and R.D. Bagwell

Abstract

Hardee, D.D., J.W. Van Duyn, M.B. Layton, and R.D.Bagwell. 2000. Bt Cotton & Management of theTobacco Budworm-Bollworm Complex. U.S. Depart-ment of Agriculture, Agricultural Research Service,ARS–154. 40 pp.

Preservation of Bt technology is critical for cottonproducers across the U.S. Cotton Belt because ofincreasing insecticide resistance and production costs.Frequent introduction of new transgenic cottonvarieties creates a need to continuously evaluate theircost-effectiveness and develop efficient plans for theirdeployment. This publication explains how Bt cottonis developed, how it controls insect pests, and how itcan most effectively be used in insect pest manage-ment. Restrictions and limitations to the use of Btcotton are discussed, such as insects’ development ofresistance to it and approaches to preserving thetechnology for long-term profits.

Audiences for the publication consist of research andextension entomologists in the public and privatesectors, consultants, and cotton producers.

Keywords: bollworm, Bt cotton, budworm, cotton,Heliothis virescens, Helicoverpa zea, refuge, resistancemonitoring, transgenic cotton

Mention of trade names, commercial products, orcompanies in this publication is solely for the purposeof providing specific information and does not implyrecommendation or endorsement by the U.S. Depart-ment of Agriculture over others not mentioned.

This publication reports research involving pesticides.It does not contain recommendations for their use nordoes it imply that uses discussed here have beenregistered. All uses of pesticides must be registered byappropriate state or Federal agencies or both beforethey can be recommended.

While supplies last, copies of this publication may beobtained at no cost from D.D. Hardee, USDA–ARS,P.O. Box 346, Stoneville, MS 38776, e-mail:dhardee@ars.usda.gov. Copies also available from F.L.Carter, National Cotton Council, P.O. Box 820285,Memphis, TN 38182.

Copies of this publication may be purchased from theNational Technical Information Service, 5285 PortRoyal Road, Springfield, VA 22161; telephone (703)605–6000.

Issued January 2001

The U.S. Department of Agriculture (USDA) prohibits discrimination in all its programs and activities on thebasis of race, color, national origin, sex, religion, age, disability, political beliefs, sexual orientation, or maritalor family status. (Not all prohibited bases apply to all programs.) Persons with disabilities who requirealternative means for communication of program information (Braille, large print, audiotape, etc.) shouldcontact USDA’s TARGET Center at (202) 720–2600 (voice and TDD).

To file a complaint of discrimination, write USDA, Office of Civil Rights, Room 326–W, Whitten Building,1400 Independence Avenue, SW, Washington, D.C. 20250–9410 or call (202) 720–5964 (voice and TDD).USDA is an equal opportunity provider and employer.

Acknowledgments

Advisory panel on resistance management plans:

C.T. Allen, University of Arkansas; J.S. Bacheler, NorthCarolina State University; J.H. Benedict, Texas A&MUniversity; J.R. Bradley, Jr., North Carolina StateUniversity; M.A. Caprio, Mississippi State University;T. W. Fuchs, Texas A&M University; D.D. Hardee,USDA–ARS; M.B. Layton, Mississippi State Univer-sity; B.R. Leonard, Louisiana State University; P.M.Roberts, University of Georgia; R.H. Smith, AuburnUniversity; and J.W. Van Duyn, North Carolina StateUniversity.

The following individuals provided invaluable inputduring multiple reviews of this publication’s contentand technical accuracy:

J.J. Adamczyk, Jr., USDA–ARS; C.T. Allen, Universityof Arkansas; G.L. Andrews, Mississippi State Univer-sity; J.S. Bacheler, North Carolina State University; J.H.Benedict, Texas A&M University; M.L. Boyd, Univer-sity of Missouri; J.R. Bradley, Jr., North Carolina StateUniversity; E. Burris, Louisiana State University; M.A.Caprio, Mississippi State University; F.L. Carter,National Cotton Council; B.L. Freeman, AuburnUniversity; T.W. Fuchs, Texas A&M University; L.D.Godfrey, University of California at Davis; F.L. Gould,North Carolina State University; M.A. Karner, Okla-homa State University; G.L. Lentz, University ofTennessee; B.R. Leonard, Louisiana State University;S.R. Matten, U.S. Environmental Protection Agency;W.J. Moar, Auburn University; J.W. Mullins, MonsantoCorporation; R.D. Parker, Texas A&M University; P.M.Roberts, University of Georgia; M.E. Roof, ClemsonUniversity; C.G. Sansone, Texas A&M University; R.W.Seward, University of Tennessee; J.E. Slosser, TexasA&M University; R.H. Smith, Auburn University; S.D.Stewart, Mississippi State University; N.P. Storer,North Carolina State University; and D.V. Sumerford,USDA–ARS.

The authors also express their appreciation to K.R.Ostlie, W.D. Hutchison, and R.L. Hellmich, authors of“Bt Corn & European Corn Borer.” 1997. University ofMinnesota, St. Paul, NCR Publication 602, for allowinguse of ideas and text.

This publication arose from a collaborative multistateand multiagency effort to provide timely guidelinesfor cotton producers, crop consultants, extension andindustry personnel, and agricultural agency adminis-trators on recommended ways to deploy Bt cottontechnology. Since this technology is changing rapidly,the publication will be updated as needed.

1

All Bt cotton plants contain one or more foreigngenes derived from the soil-dwelling bacterium,Bacillus thuringiensis; thus, they are transgenicplants.

The insertion of the genes from B. thuringiensiscauses cotton plant cells to produce crystalinsecticidal proteins, often referred to as Cry-proteins. These insecticidal proteins are effectivein killing some of the most injurious caterpillarpests of cotton, such as the larvae of tobacco

budworms and bollworms. This new technologyfor managing insect pests was approved forcommercialization in the United States by theU.S. Environmental Protection Agency (EPA) inOctober 1995 and is now available from severalseed companies in this country, as well as inmany other cotton-growing countries around theworld.

Cotton varieties containing the Cry1Ac Bt proteinprovide protection against three major U.S.

Bt CottonManagement of theTobacco Budworm-BollwormComplex

&

Tobacco budworm,Heliothis virescens(F.), larva

Bt cotton is one of the first cropprotection products from biotechnology.

2

cotton pests—tobacco bud-worms, bollworms, and pinkbollworms. Bt cotton alsoreduces survival of othercaterpillar pests such as beetarmyworms, cabbage loopers,cotton leafperforators, fallarmyworms, southern army-worms, and soybean loopers.The protection it providesagainst tobacco budworms,pink bollworms, and Europeancorn borers is greater than theprotection provided by themost effective foliar insecti-cides. Unfortunately theprotection it affords cottonagainst bollworms is generally

proper use and long-termpreservation of this valuabletechnology.

Frequent introduction ofnew transgenic cottonvarieties creates a need tocontinuously evaluate theircost-effectiveness anddevelop efficient plans fortheir deployment. A goal ofthis publication is to answerquestions about the technol-ogy by explaining how Btcotton is developed, how itcontrols insect pests, andhow it can be most effec-tively used in insect pestmanagement. Restrictionson and limitations to the useof Bt cotton are discussed,such as insects’ develop-ment of resistance to it andapproaches to preservingthe technology for long-term profits.

Tobacco budworm, Heliothisvirescens (F.), and bollworm,Helicoverpa zea (Boddie),cause more damage tocotton than any other insectpest in the U.S. Cotton Belt.The combined cost ofcontrolling these pests andthe losses they inflict oncotton production exceeds$300 million a year. Insecti-cides used against tobaccobudworms and bollworms

Tobacco budworm,Heliothis virescens(F.), moth

less than that provided byregistered insecticides.

The preservation of the Bttechnology is critical because (1)bollworms, tobacco budworms,and pink bollworms continue todevelop resistance to foliar-applied insecticides, (2) use ofinsecticides is tied to ecologicalconcerns, and (3) Bt genes arevaluable resources. This publi-cation focuses on how Bt cottonaffects the tobacco budwormand bollworm. It is intended toprovide information that willguide producers, research andextension entomologists, con-sultants, and industry in the

3

he insect-disease-causing organismBacillus thuringiensis (Bt) is a naturally

occurring soilborne bacterium foundworldwide. A unique feature is its produc-tion of crystal-like proteins that selectivelykill specific groups of insects and otherorganisms. When the insect eats these Cry-proteins, its own digestive enzymesactivate the toxic form of the protein. Cry-proteins bind to specific receptors on theintestinal walls and rupture midgut cells.Susceptible insects stop feeding within afew hours after taking their first bite, and,if they have eaten enough toxin, die within2 or 3 days.

Different Bt strains produce different Cry-proteins, and there are hundreds of knownstrains. Scientists have identified morethan 60 types of Cry-proteins that affect awide variety of insects. Most Cry-proteinsare active against specific groups of in-

often create other problems, such as higherpopulations of beet armyworms and cottonaphids and an increased pesticide load in theenvironment. Frequent exposure of insect pests toinsecticides results in the development of insecti-cide resistance, which reduces the overall effec-tiveness of available insecticides, increases croplosses, and leads to higher pest control costs andlower farm profits.

The severity of tobacco budworm and bollworminfestations and resistance to synthetic insecti-

Bollworm larvafeeding in boll

cides vary across the Cotton Belt, both betweenand within the states. Because of this variationand the price of the technology, not all areas ofthe Cotton Belt are able to economically justifythe use of Bt cotton. However, where insectinfestations are severe, Bt cotton offers a newmanagement tool for producers, helps ensureagainst yield loss in the presence of heavy infesta-tions of insecticide-resistant tobacco budworms,and aids in reducing bollworm damage.

Bt—What Is It?

T

4

sects, such as the larvae of certain kinds of flies,beetles, and moths. For example, Colorado potatobeetle larvae are affected by Cry3A proteins;Cry1Ac is used against tobacco budworms; andEuropean corn borers can be killed with Cry1Ab,Cry1F, Cry1Ac, and Cry9c proteins. Other Cry-proteins are active against mosquito larvae, flies,or even nematodes. Some Cry-proteins have been

used for more than 30 years in various liquid andgranular formulations of natural Bt insecticides,mainly to control caterpillars on a variety ofcrops. The Bt cotton varieties presently usedagainst tobacco budworms, bollworms, andcertain other caterpillars produce the Cry1Acprotein.

overcomes most of the aforementioned limita-tions. The plant-produced Bt proteins are pro-tected from rapid environmental degradationsince they are not directly exposed to the environ-ment. Incomplete coverage is not usually aproblem because the plants produce the proteinsin all tissues where larvae feed, thus ensuringthat the larvae will eat the Cry-protein. Theprotein is always present whenever newlyhatched larvae feed, eliminating the timingproblem associated with foliar application. Theresult is that Bt cotton has a built-in system thatefficiently and consistently delivers Cry-toxins tothe target pests from the time a newly hatchedlarva takes its first bite (fig. 1).

Bt cotton offers a vastly improved method fordelivering Cry-insecticides to target insects,compared to traditional Bt sprays. Bt cotton mayalso be considered a form of host plant resistance,in that the Cry-protein trait is carried in theplant’s genes, as is traditional plant resistance toinsects.

The Why’s and How’s of CreatingBt Cotton

ioinsecticides like Bt that are sprayed oncrops may perform as well as synthetic

insecticides in very limited situations, but theperformance of Bt insecticides has been inconsis-tent in many instances.

The erratic performance in cotton is attributed to

four reasons:

• The toxin is rapidly degraded by ultraviolet

light, heat, high leaf pH, or desiccation.

• Caterpillars must eat enough treated plant

tissue to get a lethal dose of the toxin, since

the toxin has no contact effect.

• The sites where tobacco budworms and

bollworms feed are difficult to cover with the

foliar-applied sprays.

• Bt Cry-proteins are less toxic to older larvae.

A cotton plant modified to produce Cry-proteinwithin the plant tissues that caterpillars eat

4

B

5

Figure 1. Mode of action for Bt toxin after eaten by a tobacco budworm larva.Modified with permission from Ostlie et al. 1997.

6

Biotechnologists created Bt cotton by insertingselected exotic DNA, from a Bt bacterium, intothe cotton plant’s own DNA. DNA is the geneticmaterial that controls expression of a plant’s or ananimal’s traits. Following the insertion of modi-fied Bt DNA into the cotton plant’s DNA, seedcompanies moved the Cry-protein trait into high-performance cotton varieties by traditional plantbreeding methods. Agronomic qualities for yield,harvestability, fiber quality, and other importantcharacteristics were preserved at the same timethe Cry-protein gene was added to commercialvarieties.

The three primary components of the geneticpackage inserted into cotton DNA include:

Protein gene. The Bt gene, modified for im-proved expression in cotton, enables the cottonplant to produce Cry-protein. The first varieties ofBt cotton produced in the United States containedone Cry-protein gene—Cry1Ac. Other varietiescontain a “stacked” gene complex, for example—one gene for insect control (Cry1Ac) and onegene to protect the cotton from application of theherbicide glyphosate. Future cotton varieties mayinclude these genes, other genes that allow theplant to produce different Cry-proteins, or insecti-cidal proteins from sources other than Bt. Thereare many possible combinations for crop im-provement traits.

Promoter. A promoter is a DNA segment thatcontrols the amount of Cry-protein produced andthe plant parts where it is produced. Some pro-moters limit protein production to specific parts ofthe plant, such as leaves, green tissue, or pollen.Others, including those used in Bt cotton andcertain Bt corn varieties, cause the plant toproduce Cry-protein throughout the plant. Pro-moters can also be used to turn on and turn offprotein production. Current varieties of Bt cottonproduce some Bt protein throughout the growingseason.

Genetic marker. A genetic marker allows re-searchers to identify successful insertion of agene into the plant’s DNA. It also assists plantbreeders in identifying and developing new cottonlines with the Bt gene. A common marker is anherbicide tolerance gene linked to the Bt gene.Following a transformation attempt to place the Btand marker gene into the plant’s DNA, plants aretreated with herbicide. Plants that were success-fully transformed have the Bt gene and theherbicide resistance gene and will survive herbi-cide treatment; plants without the marker gene,and hence without the linked Bt gene, will bekilled by the herbicide.

This genetic package—a Bt gene plus a promoterand marker—can be inserted into cotton plantDNA through a variety of plant transformationtechniques. Transformed plants may be affectedby the genetic package, as well as the location ofthe new genes in the plant DNA. The insertionsite may affect Bt protein production and otherplant functions as well. So biotechnology compa-nies carefully scrutinize each transformation toensure adequate production of Bt protein and tolimit possible negative effects on agronomictraits.

Following a successful transformation, plants areentered into a traditional backcross breedingprogram with the variety chosen to receive theforeign Bt gene package. The final product, a Btcotton variety, is developed after four or fivebackcross generations. Even though the newtransgenic Bt cotton variety may be named afterthe parent variety, agronomic qualities can beconsiderably different.

7

The Safety of Bt and Bt Cotton

Controlling Tobacco Budworms andBollworms: Does Bt Cotton Do the Job?

hen compared with otherinsecticide management

practices, Bt cotton dramaticallyimproves the control of tobaccobudworms and, to a lesser extent, thecontrol of bollworms. For example,chemical insecticides, including somenew chemical classes, often controlfrom 70 to 95 percent of a susceptibletobacco budworm population. Asindicated in table 1, the level oftobacco budworm control achievedwith Bt cotton can be very dramatic.Bt cotton varieties may provide morethan 98 percent control of tobaccobudworm throughout the growingseason. For growers whose cotton isplagued by high densities of insecti-cide-resistant tobacco budworms, Btcotton is a very welcome technology.

Bollworm/budwormeggs on a leaf. Courtesyof S. Stewart.

B

W

efore registering Bt cotton, EPA revieweddata on Bt insecticides that had been accu-

mulated for decades. Bt Cry-proteins were foundto be toxic only to certain insect groups and tohave no known negative effects to humans,domestic animals, fish, wildlife, or other organ-isms. EPA exempted Bt-produced Cry-proteins

from the requirement of tolerance in food becauseof their history of safety and because they de-grade rapidly in the environment. These Cry-proteins are considered among the safest andmost environmentally friendly insecticidesknown.

8

Bt cotton is less effective against bollworms. Still,it can eliminate as many as 60 to 90� percent ofthe bollworms infesting a cotton field (table 1).When there are high numbers of bollworms on Btcotton during the bloom stage, growers may needto apply one or more supplemental insecticidetreatments to prevent economic damage. This hasbeen well documented through field research.Fortunately, the bollworm can still be managedmore effectively and inexpensively with currentlyavailable insecticides—except perhaps in SouthCarolina where resistance to insecticides has beendetected.

Bollworm moths lay their eggs on cotton plantterminals, leaves, buds, and flowers. As the eggs

hatch, the larvae may move into open bloomswhere they feed on flower parts, includingpollen, that are known to have a lower level ofCry-protein than other plant parts. This feedingon less toxic parts may result in lower mortality.The bollworm in addition is naturally moretolerant of Cry1Ac Bt protein than the tobaccobudworm. These two factors help explain whymore bollworms than budworms survive on Btcotton. Moreover, Cry-protein expression in Btcotton decreases about 80 days postplanting,which may allow higher survival of bollworms.This late decline may also reduce Cry-proteineffectiveness against other insect pests that occurlater in the growing season and feed on matureleaves.

Table 1. Survival of tobacco budworms, bollworms, and fall armyworms onBt and non-Bt cotton genotypes

Percent survival

Insect 1994 1995

Tobacco budwormon Bt cotton leaf 1 0on Bt cotton square 2 0

on non-Bt cotton leaf 86 84on non-Bt cotton square 69 67

Bollwormon Bt cotton leaf 7 23on Bt cotton square 5 4

on non-Bt cotton leaf 80 74on non-Bt cotton square 63 52

Fall armywormon Bt cotton leaf 61 76on Bt cotton square 33 25

on non-Bt cotton leaf 76 92on non-Bt cotton square 45 42

Source: Modified and reprinted with permission from Jenkins et al. 1997.

8

9

obacco budworms and bollworms are not theonly insect pests that attack cotton. Unfortu-

nately, the Cry1Ac protein has essentially noeffect on many of them. Pests that Bt cotton doesnot directly affect include boll weevils, cottonaphids, cotton fleahoppers, cutworms, spidermites, stink bugs, tarnished plant bugs, thrips,and whiteflies. In some caterpillar species, Btcotton may provide only 10 to 50 percent control.This partial suppression may be cause for concernin the later years of an insect resistance manage-ment program because it does not provide a high-dose strategy (see Glossary) for insects such asbeet armyworms, fall armyworms (table 1),southern armyworms, soybean loopers, andyellowstriped armyworms. Bt cotton at this timeprovides good to excellent control of cabbageloopers, cotton leafperforators, European cornborers, salt marsh caterpillars, and cotton squareborers. Future varieties of Bt cotton may producedifferent Cry-proteins or other new toxins thatwill control a wider variety of pests.

The effect of Bt cotton on some insect pests maybe indirect. For example, Bt cotton does notdirectly affect the cotton aphid, but reductions ininsecticide use against tobacco budworms andbollworms allow more of the aphid’s naturalenemies to survive, and they in turn reduce aphidnumbers.

On the other hand, reducing the amount of foliarinsecticide may also allow other pests normallycontrolled by the insecticides to become moreabundant. Boll weevils, stink bugs, and plantbugs, for example, have by chance been con-trolled by foliar sprays applied against tobaccobudworms or bollworms. So, reduction in the useof insecticides on Bt cotton has allowed thesepests to increase in some areas. Offsetting thisdisadvantage, however, is Bt cotton’s usefulnesswhere the boll weevil has been eradiated.

Tarnished plant bugnymph

Southern armyworms.Courtesy of R. Smith.

T

Controlling Other Insects

10

any studies have shown that Cry1Ac inBt cotton is highly selective because it

kills only certain caterpillar species. Bt cottonhas minimal or no effect on beneficial insects,including honey bees, lady beetles, spiders, big-eyed bugs, pirate bugs, and parasitic wasps.However, laboratory research has shown thatCry1Ab protein can indirectly affect greenlacewing larvae that eat Bt-killed caterpillars. Itis not known if Cry1Ac in Bt cotton has a

similar effect on lacewing larvae. Theoretically, Btcotton may indirectly lower the general abundanceof some beneficial insects, since it causes caterpillarpopulations to decline, resulting in less food for thepredators, parasites, and the pathogens that attackthem. Offsetting this effect is the positive influencegained from reducing conventional broad spectruminsecticide use in Bt cotton. The overall balance ofthese contrasting influences is currently unknownand is difficult to predict.

Does Bt Cotton AffectBeneficial Insects?

Big-eyed bug

Lacewing larva

M

11

Table 2. Yield comparisons between Bt and non-Bt cotton

Yield (lb lint/acre)

Bt cotton non-Bt cottonYear (unsprayed) (sprayed)

1994 1369 13921995 1465 1425

Source: Reprinted with permission from Jenkins et al. 1997.

s is the case with most new technology, Btcotton offers value to the cotton farmer in

specific circumstances (table 2). Information oneconomic benefits is limited due to the short timethe technology has been available and the manynew cotton varieties introduced each year. Re-cently, the technology fee and the seed cost for Btcotton have decreased, which has affected theeconomics of growing it. Yield data are availablefrom federal and university entomologists andagronomists, as well as seed companies, in nearlyevery cotton-growing state.

Comparisons of the Bt and non-Bt cotton varietiesgenerally show that Bt cotton offers an economicadvantage in instances where effective insecticid-al control of certain caterpillar pests is difficult toachieve or is very costly.

Examples of such situations

include

• insecticide-resistant to-

bacco budworms or boll-

worms

• high populations of sus-

ceptible tobacco budworms

or bollworms, such as in

outbreak seasons or during

the initial phases of boll

weevil eradication

A • situations where a properly timed and applied

insecticide management program cannot be

achieved, such as in fields that do not allow

the proper operation of air or ground sprayers,

in remote fields, or in cases where cotton

acreage exceeds the amount of equipment or

personnel dedicated to insecticidal control

• situations where insecticidal control is exces-

sively costly (for example, more than $40/

acre), as may be the case when high infesta-

tions are coupled with newer, expensive

insecticide products—even though the insecti-

cide program may protect the crop

• situations where eliminating early tobacco

budworm sprays allows survival of beneficial

insects that reduce the risk of pest infestations

The Value of Bt Cotton to theCotton Farmer

12

associated with higher insecticide use (as with

beet armyworms or aphids).

These situations suggest that income from re-duced insecticide input, along with higher yieldsas a result of less insect damage, offset the tech-nology fee and favor Bt cotton. However, ifinfestations of tobacco budworms or bollwormsare low, or the yield of the Bt variety used is low,the technology fee may exceed the value of the Bttoxin. Also, a conventional insecticide sprayprogram may allow the farmer to grow certainhigh-yielding cotton varieties that perform betterthan available Bt varieties. Some economic com-parisons show little or no economic advantage tousing Bt cotton (table 3), whereas the economicreturns in other circumstances have been positive(table 4). Even in the areas that economicallyfavor Bt cotton, however, there are often situa-tions where high-yielding, non-Bt varietiesgrown under conventional spray programsprovide equal or greater economic returns thanBt varieties.

Regardless of whether or not a particular varietycontains the Bt gene, yield potential continues tobe the primary consideration when selecting acotton variety. The Cry1Ac gene in Bt cotton isonly one of thousands of genes that affect theyield and other characteristics of different cottonvarieties. Growers who are selecting cottonvarieties should carefully consider the traits ofeach, including yield performance, relativematurity, fiber quality, ability to withstand ad-verse weather, and harvestability.

Bt cotton may offer value to the cotton farmer inways that are hard to measure by short-termeconomic comparisons. Using insecticides in-volves complying with certain laws, such as

worker protection and pesticide label restrictions.Compliance often makes the grower’s job moredifficult and increases the risk of consequencesarising from noncompliance. Insecticide use insensitive areas—next to schools, fish ponds,dwellings, medical facilities, roads—can be aconcern to the grower and his or her neighbors. Iflegal and social risks are a concern, Bt cotton mayhave value to the grower by reducing these risks.

Adopting Bt cotton may result in a more efficiententerprise, while maintaining a high level ofinsect pest control. Insect management withinsecticides can be time-consuming and involve asignificant amount of labor and equipment. If Btcotton reduces the insect control burden anddecreases the need for labor and equipment,these resources may be diverted to other farmobligations.

Bt cotton may reduce potential resistance to foliarinsecticides in tobacco budworms andbollworms. Since resistance genes selected as aresult of one insecticide may be eliminated by adifferent and unrelated insecticide, the rotation oftoxins, including Bt Cry-protein, may be able toslow the selection of genes for resistance to anysingle toxin. Farmers will be obliged to include anon-Bt refuge to accompany any Bt cottonplanted (see refuge section, p. 23). Use of asprayed refuge provides an excellent opportunityto use newer insecticide classes, as well aseffective older chemistries, as aids in reducingdevelopment of insecticide resistance in tobaccobudworms and bollworms and in adoptingresistance management plans for otherchemistries. Maintaining effective refuge areasthat are close to Bt cotton also helps slowresistance to the current Cry1Ac toxin cottonvarieties and may decrease resistance to other Bttoxins as they are introduced.

13

Table 3. Cost of Bt cotton vs. non-Bt cotton in North Carolina, 1999

Expense Bt cotton Non-Bt cotton

Technology fee* $19.14 $ 0.00

Control costs 5.78 @ 0.76 applications/season 19.88 @ 2.65 applications/season

Damage† 0.00 @ 4.41% damage 8.50 @ 5.5% damage

Extra scouting 3.00 0.00

Total $27.92 $28.38

* Projected average cost; varies by seed rate and row spacing.† Difference in late-season bollworm damage under grower conditions (N=614 fields, 1996–1998).

Source: J. Bacheler, unpublished data.

13

Table 4. Cost of Bt cotton vs. non-Bt cotton in Mississippi, 1995–97

Expense Bt cotton Non-Bt cotton

Number of sprays* 6.7 11.7

Control costs $61.48 $68.15

lb lint/acre 876 789

Economic advantage $63.22 —

* Average of 1995–1997

Source: Reprinted with permission from Stewart et al. 1998.

14

oth Bt cotton and boll weevil eradicationhave great value for the cotton industry in

the United States. Available Bt cotton varieties arehighly effective against tobacco budworms andprovide significant suppression of bollworms andcertain other caterpillar species. Consequently,the foliar insecticide treatments required tocontrol these pests in Bt varieties are substantiallylower than in non-Bt varieties (table 5). Many ofthe treatments used to control tobacco budwormsand bollworms also are active against other pests,such as boll weevils and tarnished plant bugs,and lower insecticide use in Bt cotton reducescoincidental control of such pests. As a result, Btcotton varieties grown in boll-weevil-infestedareas typically require more foliar insecticidetreatments (table 5). Eradication of the boll weevil

is, therefore, necessary to realize the maximumpotential benefit from growing Bt cotton.

Eradication of the boll weevil reduces the numberof foliar insecticide treatments necessary tocontrol other pests in non-Bt cotton. For example,in Georgia, before the boll weevil was eradicated,the amount of insecticide required for control ofother pests was notably higher than the amountrequired following eradication. Data from Geor-gia show an increase in the need for treatment forother pests during the early years of boll weevileradication, because the insecticides used de-stroyed beneficial insects. In the absence ofbeneficial insects, populations of pests such astobacco budworms, bollworms, beet armyworms,cotton aphids, and whiteflies often increase.

Bt Cotton and Boll Weevil Eradication:

Boll weevil

Can They Work Together?

B

15

Bt cotton has proven itself to be a useful tool inminimizing the risk from certain caterpillar pestoutbreaks in the early years of a boll weevileradication program. In recent eradication pro-grams in Alabama, Mississippi, and Louisiana,producers chose to plant more than 80 percent oftheir acreage to Bt varieties, primarily to mini-mize risks of tobacco budworm outbreaks. How-ever, there has been a negative aspect to this highuse of Bt cotton. When most acreage is planted toBt varieties, growers apply fewer sprays thatprovide coincidental control of boll weevil. Thismakes boll weevil eradication somewhat moredifficult and more costly. Greatly overshadowingthis negative influence are the positive effects ofBt cotton in reducing the risks of secondary pestproblems during the initial years of an eradica-tion effort.

Where boll weevils are eradicated, the overallvalue of Bt cotton increases. In areas free of bollweevils, insecticide sprays can be cut back andbeneficial organisms more successfully relied onto reduce pest insects. The example from Georgiashows a distinct decline in the need to treat forother pests once the boll weevil was eradicated(fig. 2). An additional reduction in foliar treat-ments was observed after Bt cotton becameavailable in 1996. This example shows that bollweevil eradication and Bt cotton are complemen-tary in reducing total insecticide use and lower-ing insect control costs. Reducing insecticide useand relying more heavily on biological controlalso benefits efforts to manage insecticide resis-tance.

Table 5. Average number of annual insecticide sprays for tobaccobudworms, bollworms, and boll weevils in Mississippi, 1996–98

Insect 1996 1997 1998

Tobacco budworms andbollworms

on Bt cotton 0.3 0.9 1.2on non-Bt cotton 3.1 3.1 5.2

Boll weevilson Bt cotton — 2.6 3.3on non-Bt cotton — 1.9 1.9

Source: Modified with permission from Layton et al. 1999.

15

1616

Figure 2. Non-boll weevil treatments in Georgia. Reprinted with permission from Layton et al. 1999.

0

2

4

6

8

10

12

14

86 87 88 89 90 91 92 93 94 95 96 97 98

Boll weevil eradication began began

Av

era

ge

no

. tr

ea

tme

nts

16

Bt cotton introduced

17

t is commonly known that more than 500species of insects and mites have developed at

least some degree of resistance to insecticides(Georghiou and Saito 1983), knowledge clearlyshowing that many arthropods have the geneticpotential for rapid adaptation to chemicals intheir environment. Most scientists agree that thetobacco budworm and the bollworm will eventu-ally become resistant to the Cry1Ac protein usedin current Bt cotton varieties. The tobacco bud-worm has a well-known reputation for develop-ing resistance to chemical insecticides. Currentlyit is resistant to most conventional insecticidesused on cotton. However, for the time being, it isextremely susceptible to the Cry1Ac protein in Btcotton. The bollworm is inherently more tolerantto this toxin, and it is likely to develop resistancefaster than the tobacco budworm.

Field and laboratory studies document the devel-oped resistance of several insects to spray formu-lations of B.t. toxins. The best-known example isthe diamondback moth, a caterpillar pest thatattacks cabbage and related plants. It has shown

high levels of resistance to Bt sprays in Florida,Hawaii, North Carolina, Asia, and other locations(Tabashnik et al. 1990). It has also shown resis-tance to Bt transgenic canola plants. Researchershave already developed laboratory colonies ofColorado potato beetles, European corn borers,tobacco budworms, and bollworms that areresistant to Cry-proteins (fig. 3). The resistantlaboratory colonies of tobacco budworms andbollworms demonstrate these insects have thegenetic potential to become resistant.

Crop protection with Bt cotton is a form of hostplant resistance, like resistance of soybean variet-ies to the soybean cyst nematode. Farmers arefamiliar with resistant crops losing their protec-tion from pests, like nematodes overcomingsoybean resistance and mildew adapting toresistant wheat varieties. While the same fate ispredicted for Bt cotton, the time necessary toreach economic resistance can be greatly influ-enced by the way growers and consultants utilizethis crop.

Resistance to Bt Cotton in TobaccoBudworms and Bollworms

I

18

Figure 3. Development of resistance to Bt cotton in tobacco budworm in laboratory experiments.Reprinted with permission from Gould et al. 1992.

Cry1Ac Concentration (�g/ml)

Mo

rtal

ity

(per

cen

t)

Control strainControl female X selected maleSelected female X control maleSelected strain

100

80

60

40

20

00.01 0.1 1 10 100

19

variety of factors may influence the rate atwhich tobacco budworms and bollworms

become resistant to Cry-toxin in Bt cotton.

These factors include:

• The number of generations of tobacco bud-

worms and bollworms exposed each year to

Bt plants containing the same or similar toxins

• The percentage of each generation exposed

to Bt plants containing the same or similar Cry-

toxins

• The mortality level that Cry-toxin causes

among tobacco budworms or bollworms

carrying one copy of a resistance allele and

one copy of a susceptible allele. The mortality

level is determined by the Cry-toxin concentra-

tion in the plant, which in turn may determine

the functional dominance of the allele affecting

resistance

• The frequency with which Cry-resistance

alleles are expressed in the tobacco budworm

or bollworm population before exposure to

Cry-toxins and the dominant or recessive

nature of the resistance alleles

• The migration patterns of tobacco budworm

and bollworm moths

• The survival advantage or disadvantage that

resistance allele(s) offer tobacco budworms or

bollworms both in the presence and absence

of Cry-toxins

• The number of susceptible moths available for

mating with moths carrying resistance gene(s).

Before exposure to Cry-toxins—by sprayinginsecticides containing Bt or through planting Btcrops—very few tobacco budworms and boll-worms (perhaps 1 in 100,000 to 1 in 1 million)carry two copies of a resistance allele (RR),meaning they are fully resistant to Bt cotton.Some tobacco budworms or bollworms have asingle copy of a resistance allele and a susceptibleallele (RS); these are called heterozygotes. Theoverwhelming majority have two copies of asusceptible allele (SS).

Most of the susceptible insects (SS) are killed afterfeeding on Bt cotton, depending on the dose ofCry-toxin in the plant. The heterozygous insects(RS) usually are more difficult to kill than thesusceptible insects. Still, heterozygous insects arenot considered Bt resistant in most instances,because most will die if the toxin dose in theplant is high enough. Resistant insects (RR) arenot killed by Bt toxin. The difference in survivalrates among these three types represents a selec-tive advantage for resistant (RR) tobacco bud-worms and bollworms feeding on Bt cotton,because they will survive while susceptibleindividuals will die.

As the use of Bt cotton increases, a higher per-centage of tobacco budworm and bollwormpopulations will be exposed to Cry-toxins. Rela-tively more caterpillars carrying resistance alleleswill survive to adulthood, while fewer suscep-

How Resistance Develops

A

20

tible caterpillars will survive. As a result, moreresistant insects will pass on alleles for resistanceto new generations.

Because Cry-toxins are expressed in transgenicplants for the entire growing season—comparedto insecticides that remain active for short peri-ods—Bt cotton further prompts tobacco bud-worms and bollworms to select for resistance.This exposure can greatly enhance selection forresistant alleles and subsequently accelerate thepace that these insects develop resistance, espe-cially in areas where few alternate hosts areavailable (fig. 4).

Due to the high, season-long selection for Bt-resistant insects, scientists advise that develop-ment of field-level resistance could take a rela-tively short time. However, if Bt crops are notused over a high proportion of the total acreage,the rate that insects develop resistance should beslower and Bt crops should remain effective formany years (Gould et al. 1992). The rate thatresistance develops increases proportionately asthe acreage of Bt crops expands within a county,state, or region. The presence or absence ofalternate non-Bt food plants in a particular areamay also influence the development of resistance,which will probably occur first in a locale wherethe use of Bt cotton is high and the availability ofnon-Bt hosts, including cotton and other plants, islow. Figure 4. Increase of resistant larvae on

Bt cotton. Reprinted with permission ofD. Sumerford.

21

everal types of Bt corn use different Bttransformation events and express Cry1Ab,

Cry1Ac, Cry1F (not commercially available), orCry9c toxins. Some brands of Cry1Ab cornexpress Bt toxin in the ears, where bollwormlarvae (called corn earworms on corn) may befeeding, but other brands do not have toxin in theears. Cry9c toxin does not kill bollworms andshould not affect development of Bt resistance.

Currently, only the Yieldgard corn hybrids ex-press Bt protein in ears and may have an effect onBt resistance in bollworms (fig. 5). Tobacco bud-worms should not be affected by corn since theyvery seldom infest this crop.

Growing corn in cotton production areas couldhave a major influence on bollworms’ develop-ment of resistance to Bt cotton. Bollworm cater-

Bt Corn: Does It Hasten Resistanceto Bt Cotton?

S

Figure 5. Computer simulation showing the influence of Bt corn and Bt cotton on the ratethat bollworms develop resistance to Bt. Reprinted with permission from ILSI Health andEnvironmental Sciences Institute 1998.

100%0

10

20

30

40

50

60

% of all cotton planted to Bt cotton

Year

s to

res

ista

nce

* Spatially explicit model with 120 patches of corn, 160 patches of cotton, 40 patches of wild hosts, and 80 soybean patches. Five replicates and noncross resistance.

20% of all corn planted in Bt cotton

40% of all corn planted to Bt corn

60% of all corn planted to Bt corn

20% 40% 60%

22

rowers know that production costs canincrease as insects develop resistance toinsecticides. How much of an increase

depends on the availability and cost-effectivenessof alternative strategies. If such strategies do notexist or are expensive, production costs and croplosses may be very high. The resistance of tobaccobudworms and bollworms to Bt cotton is accom-panied by costs similar to those encountered withresistance to conventional insecticides.

There are also additional costs to resistance. Cry-toxins in Bt plants and Cry-toxin-based insecti-cides are not easily replaced when insects developresistance. In the past, growers relied on theavailability of new insecticides. There is no guar-antee this process will continue. Developing newinsecticides and transgenic insecticidal crops istime intensive, difficult, and expensive. Research-ers may not be able to develop new insecticides atreasonable costs that conform to environmentaland performance requirements as quickly as

Implications of Resistance

pillars infest the whorl and ear stages of corn.Moths are strongly attracted to silking corn to layeggs, and a major portion of the second-genera-tion population may develop in corn ears. Non-Btcorn will not select for Bt-resistant bollworms butwill act as a refuge and delay resistance. Themore non-Bt corn—other non-Bt hosts—grown ina corn and cotton production area, the slower Btresistance develops.

However, growing ear-expressing Bt corn canhasten Bt resistance in two ways: (1) if non-Bt

corn is replaced with ear-expressing Bt corn, therefuge effect that non-Bt corn provides dimin-ishes and bollworm populations are exposed toBt toxin—both from the increase in Bt crops andthe decrease in non-Bt crops; and (2) as men-tioned, greater exposure to the Bt toxin givesbollworms that are carrying a resistance allele(s) asurvival advantage and hastens the pace toresistance. If robust non-Bt refuges are main-tained for both cotton and corn, the rate thatresistance develops is reduced and should allowa prolonged life for both Bt products.

growers need them. A case in point is the highcost of recently released tobacco budworminsecticides for use on cotton.

The length of time that an insecticide or Bt cottonremains effective may depend upon how wellgrowers and pest managers follow resistancemanagement guidelines. Improper usage dra-matically decreases the effective life of a product.If Bt products are carefully used, their effective-ness may be extended for many years. But if thetechnology is abused, budworms and bollwormswill quickly become resistant. Preserving theeffectiveness of Bt cotton is one way to keep pestmanagement costs at the lowest level.

Other less obvious risks could also occur. In thepast the appearance of resistant tobacco bud-worm or bollworm infestations increased thefrequency of scouting and complicated decision-making by growers and pest managers. If new,selective insecticides are needed to combat Cry-

G

23

toxin-resistant pests, secondary insects maybecome more numerous. Also, caterpillars ex-posed to one Cry-protein (for example, Cry1Ac)may develop resistance to other similar Cry-proteins, even though they have not been ex-posed to them; this process is known as cross-resistance. When growing Bt crops, growers

should always use the refuge sizes specified inthe licensing agreement in order to slow thedevelopment of resistance to the Cry-proteinscontained in sprayable Bt insecticides and Btcrops.

logical, science-based, and proactiveresistance management strategy is neces-

sary to prevent tobacco budworms or bollwormsfrom developing resistance to Bt cotton in lessthan 10 years. All members of the cotton industryshould practice this strategy to slow developmentof resistance. EPA has registered products fromcompanies that sell Bt cotton seed, and thesecompanies are required to recommend andsupport insect resistance management (IRM)strategies for Bt cotton. IRM is a key element of agood overall integrated pest management (IPM)program.

EPA has accepted a resistance managementconcept for Bt cotton known as the “high dose/refuge strategy.” This approach has two comple-mentary principles: (1) Bt plants must produce ahigh dose of Cry-toxin throughout the season,and (2) effective IRM refuges must be maintained.An IRM refuge consists of a non-Bt host crop, andit is intended to produce susceptible tobaccobudworms or bollworms or both.

In theory, if Bt plants express a high dose of toxin,then all susceptible (SS) pests eating the plant willdie, almost all of the heterozygous (RS) insects

will die, and resistant (RR) insects will survive.When resistant survivors from the Bt crop matewith susceptible insects from a non-Bt refuge, theoffspring receive one allele—either an S or an R—from each parent. Offspring from the cross-mating will be heterozygotes (RS). If Bt plantsexpress a high dose of toxin, almost all heterozy-gotes will be killed. Eliminating most heterozy-gotes also eliminates most resistant alleles fromthe surviving populations and greatly slows thedevelopment of resistance.

Without a source for producing susceptibleinsects—an IRM refuge—the development ofresistance is proportional to the dose; that is, thehigher the dose, the more rapidly resistancedevelops. Therefore, the high dose/refuge strat-egy is a high-risk strategy, depending upon theavailability of properly functioning IRM refuges.The high dose, or in other words high effective-ness, is very good for pest control, but it cancause the rapid development of resistance in theabsence of effective IRM refuges.

As mentioned, the IRM refuge is acreage plantedwith a non-Bt crop that serves as a host for to-

Can Growers Slow Resistance?

A

23

24

Future IRM Refuge Options

bacco budworms or bollworms or both. And therefuge must be close enough to the Bt crop toensure that susceptible moths have an opportu-nity to mate with resistant ones. This means thatmoths from the Bt cotton and the refuge mustemerge at about the same time and be relativelyclose to each other.

Commercial Bt cotton varieties (Bollgard) cur-rently express enough Cry1Ac toxin throughoutmost of the season to kill all susceptible (SS) andalmost all heterozygous (RS) tobacco budworms.Only resistant (RR) tobacco budworms are ex-pected to easily survive. Thus, commercially

efuge regulations originally mandated in1995 for Bollgard varieties will remain in

effect through the 2000 growing season. When theregistration for Bollgard cotton expires after the2000 growing season, new refuge requirementsmay be forthcoming. At the time of publication,no final refuge requirements had been deter-mined for the 2001 growing season and beyond.The issue will be debated before the final decisionis made; recommendations will range fromcomplete removal of Bt technology from themarketplace, to a minimum of a 50-percent non-Bt cotton refuge, to no change from the current 4-percent-unsprayed or 20-percent-sprayed refugescenarios.

Computer simulation models, along with limitedevidence from field and laboratory studies,suggest that the 4-percent-unsprayed refuge or

available Bt cotton varieties (Bollgard) shouldqualify for the “high dose” definition for tobaccobudworms.

As mentioned, research shows that bollworms areless sensitive to Cry1Ac toxin than tobaccobudworms. Researchers estimate that from 5 to 25percent of susceptible (SS) bollworm larvaesurvive on Bt cotton varieties now in use(Bollgard), and estimates for heterozygote (RS)survival are significantly higher. So the Bt cottongrown now cannot be considered “high dose”against bollworms.

the 20-percent-sprayed refuge required on theoriginal Bt cotton label (Bollgard cotton) may notadequately delay resistance in bollworms andtobacco budworms. Studies using computermodels (Gould et al. 1992, ILSI 1998) also suggestthat bollworm resistance to Bt cotton can occurquickly if Bt cotton is extensively planted andonly small IRM refuges are used. Research indi-cates that bollworms, and to a lesser extentbudworms, have the genetic ability to adaptquickly to Bt toxin (Gould et al. 1992, Sumerfordet al. 2000, Burd et al. 2000).

In 1999, EPA and the U.S. Department of Agricul-ture, proposed for discussion the following twostructured refuge options to mitigate the resis-tance of tobacco budworms and bollworms to Bttoxins expressed in Bollgard Bt cotton:

R

25

1 An external refuge of at least 30% non-Bt cotton should be implemented. The

placement of the structured refuge

should be planted within 0.5 miles of

the farthest Bt cotton in a field to pro-

vide Bt-susceptible moths. The external

refuges of non-Bt cotton can be treated

with any other registered non-Bt insecti-

cide or other insect control measures.

2 In-field refuges of at least 10% non-Btcotton refuge should be implemented.

In-field refuges should be planted

entirely within the field as blocks,

minimum size to be determined based

on planter size, to provide Bt-suscep-

tible moths. Cotton fields may be

treated with any registered non-Btinsecticide or other control measures,

as long as the entire field is treated in

the same manner. This means that Btcotton rows cannot be treated indepen-

dently from non-Bt cotton rows with

insecticides or other insect control

measures.

In both options, (1) and (2), agronomic

practices used for farming the non-Btcotton must ensure adequate produc-

tion of susceptible [tobacco budworm

and cotton bollworm] adults to mate

with resistant adults emerging from Btcotton. In particular, termination of

growth of non-Bt cotton should not be

done until termination of growth of Btcotton has begun. In general, agro-

nomic practices for non-Bt cotton

should be similar, as practical, to those

of the Bt cotton grown in the same

management unit, especially regarding

crop nutrition, irrigation, and termination

(U.S. Environmental Protection Agency

and U.S. Department of Agriculture

1999).

EPA and USDA co-authored this position paper toprovide stakeholders with a focal point for dis-cussing future recommendations. The entiredocument may be reviewed at http://www.epa.gov/oppbppdl/biopesticides/otherdocs/bt_position_paper_618.html. Updatedversions of refuge guidelines will be available atthis website.

Entomologists in the public sector are concernedwith cotton insects’ almost 50-year history ofdeveloping resistance to sprayable insecticidesfrom almost every chemical class. Sound biologi-cal reasons indicate that Bt insecticide withinplants will be even more vulnerable to the devel-opment of insect resistance. Given the reducedpace of developing new replacement insect-control technology, the U.S. cotton industry mayface a greater risk of insecticide resistance thanever before.

The refuge plans currently in use are designed forBollgard cotton varieties that perform like thevarieties first marketed (for example, NuCotn 33and NuCotn 35). These refuges may be inappro-priate for new Bt genes or stacked gene Bt prod-ucts that may be marketed in the near future. Onehopes new Bt gene products will deliver a highdose of toxin to the bollworm and secondarylepidopterans and qualify for the high-dose refugestrategy. If these products do qualify, scientifictheory should support less restrictive IRM plansfor such future products. The development ofhighly effective transgenic insecticidal cottonplants with multiple toxic genes should beencouraged.

25

26

The Economics of Cotton Refuges

Can the Development of ResistanceBe Successfully Monitored?

he expenses and yield reduction associatedwith refuges must be viewed as costs of

using Bt cotton technology. Balancing these arethe economic and other benefits realized from Btcotton. In non-Bt cotton when insecticides areused for insect management, the chemical costs,application costs, and risks of handling pesticidesand possible complaints about them, amongother costs, are balanced against the benefit ofinsect control and higher yields. If, for example, a90/10 in-field refuge plan is used, the 10-percentrefuge may be damaged by insects, but the yieldand insect control cost over the total acreage may

be more profitable than that achieved using otherinsect management plans.

Preserving Bt cotton technology from resistancehelps ensure that growers will have effectiveinsect management options in the future. Experi-ence shows that greater crop loss and higher costinsect management typically follow the develop-ment of resistance. So, clearly, an economicbenefit is often difficult to estimate. There is noguarantee that new and cost-effective biotechnol-ogy or insecticidal products will replace existingtechnology in a timely fashion.

onitoring for the development of insectresistance to Cry-proteins is a difficult

and imprecise task. It may include surveying theannual use of Bt cotton in each county or parish,annual testing of tobacco budworm and boll-worm populations for Bt sensitivity, and checkingBt crops for any changes in the survival rate oftobacco budworms and bollworms. Monitoringtobacco budworm and bollworm populations onBt cotton is important in providing the earliestwarning that they are developing resistance; it isalso required by EPA. And it improves resistancemanagement efforts.

Companies that register the proteins in Bt cottonwith EPA are required to keep annual salesrecords on a county-by-county basis and submitsummaries for each state. Surveys of grower useof Bt cotton, conducted by Cooperative ExtensionService personnel, also may be available andcould be used as a guide for monitoring areaswhere circumstances favor the development of Btresistance by tobacco budworms and bollworms.Researchers predict that resistance is more likelyin areas where the use of Bt cotton has been highfor several years. Other characteristics of a par-

T

M

26

27

ticular region—such as the amount of non-Btcotton acreage and alternative hosts of tobaccobudworms or bollworms—may be surveyed tohelp determine the risk of resistance. Assigningresistance risk categories to unique cotton-grow-ing environments across the Cotton Belt can behelpful in monitoring the efforts at specific sites.

A major difficulty in monitoring is that resistancecan develop to an advanced stage before it iseasily detected in the field. For example, if 1 ofevery 100,000 tobacco budworms was resistantwhen Bt cotton was first marketed, resistancelevels would advance many-fold—perhaps to 1per 100 individuals—before field failures couldbe detected.

Detection of resistance in the field by scouts andgrowers would likely occur only in outbreakyears, unless the resistance level was more than 1insect per 100. The development of resistance islargely undetectable by measuring field perfor-mance of Bt cotton. This invisible phase is due, inpart, to monitoring techniques that are not sensi-tive enough to detect early shifts in tobacco bud-worm and bollworm resistance.

A first step in establishing an effective monitoringprogram is to document the initial susceptibilitylevel of tobacco budworms and bollworms to theBt toxin. With this information as a baseline, itmay be possible to spot small, early changes insusceptibility before field control failure occurs.Laboratory studies are uncovering more informa-tion about the frequency with which resistantgenes occur and the dominance of these genes.Detecting changes in susceptibility to the Cry-

toxins requires precise techniques that provideinformation from a large number of insects andmany locations. The U.S. Department of Agricul-ture and state universities are cooperating toestablish Cry-toxin susceptibility baselines andmeasure resistance in tobacco budworm andbollworm larvae collected from all across theCotton Belt, especially from the mid-South andthe Southeast (Summerford et al. 2000). Technol-ogy companies are also conducting similar tests.If these monitoring studies are sufficiently wide-spread, changes in the susceptibility of tobaccobudworms and bollworms to Bt may be success-fully detected prior to field control failures. EPAis reevaluating annual monitoring requirementsfor development of remedial action plans.

Counting the number of surviving caterpillars infields of Bt cotton may be helpful for detectingthe development of resistance. Susceptible popu-lations of tobacco budworms are decimated in Btcotton, so scouting for their survival may uncoverresistance before economic field failures occur.This method is not highly sensitive, however, andinsect resistance to Cry1Ac toxin must reach arelatively high level—for example, one resistantinsect in 500 susceptible insects—before detectionis possible.

The tolerance of bollworms to Cry1Ac toxin isalready too high to allow close measurement ofchanges in resistance using scouting and larvalsurvival. What’s more, less than adequate Btexpression in the cotton plants may produce afalse positive for resistance. Field scouting willdetect only large shifts in the survival rate ofbollworms.

28

Fitting Bt Cotton Into an InsectManagement Program

s with all conventional insecticides, Btcotton must be managed wisely and used in

combination with other insect pest and cropmanagement practices. For more than 20 years,IPM has focused on using insecticides on an as-needed basis only. The decision to use insecti-cides has generally been based on crop scoutingto determine insect pest densities and the use ofspray thresholds. Of course, Bt crops cannot beused as-needed because the Bt toxin is in theplant. For this reason, tobacco budworms andbollworms have greater opportunities for expo-sure to the toxin, and development of resistanceis more likely with Bt plants.

Consequently, resistance management must bepart of the total pest management package. Onthe positive side, Bt plants, in the absence ofinsecticide sprays, will help preserve beneficialinsects, allowing growers to take advantage ofthem as a free resource, especially in areas wherethe boll weevil has been eradicated.

When developing crop production plans, growersshould consider all insect management strategiesand tactics, with emphasis on the following areas:

Cultural practices. Early crop maturity (whichmay be optimized with computerized manage-ment programs such as COTMAN) should bepart of every pest management plan. Most grow-ers realize that early fruit set and early fruitmaturity help reduce insect pest infestations.

Combined use of optimal planting times andearly-maturing cotton varieties reduce cropdamage, minimize insecticide use, and diminishthe chances of pests’ becoming resistant to Bttoxin and insecticides. Early fruit set should beprotected so maturity is not delayed.

Rotating Bt crops with other non-Bt crops (tem-poral refuge) that also are hosts of tobacco bud-worms and bollworms is an example of resistancemanagement and offers overall pest managementbenefits. Soybeans, peanuts, and tobacco canharbor tobacco budworms, and these same crops,plus corn and grain sorghum, are all hosts ofbollworms. Yieldgard Bt corn would not beconsidered an adequate alternate host, but otheravailable Bt corn brands and non-Bt corn couldserve as alternate hosts. Certain forage crops,such as crimson clover and alfalfa, and manyweeds may also serve as hosts to tobacco bud-worms and bollworms. These alternate hosts canprovide an important refuge benefit and helpslow the development of resistance. In fact,according to computer models, the rate thatbollworms develop Bt resistance is directlyrelated to the proportion of non-Bt corn (notYieldgard) in a cotton and corn cropping system.

Beneficial insects. Conventional insecticidesoften have a significant adverse effect on benefi-cial insects in cotton. Populations of cottonaphids, beet armyworms, tobacco budworms,bollworms, fall armyworms, and other pests

A

29

frequently increase following insecticide use dueto the destruction of beneficial arthropods. Bttoxins do not have this effect. The loss of benefi-cial insects can also be reduced by using fewerapplications of insecticide or by using an insecti-cide that is less harmful.

Scouting and thresholds. Bt cotton is noteffective on pests such as boll weevils, plant bugs,and stink bugs. When bollworms, beet army-worms, fall armyworms, and looper moths laylarge numbers of eggs, damaging infestationsmay occur because the level of Bt toxin producedis not adequate to fully control them. As a result,Bt cotton must be scouted in a fashion similar tothat used with conventional cotton but usingaltered techniques and modified treatmentthresholds. Entomologists in each cotton-produc-

ing state publish variations for Bt cotton scoutingand treatment thresholds.

Bt cotton kills almost all newly hatched tobaccobudworm larvae and most bollworm larvae, sousing eggs as a criterion for insecticide applica-tions may lead to unneeded foliar treatments.When large numbers of bollworm eggs are laid,survival of the larvae may lead to damaginginfestations. As a consequence, in a few stateswhere the tobacco budworm and bollwormpopulations consist mostly of bollworms, exten-sion entomologists have adopted a high eggthreshold for Bt cotton (table 6). In other states(Arkansas, Mississippi, and Texas, for example)egg thresholds are not recommended as a treat-ment criterion. Use of the egg threshold is notpossible in states where the percentage of tobacco

Fi 14 C t i l ti h i th i fl f Bt d Bt tt th t th t b ll d l

The beneficial wasp, Microplitis croceipes, feeding on bollworm/budworm larva.Courtesy of J. Powell.

30

budworms is high, since separating tobaccobudworms from bollworms without a microscopeis difficult and time-consuming in every stage butthe adult.

A potential solution to the problem of identifyingeggs and small larvae is the development of kitscontaining species-specific monoclonal antibod-ies. These kits would allow growers and consult-ants to make better pest control decisions ifsurviving larvae longer than 1/8 to 1/4 inch arefound in Bt cotton. This technology is still beingdeveloped.

Scouting Bt cotton for tobacco budworms andbollworms must employ somewhat differenttechniques than scouting non-Bt cotton. Newlyhatched larvae do not die immediately after

feeding on Bt cotton. In fact, most states recom-mend counting caterpillars only of a certain size,such as 1/8 to 1/4 inch long. Caterpillars thatreach a minimum size have a greater ability tosurvive on Bt cotton. As the caterpillars becomelarger, they become more difficult to kill with theBt toxin and can damage the crop unless elimi-nated quickly. The larvae that survive are mostlikely to be bollworms. Caterpillars, usuallybollworms, often are found in flowers or “bloomtags,” so greater emphasis is being placed onexamining the blooms and small bolls whenscouting.

Tobacco budworm and bollworm damage in Btcotton can also be a target for scouting. Caterpil-lars that grow beyond a minimum size and feedon terminals, squares, or bolls cause recognizable

Table 6. Bollworm and tobacco budworm thresholds on cotton in South Carolina

Spray threshold

Stage % Eggs % Square damage % Larvae

Before bloomnon-Bt cotton — 20 15Bt cotton — — —

After bloomnon-Bt cotton 20 3 5Bt cotton 75 5 (bolls) 30

Source: Roof 1999

31

Cotton boll damage from budworm/bollworm larvae

damage. The superficial damage of very smallcaterpillars should not be mistaken for penetrat-ing injury that may indicate a damaging insectpopulation. In some situations, fall armywormsor beet armyworms can cause this damage, soscouts should be careful to properly identify thepest species. Some states have developed damagethresholds for Bt cotton.

Insecticide use in Bt cotton. In Bt cotton, allinsect pests not affected by the Bt toxin should bemanaged as they would be in non-Bt cotton. Itmay be necessary to apply insecticide to controlthrips, aphids, boll weevils, tarnished plant bugs,stink bugs, and a few tolerant caterpillars such asbeet armyworms and fall armyworms. Thedecision to apply insecticides to Bt and non-Bt

cotton should be based on scouting and treatmentthresholds.

Insecticide use depends on several variables,including the abundance of the pest insect andthe presence of beneficial insects. Using insecti-cides against early-season pests, such as bollweevils and tarnished plant bugs, also reducesthe number of beneficial insects. As mentionedpreviously, this reduction may result in higherpopulations of insect pests and fewer opportuni-ties for reduced insecticide use in Bt cotton. Inareas where boll weevils have been eradicatedand tarnished plant bugs are infrequent, thepresence of Bt toxin and beneficial insects greatlyreduces the amount of insecticide needed on Btcotton.

32

Key Steps—Implementing a ResistanceManagement Plan

ransgenic crops are one of the most revolu-tionary developments in agricultural pro-

duction. As with most new technology, there areexciting possibilities for the economic value of Btcotton and apprehensions about its wise use. Inorder to preserve Bt cotton well into the 21stcentury, producers, seed companies, scientists,and regulators need to foster strong collaborationto ensure longevity of the technology.

Companies are striving to develop new genes forinsertion into cotton plant DNA to provide other

he following summary, based on theprinciples outlined in this publication,

assumes that a voluntary, proactive approachby growers will provide product stewardshipfor long-term yield benefits and profitability.

1 Use Bt cotton in fields where the risk of

severe tobacco budworm and bollworm

infestations warrants the price premium for

seed.

2 Plant Bt cotton and non-Bt refuge in the

required proportions and patterns.

Bt Cotton in the Future

32

T possibilities for improving agronomic traits andpest control characteristics. Genes for new insecti-cidal toxins will be important for managing awider spectrum of insects and for slowing thepace of resistance. If future varieties express ahigh dose of toxin and the toxins do not have thesame physiological target site, the rate that insectsdevelop resistance could be greatly reduced.

3 Record where Bt and non-Bt cotton are

planted so Bt cotton performance can be

monitored and non-Bt cotton can be scouted

and treated as needed.

4 Use all cultural alternatives available for

avoiding high late-season tobacco budworm

and bollworm populations. These alternatives

include early planting, short-season varieties,

and protection of early fruit.

5 Continue using an IPM approach for all

pests. When designing scouting priorities,

keep in mind that other insects—such as

T

33

beet armyworms, cabbage loopers, cotton

square borers, European corn borers, fall

armyworms, southern armyworms, soybean

loopers, and yellowstriped armyworms—are

suppressed at various levels by Bt cotton.

Still other pests are not affected by Bt cotton,

including boll weevils, cotton aphids, cotton

fleahoppers, cutworms, spider mites, stink

bugs, tarnished plant bugs, thrips, and

whiteflies.

6 Monitor Bt cotton to verify tobacco budworm

and bollworm control throughout the season.

Since bollworms are not as readily controlled

by Bt cotton, carefully watch for a heavy egg-

lay and scout for surviving larvae lower in the

cotton plants, especially in bloom tags. If

feeding damage occurs in Bt cotton, investi-

gate the cause immediately. If needed, get

help in identifying feeding caterpillars. If

tobacco budworm or bollworm larvae or

excessive damage are discovered, resis-

tance or tolerance to Bt cotton is a possibility,

and the situation should be investigated.

Verify from field records that Bt cotton wasplanted where excessive damage or larvae areobserved. Consult your grower’s guide for theseed company’s procedure for investigatingsuspected cases of resistance or tolerance. Imme-diately notify seed company representatives orextension agents or both if evidence indicates aperformance problem.

33

34

References

Burd, A.D., J.R. Bradley, J.W. Van Duyn, and F.Gould. 2000. Resistance of bollworm, Helicoverpazea, to CryIAc toxin. In P. Dugger and D. Richter,eds., Proceedings of the Beltwide Cotton Produc-tion Research Conference, pp. 923–926. NationalCotton Council, Memphis, Tennessee.

Georghiou, G.P., and T. Saito, eds. 1983. Pestresistance to pesticides. Plenum Press, New York.

Gould, F., A. Martinez-Ramirez, and A. Ander-son. 1992. Broad-spectrum resistance to Bacillusthuringiensis toxins in Heliothis virescens. Proceed-ings of the National Academy of Sciences USA89:7986–7990.

ILSI Health and Environmental Sciences Institute.1998. Evaluation of insect resistance managementin Bt field corn; a science-based framework forrisk assessment and risk management. Report ofan expert panel. ILSI Press, Washington, DC.

Jenkins, J.N., J.C. McCarty, Jr., and R.E. Buehler.1997. Resistance of cotton with endotoxin genesfrom Bacillus thuringiensis var. kurstaki on se-lected lepidopteran insects. Agronomy Journal89:768–780.

Layton, M.B., S.D. Stewart, M.R. Williams, andJ.L. Long. 1999. Performance of Bt cotton inMississippi, 1998. In P. Dugger and D. Richter,eds., Proceedings of the Beltwide Cotton Produc-tion Research Conference, pp. 942–945. NationalCotton Council, Memphis, Tennessee.

Ostlie, K.R., W.D. Hutchison, and R.L. Hellmich,eds. 1997. Bt corn and European corn borer. NCR

Publication 602. University of Minnesota, St.Paul.

Roof, Mitchell E. 1999. Cotton insect manage-ment. Clemson University Extension InformationCard 97, Clemson, South Carolina.

Stewart, S., J. Reed, R. Luttrell, and F.A. Harris.1998. Cotton insect control strategy project:Comparing Bt and conventional cotton manage-ment and plant bug control strategies at fivelocations in Mississippi (1995–1997). In P. Duggerand D. Richter, eds., Proceedings of the BeltwideCotton Production Research Conference, pp.1199–1203. National Cotton Council, Memphis,Tennessee.

Sumerford, D.V., D.D. Hardee, L.C. Adams, andW.L. Solomon. 2001. Tolerance to CryIAc inpopulations of Helicoverpa zea and Heliothisvirescens (Lepidoptera: Noctuidae): Three-yearsummary. Journal of Economic Entomology. Inpress.

Tabashnik, B.E., N.L. Cushing, N. Finson, andM.W. Johnson. 1990. Field development of resis-tance to Bacillus thuringiensis in diamondbackmoth (Lepidoptera: Plutellidae). Journal ofEconomic Entomology 83:1671–1676.

U.S. Environmental Protection Agency and U.S.Department of Agriculture. EPA and USDAposition paper on insect resistance managementin Bt crops. 1999. Posted on the World Wide WebMay 27, 1999, minor revisions July 12, 1999.http://www.epa.gov/oppbppd1/biopesticides/otherdocs/bt_position_paper_618.html

35

Glossary

Allele. An alternative form of a gene. For ex-ample, a gene determining the reaction of aninsect to Bt toxin may occur in the form of anallele for resistance or an allele for susceptibility.

Alternate host. Non-Bt plant types other thancotton that support successful reproduction oftobacco budworms or bollworms or both; soy-bean and corn are examples.

Bacillus thuringiensis (Bt). A naturally occurringsoil bacterium that occurs worldwide and pro-duces a toxin specific to certain insects (forexample, moths, beetles, blackflies, or mosqui-toes).

Beneficial arthropods. Usually refers to insectpredators or parasites of pest insects in crops.When abundant, these insects can assist in de-creasing certain pests, although they are oftenseriously reduced by insecticide.

Biotechnology. The science and art of geneticallymodifying an organism’s DNA, such that thetransformed individual can express new traitsthat enhance the quality of a product (for ex-ample, seed oils or fiber quality) or can expressresistance to pests.

Boll weevil eradication program. A joint USDA,producer, and state program designed to elimi-nate the boll weevil as an economic pest from theCotton Belt.

Bt cotton. Commercial varieties of cotton thatcontain a gene from Bacillus thuringiensis withinits DNA. This gene allows the plant to produceCry-protein within most or all plant tissues. The

toxin makes the plant tissue—terminals, squares,or bolls—toxic to some important caterpillarpests.

Bt insecticides. Formulations of Bt Cry-proteinsmanufactured and sold for spraying a widevariety of plants in order to control certain cater-pillar or beetle pests. Some formulations are alsoused for mosquito and fly control.

Cross resistance. Resistance to one or more toxinsfrom exposure to one or more other toxins, suchas bollworms becoming resistant to Cry1Ac incotton by being exposed to CryIAb in corn, orvice versa.

Cry-proteins. Any of several crystalline proteinsfound in Bt spores that are activated by enzymesin the insect’s midgut. These proteins attack thecells lining the gut, cause gut paralysis, andsubsequently kill the insect.

Cry1Ac protein (toxin). One of many Bt crystal-line protein toxins. The Cry-protein used in thefirst varieties of Bt cotton sold commercially togrowers.

DNA. Deoxyribonucleic acid, a double-strandedmolecule, consisting of paired nucleotide unitsgrouped into genes and associated regulatorysequences. These genes serve as blueprints forprotein construction from amino acid buildingblocks.

Dominance (of an allele). The ability of one alleleto determine a characteristic in a heterozygousindividual. For example, an allele for resistance toBt toxin may be dominant. As a consequence, an

36

insect heterozygous for resistance, that is, con-taining an allele for resistance and an allele forsusceptibility, is resistant to Bt toxin.

Dose of toxin. The amount of toxin eaten perinsect. For example, the dose of toxin that acaterpillar receives by eating part of a Bt plantcan be stated as micrograms of Bt toxin eaten percaterpillar or per milligram of the caterpillar’stotal body weight.

Expression. Production of the desired trait (pro-tein concentration, for example) in a transgenicplant. Expression varies with the gene, its pro-moter, and its insertion point in the host DNA.

Gene. The basic unit of inheritance; a section ofDNA that codes for a specific product (for ex-ample, protein) or trait.

Genetic marker. See Marker.

Heterozygote. A diploid organism carrying twodifferent alleles (for example, susceptible andresistant) of a gene.

High dose. A dose of toxin high enough to kill allsusceptible target pests and nearly all heterozy-gotes. Such a dose can be delivered by plantswith sufficiently high concentrations of Bt toxin.

High-dose refuge strategy. A resistance manage-ment approach that uses plants to minimize therapid selection for resistance to transgenic plants.This strategy relies upon plants that produce Cry-proteins at a concentration sufficient to kill all butthe most resistant insects and is used in combina-

tion with a non-Bt refuge that allows susceptibleinsects to survive and mate with resistant indi-viduals.

Host plant resistance. Ability of a plant to avoidinsect damage, to kill attacking insects, or totolerate their damage.

Insect resistance management (IRM). A proac-tive approach to offset and slow insects’ resis-tance to Bt crops or insecticides by reducingselection or by counteracting the effects of selec-tion for resistance genes.

Integrated pest management (IPM). A manage-ment approach that integrates multiple, comple-mentary control tactics to manage pests in aprofitable, environmentally sound manner.Examples of the control tactics include conserva-tion of natural enemies, crop rotation, host plantresistance, and insecticides.

IPM. See Integrated pest management.

IRM. See Insect resistance management.

Larva. Immature stage of certain insect species;examples include caterpillars and grubs.

Lepidoptera. The order of insects that includesmoths and butterflies. The plant-eating immaturestages are often called worms, caterpillars, orlarvae.

Marker. A genetic flag or trait used to verifysuccessful gene transformation and to indirectlymeasure expression of the inserted genes.

37

Mode of action. Mechanism by which a toxinkills an insect. For example, the mode of action ofBt is ingestion and disruption of cells lining themidgut.

Promoter. A DNA sequence that regulates where,when, and to what degree an associated gene isexpressed.

Refuge or insect resistance management (IRM)refuge. A wild host area or an area planted withnontransgenic plants (for example, non-Bt cottonor alternative hosts for tobacco budworms orbollworms) where susceptible pests can surviveand produce a local population capable of matingwith resistant survivors from Bt cotton. Thismating, by decreasing the likelihood that Bt-resistant insects will mate with one another,dilutes resistance in the insect population.

Registration. Legal approval of pesticides andtransgenic crops for use in the United States bythe U.S. Environmental Protection Agency.Registration is granted after extensive review oftoxicology (to mammals, birds, fish, and othernontarget organisms), environmental fate, healthand safety issues, and precautions.

Resistance (by pests). The evolved capacity of anorganism to survive in response to selection fromexposure to a pesticide. The evolution of resis-tance occurs through a process of genetic accu-mulation, whereby a population becomes lesssensitive to the pesticide following repeatedexposure.

Selection. A natural or artificial process thatresults in survival and better reproductive suc-cess of some individuals over others. Selectionresults in genetic shifts if survivors are more, orless, likely to have particular inherited traits.

Stacked. Describes transgenic plants with morethan one introduced gene in a single crop plantvariety, such as a “stacked” cotton variety con-taining a Bt gene and a gene for herbicide toler-ance.

Transgenic. An organism genetically altered byaddition of foreign genetic material (DNA) fromanother organism into its own DNA.