Historic, Archive Document

Do not assume content reflects current
scientific knowledge, policies, or practices.

aszi

United States
Department of
Agriculture

Agricultural

Research

Service

ARS-106

July 1992

Cotton-Integrated
Pest Management
Proceedings of a
Symposium

September 3-9, 1990
Alma-Ata, Kazakhstan

fash J^z-hA U’z.bifc i ■

Richard S. Soper, Nikolai A. Filippov and Sultan N.
Alimukhamedov, Editors. 1992. Cotton Integrated Pest
Management. Proceedings of a Symposium sponsored by
the U.S. Department of Agriculture, Agricultural Research
Service, the All-Union Institute of Biological Methods in
Plant Protection, and the Middle Asia Institute for Plant
Protection. Alma-Ata, Kazakhstan; September 3-9, 1990.
U.S. Department of Agriculture, Agricultural Research
Service, ARS-xxx, yy pp.

The papers in these proceedings are reproduced essentially
as supplied by the authors and do not necessarily reflect the
views of the U.S. Department of Agriculture. This
symposium was held before the dissolution of the Union of
Soviet Socialist Republics; thus, current names of some
institutions and political entities in the former U.S.S.R.
may differ from those cited in papers in these proceedings.

Commercial companies and products are mentioned in this
publication solely to provide specific information. Mention
of a company does not constitute a guarantee or warranty of
its products by the U.S. Department of Agriculture or an
endorsement by the Department over products of other
companies not mentioned. This publication also reports
research involving pesticides. It does not contain recom¬
mendations for their use, nor does it imply that the uses
discussed here have been registered. All uses of pesticides
must be registered by appropriate State and Federal
agencies before they can be recommended.

While supplies last, single copies of this publication may be
obtained, at no cost, on request from the ARS National
Program Staff, Bldg. 005, BARC-West, Beltsville, Mary¬
land 20705.

Copies of this publication may also be purchased from the
National Technical Information Service, 5285 Port Royal
Road, Springfield, Virginia 22161.

United States
Department of
Agriculture

Agricultural

Research

Service

ARS-106

July 1992

Cotton-Integrated
Pest Management
Proceedings of a
Symposium

September 3-9, 1990
Alma-Ata, Kazakhstan

Edited by:

Richard S. Soper
Nikolai A. Filippov
Sultan N. Alimukhamedov

Contents

Place of Biocontrol in Integrated Pest
Management in the former Union of Soviet
Socialist Republics N. A. Filippov 1

Cotton Insect Problems in the United
States-An Overview D. E. Hardee 17

Management of Cotton Bollworm and
Tobacco Budworm Populations through
Area-Wide Application of Nuclear
Polyhedrosis Virus on Early-Season
Alternate Hosts M. R. Bell, D. D. Hardee,
J. L. Hayes and E. A. Stadelbacher 24

Production and Application of
Trichogramma in the former Union of
Soviet Socialist Republics
Sh. M. Greenberg 38

Propagation and Release of Natural
Enemies for Control of Cotton Insect and
Mite Pests in the United States, including
Status and Prospects for using Microplitis
croceipes E. G. King and J. E. Powell 47

Integration of Techniques and Methods of
Cotton Protection from Pests in Tajikistan
V. G. Koval enkov 61

Genetic Control of Cotton Insects, the
Pink Bollworm as a Working Program
R. T. Staten 66

The Problem of Cotton Pest Resistance to
Pesticides in the former Union of Soviet
Socialist Republics and Rational Systems

to Cope with this Phenomenon

G. I. Sukhoruchenko 97

Prospects and Status for Development of
Novel Chemicals for Integrated Pest
Management in Cotton J. J. Menn 105

Systems Analysis and Modeling for
Decision Making in Pest Management
T. L. Wagner 118

Boll Weevil Eradication in the Western
United States and Northwestern Mexico,

A Cooperative Program L. C. Antilla,

F. D. Myers and R. T. Staten 1 28

Use of Fungal Pathogens for Biological
Control of Insect Pests in Alfalfa

R. I. Carruthers and K. Hural 1 32

Use of Beauveria bassiana and Bacillus
thuringiensis for Suppression of the
European Corn Borer, Ostrinia nubilalis
L. C. Lewis 149

The Development of the National
Biological Control Institute

D. E. Meyer dirk 158

Cultural Control of Cotton Insect Pests,
with Emphasis on an Emerging Program of
Cultural Controls in the Lower Rio Grande
Valley of Texas K. R. Summy and
E. G. King 76

Foreword

Richard S. Soper1

Integrated Pest Management (IPM) has been an important
force in shaping crop protection strategies for the past two
decades. The evolution of this approach to controlling crop
pests has developed without very much attention being paid
to the conservation or use of the natural enemy complex.
The mass production of insect parasites and predators and
their periodic augmentative release has been important in
greenhouse culture, especially in the Netherlands, but has
not gained much acceptance in the United States. The
conservation of natural enemies through cultural practices
can be an important method for crop protection.
Development of microbial pesticides is also an important
method in biological control. These important ecologically
benign control techniques have largely been ignored in the
current implementation of IPM. The focus has been instead
on the conservative use of chemical pesticides. Through a
system of scouting to predict damaging pest population
levels, IPM technology has been able to dramatically
reduce the use of chemical pesticides. This has made a
major impact around the world in crop protection.

However, recent events in the United States and elsewhere
have necessitated a reexamination of the current
implementation of IPM. The public concerns for
environmental hazards caused by misuse of pesticides, the
implication of certain fungicides as carcinogens, the
occurrence of some herbicides in groundwater and the
hazards to endangered species caused by insecticides, all
have caused a dramatic rethinking of how IPM should be
conducted. A recent U.S. National Academy of Sciences
report urges the scientific community to develop control
schemes which use biological control as the first line of
defense and apply chemical pesticides as a last resort.

The emphasis on biological control in IPM tactics was
found to be of mutual interest between the U.S. Department
of Agriculture, Agricultural Research Service and the All-
Union Institute of Biological Methods in Plant Protection.
As part of a collaborative arrangement on biological control
between the U.S. and the former U.S.S.R., it was agreed to
co-sponsor a symposium on cotton IPM in which the major
thrust would be the maximization of natural enemies. This
was hosted in Alma-Ata, Kazakhstan, by the Middle Asia
Institute for Plant Protection September 3-9, 1990. This
symposium provided a forum for experts in cotton pest
management and biological control from both countries.
These proceedings are a compilation of papers presented at
that symposium.

1 National Program Leader, Biological Control, USDA, ARS, Beltsville,
Maryland

Place of Biocontrol in Integrated Pest
Management in the former Union of
Soviet Socialist Republics

Nikolai A. Filippov1

At present one of the most important ecological problems
for agrarian science and practice is the development and
implementation of such techniques of Integrated Pest
Management (IPM) of noxious organisms, which, along
with prevention of yield losses, would eliminate the
harmful influence of chemical pesticides on the environ¬
ment and provide production of ecologically pure agricul¬
tural produce. Under the conditions of intensification of
agricultural production in the former U.S.S.R., as well as
worldwide, the chemical strategy still predominates in plant
protection from pests, diseases and weeds.

Considering the level of current scientific projects and
species composition of noxious organisms, it is impossible
to completely eliminate chemical control of pests, vectors
of diseases and weeds on many crops, especially in cases
with fungal epiphytotics, or widespread pest outbreaks.
However, public opinion about chemical pesticides has
become increasingly negative. It is conditioned by certain
concerns about unsatisfactory assortment of pesticides, and
the violation of regulations and techniques of their use.

Areas of crops protected by chemical pesticides have been
increasing constantly: 36.4 million hectares in 1960, 80.8
million hectares in 1965, 107 million hectares in 1970,
140.3 million hectares in 1975, 167 million hectares in
1985, and 170 million hectares in 1989. To a great extent,
this increase has resulted from the implementation of the
intensive technologies in which pesticides are often used
without due regard for the actual phytosanitary conditions
(Novozhilov 1989, Kosenko 1989). The world tendencies
are known to be similar — in recent decades the use of
chemical pesticides throughout the world has increased 40
times.

In the former U.S.S.R., more than 300 pesticides are
available for application, and more than 100 plant protec¬
tion chemicals are produced in the country. Up to 250-300
thousand tons of chemical pesticides are applied annually in
the former U.S.S.R., which makes up 5.3 percent of the
world use. Over 128 million roubles worth of chemical
pesticides were imported in 1989 (Melnikov 1989).

Plant protection measures, including application of pesti¬
cides, annually prevent losses of 17-20 million tons of

1 All-Union Institute of Biological Methods in Plant Protection
(AUIBMPP), Kishinev, former U.S.S.R.

grain, 8 million tons of potatoes, 19 million tons of sugar
beets, 1.8 million tons of raw cotton, and 6-6.5 million
tons of fruits, vegetables and other agricultural products
worth more than 8 billion roubles with an expenditures of 2
billion roubles (Gulenko 1988). Although pesticide
application has a good effect economically, it has rather
serious negative effects as well. Extensive, and not always
scientifically based, use of plant protection chemicals often
leads to environmental pollution, violation of sanitary-
hygienic standards, and destruction of natural biocenoses.
Pesticide applications and their aftereffects cause consider¬
able harm to the beneficial fauna - entomophages, plant
pollinators, wild animals and birds. More frequently,
resistance development to different pesticide groups in
noxious organisms is reported.

Besides searching for new pesticides which are less toxic to
mammals and the beneficial fauna, for more extensive use
of new cultivation methods, and for disease- and pest-
tolerant varieties and hybrids of cultivated plants, we ought
to consider biological control employing microbials as an
important alternative to chemical control. Considerable
attention is being paid to biological control investigations in
the former U.S.S.R. More than 1,700 commercial
biofactories and biolaboratories in the country deal with the
production of biopreparations of about 20 species of
entomophages. More than 5,000 persons are working at
commercial biolaboratories (Nikonov 1986, 1988). Some
factories of the microbiological industry produce
biopreparations based on different strains of the
entomopathogenic bacterium. Bacillus thuringiensis
(dendrobacillin, lepidocid, bitoxibaccilin, gomelin, etc.).
More than 100 scientific and planning institutions in the
former U.S.S.R. elaborate different trends in biological
control.

In 1989, practical implementation of scientifically-based
recommendations resulted in the application of biological
methods in the former U.S.S.R. to 27 million hectares,
including 19.7 million hectares under entomophages,

7.3 million hectares under micropreparations, and green¬
house area of 18 thousand hectares (single estimate). The
share of biological control in the country is about 25
percent of the total volume of pest and disease control
(Kosenko 1989).

Insect pheromones are used to monitor the development of
pests and to improve pest management over an area of 10
million hectares, including that of cotton (Novozhilov

1

1989, Khodzhaev 1990). Biological control of a number of
agricultural pests is used in large areas (Table 1). Accord¬
ing to the estimate made by the All-Union Institute of
Biological Methods in Plant Protection (AUIBMPP),
application of biological preparations produces a favorable
ecological effect and is economically sound (Table 2).

Some Problems of Integrated Control of
Cotton Pests

Cotton is the most important crop in former U.S.S.R. The
former U.S.S.R. (alternately with the U.S.A.) is the second
or third (after China) largest cotton grower in the world
providing one sixth of the total world production. The
following are data on cotton production in the former
U.S.S.R. (million hectares): 1970 - 2.75; 1980 - 3.15;

1985 - 3.32; 1986 - 3.47; 1987 - 3.53; 1988 - 3.43. Gross
yield of raw cotton for 1981-1989 constituted 8. 3-8. 7
million tons.

Cotton yield capacity did not progress (Tables 6, 7, 8)
despite the increase of irrigated areas in cotton-growing
republics, use of fertilizers and pesticides (Tables 3, 4, 5),
and use of new varieties and technological elements. Big
economic and technological mistakes and violations have
been made in cotton growing in the former U.S.S.R. Areas
under cotton were groundlessly increased and crop rotations
were violated. Instead of recommended cotton/alfalfa crop
rotations where cotton had to cover no more than 50-55 %
of total acreage, it became a monoculture in many regions
occupying the same fields for dozens of years without
breaks. This practice resulted in soil depletion, mass
distribution of dangerous weeds, pests and diseases; and
decrease in cotton yield and quality. Cotton monopolism in
many regions has led to a complex ecological situation and
social tension.

The structure of cultivated areas in cotton-growing repub¬
lics (mainly oriented towards cotton growing) has compli¬
cated the phytosanitary situation and does not allow
maximum use of climatic conditions or possibilities of
irrigation and labor resources to be aimed at increasing the
production of fruits, grapes, vegetables, melons, cereals
and forage crops. Gross yields and yield capacities of these
crops remain low (Tables 9, 10, 11, 12).

In recent decades, cotton protection from noxious organ¬
isms became widespread in cotton growing regions, and
cotton has received as many as 12 applications of insecti¬
cides per growing season. However, this did not reduce
pests because of pesticide resistance and destruction of
beneficial insects.

Over 210 pests and 30 fungal and bacterial diseases are
recorded on cotton in the republics of Middle Asia. More
than 200 pest species and 10 disease pathogens are recorded
in association with cotton crops (Alimukhamedov 1978,
1983, 1987; Alimukhamedov et al. 1989; Maksumov and
Narzikulov 1981; Kovalenkov 1989, 1990). The following
pests are of major importance in cotton producing regions
of the former U.S.S.R.: Noctuidae, Aphidinae,
Thysanoptera, Tetranychus urticae, Bemisia tabaci, Hypera
postica ( = Phytonomus variabilis) Adelphocoris lineolatus,
Ostrinia nubilalis and others (Alimukhamedov 1987,
Alimukhamedov et al. 1989, Maksumov and Narzikulov
1981, Kovalenkov 1990, Khamraev 1989).

Scientists in the former Soviet Union have contributed
considerably to the development of integrated control of
cotton pests, including biocontrol. Economic damage
thresholds for different insect pest species and mites have
been studied. Many experimental data have been obtained
on species composition, ecology and levels of biological
effectiveness for entomophage natural populations, produc¬
tion and application technologies for reared entomophages,
and commercial microbiopreparations against cotton pests
in IPM (Adashkevich et al. 1988, Alimukhamedov et al.
1989, Alimukhamedov et al. 1989, Davletshina 1972,
Kovalenkov 1989, Khamraev 1990).

Application of Entomophages Outdoors

Trichogramma

The egg parasitoid, Trichogramma spp., is the main agent
for biological control of noxious lepidopterans (Lepidopt-
era). Species included T. evanescens, T. pintoi, T.
cacoeciae, T. semblidis, T. embryophagum. Trichogramma
is applied by seasonal colonization against noctuid pests
such as cabbage moth, Mamestra brassicae L., cotton
bollworm, Helicoverpa armigera Hubner, turnip moth,
Agrotis segetum Schiff, other species of vegetables, cereals,

2

sugar beet, cotton, and other crops, as well as against
European comborer Ostrinia nubilalis Hubner, and pea
moth, Laspeyresia nigricana Stephens. The parasitoid is
being tested against the codling moth Laspeyresia
pomonella L. and leaf rollers (Tortricidae) in orchards and
vineyards.

At present Trichogramma is applied in the former U.S.S.R.
on 17 million hectares. More than 750 mechanized
production lines are engaged in mass rearing of
Trichogramma providing the parasitoid to about 50 percent
of cotton growing areas. The AUIBMPP and other
research institutions in the former U.S.S.R. have developed
ways to increase the viability of Trichogramma reared on
Sitotroga cerealella. Research is also being done to select
more suitable laboratory hosts for Trichogramma and to
develop artificial substrates for Trichogramma rearing.

Considerable experience on the practical use of
Trichogramma has been accumulated in different regions of
the former U.S.S.R. Scientifically-based and timely
application of high-quality parasitoid material in regions
favorable for its activity yields 60-80 percent parasitization
of noctuid pest eggs and 40-60 percent of eggs of Ostrinia
nubilalis. Despite certain achievements with
Trichogramma, its effectiveness is, in some cases, unstable
and insufficient. This is mainly caused by neglect of the
recommendations for its rearing and application, poor
quality of the parasitoid material for releases, and unsuit¬
able weather during its seasonal colonization.

Tests carried out by the AUIBMPP have demonstrated that
under optimal conditions the release of the high quality
Trichogramma provides 80 percent parasitism of pests,
lower quality material 60-80 percent, third-class material
35-60 percent, and sub-standard Trichogramma less than 35
percent (Golyshin and Greenberg 1989). To increase the
effectiveness of Trichogramma, recommendations for
improving the application techniques have been prepared by
the AUIBMPP and other research-institutions. The use of
Trichogramma results in an additional yield of 2,000-3,000
kg/ha for sugar beet and cabbage and 180-230 kg/ha for
com and winter wheat grain. Application of Trichogramma
reduces chemical control (2-3 pesticide treatments) and,
hence, expenditure on chemical treatments (7-9 roubles per
ha).

At present, the release of Trichogramma in the field is done
mainly by hand which is very laborious. Work is being

done on mechanization of the process. Equipment for
ground release of Trichogramma by tractors as well as
equipment for aerial release have been designed and are
being tested.

Mechanized production lines for Trichogramma rearing at
biofactories have become obsolete and do not meet modem
technological and sanitary-hygienic requirements. For this
reason research and project-designing institutions are
developing mechanized production biofactories (2nd
generation) for Trichogramma rearing with higher produc¬
tivity and better sanitary /hygienic conditions which will
provide production of high-quality egg parasitoid material.

Bracon

In recent years, Bracon hebetor Say has been mass cultured
and widely applied to control Helicoverpa armigera in the
republics of Middle Asia. The parasitoid is reared on either
Ephestia kuehniella Zeller or Galleria mellonella L. in
biolaboratories. Bracon is released in the field at mass
emergence of H. armigera. The adults disperse rapidly and
parasitize 2nd to 4th instar larvae. The release of 500 to
1,500 individuals per hectare yields 40-50 percent or more
parasitism and mortality of H. armigera (Adashkevich and
Saidova 1988). Bracon is released on more than 2 million
hectares in the former U.S.S.R. Special equipment for its
mass rearing is being developed and tested.

Predaceous bugs

Research institutions of the former U.S.S.R. are developing
methods for mass production of the introduced predaceous
bugs, Podisus maculiventris Say and Perillus bioculatus
Fabr. for seasonal colonization against Leptinotarsa
decemlineata Say on early potatoes and eggplants. Field
tests, conducted by the AUIBMPP, demonstrated that the
predaceous bugs Perillus and Podisus, released for seasonal
colonization, effectively suppress populations of the
Colorado potato beetle. Two to three releases of 2nd-3rd
instar larvae of Perillus (40, 000-50, 000/ha) and Podisus
(80,000/ha), depending on the pest population, results in
60-80 percent or higher mortality of egg clusters and larvae
of the Colorado potato beetle. From economical and
technological view points, seasonal colonization of preda¬
ceous bugs is reasonable on early potatoes against 1st-
generation egg clusters and larvae of the Colorado potato

3

beetle and on eggplants against the 2nd-generation of the
Colorado potato beetle (Filippov 1989).

Other entomophages

The AUIBMPP and the All-Union Institute of Plant
Quarantine (AUIPQ) are carrying out investigations on the
effectiveness of the introduced egg parasitoid Edovum
puttleri, of the Colorado potato beetle. In laboratory tests,
parasitism of the Colorado potato beetle eggs in different
generations varied from 32-75%, depending on the instar
and the quality of the host. Edovum parasitizes fresh (1-
day) eggs of the Colorado potato beetle more effectively.

In field tests three releases of Edovum on the plots with
eggplants ("parasitoid-host" ratios from 1:3 to 1:7) resulted
in 72.5 percent parasitism and destruction of egg clusters.
Thus, use of the egg parasitoid against the Colorado potato
beetle is promising.

A number of entomophages are also used against pests of
citrus and fruit crops in southern parts of the former
U.S.S.R. The introduced entomophage, Aphelinus mali
Haldemann effectively suppresses the reproduction of the
wooly apple aphid, Eriosoma lanigerum Hausmann, in
100,000 hectares in southern horticultural regions. Re¬
leases of the introduced Rodalia cardinalis Mulsant provide
effective biocontrol of Icerya purchasi Maskell on citrus
and fruit crops. The parasitoid, Pseudaphicus malinus
Gahan, is successfully cultured and released against
Pseudococcus comstocki Kuwana on mulberry and fruit
crops. The entomophages, Cryptolaemus montrouzieri
Mulsant and Coccophagus gurneyi Compere, are used to
control Pseudococcus gahani Green. The introduced
predator Scutellista cyanea Motschulsky is applied against
Ceroplastes japonicus Green on citrus and ornamental
crops.

The Use of Natural Entomophage
Populations

Considerable attention is paid in the former U.S.S.R. to the
systematics and ecology of entomophages that are parasites
and predators of agricultural crop pests.

Much work has been carried out oc studies of taxonomy
and ecology of entomophages and the preparation and
publication of keys to different groups of parasites and
predators of agricultural pests. The institutions engaged in
this research include the Institute of Zoology of the
Academy of Sciences of the former U.S.S.R., zoological
institutes of some republic academies, higher educational
institutions, the AUIPP, and the AUIBMPP.

In Middle Europe more than 10,000 phytophagous species
are recorded, including 2,620 species of predators and
6,350 species of parasites. The majority of agricultural
pests recorded in the former U.S.S.R. are associated with
no less than 10-20 entomophagous species. For example,
500 species of natural enemies have been recorded on 50
species of cabbage pests. During the growing season, they
suppress 50-60 percent of the population of the cabbage
moth, cabbage root fly, whiteflies and aphids. The
entomophages of com bugs, Telenominae, often suppress
60-80 percent of the pest eggs. The aphidophages encoun¬
tered on winter wheat and maize (coccinelids, syrphids,
etc.) can (at certain "predator:host" ratios) eliminate the
aphid population completely (Tryapitsin et al. 1982,
Voronin 1988).

If insecticides are not applied, entomophagous complexes
can considerably reduce the number of pests of field,
technical and perennial crops. The process is stimulated by
special procedures for conservation and activation of
natural populations of entomophages (sowing of
nectariferous plants, sowing of grass, and creation of
entomological reservations). Chemicals are not needed
when 250-300 entomophages are encountered per 100
plants. Over 190 species of parasites and predators of
insect pests and mites occur in the cotton agroecosystem
(Narzikulov and Kovalenkov 1985). Of special importance
is the regulating role of such entomophages as Carabidae,
Staphylinidae, Coccinellidae, Chrysopidae, Syrphidae,
Ichneumonidae, Braconidae, Chalcididae, Aphidiidae,
Tachinidae and Aphelinidae. The use of natural popula¬
tions of predators and parasitoids of agricultural pests is
one of the most important trends in biological control in the
former U.S.S.R. and annually allows elimination of
chemical treatments in the area of 10-11 million hectares.

4

Microbiological Pest Control

A number of biopreparations based on different strains of
the entomopathogenic bacterium Bacillus thuringiensis
Berliner have been developed and produced by the micro¬
biological industry in the former U.S.S.R. (Stroeva and
Juschenko 1986, Kandibin 1989). Recommendations have
been developed on the use of bacterial preparations against
a number of cotton pests (Agzamova et al. 1988). At
present the following biopreparations are manufactured:

• Entobacterin - dry powder, titre not less than 30 billion
spores/g, biological activity not less than 100 Au/g.

• Dendrobacillin - dry powder, titre not less than 100
billion spores/g, biological activity 3,000 Au/g (con¬
centrated) and wettable powder, titre 60 billion spores/g,
biological activity 2,000 Au/g.

• Bitoxibacillin - dry powder, titre not less than 45 billion
spores/g, 0.6-0. 8 percent of exotoxin content, biological
activity not less than 1,500 Au/g.

• Gomelin - dry powder, titre not less than 90 billion
spores/g, biological activity not less than 2,700 Au/g.

• Lepidocid - dry powder, concentrated titre not less than
100 billion spores/g, biological activity not less than
3,000 Au/g.

The list of insect pests susceptible to Bacillus thuringiensis
includes about 150 lepidopterous species. Bacterial
insecticidal preparations are more specific than chemical
ones, are relatively safe for humans, animals and plants, do
not induce resistance in insects, have no negative effect on
the beneficial entomofauna, and are widely used against
Pieridae, Yponomeutidae, Tortricidae, Pyralidae,
Geometridae, Noctuidae, Lasiocampidae and Orgyidae.

Biopreparations based on Bacillus thuringiensis are applied
against the above-mentioned pests as aqueous suspensions
at rates of 1-5 kg per hectare. Two to three treatments of
cabbage, fruit crops, berries, grapes, sugar beet, alfalfa,
cotton, etc. are 70-90 percent effective. Bitoxibacillin,
recommended against various lepidopterous species, is
highly effective against larvae of the Colorado potato
beetle. Three treatments (2-3 kg/ha/treatment) result in 92-
97 percent reductions of pest numbers which is equally
effective as chemical insecticides. Commercial bacterial
entomopathogenic preparations are applied in the former
U.S.S.R. in an area of about 2 million hectares. Studies
have been carried out on the use of entomopathogenic fungi
against a number of agricultural pests in greenhouses and

outdoors. For example, biopreparation Boverin (a mixture
of spores and products of the vital activity of the fungus
Beauveria bassiana Bals.) has been recommended against
the Colorado potato beetle. The preparation is produced in
biolaboratories with a titre of 2-6 billion spores/g. It is
applied at the time of mass emergence of Colorado potato
beetle larvae at the rate of 2 kg/ha, with a sublethal dose of
chlorophos (0.4 kg/ha). Investigations are in progress on a
biopreparation of the fungus Entomophthora for controlling
aphids.

Intensive studies are being made on entomopathogenic
viruses which can be applied against insect pests. The
following viral preparations are being tested or have been
recommended for use in the former U.S.S.R.:

• Virin-EKS - titre not less than 1 billion polyhedra/ml.

It is recommended against small larvae of Mamestra
brassicae on cabbage, pea and beet at the rate of 0. 10-
0. 15 kg/ha.

• Virin-GYAP - titre not less than 3 billion granules/ml.

It is recommended against Laspeyresia pomonella at the
beginning and peak of larval emergence at the rate of 0.3
kg/ha.

• Virin-KSH - titre not less than 1 billion polyhedra/ml.

It is being tested against 1st and 3rd instars of
Malacosoma neustria larvae on fruit crops and forest
trees at the rate of 0.2 kg/ha.

• Virin-OS - titre not less than 3 billion granules/g. It is
being tested on cotton and other crops against 1st and
2nd instars of Scotia segetum at the rate of 0.3 kg/ha.

• Virin-HS - titre not less than 7 billion polyhedra/g. It is
being tested on cotton against 1st and 2nd instars of
Helicoverpa armigera at the rate of 0.3 kg/ha.

• Virin-ENSH - titre not less than 1 billion polyhedra/ml.
It is recommended on fruit and forest trees during
Lymantria dispar reproduction, at a density of 0.5-2 egg
clusters/tree and at the rate of 2 ml/egg cluster.

• Virin-ABB - titre 3 billion polyhedra and granules/ml.

It is recommended in orchards and forests on mulberry
against Hyphantria cunea Drury at the rate of 0. 1 kg/ha.

Methods are under development for mass rearing of
phytophagous insects necessary for the production of viral
preparations. Techniques for commercial production of
high-quality preparations are being improved.

5

The bacterial preparation of Isachenko bacterium Salmo¬
nella enteridis developed by All-Union Institute of Agricul¬
tural Microbiology has been recommended for microbio¬
logical control of rodents. It is mixed with grain having a
titre of not less than 1 billion spores/g and it is produced as
an aminobonal preparation with a titre of not less than 0. 1
billion spores/g. The grain bacterodencid is applied
without bait, while the aminobonal preparation is mixed
with bait (boiled grain, potatoes, etc.) before use. The
mice, Mus mus cuius L. and Apodemus silvaticus L. , and
the vole Microtus agrestis L. are most susceptible to
bacterodencid. The preparation is applied in the field at the
rate of 1-2 kg/ha and in storehouses at the rate of 100 g/100
m2. The biological effectiveness of bacterodencid is 75-80
percent, greatly depending on the quality of the preparation
and is produced in biolaboratories.

Conclusion

The assortment of biological preparations for plant protec¬
tion from agricultural pests allows considerable reduction in
the use of chemical insecticides. It is possible to use
biocontrol to reduce chemical treatments on vegetables and
potatoes by 50-60 percent outdoors and by 35-40 percent in
horticulture. At present, it is extremely urgent to replace
chemical pesticides with biological, cultural and selection
methods in the regions of dietetic food and child food
production, at resorts, water-preservation zones, agricul¬
tural regions around industrial centers and other zones of
great ecological danger.

To increase the scale of application of biological control of
agricultural pests, it is necessary to carry out further
investigations, especially in the following urgent trends:

• Increase in the number of biological methods for
suppression of outdoor pests such as Coleoptera,
Aphidinae, Diptera, Acridodea.

• Improvement of techniques of production and applica¬
tion of biological control of agricultural pests, including
development of up-to-date technical equipment for
biological control, and new types of industrial
biolaboratories.

• Development of methods of activation of natural
populations of entomophages and entomopathogens.

• Expansion of international cooperation in the introduc¬
tion of the beneficial organisms.

References Cited

Adashkevich, B. P., Z. Kh. Saidova Z. et al. 1988.
Perspektivi primenenia Brecon hebetor Say v borbe s
vrediteliami khlopchatnika (review) (In Russian).

Tashkent. Uxb. NIINTI.

Agzamova, X. K., K. Bababekov et al. 1988.
Microbiologicheskie sredstva zaschiti khlopchatnika (In
Russian). Zaschita rastenii 5: 16-17.

Alimukhamedov, S. N. [ed.] 1978. Integrirovannii metod
zaschity khlopchatnika i soputstvujusch kultur ot vreditelei,
boleznei i somyakov (In Russian). In Trudi
Sredneaziatskogo Nil zaschiti rastenii. Tashkent.

Alimukhamedov, S. N. 1983. Issledovania po zaschite
khlopkovogo polia (In Russian). Zaschita rastenii. 2: 17-
20.

Alimukhamedov, S. N., B. P. Adashkevich, and Z. K.
Adylov. 1989. Biologicheskaja zaschita khlopchatnika (In
Russian). Tashkent.

Alimukhamedov, S. N., Z. K. Adylov et al. 1989.
Kompetentnoje primenenie biologicheskih sredstv v
integrirovannoi zaschite khlopchatnika ot vreditelej
(recomendatsii) (In Russian). Moscov. Agropromizdat.

Davletshina, A. G. [ed.]. 1972. Entomophagi
glavneishikh vreditelei khlopchatnika v Uzbekistane (In
Russian). Tashkent.

Filippov, N. A. 1986. Problem! biologicheskoj zaschiti
rastenii (In Russian). Zaschita rastenii 12: 2-5.

Filippov, N. A. 1988. Perspektivi proizvodstva i
primenenia biologicheskih sredstv (In Russian). Zaschita
rastenii 5: 3-6.

Filippov, N. A. 1989. The present state and future
outlook of biological control in the U.S.S.R. - Acta
Entomologica Fennica 53: 11-18.

Golyshin, N. M., and Sh. M. Greenberg. 1988.
Trichogramma v zaschite rastenij (In Russian). In: Sboraik
trudov VNII biometodov. Moscov, Agropromizdat.

6

Greenberg, Sh. M., and P. V. Nikonov. 1988.
Trichogramma: vozmozhnosti, perspektivi (In Russian).
Zaschita rastenii 7: 23-27.

Gubenko, A. T. 1988. Biologicheskii metod v
proizvodstvo (In Russian). Zaschita rastenii 9: 3-7.

Kandibin, N. V. 1989. Bacterialinie sredstva borbi s
grizunami i vrednimi nasekomimi (In Russian). Moscov.
Agropromizdat.

Khamraev, A. Sh. 1990. Reservi povtshenia effectivnosti
zaschiti khlopschatnika ot vreditelei (In Russian). Zaschita
rastenii 3: 10-12.

Khamraev, A. Sh., and Sh. A. Sharafatdinov. 1989. Puti
povishenia effectivnosti biologicheskogo metoda borbi s
vrediteliami khlopchatnika (In Russian). Zaschita rastenii
1: 10-12.

Khodjaev, Sh. T. 1990. Ispolzovanie polovih pheromonov
nasekomih v Uzbekistane (In Russian). Zaschita rastenii 1:
21-23.

Kosenko, A. F. 1989. Integrirovannaja zaschita:
sostojanie i perspektivi (In Russian). Zaschita rastenii 10:
6-9.

Kovalenkov, V. G. 1989. Osnova biologicheskoi zischiti
khlopchatnika (In Russian). Zaschita rastenii 2: 13-16.

Kovalenkov, V. G. 1990. Primenenie biologicheskogo
metoda zaschiti rastenii v Tajikistane (reviw) (In Russian).
Dushanbe, Tajik NIINTI.

Maksumov, A. N., and M. N. Narzikulov. 1981.
Integrirovpnnaja zaschita knlopchatnika ot vreditelei (In
Russian). Dushanbe.

Melnikov, N. N. 1989. Pesticidi i okruzhajuschaja sreda
(In Russian). Zaschita rastenii 4: 4-8.

Narodnoie khoziaistvo SSSR v 1988. Annual statistics,
Moscow.

Narzikulov, M. N., and V. G. Kovalenkov. 1985.
Ekosistemnyi podhod k zaschite khlopchatnika (In Rus¬
sian). Zaschita rastenii 12: 19-20.

Nikonov, P. V. 1986. Razvitie biologicheskogo metoda
(In Russian). Zaschita rastenii 5: 2-4.

Nikonov, P. V. 1988. Ispolzovanie biologicheskih sredstv
(In Russian). Zaschita rastenii 6: 3-5.

Novozhilov, K. V. 1989. Zaschita rastenii - puti
optimizaisii (In Russian). Zaschita rastenii 2-6.

Orlovskaya, E. V. 1983. Virusnie preparati dlya borbi s
vrednimi nasekomimi (In Russian). Moscov,
Glavmikrobioprom.

Stroeva, I. A., and N. P. Juschenko. 1986. Bakterialinie
entomopatogennie preparati (In Russian). Zaschita rastenii.

Tryapitsin, V. A., V. A. Shapiro, and V. A.
Schepetilnikova. 1982. Paraziti i khischniki vreditelei
selskokhoziaistvennikh kultur (In Russian). Leningrad,
Kolos.

Voronin, K. E. 1988. Ohrana sredi pri intensivnih
technologiah (In Russian). Zaschita rastenii 6: 8-10.

7

Table 1. Areas of biocontrol of main pests in the former U.S.S.R.

Pest species

Volume of work in 1989 (1,000 ha)

Total

Included means

Chemical Biological

Rodents -Apodemus

11,772.4

7,500.6

4,271.8

Meriones

Microtus sp.

Pyrausta sticticalis D.

7,437.3

5,329.4

2,107.9

Agrotis sp.

4,604.8

132.6

4,472.2

Mamestra, Spodoptera sp., etc.

4,140.4

175.1

3,965.3

Ostrinia nubilalis Hb.

1,967.9

17.8

1,950.1

Helicoverpa armigera Hb.

5,421.5

453.0

4,968.5

Leptinotarsa decemlineata Say

6,492.1

6,187.0

305.1

Vegetable pests

2,088.6

1,285.1

803.5

Fruit pests

4,319.8

3,699.0

620.8

Grape pests

1,972.7

1,830.3

142.4

Pests of cotton complex

8,992.1

2,687.0

6,305.1

Table 2. Economic effectiveness of application of biocontrol in the former U.S.S.R. in 1 970-1 3881

No.

Indices 1970

1975

1980

1985

1986

1987

1988

1.

Volume of biomeans

application million hectares 4. 1

10.7

17.0

22.6

24.4

24.1

26.7

including,

entomophages 3.9

7.6

11.9

16.8

17.6

18.3

19.2

biopreparations 0 . 2

3.1

5.1

5.8

6.8

5.8

7.5

2.

Value of extra (preserved)
produce, million roubles 144.0

396.0

596.0

825.0

880.0

875.0

970.0

3.

Expenditures on application of
biological means, including costs
on harvesting and sale of extra
produce, million roubles 84.0

235.0

345.0

488.0

510.0

513.0

569.0

4.

Conditional net profit,

million roubles 60.0

161.0

251.0

337.0

360.0

362.0

401.0

1 Data obtained by the lab of economics of AUIBMPP

8

Table 3. The area of irrigated lands in the former U.S.S.R. republics growing cotton (1,000 hectares)

Republics

1970

1980

1985

1986

1987

1988

Former U.S.S.R.

11,100

17,487

19,951

20,487

20,485

20,782

Uzbekistan

2,696

3,476

3,930

4,020

4,109

4,149

Kazakhstan

1,451

1,961

2,172

2,231

2,318

2,260

Azerbaijan

1,108

1,195

1,318

1,330

1,345

1,370

Kyrgyzstan

883

955

1,009

1,020

1,028

1,037

Tajikistan

518

617

653

662

675

686

Turkmenistan

643

927

1,107

1,185

1,224

1,258

'Narodnoje khozyaistvo SSSR b 1983 godu." Annual Statistics, Moscow, 1989.

Table 4. Supplies of mineral fertilizers to the former U.S.S.R. agriculture (per 100% of nutritional
substances)

1970

1980

1985

1986

1987

1988

1 ,000 tons

All mineral fertilizers

10,317

18,763

25,395

26.514

27,412

27,196

including:

nitric

4,605

8,262

10,950

11,475

11,787

11,587

phosphorous

3,133

5,590

7,615

8,354

8,564

8,556

potash

2,574

4,904

6,822

6,677

7,052

7,044

kg/I ha

of arable

land

All mineral fertilizers

46.8

83.9

113.2

118.1

122.1

122.4

including:

nitric

20.9

.36.9

48.8

51.1

52.5

52.7

phosphorous

14.2

25.0

34.0

37.2

38.2

38.2

potash

11.7

21.9

30.4

29.8

31.4

31.5

‘Narodnoje khozyaistvo SSSR in 1988 godu."

Annual Statistics, Moscow, 1989.

Table 5. Supplies of Plant Protection chemicals to the former U.S.S.R. agriculture (in 100%

according-to a.i.) in 1,000 tons

1970

1980

1985

1986 1987

1988

All plant protection chemicals 170

279

362

346 333

308

Including herbicides 50

127

160

172 169

148

‘Narodnoje khozyaistvo SSSR in 1988 godu." Annual Statistics, Moscow, 1989.

Table 6. Gross yield of raw cotton in the former U.S.S.R. republics (1,000 tons)

Republics

Average per year

1971

-1975

1976

-1980

1981

-1985

1986

1987

1988

Former U.S.S.R.

7,667

8,547

8,314

8,234

8,084

8,689

Uzbekistan

4,895

5,359

5,159

4,989

4,858

5,365

Kazakhstan

305

317

302

333

312

325

Azerbaijan

441

627

707

784

697

616

Kyrgyzstan

205

208

87

68

73

79

Tajikistan

810

906

917

922

872

963

Turkmenistan

1,011

1,130

1,142

1,138

1,272

1,341

Including fine fibre

cotton

Former U.S.S.R.

656.7

794.0

1,087.7

1,187.6

1,182.8

1,333.5

Uzbekistan

204.6

301.0

494.5

583.7

534.2

561.0

Tajikistan

257.0

267.6

289.2

308.7

276.4

328.9

Turkmenistan

194.1

225.4

304.0

295.2

372.2

448.6

"Narodnoje khozyaistvo SSSR v 1988 godu." Annual Statistics, Moscow, 1989.

Table 7. Cotton yield in the republics of the former U.S.S.R. (centners/ha)

Average per year

Republics

1971

-1971

1976

-1976

1981

-1985

1986

1987

1988

Former U.S.S.R.

27.3

28.1

Raw cotton

25.6 23.7

22.9

25.3

Uzbekistan

28.5

29.4

26.7

24.3

23.1

26.6

Kazakhstan

26.6

27.0

23.3

25.9

24.4

25.4

Azerbaijan

21.6

27.2

23.9

26.1

23.0

20.6

Kyrgyzstan

27.6

28.3

19.1

23.5

23.5

24.7

Tajikistan

30.7

30.7

29.8

29.5

26.9

30.1

Turkmenistan

23.1

22.4

21.4

17.5

20.1

21.1

Former U.S.S.R.

23.1

22.7

Fine fibre cotton

24.6 23.9

24.2

24.6

Uzbekistan

27.6

28.0

30.2

28.8

26.9

27.6

Tajikistan

31.5

30.7

29.7

31.5

30.3

31.3

Turkmenistan

15.2

14.5

16.9

15.0

18.8

18.9

‘Narodnoje khozyaistvo SSSR v 1988 godu." Annual Statistics, Moscow, 1989.

10

Table 8. Gross yield of raw cotton (fibre) in the former U.S.S.R. republics (1,000 tons)

Rebublics

Average per year

1971

-1975

1976

-1980

1981

-1985

1986

1987

1988

Former U.S.S.R.

2,474

2,612

2.453

2,660

2,502

2,748

Uzbekistan

1,580

1,620

1,509

1,622

1,505

1,719

Kazakhstan

103

97

93

108

96

101

Azerbaijan

147

203

212

262

225

203

Kyrgyzstan

68

65

26

21

20

24

Tajikistan

262

289

278

293

276

299

Turkmenistan

314

338

335

354

380

402

Raw cotton yield (fibre) in the former U.S.S.R. republics (centners/ha)

Former U.S.S.R.

8.8

8.6

7.6

7.7

7.1

8.0

Uzbekistan

9.2

8.9

7.8

7.9

7.1

8.5

Kazakhstan

8.9

8.3

7.2

8.4

7.5

7.9

Azerbaijan

7.2

8.8

7.2

8.4

7.4

6.8

Kyrgyzstan

9.2

8.9

5.8

7.0

6.5

7.5

Tajikistan

9.9

9.8

9.0

9.4

8.5

9.3

Turkmenistan

7.2

6.7

6.3

5.4

6.0

6.3

'Narodnoje khozyaistvo SSSR v 1988 godu." Annual Statistics, Moscow, 1989.

Table 9. Agricultural crop areas in cotton growing republics of the former U.S.S.R. in 1988 (1,000 ha)

Rebublics

Total land under
cultivation

Cereals

Technical crops
including cotton

Including

Vegetable and
melon crops Forage crops

Former U.S.S.R.

211,500

114,912

14,497

8,501

73,640

Uzbekistan

4,349

1,049

2,057

235

1,008

Kazakhstan

35,658

24,290

481

327

10,560

Azerbaijan

1,468

539

321

73

535

Kyrgyzstan

1,314

552

63

51

648

Tajikistan

848

241

331

48

228

Turkmenistan

1,243

198

637

75

333

'Narodnoje khozyaistvo SSSR v 1988 godu.' Annual Statistics, Moscow, 1989.

11

Table 10. Agricultural crop yield in cotton-growing republics of the former U.S.S.R. in 1988
(centners/ha)

Republics

Cereals

Potatoes

Vegetables

Former U.S.S.R.

17.0

103

157

Uzbekistan

20.9

99

209

Kazakhstan

9.3

112

169

Azerbaijan

26.3

70

202

Kyrgyzstan

31.9

146

201

Tajikistan

15.6

163

209

Turkmenistan

22.0

100

124

"Narodnoje khozyaistvo SSSR v

1988 godu.’ Annual Statistics, Moscow, 1989.

Table 11. The area under fruit and berry crops and vineyards in cotton-growing republics of the former
U.S.S.R. in 1988 (1,000 hectares)

Republics

Fruits and berries

Grapes

Total

Including

fruit-bearing

Total

Including

fruit-bearing

Former U.S.S.R.

2,848

2,218

1,105

884

Uzbekistan

212

136

128

92

Kazakhstan

98

72

26

18

Azerbaijan

135

96

195

169

Kyrgyzstan

47

34

9

6

Tajikistan

71

51

33

25

Turkmenistan

26

15

28

21

'Narodnoje khozyaistvo SSSR v 1988 godu." Annual Statistics, Moscow, 1989.

12

Table 12. Gross yield and yielding capacity of fruit and berry crops and grapes in cotton - growing republics
of the former U.S.S.R. in 1988

Republics

Fruits and berries

Grapes

Gross yield
(1,000 tons)

Yielding

capacity

(centners/ha)

Gross yield
(1,000 tons)

Yielding

capacity

(centners/ha)

Former U.S.S.R.

8,909

39.8

5,581

61.8

Uzbekistan

623

45.6

655

70.8

Kazakhstan

275

37.9

94

49.9

Azerbaijan

427

44.3

1,250

73.7

Kyrgyzstan

145

42.3

28

44.6

Tajikistan

212

41.4

178

71.8

Turkmenistan

50

34.2

165

80.8

'Narodnojc khozyaistvo SSSR v 1988 godu." Annual Statistics, Moscow, 1989.

Table 13. Areas of biocontrol means application in the former U.S.S.R. in 1970-1989 (single estimate)

Year

Outdoors (million ha)

Greenhouse (1,000 ha)
Total entomophagesand
microbiopreparations

Total

Entomophages

Microbiopreparations

1970

4.1

3.9

0.25

0.05

1975

10.7

7.6

3.1

1.0

1980

17.0

11.9

5.1

4.2

1985

22.6

16.7

5.9

10.4

1986

24.4

17.6

6.8

11.5

1987

24.1

18.4

5.7

13.7

1988

26.4

19.3

7.4

14.3

1989

27.2

19.7

7.5

18.0

13

Table 14. Application areas of different biocontrol means in greenhouses
in the former U.S.S.R. in 1989 (single estimate)

Biocontrol means

Area - 1.000 ha

Entomophages

Phytoseiulus persimilis

6,160

Encarsia formosa

1,857

Aphidoletes aphidimyza

348

Aphidius matricariae

146

Amblyseius makenziei

399

Chrysopa earned

21

Macrolophus costalis

7

Micromus angulatus

14

Cycloneda limbifer

37

Microbiopreparations

Trichoderma lignorum

5,401

Verticillium lecanii

1,293

Salmonella enteritidis

711

Bitoxibacillin, B. thuringiensis

916

Beauveria bassiana

140

Ampelomyces sp.

184

As cher sonia sp.

31

Risoplan

49

Total

18,000

14

Table 15. Application areas of different biocontrol means outdoors
in the former U.S.S.R. in 1989

Bicontrol means

Area - 1,000 ha

Entomophages

Trichogramma sp.

17,124.6

Bracon hebetor

1,974.8

Chrysopa carnea

446.3

Phytomysa orobanchia

124.5

Metaseiulus sp.

15.2

Cryptolaemus sp.

7.6

Encarsia partenopea

5.1

Scutellista cyanea

0.11

Telenomus sp.

0.5

Total 19,685.7

Microbiopreparations

Salmonella enteritidis

5,299.7

Bacillus thuringiensis

0.0

Bitoxibacillin BT

819.0

Dendrobacillin BT

837.5

Lepidocide BT

758.3

Gomelin BT

35.0

Trichoderma lignorum

3.1

Ampelomyces sp.

0.5

Fitolavin

2.75

BVK - lisin

9.3

Total

7,778.1

15

Table 16. Volumes of biocontrol application on agricultural crops (former U.S.S.R., 1985-2005, 1,000 ha)

1985

1990

1995

2000

2005

Crops

BC1 Incl.

means MBP2
total

BC

means

total

Incl.

MBP

BC

means

total

Incl.

MBP

BC

means

total

Incl.

MBP

BC

means

total

Incl.

MBP

Cereal and leguminous

4,427.9 1,626.8

6,500

3,195

7,000

3,387

7,300

3,66

7,600

3,800

including:

com

1,795.3 57.4

2,100

820

2,200

840

2,250

865

2,300

880

etc.

1,564.6 1,554.7

2,000

1,075

2,200

1,140

2,350

1,295

2,300

1,340

legumes

925.4 14.7

2,400

1,300

2,600

1,400

2,700

1,500

3,000

1,580

Technical

8,755.5 640.9

9,900

4,220

11,000

4,256

11,250

4,400

11,500

4,500

including:

cotton

4,375.3 481.6

5,000

2,420

5,600

2,826

5,700

3,130

5,750

3,200

sugar beet

4,004.6 154.5

4,550

1,000

4,800

855

4,880

940

4,950

950

sunflower

343.3 4.8

350

20

600

175

670

225

800

350

Vegetables and melons

1,510.3 743.8

2,300

780

4.200

2,275

7,000

4,900

7,500

5,000

including:

potatoes

260.5 245.1

500

260

900

600

3,600

3,000

3,650

3,040

vegetables

1,232.8 498.7

1,750

520

3,210

1,665

3,600

1,880

3,600

1,920

melons

17.0

50

90

10

100

20

250

40

Forage crops

2,603.8 1,328.9

2,900

1,450

3,200

1,600

3,650

1,720

4,000

1,750

including:
perennial grasses

2,595.2 1,327.8

2,850

1,440

3,100

1,500

3,400

1,500

3,800

1,670

Perennial plants

919.6 762.3

1,200

810

1,500

910

1,400

1,000

1,600

1,100

including:

orchards

798.1 643.2

990

670

1,150

760

1,180

810

1,620

860

vineyards

120.1 119.1

210

140

350

150

220

190

280

240

Total

22,597.9 5,885.7

27,500

10,500

30,100

12,500

32,500

15,850

35,000

16,500

1 Biocontrol

2 Microbiopreparations

16

Cotton Insect Problems in the United
States -- An Overview

D. D. Hardee 1

Abstract

Eighteen major and eleven minor insects in 1989 caused an
estimated loss of more than $686 million to cotton produc¬
tion in the United States. The relative importance of each
of the insects is elaborated with reference to which geo¬
graphical area the total damage is done. Among several
items listed, current challenges in cotton insect management
include continuing increases in insecticide resistance, cost
of registration and re-registration, etc.

Introduction

In 1987, cotton ( Gossypium spp.) was grown in 90 coun¬
tries on 32.6 million hectares which produced 48.4 million
tons of seed cotton with an average yield of 1,485 kg seed
cotton per ha (Green and Lyon 1989). Seventy-five percent
of the area grown in cotton was in developing countries,
but they accounted for only 62% of total production; the
major producers in order were China, U.S.A., the former
U.S.S.R., Pakistan, and India. Cotton has been one of the
major agricultural crops grown in the United States since
the late 1700s (Handy 1896) and currently ranks as the
fourth most valuable crop in the U.S. after com, soybeans,
and wheat. A total of 3.86 million ha of cotton was
harvested in 1989 with an average yield of 542 kg lint per
ha or a total of 12.1 million bales (218 kg/bale). Texas,
California, and Mississippi were the three leading states in
production and accounted for almost 60 % of the total crop
(King et al. 1990). The estimated value of the 1989 crop
(from 15 states in the south and west) was near $4 billion
for lint alone, with an additional $507 million for cotton¬
seed (Anonymous 1990).

Cotton is a plant seemingly designed especially for insects,
since Young (1969) lists 13 major and 13 minor insects
which attack cotton in the U.S. In a previous review of the
evolution of cotton insect management systems in the U.S. ,
Ridgway and Lloyd (1983) noted that insects were reported
to cause extensive damage to cotton as early as the 1800s.
Since that time, depending on changes in production
systems that resulted from introduction of new pests and
improved equipment, insect damage to cotton has changed
many times. Perhaps the greatest change

1 Southern Insect Management Laboratory, USDA, ARS, Stoneville,
Mississippi

resulted from the extension of the range of the boll weevil,
Anthonomus grandis grandis Boheman, from Mexico in
1892 (Howard 1894). This insect had spread eastward
from Texas to North Carolina by 1926 leaving devastation
and poverty in its path. Since 1950, two major events
shaped future cotton insect control. These were (1)
development of resistance by the boll weevil to chlorinated
hydrocarbon insecticides and the resulting development of
organophosphates as their replacement, and (2) emergence
of the tobacco budworm, Heliothis virescens (F.), as a
major pest of cotton. A third event of importance was the
acceptance in the late 1970s through the present of pyre-
throids as the insecticides of choice for budworms and
bollworms, Helicoverpa zea (Boddie).

Arthropod pests reduced yield of the 1989 cotton crop by
9.2% (Head 1990). This amounted to a loss of 1.1 million
bales or $319.7 million in revenue. Furthermore, the
average cost for controlling arthropod pests was $93. 49/ha
across the U.S. Cotton Belt. Thus, the estimated total costs
due to arthropod pests, that is reduced yield plus cost of
treatment, was $686.1 million.

In this report, I will first present the current status and
economic importance of each major pest that occurs in
cotton from planting to harvest. Next, I will describe the
minor insect problems (that must be dealt with) and their
potential for becoming major problems. Finally, I will
relate to you the current challenges that we face in cotton
insect management in the U.S.

Major Insect Problems

Major insect pests are presented in the order in which they
attack cotton (from seedling stage to harvest).

Cutworms (Various species; Lepidoptera: Noctuidae)

Cutworms damage cotton in the seedling stage by cutting
stems completely at the soil surface and occasionally
feeding on the leaves. Several species of cutworms,
including the variegated cutworm, Peridroma saucia
(Hubner), may develop in weeds or other crops, especially
legumes, and subsequently attack cotton planted on land
previously in weeds or legumes (Young 1969). Cotton
planted in wet, grassy areas or growing in an area receiving

17

excessive rainfall may have to be replanted due to feeding
by cutworms resulting in stand reduction. Cutworms are
considered minor pests in some areas, and no reliable cost
estimates are available.

Seedling Thrips

Onion thrips, Thrips tabaci Linde man; Tobacco thrips,
Frankliniella fusca (Hinds) (Thysanoptera: Thripidae).
Adult and nymphal thrips attack cotyledons and new leaves
of the cotton plant by piercing leaf tissues with their sharp
mouthparts and sucking the juices (Head and Williams
1983). This feeding damage can substantially reduce cotton
stand and necessitate occasional replanting. Excessive
feeding may cause "possum-eared," stunted cotton which
may ultimately yield less and delay maturity by as much as
2-3 weeks.

Mirids

Cotton fleahopper, Pseudatomoscelis seriatus (Reuter)
Plant bugs, Lygus lineolaris (Palisot de Beauvois); Lygus
hesperus Knight (Heteroptera: Miridae). The entire cotton
plant may be injured during feeding by mirids which
remove plant sap and inject insect enzymes into the plant.
Feeding may occur at all stages of growth of the plant, but
most damage occurs from the 4th leaf stage to bloom,
during which time leaves become deformed and ragged,
pinhead squares turn brownish black and are shed (blasted
squares), and developing pollen is destroyed. Feeding on
mainstem terminals can produce abnormal branching
("crazy cotton") and delayed fruiting in the cotton plant.
Lygus hesperus is prominent in the western U.S. (Leigh et
al. 1988), P. seriatus is the predominant species in Texas
(Sterling et al. 1988), and L. lineolaris primarily fills the
niche in the rest of the Cotton Belt (Scott et al. 1986).

Boll Weevil

Anthonomus grandis grandis Boheman (Coleoptera:
Curculionidae). Uncontrolled populations of the boll
weevil can effectively cause 100% damage in a cotton field,
which accounts for its status as the number one insect
problem in the U.S. (Cross 1973). As seen in Table 1, the
boll weevil was the leading pest of cotton in 1989. Adults

may cause minor damage by feeding on seedlings and
tender foliage, but the main concern is oviposition in
squares (flower buds) or bolls, followed by larval develop¬
ment inside fruiting forms with subsequent abscission of
squares and lint damage in bolls.

As of this writing, parts of the U.S. are involved in a boll
weevil eradication program financed jointly by the produc¬
ers and the USDA Animal and Plant Health Inspection
Service (APHIS). The boll weevil essentially has been
eliminated from Virginia, North Carolina, and part of
South Carolina, and these states are currently part of a
containment program. Current APHIS-coordinated boll
weevil cooperative programs include the Southeast Boll
Weevil Eradication Program (approximately 500,000 acres
of cotton in Florida, Georgia, South Carolina, and southern
Alabama), the West Texas Containment Program (270,000
acres of cotton in the High and Rolling Plains of West
Texas), and the Southwest Boll Weevil Eradication Pro¬
gram (including approximately 300,000 acres of cotton in
western Arizona, southern California and northwestern
Mexico) (Anonymous 1989). Utilizing insecticides and
pheromone traps, and through a series of grower referenda
and expanding technologies and successes, the long-range
plan is to free the U.S. entirely of boll weevils over the
next 10-15 years by forcing it back into Mexico, followed
by a containment barrier across northern Mexico.

Boll worms/Bud worms

Bollworm, Helicoverpa zea (Boddie); Tobacco budworm,
Heliothis virescens (Fabricius) (Lepidoptera: Noctuidae).
The bollworm and the budworm together are the "boll-
worms" attacking cotton. Larvae hatch and feed wherever
eggs are laid, and they may destroy terminals, buds, and
young squares, as well as eat significant portions of older
squares and bolls to the extent that fruiting forms are
destroyed (Head 1981). The bollworm occurs throughout
the Cotton Belt, whereas the budworm is primarily a pest of
cotton from Texas eastward. The budworm is more
difficult to kill than the bollworm, and resistance to
organophosphate and pyrethroid insecticides is more
common in the budworm.

Resistance management is currently being recommended
(Luttrell and Roush 1987) and used to help curb increasing
occurrences of resistance to pyrethroid insecticides. This
program consists of (1) use of non-pyrethroid insecticides

18

from planting until bloom; (2) where needed, selective use
of pyrethroids for a 6-week period after bloom; and (3) use
of non-pyrethroids thereafter in late season. Data for 1989
(Table 1) show that the bollworm/budworm complex was
third in importance in 1989.

Pink Bollworm

Pectinophora gossypiella (Saunders) (Lepidoptera:
Gelechiidae). The pink bollworm is of major importance in
Arizona and California, but can be found in Texas, Okla¬
homa, New Mexico, Arkansas, and Louisiana (Henneberry
1986). This insect may affect the yield and quality of
cotton in several ways, including shedding of squares and
small bolls, increase of boll rot from larval feeding,
reduction in seed weight and viability from larval feeding,
and discoloration of lint by feeding and movement of larvae
inside the boll. In addition, strength and maturity of fibers
may be diminished as a result of larval feeding.

Cotton Aphid

Aphis gossypii Glover (Homoptera: Aphididae. Aphids
cause yield losses in cotton by sucking juices necessary for
normal boll development (Young 1969). Aphids also
secrete a honey-like substance ("honey dew") which causes
leaves to be sticky, and contributes to the formation of a
black, sooty mold on cotton lint which lowers quality and
causes problems in spinning. Until recently, the cotton
aphid was considered a minor and sporadic pest of cotton.
Increased usage of pyrethroid insecticides, which were
ineffective against aphids in the late 1970 's and early
1980's, and confirmed development of resistance to
commonly used aphicides (O'Brien et al. 1990), have led to
the emergence of this insect as a major problem in cotton
production. Serious aphid infestations have developed
earlier in cotton and in higher numbers in the mid-south
during the past three years. These infestations have often
remained throughout the season, and are difficult and costly
to control (Hardee and O'Brien 1990).

Armyworms

Beet armyworm, Spodoptera exigua (Hubner) Fall army-
worm, Spodoptera frugiperda (J. E. Smith) (Lepidoptera:
Noctuidae). Beet and fall armyworms tend to cause
sporadic pest problems in cotton, but recently they have
become a problem somewhere in the U.S. every year.
Historically, when armyworms appeared on cotton they
were known as foliage feeders, but their behavior has now
changed to the point of feeding on squares, blooms, and
bolls, as well as foliage (Smith 1989). Control with
insecticides is costly, and the problem is amplified by their
habit of feeding inside closed blooms which decreases the
effectiveness of insecticides.

Whiteflies

Sweetpotato whitefly, Bemisia tabaci (Gennadius);
Bandedwinged whitefly, Trialeurodes abutilonea
(Haldeman) (Homoptera: Aleyrodidae). The bandedwinged
whitefly occurs throughout the Cotton Belt, but the
sweetpotato whitefly occurs on cotton primarily west of
Texas. Both species can be a pest in cotton, especially in
late season, and may create "sticky cotton" problems
similar to, but often greater than, that of cotton aphids.

Both species (especially the sweetpotato whitefly) are
difficult and costly to control because of their distribution
inside the plant canopy on the underside of leaves, and their
apparent tolerance to all major insecticides (Butler et al.
1986).

Two Spotted Spider Mite

Tetranychus urticae Koch (Acari: Tetranychidae). A
number of spider mites attack cotton and are capable of
causing serious damage, but the two-spotted spider mite is
most common (Trichilo and Leigh 1986a). Feeding by
spider mites causes blotched and discolored leaves which
may abscise prematurely in heavy infestations. Discolora¬
tion and leaf drop reduce yield by decreasing photosynthe¬
sis by the plant. Spider mites may occur in all areas of the
Cotton Belt, but they usually are worse in the western U.S.
Outbreaks of spider mites usually occur late in the season
and are associated with dry, hot weather, and after insecti¬
cide applications have destroyed beneficial insects and
mites. Chemical control is costly and difficult because of
increasing resistance to available miticides (Leigh 1963).

19

Western Flower Thrips

Frankliniella occidentalis (Pergande) (Thysanoptera:
Thripidae). Western flower thrips occasionally attack
cotton after initiation of blooming with 300 thrips per
bloom not uncommon. Feeding on flower bracts and pollen
may result in premature bloom shed and incomplete boll
development, with subsequent decreased yield (Terry and
Barstow 1988). However, Trichilo and Leigh (1986b)
suggest that western flower thrips are an important predator
on spider mites in the western U.S.

Cotton Leafperforator

Bucculatrix ihurberiella Busck (Lepidoptera: Lyonetiidae).
The cotton leaf perforator is found primarily in Arizona and
California where it thrives in hot, dry climates (Young
1969). It feeds on cotton leaves causing premature defolia¬
tion in heavy infestations, resulting in square and boll drop
and subsequent loss in yield. Control with insecticides is
difficult and costly due to its resistance to many insecti¬
cides.

Minor Insect Problems

Each year several minor insects are reported to occur in
isolated areas of the U.S. Cotton Belt (Young 1969).
Damage and losses due to these insects are grouped in
Table 1 as "minor pests" and "new pests."

Cotton Square Borer

Strymon melinus Hubner (Lepidoptera: Lycaenidae). The
cotton square borer is seldom of economic importance since
its natural enemies usually hold it in check. Like the more
important bollworm, the square borer consumes the
contents of the square which causes it to flare and drop.

Garden Webworm

Achyra rantalis (Guenee) (Lepidoptera: Pyralidae). This
insect is present in most years in some states, but chemical
control is seldom exercised. They can be found feeding in
terminals or on young leaves, usually enclosed in their
webbing.

European Corn Borer

Ostrinia nubilalis (Hubner) (Lepidoptera: Pyralidae).
Damage to cotton from this insect has been reported in
several states in the southeast, especially in North Carolina
after the Boll Weevil Eradication Program was completed
in this state (Savinelli et al. 1986). Borers are known to
enter the stalk 5 to 15 m above ground and burrow up
through the center resulting in lodged plants and improper
maturing of bolls. At the end of the growing season, larvae
are also known to attack bolls in the middle third of the
cotton plant.

Loopers

Cabbage Looper, Trichoplusia ni (Hubner) Soybean
Looper, Pseudoplusia includens (Walker) (Lepidoptera:
Noctuidae). The cabbage looper has been an occasional
leaf feeder on cotton for many years, but the soybean
looper was identified for the first time in 1989 feeding on
cotton in several areas in the southeast (King et al. 1990).
The concern for this development is that soybean loopers
feeding on soybean are now resistant to most available
insecticides.

Mexican Mealybug

Phenacoccus gossypii Townsend & Cockerell (Homoptera:
Pseudococcidae). This insect has become a late-season pest
of cotton since 1987 and seems to be spreading each year
(T. F. Watson, Univ. of Arizona, personal communica¬
tion). It is currently limited to spotty infestations in about
40,000 acres of cotton in Arizona and part of Texas.
Damage occurs late in the season and can result in death of
the plant from excessive numbers in terminals, as well as
very heavy "honey dew" which aggravates the "sticky"
cotton problem from whitefly feeding.

Saltmarsh Caterpillar

Estigmene acrea (Drury) (Lepidoptera: Arctiidae). These
larvae usually feed on cotton foliage late in the season,
primarily in western irrigated areas.

20

Stink Bugs

Brown - Euschistus servus (Say), Green - Acrosternum
hilare (Say), Southern green - Nezara viridula (Linnaeus)
(Heteroptera: Pentatomidae). Damage from stink bugs
varies from year to year and from area to area. Feeding on
bolls results not only in lower yields through a reduced
number of harvestable locks, but also in lower quality of
lint and seed (Barbour et al. 1989).

Yellowstriped Army worm

Spodoptera ornithogalli (Guenee) (Lepidoptera:
Noctuidae). This armyworm is a day-feeding insect which
damages cotton by feeding on the leaves, usually in early
season, and occasionally by boring into squares and bolls
later in the season.

Current Challenges in Cotton Insect
Management in the United States

We are currently operating in the U.S. with several
challenges which will greatly affect our ability to produce
cotton efficiently in the future. These include:

• Continuing increase in insecticide resistance: Without
exception, all of the major cotton insect pests are
confirmed or suspected of being resistant to all or some
of the currently used insecticides.

• Loss of available insecticides: Several of our most
important insecticides used for cotton insects have been
removed from the market for environmental or legisla¬
tive reasons, as well as from lack of interest by private
industry because of re-registration costs.

• Cost of registration: The current estimated cost of
registering a new product from initial testing to avail¬
ability to the producer is estimated at $40-50 million.

• Lack of new chemistry: Very few new products are in
the testing stage by industry, and to my knowledge only
one company is testing potential products with new
chemistry.

• Current environmental issues: Public awareness and
concern for food safety, global warming, water quality,
and environmental safety are at an all-time high. We
must, therefore, be cautious and judicious in our
recommendations and use of agricultural chemicals to
prevent environmental damage (real or imagined) and
subsequent adverse publicity and litigation.

• Lack of alternative control strategies-methods: Other
control measures, such as genetic control, use of
microbials, pheromones, host plant resistance, beneficial
insects, or a combination of these, are all in their infancy
and offer few, if any, immediate solutions.

References Cited

Anonymous. 1989. National Boll Weevil. Cooperative
Control Program: Draft Environmental Impact Statement -
1989. USD A, APHIS DEIS 89-001, p. S-3.

Anonymous. 1990. Crop Values: 1989 Summary. USDA
National Agric. Statistics Service, Agric. Statistics Board,
Washington, D.C.

Barbour, K. S., J. R. Bradley, Jr., and J. S. Bacheler.

1989. Phytophagous stink bugs in North Carolina cotton:
An evaluation of damage potential. Proc. 1988 Beltwide
Cotton Prod. Res. Conf., pp. 280-281.

Butler, G. D., Jr., T. J. Henneberry, and W. D.

Hutchison. 1986. Biology, sampling and population
dynamics of Bemisia tabaci. Agric. Zool. Rev. 1: 167-95.

Cross, W. H. 1973. Biology, control, and eradication of
the boll weevil. Annu. Rev. Ent. 18: 17-46.

Green, M. B., and D. J. de B. Lyon. 1989. Pest Manage¬
ment in Cotton. Ellis Horwood Limited, Chichester,
England. 259 p.

Handy, R. B. 1896. History and general statistics of
cotton. In The Cotton Plant, U.S. Department of Agricul¬
ture Bulletin, p. 17-66.

Hardee, D. D., and P. J. O'Brien. 1990. Cotton aphids:
Current status and future trends in management. In 1990
Proc. Beltwide Cotton Prod. Research Conf. (In Press).

21

Head, R. B. 1981. Control cotton bollworms and tobacco
budworms. Miss. Coop. Ext. Serv. Information Sheet
1198, 2 p.

Head, R. B. 1990. Cotton Insect Losses - 1989. 43rd
Annual Conf. Report on Cotton Insect Research and
Control. Proc. 1990 Beltwide Cotton Prod. Res. Conf. (In
Press.)

Head, R. B., and M. R. Williams. 1983. Biology and
control of thrips. Miss. Coop. Ext. Serv. Information
Sheet 1226, 2 p.

Henneberry, T. J. 1986. Pink bollworm management in
cotton in the southwestern United States. U.S. Department
of Agriculture, Agr. Res. Serv., ARS-51, 45 p.

Howard, L. O. 1894. A new cotton insect in Texas.
U.S.D.A. Div. Ent., Insect Life 7: 273.

King, E. G., Jr., G. A. Herzog, and R. B. Head. 1990.
43rd Annual Conference Report on Cotton Insect Research
and Control. Proc. 1990 Beltwide Cotton Prod. Res. Conf.
(In Press).

Leigh, T. F. 1963. Considerations of distribution,
abundance, and control of acarine pests of cotton. Adv.
Acarol. 1: 14-20.

Leigh, T. F., T. A. Kerby, and P. F. Wynholds. 1988.
Cotton square damage by the plant bug, Lygus hesperus
(Hemiptera: Heteroptera: Miridae), and abscission rates. J.
Econ. Entomol. 81: 1328-1337.

Luttrell, R. G., and R. T. Roush. 1987. Strategic ap¬
proaches to avoid or delay development of resistance to
insecticides. Proc. Beltwide Cotton Prod. Res. Conf., pp.
31-33.

O'Brien, P. J., D. D. Hardee, and B. Grafton-Cardwell.
1990. Laboratory screening of Aphis gossypii for insecti¬
cide tolerance, 1989. Insecticide and Acaricide Tests (In
Press).

Ridgway, R. L., and E. P. Lloyd. 1983. Evaluation of
cotton insect management in the United States, pp. 3-28.

In R. L. Ridgway, E. P. Lloyd, and W. H. Cross [eds.].

Cotton Insect Management With Special Reference to the
Boll Weevil. U.S. Department of Agriculture, Agriculture
Handbook 589.

Savinelli, C. S., J. S. Bacheler, and J. R. Bradley, Jr.

1986. Nature and distribution of European com borer
(Lepidoptera: Pyralidae) larvae feeding damage to cotton in
North Carolina. Environ. Entomol. 15: 399-402.

Scott, W. P., J. W. Smith, and G. L. Snodgrass. 1986.
Impact of early season use of selected insecticides on cotton
arthropod populations and yield. J. Econ. Entomol. 79:
797-804.

Smith, R. H. 1989. Beet armyworms on cotton. Alabama
Coop. Ext. Serv. Circ. ANR-538, 4 p.

Sterling, W. L., L. T. Wilson, A. P. Gutierrez, D. R.
Rummel, J. R. Phillips, N. D. Stone, and J. H. Benedict.
1989. Strategies and tactics for managing insects and
mites, pp. 267-325. In R. E. Frisbie, K. El-zik, and L. T.
Wilson [eds.]. Integrated Pest Management Systems and
Cotton Production.

Terry, L.I., and B.B. Barstow. 1988. Susceptibility of
early season cotton floral bud types to thrips
(Thysanoptera: Thripidae) damage. J. Econ. Entomol.
81(6): 1785-91.

Trichilo, P. J., and T. F. Leigh. 1986a. The impact of
cotton plant resistance on spider mites and their natural
enemies. Hilgardia 54: 1-20.

Trichilo, P. J., and T. F. Leigh. 1986b. Predation on
spider mite eggs by the western flower thrips, Frankliniella
occidentalis (Thysanoptera: Thripidae), an opportunist in a
cotton agroecosystem. Environ. Entomol. 15: 821-825.

Young, D. F., Jr. 1969. Cotton Insect Control. Oxmoor
House, Birmingham. 184 p.

22

Table 1. Cotton Insect Losses for 1989 for All States1

Number of Acres

Above Treatment Insecticide Cost of one % Yield No. Bales

Pest

Infested

Thresholds

Applications

Application

Reduction

Lost

Boll weevils

5,865,741

4,783,596

2.6

3.85

2.75

331,466

Lygus bugs

4,072,826

2,346,507

0.3

4.16

2.05

246,667

Bollworm/Budworms

7,478,934

4,683,150

1.8

7.14

1.87

225,115

Spider mites

2,659,204

1,599,112

0.2

11.02

1.11

133,838

Aphids

6,021,428

3,798,381

0.8

5.06

0.55

65,805

Thrips

6,581,397

3,593,701

0.5

4.09

0.36

42,782

Beet armyworms

1,194,911

529,996

0.1

10.47

0.15

18,629

Pink bollworm

416,775

350,505

0.2

8.94

0.14

16,438

Fleahoppers

4,404,730

1,032,670

0.1

3.09

0.11

12,951

Fall armyworms

584,400

170,336

0.0

7.80

0.03

3,596

Western flower thrips

2,198,300

158,857

0.0

7.36

0.03

3,388

Leaf perforators

234,000

109,000

0.0

9.04

0.00

267

Minor pests

1,004,227

371,727

0.1

9.86

0.05

5,741

New pests

319,805

181,705

0.0

8.96

0.03

3,223

'Modified from Head (1990).

23

Management of Cotton Bollworm and
Tobacco Bud worm Populations
through Area-Wide Application of
Nuclear Polyhedrosis Virus on Early-
Season Alternate Hosts

M. R. Be//* 1, D. D. Hardee1, J. L. Hayes2, and E . A.
Stade/bacher1

Abstract

When all of the early-season areas within a 100 square mile
area of the Mississippi Delta were treated with a single
application of the nuclear polyhedrosis virus from the
bollworm Helicoverpa zea, pheromone trap data indicated
the rate of increase of the population of bollworms and
tobacco budworms, Heliothis virescens, was less in the
treated area compared to control areas. The reduction in
population was estimated at 25-38% fewer budworms and
19-31 % fewer bollworms in the treated areas. Bioassays of
plants indicated that only ca. 12% as much virus was
deposited in the target areas compared to previous studies,
and the loss of virus coverage was believed to have resulted
in reduced control effectiveness. In cage testing areas,
where 22 % as much virus was deposited compared to
previous studies, there was a 41 % overall reduction in the
numbers of emerging budworm/bollworm adults. Results
of the study demonstrated the need for application and
formulation studies prior to any future tests.

Introduction

Two Lepidopterous species, the cotton bollworm,
Helicoverpa zea (Boddie), and the tobacco budworm,
Heliothis virescens (F.), are major pests of cotton in the
Mid- and Southeastern U.S. and serious pests of many
food, fiber, and forage crops in the western hemisphere.
Their continuing propensity for developing resistance to
insecticides, and the increasing problems associated with
the use of chemicals for their control, indicate an essential
need for the development of alternative methods of manage¬
ment or control of these two pests. One such method is
through the suppression of the first generation of adults that
are produced in wild, early-season alternate hosts
(Stadelbacher 1981).

'Southern Insect Management Laboratory, USDA, ARS, Stoneville,
Mississippi

1 USDA Forest Service, Pineville, Louisiana

The importance of several introduced and naturalized early-
season host plants of bollworm/budworm in the delta of
Mississippi in the buildup of the F, larval generation which
subsequently invades cotton was discussed by Stadelbacher
(1979, 1981). These investigations showed that a species
of wild geranium, Geranium dissectum L. , was the major
alternate host in the area. Based on a quantitative study of
larval and adult populations (Stadelbacher 1979), it was
estimated that as many as 450,000 bollworm/budworm
larvae and 17,000 adults were produced per hectare of this
host. His studies indicated that moths emerging in the
spring were restricted to and concentrated in early-season
alternate host plants occupying less than 5 % of the total
rural area. The rationale of attacking bollworm/budworm
populations during this first generation, and possibly the
second, was later addressed by Knipling and Stadelbacher
(1983). They envisioned that a 90% reduction in area-wide
emergence in the first generation could be an effective
management tool.

A simple dye method of marking bollworm/budworm
populations emerging from areas of early-season weeds was
used in a field study to demonstrate a reduction in moth
emergence due to treatment with an environmentally safe,
nuclear polyhedrosis virus (NPV) (Bell 1988). In a
subsequent cage study, a single, hand-sprayed application
of the bollworm NPV in water resulted in greater than a
90% reduction in emergence (Bell 1990a). Additional cage
studies demonstrated that aerial applications of virus
suspended in water to early season hosts resulted in similar
reductions in emergence (Bell 1990b). All of these studies
provided the basis for a pilot test conducted in the spring of
1990 to investigate the effect on bollworm/budworm
populations when a nuclear polyhedrosis virus is applied
over a large area during the period that bollworm/budworm
are known to be developing on early-season alternate hosts.
Here we report the results of that test.

Research Plan

The plan consisted of aerially applying a viral insecticide to
all of the early-season alternate host plants of bollworm/
budworm within a single area of 100 square miles (64,000
acres) and comparing the emerging adult populations to
those in a similar untreated area, separated by a 10 mile
buffer area. The 2 areas were chosen based on similarity of
both crop and weed hosts of bollworm/ budworm. The
treatment area was located in parts of Washington and

24

Sunflower Counties in the Mississippi Delta, between
Leland and Indianola, Mississippi. This particular treat¬
ment area was chosen because a 3-year history of budworm/
bollworm populations in the area was already available
(Hayes 1990). The virus treatment (Elcar) the nuclear
polyhedrosis virus from Helicoverpa zea was to be applied
at a time of maximum immature seed production in a
majority of the wild Geranium spp. in the test area. Data
would then be collected to determine spray coverage,
infection rates of larvae on pre-season wild hosts, persis¬
tence of the virus on the plants, the number and species of
moths emerging from alternate hosts within the areas, and
the number and species of the first bollworm/budworm
adult populations invading cotton.

Obtaining data on spray coverage, viral persistence, and
infection rates utilizes relatively standard procedures in
insect pathology. However, the primary evaluation of the
success of the wide area suppression tactic in an open and
heterogeneous agroecosystem, where the target is a highly
mobile pest like Heliothis or Helicoverpa, presents some
serious challenges to standard scientific procedures.
Available funding and practical considerations preclude the
level of replication necessitated by the presumed plot to
plot variability. For large-scale entomological studies,
Schneider (1989) reviewed three approaches. The optimal
approach was a 2-yr study with reversal of treatment and
control sites in alternate years. In this approach both year
and location serve as replicates, or more accurately,
standards for comparison. It is common practice in other
fields, such as water shed studies, to utilize historical data
and reference sites as standards for comparison.

The evaluation design for the HNPV pilot test followed the
design outlined by Schneider (1989). Two 100 square mile
(10 mi X 10 mi) areas of the Delta were identified, one was
designated the control area and the other the treatment area
for yr 1. These designations will be reversed in yr 2.
Historical data were already available for designated
treatment area for year 1 from a long-term movement study
conducted in approximately the same location from 1987-
1989. In this previous study, both male adult and egg
count data were taken and thus could be used to assess the
impact of the HNPV treatment in 1990 (Hayes 1990).

Egg count data could be the most valuable information,
since current economic thresholds may be based on egg
numbers (no. eggs/plant) and the decision to treat is often
based on egg counts (e.g., see Miss. Control Guide, Miss.

Coop. Extn. Ser.). However, collection of these data is
labor-intensive and impractical on a large scale. During the
early season, data analysis is complicated by the rapid
changes in host attractiveness. Field to field variability is
high since planting times vary among growers and is
weather dependent. Until cotton reaches the pinhead square
stage, oviposition is relatively rare. At the same time
velvetleaf, Abutilon theophrastii Medicus, a frequent weed
in early season cotton, will be heavily utilized by oviposit¬
ing females. Later, as cotton becomes more attractive,
cultivation of fields eliminates much of the velvetleaf,
which may result in increased utilization of cotton by
ovipositing females.

Pheromone traps are extremely valuable sampling tools
despite inability to ascribe an economic value to the number
of moths captured on a day to day basis. Extensive studies
of trap performance (Hayes and Coleman 1989, Lopez et
al. 1988), have shown that trap capture data accurately
reflect the relative abundance and fluctuations in the adult
male populations in the immediate area. Also, while all
trapping devices have inherent biases, they are far more
reliable and consistent than human sampling efforts
(Hartstack et al. 1978).

Both long-range and mesoscale movement of Heliothis/
Helicoverpa play significant roles in area-wide control.

For the treatment area, sampling of both adults and eggs
was extended beyond the treatment boundary to assess the
impact of migration. Previous studies (Hayes 1990) have
shown that Heliothis /Helicoverpa can move as far as 12
miles and may typically move 2-5 miles per generation,
depending on environmental conditions. Thus, the expecta¬
tion is that the impact of HNPV application will be most
apparent in the center of the treatment area and will
dissipate at the borders. However, spatial heterogeneity in
the area is high and can obliterate the impact of treatment.
Additionally, any indicated suppression in a small treatment
area is likely to be diminished from one generation to the
next due to migration.

The untreated extension of a treatment plot is more likely to
be similar in host and pest composition to the treated plot
than the control area, but it is not independent of the treated
plot because they share a border and are directly influenced
by migration (emigration and immigration).

25

The impact of treatment will be apparent in deviations in
trap capture patterns between years (in the treatment area)
and within-year between treated and untreated areas, both
the control and untreated area beyond the actual treatment
plot. It is unlikely that any two, 100 square mile areas are
similar in crop or pest patterns; however, the capture
patterns should be more consistent with the treated area
when compared to the previous year patterns. Also, in year
2, the control plot will become the treatment plot, and the
1990 data provide the standard for comparison in 1991.

To determine the level of suppression where spatial
heterogeneity is high, the overall rate of increase in the area
from parental to F. generation is a critical measure. An
estimate of the actual rate of increase can be obtained from
trap capture data.

Public Information Strategy

Since no large, populated, and non-crop area was known to
have been treated with a microbial pesticide in the past, the
reaction of the public to the project was of utmost concern.
We believed, however, that an informed public would be an
approving public. The strategy for the public information
campaign was developed jointly by USD A scientists
involved with the project, the USDA, ARS Information
Office, Delta Council, and personnel of Sandoz Crop
Protection, Inc. This process began with the preparation of
an information brochure and a slide presentation prior to
the first meeting with those not involved with the project.
The first meeting was held ca. 8 months prior to the
beginning of the test. The purpose of this meeting was to
inform those persons identified as the most likely to be
approached for comment by the general public or the media
of the details of the test, as well as the safety of the project.
The meeting included a cross section of persons, such as
Mississippi State regulatory officers, large-scale farmers,
congressional representatives, scientists working in similar
areas of study, environmental and health regulatory
personnel, representatives of action groups, extension and
university personnel, and heads of research departments.
This core group would be used if any pre-test media
publicity occurred in which comments from other informed
persons were needed.

Approximately 5 months before the test, a meeting was
arranged with local physicians to acquaint them with
nuclear polyhedrosis viruses and answer their safety

questions. After that meeting, all of the physicians in
attendance agreed to allow their names to be used as
contacts for any questions from media sources and from
their patients. About the same time, meetings were
arranged with local city and county officials to present the
proposed test to them. These information contacts were
made by a well-known local city councilman in the area
after he became familiar with the project.

A date approximately 3 months prior to the beginning of
the test was chosen as the date to begin informing the
general public of the pending test. A national press release
was distributed to the media from 1 to 3 months prior to the
test, and contacts were made with the local media to supply
them additional information regarding the test if they
desired. Many presentations were given during that same
period to local civic clubs, garden clubs, environmental
interest groups, and private interest groups. The virus to
be sprayed, its safety, and how it was to be used was
discussed at all presentations. Special meetings were also
held with specific groups of invited participants working in
local agriculture, such as consultants, aerial applicators,
and extension personnel.

The last month before the test was devoted specifically to
contacting those persons living within the test area. To
accomplish this goal, each residence in the test area was
visited, given an information brochure, and the project was
discussed when questions were asked. The brochures also
contained invitations to meetings specifically for the
residents and general public to discuss the test or ask
questions. Two meetings were held, but none of the
residents attended. A special effort was made to contact
and visit the five catfish producers in the test area. After
learning more about the project, and specific safety testing
of the virus on possible effects on catfish, all of the
producers gave their permission for viral treatment in and
around their pond areas. No complaints or negative
publicity was detected or received throughout the entire
test.

26

Methods and Materials

Virus

The NPV from H. zea (Elcar) was provided by Sandoz
Crop Protection, Inc. The virus was standardized prior to
delivery and contained 4 X 1 09 polyhedral inclusion bodies
(PIB)/g. The Elcar was supplied in 20 lb lots. Although
the virus was bioassayed during the standardization proce¬
dures by Sandoz personnel, additional bioassays were
conducted at this laboratory and by other cooperators to
verify its activity. Our bioassays were conducted using
methods described by Dulmage et al. (1976) as modified by
Bell and Romine (1986).

Application

The virus was applied both by aircraft and by trucks
equipped with mist blowers. Both methods of application
were calibrated to deliver virus at a rate of 40 larval
equivalents (40 L.E.) or 2.4 X 10" PIB in 2 gallons of
water per acre. Two aircraft (Cessna Ag Trucks) and pilots
were supplied by USDA, APHIS Aircraft and Equipment
Operations, Edinburg, Texas for application services. The
aircraft were calibrated at their home base to deliver 2 GPA
and 50-60 droplets per square centimeter with a VMD of
300 micron. This application rate was shown in a previous
study to provide good spray coverage of the wild geranium
(Bell 1990b). The 100 square mile area was divided into
20 plots containing ca. 5 square miles each in order to have
a workable size area for treatment. Aerial photos (USDA,
ASCS, Aerial Photography Field Office, Salt Lake City,
Utah) were used to prepare booklets with 1 plot/page for
the pilots use during the application. Additional maps were
prepared and supplied to all aerial applicators working in
the treatment area, who were then notified which areas the
aircraft would be treating as the application proceeded. A
treatment operations headquarters was set up at the munici¬
pal airport at Indianola, Mississippi, located near the
northeast comer of the test area. This facility was chosen
due to the availability of water for mixing, aviation fuel,
and a hard-surface runway.

The virus for application was mixed with thorough agita¬
tion in 1000 gallon tanks, preparing 600 gallons each time
to allow space for mixing. The tanks were equipped with
gasoline-driven pumps and normal aircraft fittings for

sprayer loading in minimum time (200 gallons per 3
minutes). Since the plan was to treat only the possible
alternate host habitats, the pilots attempted to treat only
along field edges, borders, fallow fields, treelines, etc.
Water-sensitive cards were placed randomly among the
weeds to determine sample spray coverages in some test
areas. Aerial application began on 24 April, a time
believed to be ca. 7 days prior to the maximum overall seed
production in the G. dissectum, the primary alternate early
season host of bollworm/budworm (Stadelbacher 1979).

All host areas treated by aircraft within the test area, except
the fallow fields, were treated by 8 May, although much of
the application was done during weather conditions,
primarily too much wind, not favorable to spot spraying
sites. Due to a series of early spring weather fronts,
farmers in the area were late in tilling many of their fields.
This resulted in an unusual amount of weed areas where it
was questionable as to whether or not they should be
treated in our test. During the period of 8-1 1 May,
applications were made to fields that were not treated
during the previous application, with the pilots determining
which fields would be treated based on weed populations
and the likelihood that the particular area would be plowed
in the near future. As shown in the spray data, the aircraft
applied a total of 1,653 lbs of Elcar in 25,020 gal of water.
This volume would treat ca. 12,510 ac, or 19.5% of the
treatment area.

Application by the truck-mounted mist blowers was done
between 30 April and 4 May. These blowers were cali¬
brated to deliver 2 gallons per acre from the side of the
truck driven at 5 MPH with a 60-ft swath width. This
application procedure was used to treat along the major
roadways within the test area, a distance of 335 miles total.
Although these areas are known to be major sites for
geranium growth (Stadelbacher 1982), they are also
generally more difficult to treat by aerial application, either
because of vehicle traffic or nearby power lines. The
trucks applied a total of 145 lb of Elcar along the road¬
ways, enough to treat a total of 1,097 acre.

27

Evaluation Techniques
Field bioassays

Bioassays were made of field collected material to estimate
the efficiency of the spray application and the persistence of
the virus on the wild geranium. To estimate the quantity of
virus on the alternate hosts, plant terminals were cut at
random on various days from G. dissectum in 8 plots in the
treatment area as soon as possible after the aircraft com¬
pleted the treatment of the plot. The terminals were then
fed to 6-day-old tobacco budworm larvae (avg. wt.: 21.6
mg) for a period of 48 hr (72 terminals from each location),
after which the larvae were then placed on artificial diet.
Larval mortality was recorded 10 days after the beginning
of the feeding period. Similar samples were taken at the 3
locations used for the cage study below, plus the sampling
procedure was repeated 1, 2, 3, 5, 7, and 9 days after
application to estimate the persistence of the virus on the
plants. A final bioassay was conducted on 29 May to
determine the residual virus on the weeds. Random
samples of remaining wild geranium were collected and fed
to larvae as before, with 144 larvae fed material from
treated plots and 144 fed material from untreated areas.

Cage study

The methodology used for the cage tests was similar to that
used in a previous study (Bell 1990b) with some modifica¬
tion. Plots of early-season weeds, primarily wild gera¬
nium, were chosen within the treatment area at 3 locations
and measured for cages in late March. Plot edges were
mowed to facilitate cage installation. Plots at each location
were arranged in RCB design with 3 replicates of 2 treat¬
ments each plus an untreated control. One treatment
consisted of NPV applied aerially only on the date that the
test area was treated. The second treatment at each location
received the aerial application, plus a ground application of
the virus at similar rates, to test the possibility of a timing
effect on control results. Aerial applications were made in
plot numbers 5, 10, and IS (the areas in which the cage
tests were done) on 26 April, 30 April, and 4 May,
respectively. The untreated control plots were covered with
plastic sheeting during application. Cages made of plastic
insect screen (4m X 8m X 2m) were placed over each plot
and secured on 8 May. Periodic external inspections were
made to assure that the cages remained secure. Visual
examinations for the beginning of moth emergence were
also made from outside the cages. Searches were conducted
daily within the cages after the first budworm/bollworm

moth was observed, and moths were captured and removed.
Numbers, species, and sex of the moths were recorded.

The data were analyzed by the procedures of the Statistical
Analysis System (SAS Institute 1985) using ANOVA, and
means were examined using the method of Least Significant
Difference.

Adult Heliothis/Helicoverpa sampling
Pheromone traps, standard 75-50 hardware cloth cone traps
(Hartstack et al. 1979), were used to monitor the relative
abundance and fluctuations in the adult male Heliothis/
Helicoverpa populations in the treatment and control areas.
In both areas, four trap locations (1 trap/sp./loc.) were
identified near cotton fields within each 1 mile interval/
radius from the center; at radii from 1-5 miles in the
control plot, and from 1-10 miles in the treatment area, to
assess the impact of migration on the treated area (encom¬
passed in radii 1-5). The traps were baited with either H.
zea lure (Hereon Laboratories, Inc. , Emigsville, Pennsylva¬
nia) or H. vires cens lure (Scentry, Inc., Phoenix, Arizona).
Depending on availability of an accessible cotton field,
attention was given to spacing trap locations in different
quadrants. A trap location consisted of 1 trap/sp. separated
by ca. 100 meters, placed at the edge of a cotton field along
an accessible roadway. Because fields are frequently
cultivated and roadsides are often mowed or burned, traps
were placed near power poles and in other protected sites.

Traps were routinely monitored from 1 April to 1 August,
which encompassed and/or included the flight of the
parental, F, and F2 generations. During that time, traps
were emptied three times per week (MWF), and the
captured moths counted and recorded; lures were replaced
biweekly. Traps were repaired as needed, and the area
under each trap was kept clean of vegetation.

Heliothis/Helicoverpa egg sampling
In both control and treatment areas, one to three cotton
fields in the neighborhood of each trap location were
searched routinely for eggs (and occasionally larvae) during
the F, generation. Multiple sites were sampled to amelio¬
rate the site-to-site variability in crop phenology. The
developmental stages of cotton ranged from cotyledon to
pinhead square at the onset of sampling to pinhead square
to flowering at the end of sampling. Velvetleaf was
initially prevalent in many fields and was sampled along
with cotton. Some com and velvetleaf-only sites were
sampled in order to assess more accurately the overall rate
of oviposition in each area.

28

Each field was searched, by terminal inspection, for 30
minutes at 2-3 day intervals. All eggs encountered were
collected in 30 ml plastic cups, kept separate by location,
and returned to the laboratory for processing. Eggs were
counted and all (if <20) or a portion (not less than 20%)
were set up on diet to rear for species identification.

The control area was arranged similarly except that the
traps did not extend outside the 100 square mile area. Each
test area contained 24 trap locations, and there were 17 trap
locations in the extended area surrounding the treated area
for a total of 130 traps. The traps were baited on 14 March
(Julian date 75) and trapping continued throughout the
entire test. All traps were checked 3 times per week and
the numbers of moths recorded. The lures were replaced
every other week.

In addition to the estimation of the emerging population by
adult trapping, attempts were made to estimate the number
of adults by counting the egg deposition on early hosts and
cotton within the test areas. This method required visually
checking plants for a total of 30 minutes in each sample
area and recording the numbers and relative population of
species of budworm/bollworm eggs located in a fixed time.

Personnel Requirements

The number of personnel needed for this test was quite
extensive. During the application period, 3 ground crew
persons plus 2 pilots worked at the airport, while an
additional 2 persons marked the comers of the plots on the
ground when needed. Two persons were needed for
conducting the bioassays during the entire test period. The
ground application with the mist blowers required a
supervisor and 3 additional persons during that week only.
Setting up, baiting, and checking the pheromone traps
required 4 persons plus a technical supervisor. Checking
the plants for presence of eggs during the emergence period
of the moths that developed as larvae on the early season
hosts required 7 persons. The cage studies required 2
technical supervisors and 2 additional persons. To conduct
the test, a total of 21 temporary persons was employed at
various times; in addition, 3 supervisors and 4 technical
supervisors were also needed.

Results and Discussion

Field Bioassays

When G. dissectum terminals were randomly collected from
8 different plots immediately after aerial application, and
then fed to 6-day-old tobacco budworm larvae for 48 hours,
mortalities at 10 days averaged 75.1% and ranged from
67.7 % to 88.7 % compared to a background mortality of
9.3 % when larvae were fed terminals from untreated areas.
Although the mortalities of larvae fed treated terminals
were significantly greater than when larvae were fed
untreated terminals, the 75.1% mortality was also signifi¬
cantly less than that reported in a previous study (Bell
1990b). In that study the treated area was in an open field,
and the application was made in broadcast fashion and not
to a strip target area. When those terminals were tested,
larval mortalities averaged 95.7%. The random samples
tested here were taken from along tree lines, field borders,
etc. and not from open fields. The difference in the
concentration of virus required to produce 75 % mortality in
a laboratory bioassay (ca. 760 PIB/ml) compared to the
dose that results in a 96% mortality (ca. 6130 PIB/ml)
indicates a reduction of ca. 88% in virus deposition on the
target hosts in this study compared to the earlier application
test. This lack of virus on the target hosts was also
indicated by the numbers of droplets found on water-
sensitive cards. A similar application rate and droplet size
in the earlier test resulted in an average of 54.4 droplets per
cm2, whereas in the present test, there was an average of
3.9 droplets per cm2 on 20 cards placed at random in strips
of weeds within the test areas, indicating a 93% reduction
in droplets hitting the plants.

In areas where the cage tests were conducted, water-
sensitive cards contained an average of 28.2 droplets per
cm2 on 30 cards placed within the treatment area (10 cards
per area). The virus was applied in these areas as a
broadcast spray, sometimes from a higher altitude due to
the presence of nearby trees. Although the number of
droplets on the cards indicated increased coverage com¬
pared to the random samples, it was still only ca. half the
54.4 droplets deposited in the previous study. When 6-day-
old tobacco budworm larvae were allowed to feed on the
terminals from the cage test areas for 48 hours, mortality at
10 days averaged 81.0%, only a 6% increase compared to
the random samples. The difference in the viral concentra¬
tion required to produce 81 % mortality in a laboratory

29

bioassay (ca. 1305 PIB/ml) compared to that resulting in a
96% mortality (ca. 6130 PIB/ml) indicates a reduction of
ca. 78% in virus deposition on the target hosts in this test
compared to the previous application study. Although the
numbers of droplets were reduced by only ca. half, the
bioassay data here indicate that the amount of virus depos¬
ited on the plants was reduced by almost 80 % . Mortality
due to viral infection at various times after application and
the indicated persistence of the virus on the plants are
shown in Fig. 1. When the larvae were fed on terminals
collected at 1, 2, 3, 5, 7, and 9 days after treatment, viral
mortality averaged 77, 69, 64, 57, 41, and 38 percent,
respectively. The original activity remaining (OAR -
obtained by mortality each day being compared to the
mortality on day 0) on those dates after application was 95,
85, 79, 70, 50, and 47 percent, respectively. This persis¬
tence compares very favorably to the data reported in the
previous aerial application test (Bell 1990b).

When wild geranium collected from the treated area on 29
May, 18 days after the end of all spraying, was fed to
larvae as before, only 5 % of the larvae died from NPV.
Samples from outside of the test area produced no viral
mortality. Very little of the wild geranium remained on
that date.

Cage Study

The emergence and capture of H. virescens and H. zea
adults in the cages are illustrated in Fig. 2. There were no
differences between treatments due to the date of applica¬
tion. Since the Date*Treatment error component was not
significant, a simplified ANOVA was performed across the
9 replicates for RCB design. The analysis indicated that
both of the viral treatments significantly reduced moth
emergence compared to the control (Table 1); however
there was no difference between the 2 virus treatments.
There was a 41 % total overall average reduction in emer¬
gence in the cages treated by aerial application alone, and a
50% average reduction in cages treated by air plus a
treatment using hand sprayers on 10 May. These results
indicate much less effect than in the previous tests where
the least overall reductions have been ca. 88% (Bell
1990b). In the previous study, however, the moth emer¬
gence was reduced only by 64 % when the virus dosage was
halved (20 L.E./ac). The bioassay data in the cages
indicate that the aerial application resulted in ca. 22 % as
much virus on the plants in this study as in the previous

study. If these are factual indications, then the 41 %
reduction in moth emergence in this study would be
reasonable. These data all indicate that deposition of the
virus to the target area was a major problem in this test,
even in the cage test areas where application was somewhat
easier than in other host plant habitats.

The average number of moths emerging in the untreated
cages in this test (9.6/cage) was less than the average
captured in the earlier study (25.3). This was expected
since the cage areas used for the earlier tests were treated
with insecticide to reduce predator and parasite populations
(Bell 1990b). In the present study, the total number of
moths captured in individual untreated cages ranged from 5
to 17, and the number in cages that were aerially treated
ranged from 2 to 17.

Adult He/iothis/Helicoverpa sampling

Visual inspection of the trap capture patterns revealed
deviations from expected in the treatment area (1990
treated) compared to the previous yr (1989 treated), the
control area, and 1990 untreated (Figs. 3 and 4). These
graphs show the mean counts per day, by species, within
the different areas. The vertical dotted lines define the
generations: N1 = the first generation to emerge in the
spring which lays its eggs on early-season wild hosts, often
referred to as the P generation; and N2 = the first genera¬
tion to move into cotton, often referred to as the F, genera¬
tion. The egg data along with the trap data were used in
determining the positioning of these dates. Egg data can be
more reliable than trap data because egg deposition usually
drops to zero between generations while trap captures
"bounce", and rarely return to zero after the first (P)
generation.

Under normal conditions (see Fig. 3, 1989) trap captures
for both spp. increase with each successive generation until
mid-season. From the parental to F, generations peak mean
captures/day may increase by 2-fold. In the treatment area
(Fig. 4) H. zea showed an initial increase during the F,
generation; however, H. virescens failed to reach a mean
capture rate/day during the F, that exceeded the parent (P)
generation peak. At the same time, in the adjacent
untreated area, both H. zea and H. virescens patterns appear
normal, i.e., peak mean captures/day during the F, genera¬
tion exceeded the P generation peaks. In the control area,
the mean captures/day peak exceeded the P generation

30

peak, and continued to climb after the treated and untreated
plot values showed a decline. Trap captures of H. zea in
the treated plot dropped off dramatically on 6 June, while
the rate of decline in the untreated plot was more gradual .
Trap captures of H. vires cens in the control area were
relatively low in both generations (compared with the
treated and untreated plots) and reached nearly equal peak
values in the P and F, generations.

Data collected in the F2 generation suggested that activity in
the control plot is normal, i.e., mean captures/day will
exceed those in the previous generation (F(). However,
activity in the treated and untreated plots remained rela¬
tively low. Further trap capture data are needed, however,
to determine whether the apparent impact on both spp., as
measured by deviations in trap capture patterns, will be felt
in the subsequent generation(s).

Correlations analysis of 1990 trap capture patterns revealed
no significant differences, although the correlation coeffi¬
cients differ in anticipated directions (e.g., for the P
generation, r ranges from 0.96 to 0.99 for treated,
untreated, and control; for the F generation r ranges from
0.89 to 0.96). It is unlikely that correlations analysis
would expose the impact of a systematic suppression tactic
such as use of HNPV, since the population should show a
characteristic cycle of emergence and activity but at much
reduced densities.

Calculations of rate of increase (r,) between the generations
confirmed the visual assessments of trap capture data and
were perhaps the most indicative data collected. These
values are calculated by taking the natural log of the total
number of moths captured in generation 2 divided by the
total number captured in generation 1 . The rates of
increase (r values) between the first two generations are
given for each species (Figs. 3 and 4). In general, the rate
of increase in all plots in 1990 were lower than those
measured for 1989, and the rate of increase for H. zea in
both years was higher than that for H. virescens. In 1989,
H. virescens increased by 38%, and H. zea increased by
95 % . These increases are typical from generation to
generation over the season through at least the F3 genera¬
tion.

The rates of increase of both species in the treated plot was
lower in comparison than the increases in the untreated and
control (Fig. 4). Both species realized lower rates of

increase in the treated area compared to the expected rate of
increase based on 1989 results and compared to either the
1990 control or untreated extension of the treated area.

The rate of increase for H. virescens in the treated area was
only 13 % compared to 38 % for the control and 5 1 % for the
untreated area. The rate of increase for H. zea in the
treated area was only 36 % compared to 55 % for the control
area and 67 % for the untreated area. The treated area
produced 38% fewer H. virescens and 31 % fewer H. zea
overall than the untreated area, and 25 % fewer H. virescens
and 19% fewer H. zea than the control area.

Neither linear nor non-linear regression analysis revealed
significant relationships between mean captures of either
species and distance from the center of the treated area.

This result is not surprising given the high degree of spatial
heterogeneity on the area, and also the high degree of
temporal heterogeneity imposed by the elongated applica¬
tions period (21 days). Analyses that account for this
heterogeneity are planned.

Conclusions

The results of this test indicate that the single virus applica¬
tion reduced the adult Heliothis virescens population
emerging from early season alternate hosts by ca. 25-38%,
and the Helicoverpa zea population by ca. 19-31 % in the
100 square mile area. In cage test areas where there was
somewhat increased virus coverage compared to the
remainder of the test area, the overall bollworm/budworm
adult emergence was reduced by 41 %. Although the
treatment failed to reduce the adult population as much as
was expected, these results were still encouraging. The
indicated reduction in the number of moths was understand¬
able when all factors of the study were taken into consider¬
ation. The primary factor was the lack of spray coverage,
and a secondary factor was the timing of the application.
Although aerial application studies had been done in which
the virus was applied to alternate host areas, this test was
different in that much of the application had to be done
under windy weather conditions. Bioassays were very
informative in indicating the presence of the virus on the
plants, and based on that data, the reductions in the
populations was comparable with prior tests. Although
timing was less a factor than spray coverage, an application
made over a 21 day period would be considered too long
for most seasons when the growth of the alternate hosts is

31

considered. The length of time needed to apply the virus in
this test could easily be corrected by increasing the number
of spray aircraft.

In general, we feel that the test reported here accomplished
some valuable goals, e.g. the education of and acceptance
by a large segment of the general public regarding the use
and safety of microbial insecticides in insect pest manage¬
ment systems. We further believe that the factors which
probably reduced the effectiveness of the virus application
in this test can be overcome, and that, following further
testing, this method of pest management will be proven
effective in the future.

Acknowledgment

Many persons played instrumental parts in this pilot test,
without whose aid the successful completion of this test
would have probably been impossible. Many personnel of
the Mississippi Agricultural and Forestry Station aided in
the public information campaign as well as provided
support. Sandoz Crop Protection personnel aided in several
ways prior to and during the test. We wish to especially
acknowledge and thank those named below as well as
countless others not named here who gave their time and
support during the test.

D.L. Boykin, USDA, ARS
J.R. Brazzel, USDA, APHIS

D. W. Hubbard, USDA, ARS
C.M. Ignoffo, USDA, ARS
R.G. Jones, USDA, APHIS
M.T. Misner, USDA, ARS

E. M. Morgan, Delta Council

R. M. Parry, USDA, ARS
B.D. Radsick, USDA, APHIS

J.C. Schneider, Mississippi State Univ.

B.F. Tanner, USDA, APHIS

S. A. Thomas, USDA, ARS

References Cited

Bell, M. R. 1988. Oil-soluble dyes for marking and
estimating the effectiveness of microbial control of
Heliothis spp. developing on early-season wild geraniums.
Proc. Beltwide Cotton Prod. Conf. pp. 254-257.

Bell, M. R. 1990a. Effectiveness of microbial control of
Heliothis spp. developing on early-season wild geraniums:
field and field cage tests. J. Econ. Entomol, in press.

Bell, M. R. 1990b. Use of Baculovirus for the Control of
Heliothis spp. in Area-Wide Pest Management Programs.
Proc. Vth International Colloquium of Invertebrate Pathol¬
ogy and Microbial Control, August 20-24, Adelaide, South
Australia.

Bell, M. R., and C. L. Romine. 1986. Heliothis virescens
and H. zea (Lepidoptera: Noctuidae): Dosage effects of
feeding mixtures of Bacillus thuringiensis and a nuclear
polyhedrosis virus on mortality and growth. J. Environ.
Entomol. 15: 1161-1165.

Dulmage, H. R., A. J. Martinez, and T. Pena. 1976
Bioassay of Bacillus thuringiensis (Berliner) a-toxin using
the tobacco budworm. U.S. Dept. Agric. Tech. Bull.

1528.

Hayes, J. L. 1989. Relating Heliothis spp. pheromone
trap captures to egg counts in cotton: Third year data from
the Delta of Mississippi. In 1990 Proc. Beltwide Cotton
Prod. Research Conf., in press.

Hayes, J. L., and R. J. Coleman. 1990. Relating
Heliothis spp. pheromone trap captures to egg counts in
cotton: data from the Delta of Mississippi. In 1988 Proc.
Beltwide Cotton Prod. Research Conf. pp. 265-268.

Knipling, E. F., and E. A. Stadelbacher. 1983. The
rationale for area-wide management of Heliothis popula¬
tions (Lepidoptera: Noctuidae). Bull. Ent. Soc. Amer. 29:
29-37.

Lopez, J. D., J. L. Goodenough, T. N. Shaver, and others.
1988. Potential use of sex pheromone traps for monitoring
Heliothis zea and virescens. In Proc. Beltwide Cotton Res.
Conf. New Orleans, Louisiana, paper no. G4.

Schneider, J. C. 1989. Role of movement in evaluation of
area-wide insect pest management tactics. Environ.
Entomol. 18(5): 868-874.

32

Stadelbacher, E. A. 1979. Geranium dissectum: an
unreported host of the tobacco budworm and bollworm and
its role in their seasonal and long term population dynamics

in the delta of Mississippi. Environ. Entomol. 8: 1152-
1156.

Stadelbacher, E. A. 1981. Role of early-season wild and
naturalized host plants in the buildup of the F, generation
of Heliothis zea and H. vires cens in the Delta of Missis¬
sippi. Environ. Entomol. 10: 766-770.

Stadelbacher, E. A. 1982. An overview and simulation of
tactics for management of Heliothis spp. on early-season
wild host plants. In Beltwide Cotton Prod. Res. Conf., Las
Vegas, NV, pp. 209-221.

SAS Institute. 1985. SAS users' guide: statistics, vers. 5.
ed. SAS Institute Inc. Cary, N.C.

Table 1. Number of moths emerging per cage after treatment with nuclear polyhedrosis virus (NPV)
(average of 9 cages per treatment)

Treatment

H. vires can s

H. zea

Totals

Untreated

8.11 a1

1.44 a

9.56 a

Aerial Application

5.00 b

0.67 ab

5.67 b

Aerial and Ground

4.70 b

0.11 b

4.78 b

Application

LSD

3.04

1.08

3.32

1 Numbers in columns not followed by the same letter are significantly different at the 5 % level

33

Viral Activity in Cage Testing Area

PERCENT

NO DAYS AFTER APPLICATION

MORTALITY

ACTIVITY REMAINING

frg. 1. Amount of active virus remaining at indicated number of days after application based on mortality of
6-day-old larvae 10 days after feeding for 48 hours on treated plant terminals (avg. 3 reps). Activity
remaining is comparison of viral activity to activity on day 0.

34

Heliothis Emergence

Untreated - H.v.
Aerial NPV - H.z

Untreated - H.z

111 Aerial-*- Gnd. - H.v.

I .... 1 Aerial NPV - H.v.

1 1 Aerial-*- Gnd. - H.z

Adults per 3 cages

35 Y

4/26 4/30 5/4

Treatment Dates

Fig. 2. Number of adult moths emerging in cage areas treated with 40 Larval Equivalents of NPV/acre on 3
different dates.

35

MEAN COUNT/DAY MEAN COUNT/DAY

HNPV TREATMENT AREA 1989

H. virescens

H. zea

Fig. 3. Captures and rates of increase (r^ of budworm and bollworm adults in 1989 in area used for
treatment in 1 990.

36

HNPV TREATMENT AREAS 1990

H. virescens H. zea

Fig. 4. Captures and rates of increase (r,) of bud worm and boll worm adults in treated area compared to the
control area and the untreated area immediately surrounding the treated area (untreated) in 1 990.

37

Production and Application of
Trichogramma in the former Union of
Soviet Socialist Republics

Sh. M. Greenberg1

In the former U.S.S.R., the scale of biocontrol applications
has increased dramatically from 200,000 to 39,000,000 ha
in the past 30 years. The areas are expected to reach 48.5
million ha in 2000 and 55 million ha in 2005 (Bondarenko,
Voronin and Greenberg 1979; Novoxhilov 1989).

Trichogramma is the main means of biocontrol of the
lepidopterous complex on cereal, leguminous, technical and
vegetable crops and perennial grasses. Large scale investi¬
gations and production evaluation of Trichogramma
effectiveness in orchards and vineyards have been carried
out. In the recent 5 years, it has been annually applied on
60-65% of the areas (in 1988 - on 16.5 million ha) pro¬
tected by active biocontrol (Greenberg and Nikonov 1988).
In the near future, i.e. in 2000, the former U.S.S.R. plans
to increase Trichogramma application volume 2.5 times. It
is promoted by such factors as the simplicity of laboratory
rearing, ability to accumulate quickly at rearing and
release, the use of the parasite to eliminate pests on a
nonharmful stage of their development, and high biological
and economic effectiveness of the entomophage in a number
of the former U.S.S.R. regions. It has been established by
the All-Union Institute of Biological Methods in Plant
Protection (AUIBMPP) and All-Union Institute of Plant
Protection (AUIPP) that scientifically grounded and timely
application of Trichogramma in the regions favorable for its
activity allows elimination of up to 60-80% of pest eggs
and increases winter wheat yield 1. 7-2.0 centners/ha, com
1.8-2. 3 centners/ha, sugar beet 20-35 centners/ha; and
cabbage 20-30 centners/ha (Fadeev and Greenberg 1985).
Trichogramma production biofactories have been built and
are functioning successfully at present. Research and
production experience has been accumulated.

Trichogramma is a good test object and research results can
be used to reveal regularities for production and application
of other entomophages.

The experience in Trichogramma research accelerates the
development of a new branch - commercial entomology -
similar to research experience on Drosophila which
promoted developments in genetics.

Home and foreign experience demonstrates promising
abilities for Trichogramma application against lepidopter¬
ous pests. However, wide variability of the data obtained

1 Department of Entomophages, All-Union Institute of Biological Methods
in Plant Protection, Kishinev, former U.S.S.R.

in experiments and during practical usage of Trichogramma
introduces a set of new tasks for science. They are related
to critical reevaluation and improvement of the existing
Trichogramma production and application technologies.
Under modem farming conditions phytosanitary problems
become more complex. Thus, Trichogramma research
aspects can't be considered without taking into account
peculiarities of host-parasite relationships in different
agrobiocenoses. Transition to the commercial
Trichogramma production technology has to interact with
biological and ecological changes both of the entomophage
and its laboratory host. Genetic consequences of insect
mass rearing have not been evaluated properly. It is
necessary to establish the role and to determine the place of
Trichogramma in integrated pest management (IPM). To
solve the above mentioned problems and to get permanent
high effectiveness of Trichogramma, it is necessary to
intensify fundamental and applied studies on the problem
by attracting different discipline specialists (entomologists,
ecologists, physiologists, biochemists, geneticists, math¬
ematicians, engineers). In the former U.S.S.R., 45
researchers from 15 institutions deal with the
Trichogramma research project.

Selection of effective species plays considerable role in the
use of Trichogramma in programs of biocontrol of noxious
organisms. In the past 20 years, a revision of
Trichogramma genus in the former U.S.S.R. fauna has
been completed. Twenty-five Trichogramma species have
been revealed and described, and key tables have been
composed (Sugonyaev and Sorokina 1975, 1976; Sorokina
1986). Species promising for a number of regions have
been determined: Trichogramma evanescens Westw., T.
pintoi, T. voegele, T. semblidis Gir. , T. principium Sug. et
Sor., T. cacoeciae March., T. embryophagum Htg. The
first four species are, as a rule, the inhabitants of field
agrocenoses, the latter two species are encountered in
perennial grasses. T. evanescens mainly parasitizes eggs of
Pyralidae and Noctuidae; T. pintoi, eggs of cutting noctu-
ids. These species are ecologically plastic and develop at
18-30°C, relative air humidity 60-90%. T. semblidis is a
hygrophilic species, promising against Noctuidae, Cossida
nebulosa, C. nobilis, and Pegomyia heosciami. T.
principium is xerophilic and resistant to high temperature.

It prefers noctuid eggs and is effective on cotton under
conditions of Middle Asia. T. cacoeciae and T.
embryophagum parasitize eggs of Tortricidae, Malocosoma
neustria, Orgia antiqua and others. T. cacoeciae is
hygrophilic (70-80 %RH) and prefers moderate tempera-

38

tures (18-23°C), T. embryophagum is xerophilic 40-60%
RH) and develops optimally at high temperatures (26-
28°C). Biochemical criteria of taxonomy have been
developed in the former U.S.S.R. for economically
important Trichogramma species (Yazlovetski et al. 1981).

Trichogramma effectiveness is, to a considerable extent,
determined by its quality. In AUIBMPP it has been
proposed to evaluate quality using the united quantitative
criterion (D), which unites 4 main Trichogramma indices:
surviving ability, sex index, and fecundity and activity for
search of host eggs (Greenberg, Mencher and
Podberezskaya 1979). Trichogramma standards differing
according to the quality class had been developed
(Greenberg, Podberezskaya and Mencher 1980a, 1980b).
Dependence has been revealed [1] between the united
quality criterion and the expected biological effectiveness of
Trichogramma in the field (E) under optimal hydrothermal
environmental condition (Table 1).

A system has been created in which Trichogramma quality
is evaluated by the integral index under a certain environ¬
mental regime. It includes 7 biological parameters of the
entomophage: ability to parasitize, ability to survive,
deformed individuals, sex index, reproductive abilities,
migrational abilities, and searching abilities (Chernyshev et
al. 1986).

The selection of laboratory hosts is an important techno¬
logical problem during Trichogramma mass production.

The Angoumois grain moth ( Sitotroga cerealella Oliv.)
used at present in the former U.S.S.R. as Trichogramma
laboratory host doesn't completely meet nutritional require¬
ments of larval parasitoids both in egg size and their
biochemical composition (Ageeva et al. 1978). Under long
term rearing, the size of the parasite decreases as well as
female searching capacity. Proper selection of the labora¬
tory substrate has great potential to increase Trichogramma
application effectiveness. There are several research
possibilities to solve the problem such as selection of
effective laboratory hosts reared on artificial nutritional
diets and development of artificial nutritional substrates
(Greenberg and Yazlovetski 1983). Data are scarce in
literature from the former Soviet Union and foreign
literature on comparative quality evaluation of
Trichogramma reared on different hosts. They are also
difficult to compare since they are expressed by many
indices. It has been established that the united quality
criterion of Trichogramma bred on cabbage moth

(Mamestra brassicae L.) eggs constituted 0.96, on the flour
moth ( Ephestia kuehniella Zeller) - 0.53, and on the
Angoumois grain moth ( Sitotroga cerealella Oliv.) - 0.43.
Trichogramma activity in searching for main host eggs has
been 2 times higher after rearing it on Mamestra brassicae
than on unusual hosts (Table 2).

We consider that the problem can be effectively solved by
replacing the Angoumois grain moth either by natural hosts
or by artificial nutritional substrates. Wide introduction
into practice of the commercial technology for
Trichogramma rearing induces improvement of the
entomophage survivability on the given laboratory host.
There are 750 mechanized production lines functioning in
the former U.S.S.R. which supply Trichogramma for 8-9
million ha (Abashkin et al. 1988). Research and produc¬
tion experience accumulated in our country formulates the
scheme for a year cycle of Trichogramma rearing based on
the collection of natural host eggs colonized by
Trichogramma (at least 1,000-2,000) and the completion of
2 reproductions on natural host eggs and not less then 7 on
the eggs of Sitotroga cerealella before releasing into the
field.

Guaranteed collection of a great quantity of Trichogramma
starting culture in nature can be provided by organizing
nurseries — places of artificial concentration of
Trichogramma under natural conditions. Nurseries are
crop conveyers where phytophage and entomophage
monitoring is done during vegetation period. It consider¬
ably improves creation of a colony founder, one of the
main lines in the scheme of Trichogramma rearing.
Interpopulation crossings are one of the possible ways to
increase the heterogeneity level of nursery material under
mass production. Gene exchange between populations
prevents their degeneracy and promotes quick adaptation
changeability (Rusnak and Greenberg 1982; Greenberg et
al. 1984; Greenberg, Rusnak, Djurich et al. 1986;
Greenberg and Gavrilitsa 1989). The scheme proposed
provides stable production of the second class
Trichogramma (D = 0.6-0. 7) with 60-80% biological
effectiveness under optimal hygrothermal conditions during
parasite release into the field. Republic or regional
biolaboratories in the former U.S.S.R. are responsible for
the production of natural host eggs and rearing of
Trichogramma on them, i.e. for the preparation of the
founding culture. Regional or collective farm

39

biolaboratories similarly existing in the former U.S.S.R.
net of pedigree and seed farming are responsible for
commercial production of Trichogramma.

Species competition gains importance under mass produc¬
tion and field release of Trichogramma. We have revealed
mechanisms of competition between T. evanescens and T.
pintoi under laboratory and field conditions and ways to
regulate it (Greenberg and Gavrilitsa 1983, Greenberg et al.
1984).

Techniques, regimes and equipment have been developed
for long term storage (up to 7-8 months) of diapausing
Trichogramma, its reactivation and for short-term storage
(up to 30 days) of active Trichogramma (Shlyakhtich et al.
1989).

Recommendations have been prepared on long term storage
of Sitotroga cerealella eggs in liquid nitrogen as well as
sterile noctuid eggs (Gennadiev et al. 1985, Degtyarev and
Yanishevskaya 1986). Scientifically grounded technology
for Trichogramma application must take into account
parasitoid behavioral peculiarities. It was established on
the basis of observations and experiments that the labora¬
tory population of Trichogramma has weak searching and
migrational capacities. The parasite is searching for its host
without any system. It finds and attacks the egg only at the
distance of 1.5-10 mm. Trichogramma movements are
occasional and are not associated with host distribution.
Entomophage searching area is inversely proportional while
frequency of occurrence with the host and biological
effectiveness are directly proportional to cluster densities of
pest eggs. Low host density reduces Trichogramma
reproduction tempos. Parasite overcrowding increases
iatraspecific competition and mortality. Relationships of
Trichogramma and Ostrinia nubilalis can serve as an
example. Searching area of female parasitoids is 2.8 times
higher at parasite-host ratio =1:1 than at that of 1 1 : 1
(constituting 1,150 and 415 sq m respectively). The
frequency of occurrence of the parasitoid with the host
(parasite-host ratio = 1:1) at 2.5 egg clusters/100 plants is
9 times less than at 13 100 plants (the quantity of eggs
parasitized constituting 33.1% and 68.4%, respectively).
Trichogramma mortality at 13 egg clusters of Ostrinia
nubilalis! 100 plants is 4 times higher at parasite-host ratio
= 11:1 than at that equal to 1 : 1 . Moderate competition
(parasite-host ratio = 1 1 : 1 or 6: 1) regulates the population
number maintaining it at about the same level (Greenberg
and Boubetrin 1988). Trichogramma searching activity is

determined by endogenic and exogenic factors. The
parasitoid is stimulated to host searching by the presence of
eggs in the female ovariole and by the physiological age of
host eggs. We noticed that under laboratory conditions T.
evanescens parasitized 80-85% of 1-2 day old eggs of
Mamestra brassicae, 40-50% of 3 day old eggs and 10-
15% of 4 day old ones (Greenberg 1983). Abstention from
oviposition results in the decrease of Trichogramma
fecundity and reduction of the rate of their moving on the
substrate (according to our data from 46.8 to 16.9 eggs/
female and from 0.38-0.43 to 0.16-0.20 cm/sec. respec¬
tively). Leaf surface area and distance between plants
greatly influence the rate and degree of host eggs infesta¬
tion. For example, com leaf biomass (for the period of
Trichogramma release) increases from 13.5 to 26.0 thou¬
sands of square meters/ha and cabbage - from 16.9 to 54.2
thousands of square meters/ha. Total plant quantity/ha
varies respectively from 50 to 70 and from 40 to 60
thousands. There was feedback between the variables
mentioned.

Experiments demonstrated that the radius of Trichogramma
effective activity when it moves independently constitutes
about five meters, while when it is passive (with the wind)
these distances can reach in average 10-15 m. Useful
activity of adult Trichogramma from laboratory population
is limited by 3-5 days.

One of the reasons of unstable entomophage effectiveness is
pest distribution in time and space. Pests against which
Trichogramma are used, as a rule, lay eggs evenly in the
field while the parasitoid is mainly released into foci (50-
100 points/ga). Flight and oviposition of Noctuidae,
Pieridae and Tortricidae goes on continuously during 30-35
days of the first generation and up to two months of the
second one. The parasitoid is released three times for this
period. The number of releases depends on entomophage
reproduction in nature. The population biotic potential and
its growth rate directly relate to abiotic environment
conditions and pest density. The role of daughter progenies
of Trichogramma released into the field is manifested by
the end of phytophage oviposition. According to our
observations, the number of parasitized eggs varies from
60% to 100% which constitutes 10-15% of the total
number of eggs killed by Trichogramma for one genera¬
tion. One can correct release rates (Greenberg et al. 1989)
only on the basis of models of dependence of fecundity,
number of females and Trichogramma survival on abiotic
and biotic factors. Thus, the scheme leaves time and space

40

when there's almost no danger for normal development of
pests embryos. According to literary data, entomophage
release rates vary from 20,000 to 600,000 parasitoid/ha and
often don't agree with the density of pest egg clusters and
parasitoid ecology. However, scientifically grounded
Trichogramma release rates can be developed only taking
into account density of phytophage egg clusters, quality of
released Trichogramma , climatic conditions during parasi¬
toid application and plant state. We developed mathemati¬
cal models relating forecasted effectiveness of
Trichogramma application with the above mentioned factors
using as examples Mamestra brassicae on cabbage,

Ostrinia nubilalis on com, and Lobesia botrana (Den. et
Schiff.) on grapes, with graphical and mechanical nomo¬
grams on the base of which one can calculate optimal
parasitoid release rate. Once pest threshold number is
reached, Trichogramma releases begin. Subsequent
releases must be done every 5-6 days during the whole
period of pest egg laying. Table 3 presents release rates of
the 2nd class entomophage (D = 0.5-0. 7) under optimal
hygrothermal conditions during release and average leaf
surface of plants.

Guideline have been confirmed for the distribution of
different aged Trichogramma which provide reliable
effectiveness and a reduction of the number of releases from
5-6 to 2-3 (Greenberg et al. 1983). Traps baited with pest
pheromones are used to establish optimal time and rates for
Trichogramma release. The short term forecast of oviposi-
tional intensities has been developed on the basis of average
quantity of males caught per one trap for 5 days and
biological peculiarities of Ostrinia nubilalis and Mamestra
brassicae (Methodicheskie ukazania 1985). Comparison of
forecasted and factual data on egg quantity shows correla¬
tion (r) at the level of 0.75-0.80. In the former U.S.S.R.,
adult Trichogramma are distributed by hand (100-200
points/ha) and eggs by mechanized devices 12 or 24 hours
before adult emergence. At present some research institu¬
tions of the country developed and introduced technical
means for ground and plane dispersal of Trichogramma
(Recomendatsii po primeneniyu trichogrammi v SSSR
1985).

Effectiveness of Trichogramma application is determined
by the number of eggs parasitized. The higher the index is
the better the effect of the procedure. However, it is not
legitimate in our opinion. For example, Ostrinia nubilalis
density equal to 4 egg clusters/ 100 plants is enough to get
Trichogramma effectiveness in the range of 30 % (taking

into account natural mortality of pest eggs) and bring its
number to the threshold, 55 % at 6 eggs clusters. At the
same time, when Ostrinia nubilalis density constitutes more
than 11 egg clusters/100 plants, even the effectiveness at
the level of 70% doesn't provide density reduction to the
threshold. Other target pests are characterized by a similar
picture. A different approach to the evaluation of the
effectiveness of Trichogramma and other plant protection
means is sought to develop an application strategy that
won't violate existing relationships between pests and
beneficial organisms in agricultural biocenoses. The All-
Union Institute of Biological Methods in Plant Protection -
AUIBMPP) is developing a united criterion of
Trichogramma effectiveness against Ostrinia nubilalis,
Mamestra brassicae and Lobesia botrana. This criterion
establishes biological and economic effectiveness of the
entomophage.

Scientific and production experience shows that the greatest
effectiveness can be achieved when Trichogramma is used
in combination with other protection procedures. This
includes the combination of multiple Trichogramma
releases with rational use of pesticides, microbiological
preparations, and an increased role of natural beneficial
organisms. Great importance ought to be given to the
development of the agrotechnical phone promoting increase
of the effect of both chemical and biological techniques.
Thus, timely and qualitative soil treatment permits 5-6
times reduction of cutting noctuid larvae, 30-40% reduction
of Mamestra brassicae, Plutella maculipennis and Pieris
sp. pupae, 40-50% reduction of Loxostege sticficalis pupae,
and up to 80 % reduction of diapausing pupae of
Laspeyresia nigricana. Low cutting of com (up to 10 cm
from soil surface) followed by burning of plant remnants
reduces Ostrinia nubilalis density 8 times because the main
mass of diapausing pest larvae (up to 77.6%) lives higher
than this level. Crop rotation and space isolation of plants
from the places of pest reservation, sowing time, density of
planting, and use of mineral fertilizers influence pest
number considerably (Greenberg et al. 1987). Table 4
shows the effect of the crop cultivation agrophone on
Trichogramma effectiveness against Ostrinia nubilalis on
com.

Microbial preparations must be components of !PM. It was
established by AUIBMPP, Ukrainian Institute of Plant
Protection and Belorussian Institute of Plant Protection that
microbiopreparations (dendrobacillin, VIRIN-KS, VIRIN-
XS, VIRIN-OS) had no toxic effect on Trichogramma 24

41

hours after application. Table 5 presents biological
effectiveness of mutual and separate use of Trichogramma
and lepidocid against Ostrinia nubilalis. The use of
lepidocid against this pest (first use at larvae emergence and
subsequent ones in 10 days intervals at a concentration of
0.5%) permits 2 times reduction of the population density.
However, the effectiveness of the biopreparation is lower
than that of the parasitoid which is explained by short
period (several hours) of exposed mode of life of Ostrinia
nubilalis larvae. The most effective was the variant where
Trichogramma was combined with lepidocid. The intro¬
duction of Bracon, Chrysopa, Aphidoletes, Aleochara,
Tribliographa and others will widen the assortment of
biocontrol means used in IPM on cotton, vegetables and
cereals.

Sowing of nectariferous grasses in field margins improves
the activity of natural entomophages and increases
Trichogramma effectiveness. The most difficult combina¬
tion is Trichogramma with chemical plant protection means
due to the fact that Trichogramma is characterized by
having the least resistance to pesticides. It was established
by AUIBMPP and AUIPP that the egg and larva stages of
Trichogramma are more resistant to pesticides at concentra¬
tions close to commercial (but not ovicides). Susceptibility
increases at the stage of pupae and especially at the stage of
adults. Acaricides and aphicides pose considerable selectiv¬
ity for Trichogramma. Lethal doses of the majority of
broad spectrum insecticides are much less than those
applied. Captan and cineb are the only fungicides selective
for Trichogramma, others induce 50-100% mortality. It
was established by the All-Union Institute of Plant Protec¬
tion that methylmercaptafos, syfos, tedion, captan, cineb,
and benlate can be referred to as pesticides with low
persistence; rogor, phosolon, metafos, chlorofos, keltan,
and uzgen have moderate ones. The rest have high residual
toxicity for Trichogramma. When Trichogramma is
applied in combination with chemicals, it is advised that
broad spectrum insecticides with high initial and residual
toxicity be excluded giving preference to selective or low
persistent compounds. Chemicals ought to be used at least
3-4 days after parasitoid release. Trichogramma dispersal
immediately after chemical treatments reduces the effective¬
ness of the procedure. It is necessary to take into account
the duration of pesticide toxic effect.

In conclusion I'd like to mention that 5-12 g of Sitotroga
cerealella eggs and 3-5 g of Trichogramma are produced
from 1 kg of grains. The cost of 1 g of Sitotroga cerealella

eggs (50,000 eggs) constitutes 0.49 roubles, and 1 g of
Trichogramma (80,000 eggs) costs 0.85 rouble. Total
expenditures for Trichogramma application (field, veg¬
etable crops, etc.) varies from 2.0 to 2.6 roubles/ha.

References Cited

Abashkin A. S., Sh. M. Greenberg, and V. P. Gorban.
1988. Sostoianie voprosa i perspektivi proizvodstva
trichogrammi na biofabrikah. In Trichogramma v zaschite
rastenii. (In Russian). M., 3-12.

Ageeva, L. I., L. B. Kaplan, V. V. Sumenkova, and I. G.
Yazlovetski. 1978. Aminokisloti, belki i lipidi
tcheshuekrilyh-khoziaev trichogrammi. In Tezisi dokladov
I Vsesojuznogo soveschania po primeneniju trichogrammi v
zaschite rastenii. (In Russian). Kishinev, 51-52.

Bondarenko, N. V., K. E. Voronin, and Sh. M.

Greenberg. 1979. Sostojanie i perspektivi razvitia
biologicheskogo metoda zaschity rastenii v SSSR. (In
Russian). Selskokhozyaistvennaja biologia, XIV, 6: 675-
682.

Chemishev, V. B., V. M. Afonina, and Sh. M. Greenberg.
1986. Usovershenstvovanie metodiki i ustroistva dlya
opredelenia kachestva trichogrammi. (In Russian). In
Tezisi dokladov I Vsesojuznogo soveschania po
promishlennomu razvedeniu nasekomyh, M., 55-56.

Degtyarev, B. G. and L. V. Janishevskaya. 1986.
Povishenie effektivnosti trichogrammi. (In Russian).
Zaschita rastenii, 5:21.

Fadeev, Ju. N. and Sh. M. Greenberg. Sostojanie i
perspektivi issledovanii i prakticheskogo ispolzovania
trichogrammi. (In Russian). Bull.EPS IOBC, 11: 45-56.

Gennadiev, V. G., E. D. Khlistovski, and L. A. Popov.
1985. Metod konservatsii jaits zemovoi moli dlya uslovii
promishlennogo razvedenia trichogrammi. (In Russian).
Selskokhoziaistvennaja biologia, 3: 118-124.

Greenberg, Sh. M., I. N. Boubertin, and B. V. Pynzar.
1983. O rezultatah laboratomogo razvedenia i ispolzovania
raznovozrastnoi trichogrammi. In Massovoe razvedenie
nasekomyh. Kishinev, Shtiintsa, 60-64.

42

Greenberg, Sh. M. and L. F. Gavrilitsa. 1989. Rol
passajei trichogrammi cherez jaitsa ejo estestvennih
khoziaev. (In Russian). In Parasiti i khischniki v zaschite
rastenii. Kishinev, Shtiintsa, 38-45.

Greenberg, Sh. M. and L. F. Medoni. 1983.

Trichogramma na kapustnyh poljah. Zaschita rastenii. (In
Russian). 12:18.

Greenberg, Sh. M., E. M. Mencher, and L. V.
Podberezskaja. 1979. Metodika otsenki trichogrammi. (In
Russian). Kishinev, 15 p.

Greenberg, Sh. M., E. M. Mencher, and L. V.
Podberezskaja. 1980a. Standart na trichogrammu v borbe
s sovkami. (In Russian). In Trichogramma, part 2,
Kishinev, Shtiintsa, 32-35.

Greenberg, Sh. M., L. V. Podberezskaja, and E. M.
Mencher. 1980b. Razrabotka testa poiskovoi sposobnosti
trichogrammi. (In Russian). Selskokhoziaistvennaja
biologia, 3: 132-135.

Greenberg, Sh. M., B. V. Pynzar, and L. P. Zilberg.

1989. Biologicheskoje obosnovanie i optimizatsia
ispolzovania trichogrammi v borbe s kapustnoi sovkoi. (In
Russian). In Biologicheskii metod borbi s vrediteliami
ovoschnikh kultur. M. 177-88.

Metodicheskie ukazania po nadzoru za chislennostju i
signalizatsia srokov borbi so steblevim motilkom s
pomoschiu feromonnyh lovushek. (In Russian). 1981,
Kishinev, 7.

Metodicheskie ukazania po primeneniu feromonnih
lovushek dlya nadzora za kapustnoi sovkoi i opredelenia
neobkhodimih srokov borbi s nei. (In Russian). 1985, M.,

14.

Novozhilov, K. V. 1989. Zaschita rastenii: puti
optimizatsii. (In Russian). Zaschita rastenii, 3: 2-6.

Rekomendatsii po primeneniu trichogrammi v SSSR. 1985.
(In Russian). M., 85.

Rusnak, A. F. and Sh. M. Greenberg. 1982. Razvedenie
matochnoi kulturi trichogrammi. (In Russian). Kishinev,

8.

Sorokina, A. P. 1986. Opredelitelnie tablitsi vidov roda
Trichogramma Westw (Hymenoptera, Trichogrammatidae)
if SSSR. (In Russian). Bull.EPS IOBC, 15: 7-20.

Sugonyaev, E. S. and A. P. Sorokina. 1975. Systematika
roda trichogramma. (In Russian). Zaschita rastenii, 6: 33-
35.

Sugonyaev, E. S. and A. P. Sorokina. 1976. Novie vidi
roda Trichogramma (Hymenoptera, Chalicidoidae) if
Srednei Azii i Altaiskogo kraja. (In Russian).
Zoologicheskii zhumal, V. IV, 5: 777-779.

Yazlovetski, I. G., V. V. Sumenkova, and L. I. Ageeva.
1981. Opredelenie vidovoi prinadlezhnosti trichogrammi
metodom electroforeza v poliakrilamidnom gele. (In
Russian). Dokladi VASKHNIL, 9: 27-29.

43

Table 1 . Biological effectiveness of Trichogramma depending on the quality of the biomaterial used

Intervals of values of united
quality criterion (D)

Classes according to
the quality

Forecasted biological
effectiveness of Trichogramma (E)%

0.7 - 1.0

1

>80

0.5 -0.7

2

60 - 80

0.3 -0.5

3

33-60

0.0 -0.3

nonstandard

<33

E = -0.527 + 1 12.99 D

Table 2. Comparative evaluation of the quality of Trichogramma reared on different hosts

Trichogramma hosts

Indices

Symbols

Mamestra

brassicae

Ephestia

kuehniella

Sitotroga

cereaiella

Emerged individuals (%)

al

97.0

91.2

86.8

Number of females (%)

a2

62.1

57.0

52.2

Fecundity (individuals/female)

n

116.5

74.6

59.4

Statistical quality criterion
(al/100-a2/100-n)

yi

70.2

38.8

26.9

Activity for searching host eggs (%)

y2

55.5

26.9

20.4

United quality criterion

A

0.96

0.63

0.43

44

Table 3. Release rates of Trichogramma against lepidopterous pests

Pest

Crop

Pest threshold

Pest density

Ratio 1 1 :X ( Trichogramma
females:pest eggs)

Cutting

Sugar beet,

0.4-0. 6

1st release - 20,000 females/ha

noctuids

Perennial

individuals/m2

Subsequent releases:

grasses,

0.7 - 4 ind./m2

1:1

Full fallow

4.1 - 10 ind./m2

1:3

10. 1-15 ind./m2

1:5

> 15 ind./m2

1:10

Leaf chewing

Vegetables

4-5 eggs/m2

1st release - 25-30,000 females/ha

noctuids with

Sugar beet

at 1st pest

Subsequent releases:

group egg

generation.

up to 10 ind./m2

1:1

clusters

7-8 eggs/m2

10.1 - 20 ind./m2

1:5

at 2nd pest

20. 1 -50 ind./m2

1:10

generation

>50 ind./m2

1:20

Leaf chewing

Cotton,

up to 20 eggs/ 100 plants

70-75,000 females/ha

noctuids w /

Tomato,

over 20 eggs/ 100 plant

60,000 females/ha

single egg

Com

clusters

Ostrinia

Com

2 ovipositions/

up to 6 egg clusters/ 100 plants

5:1

nubilalis

100 plants

6.1-10 egg clusters/ 100 plants

1:1

over 10 egg clusters/ 100 plants

1:2

Loxostege

Agrocenoses

10-15 eggs/m2

15-20 eggs/m2

55,000 females/ha

sticticalis

with the pest

21-40 eggs/m2

50,000 females/ha

41-50 eggs/m2

45,000 females/ha

Pier is sp.

Vegetable

10 eggs/m2

1st release - 20,000 females/ha

Subsequent releases: — —

1:10

45

Table 4. Survival of Ostrinia nubilalis under low and high agrophone of corn cultivation using
Trichogramma (per 100 m2)

Index

Low agrophone

Mortality of the
preceding stage individuals

number %

Number of live log nos. live
individuals individuals

Mortality

(K)

Diapausing larvae after

overwintering 1984/85

2.0

15.4

13.0

1.10

0.06 (Kl)

Pupated

2.0

18.2

11.0

1.04

0.09 (K2)

Adults emerged

9.0

0.95

Eggs laid

1477.0

83.0

1780.0

3.25

0.77 (K3)

Larvae emerged

63.0

20.8

303.0

2.48

0.10 (K4)

Active larvae

216.0

9.0

240.0

2.38

1.00 (K5)

Larvae diapausing

K = 2.02

in plant remnants

24.0

1.38

F = 2. 17

Index

High agrophone

Mortality of the
preceding stage individuals

number %

Number of live
individuals

log nos. live
individuals

Mortality

(K)

Diapausing larvae after

overwintering 1984/85

2.1

20.8

10.1

1.00

0.10 (Kl)

Pupated

2.0

25.0

8.0

0.90

0.12 (K2)

Adults emerged

6.0

0.78

Eggs laid

1257.0

83.5

1506.0

3.18

0.78 (K3)

Larvae emerged

50.0

20.1

249.0

2.40

0.10 (K4)

Active larvae

192.7

96.8

199.0

2.30

1.50 (K5)

Larvae diapausing

K = 2.60

in plant remnants

6.3

0.80

F = 2.17

Table 5. Effectiveness of means for biocontrol of Ostrinia nubilalis Hbn.

Variants

% damaged
plants

Intensity
of damage

Number larvae/
damaged plant

Population density
larvae/m2

Standard

72.0

54.5

2.0

7.20

Trichogramma

33.0

16.3

0.8

1.32

Lepidocid

53.0

22.3

1.2

3.18

Trichogramma +

25.0

11.4

0.5

0.62

lepidocid

46

Propagation and Release of Natural
Enemies for Control of Cotton Insect
and Mite Pests in the United States,
including Status and Prospects for
using Microplitis croceipes

Edgar G. King1 and Janine E. PowelP

Introduction

The aggregate damage attributed to insect and mite pests in
United States cotton ranged from 7 to 14 percent, annually,
during the last decade (Suguiyama and Osteen 1988).

These losses occurred in spite of the best control efforts. In
1989, arthropod pests reduced cotton yields by 9.2 percent
in the United States, resulting in a loss of 1 . 1 million bales
from potential yield and $319 million in revenue. Addi¬
tionally, the cost for insecticides and miticides to control
the pests was $366 million. The total cost of arthropod
pests to United States cotton production in 1989 was $685
million (King and Herzog 1990).

Biological Control Drivers

Obviously, the cost of yield losses coupled with the cost of
control drastically reduces the profitability of producing
cotton in the United States. Nevertheless, United States
cotton producers have and still are relying heavily on
synthetic pesticides to control arthropod pests (Herzog and
King 1990). These chemicals are relatively fast-acting,
often control a complex of pests, can be used at a grower's
discretion on a field-by-field basis, and generally are cost
effective as compared to alternative control methodologies.

Nevertheless, the availability of effective pesticides is
declining precipitously. On the one hand, many pesticides
have been rendered obsolete because of resistance buildup
beginning with the organochlorines, followed by the
organophosphate and carbamate compounds, and most
recently the pyrethroids. On the other hand, the United
States public has become more and more alarmed about the
safety of their food and water as well as impact of pesticide
residues on animal and plant life threatened by extinction
(King et al. 1988). Consequently, many chemicals have
been banned from usage. Moreover, use of pesticides is
being restricted by The Endangered Species Act of 1973,
which requires the Environmental Protection

'Subtropical Agricultural Research Laboratory, USDA, ARS, Weslaco,
Texas

2 Northern Grain Insects Research Laboratory, USDA, ARS, Brookings,
South Dakota

Agency (EPA) to protect endangered and threatened species
under the Federal Insecticide, Fungicide, and Rodenticide
Act.

These public concerns, coupled with the cost of searching
for new, effective compounds have escalated the cost of
registering new chemicals. The cost of acquiring the first
label for registered use of a single pesticide has been
estimated at $20 to 50 million. Many chemicals registered
prior to November 1984 are not being re-registered under
EPA requirements because of putative chronic health effects
and ground water leaching.

Finally, synthetic pesticides are petroleum based, and this
non-renewable resource is becoming less plentiful and more
expensive. The effects of synthetic pesticides on natural
enemy populations with consequent elevation of secondary
pests to primary pest status, previously innocuous
arthropods to pest status, and pest resurgence are discussed
in subsequent sections of this paper.

There is renewed interest in the United States in the
development of non-chemical methods for controlling
arthropod pests. Certainly, biological control must receive
high priority. Biological control involves managing natural
enemies and their products to reduce pest populations and
their effects. (Natural enemies refers to predators, parasi-
toids, and pathogens, but in this paper the term is restricted
to predators and parasitoids.) Interrelated strategies are
conservation, importation, and augmentation. Conserva¬
tion is preservation of indigenous natural enemies. Classi¬
cal biocontrol involves importation for release and estab¬
lishment where natural enemies are lacking. Augmentation
is propagation and release of natural enemies in selected
areas where natural enemies occur in too few numbers.

Integrated Pest Management

Biological control is an integral component of the cotton
integrated pest management strategy in the United States
(King 1986). Integrated Pest Management (IPM) has been
defined as a system in which all available techniques are
evaluated and consolidated into a unified program for
managing pest populations to avoid economic damage and
minimize adverse side effects on the environment (NAS
1969). For example, with regard to controlling the
bollworm, Helicoverpa ( = Heliothis) zea (Boddie) and the
tobacco budworm, Heliothis virescens (F.), United States

47

cotton-insect control guides often provide a list of the
predators that may be encountered while surveying insect
pest infestations. Parasitoids (parasites in the class Insecta)
are usually mentioned, but not by name. Often these
guides provide illustrations of important natural enemies.
Typically, efforts are made to spare natural enemies by
restricting insecticide applications thereby maximizing pest
suppression action (King 1986).

Avoidance of pesticide usage has often been cited as
precluding the buildup of pest populations such as boll-
worm and tobacco budworm, aphids, whiteflies, and mites
because of pesticide-related mortality to the natural enemies
(Bottrell and Adkisson 1977). A particularly outstanding
book outlining the IPM strategy for managing cotton pest
populations is "Integrated Pest Management for Cotton in
the Western Region of the United States" (Anonymous
1984).

Unfortunately, natural enemy populations often occur in
too few numbers to prevent the buildup of pest arthropod
populations to damaging levels. Certain pests such as the
boll weevil, Anthonomus grandis grandis Boheman, plant
bugs such as Lygus spp., the cotton fleahopper,
Pseudatomoscelis seriatus (Reuter), and the pink bollworm,
Pectinophora gossypiella (Saunders), generally do not have
effective natural enemies in cotton fields. Efforts have
been made to import and establish effective natural enemies
of these pests (boll weevil - Cate 1985; plant bugs -
Coulson 1987; pink bollworm - Clausen 1978 and Legner
1979) as well as the bollworm and tobacco budworm
(Powell 1989), but these efforts have been unsuccessful.

E. F. Knipling (1979, Unpubl. Data) has developed
numerous theoretical models demonstrating the technical
feasibility of controlling pest populations by augmenting
predator or parasitoid populations. Knipling (personal
communication, 4 April 1990) conceptualized his thesis for
augmentation as follows: "Host dependent parasites have
evolved systems allowing reproduction without jeopardiz¬
ing existence of the host. The coexistence patterns, in
numerical terms, operate within rather narrow limits.
However, addition of the parasites in sufficient numbers at
strategic times can result in virtual elimination of the host. "

The technical feasibility of propagating and releasing
parasitoids and predators to control arthropods has been
demonstrated repeatedly. This subject has been reviewed
extensively (DeBach and Hagen 1964, Ridgway and Vinson

1977, King et al. 1984, King et al. 1985a, King et al.
1985c). Van Lenteren (1986) reviewed the augmentative
releases of parasitoids in greenhouses. Greany et al. (1984)
briefly review the potential for using semiochemicals to
manipulate parasitoids and the need for artificial culture
techniques to mass propagate them. In general, the use of
predators and parasitoids to control cotton pests was
included in these reviews, but no review was devoted
exclusively to this topic. Thus, the following discussion
will focus on state-of-the-art control of four primary pests
or pest complexes in the United States by augmentative
releases of predators and parasitoids: bollworm/tobacco
budworm complex; the boll weevil; the pink bollworm; and
plant bugs, specifically Lygus spp.

Control of the Bollworm and Tobacco
Budworm

Parasitoids

The principal parasitoids that contribute to mortality of
bollworm and tobacco budworm eggs and larvae are given
in the following list.

Important parasitoids and predators of H. zea and
H. virescens

Parasitoids

Predators

Microplitis croceipes

Geocoris spp.

Cardiochiles nigriceps

Orius spp.

Cotesia marginiventris

Chrysoperla spp.

Campoletis sonorensis

Coccinellids

Trichogramma spp.

Nabids

Eucelatoria bryani

Spiders

Of these parasitoids, primary attention has been given to
augmenting Trichogramma populations. More recently,
major emphasis is being placed by three Agricultural
Research Service (ARS) laboratories on the development of
augmentative technology for the larval endoparasitoid
Microplitis croceipes (Cresson). One laboratory also is
researching the use of the tachinid, Archytas marmoratus

48

(Townsend), and considerable progress at another labora¬
tory has been made on the in vitro rearing of the tachinids,
Eucelatoria bryani Sabrosky and Palexorista laxa (Curran).

Trichogramma

Biological control of the bollworm and tobacco budworm in
cotton by augmentative releases of Trichogramma in the
United States, particularly T. pretiosum Riley, is compre¬
hensively reviewed in King et al. (1985a). All aspects are
reviewed including rearing (Morrison 1985a, 1985b),
transport, storage, and parasitoid release technology (Bouse
and Morrison 1985), behavioral manipulation (Lewis et al.
1985) and parasitoid movement (Keller and Lewis 1985),
efficacy (King et al. 1985b, Lopez and Morrison 1985),
pesticide effects (Bull and Coleman 1985), and modeling of
augmentative releases (Goodenough and Witz 1985).
Another recent popularized review of the state-of-the-art
technology for identifying, propagating, and augmenting
Trichogramma populations is given by Olkowski and
Zhang (1990).

Olkowski and Zhang (1990) list seven commercial produc¬
ers of Trichogramma in the United States. These parasi-
toids are released over a total of about 81,000 hectares.

The parasitoid most commonly reared and released in
cotton is T. pretiosum, and the Angoumois grain moth,
Sitotroga cerealella (Olivier), is the host of choice.

The technical feasibility of suppressing bollworm and
tobacco budworm populations in cotton by augmentative
releases of Trichogramma has been repeatedly demonstrated
in the United States (Table 1). Lingren and Kim (unpub¬
lished data) manually released 494,000 adult
Trichogramma! ha resulting in about 60% parasitism of
bollworm and tobacco budworm for over a month in and
near the release area. Aerial releases of 123,500-247,000
adult Trichogramma/ha resulted in an average 5 1 % parasit¬
ism of bollworm and tobacco budworm eggs on five Texas
cotton farms (Ridgway et al. 1977).

Stinner et al. (1974) evaluated the technical feasibility of
reducing bollworm and tobacco budworm larval popula¬
tions in cotton by releasing T. pretiosum. Parasitoid
release rates were high (up to 957,000/ha), but larval
populations were suppressed. King et al. (1985b) reported
three years of data following releases of T. pretiosum in
cotton. In each year, egg parasitism was increased as a
consequence of the release of parasitoids, but these parasit¬
ism rates could not be correlated with larval suppression.

In the third year, parasitoid release plots yielded more
cotton than untreated plots, though only 77 % as much lint
as in the insecticide-treated plots.

Perhaps the major factor limiting Trichogramma use for
boll worm/ tobacco budworm control in cotton in the United
States is the preferential use of effective chemicals for
controlling a complex of pests including plant bugs and the
boll weevil. Since Trichogramma is extremely sensitive to
nearly all synthetic chemicals, its use is prohibited. This
situation may change as chemicals become less available,
boll weevils are eliminated as a pest from parts of the
United States, and other nonchemical control strategies are
developed. More research is needed to develop technology
for managing parasitoids after release and for producing
more vigorous parasitoids with longer life spans. Current
procedures call for releasing the parasitoid at rates exceed¬
ing 100, 000/ha at 3-day intervals during moth oviposition
periods.

Larval Parasitoids

Larval parasitoids are an important part of the environmen¬
tal resistance to increase of bollworm and tobacco budworm
populations. Unique complexes of hymenopterous and
tachinid parasitoids have been recorded in the various
regions of the world (King and Jackson 1989). Cumulative
rates of larval parasitism are often high but the predominant
species vary between regions of the country in the United
States as well as crop (King et al. 1982). Special attention
will be given to M. croceipes, one of the most important
parasitoids of bollworm and tobacco budworm larvae in
cotton and wild host plants in the United States (King and
Powell 1989). Over 50% parasitism of larvae was recorded
during 1981 and 1982 in a series of cotton fields in the
midsouthem United States despite application of
insecticides.

Potential for using larval parasitoids in augmentative
releases has been indicated in small-scale tests. Lingren
(1969) reported that Cotesia (= Apanteles ) marginiventris
(Cresson) had considerable potential for use in augmenta¬
tive programs. Also, Campoletis sonorensis (Carlson)
released at the rate of 680/day for 10 consecutive days in a
0.2-ha cage (34,000 wasps/ha equivalent) infested with
tobacco budworm larvae resulted in 85 percent parasitiza-
tion for nine consecutive weeks (Lingren 1977). Jackson et
al. (1970) reported 58 percent parasitization of third-instar
tobacco budworm larvae in cages when 2964 (equivalent)
M. croceipes female wasps were released per hectare.

49

Jackson et al. (1970) reported that if the tachinids, E.
bryani and P. laxa, were released at the rate of 6,175
female flies/ha on cotton containing 12,350 bollworm and
tobacco budworm larvae/ha, about 50 percent parasitization
should occur in two days.

Research on M. croceipes has been extensively reviewed
(Powell et al. 1989). Basic biology including host relation¬
ship physiology was recently reviewed by Powell and
Elzen (1989) and Vinson and Dahlman (1989). Behavioral
aspects relating to habitat and host location, mate finding,
and mating were reviewed by Nordlund et al. (1989), Elzen
and Powell (1989), and Jones (1989). Other research vital
to development of augmentative technology for M.
croceipes is the effect of insecticides on the parasitoid (Bull
et al. 1989), genetic characterization and genetic improve¬
ment (Steiner and Teig 1989), and the possibility of
developing an in vitro rearing system for the parasitoid
(Greany et al. 1989).

Hopper (1989) surmised that augmentation of M. croceipes
for control of the bollworm and tobacco budworm is
technically feasible. Of the principal parasitoids of the
bollworm and tobacco budworm, across host plants,

M. croceipes has emerged as one of the most important
(King et al. 1985c, King and Powell 1989). King et al.
(1985d) hypothesized, based on dramatically higher rates of
bollworm and tobacco budworm larval parasitism, particu¬
larly by M. croceipes , that M. croceipes was highly tolerant
of many commonly used insecticides, particularly the
pyrethroids. In general, the parasitoids are more tolerant of
certain pyrethroids (e.g., esfenvalerate and cypermethrin)
and carbamates (e.g., thiodicarb and oxamyl) and least
tolerant of certain organophosphates (e.g., acephate and
profenofos) (Powell and Scott 1985).

Microplitis croceipes prefers to parasitize third instars
(Hopper and King 1984b), but all parasitized instars move
and feed less on the cotton plant (Hopper and King 1984a).
Consequently, less damage is caused by parasitized larvae.
Hopper et al. (1991) report that releasing 2,000 female M.
croceipes per hectare of cotton resulted in 75 % parasitized
bollworm and tobacco budworm larvae after six days, with
an estimated 38% reduction in damage. Hopper (1989)
suggested that releases over large areas, particularly during

the time that bollworm and tobacco budworm are restricted
to wild host plants (valid only for the Midsouth United
States), might be an effective population suppression tactic.

Growers may not use M. croceipes in cotton even if it
could be mass produced because apparently the presence of
feeding larvae in cotton would not be tolerated, though
feeding is reduced. Perhaps the best approach would be
releasing parasitoids on an area-wide basis while bollworm/
tobacco budworm populations are still on early-season wild
host plants along roadsides and field margins, as stated in
this proceedings by D. D. Hardee. This area-wide suppres¬
sion would provide an additive mortality factor to virus
application, and parasitoids would search where the virus
had not been applied.

Predators

No predators are currently being propagated and released
for controlling the bollworm and tobacco budworm in
cotton in the United States. Most management models do
include predator-caused mortality, at least indirectly if not
directly (See T. R. Wagner, these proceedings).

Releases of several hemipteran predators indicate that it
might be feasible to augment their populations if economi¬
cal procedures for mass producing them could be devel¬
oped. Field-cage studies by Lingren et al. (1968), van den
Bosch et al. (1969), and Lopez et al. (1976) with Geocoris
punctipes (Say), Nabis americoferus Carayon, and Podisus
maculiventris Say, respectively, in cotton demonstrate their
ability to suppress bollworm and tobacco budworm
populations.

Ridgway and coworkers demonstrated the technical feasibil¬
ity of suppressing bollworm and tobacco budworm larval
populations in cotton by periodic release of C. carnea eggs
or larvae (Table 2). Release of 2- to 3-day-old larvae
consistently produced significant reductions of bollworm
and tobacco budworm populations on cotton. Reductions
in bollworm and tobacco budworm larval populations were
obtained by releasing as few as 24.7 thousand C. carnea
larvae/ha, and high levels of reduction were obtained in the
field by releasing 247,000 to 494,000/ha (Ridgway et al.
1977).

50

Control of the Boll Weevil

Successful colonization of cotton, a new host plant, enabled
the boll weevil to migrate from its original distribution area
in southeastern Mexico and north- central South America to
the United States through the wild and cultivated cottons of
Mexico (Burke et al. 1986). The boll weevil escaped its
effective natural enemies, and attempts to establish effective
species on it have failed so far.

At least 55 indigenous parasitoid species are known to
attack the boll weevil in the United States (Pierce 1908,
Hunter 1910, Pierce et al. 1912, and Chestnut and Cross
1971). These parasitoids characteristically have a wide host
range and are facultative parasitoids of the boll weevil
(Pierce 1908, Cushman 1911). Consequently, they do not
respond to boll weevil population dynamics as would a
more host-specific parasitoid. One of these parasitoids,
Bracon mellitor Say, has been studied (Adams et al. 1969,
Barfield et al. 1977), and is often recorded from field-
collected, weevil-infested fruit (Bottrell 1976). Bracon
mellitor parasitism rates of the boll weevil can be high
(Marlatt 1933), and these rates are affected by cotton
variety (McGovern and Cross 1976, Adams et al. 1969).
However its facultative host selection behavior, failure to
search the ground for weevil larvae, and preference for late
instar larvae reduces its effectiveness (Cate 1985, Meinke
and Slosser 1982).

Cate and associates have identified at least 14 parasitoids
attacking the boll weevil and a closely related species,
Anthonomus hunteri Burke and Cate, in southern Mexico.
Rearing techniques have been developed for several of these
species (Cate 1987).

One of the most promising exotic species is the pteromalid
Catolaccus grandis Burks. Cate (1985) reported that
experimental field releases of C. grandis in Mississippi
resulted in a five-fold population increase in larval parasit¬
ism, but the parasitoid failed to survive the winter. In a
south Texas study, low rates of C. grandis released on low
density host populations resulted in high parasitization rates
(Morales-Ramos and King, 1991). Certainly, indications
are that C. grandis is an efficient host searcher and it is
capable of parasitizing a high percentage of boll weevil-
infested fruit. So, its failure to overwinter can be overcome
by annual propagation and release. Knipling (1990,
personal communication) demonstrated in a simulation
model that release of as few as 20 female C. grandis per

acre of cotton could result in high rates of parasitism and
virtual elimination of damage from progeny of F: genera¬
tion adult weevils. Successful nonchemical control of this
key pest might allow successful biological control of
bollworm, tobacco budworm and other secondary pest
populations.

Control of the Pink Bollworm

Several native predators and exotic parasitoids have been
evaluated as potential candidates for augmentation against
pink bollworm. Studies by Irwin et al. (1974) suggested
that most predators tend to be effective against the pink
bollworm only at relatively high densities. Strategies have
been developed to manipulate certain crops as field "nurser¬
ies" for native predators, e.g., alfalfa (Stem et al. 1967)
and sorghum (Fye 1971, DeLoach and Peters 1972, Fye
and Carranza 1972, Robinson et al. 1972).

Bryan et al. (1973a, 1973b) released more than two million
Bracon kirkpatricki and ca. 280,000 Chelonus blackburni in
approximately 113 acres (46 ha) of Arizona cotton during
1972, and documented a significant reduction in the need
for insecticidal treatment in release plots (one treatment for
Lygus and one for pink bollworm) versus control plots
(four treatments for pink bollworm and one for the boll¬
worm. Parasitism by B. kirkpatricki ranged up to 25
percent (which the authors considered an underestimation),
whereas C. blackburni appeared to be largely ineffective
(which the authors attributed to release in insufficient
numbers). Later, Bryan et al. (1976) documented parasit¬
ism of about 32 percent by B. kirkpatricki and about nine
percent by C. blackburni, but concluded that such rates
were insufficient to control pink bollworm.

Inundative release of several parasitoid species in the lower
Colorado Desert of Arizona and California produced
variable levels of pink bollworm control (Legner and
Medved 1979). Most effective was Chelonus sp. nr.
curvimaculatus Cameron, which was credited with an
adjusted 69.6 percent infested boll reduction at the equiva¬
lent release rate of 2,667 females/ha (Legner and Medved
1979). The prospects of parasitoid augmentation as an
effective control strategy have been enhanced by the fact
that artificial diets and/or mass-rearing procedures have
been developed for pink bollworm (Adkisson et al. 1960,
Stewart 1984) and several of its associated hymenopterous
parasitoids (Bryan et al. 1969).

51

Control of Lygus Species

Lygus are attacked by a wide range of predators and
parasitoids, particularly while on host plants other than
cotton. A particularly effective predator of plant bugs is
Geocoris spp. (Leigh and Gonzalez (1976). High rates of
egg parasitism (36 percent) by the mymarid, Anaphes iole
(Girault) (=A. ovijentatus (Crosby and Leonard)) has been
recorded in cotton (Graham et al. 1986), but most studies
indicate highest rates of parasitism in agriculturally undis¬
turbed areas (Scales 1973, Sillings and Broersoma 1974).

Debolt (1987) discusses in some detail the potential for
using augmentative releases of parasitoids to control Lygus.
The development of an artificial diet for L. hesperus
(Debolt 1982) has made the production of large numbers of
host Lygus nymphs and eggs possible. An ARS Pilot Test
is presently being conducted to assess the technical feasibil¬
ity of controlling L. hesperus in alfalfa by augmentative
releases of the mymarid egg parasitoid, Anaphes iole, the
braconid nymphal parasitoid Leiophron uniformis (Gahan).
Control of L. hesperus in alfalfa would limit migration of
the bugs into cotton after the plant is cut for hay or matures
for seed production.

In small field plots in 1990, average egg parasitism rates of
57 % were achieved with A. iole with single releases of
130,000 parasitoids (male and female) per acre. Average
parasitism levels of 29 % of the nymphs occurred following
one-time releases of 22,000 female L. uniformis per acre
(C. G. Jackson, Personal communication, 23 August 1990).

Insect Rearing

Efficient and cost effective methods of rearing predators
and parasitoids must be developed if augmentative releases
are to be feasible (Beime 1974). Literally millions of
natural enemies, available at somewhat unpredictable times,
may be required for commercial augmentative releases.
Finney and Fisher (1964) discussed problems associated
with the culture of predators and parasitoids. Considerable
attention has been devoted to development of techniques to
produce quality insects in large numbers (King and Leppla
1984, Smith 1966). The genetic implications of long term
laboratory rearing of insects are addressed elsewhere
(Bouletreau 1986, Mackauer 1972, 1976).

Powell and Hartley (1987) described techniques for
manually producing large numbers of parasitoids effi¬
ciently. These authors adapted a multicellular host rearing
tray technique (Hartley et al. 1982) to rear M. croceipes
and several other species of parasitoids. Techniques,
involving use of low temperature storage, allowed simulta¬
neous release of nearly 17,000 wasps. They noted several
factors that were important for maintaining the rearing
program which may be applicable to others. The quality of
mass-reared insects also should be assessed (Boiler and
Chambers 1977).

Our present inability to mass propagate parasitoids and
predators of assured quality for a competitive cost is a
major constraint to the augmentative approach in the United
States. However, this may not pose a challenge in the near
future as scientists (J. L. Roberson, D. D. Hardee) at the
USDA Gast Rearing Laboratory, Southern Insect Manage¬
ment Laboratory, come closer to making automated mass
production of M. croceipes a reality (Powell and Roberson,
In press). Commercialization may be practical only for
selected organisms for which suitable diets and storage
methods are developed. Artificial rearing is the ' ' break¬
through" that would potentiate practical use of augmenta¬
tion for widespread pest control. Various groups have
made progress in in vitro rearing of parasitoids, but a
feasible system remains to be developed. As many as 22
entomophagous species have been reared in vitro. Several
Hymenoptera (1 ectoparasite, 4 pupal parasitoid species, 4
species of Trichogramma) and 3 species of Diptera have
been cultured with varying success (King et al. 1984).
Predators, notably C. carnea, have been reared on artificial
diets (Vanderzant 1973, Martin et al. 1978).

While there have been numerous successes in oviposition
stimulant identification or partial rearing (Nettles and Burks
1975, Nettles 1982), definitive development of a feasible in
vitro rearing system for parasitoids has yet to be developed.
Although considerable advances have been made in in vivo
rearing, the advances have not been achieved to such an
extent with in vitro rearing. The recent work of Wu et al.
(1982) illustrated an instance in which a completely
synthetic artificial host egg was produced which contains no
insect derivatives, and supports Trichogramma oviposition
and development.

Hymenopterous larval endoparasitoids have not been
successfully reared to the adult stage on artificial diet.
However, C. marginiventris and M. croceipes have been

52

reared on artificial media through the first instar (P.

Greany, pers. com.). Larval endoparasitoids have evolved
complex mechanisms that interact with the host's internal
dynamics and organs without damaging this environment or
causing untimely demise of the host. The workings of
these interacting factors must be at least understood
minimally for in vitro rearing of larval endoparasitoids to
become a reality (Thompson, 1986).

Developments in rearing natural enemies on artificial diet
may allow us to produce sufficient numbers to further
evaluate natural enemies as biological control agents. An
artificial diet that can be encapsulated has been developed
for Chrysopa, and G. punctipes has been reared for more
than 65 generations on an artificial diet (Cohen, In Press).

In both cases, mass propagation requires significant
engineering input.

Behavioral Manipulation

A variety of behavior-modifying chemicals
(semiochemicals) affect the actions of predators and
parasitoids in cotton and surrounding fields. These
chemicals function at different trophic levels. In addition
to chemicals from the parasitoid or predator and from the
host or prey, chemical cues emanating from the host habitat
affect the behavior of the natural enemies. Moreover,
chemical cues from the parasitoid or predator may be used
in host seeking by hyperparasitoids.

Understanding this complex chemical communication by
parasitoids and predators, and between the different trophic
levels, is critical for managing natural enemies and predict¬
ing levels of biological control.

Numerous kairomonal relationships between cotton arthro¬
pod pests and their natural enemies have been reported.

For example, dispersal from the target area after augmenta¬
tive release of predators or parasitoids often reduces the
effectiveness of the augmentative approach. Provision of
supplemental resources such as kairomones to attract,
arrest, retain, or stimulate the natural enemy to search more
intensively for the host or prey could provide mechanisms
for managing parasitoids and predators (Nordlund and Sauls
1981). Tricosane and other hydrocarbons extracted from
moth scales have been shown to increase parasitization by
Trichogramma spp. (Gross et al. 1975, Jones et al. 1973).
Oviposition stimulating kairomones have been extracted

from the accessory gland of the female bollworm (Nordlund
et al. 1987). The sex pheromone, gossyplure, from the
pink bollworm caused increased parasitization of pink
bollworm eggs by T. pretiosum (Zaki 1985). Kairomones
have been extracted from the frass of bollworm and tobacco
budworm larvae that cause host seeking by the larval
parasitoids M. croceipes and Cardiochiles nigriceps
Viereck; one chemical identified was 13-
methylhentriacontane (Jones et al. 1971, Nordlund and
Lewis 1985, Heath et al. 1990). A proteinaceous material
found in the frass or hemolymph of tobacco budworm
larvae stimulated larviposition by the tachinid A.
marmoratus (Nettles and Burks 1975).

Other behavioral chemicals termed synomones because they
facilitate location of the host by the parasitoid and conse¬
quently benefit the parasitoid and plant, have been found in
cotton. The ichneumonid, Campoietis sonorensis Cameron
was shown to orient to and search host-free cotton plants
(Elzen et al. 1983). Host-searching behavior by C.
sonorensis (Elzen et al. 1984) and M. croceipes (Nordlund
and Sauls 1981) increased on artificial diet-fed bollworm
and tobacco budworm larvae if cotton-plant material was
put in the diet. Turlings et al. (1990) discovered terpenoid
volatiles from com plants after feeding by Spodoptera
exigua (Hubner) caterpillars that attracted females of the
parasitic wasp Cotesia marginiventris (Cresson) for host
finding, they concluded that these terpenoids may be
produced in defense against herbivores but may also serve a
secondary function in attracting herbivore natural enemies.

Predators respond to many of the same chemical cues.

Lewis et al. (1977) demonstrated an increased search rate
by the common green lacewing, Chrysoperla cornea
Stephens, on eggs of the bollworm when bollworm moth
scales or extracts of the scales were applied to the search
area. A compound in cotton, B-caryophyllene, was
attractive to adult female green lacewings (Flint et al.

1979). Sabelis and Dicke (1985) summarized data showing
that the phytoseid mite, Phytoseiulus persimilis Athias-
Henriotuses, uses kairomones from its prey to locate it.

Summary

Much of the basic information required to implement
biological control by augmentation in the United States is
available. Enhanced knowledge on management of preda¬
tor and parasitoid behavior coupled with use of genetically

53

improved strains will lead to increased effectiveness of
biological control. The development of artificial diets and
in vitro rearing procedures for parasitoids and predators
will open the path for their mass production and commer¬
cial distribution for augmentative releases. Over the next
decade, social and environmental pressures in many
countries will expedite implementation of technology for
biological control of arthropod pests of cotton as well as
other crops by propagation and augmentative releases of
their natural enemies (predators and parasitoids).

References Cited

Abies, J. R., S. L. Jones, R. K. Morrison, V. S. House,

D. L. Bull, L. F. Bouse and J. B. Carlton. 1979. New
developments in the use of Trichogramma to control
lepidopteran pests of cotton, p. 125-127. Proc. Beltwide
Cotton Prod. Res. Conf., National Cotton Council,
Memphis, TN.

Adams, C. H., W. H. Cross and H. C. Mitchell. 1969.
Biology of Bracon mellitor, a parasite of the boll weevil. J.
Econ. Entomol. 62: 889-895.

Adkisson, P. L., E. S. Vanderzant, D. L. Bull and W. E.
Allison. 1960. A wheat germ medium for rearing the pink
bollworm. J. Econ. entomol. 53: 759-761.

Anonymous. 1984. Integrated pest management for cotton
in the Western Region of the United States. Univ. of
Calif., Berkeley, Div. Agric. Sci., Publ. 3305.

Barfield, C. S., P. J. H. Sharpe and D. G. Bottrell. 1977.
A temperature-driven developmental model for the parasite
Bracon mellitor (Hymenoptera: Braconidae). Can.

Entomol. 109: 1503-1514.

Beime, B. P. 1974. Status of biological control proce¬
dures that involve parasites and predators, pp. 69-76. In F.
G. Maxwell and F. A. Harris, [eds.], Proceedings of the
summer institute on biological control of plant insects and
diseases. Univ. Press of Mississippi, Jackson.

Boiler, E. F., and D. L. Chambers. 1977. Quality aspects
of mass-reared insects. In R. L. Ridgway and S. B. Vinson
[eds.], Biological Control by Augmentation of Natural
Enemies, pp. 219-236. Plenum Press, New York.

Bottrell, D. G. 1976. Biological control agents of the boll
weevil, pp. 22-25. In Proc. Conference on boll weevil
suppression, management and elimination technology,
February 13-15, 1974, Memphis, Tennessee, USDA, ARS-
S-71.

Bottrell, D. G., and Adkisson P. C. 1977. Cotton insect
pest management. Ann. Rev. Entomol. 22: 451-481.

Bouletreau, M. 1986. The genetic and coevolutionary
interactions between parasitoids and their hosts, pp. 169-
200. In Waage, J., and D. Greathead [eds.], Insect
parasitoids: 13th symposium of the Royal Entomological
Society of London, 18-19 September 1985 at the Depart¬
ment of Physics Lecture Theatre, Imperial College,

London. London: Academic.

Bouse, L. F., and R. K. Morrison. 1985. Transport,
storage, and release of Trichogramma pretiosum. South¬
west. Entomol. 8: 36-48.

Bryan, D. E., R. E. Fye, C. G. Jackson and R. Patana.
1976. Nonchemical control of pink bollworms. USDA,
ARS W-39.

Bryan, D. E., R. E. Fye, C. G. Jackson and R. Patana.
1973a. Releases of parasites for suppression of pink
bollworms in Arizona. USDA, ARS W-7.

Bryan, D. E., R. E. Fye, C. G. Jackson and R. Patana.
1973b. Releases of Bracon ldrkpatricki (Wilkinson) and
Chelonus blackburni Cameron for pink bollworm control in
Arizona. USDA Prod. Res. Rep. 150.

Bryan, D. E., C. G. Jackson and A. Toner. 1969. Rearing
cotton insect parasites in the laboratory. USDA Prod. Res.
Rep. 109.

Bull, D. L., and R. J. Coleman. 1985. Effects of pesti¬
cides on Trichogramma spp. Southwest. Entomol. 8: 156-
168.

Bull, D. L., E. G. King and J. E. Powell. 1989. Effects
and fate of selected insecticides after application to
Microplitis croceipes. Southwest. Entomol. Suppl. 12.
pp. 59-70.

54

11

Burke, H. R., W. E. Clark, J. R. Cate and P. A. Fryxell.
1986. Origin and dispersal of the boll weevil. Bull.

>• Entomol. Soc. Amer. 32: 228-238.

Cate, J. R. 1985. Cotton: Status and current limitations to
biological control in Texas and Arkansas, pp. 537-556. In
M. A. Hoy and D. C. Herzog [eds.]. Biological control in
Agricultural IPM systems. Academic Press, Inc., New
York.

Cate, J. R. 1987. A method of rearing parasitoids of boll
weevil without the host plant. Southwest. Entomol. Suppl.
12. pp. 211-215.

Chestnut, T. L., and W. H. Cross. 1971. Arthropod
parasites of the boll weevil, Anthonomus grandis : Compari¬
son of their importance in the United States over a period
of thirty-eight years. Ann. Entomol. Soc. Amer. 64: 549-
557.

Clausen, C. P. 1978. Introduced parasites and predators
of arthropod pests and weeds: A world review. USDA
Agric. Handb. 480.

Cohen, A. C. In Press. Using a systematic approach to
develop artificial diets for predators. In N. C. Leppla and
T. L., anderson [eds.]. Advances in Insect Rearing for
Research and Pest Management.

Coulson, J. R. 1987. Studies on the biological control of
plant bugs (Heteroptera: Miridae): An introduction and
history, 1961-83, pp. 1-12. In R. C. Hedlund and H. M.
Graham [eds.], Economic Importance and Biological
Control of Lygus and Adelphocoris in North America.
USDA-ARS. ARS-64.

Cushman, R. A. 1911. Studies in the biology of the boll
weevil in the Mississippi Delta Region of Louisiana. J.
Econ. Entomol. 4: 432-448.

DeBach, P., and K. S. Hagen. 1964. Manipulation of
entomophagous species, pp. 429-58. In P. DeBach (ed.),
Biological Control of Insect Pests and Weeds. Chapman
and Hall, London.

Debolt, J. W. 1982. Meridic diet for rearing successive
generations of Lygus hesperus. Ann. Entomol. Soc. Amer.
75: 119-122.

Debolt, J. W. 1987. Augmentation: rearing, release, and
evaluation of plant bug parasites, pp. 82-87. In Hedlund,

R. C., and H. M. Graham [eds.], Economic Importance
and Biological Control of Lygus and Adelphocoris in North
America. USDA-ARS. ARS-64.

DeLoach, C. J., and D. G. Peters. 1972. Effect of strip¬
planting and solid- planting on predators of cotton insects
in southeastern Missouri. Environ. Entomol. 1: 94-102.

Elzen, G. W., and J. E. Powell. 1989. Mate finding and
mating behavior of Microplitis croceipes, pp. 49-52. In J.
E. Powell, D. L. Bull, and E. G. King [eds.], Biological
Control of Heliothis spp. by Microplitis croceipes. South¬
west. Entomol. Supplement No. 12.

Elzen, G. W., H. J. Williams and S. B. Vinson. 1984.
Isolation and identification of cotton synomones mediating
searching behavior by parasitoid Campoletis sonorensis.

J. Chem. Ecol. 10: 1251-1264.

Elzen, G. W., H. J. Williams and S. B. Vinson. 1983.
Response by the parasitoid Campoletis sonorensis (Hym-
enoptera: Ichneumonidae) to chemicals (synomones) in
plants: Implications for host habitat location. Environ.
Entomol. 12: 1873-1877.

Finney, G. L and T. W. Fisher. 1964. Culture of ento¬
mophagous insects and their hosts, pp. 329-55. In P.
DeBach (ed.). Biological Control of Insect Pests and
Weeds. Chapman and Hill, Ltd., London.

Flint, H.M., S.S. Salter and S. Walters. 1979.
Caryophyllene: An attractant for the green lacewing.
Environ. Entomol. 8: 1123-1125.

Fye, R. E. 1971. Grain sorghum: A source of insect
predators for insects on cotton. Prog. Agr. in Ariz. 23: 12-
13.

Fye, R. E., and R. L. Carranza. 1972. Movement of
insect predators from grain sorghum in cotton. Environ.
Entomol. 1: 790-791.

Goodenough, J. L., and J. A. Witz. 1985. Modeling
augmentative releases of Trichogramma pretiosum. South¬
west. Entomol. Suppl. 8. pp. 169-189.

55

Graham, H. M., C. G. Jackson and J. W. Debolt. 1986.
Lygus (Hemiptera: Miridae) and their parasites in agricul¬
tural areas of southern Arizona. Environ. Entomol. 15:
132-142.

Greany, P. D., S. M. Ferkovich and W. R. Clark. 1989.
Progress towards development of an artificial diet and an in
vitro rearing system for Microplitis croceipes. Southwest.
Entomol. Suppl. 12. pp. 89-94.

Greany, P. K., S. B. Vinson and W. J. Lewis. 1984.

Insect parasitoids: finding new opportunities for biological
control. BioScience. 34: 690-696.

Gross, H. R., Jr., W. J. Lewis, R. L. Jones and D. A.
Nordlund. 1975. Kairomones and their use for manage¬
ment of entomophagous insects: III. Stimulation of
Trichogramma achaeae, T. pretiosum, and Microplitis
croceipes with host-seeking stimuli at time of release to
improve their efficiency. J. Chem. Ecol. 1: 431-438.

Hartley, G. G., E. G. King, F. D. Brewer and C. W.

Gantt. 1982. Rearing of the Heliothis sterile hybrid with a
Heliothis multicellular larval rearing container and pupal
harvesting. J. Econ. Entomol. 75: 7-10.

Heath, R. R., S. M. Ferkovich, P. D. Greany, F. J. Eller,
B. D. Dueben and R. L. Tilden. 1990. Progress in the
isolation and characterization of a host hemolymph oviposi-
tional kairomone for the endoparasitoid Microplitis
croceipes. Arch. Insect Biochem. Physiol. 13: 255-265.

Herzog, G. A., and E. G. King. 1990. Highlights of the
1990 Cotton Insect Research and Control Conference, pp.
84-86 In Proceedings, Beltwide Cotton Prod. Res. Conf.,
January 10-11, 1990, Las Vegas, Nevada.

Hopper, K. R. 1989. Conservation and augmentation of
Microplitis croceipes for controlling Heliothis spp. South¬
west. Entomol. Suppl. 12. pp. 95-115.

Hopper, K. R., and E. G. King. 1984a. Feeding and
movement on cotton of Heliothis species (Lepidoptera:
Noctuidae) parasitized by Microplitis croceipes (Hym-
enoptera: Braconidae). Environ. Entomol. 13: 1654-1660.

Hopper, K. R., and E. G. King. 1984b. Preference of
Microplitis croceipes (Hymenoptera: Braconidae) for

instars and species of Heliothis (Lepidoptera: Noctuidae).
Environ. Entomol. 13: 1145-1150.

Hopper, K. R., J. E. Powell and E. G. King. 1991.

Spatial density dependence in parasitism of Heliothis
virescens (Lepidoptera: Noctuidae) by Microplitis croceipes
(Hymenoptera: Braconidae) in the field. Environ.

Entomol. 20: 292-302.

Hunter, W. D. 1910. The status of the boll weevil in
1909. USDA Bur. Entomol. Circ. 122.

Irwin, M. E., R. E. Gill and D. Gonzales. 1974. Field-
cage studies of native egg predation of the pink bollworm
in southern California cotton. J. Econ. Entomol. 67: 193-
196.

Jackson, C. G., D. E. Bryan, E. G. Neeman and A. L.
Wardecker. 1970. Biological control: Results of field cage
test with parasites of Heliothis spp. Third Qtr. Rpt.,

Cotton Insects Biological Control Investigations, U.S.

Dept. Agric., Tucson, AZ.

Jones, R. L. 1989. Semiochemicals mediating Microplitis
croceipes habitat, host, and mate finding behavior. South¬
west. Entomol. Suppl. 12. pp. 53-57.

Jones, R. L., W. J. Lewis, M. Beroza, B. A. Bierl and A.
N. Sparks. 1973. Host seeking stimulants (kairomones)
for the egg parasite Trichogramma evanescens. Environ.
Entomol. 2: 593-596.

Jones, R. L., W. J. Lewis, M. C. Bowman, M. Beroza and
B. A. Bierl. 1971. Host-seeking stimulant for parasite of
com earworm: Isolation, identification, and synthesis.
Science 173: 842-843.

Jones, S. L., R. K. Morrison, J. R. Abies, L. F. Bouse, J.
B. Carlton and D. L. Bull. 1979. New techniques for the
aerial release of Trichogramma pretiosum. Southwest.
Entomol. 4: 14-19.

Keller, M. A., and W. J. Lewis. 1985. Movements by
Trichogramma pretiosum (Hymenoptera:
Trichogrammatidae) released into cotton. Southwest.
Entomol. Suppl. 8: 99-109.

56

King, E. G., J. E. Powell, and J. W. Smith. 1982.
Prospects for utilization of parasites and predators for
management of Heliothis spp., pp. 103-122. In Reed, W.
(ed.), ICRISAT (International Crops Research Institute for
the Semi-Arid Tropics): Proceedings of the International
Workshop on Heliothis Management, 15-20 November
i 1981. ICRISAT Center, Patancheru, A.P., India.
Patancheru, A.P., India: ICRISAT.

King, E. G., and N. C. Leppla. 1984. Advances and
Challenges in Insect Rearing. USDA, Agric. Res. Serv.,
New Orleans.

King, E. G., D. L. Bull, L. F. Bouse and J. R. Phillips.
1985a. Introduction: Biological control of Heliothis spp. in
cotton by augmentative releases of Trichogramma. South¬
west. Entomol. 8: 1-10.

King, E. G., R. J. Coleman, J. R. Phillips and W. A.
Dickerson. 1985b. Heliothis spp., and selected natural
enemy populations in cotton: A comparison of three insect
control programs in Arkansas (1981-82) and North Caro¬
lina (1983). Southwest. Entomol. Suppl. 8: 71-98.

King, E. G., K. R. Hopper and J. E. Powell. 1985c.
Analysis of systems for biological control of crop arthropod
pests in the U.S. by augmentation of predators and para¬
sites. See Ref. 52, pp. 201-227.

King, E. G., J. E. Powell and R. J. Coleman. 1985d. A
high incidence of parasitism of Heliothis spp. larvae in
cotton in southeastern Arkansas. Entomophaga. 30: 419-
426.

King, E. G. 1986. Insecticide use in cotton and the value
of predators and parasites for managing Heliothis , pp. 155-
162 In Proc. Beltwide Cotton Prod. Res. Conf., Jan. 4-9,
1986. Las Vegas, Nevada.

King, E. G., J. R. Coulson and R. J. Coleman. 1988.

ARS National Biological Control Program. In Proceed¬
ings, Workshop on Research Priorities, July 14-15, 1987,
Beltsville, Maryland.

King, E. G., and R. D. Jackson, [eds.]. 1989. Proceed¬
ings of the Workshop on Biological Control of Heliothis :
Increasing the Effectiveness of Natural Enemies. 11-15
Nov. 1985, New Delhi, India. Far Eastern Regional
Research Office, U.S. Department of Agriculture, New
Delhi, India.

King, E. G., and J. E. Powell. 1989. Biological control of
Heliothis spp. emphasizing the present and potential role of
Microplitis croceipes. Southwest. Entomol. Suppl. 12. pp.
1-9.

King, E. G., and G. A. Herzog. 1990. 43rd Annual
Conference Report on Cotton Insect Research and Control,
pp. 140-154. In Proceedings, 1990 Beltwide Cotton
Production Research Conferences, January 9-14, 1990, Las
Vegas, Nevada.

Legner, E.F. 1979. Emergence patterns and dispersal in
Chelonus spp. near curvimaculatus and Pristomerus
hawaiiensis parasitic on Pectinophora gossypiella. Ann.
Entomol. Soc. Amer. 72: 681-686.

Legner, E. F., and R. A. Medved. 1979. Influence of
parasitic hymenoptera on the regulation of pink bollworm,
Pectinophora gossypiella, on cotton in the Lower Colorado
Desert. Environ. Entomol. 8: 922-930.

Leigh, T. F., and D. Gonzalez. 1976. Field cage evalua¬
tion of predators for control of Lygus hesperus Knight on
cotton. Environ. Entomol. 5: 948-952.

Lewis. W. J., H. R. Gross and D. A. Nordlund. 1985.
Behavioral manipulation of Trichogramma (Hymenoptera:
Trichogrammatidae). Southwest. Entomol. Suppl. 8. pp.
49-55.

Lewis, W. J., D. A. Nordlund, H. R. Gross, Jr., R. L.
Jones and S. L. Jones. 1977. Kairomones and their use for
management of entomophagous insects. V. Moth scales as
a stimulus for predation of Heliothis zea (Boddie) eggs by
Chrysopa carnea Stephens larvae. J. Chem. Ecol. 3: 483-
487.

Lingren, P. D., R. L. Ridgway and S. L. Jones. 1968.
Consumption by several common arthropod predators of
eggs and larvae of two Heliothis species that attack cotton.
Ann. Entomol. Soc. Am. 61: 613-18.

57

Lingren, P. D. 1969. Approaches to the management of
Heliothis spp. in cotton with Trichogramma spp. pp. 207-
217. In Proceedings, Tall Timbers Conference on Eco¬
logical Animal Control by Habitat Management, 26-28
Feb. 1969, Tallahassee, Fla. U.S.A.

Lingren, P. D. 1977. Campoletis sonorensis: Mainte¬
nance of a population on tobacco budworm in a field cage.
Environmental Entomology 6: 72-76.

Lopez Jr., J. D., and R. K. Morrison. 1985. Parasitiza-
tion of Heliothis spp. eggs after augmentative releases of
Trichogramma pretiosum Riley. Southwest. Entomol.
Suppl. 8. pp. 110-137.

Lopez Jr., J. D., R. L. Ridgway and R. E. Pinnell. 1976.
Comparative efficacy of four insect predators of the
bollworm and tobacco budworm. Environ. Entomol. 5:
1160-1164.

Luttrell, R. G., M. Crawford, W. C. Yearian, S. Y.

Young and A. J. Mueller. 1980. Aerial release of
Trichogramma pretiosum for control of Heliothis on cotton.
Ark. Farm Research. 29: 13.

Mackauer, M. 1972. Genetic aspects of insect control.
Entomophaga. 17: 27-48.

Mackauer, M. 1976. Genetic problems in the production
of biological control agents. Annu. Rev. Entomol. 21:
369-385.

Marlatt, C. L. 1933. Report of the chief of the Bureau of
Entomology. USDA, Washington, D.C.

Martin, P. B., R. L. Ridgway and C. E. Schuetze. 1978.
Physical and biological evaluations of an encapsulated diet
for rearing Chrysopa carnea. Fla. Entomol. 6: 145-152.

McGovern, W. L., and W. H. Cross. 1976. Effects of
two cotton varieties on levels of boll weevil parasitism
(Col.: Curculionidae). Entomophaga 21: 123-125.

Meinke, L. J., and J. E. Slosser. 1982. Fall mortality of
the boll weevil in fallen cotton squares, with emphasis on
parasite-induced mortality. Environ. Entomol. 11: 318-
323.

Morales-Ramos, J. A., and E. G. King. 1991. Evaluation
of Catolaccus grandis( Burks) as a biological control agent
against the cotton boll weevil, pp. 724. Proc. Beltwide
Cotton Prod. Res. Conf., Jan. 8-12, 1991, San Antonio,
Texas.

Morrison, R. K. 1985a. Effective mass production of eggs
of the Angoumois grain moth, Sitotroga cerealella
(Oliver). Southwest. Entomol. Suppl. 8. pp. 28-35.

Morrison, R. K. 1985. Mass production of Trichogramma
pretiosum Riley. Southwest. Entomol. Suppl. 8. pp. 21-
27.

National Academy of Sciences. 1969. Principles of plant
and animal pest control. Vol.3. Insect-pest management and
control. Publ. 1695.

Nettles Jr., W. C. 1982. Contact stimulants from
Heliothis virescens that influence the behavior of females of
the tachinid, Eucelatoria bryani. J. Chem. Ecol. 8: 1183-
1191.

Nettles, W. C., Jr., and M. L. Burks. 1975. A substance
form Heliothis virescens larvae stimulating larviposition by
females of the tachinid, Archytas marmoratus. J. Insect
Physiol. 21: 965-978.

Nordlund, D. A., and W. J. Lewis. 1985. Response of
the braconid parasitoid Microplitis demolitor to frass of
larvae of the noctuids, Heliothis zea and Trichoplusia ni
and to 13-methyihentriacontane. Entomol. Exp. Appl. 38:
109-112.

Nordlund, D. A., W. J. Lewis and J. H. Tumlinson.

1989. Habitat and host location behavior of Microplitis
croceipes. Southwest. Entomol. Suppl. 12. pp. 39-48.

Nordlund, D. A., and C. E. Sauls. 1981. Kairomones and
their use for management of entomophagous insects. XI.
Effect of host plants on kairomonal activity of frass from
Heliothis zea larvae for the parasitoid Microplitis croceipes.
J. Chem. Ecol. 7: 1057-1061.

Nordlund, D. A., M. R. Strand, W. J. Lewis and S. B.
Vinson. 1987. Role of kairomones from host accessory
gland secretion in host recognition by Telenomus remus and
Trichogramma pretiosum, with partial characterization.
Entomol. Exp. Appl. 44: 37-43.

58

Olkowski, W., and A. Zhang. 1990. Trichogramma - A
modern day frontier in biological control. The IPM
Practitioner 12: 1-15.

» I

Pierce, W. D. 1908. Studies of parasites of the cotton boll

weevil. USDA Bur. Entomol. Bull. 73.

:gs

Pierce, W. D., R.A. Cushman, C.E. Hood and W.D.
Hunter. 1912. The insect enemies of the cotton boll

weevil. USDA Bur. Entomol. Bull. 100.

w

Powell, J. E., and G. G. Hartley. 1987. Rearing
Microplitis croceipes (Hymenoptera: Braconidae) and other
parasitoids of Noctuidae with multicellular host-rearing

trays. J. Econ. Entomol. 80: 968-971.

id

Powell, J. E. 1989. Importation and establishment of
predators and parasitoids of Heliothis into the USA, pp.
387-395. In E. G. King and R. D. Jackson [eds.], Proc.
Workshop on Biological Control of Heliothis: Increasing
the Effectiveness of Natural Enemies, 11-15, November
1985, New Delhi, India.

Powell, J. E., and G. W. Elzen. 1989. Biological and
host relationships of Microplitis croceipes. Southwest.
Entomol. Suppl 12. pp. 11-16. Powell, J. E., and J. L.
Roberson. In Press. Status of rearing technology for
cotton insects. In E. G. King and J.R. Phillips [eds.],
Cotton Insects and Mites: Characterization and Manage¬
ment. The Cotton Foundation.

Powell, J. E., and W. P. Scott. 1985. Effects of insecti¬
cide residues on survival of Microplitis croceipes adults
(Hymenoptera: Braconidae) in cotton. Fla. Entomol. 68:
692-693.

Ridgway, R. L., E. G. King and J. L. Carillo. 1977.
Augmentation of natural enemies for control of plant pests
in the western hemisphere, pp. 379-428. In R. L. Ridgway
and S. B. Vinson [eds.], Biological Control by Augmenta¬
tion of Natural Enemies. Plenum Press, New York.

Robinson, R. R., J. H. Young and R. D. Morrison. 1972.
Strip-cropping effects on abundance of predatory and
harmful cotton insects in Oklahoma. Environ. Entomol. 1:
145-149.

Sabelis, M. W., and M. Dicke. 1985. Long-range
dispersal and searching behavior, pp. 141-160. In W.

Helle and M. W. Sabelis [eds.], Spider Mites: Their
Biology, Natural Enemies and Control. Elsevier,
Amsterdam.

Scales, A. L. 1973. Parasites of the tarnished plant bug in
the Mississippi Delta. Environ. Entomol. 2: 304-305.

Sillings, J. O., and D. B. Broersma. 1974. The parasites
of the tarnished plant bug Lygus lineolaris in Indiana. Proc.
North Cent. Branch Entomol. Soc. Amer. 29: 120-125.

Smith, C. N. (ed.). 1966. Insect Colonization and Mass
Production. Academic Press, New York and London. 618

pp.

Steiner, W. W. M., and D. A. Teig. 1989. Microplitis
croceipes (Cresson): Genetic characterization and develop¬
ing insecticide resistant biotypes. Southwest. Entomol. 12:
71-87.

Stem, V. M., R. van den Bosch, T. F. Leigh, O. D.
McCutcheon, W. R. Sallee, E. E. Houston and M.

J. Garber. 1967. Lygus control by strip cutting alfalfa.
Univ. Calif. Agric. Ext. Serv. AXT-241.

Stewart, F. D. 1984. Mass rearing the pink bollworm,
Pectinophora gossypiella, pp. 176-187. In E. G. King and
N.C. Leppla [eds.]. Advances and Challenges in Insect
Rearing. USDA, Agric. Res. Serv., Washington D.C.

Stinner, R. E., R. L. Ridgway, J. R. Coppedge, R. K.
Morrison and W. A. Dickerson, J.R. 1974. Parasitism of
Heliothis eggs after field releases of Trichogramma
pretiosum in cotton. Environ. Entomol. 3: 497-500.

Suguiyama, L., and C. Osteen. 1988. The economic
importance of cotton insect pests on U.S. agriculture.
USDA, Econ. Res. Serv.

Thompson, S. N. 1986. Nutrition and in vitro culture of
insect parasitoids. Ann. Rev. Entomol. 31: 197-219.

van den Bosch, R., T. F. Leigh, D. Gonzalez and R. E.
Stinner. 1969. Cage studies on predators of the bollworm
in cotton. J. Econ. Entomol. 62: 1486-1489.

59

I

Vanderzant, E. S. 1973. Improvements in the rearing diet
for Chrysopa carnea and the amino acid requirements for
growth. J. Econ. Entomol. 66: 336-338.

Van Lenteren, J. C. 1986. Parasitoids in the greenhouse:
Successes with seasonal inoculative release systems, pp.
341-374. In Waage, J., and D. Greathead [eds.], Insect
Parasitoids: 13th Symposium of the Royal Entomological
Society of London. Academic Press.

Vinson, S. B., and D. L. Dahlman. 1989. Physiological
relationship between braconid endoparasites and their hosts:
the Microplitis croceipes - Heliothis spp. system. South¬
west. Entomol. 12: 17-37.

Zaki, F. N. 1985. Reactions of the egg parasitoid
Trichogramma evanescens to certain insect sex phero¬
mones. Z. Angew. Entomol. 99: 448-453.

Zhi-xin, Wu and Qin Junde. 1984. Culturing
Trichogramma dendrolimi in vitro with media devoid of
insect materials, pp. 431-435. In Adkisson, P.L., and Ma
Shijun [eds.], The proceedings of the Chinese Academy of
Sciences-United States National Academy of Sciences Joint
Symposium on Biological Control of Insects, September
25-28, 1982, Beijing, China. Science Press, Beijing,
China.

Table 1 . Biological control of H. zea and H. virescens in cotton by augmentative releases of Trichogramma

Release

Rate/ha
(x 1,000)

%

Parasitism

Control

Evidence

Reference

46-957

33-81

66-80% larval reduction

Stinner et al. 1974

112-178

55-84

21% larval reduction

Jones et al. 1979

110

15-90

larval reduction

Abies et al. 1979

47

42-80

inadequate suppression

Luttrell et al. 1980

413

28-60

3 1 % increased yield

King et al. 1985b

Table 2. Augmentative release of Chrysoper/a carnea for biological control of H. zea and H. virescens in

cotton fields

Release rate/ha

% Larval

(x 1,000)

reduction

494

82

227

89

25

33

74

54

247

83

(Ridgway, King and Carrillo 1977)

60

Integration of Techniques and
Methods of Cotton Protection from
Pests in Tajikistan

Vyacheslav G. Kovalenkov1

Young cotton is infested by a complex of insect pests and
mites which develop for 3-24 generations and feed on
plants during the entire growing season until harvest.
Potential losses (without control measures) can constitute
60% of the yield and prevent the need for cultivation.
Hence, plant protection is an important component of
cotton growing technology.

A historical review of cotton improvement systems reveals
several stages. The first and most difficult stage was
characterized by transition from total aerial treatments with
insectoacaricides in 1950-1960 to mainly ground applica¬
tions based on damage thresholds of phytophages. Treat¬
ment thresholds include 10-12 Helicoverpa armigera
larvae/ 100 plants of average-staple cotton, 3-5 pests/ 100
plants of fine-staple cotton, 0.3-0. 5 of Agrotis segetum
larvae/ 1 m2, 10% cotton infestation with aphids and 5%
infestation with Spodoptera exiqua Hb.

Limited use of chemicals was also conditioned by necessity
to reduce resistance of chewing pests to chlorinated
hydrocarbons and sucking pests to organophosphates. By
the end of 1960, pest resistance to these preparations
became a serious problem: every treatment stimulated pest
reproduction and killed natural enemies (parasitoids and
predators) resulting in losses in yield. The use of new
chemicals and their rotation according to damage thresholds
reduced pesticide effect on plants. It decreased insecticide
treatments 2.5 times and acaricide treatments 3.5 times
from 1967 to 1976.

To strengthen this tendency, special attention was paid to
non-chemical means of pest suppression. Cultural practices
which either suppressed or stimulated development of
noxious insects and mites have been evaluated. Regimes
for fertilization and irrigation were found to play key roles
in irrigation farming. Studies carried out in Guissar Valley
revealed relationships of aphid number and the quantity of
nitrogenous fertilizers applied. The use of 400-500 kg of
fertilizers/ha (instead of recommended 200-250 kg/ga)
allowed cotton aphid density to be maintained in the range
of 800-4,000 aphids/plant. Three to four insecticidal
treatments did not reduce the resistance development.

Proper use of nitrogenous fertilizers and

1 All-Union Institute of Biological Methods in Plant Protection, Kishinev,

former U.S.S.R.

elimination of weeds surrounding cotton fields maintained
aphid density at the level of 47-296 pests/plant using only
one chemical treatment.

Proper application of pesticides promotes reproduction and
accumulation of natural enemies by stimulating develop¬
ment of natural insect populations. According to the data
obtained by M. N. Narzikulov and Sh. A. Umarov (1961),
226 natural enemies including 192 parasites and predators
(84%) were recorded in cotton agrobiocenoses. Each of
them plays a certain role in phytophage management. Their
total ability to protect cotton without chemicals was named
by the authors as the level of entomophage effectiveness.

To provide reliable pest control it is necessary to have 250-
300 entomophages/100 cotton plants. These figures differ
when dealing with different pests: e.g., 150-200 against
Helicoverpa armigera, and Tetranychus urticae but only
100-150 against aphids.

Since predators and parasitoids regulate their numbers and
developmental tempos depend on pest density. Low pest
density can fail in attracting entomophages; a high one can
result in yield loss. To guarantee the yield, we consider it
reasonable to use the concept of "minimum necessary pest
number. " This is the ecological criterion reflecting the
number of insect pests and mites needed to guarantee
establishment and accumulation of natural entomophages.
This pest number does not endanger the crop yield. For
example, the first Coccinellidae were recorded on cotton
after infestation by aphids of 5-7 % of plants; mass flight -
at 12-18% with the density of 10 and more aphids/leaf.

The following procedures increase activity of natural
entomophages: minimum soil treatment, adherence to
damage thresholds, strict focus on treatments with
insectoacaricides (preferably of selective activity), and
increased alfalfa acreage (especially for seeds). The latter
is of special importance since alfalfa unplowed for 3 years
and untreated with chemicals serves as a place for accumu¬
lation and overwintering of entomophages. Strip mowing
improves ecological functions of this crop.

Artificial rearing and mass applications of entomophages
are very important for intensive cotton growing and
represent the second component of biocontrol which is
based on the knowledge that:

61

• multifactor mechanisms of agroecosystem self-regula¬
tion where (difficult for management) forces of nature
predominate and are unreliable, and

* there is a necessity tQ change ratio of beneficial and
noxious species by releasing laboratory reared parasi-
toids and predators.

In many cases, according to K. E. Voronin (1984),
entomophages released together with other components
regulate seasonal pest numbers and provide an ecological
balance sooner than under natural conditions.

Trichogramma pintoi and Bracon hebetor are used on
cotton in conjunction with parameters developed at the
Plant Protection Department of Tajik Institute of Plant
Protection. They are produced by 3 1 biolaboratories of the
republic and are released at an optimum rate of 60,000
Trichogramma females/ha at 20-25 Helicoverpa armigera
eggs/ 100 plants assuming 50% hatch. Such a density
corresponds to 10-12 larvae/100 plants which is permis¬
sible. On hatch of 60% or more larvae, 70,000-75,000
Trichogramma are released when 15 pest eggs/ 100 plants
are found. Similar to natural entomophages, 15
Helicoverpa armigera eggs are considered the maximum
necessary rate; complete dispersal of preimaginal stage
guarantees success. One release of Trichogramma elimi¬
nates an average of 50% of host eggs. To provide control
of egg laying of one generation of Helicoverpa armigera,
"flood" with Trichogramma by releasing 3-4 times every 4-
5 days.

Unlike Trichogramma, Bracon released into the field
establishes itself quickly by feeding on flower pollen and
preying on hemolymph: Bracon completes 2-5 generations
during development of one Helicoverpa armigera genera¬
tion. The optimal rate of 800-1,000 parasitoids/ha provides
80% mortality of Helicoverpa armigera larvae. Bracon is
very susceptible to changes in host number; i.e., when plots
vary in host density, the aggregate in areas where there are
at least 5-7 larvae/100 plants; this host density is the signal
for the parasitoid release.

Bracon effectiveness depends on several factors, including
what crop borders cotton. Thus, if Bracon is released in a
tomato field surrounded by cotton, the parasitoid remains
on tomatoes, the more attractive crop, and the biocontrol
effect on cotton is the biggest. If it is surrounded by

vineyards, blossoming com or orchards, less migration
from tomatoes occurs, resulting in less effectiveness. These
differences must be taken into account in making Bracon
releases.

Parasitization of 15-20% of Helicoverpa armigera larvae by
naturally Bracon will require in a release of 800 parasi¬
toids/ha. The rate is higher (1,000 parasitoids) at 8-12% of
parasitized larvae. When cotton borders tomatoes or
vineyards, do not release less than 1 ,500 parasitoids. Rates
must vary depending on the phenological state of the crop,
e.g., in bloom stage, Bracon is retained in the field by the
host and by carbohydrate feeding.

Commercial preparations of Bacillus thuringiensis (the third
biocontrol component) effectively supplements
entomophages. The use of dendrobacillin and bitoxibacillin
can eliminate 82% of Helicoverpa armigera. Subsequently,
the quantity of eggs laid (the majority of which are sterile)
by the 22-37.9% of deformed adults surviving after
treatments of pupated larvae is reduced 2-5 times. Experi¬
ments show that it is impossible to protect the crop using
one method or technique. Therefore, multiple components
may prevent yield losses in any situation in every ecological
zone. It is important to carry out control procedures during
vegetative periods during overwintering, spring emergence
of pests, and autumn formation of overwintering pest
reserves.

The complex of introduced cultural practices (soil prepara¬
tion, weed elimination, and treatment of mulberry trees
surrounding cotton fields) results in 27-69% reduction of
the overwintering pests. Trichogramma are released twice
in April and May when noctuid pests emerge and lay eggs
onto weeds, young plants of com, sorghum, tomatoes and
alfalfa. Up to four Trichogramma releases are recom¬
mended against the first generation of Helicoverpa
armigera in all zones. High temperatures (32 °C) during
development of the second generation prevent
Trichogramma from manifesting itself properly. Thus, one
treatment with one of the insecticides (sevin, fosolon,
thiodan) is carried out in southern regions of Tajikistan (the
zone of the most active pest development), while in
northern and central regions, Bracon are released. The
development of the third generation is controlled by
dendrobacillin in the south and by 3 Trichogramma releases
in other zones of the republic. Bacterial preparations are
not applied in the zones where Bracon is released.

62

During August or September (at the beginning of cotton
harvesting), it is necessary to start the control of autumn
formation of overwintering pest reserves. At this time,
adults continue oviposition while Helicoverpa armigera
larvae of the last generation feed and pupate on the late
sown and resown fields in plots either with prolonged
vegetation period (due to excess of nitrogen fertilizers or
close subsoil waters) or neighboring attractive crops (com,
vegetables). In autumn of 1980-1986 in Guissar Valley,
Trichogramma releases resulted in 28-47 % reduction of
pest eggs, Bracon releases resulted in 32-78% decrease of
pest larvae, and dendrobacillin killed 48-76.8% of pests.
Bracon released in the summer maintained activity in the
autumn producing mortality of pest larvae.

Fields receiving few chemical treatments attract more
natural parasitoids and predators which create natural
barriers while suppressing the reproduction of aphids,
spider mites and bugs. Entomophages are not allowed to be
used after Bi-58, antio, and fosolon (preparations applied to
control aphids and thrips), thus inhibiting pest resistance
development. Continuous use of organophosphate insecti¬
cides for 30 years has resulted in resistance of aphids,
thrips and phytophagous bugs with poor control of
Helicoverpa armigera and Tetranychus urticae as well.

In 1985 (under conditions of the melon aphid outbreak), we
managed to follow dynamic changes in resistance level
depending on the number of treatments with the organo-
phosphorus preparation, Bi-58. We established that with
every subsequent application of the chemical, loss of aphid
susceptibility occurred. In the Guissar region resistance
increased 16 (LD50 - 0.003 g), 34.5 (0.0069 g), 48
(0.0096 g), and 56 (0.011 g) after 1, 2, 3 and 4 treatments,
respectively. In the Tursunzadev region it increased from
94 (0.0039 g) to 94 (0.0186 g) resulting in a decrease in
pest mortality decrease from 72.8% to 23.6% in Guissar
region and from 71.4% to 14.8% in Tursunzadev region.

At the same time, pest numbers increased 4.3-11.5 times
and 6.1-17.6 times, respectively. A similar situation
occurs in the cotton aphid. Resistance levels were as
follows after unreplaced continuous use of Bi-58 and
fosolon: from 83 (0.0083) to 92.3 (0.0018) in Guissar
region; in Tursunzadev region it constituted 180 (0.018)
and 245 (0.0049). Treatments with Bi-58 modelled the
situation when the insecticide lost its decreasing yield.
Under these conditions, farms were forced to use
sumicidin, which reduces aphid numbers by 93.2-98.4% in
5 days after one treatment.

Excess use of Bi-58 against cotton aphid promotes resis¬
tance development in the spider mite. Phytophagous bugs
began to develop intensively (in July and August - 450-600
bugs/100 plants). Replacement of Bi-58 with sumicidin
resulted in the outbreak of the spider mite. In 5, 10, and
15 days after treatment (0.5 kg/ha) pest density increased
2. 5-3. 8 times, 6.3-11.2 times, and 8.1-14.1 times respec¬
tively. This phenomenon cannot be explained by the
elimination of entomoacariphages alone. Our experiments
show that sumicidin stimulates development of the spider
mite creating conditions for formation of 2.5-3 generations
more than application of organophosphorus materials. The
negative role of chemicals is manifested most of all when
choice of pesticides is limited. The following scheme was
found to be efficient against aphids:

• weed elimination around fields before cotton sprouting,

• one treatment of marginal stripe of the field with either
fosolon or Bi-58 at 10-15% of infested plants, and

• treatment of appearing foci with sumicidin in 20-25 % of
damaged plants.

Fields with weakened chemical pressure are more eagerly
colonized with entomophages (780-920 entomophages/ 100
plants in the experiments made), which create natural
barriers suppressing aphid development. It resulted in an
81-86% increase of Bi-58 effectiveness and 79.7-83.7%
rise of that of fosolon. Pest resistance decreased to
8.2(0.00164)-22(0.0044) level.

Application of selective acaricides can control spider mite
development and neutralize negative aftereffects of
sumicidin. The data obtained show that limited pesticide
choice results in the reduction of effectiveness of materials
used, and pest outbreaks become inevitable and are then
treated again, thereby strengthening negative aftereffects.
Rotation of different classes of chemicals increases their
effectiveness, and the intervals between treatments may
become longer. Foci application of pesticides provides a
cleaner environment, more attraction to parasitoids and
predators, and formation of the dynamic balance needed for
noxious and beneficial species. The number of the latter
increases so that it becomes capable of establishing long
term control of pest development. Pesticide replacement by
Trichogramma, Bracon and microbiopreparations will
increase natural control, and this is the ecological essence
of tactics tested in Tajikistan for overcoming negative
results of chemicals use and their integration with
biocontrol means. Hence, widening of pesticide assortment

63

is considered necessary and is the key to optimization of
chemical pressure in the cotton field. The scheme ap¬
pended (Kovalenkov 1990) illustrates aftereffects of two
approaches: limited and excess use of pesticides.

The area treated is important in stabilizing the effect of
biocontrol. In an area more than 100 ha where
entomophages are released and interact with natural ones,
they exercise more rapid control of the pest, thereby
affecting both current and subsequent generations. Mortal¬
ity of Helicoverpa armigera progresses, and pest numbers
become economically unimportant. If biocontrol agents are
used only in 1-2 fields, there are considerable fluctuations
in parasitism parameters because of migration of pests from
other fields. In data obtained from the center of a field
with the area of 280-690 ha, the effectiveness of
entomophages, Trichogramma and Bracon, was 11-20%
and 15-26% respectively -- higher than in the field borders.
Hence, the bigger the area under protection colonized with
entomophages, the higher and more stable parasitism
parameters are, and the more guaranteed the formation of
the agrocoenosis is with a balanced ratio of noxious and
beneficial pests.

The difference between upper and lower effectiveness
thresholds is minimal if Trichogramma and Bracon disperse
simultaneously on several bordering crops thereby improv¬
ing activity of entomophages. Two to three year integra¬
tion of biological and chemical methods is necessary for
transition from chemical to biological control. This has
been observed on cotton, com and tomatoes.

Chances of higher yield are greater if IPM is carried out on
all crops of cotton rotation including com, alfalfa, veg¬
etables. These other crops are reservoirs for polyphagous
pests which accumulate and (under excess density or at the
end of vegetation period) migrate to cotton creating
additional damage. For example, com sown for grain and
attracting Helicoverpa armigera is recommended to be
treated with dendrobacillin at the silk stage during mass
egg laying and larva emergence. When com is grown for
silage, it should be harvested before pupation of larvae.

Late protection measures result in reduced pest develop¬
ment.

On tomatoes, Helicoverpa armigera has one more genera¬
tion and 5-10 times more numerous than on cotton. When
250-300 larvae are present/ 100 plants, expect migration of
subsequent adults to cotton. Thus, it is very important to

control this pest in tomatoes. This problem is successfully
solved by release of Trichogramma and Bracon. Such
tactics prevent damage to cotton and produce 13-16
centners of com grain/ha and 30-100 centners of tomatoes/
ha higher yield than chemical treatments.

Pest control effectiveness is affected by time of application
of chemical or biological means. Up-to-date control tactics
require adequate sampling techniques and determination of
optimal time for application. To meet these requirements
studies were made of sex pheromones of noctuid pests, and
technologies for their application were developed. Predic¬
tion of time of chemical application with pheromones
reduced chemical treatment numbers by 0.5-1. In 1988, the
use of pheromone traps in the area of 97,000 ha reduced
chemical costs 1-6 million roubles.

Variation of application time for natural enemies, and their
proper selection and rational combination constitute up-to-
date scientifically-based integration of knowledge. If
chemical treatments are used, they must be carried out to
decrease maximum pest numbers as well as preserve
entomophages. Different procedures and combinations add
to our knowledge of cotton production. Reliable sampling
data by hand and with pheromone traps helps decision
making on which chemicals to use and the sequence of their
use. This tactic determines the degree of biocontrol
effectiveness, making IPM flexible, multivariate and
capable of increasing yield under any situation in every
ecological zone.

Proper selection and rational combinations for protection of
crops in cotton have been widely adopted in state and
collective farms. In recent years it has become possible to
prevent 10-15 centners of cotton losses/ha in the republic
and to reduce chemical treatments to 0. 5-0.7. Annually the
use of insectoacaricides is abolished in the area of 120-180
thousand hectares. The area for release of laboratory reared
entomophages increased from 5,000 ha in 1980 to 1 14.000
ha in 1989. The environment has considerably improved,
and, in the process, has allowed the development of
apiculture.

64

References Cited

Narzikulov, M. N., and Sh. A. Umarov 1981.

Kovalenkov, V. G. 1990. Primenenie biologicheskogo
metoda zaschiti ras tenii v Tajikistane. Obzomaya
informatsia. (In Russian). Dushanbe, Gosplan of Tajik

SSR, Tajik INTI, 76.

1

1

Entomophagi kak component khlopkovogo agrobiocenosa.
(In Russian). In Integrirovannaja zaschita khlopchatnika ot
vreditelei, Dushanbe, donish, 33-42.

Voronin, K. E. 1984. Nasekomie-entomophagi v
integrirovannoi zaschite rastenii. (In Russian). In

Nauchnie osnovi zaschiti rastenii. M., 152-174.

Table 1. The scheme of positive and negative aftereffects of limited and excess application of pesticides.

Limited application

Excess application

Preferential use of biocontrol agents,
local treatments with selective chemicals.

Excess chemical pressure.

Stimulation of the activity of natural
and colonized entomoacariphages.

Elimination of natural entomoacariphages.

Reduction of pest contacts with pesticides.

Increase of their susceptibility to toxicants.

Formation of resistant pest populations.

Controlled development of pests

in the range of permissible damage thresholds.

Management pest outbreaks.

Guaranteed yield.

Yield losses.

65

Genetic Control of Cotton Insects
the Pink Bollworm as a Working
Program

R. T. Staten1

Introduction

The use of genetic control mechanisms for cotton insects
has been explored to varying degrees with several insects
including the tobacco budworm, Heliothis virescens (F.),
and the boll weevil, Anthonomus grandis Boheman. Its use
and actual application on a large scale has been limited,
however, to the pink bollworm, Pectinophora gossypiella
(Saunders). This report will cover that usage and its
potential for expansion.

It is important to have a historical perspective of the Pink
Bollworm problem in the U.S. At the present time it is a
western United States problem restricted to the irrigated
desert areas. Pink bollworms are occasionally recovered
from the south, east of Texas with only one endemic
population on wild cotton in the southern extreme of
Florida. The Floridian population has total geographical
isolation from any commercial cotton growing area in the
U.S.

The pink bollworm is most destructive in Arizona, southern
California, and the northwestern Mexican desert cotton
growing areas. (Figure 1). Its introduction and establish¬
ment in central Arizona in the mid 1950s and into the
Colorado River Basin of western Arizona, southern
California, and northwestern Mexico in 1965, has led to the
development of a major pest problem with extreme side
effects on the whole ecosystem. This is, of course,
exacerbated by a long growing season which when pushed
for maximum production results in tremendous population
potential. The extensive use of broad spectrum conven¬
tional pesticides including the newer synthetic pyrethroid
compounds has resulted in expensive control systems with
secondary pest problems. The only major area in the
western cotton belt where the impact of the pink bollworm
has not been felt by the grower is in the central •California
San Joaquin Valley. In this area the establishment of pink
bollworm has been prevented through the use of sterile
insect technology and adequate cultural control. This paper
will briefly outline the details and provide program statis¬
tics for that project. In addition, it will discuss efforts to
integrate sterile release technology with pheromone systems
and optimum cultural practices in the heavily infested
Southern California desert.

'Science & Technology, USDA, APHIS, Phoenix, Arizona

The San Joaquin Exclusion Program

The San Joaquin Valley has never had a known continuous
endemic population of the pink bollworm even after the
expansion of pink bollworm into the Colorado River desert
growing areas in 1965 and 1966. The San Joaquin Valley
is separated by approximately 175 miles (direct line) from
any generally infested areas of southern California. The
entire 175 miles is a host free zone. This expanse of desert
and mountains has formed a partial barrier to pink boll¬
worm movement. Movement within the desert region
between the generally infested southern California deserts
and the San Joaquin Valley has been documented by Stem
et al. (1978) and by trap lines maintained by program
personnel in the mountain passes between the desert cotton
growing areas and the San Joaquin Valley. Extensive
movement into the desert area has been documented early in
the spring, late in the summer, and throughout the fall.
Movement over mountain passes into the adjacent San
Joaquin Valley has been documented in the fall. Program
trap data further supports evidence of movement within the
San Joaquin Valley since late season increases in detection
have coincided with captures in the desert trap lines. All of
this data has been gathered by using gossyplure baited traps
placed along highways in the deserts between cotton
growing areas.

Sterile insects are now reared in Phoenix, Arizona using
techniques or modifications of techniques described by
Stewart (1984). Attention to detail and constant refine¬
ments of these rearing processes since 1971 has constantly
improved the quantity and quality of insects available for
shipment.

The increase in quality is illustrated in Table 1, which
summarizes quality control mating tests by program
personnel. Twenty pairs of sterile males and females are
held in cages. Samples are taken each day of shipment (7
days a week) throughout season. Those females which
contain spermatophores are considered mated. Table 1
shows an increased mating propensity between 1985 and
1989 after small procedural improvements were made in
shipping.

Similar laboratory tests were used to check mortality over a
14 day period. Mortality of 8-10% was normal in 1985
and 1986. However, season-long moth longevity was
improved enough to reduce season-long mortality to 5 to
6% in 1989.

66

Production capability of the Phoenix facility is illustrated in
Table 2. Laboratory space was increased by approximately
15% during the 1970 - 1989 period. All other production
increases were technical.

The Phoenix USDA/Califomia Cotton Growers funded
facility was the only provider of sterile insects for the San
Joaquin Valley from 1970 through 1975 and 1983 through
1990. It has also provided significant numbers of pink
bollworm for research work including sterile insect re¬
search. These trials included those in the Moapa Valley of
Nevada (Staten, unpublished), St. Croix (Henneberry, et
al. 1985), and in the Coachella Valley of California.
Pheromone isolation and identification projects (Hummel,

1973 et al., Bierl et al. 1974) have also been supported.

From 1976 through 1982 a commercial "satellite" PBW
rearing facility was operated for 8 months of each year.

This facility provided moths for the San Joaquin Valley.

Sterile insects are shipped at 40 _+ 5 degrees fahrenheit
from Phoenix, Arizona to Bakersfield, California where
they are transferred (still in a chilled condition) to aircraft
and released at approximately 500 feet above the cotton
fields. The time of collection from eclosion chambers to
the time of release is from 24 to 48 hours. Insects are held
in an immobile state by chilling during this time.

The sterile and native populations are currently monitored
with pheromone traps baited with gossyplure in a controlled
release formulation. The Delta Trap (Foster et al. 1977)
has been used since 1974.

With the advent of gossyplure and its availability in 1974
the conduct of this program has become much more precise.
For that reason this document will discuss program opera¬
tions after that time. A summary of pertinent information
is shown in Table 3.

Numbers of traps have varied from 12,850 to 33,711.

Traps have been placed at average densities varying
between one trap per 39.5 acres to one trap per 79.8 acres.
Traps are serviced weekly. The data produced by these
traps is used to schedule the release of sterile insects. From

1974 to 1989, the area requiring release of sterile insects
has ranged from 4.6 percent to 29.5 percent of the total San
Joaquin cotton production area.

Each year the season starts with releases on cotton in
sections which had moth finds from the previous year.

Thus, the percent of the total acres released has varied. As
is appropriate (dependent upon native moth capture rates
the previous year), buffer zones are also released around
these sections. Moth releases are controlled at the section
level (640 acres). As new unmarked (native) pink boll-
worm are found, the releases are adjusted accordingly.

Moth release typically begins before the first flower buds
are present on cotton plants (pin square) and continue with
daily shipments of moths throughout the season. Releases
of sterile insects have been terminated typically each year
during the last two weeks of October or the first two weeks
of November.

A general overview of the most pertinent trap capture and
sterile release data is shown in Table 4. Major improve¬
ments in rearing technology and management systems in
1975 represented a significant improvement in program
operations. A very dramatic degree of improvement in
rearing is seen when one compares the numbers of moths
shipped to the San Joaquin in 1975 compared to 1974
(Table 4).

Shipments increased from 37.01 x 106 moths in 1974 to
150.11 x 106 in 1975, and 1976 at 173.6 x 106 in 1976.
Although native unmarked populations were low early in
1974, it became apparent that a major problem existed as
the season progressed. The total sterile moth captures of
282,897 was 647 times higher than the non sterile recovery,
but acceptable working ratios at the individual field level
could not be maintained when native populations became
very heavy in August and September. (Note: Through
experience, the ratio of 60 sterile moths to one native moth
has been established as a working, safe ratio.) It was very
fortunate that improvements were brought on line before
the 1975 season. The production in 1975 was able to cover
the entire infested area adequately.

Relatively high native moth capture (1,474) late in the
season in 1976 followed tropical storms off Baja California,
Mexico and resulted in a decision to augment sterile moth
production at the USDA facility in 1977 by outside contract
production. This contract provided 28 percent of the 412
million moths used in the San Joaquin in 1977 and provided
an average of 39 percent of all production until the 1983
season. In 1983, the 586 million moths shipped were
provided exclusively by the USDA facility in Phoenix.

67

In 1980 we captured the biggest number of native moths in
program history. Unlike 1974, however, coverage of the
entire area where moths were detected was achievable
throughout the season. Major efforts were also made
through grower meetings throughout this area to optimize
better cultural controls.

The following year 24% of the Valley received sterile
insects. This rate is high in comparison to the preceding
few years of release, but with good winter mortality
between the 1980 and 1981 season it was apparently
adequate. We have not had buildups of these levels again.

When one tries to asses this program with its massive
number of traps and extensive amounts of data they
generate, it is apparent any analysis would be highly
speculative in nature. There are no check plots and/or
adjacent identical infested areas.

It should also be noted that until the present time we have
used resources on an as available basis. We are now
constructing in APHIS, with cooperation from Texas A&M
University a heuristics-based expert system to help the
decision maker utilize the extensive data base that this
project generates. So, although the project has its limita¬
tions, two pertinent points have not been missed by the
cotton industry. The first is that there is not a detectable
ongoing pink bollworm population in the San Joaquin
Valley, and second that our rearing technology has im¬
proved significantly and consistently over the years.

The Integration of Sterile Insect System
With Pheromone Disruption

Once the pink bollworm spread into the Colorado River
Basin (Figure 1), the entire cotton pest management system
changed drastically. Cotton growers were determined to
continue maximizing yields and were quick to utilize
regularly scheduled insecticide application. This philoso¬
phy continued essentially unabated for approximately 20
years. In the most severely infested areas (such as the
Imperial Valley) grower leadership is now seeking other
remedies. In addition to working with sterile insect
technology, our laboratory, the Agricultural Research
Service, and industrial cooperators have worked extensively
for the development of pheromone communication disrup¬
tion. The large scale field testing of the Mitsubishi PBW
rope in 1985 (Staten et al. 1987) in the Imperial Valley

provided a breakthrough. This breakthrough led to the
conduct of an "area wide" management project in the
Coachella Valley on its restricted acreage.

The Coachella Valley is reasonably isolated during the main
growing season and measurably affected by migration only
during storm conditions that include surface winds from the
southeast, usually late in the season.

In 1986 we started working in a cooperative venture with
the growers of the area and their pest control advisor (M.
Grummet of Foster Gardner Chemical Co.). Our first step
in 1986 was to treat the entire valley with the high rate
PBW rope system, at 30 grams active ingredient (AI) per
acre, in one application, at the 6-8 leaf cotton development
stage. Our results were encouraging enough that in 1987-
1989 we started releases of sterile insects at 500+ sterile
pink bollworms per day. This was done in connection with
a minimum shortening of the growing season. Under this
system, sterile insect release started at the 3-4 leaf cotton
stage. The pheromone was only applied at the 6-8 leaf
cotton stage. If a ratio of 60: 1 sterile to native insects was
not achieved in a given field at a 6 leaf stage, that field
received a pheromone treatment. Thus, while not all fields
would receive pheromone, all fields received sterile insects.
At any time in the season that chemical control thresholds
were exceeded, the grower and his pest control advisor
were expected to utilize chemical control to complete the
growing season. This treatment was on a field by field
basis as needed.

Pink bollworm sterile and native populations were moni¬
tored with Delta traps at 2 traps per 40 acres and with boll
samples collected in each field (80 to 100 bolls per field
each week). Bolls were then returned to the laboratory for
examination under large lighted 3x magnification lenses.
Samples were taken from the second week in July through
August.

Management Trial Results: The overall treatment strate¬
gies, pheromone usage, and insecticide use patterns are
shown in accompanying Table 5.

Before 1986 the preponderance of the cotton growers in the
valley used expensive multiple chemical pesticide applica¬
tions for control of PBW even though in this valley other
crops such as grapes, dates, citrus, and vegetables were
considered more important and placed on the better land.
Cotton is usually grown on more marginal ground at as

68

low a cost as possible. In the most successful year, 1988,
no conventional organic insecticides were used. In this
year, pink boll worm movement from the Imperial Valley to
the Coachella Valley was not significant until the last weeks
of August as detected by traps in the desert between the two
valleys. In 1989 the Imperial Valley terminated its cotton
by cessation of all irrigation in early to mid August. All
fields were defoliated chemically by September 1, with 12
exceptions out of over 130 fields. Movement was very
apparent across the desert from the last weeks of July and
throughout August. Our control with minimal insecticide
in the Coachella Valley was still acceptable as illustrated by
both pesticide use patterns and by low larval populations as
shown in Table 6, which summarizes larval population
levels through August in all four years.

It should be noted that, although usage of conventional
insecticides was being restricted significantly, the overall
population levels were held at drastically reduced levels.
This is particularly true in comparison to the larval popula¬
tions in the Imperial and Parker Valleys in 1989. This may
be noted in comparing our data with that from cooperative
monitoring projects in Imperial (D. Weddle, C. C. Chu, R.
T. Staten, T. J. Henneberry, and S. Birdsall, unpublished)
and a Parker, Arizona project (R. T. Staten, L. Antilla, F.
Myers, and R. Crout, unpublished) Table 7.

The population of larvae in Table 5 are from boll samples
taken season long in 45 fields in Parker, Arizona, and 55
fields in the Imperial Valley, California. All fields were
under conventional insecticide treatment.

Summary - Future Prospects

The twenty three year history of the San Joaquin Valley
exclusion project is considered by the California cotton
industry as a major success. The industry funds their
program at better than 85 % of its total cost, and has opted
to spend 2. 1 million dollars for two new rearing buildings
with options for major expansion to continue this program.

In addition, these same growers are in pursuit of an
expanded trial in the Imperial Valley. The Imperial Valley
is ideal as a target to expand these techniques because it has
a much larger acreage than Coachella and is, as a cotton
growing area, under extreme duress. Normal cotton
production of 40-80,000 acres has been the rule in the past
years. Cotton production has slipped to approximately

10,000 acres largely due to the extreme cost of treating
pink bollworm infestations. In addition, the growers in
that community have opted to drastically shorten the
growing season with marked impact on pink bollworm
population. Our laboratory, in cooperation with the
Imperial County Commissioner of Agriculture, and
Agricultural Research Service, monitored the pink boll¬
worm population extensively in 1989 before any winter
cultural program could affect the population of that year.
Populations were thus mentioned in an identical fashion in
the season of 1990, following a mandatory fall termination
in 1989. Populations have been drastically reduced as
evidenced by adult moth counts in delta traps and boll
infestation data. In addition major test releases this year
clearly indicate a 500 sterile moth per acre per day release
is feasible. I expect that by next year an even lower release
rate would be effective at achieving a 60: 1 ratio. My
reason for this optimism is found in a comparison of boll
sample data collected from boll incubation boxes. In 1989
versus this year, 1990, drastically lower populations have
already been noted. Our data is far from complete. Table
8 contains a summary of July's data.

This data was taken largely from a non-sterile release zone.
As these population trends advance, I expect an even lower
overwintering survival for the 1991 season. Sterile releases
will become more economically feasible.

The importance of these approaches and the necessity of
finding a better management scenario are imperative in
today's ecologically concerned society. We can no longer
live with major insecticide usage for pink bollworm and its
associated secondary pest problems. In addition, we can
not keep ahead of insecticide resistance for key pests such
as the pink bollworm and the white fly species now found
in cotton. We must reduce environmental hazards from
pesticide use, including drift and residue. The pink
bollworm problem with today's knowledge can only be
dealt with in an organized area wide rational basis.

69

References Cited

Bierl, B. A., M. Beroza, R. T. Staten, P. E. Sonnet, and
V. E. Adler. 1974. The Pink Bollworm Sex Attractant. J.
Econ. Entomol. 67:211-216.

Foster, R. N., R. T. Staten and E. Miller. 1977. Evalua¬
tion of Traps for Pink Bollworm. J. Econ. Entomol. 70:
289-291.

Henneberry, T. J. and D. F. Keaveny. 1985. Suppression
of Pink Bollworm By Sterile Moth Releases. U.S. Depart¬
ment of Agricultural Research Service. 74 pp.

Hummel, H. E., L. K. Gaslon, H. H. Shorey, R. S. Kane,
K. S. Byme and R. M. Silverstein. 1973. Clarification of
the chemical status of the pink bollworm sex pheromones.
Science 181: 873-875.

Staten, R. T., H. M. Flint, R. C. Weddle, E. Quintero, R.
E. Zarte, C. M. Finnell, M. Hemandes and A. Yamamoto.
1987. Pink Bollworm (Lepidoptera: Gelechiidae): Large
scale field trials with a high rate gossyplure formulation. J.
Econ. Entomol. 80: 1267-1271.

Stem, V., and V. Sevacherian. 1978. Long range dis¬
persal of pink bollworm into the San Joaquin Valley.
California Agriculture 32: 4-5.

Stewart, F. D. 1984. Mass Rearing the Pink Bollworm.
Pectinophora gossypiella, pp. 176-178. In Advances and
challenges in insect rearing. E. G. King and N. C. Leppla
[eds.]. U.S. Department of Agriculture, Agricultural
Research Service. 306 pp.

70

Table 1 . Annual Summary of Mating propensity of sterile Pink Bollworm females shipped
to the San Joaquin as determined in the laboratory for 1985 through 1989

Year

% Mated

1985

43

1986

70

1987

81

1988

76

1989

86

Table 2. Annual Production of the Phoenix Pink Bollworm Rearing Facility. 1970 through 1989

Year

Total Moths x 10®

1970

135

1971

120

1972

113

1973

99

1974

41

1975

154

1976

177

1977

192

1978

301

1979

385

1980

372

1981

401

1982

497

1983

594

1984

579

1985

489

1986

702

1987

735

1988

804

1989

826

71

Table 3. General Statistics for the San Joaquin program for 1974 through 1987

Year

no. acres no. traps

mapped xIO3 maintained xIO3

acres per maximum acres percent acres
trapxlO3 released xIO3 released

mean no.
steriles per acre

1974

1,275

32.0

39.5

101

7.9

366

1975

866

22.0

40.0

117

13.1

1,287

1976

1,130

27.0

41.0

90

7.9

2,141

1977

1,353

33.7

40.0

400

29.5

1,030

1978

1,501

19.6

76.3

355

26.6

1,286

1979

1,625

20.6

78.8

254

15.6

2,512

1980

1,463

20.8

70.2

177

12.1

2,880

1981

1,451

19.4

74.8

352

24.3

2,251

1982

1,326

16.9

78.3

147

11.0

5,260

1983

1,005

12.9

78.2

132

13.1

4,452

1984

1,392

17.6

79.0

42

6.1

13,329

1985

1,336

17.2

77.9

59

4.6

8,167

1986

1,061

13.3

79.8

44

4.1

14,930

1987

1,368

22.7

60.4

71

5.2

10,655

1989

1,128

19.2

58.8

104

9.2

7,258

Table 4.

Sterile Moth Release and Total Moth Capture Data in the San Joaquin Valley

Year

Sterile Moths
Released xIO6

Percent Sterile
Recovered

(Native) Sterile
Moths Recovered

Non-Sterile Sterile Insects

Recovery (Native) Captured per Native

1974

37.01

0.76

282,897

437

647.0

1975

150.11

1.01

1,528,260

245

623.8

1976

173.64

0.63

1,229,742

1,474

834.0

1977

412.17

0.40

1,677,900

7,402 1

22.0

1978

455.99

0.09

429,063

69

621.8

1979

636.98

0.09

545,295

754

723.0

1980

510.49

0.11

566,170

4,492 1

126.0

1981

794.26

0.11

864,861

677

127.7

1982

772.12

0.13

1,041,280

120

8,677.3

1983

586.77

0.18

1,057,735

863 '

1,225.6

1984

571.50

0.38

2,144,018

351

6,108.3

1985

483.74

0.09

434,739

160

2,717.1

1986

660.35

0.05

352,825

62

5,690.7

1987

670.07

0.15

1,057,925

294

3,598.7

1988

754.87

0.16

1,188,335

891

1,335.7

1 Finds of 4, 9, and 4 larvae in respective years

72

Table 5. Control strategies and conventional insecticide use patterns for pre-management and area-wide
management systems in the Coachella Valley.

Treatment System

Fields Treated
with Pheromone

x No. Conv.
Insecticides

No. Fields Treated
with Insecticides

Cumulative Conv.
Insect. Treatments

1985

Insecticide only

(Pre-trial data)

7.27

56/57

414

1986

Pheromone and

Insecticide

31/31

1.80

17/31

56

1987

Sterile Insects, Pheromone

and Insecticides

17/27

1.03

7/27

28

1988

Sterile Insects

and Pheromone

4/31

0.00 1

0/31

0

1989

Sterile Insects, Pheromone

and Insecticides

3/23

1.90

18/23

44

1 Most importantly, we have achieved major reductions in conventional pesticide usage in the four years that we have operated this program with
combined pheromone and sterile insect strategies as shown in table 5.

Table 6. Summary Boll Data - Coachella Valley Pink Bollworm Management Study Showing Total Larvae per
Sample and Number of Bolls Sampled

Week

1986

1987

1988

1989

No. Bolls
Sampled

Total

Larvae

No. Bolls
Sampled

Total

Larvae

No. Bolls
Sampled

Total

Larvae

No. Bolls
Sampled

Total

Larvae

1

50

3

370

0

2

500

3

500

0

726

0

560

0

3

960

2

1,700

0

1,200

0

720

0

4

1,180

9

2,400

7

1,600

1

960

0

5

1,520

1

2,400

33

2,000

0

1,520

0

6

1,520

3

2,500

34

2,160

0

1,600

1

7

1,600

0

2,700

11

2,320

0

1,760

0

8

1,520

1

2,700

6

2,480

0

1,564

18

9

1,600

1

2,700

5

2,480

5

1,860

39

10

1,760

43

1,350

25

2,480

1

1,960

39

11

2,200

138

1,350

41

2,480

0

1,240

47

73

Table 7. Data from Boll Surveys Selected Randomly from Fields in Imperial Valley, California,
and Parker, Arizona, and all Coachella Fields in 1989

Date

Number of Larvae per 100 Bolls

Imperial Valley1

Coachella

Parker

7/01 - 7/05

_

0.00

_

7/06 - 7/14

7.84

0.00

....

7/17-7/21

8.84

0.00

....

7/24 - 7/28

9.50

0.06

3.16

7/31 - 8/04

11.40

0.00

5.40

8/07 - 8/11

15.34

1.15

11.05

8/14 - 8/18

25.10

1.99

16.20

8/21 - 8/25

30.70

1.98

18.86

8/28 - 9/01

34.34

3.80

21.89

9/03 - 9/06

— —

....

23.02

9/10-9/13

— —

....

16.98

9/24 - 9/27

— —

....

14.15

1 Data adjusted from Boll Box Survey to equate with standard cutting procedures (Henneberry et al. 1985).

Table 8. Comparative Larval Infestation in the Imperial Valley before and after reduction in
season length, expressed as larva per 100 bolls.

1989

1990

July 6 - July 14

15.6

0.67

July 17 - July 21

17.6

0.37

July 24 - July 28

19.0

2.46

July 30 - August 4

22.8

3.20

74

Fig. 1. Cotton growing areas of the western United States (Arizona and California) and northwestern Mexico

Cultural Control of Cotton Insect
Pests, with Emphasis on an Emerging
Program of Cultural Controls in the
Lower Rio Grande Valley of Texas

Kenneth R. Summy and Edgar G. King1

Abstract

During the past two decades, cotton production in the
United States has largely shifted from production systems
heavily dependent on Broad-spectrum pesticides to systems
utilizing an array of cultural and biological controls.
Cultural controls embody an array of potential pest-control
tactics, ranging from initial cultivar selection to a sequence
of agronomic practices which commence prior to planting
and terminate following harvest. Many of these strategies
are singularly effective against one or more cotton insect
pests, and may become particularly potent when used in
conjunction with other cultural practices in an organized
community-wide pest management effort. The principal
cultural practices employed against cotton insect pests in
the United States are discussed, and an emerging program
of cultural controls in the Lower Rio Grande Valley of
Texas (U.S.A.) is summarized.

Introduction

Cotton production in the United States has traditionally
been plagued by serious losses to insect pests, the most
important of which are the boll weevil, Anthonomus
grandis Boheman, pink bollworm, Pectinophora
gossypiella (Saunders), bollworm, Helicoverpa
( = Heliothis) zea (Boddie), tobacco budworm, Heliothis
virescens (Fabricius), sweetpotato whitefly, Bemisia tabaci
(Gennadius) and a complex of plant bugs which includes
the cotton fleahopper, Pseudatomoscelis seriatus (Reuter),
tarnished plant bug, Lygus lineolaris (Palisot de Beauvois),
and L. hesperus Knight. During 1974-1976, an estimated
201.6 million bales were destroyed by boll weevil alone,
which represented a total annual loss of ca. $145.5 million
(DeBord 1977, Schwartz 1983). During 1989, arthropods
accounted for a loss of ca. 1.1 million bales and a total
expenditure of ca. $686 million (Herzog and King 1990).

Newsom and Brazzel (1968) partitioned the history of
cotton pest control in the United States into two distinct
periods: 1) prior to the invasion of boll weevil (ca. 1892),
in which losses to insect pests were generally moderate and

emphasis was placed on cultural and biological controls,
and 2) thereafter, in which losses to insect pests became
substantial and chemical control became predominant. The
chemical control era was further separated into four distinct
periods characterized by 1) a continuing emphasis on
cultural and biological controls (1908-1923), 2) extensive
use of calcium arsenate (1924-1945), 3) dependency on
chlorinated hydrocarbon insecticides (1946-1955), and 4)
unilateral reliance on organophosphate insecticides (1956-
1968). The latter period culminated with the development
of organophosphate-resistant populations of tobacco
budworm, which precipitated a series of crop disasters and
shattered the illusion that cotton could be produced indefi¬
nitely in this manner (Adkisson 1971).

The concept of "integrated control" was originally defined
as the use of insecticides in a manner least detrimental to
biological control agents (Bartlett 1956). More recently,
this definition was broadened to the effect that "all neces¬
sary techniques are consolidated into a unified program, so
that populations can be managed in such a manner that
economic damage is avoided and adverse side effects are
minimized" (National Academy of Sciences 1969). These
necessary techniques include a broad range of cultural
controls, defined as the "purposeful manipulation of the
environment to make it less favorable, thereby exerting
economic control of the pests or at least reducing their
rates of increase and damage" (National Academy of
Sciences 1969).

Considerable emphasis has been placed on cultural control
of cotton insect pests during the past two decades (see
reviews in Noble 1969, Walker 1984, Bradley et al. 1986,
Luttrell et al. 1986, Rummel et al. 1986, Henneberry 1986,
Walker and Smith, in press). Cultural strategies encompass
a potential arsenal of tactics, ranging from the initial
cultivar selection to a sequence of agronomic practices
commencing prior to planting and terminating after harvest.
These methods, in conjunction with the fundamentals of
insect pest and beneficial insect sampling and economic
thresholds, provide a strong base for ecologically-oriented
pest management systems.

'Subtropical Agricultural Research Laboratory, USDA, ARS, Weslaco,
Texas

76

Components of Cultural Control

Cultivar Selection

Cotton breeders have expended considerable effort in
attempts to develop cultivars resistant to arthropod pests
and pathogens (see reviews by Jenkins 1982a, 1982b,
Jenkins et al. 1969, Maxwell et al. 1972, Schuster and
Frazier 1976, Jones et al. 1978, Lukefahr 1977, Schuster
1979, Niles 1980, Adkisson and Dyck 1980, Adkisson et
al. 1982, Wilson 1982, Namken et al. 1983, Walker 1984,
Bridge and McDonald 1987, El-Zik and Thaxton 1989,
Jenkins and Wilson, in press). These efforts have tradition¬
ally concentrated on two principal approaches: varietal
resistance and earliness.

Varietal Resistance

Painter (1951) defined the components of host plant
resistance as 1) antibiosis, i.e., factors which reduce pest
survival, prolong developmental time, or produce other
effects detrimental to the pest, 2) nonpreference, i.e.,
factors which tend to reduce the incidence of pest attack
relative to a more susceptible counterpart, and 3) tolerance,
i.e., factors which allow a cultivar attacked by the pest to
produce a yield greater than that of a less tolerant counter¬
part. A considerable number of resistance factors have
been identified and incorporated into experimental and
commercial cotton cultivars. Several of the important
insect-resistance characters in cotton have involved modifi¬
cation of leaf shape, structure or color. Pilosity (pubes¬
cence) confers resistance to boll weevil, pink boll worm and
plant bugs (Wannamaker 1957, Wilson and Wilson 1975,
Meredith and Schuster 1979), while glabrous (smooth leaf)
cottons exhibit variable degrees of resistance to bollworm,
pink bollworm and whiteflies (Lukefahr et al. 1971,
Schuster 1979, Butler and Wilson 1984, Butler and
Henneberry 1984, Butler et al. 1986, Berlinger 1986). The
okra-leaf and super okra-leaf traits facilitate penetration of
insecticides within plant canopies and appear to increase
resistance to pink bollworm and whiteflies (Jones et al.
1975, Berlinger 1986, Wilson 1987). Red pigment confers
resistance to bollworm (Bhardwa and Weaver 1983) and
appears to hinder detection of plants by boll weevil (Isely
1928); once detected, however, red cottons are as suscep¬
tible to boll weevil damage as green cottons (Hunter et al.
1965). Nectariless cotton (devoid of extrafloral nectaries)
has been shown to reduce damage by bollworm, pink
bollworm, cotton leaf perforator, Bucculatrix thurberiella

Busck, and plant bugs (Lukefahr and Rhyne 1960,

Lukefahr et al. 1971, Henneberry et al. 1977, Meredith et
al. 1973, Schuster et al. 1976, Bailey et al. 1980, Bailey
1982, Bailey et al. 1984). Several resistance characters
have involved modifications of the floral structures and
levels of secondary plant substances, i.e., allelochemicals.
Examples of the former include male sterility (reduction in
anthers) which confers resistance to boll weevil (Jones et al.

1978) , yellow or orange pollen mutants which exhibit
resistance to tobacco budworm (Bailey 1981, Hanney et al.

1979) , and frego bract, which appears to be one of the
more promising resistance characters for boll weevil
(Jenkins et al. 1969). Examples of the latter include high
levels of gossypol, which confers resistance to bollworm
and plant bugs (Bottger and Patana 1966, Lukefahr and
Houghtaling 1969, Schuster and Frazier 1976), and
condensed tannins, which confer resistance to bollworm
and spider mites (Schuster 1979, Lane and Schuster 1979).

One of the recurrent problems associated with plant
breeding has been the fact that resistance factors effective
against one pest species may increase susceptibility to
another pest or group of pest species. Schuster (1979)
summarized some of the well-documented examples of this
apparent antagonism: 1) pilosity, which confers resistance
to boll weevil, pink bollworm and plant bugs, increases
susceptibility to bollworm, 2) the glabrous trait, which
confers resistance to pink bollworm and bollworm, in¬
creases susceptibility to plant bugs (Meredith and Schuster
1979), and 3) frego bract, which reduces oviposition by
boll weevil, increases susceptibility to plant bugs (Scales
and Hacskaylo 1974, Jones 1972, Tingey et al. 1975).

This dilemma has been addressed by the concept of multi¬
adversity resistance (MAR), which involves the direct or
indirect selection for genes or traits that provide resistance
to a broad range of adversities, including insect pests and
pathogens (El-Zik and Thaxton 1989). An example of the
MAR approach would be the combination of frego bract
and the nectariless trait in a single cultivar to provide
concurrent resistance to boll weevil and plant bugs (Jones et
al. 1983).

A second potential problem associated with breeding pest-
resistant cultivars involves the possibility that certain
resistance factors might adversely impact populations of
beneficial predators and parasites. For example, nectariless
cottons have been shown to reduce densities of certain
predators which habitually feed on extrafloral nectaries,
e.g., Chrysoperla spp. (Schneider et al. 1986). Field

77

studies of natural enemy performance on resistant cultivars
have produced variable results: for example, Lingren et al.
(1978) concluded that resistant cottons did not adversely
impact the bollworm parasite Campoletis sonorensis
(Cameron), while Mussett et al. (1979) noted a 69 percent
reduction in predatory arthropods on resistant cultivars
compared to a susceptible standard. While generalities
regarding the compatibility of plant resistance characters
and biological control agents appear somewhat problematic
(see Painter [1951] and Adkisson and Dyck [1980] for
conflicting viewpoints), any given effect will most probably
depend on the particular trait and natural enemy species
involved (see reviews in Boethel and Eikenbary [1986]).

Eariiness

The second major approach to cotton breeding has involved
selection for rapidly-maturing cultivars which (while
technically not considered resistance per se ) provide the
means to escape or evade pest damage. The development of
rapidly-maturing cultivars was among the initial recommen¬
dations for control of boll weevil (Howard 1896, Bennett
1904, 1908, Mally 1901, Newell and Rosenfield 1909) and
received major emphasis prior to the synthetic pesticide era
commencing shortly after World War II. Rapidly-
maturing cultivars were again reemphasized following the
"Heliothis" disasters of the 1970s (Adkisson 1971, Wilson
1974) and currently provide the foundation of the "short-
season" production systems that have been widely adopted
in many areas of the United States (Adkisson et al. 1982).
Bridge and McDonald (1987) noted that rapidly-maturing
cultivars developed during the past two decades have
reduced the growing season (time from planting to harvest)
by 28-33 days and have provided higher average yields.
Although entomologists at the turn of the century obviously
appreciated the value of earliness for control of pests such
as boll weevil, the ecological basis of the "short-season
effect" was elucidated much later in two important studies.
Walker and Niles (1971) detailed the process of boll weevil
reproduction as it interfaces with fruiting patterns of cotton.
These authors documented the capability of rapidly-
maturing cultivars to set an adequate crop of 12-day -old
bolls (which tend to be less susceptible to boll weevil
damage than younger counterparts) provided two conditions
are met: 1) that fewer than ca. twenty female boll weevils
per acre are initially present, and 2) that boll weevil
densities remain below damaging levels during a critical 30-
day blooming period. Parker et al. (1980) further docu¬
mented that the second condition could be achieved if
weevil densities were maintained at or below ca. 1,500

punctured squares per acre during the overwintered genera¬
tion, i.e., during the prebloom period. Results of the latter
study clearly demonstrated that the use of new rapidly-
maturing cultivars, in conjunction with narrow-row
planting ( < 40-in) and preemptive insecticidal applications
(1-3 treatments for overwintering boll weevils), could result
in acceptable yields (500-750 lbs/acre) in the presence of
the pest. In addition to these effects on boll weevil,
earliness has long been recognized as a means to reduce
damage by pink bollworm (Fenton and Owen 1931,
Chapman and Cavitt 1937, Fife et al. 1947). The develop¬
ment and adoption of rapidly-maturing cotton cultivars, in
conjunction with other cultural strategies such as the use of
defoliants and desiccants, has largely eliminated pink
bollworm as a pest of cotton in most regions of Texas
(Adkisson et al. 1982). This same approach appears to
have considerable potential for reducing pink bollworm
damage in the arid production regions of the Far West
(Walhood et al. 1983), and currently provides a basis for
management of bollworm and tobacco bud-worm popula¬
tions in much of the south-western United States (Rummel
et al. 1986).

Transgenic Cottons

A more recent and novel approach to plant breeding has
utilized the techniques of recombinant DNA technology in
an attempt to genetically engineer cultivars resistant to
insect or disease pests (Schmidt 1990). Using a transforma¬
tion system based on the soil microbe Agrobacterium
tumefaciens, researchers have successfully incorporated
genes of Bacillus thuringiensis which code for the produc¬
tion of the delta-endotoxin (toxic to many lepidopterous
pests) and the cowpea trypsin inhibitor (which encodes a
serine protease inhibitor) into cotton plants (Fillatti et al.
1989). Such toxin-producing plants produce an effect
similar to that of conventional insecticides; however, the
relative specificity of the B. t. delta-endotoxin is expected to
be considerably less disruptive to biological control agents
than synthetic insecticides (Gould and Weissinger 1990).
The probability that commercial adoption of genetically-
engineered cultivars will intensify selection pressure for
B.t. -resistance has been recognized (McGaughey 1990) and
several strategies to ameliorate the problem have been
proposed: 1) planting mixtures of resistant and susceptible
plants, 2) expression of toxins at levels that slow pest
growth and reproductive rates, 3) use of tissue-specific
promoters that cause only high levels of toxin-gene expres¬
sion in damage-sensitive plant parts, and 4) use of inducible
promoters that are triggered only when a threshold of

78

damage is exceeded, or by an inducing agent applied by the
producer (Gould and Weissinger 1990).

In addition to the development of pest-resistant cultivars, a
major emphasis of genetic engineering involves increasing
tolerance to various herbicides (Fillatti et al. 1989, Collins
1990). The role of biotechnology in agriculture, and recent
developments in cotton culture, are summarized in a
number of recent articles (Baker and Dunn 1990, Jenkins
and Parrott 1990, Leemans et al. 1990, Umbeck et al.
1990).

Agronomic Practices

Planting Date

Entomologists concerned with the control of boll weevil at
the turn of the century were adamant in their recommenda¬
tions for planting at the earliest possible date as a means to
enhance earliness (Howard 1896, Mally 1902, Bennett
1904). While early planting remains a basic tenet of most
"short-season" production schemes (Namken et al. 1983),
producers in at least one major production region (the
Rolling Plains of Texas) have adopted a somewhat different
approach. Based on detailed studies of boll weevil diapause
and overwinter survival (Rummel and Adkisson 1970,

White and Rummel 1978, Slosser 1978, Rummel and
Carrol 1985), a uniform delayed planting strategy was
developed in the Texas Rolling Plains as a means to
maximize suicidal emergence of postdiapause weevils
(Slosser 1978). During the past decade, this program has
significantly reduced losses to boll weevil and substantially
increased the profitability of cotton production in the region
(Walker 1984, Masud et al. 1984). Several workers have
suggested that a similar program of uniform delayed
planting might effectively suppress damage by pink
bollworm in the southwestern United States (Adkisson et
al. 1962, Henneberry 1986).

Planting Density and Row Configuration

Recommendations for planting configuration at the turn of
the century generally stressed relatively low plant densities
and wide row spacing as a means to increase mortality of
developing boll weevils by direct exposure to sunlight
(Mally 1902, Hunter 1904a, 1909, 1912, Hunter and Pierce
1912) or to enhance the effectiveness of indigenous natural
enemies (Pierce et al. 1912). Eventually, the advantages of
relatively high plant densities and narrow row spacing were

recognized (as means to enhance earliness) and the conven¬
tional 40-in row configuration was widely adopted (Cook
1913, Martin et al. 1923, Reynolds 1926, Ware 1930,
Cotton and Brown 1934). Narrow-row (30-in) configura¬
tions have been shown to accentuate the effect on earliness
(Niles 1970; Heilman and Namken 1987; Heilman et al.
1986, 1989; Kerby et al. 1988) and are becoming increas¬
ingly commonplace as the appropriate implements become
available from industry.

Fertilization and Irrigation

Efficient cotton production in many areas of the United
States is dependent on fertilization and irrigation; however,
excessive use of either may create conditions conducive to
outbreaks of certain pest species, e.g., members of the
bollworm-budworm complex (Fletcher 1941, Adkisson
1958, Wilson et al. 1980). Thus, current recommendations
generally stress moderation in the use of both inputs.

Recent studies have demonstrated that the timing of
irrigation may be used to considerable advantage: for
example, Slosser (1980) noted that bollworm damage in the
Texas Rolling Plains could be reduced substantially by
abstaining from irrigation for ca. 7-10 days prior to peak
moth flights, and for 3-4 days thereafter (irrigations were
scheduled using the simulation model MOTHZV-2
[Hartstack et al. 1976], one of several simulation models
currently available to producers for similar purposes).
Moreover, early termination of irrigation has been shown
to accelerate crop maturity and appears to be of consider¬
able importance in control of late-season pink bollworm
infestations in the southwestern United States (Watson et al.
1978, Henneberry 1986).

Plant Growth Regulators and Harvest-Aid Chemicals

Several of the plant growth regulators and harvest-aid
chemicals registered for use on cotton appear to have
considerable potential for suppression of late-season pest
infestations. Mepiquat chloride (PixR) applied to curtail
vegetative growth may compensate, in part, for the adverse
effects of high fertilization and irrigation rates, although the
impact on pest infestations is not entirely clear (Bradley et
al. 1986). Defoliants and desiccants applied to facilitate
harvest also effectively remove late-season squares and
bolls, which contribute little to harvest but are critically
important in the initiation of adult diapause in boll weevil
(Brazzel and Newsom 1959) and larval diapause in pink
bollworm (Lukefahr 1961, Adkisson et al. 1966). Timely
application of harvest-aid chemicals has been shown to
effectively suppress late-season infestations of both pest

79

species (Adkisson 1962, Cleveland and Smith 1964, Rice et
al. 1971, Mauney et al. 1972, Kittock et al. 1978, Bariola
et al. 1976, 1981, 1987, 1990, Hopkins and Moore 1980,
Henneberry and Bariola 1985). Henneberry (1986) noted
that chemical crop termination is among the most promising
approaches to late-season management of pink bollworm in
the Far West.

Stalk Destruction

Early destruction of cotton stalks was among the initial and
most adamant recommendations for control of boll weevil
(Howard 1986, Mally 1901, 1902, Sanderson 1903, Hunter
1904a, 1904b, 1907, 1909, Newell 1906, Hinds 1908) and
pink bollworm (Orlendorf 1926, Gaines 1952, Adkisson
and Gaines 1960, Noble 1969). The rationale of stalk
destruction in temperate environments is similar to that
discussed for harvest-aid chemicals, i.e., to destroy the
food resources required for initiation of adult diapause in
boll weevil and larval diapause in pink bollworm. In
addition to these effects, early stalk destruction in subtem-
perate and subtropical environments provides one of the
few means available to curtail or eliminate boll weevil
reproduction (Summy et al. 1988a) and overwinter survival
in desiccated bolls, a phenomenon which has been docu¬
mented in temperate environments (Cowan et al. 1963,
Walker and Shipp 1963) and is particularly commonplace in
the Arizona-Califomia region and much of southern Texas
(Fye et al. 1970, Bergman et al. 1982, 1983a, Bariola et al.
1984, Guerra et al. 1984, Summy 1990, Summy et al.
1988a, 1988b, Henneberry et al. 1989).

The few organized stalk destruction programs that have
been adequately evaluated have generally indicated a
dramatic suppression of overwintering boll weevils (Hunter
1907, 1912, Newell 1909, Isely 1930, Gaines 1952, Gaines
and Johnson 1949, Summy et al. 1988b). The importance
of early stalk destruction is perhaps best exemplified by the
experience with "stub" or "ratoon" cotton in the southwest¬
ern United States, which has invariably generated major
outbreaks of both species (e.g., Bergman et al. 1983b).
Early stalk destruction, in conjunction with other cultural
controls, played a key role in the demise of pink bollworm
as a cotton pest in most regions of Texas (Reynolds et al.
1975, Adkisson et al. 1982).

Tillage

Plowing and other forms of tillage may cause considerable
mortality of insect pests overwintering as diapausing larvae
or pupae within the former cotton environment, e.g., pink
bollworm (Watson 1980) and members of the bollworm-
budworm complex (Barber and Dicke 1937, Fife and
Graham 1966, Hopkins et al. 1972, Young and Price 1977,
Rummel and Neece 1989). Barber and Dicke (1937) noted
that overwintering survival of bollworm was reduced most
by fall plowing and least by fall disking, and documented a
significantly greater overwintering survival in red clay soils
compared to sandy loams. Overwintering survival in tilled
soils at two locations (0.3- 1.6 percent in Virginia and 7.8-
13.5 percent in Georgia) was significantly lower than that
occurring in untilled soils at the same locations (9-10
percent and ca. 30 percent, respectively). Roach (1981)
documented a significant reduction in bollworm emergence
from plots subjected to double-disking (1,746 moths/ha)
and subsoiling-bedding (3,007 moths/ha) compared to
emergence from undisturbed plots (7,760 moths/ha). More
recently, Rummel and Neece (1989) monitored overwinter¬
ing survival of bollworm during a three-year period in the
Texas High Plains and documented a significant reduction
in overwinter survival in cultivated plots (0-0.2 percent)
compared to noncultivated counterparts (2.0-30.9 percent).

Deep plowing has been shown to significantly reduce
overwinter survival of boll weevil and pink bollworm
infesting cotton bolls (Watson 1980, Henneberry 1986,
Henneberry et al. 1990). The practice has been mandated
by law in southern Texas as a means to destroy root
systems of cotton stalks which would otherwise produce
lush regrowth (which constitutes a nearly-optimal overwin¬
tering habitat for pests such as boll weevil) during much or
all of the fall and winter period (Summy et al. 1986a,

1988a, 1988b). Moreover, deep plowing may significantly
reduce the incidence of certain cotton pathogens, e.g.,
bacterial blight, Xanthomonas malvacearum, and root rot,
Phymatotrichum omnivorum (Presley and Bird 1968).

Habitat Manipulation

The cultural strategies discussed thus far involve manipula¬
tion of various elements within a single crop environment
to render conditions less favorable to one or more pest
species. An additional group of strategies has been
designed to manipulate pest densities to advantage within
the crop ecosystem (including native vegetation and

80

unrelated crop species) and/or to increase the effectiveness
of natural enemies (see reviews by Herzog and Funderbunk
1985, Stinner and Bradley 1989). Several of these strate¬
gies have been shown to significantly reduce damage by
cotton pests.

Strip harvesting

The management program developed for Lygus hesperus in
the San Joaquin Valley of California (Stem et al. 1964,
1967) provides an excellent example of the fact that
management practices on one crop may profoundly influ¬
ence pest damage on another. These authors noted that
conventional harvesting of alfalfa, the preferred host of
Lygus in California, was usually followed by a mass influx
of the pest into cotton. By harvesting alfalfa in strips,
rather than entire fields, Lygus infestations remained largely
concentrated on alfalfa and the need for insecticidal
treatment on cotton was greatly reduced (Stem et al. 1964,
1967, Reynolds et al. 1975).

Strip-cropping

Laster (1974) defined strip-cropping as "the interplanting of
primary crops with uniform parallel strips of secondary
crops in sufficient density to harbor ample beneficial insect
populations to combat insect pests of the primary crop. "
Interplanting of alfalfa and cotton in California achieved a
twofold effect: Lygus hesperus preferentially infested alfalfa
(as long as the latter remained lush) and densities of certain
natural enemy species increased substantially (Stem 1969).
Similar natural enemy increases were evident in Oklahoma
cotton fields interplanted with sorghum (Robinson et al.
1972, Burleigh et al. 1973, Johnson et al. 1976), which has
been formally recommended as a control strategy for that
area (Coakley and Young 1975). Pierce et al. (1912)
recommended the planting of certain cultivated crops (e.g.,
cowpeas) and native vegetation (e.g., dewberries) in close
proximity to cotton as a means to generate field nurseries of
boll weevil parasites. The latter strategy, while conceptu¬
ally sound, has not been tested empirically or adopted.

Provision of Supplemental Nutrients

Since commercial cotton cultivars (e.g., nectariless cottons)
may be deficient in nectar sources required by certain types
of natural enemies, several strategies have been proposed to
supply such resources artificially (see review in Johnson et
al. 1986). Certain types of native vegetation may provide
such nutrients (King and Holloway 1930), although it
should be recognized that these same species may conceiv¬
ably harbor important plant pathogens (van Emden 1965).

The application of artificial honeydew, a mixture of sugar
and Wheast (Hagen et al. 1976), has been shown to attract
and increase oviposition of a number of important predators
(Hagen and Tassan 1970, Hagen et al. 1971). Application
of artificial honeydew resulted in a 3-fold increase in
oviposition by Chrysoperla ( = Chrysopa ) and a significant
reduction in bollworm damage to cotton (Hagen et al.

1971). A similar effect was documented in Arkansas,
where application of a sugar- WheastR mixture to cotton
was accompanied by a one-month delay in the need for
insecticidal treatment for bollworm on cotton (Denver
1974).

Weed Management

The significance of wild hosts in the seasonal population
dynamics of highly polyphagous cotton pests is somewhat
conjectural and appears to vary regionally (Stadelbacher et
al. 1986). Surveys in certain areas e.g., southern Texas
(Harding 1976), have suggested only moderate densities of
Heliothis and Helicoverpa on native hosts during the early-
spring period, while major concentrations of both pest
species have been documented on plants such as crimson
clover. Trifolium incarnatum L., and Carolina geranium,
Geranium carolinianum L., in other areas, e.g., the south¬
eastern states (Snow and Brazzel 1965, Snow et al. 1966).
Stadelbacher (1982) demonstrated that bollworm densities
on Geranium dissect urn L. could be reduced by ca. 95.9
percent by a single properly-scheduled mowing, and by ca.
99.6 percent by a single herbicide application. Knipling
and Stadelbacher (1983) noted that bollworm infestations
on wild hosts during the early-season period tend to be
concentrated within a relatively restricted area, and pro¬
jected that 90 percent control of reproduction by first-
generation bollworms on such hosts might be obtained
through mowing, cultivation and application of selective
herbicides. Destruction of weeds such as Croton spp. has
also been proposed as a means to control cotton fleahopper
(Reinhard 1926, Eddy 1927, Folsom 1932), although the
efficacy of this approach has not been demonstrated
(Sterling et al. 1989).

Three possible adverse effects of weed management should
be recognized: stands of native vegetation may be important
aesthetically and/or for erosion control, may constitute
important reservoirs of natural enemies (Altieri et al. 1977,
Altieri and Whitcomb 1979, King et al. 1982) and may also
serve as important food sources for wildlife (Sterling et al.
1989).

81

Destruction or Modification of Overwintering Sites

In contrast to the majority of serious cotton pests, diapaus-
ing adult boll weevils (particularly in temperate environ¬
ments) overwinter in a state of dormancy in certain types of
preferred habitat, the most important of which is deciduous
leaf litter (Bottrell et al. 1972). Isely (1930) conducted a
brush-clearing experiment in a 600-acre tract in Arkansas,
and documented a substantial decline in boll weevil damage
during the following season. More recently, Slosser and
Boring (1980) studied the effects of pruning or thinning
trees in "shelter-belts" (rows of trees planted during the
1930s for erosion control) and documented a significant
reduction in overwinter survival of boll weevil which they
attributed to this type of "shelter-belt management."

Cultural Controls and Pest Population
Management

Many of the cultural strategies discussed previously are
singularly effective against one or more cotton insect or
disease pests. Most become considerably more effective
when used in combination with other cultural strategies,
and may become particularly potent when applied in
combination with other strategies in an organized commu¬
nity-wide pest management effort. The current trend
towards management of pest populations through a re¬
gional, rather than field-by-field, approach has long been
advocated (Knipling 1979) and ranks among the most
significant changes in the philosophy of pest management
which have occurred during the present century.

Regional Pest Control Programs

During the past two decades, regional control programs for
cotton pests have been implemented in several areas of the
United States (Henneberry et al., in press). The more
important of these include the Heliothis Management
Program in Arkansas (Phillips and Nicholson 1979) and
several management or eradication programs for boll weevil
(Davich 1976, Andrews 1981, Hamer et al. 1983, Masud et
al. 1984, Farr and Lane 1987, Summy et al. 1988a, Brazzel
1989). Operational differences notwithstanding, these
programs share one common denominator, i.e., the
objective of obtaining effective control through an orga¬
nized and concerted attack on the pest population, rather
than individual infestations.

A Case Study: the Lower Rio Grande Valley of
Texas

One of the most recent regional management programs for
cotton pests was directed against boll weevil in the Lower
Rio Grande Valley (LRGV) of Texas (Summy et al.

1988a). Many aspects of the LRGV program are unique to
the subtropical south Texas environment; however, others
(particular the logistical difficulties involved) are likely to
be encountered in similar programs established in environ¬
ments ranging from tropical to temperate. The LRGV
region encompasses a four-county area in extreme southern
Texas and an equivalent area of northern Tamaulipas,
Mexico. Cotton has traditionally ranked among the
principal and most profitable agricultural crops of the
LRGV region, although problems with insect pests have
been among the most severe in the Nation. Outbreaks of
organophosphate-resistant tobacco budworms during the
early 1970s caused unprecedented damage to LRGV cotton
and destroyed the 275, 000-ha cotton industry of northern
Tamaulipas, Mexico, within a period of ca. three years
(Reynolds et al. 1975). These disasters, more than any
encountered previously, emphasized the dangers of unilat¬
eral reliance on insecticides and the need for ecologically-
sound management practices for cotton pests.

The rejuvenation of LRGV cotton production may be
attributed in part to the development of a "short-season"
production scheme, which required nearly two decades of
research and involved the merger of several important
cultural strategies (Namken and Heilman 1973, Larson et
al. 1975, Namken et al. 1975, 1979, Gausman et al. 1979,
Heilman 1987, Heilman et al. 1979, Namken et al. 1983).
Namken et al. (1983) outlined the salient features of short-
season cotton production in areas such as the LRGV: 1) the
use of rapidly-maturing cultivars in combination with early-
season insecticidal treatment to minimize damage by
overwintered boll weevils, and 2) incorporation of agro¬
nomic practices designed to hasten crop maturity and thus
evade much of the late-season "Heliothis" infestation. The
transition from indeterminate to determinate cotton culti¬
vars (Bird 1976) proceeded at a somewhat slower pace than
in other areas of Texas because of initial problems with
fiber quality with the new short-season types. However,
once these problems were resolved, the transition proceeded
rapidly and virtually all cotton currently produced in the
LRGV region is classified as "short-season" (Heilman
1987).

82

A second major development in cotton pest management
involved the implementation of a regional stalk destruction
program directed primarily against boll weevil (Summy et
al. 1988a). Contrary to popular opinion, stalk destruction
in much of south Texas (the LRGV in particular) has, until
recently, been largely a matter of rhetoric rather than one of
concerted action. For several decades previously, the
Texas Pink Bollworm Quarantine provided a legal basis for
mandatory stalk destruction, but was fraught with legal
technicalities (for example, the absence of any provisions
for control of boll weevil) and was only rarely enforced
effectively. Mandatory stalk destruction was strongly
reemphasized following a series of destructive boll weevil
outbreaks during the late 1970s and early 1980s, and (after
the limitations of the Pink Bollworm Quarantine became
evident) a decision was eventually made by the local cotton
industry to pursue entirely new legislation. During 1987,
the Texas Legislature enacted the Texas Boll Weevil
Control Act, which provides stringent regulations on early
stalk destruction and, in contrast to previous legislation,
holds producers and landowners jointly liable for stalk
destruction violations. Under provisions of the new
legislation, pest management zones for control of boll
weevil have been implemented in 31 south and west Texas
counties (Summy et al. 1988a). Regulations applicable to
the LRGV region originally established a stalk destruction
deadline of 15 September, which has more recently been
reverted to 1 September.

Despite the relatively simple rationale of early stalk
destruction, i.e., as a means to curtail or eliminate boll
weevil reproduction and overwintering via diapause or
survival in desiccated bolls (Summy et al. 1988a), logistical
difficulties involved in the implementation of a regional
stalk destruction program proved to be formidable (Summy
et al. 1989). One of the principal problems involved
detection of undestroyed cotton, a significant fraction of
which commonly occurred in relatively isolated sites
inaccessible by public roads. A second problem involved
delays in stalk destruction due to inclement conditions,
e.g., precipitation, which tends to increase substantially
during September, and soil compaction, which may
preclude plowing by conventional implements. Such delays
were almost invariably accompanied by lush growth of
cotton plants (volunteer and regrowth) and the development
of dense weevil infestations which generally proved to be
difficult, if not impossible, to destroy during the fall and
winter period (Summy et al. 1988a). These problems were
addressed by the development of an aerial surveillance

system to facilitate efficient detection of cotton (Summy et
al. 1984, 1988a) and several strategies to ameliorate the
effects of inclement conditions (Summy et al. 1986a,

1986b). Problems with the small minority of producers
who, through ignorance or apathy, simply refused to abide
by stalk destruction guidelines was effectively addressed by
the new stalk destruction legislation.

The merger of rapidly-maturing cotton cultivars and early
stalk destruction into a single "short-season'' production
system completed the scheme of cultural controls recom¬
mended by cotton entomologists nearly a century ago (e.g.,
Howard 1896). The effect has been a significant reduction
in boll weevil damage and a concomitant reduction in
insecticide usage on cotton. Summy et al. (1988a) noted
that a series of destructive boll weevil outbreaks during the
late 1970s and early 1980s (which were largely the result of
lax stalk destruction) caused significant yield reductions and
a substantial increase in insecticide usage on LRGV cotton.
For example, production data for irrigated cotton during
1980 indicated an average yield of ca. 400 lbs/ac and an
average number of insecticide treatments approaching 15.0,
a figure comparable to that documented during the earlier
tobacco budworm outbreaks (Reynolds et al. 1975).

During a brief period in which the degree of stalk destruc¬
tion became increasingly efficient, average yields increased
while insecticide usage declined appreciably (e.g., by 1984,
average yields had increased to ca. 766 lbs lint/ac while the
average number of insecticide treatments to irrigated cotton
had declined to ca. 6.5). Conversely, such trends were
reversed entirely during a period in which stalk destruction
became increasingly lax (e.g., by 1986, average yields had
declined to ca. 605 lbs lint/ac while insecticide treatments
had increased to ca. 1 1.0). With the advent of new stalk
destruction legislation during 1988, and renewed emphasis
by the regulatory agency charged with enforcement, the
average number of required insecticide applications
declined appreciably, e.g., to 5-6 for irrigated and <1.0
for dry land cotton during 1990 (J. W. Norman, Texas
Agricultural Extension Service, pers. comm.).

The latter trend is particularly timely as a means to lessen
selection pressure on the local population of boll weevil,
which appears to be considerably less susceptible to
azinphosmethyl than other populations in Texas (Teague et
al. 1983), and tobacco budworm, which has recently
exhibited evidence of resistance to two of the synthetic
pyrethroid insecticides (Plapp and Campanhola 1986,

Roush and Luttrell 1987). Moreover, the reduction of

83

pesticide usage on LRGV cotton has created an environ¬
ment more conducive to the utilization of biological control
agents for cotton pests and has fostered an unusual degree
of cooperation between LRGV cotton producers and local
environmentalists (Graham 1990). The latter trend stands
in marked contrast to the controversies which characterized
boll weevil control and eradication efforts in other areas of
the United States during recent years.

Much of the technology employed during the stalk destruc¬
tion effort has much broader applications in agriculture.

For example, the color infrared photography used in aerial
surveillance has been used extensively to estimate crop
yields (Thomas and Oerther 1977), evaluate the effects of
nutrient stress and freeze damage (Escobar et al. 1983,
Thomas and Learner 1987), and to assess damage caused by
insect and disease pests (Norman and Fritz 1965, Hart and
Meyers 1968, Heald et al. 1972, Hart et al. 1977, Manzer
and Cooper 1982, Wildman 1984). In conjunction with
more recent innovations such as color video imagery
(Nixon et al. 1987) and satellite imagery (Richardson et al.
1989), such remote sensing technology forms the basis for
the development of a Geographic Information System for
the LRGV and other areas (J. R. Richardson, pers.
comm.). Perhaps the most significant result of the stalk
destruction effort has been acceptance by the majority of
LRGV producers of the need for regional management of
boll weevil and other cotton pests. The concept of regional
management has more recently been extended to the newly-
emerging cotton industry of Tamaulipas, Mexico, which
has escalated from ca. 1,200-ha during 1989 to ca. 60,000-
ha during 1991. Several recent meetings between cotton
producers from Texas and Mexico stressed the theme that
effective management of boll weevil in the LRGV region
will be possible only through a sustained cooperative effort
involving both countries. As a result, stringent stalk
destruction regulations have been reinstituted in northern
Mexico, although a cooperative management program with
the United States has yet to evolve.

To summarize, the current management program for LRGV
cotton pests involves an array of cultural strategies which
have been voluntarily adopted as part of the changing
technology of cotton production (e.g., determinate cotton
cultivars, plant growth regulators and defoliants or desic¬
cants) and others which have been mandated by law (e.g.,
planting and stalk destruction deadlines). Many of the
important cultural strategies discussed herein have received

little emphasis (e.g., minimizing rates of irrigation and
fertilization) but could greatly increase efficacy of the
overall program if adopted by the majority of producers.

Summary

Cotton production in the United States during the past two
decades has shifted largely from production systems heavily
dependent on broad-spectrum insecticides to systems
utilizing an array of ecologically-sound cultural and
biological controls. Widespread adoption of cultural
strategies, which range from use of resistant cultivars to a
sequence of agronomic practices which render the environ¬
ment less hospitable to one or more pests, has greatly
diminished damage by key pests, reduced the incidence of
secondary outbreaks, and is expected to create an environ¬
ment conducive to the success of additional strategies such
as biological control. Cultural strategies become particu¬
larly potent when incorporated into organized regional pest
management efforts, and provide a means to produce cotton
profitably without the serious economic losses and environ¬
mental degradation which has characterized cotton produc¬
tion in the past.

Acknowledgment

The authors express appreciation to Drs. T. J. Henneberry,
T. F. Leigh and J. K. Walker for critical reviews of this
manuscript.

References Cited

Adkisson, P. L. 1958. The influence of fertilizer applica¬
tions on populations of Heliothis zea (Boddie)and certain
insect predators. J. Econ. Entomol. 51: 757-759.

Adkisson, P. L. 1962. Timing of defoliants and desiccants
to reduce populations of the pink bollworm in diapause. J.
Econ. Entomol. 55: 949-951.

Adkisson, P. L. 1971. Objective uses of insecticides in
agriculture, pp. 43-51. In J. E. Swift [ed.], Proc. Symp.
on Agricultural Chemicals - Harmony or discord for food,
people and the environment. Univ. Calif.

84

Adkisson, P. L., and J. C. Gaines. 1960. Pink boll worm
control as related to the total cotton insect control program
of Central Texas. Texas. Agric. Exp. Stn. Misc. Publ.

444.

Adkisson, P. L., and V. A. Dyck. 1980. Resistant
varieties in pest management systems, pp. 233-251. In F.
G. Maxwell and P. R. Jennings [eds.], Breeding plants
resistant to insects. John Wiley, New York.

Adkisson, P. L., O. T. Robertson, and L. C. Fife. 1962.
Planting date as a factor involved in pink boll worm control,
pp. 16-20. In D. F. Martin and R. D. Lewis [eds.], A
summary of recent research basic to the cultural control of
pink bollworm. Texas Agric. Exp. Stn., Misc. Pub. 579.

Adkisson, P. L., C. F. Bailey, and A. Niles. 1966.

Internal clocks and insect diapause. Sci. 154: 234-241.

Adkisson, P. L., G. A. Niles, J. K. Walker, L. S. Bird,
and H. B. Scott. 1982. Controlling cotton insect pests: A
new system. Science 216: 19-22.

Altieri, M.A., and W. H. Whitcomb. 1979. The potential
use of weeds in the manipulation of beneficial insects.

Hort. Sci. 14: 12-18.

Altieri, M. A., A. van Schoonhoven, and J. Doll. 1977.
The ecological role of weeds in insect pest management
systems: A review illustrated with bean ( Phaseolus
vulgaris L.) cropping systems. PANS 23: 185-206.

Andrews, G. L. 1981. Optimum pest management trial,
pp. 41-44. In Proc. Beltwide Cotton Prod. Res. Conf.

Bailey, J. C. 1981. Growth comparison of Heliothis
virescens (F.) larvae fed white, yellow and orange pollen.
Proc. Beltwide Cotton Prod. Res. Conf.

Bailey, J. C. 1982. Influence of plant bug and leafhopper
populations on glabrous and nectariless cottons. Environ.
Entomol. 11: 1011-1013.

Bailey, J. C., B. W. Hanney, and W. R. Meredith, Jr.
1980. Combinations of resistant traits and insecticides:
Effect on cotton yield and insect populations. J. Econ.
Entomol. 73: 58-60.

Bailey, J. C., A. L. Scales, and W. R. Meredith, Jr. 1984.
Tarnished plant bug (Heteroptera: Miridae) nymph numbers
decreased on caged nectariless cotton. J. Econ. Entomol.
77: 68-69.

Baker, R. R. and P. E. Dunn. 1990. New directions in
biological control. Alan R. Liss, Inc., New York.

Barber, G. W., and F. F. Dicke. 1937. The effectiveness
of cultivation as a control for the com earworm. U. S.

Dep. Agric. Tech. Bull. 561.

Bariola, L. A., D. L. Kittock, H. F. Arle, and others.

1976. Controlling pink bollworms: Effect of chemical
termination of cotton fruiting on populations of diapausing
larvae. J. Econ. Entomol. 69: 633-636.

Bariola, L. A., T. J. Henneberry, and D. Bergman. 1984.
Boll weevils (Coleoptera: Curculionidae): Emergence from
bolls during the spring, and trapping of adults in Arizona.

J. Econ. Entomol. 77: 1166-1170.

Bariola, L. A., T. J. Henneberry, and D. L. Kittock. 1981.
Chemical termination and irrigation cut-off to reduce over¬
wintering populations of pink bollworms. J. Econ.
Entomol. 74: 106-109.

Bariola, L. A., T. J. Henneberry, and C. C. Chu. 1987.
Prep and Dropp for pink bollworm and boll weevil control
in Arizona and southern Arizona, p. 340. In Proc.

Beltwide Cotton Prod. Res. Conf.

Bariola, L. A., C. C. Chu, and T. J. Henneberry. 1990.
Controlling pink bollworms: Timing applications of plant
growth regulators for chemical termination. J. Econ.
Entomol. 83: 1074-1079.

Bartlett, B. R. 1956. Natural predators. Can selected
insecticides help to preserve biotic control? Agric. Chem.
11: 42-44, 107-109.

Bennett, R. L. 1904. Early cotton. Texas Agric. Exp.

Stn. Bull. 75.

Bennett, R. L. 1908. A method of breeding early cotton
to escape boll weevil damage. U. S. Dep. Agric. Farmers'
Bull. No. 314.

85

I

'

■

Bergman, D., T. J. Henneberry, and L. A. Bariola. 1982.
Distribution of the boll weevil in southwestern Arizona
cultivated cotton, pp. 204-207. In Proc. Beltwide Cotton
Prod. Res. Conf.

Bergman, D., T. J. Henneberry, and L. A. Bariola. 1983a.
Overwintered boll weevil populations in southwestern
Arizona cultivated cotton, pp. 182-185. In Proc. Beltwide
Cotton Prod. Res. Conf.

Bergman, D., T. J. Henneberry, L.A. Bariola, and J.M.
Gillespie. 1983b. Studies of pest and beneficial insects in
Arizona stub and planted cotton. U. S. Dep. Agric.,

Agric. Res. Ser. ARS-W-32.

Berlinger, M. J. 1986. Host plant resistance to Bemisia
tabaci. Agric. Ecosys. Environ. 17: 69-82.

Bhardwa, H. L., and J. B. Weaver, Jr. 1983. Bollworm
resistance to red leaf cotton, pp. 1 17-1 19. In Proc.

Beltwide Cotton Prod. Res. Conf.

Bird, L. S. 1976. Registration of TAMCOT SP21,
TAMCOT SP23, and TAMCOT SP37 cottons. Crop Sci.
16: 884.

Boethei, D. J., and R. D. Eikenbary. 1986. Interactions
of plant resistance and parasitoids and predators of insects.
Ellis Horwood, Chichester.

Bottger, G. T., and R. Patana. 1966. Growth, develop¬
ment and survival of certain lepidoptera fed gossypol in the
diet. J. Econ. Entomol. 59: 1166-1168.

Bottrell, D. G., J. R. White, D. S. Moody, and D. D.
Hardee. 1972. Over-wintering habitats of the boll weevil
in the Rolling Plains of Texas. Environ. Entomol. 1: 633-
638.

Bradley, J. R., Jr., G. A. Herzog, S. H. Roach, R. E.
Stinner, and L. I. Terry. 1986. Cultural control in
southeastern U. S. cropping systems, pp. 22-27. In S. J.
Johnson, E. G. King, and J. R. Bradley [eds.], Theory and
tactics of Heliothis population management: I - Cultural and
biological control. Southern Coop. Series Bull. 316.

Brazzel, J. R. 1989. Boll weevil eradication - An update,
pp. 218-220. In Proc. Beltwide Cotton Prod. Res. Conf.

Brazzel, J. R., and L. D. Newsom. 1959. Diapause in
Anthonomus grandis Boh. J. Econ. Entomol. 52: 603-611.

Bridge, R. R., and L. D. McDonald. 1987. Beltwide
efforts and trends in development of varieties for short-
season production systems, pp. 81-85. In Special Session,
Beltwide Cotton Prod. Res. Conf.

Burleigh, J. G., J. H. Young, and R. D. Morrison. 1973.
Strip-croppings effect on beneficial insects and spiders
associated with cotton in Oklahoma. Environ. Entomol. 2:
281-285.

Butler, G. D., Jr., and T. J. Henneberry. 1984. Bemisia
tabaci'. Effect of cotton leaf pubescence on abundance.
South-western Entomol. 9: 91-94.

Butler, G. D., Jr., and F. D. Wilson. 1984. Activity of
adult whiteflies (Homoptera: Aleyrodidae) within plantings
of different cotton strains and cultivars as determined by
stick-trap catches. J. Econ. Entomol. 77: 1137-1140.

I

Butler, G. D., Jr., T. J. Henneberry, and F. D. Wilson.

1986. Bemisia tabaci (Homoptera: Aleyrodidae) on cotton:
Adult activity and cultivar oviposition preference. J. Econ.
Entomol. 79: 350-354.

I

Chapman, A. J., and H. S. Cavitt. 1937. Possibilities of
reducing overwintered pink bollworm population in the soil
as shown by stripping tests. J. Econ. Entomol. 30: 837-
838.

Cleveland, T. C., and G. L. Smith. 1964. Effects of post
season applications of insecticides, defoliants, and desic¬
cants on diapausing boll weevils. J. Econ. Entomol. 57:
527-529.

Coakley, J., and J. Young. 1975. Strip-cropping cotton
and sorghum for bollworm control. Okla. State Univ.

Current Report, pp. 7151-7151.1.

Collins, H. B. 1990. Biotechnology and cotton improve¬
ment in the 1990s, pp. 7-9. In Special Session, Beltwide
Cotton Prod. Res. Conf.

Cook, O. F. 1913. A new system of cotton culture. U. S.
Dep. Agric., Bur. Plant Ind. C. 115.

86

Cotton, L. R., and H. B. Brown. 1934. Cotton spacing in
southern Louisiana in relation to certain plant characters.
Louisiana Agric. Exp. Stn. Bull. 246.

Cowan, C. B., Jr., J. W. Davis, and C. R. Parencia.

1963. Winter survival of the boll weevil in cotton bolls in
central Texas. J. Econ. Entomol. 56: 494-496.

Davich, T. B. 1976. Foreward, pp. 1-2. In Conf. Proc.
Boll Weevil Suppression, Management and Elimination
Technology. U. S. Dep. Agric., Agric. Res. Serv. ARS-S-
71.

DeBord, D. V. 1977. Cotton insect and weed loss
analysis. The Cotton Foundation. Mimeographed, 122 pp.

Denver, C. 1974. The use of biological control in
Arkansas farming operation. Proc. Tall Timbers Conf. on
Ecol. Anim. Cont. Habit. Mgmt. 5: 61-65.

Eddy, C.O. 1927. The cotton fleahopper. S. C. Agric.
Exp. Stn. Bull. 235.

Escobar, D. E., R. L. Bowen, H. W. Gausman, and G. R.
Cooper. 1983. Use of near-infrared video recording
system for the detection of freeze-damaged citrus leaves. J.
Rio Grande Valley Hort. Soc. 36: 61-66.

El-Zik, K. M., and P. M. Thaxton. 1989. Genetic
improvement for resistance to pest and stresses in cotton,
pp. 191-224. In R. E. Frisbie, K. M. El-Zik, and L. T.
Wilson [eds.], Integrated pest management systems and
cotton production. John Wiley & Sons, New York.

Farr, C. R. and M. L. Lame. 1987. Late-season agro¬
nomic evaluation of boll weevil control programs, p. 341,
In Proc. Beltwide Cotton Prod. Res. Conf.

Fenton, F. A., and W. L. Owen. 1931. Hibernation of
Pectinophora gossypiella in Texas. J. Econ. Entomol. 24:
1197-1207.

Fife, L. C., and H. M. Graham. 1966. Cultural control of
overwintering bollworm and tobacco budworm. J. Econ.
Entomol. 59: 1123-1125.

Fife, L. C., I. Schiller, and A. J. Chapman. 1947. Pink
bollworm carryover from one cotton crop to the next in the
lower Rio Grande Valley. J. Econ. Entomol. 40: 540-545.

Fillatti, J., C. McCall, L. Comai, J. Kiser, K. McBride
and D. Stalker. 1989. Genetic engineering of cotton for
herbicide and insecticide resistance, pp. 17-19. In Proc.
Beltwide Cotton Prod. Res. Conf.

Fletcher, R. K. 1941. The relation of moisture content of
the cotton plant to oviposition by Heliothis armigera
(Hbn.) and to survival of young larvae. J. Econ. Entomol.
34: 856-858.

Flynn, C. W., Jr. 1907. The boll weevil. La. Crop Pest
Comm. C. 11.

Folsom, J. W. 1932. Insect enemies of the cotton plant.

U. S. Dep. Agric. Farmers' Bull. 1688.

Foudin, A. S. and S. Shantharam. 1989. USDA's role in
regulating genetically engineered organisms, pp. 19-20. In
Proc. Beltwide Cotton Prod. Res. Conf.

Fye, R. E., J. E. Leggett, and C. D. Bonham. 1970.
Winter survival of the boll weevil complex in Arizona. J.
Econ. Entomol. 63: 1071-1074.

Gaines, R. C. 1952. The boll weevil, pp. 501-504. In F.
C. Bishopp, and others [eds.], The yearbook of agriculture
- 1952. U. S. Dep. Agric.

Gaines, J. C., and H. G. Johnson. 1949. Report to a
meeting on 15-18 June. Acco. Press., Houston, Texas.

Gausman, H. W., L. N. Namken, M. D. Heilman, J.
Walter, and F. R. Rittig. 1979. Physiological effects of a
growth regulator (Pix) on the cotton plant, pp. 51-52. In
Proc. Beltwide Cotton Prod. Res. Conf.

Gould, F. and A. Weissinger. 1990. Roles for public and
private sector scientists in developing pest-resistant crops,
pp. 641-648. In R. R. Baker and P. E. Dunn (eds.) New
directions in biological control. Alan R. Liss, Inc., New
York.

Graham, F. 1990. Evil weevil. Audubon 92:8-10.

Guerra, A. A., R. F. Garcia, P. R. Bodegas V., and M. E.
De Coss F. 1984. The quiescent physiological status of
boll weevils (Coleoptera: Curculionidae) during the
noncotton season in the tropical zone of Soconusco in
Chiapas, Mexico. J. Econ. Entomol. 77: 595-598.

87

Hagen, K. S., and R. L. Tassan. 1970. The influence of
food Wheast and related Saccharomyces fragilis yeast
products on the fecundity of Chrysopa carnea. Can.
Entomol. 102: 806-811.

Hagen, K. S., E. F. Sawall, Jr., and R. L. Tassan. 1971.
The use of food sprays to increase effectiveness of ento-
mophagous insects. Proc. Tall Timbers Conf. Anim. Cont.
Habit. Mgmt. 2: 59-81.

Hagen, K. S., P. Greany, E. F. Sawall, Jr., and R. L.
Tassan. 1976. Tryptophan in artificial honeydews as a
source of an attractant for adult Chrysopa carnea. Environ.
Entomol. 5:458-468.

Hamer, J. L., G. L. Andrews, R. W. Seward, D. F.

Young, Jr., and R. B. Head. 1983. Optimum pest
management trial in Mississippi, pp. 385-407. In R. L.
Ridgway, E. P. Lloyd and W. H. Cross (eds.), Cotton
insect management with special reference to the boll
weevil. U. S. Dep. Agric., Agric. Handb. 589.

Hanney, B. W., J. C. Bailey, and W. R. Meredith, Jr.

1979. Yellow cotton pollen suppresses growth of tobacco
budworm larvae. Proc. Beltwide Cotton Prod. Res. Conf.

Harding, J. A. 1976. Heliothis spp.: Seasonal occurrence
of hosts , and host importance in the lower Rio Grande
Valley. Environ. Entomol. 5: 666-668.

Hart, W. G., and V. I. Myers. 1968. Infrared aerial
photography for detection of populations of brown soft
scale in citrus groves. J. Econ. Entomol. 61: 617-624.

Kart, W. G., S. J. Ingle, and M. R. Davis. 1977. Remote
sensing to detect insect populations attacking citrus, pp.
285-287, In Proc. Int. Soc. Citriculture.

Hartstack, A. W., Jr., J. A. Witz, J. P. Hollingsworth, R.
L. Ridgway, and J. D. Lopez. 1976. MOTHZV-2: A
computer simulation of Heliothis zea and Heliothis
virescens population dynamics. U. S. Dep. Agric., Agric.
Res. Ser. ARS-S-127.

Heald, C. M., W. H. Thames, and C. L. Wiegand. 1972.
Detection of Rotylenchulus reinformis infestations by aerial
infrared photography. J. Nematology 4: 298-300.

Heilman, M. D. 1987. The Texas short-season program,
pp. 13-18. In Western Cotton Prod. Conf., El Centro,
Calif.

Heilman, M. D, and L. N. Namken. 1987. Integrated
cotton production system with 30-inch row culture for
improved efficiency and quality, pp. 95-97. In Proc.
Beltwide Cotton Prod. Res. Conf.

Heilman, M. D., L. N. Namken, J. W. Norman, and M. J.
Lukefahr. 1979. Evaluation of an integrated short-season
management production system for cotton. J. Econ.
Entomol. 72: 896-900.

Heilman, M. D., T. D. Valeo, A. W. Scott, and J. W.
Norman. 1986. Evaluation of cultivars, row configuration
and harvesting methods for 30-inch production systems in
the Lower Rio Grande Valley of Texas, pp. 111-113. In
Proc. Beltwide Cotton Prod. Res. Conf.

Heilman, M. D., L. N. Namken, and T. D. Valeo. 1989.
Comparison of 30- vs 40-inch row spacing on lint cotton
yield and quality, pp. 125-127. In Proc. Beltwide Cotton
Prod. Res. Conf.

Henneberry, T. J. 1986. Pink bollworm management in
cotton in the southwestern United States. U. S. Dep.
Agric., Agr. Res. Serv. ARS-51.

Henneberry, T. J., and L. A. Bariola. 1985. The potential
of plant growth regulators for selective termination of late
season cotton fruiting and pink bollworm population
management, pp. 22-27. In E. Natwick, ed.) Proc., 1985
Cotton Insect and Prod. Meeting, Univ. Calif. Coop. Ext.
Serv., El Centro, Calif.

Henneberry, T. J., and J. R. Phillips. Suppression and
management of cotton insect populations on an areawide
basis. In Cotton insects and mites: Their characterization
and management. The Cotton Foundation Reference Series
No. 3. , in press.

Henneberry, T. J., L. A. Bariola, and D. L. Kittock.

1977. Nectariless cotton: Effect on cotton leafperforator
and other cotton insects in Arizona. J. Econ. Entomol. 70:
797-799.

88

Henneberry, T. J., T. Meng, and L. A. Bariola. 1989.
Late season boll weevil infestations and survival in cotton
bolls in Arizona, pp. 197-199. In Cotton, Series P-77.
Arizona Coop. Ext. Serv.

Henneberry, T. J., T. Meng, Jr., and L. A. Bariola. 1990.
Overwintering survival and emergence of boll weevil
(Coleoptera: Curculionidae) in cotton bolls in Arizona. J.
Econ. Entomol. 83:1879-1882.

Herzog, D. C., and J. E. Funderbunk. 1985. Plant
resistance and cultural practice interactions with biological
control, pp. 67-81 In M. A. Hoy and D. C. Herzog [eds.],
Biological control in agricultural IPM systems. Academic
Press, New York.

Herzog, G. A., and E. G. King. 1990. Highlights of the
1990 cotton insect research and control conference, pp. 84-
86. In Special Sessions, Proc. Beltwide Cotton Prod. Res.
Conf.

Hinds, W. E. 1908. The first and last essential step in
combating the boll weevil. J. Econ. Entomol. 1: 233-243.

Hopkins, A. R., and R. F. Moore. 1980. Thidiazuron:
Effect of applications on boll weevil and bollworm popula¬
tions, leaf abscission and growth of cotton plants. J. Econ.
Entomol. 73: 768-770.

Hopkins, A. R., H, M. Taft, and W. James. 1972.
Comparisons of mechanical cultivation and herbicides on
emergence of bollworms and tobacco budworms. J. Econ.
Entomol. 65:870-872.

Howard, L. O. 1896. Insects affecting the cotton plant.

U. S. Dep. Agric. Office of Exp. Stn. Bull. 33.

Hunter, W. D. 1904a. Information concerning the
Mexican cotton-boll weevil. U. S. Dep. Agric. Farmers'
Bull. 189.

Hunter, W. D. 1904b. The most important step in the
cultural system of controlling the boll weevil. U. S. Dep.
Agric. Bur. Entomol., C. 56.

Hunter, W. D. 1907. The most important step in the
control of the boll weevil. U. S. Dep. Agric., Bur.
Entomol. C. 95.

Hunter, W. D. 1909. The boll weevil problem with
special reference to means of reducing damage. U. S. Dep.
Agric. Farmers' Bull. 344.

Hunter, W. D. 1912. The boll weevil problem, with
special reference to means of reducing damage. U. S. Dep.
Agric., Farmers' Bull. 512.

Hunter, W. D. and W. D. Pierce. 1912. The Mexican
cotton boll weevil; A summary of the results of the investi¬
gations of this insect up to December 31, 1911. U. S. Dep.
Agric., Bur. Entomol. Bull. 114.

Hunter, W. D. and B. R. Coad. 1923. The boll weevil
problem. U. S. Dep. Agric., Farmers' Bull. 1329.

Hunter, R. C, T. F. Leigh, C. Lincoln, B. A. Waddle, and
L. A. Bariola. 1965. Evaluation of a selected cross¬
selection of cottons for resistance to the boll weevil. Ark.
Agric. Exp. Stn. Bull. 700.

Isely, D. 1928. The relation of leaf color and leaf size to
boll weevil infestation. J. Econ. Entomol. 21: 553-559.

Isely, D. 1930. Ark. Agric. Exp. Stn., 42nd Ann. Rep. B.
257: 54-55.

Jenkins, J. N. 1982a. Plant-pest interaction with environ¬
mental stress and breeding for pest resistance, pp. 365-373.
In D. Christianson [ed.] Breeding plants for less favorable
environments. John Wiley & Sons, New York.

Jenkins, J. N. 1982b. Present state of the art and science
of cotton breeding for insect resistance in the southeast, pp.
117-125, In Proc. Beltwide Cotton Prod. Res. Conf.

Jenkins, J. N. and W. L. Parrott. 1990. Field and
laboratory evaluations of transgenic cotton strains contain¬
ing a Gene from Bacillus thuringiensis Kurstaki strain
HD-1, p.635. In Proc. Beltwide Cotton Prod. Res. Conf.

Jenkins, J. N., and F. D. Wilson. Host plant resistance.

In E. G. King and J. R. Phillips [eds.], Cotton insects and
mites: Their characterization and management. The Cotton
Foundation Reference Series No. 3. , in press.

Jenkins, J. N., F. G. Maxwell, W. L. Parrott, and W. T.
Buford. 1969. Resistance to boll weevil ( Anthonomus
grandis Boh.) oviposition in cotton. Crop Sci. 9: 369-372.

89

Johnson, E. K., J. H. Young, D. R. Molnar, and R. D.
Morrison. 1976. Effects of three insect control schemes
on populations of cotton insects and spiders, fruit damage,
and yield of Westbum 70 cotton. Environ. Entomol. 5:
508-510.

Johnson, S. J., H. N. Pitre, J. E. Powell, and W. L.
Sterling. 1986. Control of Heliothis spp. by conservation
and importation of natural enemies, pp. 132-154. In S. J.
Johnson, E. G. King, and J. R. Bradley [eds.], Theory and
tactics of Heliothis population management. Southern
Coop. Series Bull. 316.

Jones, J. E. 1972. Effect of morphological characters of
cotton on insects and pathogens, pp. 88-92. In Proc.
Beltwide Cotton Prod. Res. Conf.

Jones, J. E., D. F. Clower, M. R. Milam, W. D.

Caldwell, and D. R. Melville. 1975. Resistance in upland
cotton to the bandedwinged whitefly, Trialeurodes
abutilonea (Haldeman), pp. 98-99. In Proc. Beltwide
Cotton Prod. Res. Conf.

Jones, J. E., J. B. Weaver, and M. F. Schuster. 1978.

Host plant resistance to the boll weevil, pp. 50-73. In The
boll weevil: Management strategies. Southern Coop. Series
Bull. 228.

Jones, J. E., D. F. Clower, E. Burris, J. G. Marshall, and
S. Stringer. 1983. Progress in breeding frego-nectariless
cottons for reduced plant bug sensitivity. In Proc. Beltwide
Cotton Prod. Res. Conf.

Kerby, T. A., S. Johnson, and H. Yamada. 1984. Fruit¬
ing pattern modification by irrigation regime, nitrogen
levels and Pix, pp. 65-66. In Proc. Beltwide Cotton Prod.
Res. Conf.

Kerby, T. A., K. G. Kassman, L. Urie, and M. Keeley.
1988. Phenotypes and a production system for high-
yielding 30-inch, pp. 119-120. In Proc. Beltwide Cotton
Prod. Res. Conf.

King, J. L., and J. K. Holloway. 1930. Tiphia
popilliavora Rohwer, a parasite of the Japanese beetle.

U. S. Dep. Agric., Circ. 145.

King, E.G., J. E. Powell, and J. W. Smith. 1982. Pros¬
pects for utilization of parasites and predators for manage¬
ment of Heliothis spp, pp. 103-122. In Proc. Intern.
Workshop on Heliothis Management, ICRISAT,

Patancheru, India.

Kittock, D. L., H. F. Arle, T. J. Henneberry, and L. A.
Bariola. 1978. Chemical termination of late-season cotton
fruiting in Arizona and California, 1972-76. U. S. Dep.
Agric., Agric. Res. Serv. ARS-W-52.

Knipling, E. F. 1979. The basic principles of insect
population suppression and management. U. S. Dep.

Agric., Agric. Handb. 512.

Knipling, E. F., and E. A. Stadelbacher. 1983. The
rationale for areawide management of Heliothis (Lepidopt-
era: Noctuidae) populations. Bull. Entomol. Soc. Am. 29:
29-37.

Lane, H. C., and M. F. Schuster. 1979. Tannins in cotton
leaves as deterrents to spider mite feeding. Proc. Beltwide
Cotton Prod. Res. Conf.

Larson, J. L., R. D. Lacewell, J. E. Casey, M. D.
Heilman, L. N. Namken, and R. D. Parker. 1975. Eco¬
nomic environmental and energy use of implications of
short-season cotton production: Texas Lower Rio Grande
Valley, pp. 171-177. In Proc. Beltwide Cotton Prod. Res.
Conf.

Laster, M. L. 1974. Increasing natural enemy resources
through crop rotation and strip-cropping. In F. G. Max¬
well and F. A. Harris [eds.], Proc. Summer Institute of
Biol. Cont. Univ. Miss. Press.

Leemans, J., A. Reynaerts, H. Hofte, M. Peferoen, H.
van Mellaert and H. Joos. 1990. Insecticidal crystal
proteins from Bacillus thuringiensis and their use in
transgenic crops, pp. 573-581, In R. R. Baker and P. E.
Dunn (eds.) New directions in biological control. Alan R.
Liss, Inc., New York.

Lingren, P. D., M. J. Lukefahr, M. Diaz, and A.

Hartstack. 1978. Tobacco budworm control in caged
cotton with a resistant variety, augmentative releases of
Campoletis sonorensis and natural control by other benefi¬
cial species. J. Econ. Entomol. 71: 739-745.

90

Lopez, E. G., and G. L. Teetes. 1979. Selected predators
of aphids in grain sorghum and their relation to cotton. J.
Econ. Entomol. 69: 198-204.

Lukefahr, M. J. 1961. Factors related to the induction of
diapause in the pinjk bollworm. Ph.D. Diss., Texas A&M
Univ.

Lukefahr, M. J. 1977. Varietal resistance to cotton
insects, pp. 236-237. In Proc. Beltwide Cotton Prod. Res.
Conf.

Lukefahr, M. J., and C. Rhyne. 1960. Effects of
nectariless cottons on populations of Lepidopterous insects.
J. Econ. Entomol. 53: 242-244.

Lukefahr, M. J., and J. E. Houghtaling. 1969. Resistance
of cotton strains with high gossypol content to Heliothis
spp. J. Econ. Entomol. 62: 588-591.

Lukefahr, M. J., J. E. Houghtaling, and H. M. Graham.
1971. Suppression of Heliothis populations with glabrous
cotton strains. J. Econ. Entomol. 64: 486-488.

Luttrell, R. G., J. R. Phillips, and T. R. Pfrimmer. 1986.
Cultural control in mid-south U. S. cropping systems, pp.
28-37. In S. J. Johnson, E. G. King, and J. R. Bradley
[eds.], Theory and tactics of Heliothis population manage¬
ment: I - Cultural and biological control. Southern Coop.
Series Bull. 316.

Mally, F. W. 1901. The Mexican cotton boll weevil. U.
S. Dep. Agric., Farmers' Bull. No. 130.

Mally, F. W. 1902. Report on the boll weevil. The
Agricultural and Mechanical College of Texas. Von
Boeckmann, Schultze and Co., Austin, Texas.

Manzer, F. E., and G. R. Cooper. 1982. Use of portable
videotaping for aerial infrared detection of potato diseases.
Plant Disease 66: 665-667.

Martin, R. D., W. W. Ballard, and D. M. Simpson. 1923.
Growth of fruiting parts in cotton plants. J. Agr. Res. 25:
195-208.

Masud, S. M., R. D. Lacewell, E. P. Boring, and T. W.
Fuchs. 1984. Economic implications of a delayed uniform

planting date for cotton production in the Texas Rolling
Plains. Texas Agric. Exp. Stn. Bull. 1481.

Mauney, J. R., D. L. Kittock, and L. A. Bariola. 1972.
Limitation of late-season cotton bolls: A possible new
control for pink bollworm, pp. 103-105. In Proc. Beltwide
Cotton Prod. Res. Conf.

Maxwell, F. G., J. N. Jenkins, and W. L. Parrott. 1972.
Resistance of plants to insects, pp. 187-265. In Advances
in agronomy. Am. Soc. Agron., Madison, Wisconsin.

McGaughey, W. H. 1990. Insect resistance to Bacillus
thuringiensis O-endotoxin. pp. 583-598. In R. R. Baker
and P. E. Dunn (eds.) New directions in biological control.
Alan R. Liss, Inc., New York.

Meredith, W. R., and M. F. Schuster. 1979. Tolerance of
glabrous and pubescent cottons to tarnished plant bugs.

Crop Sci.

Meredith, W. R., Jr., C. D. Rainey, M. L. Laster, and R.
R. Bridge. 1973. Agronomic potential of nectariless
cotton. J. Environ. Quality 2: 141-144.

Mussett, K. S., J. H. Young, R. G. Price, and R. D.
Morrison. 1979. Predatory arthropods and their relation¬
ship to fleahoppers on Heliothis resistant cotton varieties in
southwestern Oklahoma. Southwestern Entomol. 4: 35-39.

Namken, L. N., and M. D. Heilman. 1973. Determinate
cotton cultivars for more efficient cotton production on
medium-textured soil in the Lower Rio Grande Valley of
Texas. Agron. J. 65: 953—956.

Namken, L. N., and H. W. Gausman. 1978. Practical
aspects of chemical regulation of cotton plant growth and
fruiting, pp. 23-25. In Proc. Cotton Prod. Mech. Conf.

Namken, L. N., M. D. Heilman, and R. G. Brown.

1975. Flowering intervals, days of initial flower, and
seedling uniformity as factors for development of short-
season cotton cultivars, pp. 80-85. In Proc. Beltwide
Cotton Prod. Res. Conf.

Namken, L. N., M. D. Heilman, and R. H. Dilday. 1979.
Arrangement of sympodia and earliness potential of cotton.
Crop Sci. 19: 620-622.

91

Namken, L. N., M. D. Heilman, J. N. Jenkins, and P. A.
Miller. 1983. Plant resistance and modified cotton
culture, pp. 73-101. In R. L. Ridgway, E. P. Lloyd, and
W. H. Cross [eds.], Cotton insect management with special
reference to the boll weevil. U. S. Dep. Agric., Agric.
Handb. No. 589.

National Academy of Sciences. 1969. Cultural control,
pp. 208-242. In Insect-pest management and control. Nat.
Acad. Sci.

Newell, W. 1906. The remedy for the boll weevil. La.
Crop Pest Comm. C. 3: 20.

Newell, W. 1909. Destroying the boll weevils before they
enter hibernation. Third Biennial Report to the Secretary,
1908-1909. C. 24

Newell, W., and A. H. Rosenfield. 1909. Report upon
variety and fertilizer experiments with cotton in the boll
weevil infested sections of Louisiana. State Crop Pest
Comm, of Louisiana C. 26.

Newsom, L. D. 1974. Pest management: History, current
status and future progress, pp. 18. In F. G. Maxwell and
F. A. Harris [eds.], Proc. Summer Institute of Biol.

Control of Plant Insects and Diseases.

Newsom, L. D., and J. R. Brazzel. 1968. Pests and their
control, pp. 367-405. In F. C. Elliot, M. Hoover, and W.
K. Porter, Jr. [eds.], Advances in production and utiliza¬
tion of cotton: Principles and practices. Iowa State Univ.
Press, Ames, Iowa.

Niles, G. A. 1970. Development of plant types with
special adaptation to narrow row culture, pp. 63-64. In
Proc. Beltwide Cotton Prod. Res. Conf.

Niles, G. A. 1980. Breeding cotton for resistance to insect
pests, pp. 336-369. In F. G. Maxwell and P. R. Jennings,
eds) Breeding plants resistant to insects. Wiley-
Interscience, New York.

Noble, L. W. 1969. Fifty years of research on the pink
bollworm in the United States. U. S. Dep. Agric., Agric.
Handb. 357.

'

i'

Norman, J. W., Jr., K. R. Summy, and J. R. Cate. 1984.
Boll weevil management through cotton stalk destruction,
pp. 216-218. In Proc. Beltwide Cotton Prod. Res. Conf.

Norman, G. G., and N. L. Fritz. 1965. Infrared photog¬
raphy as an indicator of disease and decline in citrus.

Proc. Fla. State Hort. Soc. 78: 59-63.

Orlendorf, W. 1926. Studies of the pink bollworm in
Mexico. U. S. Dep. Agric. Bull. 1374.

Painter, R. H. 1951. Insect resistance in crop plants.

Univ. Press of Kansas, Lawrence, Kans.

Parencia, C. R., C. B. Cowan, and J. W. Davis. 1962.
Relationship of Lepidoptera light-trap collections to cotton
field infestations. J. Econ. Entomol. 55: 692-695.

Parker, R. D., J. K. Walker, G. A. Niles, and J. R.

Mulkey. 1980. The 'short-season effect' on cotton and
escape from the boll weevil. Texas. Agric. Exp. Stn. Bull.
1315.

Perkins, J. H. 1980. The quest for innovation in agricul¬
tural entomology, 1945-1978, pp. 23-80. In D. Pimentel
and J. H. Perkins [eds.], Pest control: Cultural and envi¬
ronmental aspects. Westview Press, Boulder, CO.

Phillips, J. R., and W. F. Nicholson. 1979. Coping with
the budworm/tobacco budworm problem: Communitywide
bollworm management. Proc. Beltwide Cotton Production-
Mechanization Conf.

Pierce, W. D., R. A. Cushman, C. E. Hood, and W. D.
Hunter. 1912. The insect enemies of the cotton boll
weevil. U. S. Dep. Agric., Bur. Entomol. Bull. 100.

i

Plapp, F. W., Jr., and C. Campanhola. 1986. Synergism
of pyrethroids by chlordimeform against susceptible and
resistant Heliothis, pp. 167-169. In Proc. Beltwide Cotton
Prod. Res. Conf.

Presley, J. T., and L.S. Bird. 1968. Diseases and their
control, pp. 347-366. In F. C. Elliott, M. Hoover, and W.
K. Porter [eds.], Advances in production of quality cotton:
Principles and practices. Iowa State Univ. Press, Ames,
Iowa.

92

Reed, W., S. S. La tee f, and S. Sithanantham. 1986.
Compatibility of host-plant resistance and biological control
of Heliothis spp. (Lep.: Noctuidae), pp. 529-535. In E. G.
King and R. D. Jackson [eds.], Proc. workshop on biologi¬
cal control of Heliothis : Increasing the effectiveness of
natural enemies. FERRO/OICD, New Delhi, India.

Reinhard, H. J. 1926. The cotton flea-hopper. Texas
Agric. Exp. Stn. Bull. B-339.

Reynolds, E. B. 1926. The effect of spacing on the yield
of cotton. Texas Agric. Exp. Stn. Bull. 340.

Reynolds, H. T., P. L. Adkisson, and R. F. Smith. 1975.
Cotton insect pest management, pp. 379-443. In R. L.
Metcalf and W. H. Luckmann [eds.], Introduction to
insect pest management. John Wiley & Sons, New York.

Rice, R. E., A. J. Mueller, H. T. Reynolds, and others.

1971. Reduction of pink bollworm moths in southern
California by early crop termination. Calif. Agric. 25: 6-7.

Richardson, A. J., M. M. Crawford, J. H. Everitt, L.

Dron, and R. Summy. 1989. Evaluation of agricultural
management applications in the Rio Grande Valley using
satellite imagery, pp. 816-818. In Quantitative remote
sensing: An economic tool for the nineties. Proc. IGARSS
'89 - 12th Canadian Symp., Vancouver, Canada.

Ridgway, R. L., and E. P. Lloyd. 1983. Evolution of
cotton insect management in the United States, pp. 3-27.

In R. L. Ridgway, E. P. Lloyd, and W. H. Cross [eds.],
Cotton insect management with special reference to the boll
weevil. U. S. Dep. Agric., Agric. Handb. 589.

Roach, S. H. 1981. Emergence of over-wintered Heliothis
spp. moths from three different tillage systems. Environ.
Entomol. 10: 817-818.

Robinson, R. R., J. H. Young, and R. D. Morrison.

1972. Strip-cropping effects on abundance of predatory
and harmful cotton insects in Oklahoma. Environ.

Entomol. 1: 145-149.

Roush, R. T., and R. G. Luttrell. 1987. The phenotypic
expression of pyrethroid resistance in Heliothis and
implications for resistance management, pp. 220-224. In
Proc. Beltwide Cotton Prod. Res. Conf.

Rummel, D. R., and P. L. Adkisson. 1970. Distribution
of boll weevil-infested cotton fields in relation to overwin¬
tering habitats in the High and Rolling Plains of Texas. J.
Econ. Entomol. 63: 1906-1909.

Rummel, D. R., and S. C. Carrol. 1985. Longevity of
overwintered boll weevils (Coleoptera: Curculionidae)
following emergence in spring and early summer. Environ.
Entomol. 14: 127-130.

Rummel, D. R., and K. C. Neece. 1989. Winter survival
of Heliothis zea (Boddie) in cultivated and noncultivated
soil in the southern Texas High Plains. Southwestern
Entomol. 14: 117-125.

Rummel, D. R., J. F. Leser, J. E. Slosser, G. J. Puterka,
C. W. Neeb, J. K. Walker, J. H. Benedict, M. D.

Heilman, L. N. Namken, J. W. Norman, and J. H. Young.
1986. Cultural control of Heliothis spp. in southwestern
U. S. cropping systems, pp. 38-53. In S. J. Johnson, E.

G. King, and J. R. Bradley, eds) Theory and tactics of
Heliothis population management: I - Cultural and biologi¬
cal control. Southern Coop. Series Bull. 316.

Sanderson, E. D. 1903. How to combat the Mexican
cotton-boll weevil in summer and fall. Texas Agric. Stn.

C. 4.

Scales, A. L., and J. Hacskaylo. 1974. Interaction of
three cotton cultivars to infestations of the tarnished plant
bug. J. Econ. Entomol. 67: 602-604.

Schmidt, R. R. 1990. Investigation of mechanisms: the
key to successful use of biotechnology, pp. 1-22, In R. R.
Baker and P. E. Dunn (eds.) New directions in biological
control. Alan R. Liss, Inc., New York.

Schneider, J. C., J. R. Benedict, F. Gould, W. R.
Meredith, Jr., M. F. Schuster, and G. R. Zummo. 1986.
Interaction of Heliothis with its host plants, pp.3-21. In S.
J. Johnson, E. G. King and J. R. Bradley, Jr., eds) Theory
and tactics of Heliothis population management: I - Cul¬
tural and biological control. Southern Coop. Series Bull.
316.

Schuster, M. F. 1979. Insect resistance in cotton, pp.
101-112. In M. K. Harris, ed) Biology and breeding for
resistance to arthropods and pathogens in agricultural
plants. Texas A&M Univ., College Station, Texas.

93

i

i

i

i

Schuster, M. F., and J. L. Frazier. 1976. Mechanisms of
resistance to Lygus spp. in Gossypium hirsutum L, pp. 129-
135. In EUCARPIA/OILB Host Plant Resistance Proc.,
Wageningen, the Netherlands.

Schuster, M. F., M. J. Lukefahr, and F. G. Maxwell.

1976. Impact of nectariless cotton on plant bugs and
natural enemies. J. Econ. Entomol. 69: 400-402.

Schwartz, P. H, 1983. Losses in yield of cotton due to
insects, pp. 329-358. In R. L. Ridgway, E. P. Lloyd, and
W. H. Cross [eds.]. Cotton insect management with special
reference to the boll weevil. U. S. Dep. Agric., Agric.
Handb. 589.

Slosser, J. E. 1978. The influence of planting date on boll
weevil management. Southwestern Entomol. 3: 241-246.

Slosser, J. E. 1980. Irrigation timing for bollworm
management in cotton. Econ. Entomol. 73: 346-349.

Slosser, J. E., and E. P. Boring, III. 1980. Shelterbelts
and boll weevils: A control strategy based on management
of overwintering habitat. Environ. Entomol. 9: 1-6.

Snow, J. W., and J. R. Brazzel. 1965. Seasonal host
activity of the bollworm and tobacco budworm during 1963
in northeast Mississippi. Miss. State Coll. Agric. Exp.

Stn. Bull. 712.

Snow, J. W., J. H. Hamm, and J. R. Brazzel. 1966.
Geranium carolinianum, as an early host for Heliothis zea
and H. virescens (Lepidoptera: Noctuidae) in the southeast¬
ern United States, with notes on associated parasites. Ann.
Entomol. Soc. Am. 59: 506-509.

Stadelbacher, E. A. 1982. An overview and simulation of
tactics for management of Heliothis spp. on early-season
wild host plants, pp. 209-212. In Proc. Beltwide Cotton
Prod. Res. Conf.

Stadelbacher, E. A., H, M. Graham, V. E. Harris, J. D.
Lopez, J. R. Phillips, and S. H. Roach. 1986. Heliothis
populations and wild host plants in the southern U. S, pp.
54-74. In S. J. Johnson, E. G. King, and J. R. Bradley
[eds.], Theory and tactics of Heliothis population manage¬
ment: I- Cultural and biological controls. Southern Coop.
Series Bull. 316.

Sterling, W. L., L. T. Wilson, A. P. Gutierrez, D. R.
Rummel, J. R. Phillips, N. D. Stone, and J. H. Benedict.
1989. Strategies and tactics for managing insects and
mites, pp. 267-325. In R. E. Frisbie, K. M. El-Zik, and
L. T. Wilson [eds.], Integrated pest management systems
and cotton production. John Wiley & Sons, New York.

Stem, V. M. 1969. Interplanting alfalfa in cotton to
control Lygus bugs and other insect pests. Proc. Tall
Timbers Conf. on Ecol. Animal Cont. by Habitat Mgmt. 1:
55-69.

Stem, V. M., R. van den Bosch, and T. F. Leigh. 1964.
Strip-cutting alfalfa for lygus bug control. Calif. Agric.

18: 4-6.

Stem, V. M., R. van den Bosch, T. F. Leigh, O. D.
McCutcheon, W. R. Sallee, C. E. Houston, and M. J.

Garber. 1967. Lygus control by strip-cutting alfalfa.

Univ. Calif. Agric. Ext. Serv. AXT-241.

Stinner, R. E., and J. R. Bradley, Jr. 1989. Habitat
manipulation to increase effectiveness of predators and
parasites, pp. 519-527 In E. G. King and R. D. Jackson
[eds.]. Proceedings of the Workshop on biological control
of Heliothis: Increasing the effectiveness of natural en¬
emies. Far Eastern Regional Research Office, U. S.
Department of Agric., New Delhi, India.

!

Summy, K. R. 1990. Significance of cotton residue as
overwintering habitat for boll weevil in southern Texas, pp.
296-297. In Proc. Beltwide Cotton Prod. Res. Conf.

Summy, K. R., W. G. Hart, and J. R. Cate. 1984. Aerial
surveillance for regrowth cotton in the Lower Rio Grande
Valley of Texas, p. 191. In Proc. Beltwide Cotton Prod.

Res. Conf.

Summy, K. R., W. G. Hart, J. R. Cate, and D. Bar. 1985.
Research on areawide cotton stalk destruction for cultural
control of boll weevil in the Lower Rio Grande Valley, p.
145. In Proc. Beltwide Cotton Prod. Res. Conf.

Summy, K. R., M. D. Heilman, L. N. Namken, and W.

G. Hart. 1986a. Control of boll weevils (Coleoptera:
Curculionidae) through crop residue disposal: Destruction
of subtropical conditions. J. Econ. Entomol. 79: 1662-
1665.

94

Summy, K. R., W. G. Hart, M. D. Heilman, and J. R.
Cate. 1986b. Late-season boll weevil control: Combined
impact of stalk shredding and lethal soil temperatures, pp.
233-235. In Proc. Beltwide Cotton Prod. Res. Conf.

Summy, K. R. 1987. Logistics of cotton stalk destruction
for cultural control of pink bollworm and boll weevil, pp.
50-51. In Proc., 1987 Cotton Insect and Prod. Conf.,

Calif. Coop. Ext. Serv., El Centro, Calif.

Summy, K. R., J. R. Cate, and W. G. Hart. 1988a.
Overwintering strategies of boll weevil in southern Texas:
Reproduction on cultivated cotton. Southwestern Entomol.
13: 159-164.

Summy, K. R., W. G. Hart, M. R. Davis, J. R. Cate, J.

W. Norman, Jr., C. W. Wofford, Jr., M. D. Heilman, and
L. N. Namken. 1988b. Regionwide management of boll
weevil in southern Texas, pp. 240-247. In Proc. Beltwide
Cotton Prod. Res. Conf.

Summy, K. R., J. H. Everitt, W. G. Hart, D. E. Escobar,
and M. R. Davis. 1989. Applications of remote sensing
technology in cotton pest management, pp. 226-231. In
Proc. Beltwide Cotton Prod. Res. Conf.

Thomas, J. R. and R. W. Learner. 1987. Photographic
response of grass canopies to nitrogen stress, pp. 170-183.
In Proc. 11th Bienn. Workshop on Color Aerial Photogra¬
phy and Videography in the Plant Sciences. Am. Soc.
Photogrammetry .

Thomas, J. R. and G. F. Oerther. 1977. Estimation of
crop conditions and sugar cane yields using aerial photogra¬
phy. Proc. Am. Soc. Sugar Cane Tech. 6: 93-99.

Tingey, W. M., T. F. Leigh, and A. H. Hyer. 1975.

Lygus hesperus : Growth, survival and egg laying resistance
of cotton genotypes. J. Econ. Entomol. 68: 28-30.

U. S. Department of Agriculture, Statistical Reporting
Service. 1985. Agricultural statistics. USGPO

Umbeck, P. F., K. A. Barton, J. N. Jenkins and W. L.
Parrott. 1990. Transgenic cotton plants: Field test
preparations and new developments in insecticidal proteins,
p. 635. In Proc. Beltwide Cotton Prod. Res. Conf.

van Emden, H. F. 1965. The role of uncultivated land in
the biology of crop pests and beneficial insects. Sci. Hort.
17: 121-136.

Walker, J. K. 1984. The boll weevil in Texas and the
cultural strategy. Southwestern Entomol. 9: 444-463.

Walker, J. K., and C. W. Smith. Cultural control. In E.

G. King and J. R. Phillips [eds.], Cotton insects and mites:
Their characterization and management. The Cotton
Foundation Reference Series, No. 3, in press.

Walker, J. K, and O. E. Shipp. 1963. Occurrence of
developmental stages of the boll weevil in cotton bolls
during the dormant season of 1962-1963 at College Station.
Texas Agric. Exp. Stn., Prog. Rep. 2282.

Walker, J. K., and G. A. Niles. 1971. Population
dynamics of the boll weevil and modified cotton types.
Texas Agric. Exp. Stn. Bull. 1109.

Walwood, V. T., T. J. Henneberry, L. A. Bariola, and
others. 1983. Insect pest management potential of cotton
produced in narrow-row, short-season cultures in the
irrigated Far West. U. S. Dep. Agric., Agric. Res. Serv.
ARR-W-30.

Wannamaker, W. K. 1957. The effect of plant hairiness of
cotton strains on boll weevil attack. J. Econ. Entomol. 50:
418-423.

Ware, J. O. 1930. Cotton spacing. II. Effect of blooming
on earliness, fruit set, and yield. Ark. Agric. Exp. Stn.
Bull. 253.

Watson, T. F. 1980. Methods for reeducing winter
survival of the pink bollworm, pp. 29-34. In H. Graham
(ed.), Pink bollworm control in the western United States.
U. S. Dep. Agric.,

Science and Education Admin., ARM-W-16.

Watson, T. F., F. M. Carasso, D. T. Langston. 1978.

Pink bollworm suppression through crop termination. J.
Econ. Entomol. 71: 638-641.

95

White, J. R., and D. R. Rummel. 1978. Emergence
profile of overwintered boll weevils and entry into cotton.
Environ. Entomol. 7: 7-14.

Wildman, W. E. 1984. The devastation of a vineyard by
Phylloxera , pp. 111-119. In Proc. 9th Bienn. Workshop
on Color Aerial Photography in the Plant Sciences.

Wilson, A. G. L. 1974. Resistance of Heliothis armigera
to insecticides in the Ord Irrigation Area, North Western
Australia. J. Econ. Entomol. 67: 256-258.

Wilson, F. D, 1982. Present state of the art and science of
cotton breeding for insect resistance in the west, pp. 111-
116. In Proc. Beltwide Cotton Prod. Res. Conf.

Wilson, F. D. 1987. Release of three cotton germplasm
lines. Crop Sci. 27: 820-821.

Wilson, R. L., and F. D. Wilson, 1975. Effects of pilose,
pubescent and smooth cottons on the cotton leafperforator.
Crop Sci. 15: 807-809.

Wilson, R. L., and F. D. Wilson. 1976. Nectariless and
glabrous cottons: Effect on pink bollworrn in Arizona. J.
Econ. Entomol. 69: 623-624.

Wilson, L. T., A. P. Gutierrez, and T. F. Leigh. 1980.
Within-plant distribution of the immatures of Heliothis zea
(Boddie) on cotton. Hilgardia 48: 12-23.

Young, J. H., and R. G. Price. 1968. Effect of irrigation
and submersion in water on pupal survival of the bollworrn.
J. Econ. Entomol. 61: 959-961.

Young, J. H. and R. G. Price. 1977. Overwintering of
Heliothis zea in southwestern Oklahoma. Environ.

Entomol. 6: 627-628.

96

The Problem of Cotton Pest
Resistance to Pesticides in the former
Union of Soviet Socialist Republics
and Rational Systems to Cope with
this Phenomenon

Galina I. Sukhoruchenko1

Cotton is the first crop in the former Soviet Union on
which resistant pest populations have emerged and caused
serious economical losses. The importance of resistance for
this crop is therefore obvious.

The first reports on the emergence of resistance on cotton
appeared in the early sixties when resistance to metasystox
in populations of the two-spotted spider mite, Tetranychus
urticae Koch, was noted in the Yangiulsky region of
Uzbekistan (Ivanova and Kornilov 1964). However, in
1965, observations of resistance in cotton-growing regions
of the Republic revealed a 350-fold level in Fergana valley
(Zainutdinova and Slivochkina 1968). The reports on
losses in efficiency of organophosphates (OPhs) not only in
Uzbekistan but also in some regions of South Tajikistan
occurred during this period (Manina 1968; Smirnova
1968). In the latter case, 200-500-fold levels of resistance
to metasystox, intration, kilval and other chemicals were
shown (Smirnova et al. 1971).

In the early seventies a substantial increase in the resistant
area of Tetranychus urticae to OPhs was observed in
regions sustaining intensive application in Uzbekistan and
Tajikistan. Simultaneously, however, the susceptibility of
T. urticae to this group of toxicants was maintained in areas
with moderate number of pests and low number of applica¬
tions of OPhs (in several districts of the Northern
Uzbekistan, Central Tajikistan, and Azerbaijan) (Smirnova
et al. 1972).

In the late sixties and early seventies, the All-Union
Institute for Plant Protection (VIZR) devised a rotation of
specific contact acaricides (kelthane, neoron, tedion, acrex
and milbex) that not only prevented economical loss in
regions with maximum resistance, but also retarded the
spread of resistance into safer regions. The tactic of
acaricide application has also changed from aerial treat¬
ments with systemic compounds to ground treatments with
contact chemicals by means of a ventilator apparatus. This
permits treatments of fields on the basis of economical
thresholds thereby markedly decreasing the number of
applications of highly-toxic pesticides (Smirnova et al.
1975).

1 All-Union Institute for Plant Protection, Leningrad, former U.S.S.R.

The substantial decrease in the usage of OPhs for control¬
ling Tetranychus urticae in South Tajikistan has consider¬
ably delayed the development of resistance to these chemi¬
cals in populations of Aphis gossypii Glov. and
Acyrthosiphon gossypii Mordv., though an initial stage of
its forming to metasystox and rogor in the early seventies in
the Kulyab region (Ivanova 1975). However, in the early
eighties, the emergence of aphid populations resistant to
these chemicals was observed in the regions of Central
Tajikistan (Kovalenkov 1984). The high levels of resis¬
tance (200-fold) to OPhs in cotton aphid populations in the
Tashkent region of Uzbekistan and their cross-resistance to
the carbamate, pirimor, was noted in the same period
(Hodjayer et al. 1985). An unprecedented outbreak in
cotton aphid reproduction was observed that year in the
farms of South Tajikistan due to intensive use of rogor and
pyrethroids. As a result, at the end of season in the Parhar
region we found 500-fold level of cotton aphid resistance to
rogor, 67-fold level to phosalone and the development of
cross-resistance to the pyrethroids permethrin,
cypermethrin and fenvalerate. The latter were used during
some years against the other key pest, the cotton bollworm,
Helicoverpa armigera Hubner, in this region.

Resistance in populations of the bollworm, H. armigera ,
has been the slowest to appear because of its development
on a number of crops in addition to cotton (chick pea,
tomato, com, potato and alfalfa) which are seldom, if ever,
treated with chemicals in the former U.S.S.R. As a result,
only a part of its population is treated with insecticides.
Moreover, only one or two generations from 3-5 are subject
to intensive treatments, depending on the region and
weather conditions. All this affects selection rate of
resistant genotypes in populations of this insect and retards
the development of its resistance compared to other pests.
Resistance of the bollworm, to DDT compounds was
experimentally produced in the early seventies in
Azerbaijan (Ragimov et al. 1975) and in South Tajikistan
(Smirnova et al. 1972). In the middle of the 1970s, cross
resistance to polydophene, polychlorcamphene and dilor
was recorded in the same regions which coincided with a
sharp decrease in the effectiveness of organochlorine
compounds in the control of the same insect
(Sukhoruchenko 1976). In 1977 the 97-fold level of
resistance of this pest to DDT and the complete loss of
effectiveness of this insecticide were noted in the Parhar
region. It was also discovered that this population was
resistant to sevin (32-fold index) and tolerant to phosalone
(Sukhoruchenko et al. 1980). At the same time, data in

97

i

I

Azerbaijan (Muganskaya region) indicated a considerable
decrease in effectiveness of sevin (Agaeva and Babaeva
1980). Thus, in the late seventies and in the early eighties
the populations of cotton pests in the former U.S.S.R. were
at different stages of development of resistance to pesticides
used (Table 1). In some cases, high levels of resistance
were noted, and as a result, whole groups of pesticides
were lost in practice (e.g., OPhs toward spider mites and
organ ochlorine insecticides toward the boll worm). In other
cases tolerance was observed (bollworm to sevin and
phosalone; cotton aphids to organophosphates) when insect
pests responded to pesticide treatments, but this effective¬
ness was lost rapidly due to increasing levels of resistance.

Duration and intensity of pesticide application had a large
influence on population development in beneficial insects.

In this connection, the most difficult situation concerning
resistance was in cotton-growing areas where the optimum
combination of factors favored increase in the numbers of
pests, and accordingly, increase in pesticide treatments was
observed. The South Tajikistan is one of these regions
where populations of T. urticae, H. armigera, A. gossypii
and Ac. gossypii were found resistant because VIZR chose
this region to develop a system for delaying development of
resistance in populations of several pest species simulta¬
neously. Complicating this system was the fact that ways
to cope with resistance in separate pest species had to be
combined in such a way in order to prevent resistance in the
whole complex of arthropoda.

In keeping with the idea that resistance is a microevolution¬
ary phenomenon based on genotypical changes in popula¬
tions under selective pressure of pesticides, the best strategy
to cope with resistance is management by decreasing the
toxic pressure on the agrobiocenosis of cotton fields.

The second strategic principle based for this system is
management of resistance by noting all positive and
negative sides of pesticide effects on populations of noxious
and beneficial arthropods in cotton fields. Most attention
must be paid to the preservation of natural beneficial
entomofauna that under conditions of less pesticide pressure
would contribute to the reversion of resistance by killing
resistant individuals. We can also combine efficiency of
chemical control with pesticide aftereffects resulting in less
toxic pressure on agrobiocenoses.

Based on these strategies, an integrated resistance manage¬
ment program has been developed by rotating pesticides
based on modes of action. Pesticides have been chosen for
rotation based on the spectrum of cross-resistance and its
possible outcomes. Pesticides from seven chemical groups
have been chosen: organophosphates, organochlorine
compounds, carbamates, pyrethroids, and nitrogen-tin-
sulphur-containing compounds. These are acaricides:
kelthane, neoron, mitran, acrex (isophen), omite, plictran,
sulphur; and the insecticides rogor, phosalone, pirimor,
gardona, dilor, thiodan, etaphos, sevin, kroneton, ambush,
corsar, rovicurt, ripcord, cymbush, sherpa, decis,
sumicidin. Viral (elcar, virin-XS) and biological prepara¬
tions (dendrobacillin, bitoxibacillin) have been included in
the control program as alternative agents to reduce toxic
pressure. The use of every pesticide in the rotation
program was based on economic thresholds and with
constant monitoring of the development of resistance to
pesticides in populations of major pest species.

The need for preservation of beneficial insects was taken
into account while working out the seasonal control
program. An index was developed which showed not only
the initial toxicity of pesticides but also the recovery rate of
beneficial species in treated fields. After long-term studies, j
the following toxicity classes to define the hazard rating
associated with the use of pesticides for beneficial entomo¬
fauna have been established:

'

• first class - slightly toxic pesticides cause 20 % decrease
in numbers of populations - 10 days after treatment
under field conditions;

• second class - moderately toxic pesticides - cause 21-
50% decrease in numbers of populations 10 days after
treatment;

• third class - highly toxic pesticides - cause more than
50% decrease in populations 20 days after treatment.

Pesticides selected for rotation on the basis of the toxicity
rating are classified in the following way (Table 2):

• sevin, rogor and pyrethroids belong to the class which is
dangerous for entomophages despite a slight toxicity for
some species;

98

• gardona, phosalone, thiodan and pirimor belong to the
class of pesticides which are moderately toxic for
zoophages;

• dilor and croneton are the safest insecticides, which are
similar in accord with the toxicity rating to specific
acaricides: kelthane, milbex and biopreparations which
belong to the slightly toxic class.

j|

The ability of broad-spectrum insecticides to stimulate the
development of sucking phytophages, especially the spider
mite, has also been taken into consideration (Fig. 1).

The characteristics of a given pesticide selected for rotation
were used to determine its place in seasonal control
programs. Thus, pesticides slightly toxic or moderately
toxic to zoophages which do not cause an increase in
number of sucking pests (microbiopreparations, phosalone,
dilor, thiodan, croneton) may be used for the first half of
the season (up to the middle of July). For the second part
of the season (from the middle of July to the end of August)
more toxic pesticides may be used which may result in the
increase in populations of sucking species. Specific
acaricides may be safely used for agrocenosis during the
entire vegetative period.

The position in rotation of insecticides used against the
bollworm is determined by their rate of their inhibiting
action on pest populations. The marked reduction of
population indices in survival of preimaginal stages, net
reproductive rate, biotic potential, etc., from usage of
thiodan and pyrethroids compared to gardona and
phosalone as treated in filial generations (Table 3 - example
with thiodan, phosalone and gardone) has shown a need to
use these materials during maximum concentration of the
bollworm in cotton. This time coincides with the develop¬
ment of second generation of the pest, and a reduction of
the number of insecticide treatments in the third generation
will follow.

Six-years of using rotation schedules based on the above-
mentioned approaches have shown that a high protective
effect may be achieved if rogor and phosalone are rotated
with any specific acaricide in zones where spider mites are
not resistant to organophosphates. Organohaloid com¬
pounds (kelthane, neoron, mi t ran) are rotated with mem¬
bers from nitrogen-sulphur tin-containing groups (acrex,
omite, plictran, sulphur compounds) in zones where spider
mites are resistant to organophosphates. Organophosphates
(gardona, phosalone) should be rotated with biopreparations

(dendrobacillin, bitoxibacillin, elcar, virin-XS) in the first
generation of bollworm, chemicals of diene synthesis or
pyrethroids in the second generation, and carbamate sevin
in the third generation. However, gardona, phosalone, and
dilor are not used in regions where an insect pest is resis¬
tant to organophosphorus or organochlorine insecticides.
Any compounds from organophosphates or carbamates may
be used for controlling aphids in regions where resistance is
not observed. Rotation of etaphos and phosalone with
pyrethroids is recommended in areas where resistance is
known.

The main positive effect in the use of resistance manage¬
ment programs is reducing or inhibiting the development of
resistance to pesticides. For almost 10-years of rotation,
resistance to rogor in spider mites has been reduced nearly
10 times; phosalone, kelthane, acrex and omite remain
effective at the same dose; usage of plictran has increased
by 14.5 times, but its effectiveness has not been reduced
(Table 3). The sensitivity of Acyrthosiphon gossypii to
aphicides remains at the same level (Table 3). However, an
increase in resistance of Aphis gossypii to phosalone and
rogor has been observed, but the efficiency is not decreas¬
ing under these indices. Resistance to DDT in the boll¬
worm was reduced from 97- to 7-fold level, and resistance
to sevin - from 29.9 - to 0.7 level. Resistance to phosalone
has remained the same, and sensitivity to pyrethroids in
bollworm populations has decreased similarly to DDT.

The decrease in indices of resistance to pesticides applied in
pest populations has shown a remarkable turnaround within
ten years. It has resulted in the reduction of toxic pressure
on agrobiocenosis to an extent that beneficial arthropods
have recovered in high enough numbers to aid in control of
pests. Numbers of treatments per season using resistance
management has been no more than five during years with
low populations of the pest.

These positive aspects of insecticide control improve the
economic situation. The average annual benefit has been
59.84 rbl/ha increase through resistance management. In
1984 this program was approved by the Scientific and
Technical Council of the Ministry of Agriculture in the
former U.S.S.R. In 1985, "Recommendations on rational
rotations of insecticides, acaricides and biopreparations in
control of resistant populations of cotton pests" was
published. From 1985 until 1988 the rotation of pesticides
was introduced in farms of South Tajikistan on the area of
75, (XX) hectares of cotton.

99

In addition to an educational program, providing adequate
pesticides is necessary for the program to work. However,
the plan worked out 10 years ago has undergone substantial
changes. Acrex, pirimor and sevin were banned from use
because of environmental considerations. Thiodan may be
banned for the same reasons; thus, we must search for
efficacious alternatives.

The intensive use of organophosphates and pyrethroids for
the control of the bollworm in 1989 resulted in the high
levels of resistance to rogor and cross-resistance to some
pyrethroids. Thus, it is also necessary to search out
aphicides among pesticides from other groups including
microbiopreparations; good results were obtained with
pegas - a member of the thiourea. The field bug Lygus
pratensis L. is a problem at present time and
biopreparations are being tested for this insect as substitute
insecticides; dipel, legend, and aim are examples of
products of interest for resistance management of the
bollworm.

Because emphasis has been placed on the usage of pyre¬
throids with acaricidal activity (e.g., dan i to!, mavric, and
rufast), resistance management programs have changed
considerably these past years. We feel that in the future
this program will improve as advancements in biological
control are made. Knowledge of physiological, biochemi¬
cal and genetic mechanisms of resistance will be the basis
for more fundamental changes in similar control programs.

References Cited

Agaeva, Z. M., and B. Kh. Babaeva. 1980. Ob
ustoichivosti khlopkovoi sovki k insectitsidam v
Azerbaidzhanskoi SSR. (In Russian). Sostojanie i
perspektivi razvitia po predotvrascheniu rezistentnosti u
vreditelei, vozbuditelei boleznei i somjakov k-pestitsidam i
razrabotka effektivnikh mer borbi s bakterialnimi
boleznjami rastenii: Tezisi dokladov Pjatogo Vsesojuznogo
soveschania, Erevan, 1980. Leningrad. 31-36.

Ivanova, G. P. 1975. Ustoichivost khlopkovikh tlei k
fosfororganicheskim insektitsidam i preduprezhdenie eje
razvitia. (In Russian). Khimia v selskom khozjaistve. 13:
47-50a.

Ivanova, N. A., and V. G. Kornilov. 1964. O privikanii
pautinnogo klescha na khlopchatnike k merkaptofosu. (In
Russian). In Trudi VNII Zaschiti Rastenii. 20: 12-17.

Khodzhaev, Sh. M., S. A. Roslavtseva et al. 1985.
Chuvstvitelnost khlopkovoi tli k insektitsidam. (In Rus¬
sian). Khimia v selskom khozjaistve. 23: 33-35.

Kovalenkov, V. G. 1984. Biometod v integrirovannoi
zaschite khlopchatnika. (In Russian). Zaschita rastenii. 8:
12- 14a.

Manina, L. I. 1968. Uslovia obrazovania ustoichivik k
jadam populjatsii pautinnogo klescha. (In Russian). Tezisi
dokladov soveschania po rezistentnosti kleschei k
akaritsidam. Leningrad. 29-30.

Ragimov, Z. A., R. F. Papijan, et al. 1975. Vozmozhnost
razvitia resistentnosti u khlopkovoi sovki ( Chloridea
armigera Hb.) k polidofeau i sevinu. Rezistentnost
vreditelei i vozbuditelei boleznei k khimicheskim sredstvam
zaschiti rastenii: Tezisi dokladov Chetvertogo
Vsesojuznogo soveschania. Leningrad. 1975. Moscow.
53-56.

Smirnova, A. A. 1968. Sostoianie vo osa ob ustoichivosti
pautinnogo klescra na khlopchatnike k fosfororganicheskim
akaritsidam. (In Russian). Tezisi dokladov soveschania po
rezistentnosti kleschei k akaritsidam. Leningrad. 3-6.

Smirnova, A. A., V. G. Kornilov et al. 1970.

Ustoichivost pautinnogo klescha k fosfororganicheskim
preparatam v osnovnikh zonakh khlopkosejania k
khimicheskim sredstvam borbi s ustoichivimi populajtsijami
vreditelej. (In Russian). Vtoroe soveschanie po
rezistentnosti vreditelei k khimicheskim sredstvam zaschiti
rastenii: Tezisi dokladov. Leningrad. 3-12.

Smirnova, A. A., V. G. Kornilov et al. 1972. O razvitii
ustoichivosti k DDT khlopkovoi sovki ( Heliothis armigera )
v Uzhnom Tadzhikistane. Kratkie tezisi dokladov Tretjego
soveschania po rezistentnosti vreditelei k khimich— skim
sredstvam zaschiti rastenii. Leningrad. 87-89.

100

Smirnova, A. A., V. G. Kornilov et al. 1972. Razviyie
ustoichivosti k fosfororganicheskim akaritsidam u
obiknovennogo pautinnogo klescha na khlopchatnike i
khimicheskie meropriyatia po borbe s ustoichivimi
populjatsijami vreditelja'. Trudi VNII. Zaschiti rastenii.

35: 189-208.

Smirnova, A. A., G. V. Smirnova et al. 1975. Ispitanie
insektitsidov (karbofosa, 40%; ritsifona, 30%; dilora,

15%; antio, 70%; galekrona, 25%; despirolja, 5%;
elokrone, 50%; (metodom UMO). (In Russian). Itogi
gosudarstvennikh ispitanii insektitsidov i akaritsidov v 1974
g. Moskva. 110-132.

Sukhoruchenko, G. I., O. D. Nijazov et al. 1976. Vlijanie
khimicheskikh obrabotok na vrednuju i poleznuju faunu
khlopchatika. (In Russian). In Ecologia i khozjaistvennoe
znachenie nasekomikh Turkmenii. Ashkhabad. 46-61.

Sukhoruchenko, G. I., A. A. Smirnova et al. 1980.
Effektivnost piretroidov v borbe s ustoichivimi popul
atsijami khlopkovoi sovki. (In Russian). Sostojanie i
perspektivi razvitia nauchnikh issledovanii po
predotvrascheniu rezistentnosti u vreditelei, vozbuditelei
boleznei i somjakov k pestitsidam i razrabotka effektivnikh
mer borbi s bakterialnimi boleznjami rastennii: Tezisi
dokladov Pjatogo Vsesojuznogo soveschania. Erevan.

1980. Leningrad. 43-45a.

Zainutdinova, K. M., and L. A. Slivochkina. 1968.
Ustoichivost pautinnogo klescha kak resultat
mnogokratnogo primenenia akeritsidov. (In Russian).

Tezisi dokladov soveschania po rezistentnosti kleschei k
akeritsidam. Leningrad. 19-20.

Table 1. Reversion and retarding the resistance in cotton pest populations after rotation of pesticides

Levels of resistance

Pest

Pesticide

1977-1978

1986-1987

Tetranychus

Rogor

250.0

21.0

urticae

Phosalone

39.0

20.0

Kelthane

22.0

32.0

Acrex

4.3

2.3

Omite

4.0

0.7

Plictran

1.7

14.5

Aphis

Rogor

0.5

5.7

gossypii

Phosalone

3.4

22.0

Kroneton

1.3

1.0

Pirimor

1.0

1.0

Acyrthosiphon

Rogor

3.5

3.6

gossypii

Phoselone

8.5

8.5

Kroneton

1.0

1.0

Helicoverpa

DDT

97.3

7.0

armigera

Sevin

29.9

0.7

Phosalone

4.0

5.4

Gardona

1.0

1.2

Thiodan

1.0

2.0

Ambush

1.0

0.1

Decis

1.0

0.1

Gymbush

1.0

0.1

Sumicidin

1.0

0.3

102

Table 2. The degree of hazard of pesticides to beneficial arthropods

Classes of hazard for

Pesticide

Adonia

variegata

Onus

niger

Nabis Deraeocoris Aeo/othrips
pa lifer punctulatus intermedius

Chrysoper/aAraneina Trichogramma
carnea sp. evanescens

zoophages

Rogor

3

2

3

3

3

3

1

3

2.6

Phosalone

1

2

2

2

3

2

1

0

1.8

Gardona

3

1

1

2

3

0

1

0

1.4

Dilor

1

1

1

1

1

0

1

0

0.8

Thiodan

3

0

3

2

3

1

2

1

1.8

Kelthane

2

0

0

0

0

0

0

1

0.6

Milbex

1

0

2

0

0

2

1

1

0.9

Sevin

3

3

3

3

3

2

3

3

2.9

Kroneton

0

1

1

2

1

1

0

2

1.0

Pirimor

1

2

1

2

2

2

2

2

1.8

Ambush

3

2

3

2

3

3

3

2

2.6

Cymbush

3

2

3

1

3

2

3

3

2.5

Decis

3

2

3

2

3

3

3

3

2.8

Sumicidin

3

2

3

2

3

1

3

3

2.5

Biopreparations 3

1

2

0

0

2

2

0

1.3

Table 3. Demographic characteristics of a population of H. armigera untreated and treated by insecticides.

Population Parameters

Pesticide

Generation

R

o

T

r

m

Thiodan

II

9.32

44.5

0.0503

III1

0.00

0.0

0.0000

Untreated

II

37.30

33.5

0.1090

III1

51.58

41.4

0.0950

Phosalone

I

27.23

41.3

0.0818

II1

2.24

43.5

0.0190

Untreated

I

99.78

41.7

0.1111

II1

2.67

46.4

0.0250

Gardona

I

95.94

39.4

0.1168

II1

24.98

34.5

0.0935

Untreated

I

101.40

36.3

0.1261

II1

102.71

34.2

0.1362

1 Daughter generation

103

Fig. 1. The dynamics of two spotted spider mite through insecticide treatments. Numbers on the y-axis are
changes in population sizes (in natural logarithms) compared with the initial size.

104

Prospects and Status for
Development of Novel Chemicals for
Integrated Pest Management in
Cotton

Julius J. Mean1

Abstract

Increasing concerns regarding food and feed safety, quality
of groundwater and surface water, and resistance of insects
to chemical pesticides have provided added impetus to the
development of integrated pest management (IPM) systems
for cotton. The chemical industry is cognizant of these
concerns and is directing increasing research resources to
develop more selective, IPM-compatible pesticides through
more rational approaches than those used previously. This
paper discusses the status of current chemical control
practices in cotton and describes several promising candi¬
date control agents that would be compatible with IPM.

The developing use of semiochemicals for insect control is
also briefly discussed in terms of the potential of the IPM
approach to reduce the application volume and frequency of
conventional insecticides.

Introduction

Com, soybeans, cotton, and cereals constitute the most
important crop plants in the United States. The Economic
Research Service (ERS) of the U.S. Department of Agricul¬
ture (USDA) forecasted that approximately 5.5 million
hectares (ha) would be planted to cotton in the United
States in 1990 (ERS 1990).

Of the 64 million pounds of insecticides used on major field
crops in 1990 in the United States, 19 million pounds, or
30 percent was applied on cotton (ERS 1990). The latter
figure translates to 1.5 kg of active ingredient per hectare
(ha). Synthetic pyrethroids (SP's) constitute about one-half
the volume of all insecticides applied on cotton, and their
application rates are low, ranging from O.OZto 0.22 kg/ha
(Menn et al. 1989). This high use of SP's on cotton
assumes even greater significance, considering that in some
parts of the Cotton Belt, especially in the Southeast, they
may be applied as many as 10 times per growing season.
The primary targets of these applications in order of
importance are the boll weevil [Anthonomus grandis
grandis Boheman] and the "Heliothis" complex, which

1 Plant Sciences Institute, USDA, ARS, Beltsville, Maryland

comprises the tobacco budworm [ Heliothis virescens
(Fabricius)] and the bollworm [Helicoverpa ( Heliothis ) zea
(Boddie)].

Despite intensive efforts in the integrated pest management
(IPM) arena (Frisbie 1985), heavy pesticide use in cotton
still continues. Perhaps the greatest difficulty in develop¬
ing IPM strategies that incorporate the use of conventional
chemicals is rooted in the great diversity of cotton insect
pests. These include not only the boll weevil and
"Heliothis" complex but also intermediary and more
regional pest species, such as the sweet potato white fly
[Bemisia tabaci (Gennadius)], Lygus species, including the
tarnished plant bug [Lygus lineolaris (Palisot de Beauvois)],
the pink bollworm [Pectinophora gossypiella (Saunders)],
and spider mites. Complicating IPM efforts is the ability of
the "Heliothis" complex to migrate long distances and the
invasion of pest species from other croppings. The
chemical industry, the major discoverer and developer of
chemical insecticides, has long recognized cotton as a
primary target for field testing new insecticides. In fact,
many insecticides in the United States were first registered
for use on cotton, due to the diversity of insect pests and
relative ease of registration. However, because cottonseed
oil is classified as a raw agricultural food commodity,
registration can be more readily obtained for insecticides to
be used at the preflowering stage of cotton than for those to
be used at later stages. Developing first prebloom uses
would avoid the need for extensive compilation of data on
residue in cottonseed oil. Such a compilation is required
for full tolerance registration by the U.S. Environmental
Protection Agency (EPA).

Current Chemical Control in Cotton

An estimated 35 chemical insecticides are in current use to
control various pests of cotton in the United States
(Suguiyama and Osteen 1988). The insecticides in greatest
use, based on 1988 data, are shown in Figure 1. Parathion
(22.7%), methyl parathion (10.7%), azinphos-methyl
(4.6%), dicrotophos (6.4%), acephate (2.6%), and SP's
(21 %) account for 77.2% of all insecticides applied on
cotton (C. Giedraitis, personal communication, 1990).

While the organophosphorus esters (OP's) are still the
primary chemical weapons against the boll weevil and
sucking insects, the SP's have in recent years become the
major weapons against the "Heliothis" complex. The SP's

105

appeared to be the ideal insecticides: effective at low rates
and yet relatively safe to beneficial insects; fairly nonpersis-
tent in the environment; and low in cost, based on use
pattern.

However, as early as 1985-1986, control failures with SP's
were reported in Texas (Bull and Menn 1990). It appears
now that genes for SP resistance are present in populations
of the "Heliothis" complex in the United States, and it is
likely that resistance to these insecticides will continue to
build as a consequence of intensive use pattern (Bull and
Menn 1990). Although the resistance phenomenon por¬
tends economic difficulties for cotton production, it may in
the long-term catalyze the development of biological
control agents and more selective chemical insecticides that
can be used in IPM systems for cotton.

Discovery of New Chemical Insecticides

Discovery and successful market introduction of a synthetic
chemical pesticide has been likened to the development and
successful launching of a spaceship. Both projects are
highly creative, interdisciplinary, costly, and unpredictable.
In the case of the pesticide, more competition and time are
involved due to the uncertainties of the registration process.

Details of the research and development process to bring to
market a new pesticide were extensively reviewed in
numerous articles, including those by Menn (1983),
Braunholtz (1981), and Menn and Henrick (1985).

Table 1 shows the four stages for producing and bringing
one successful candidate to market. It had been estimated,
based on previous publications (Menn and Henrick 1981;
Braunholtz 1981), that only one in 20 thousand synthetic
compounds will reach the marketplace. According to R.F.
Flattum (personal communication 1990), this ratio has
escalated to one in 40 thousand compounds in 1990. In the
late 1970s, the activities shown in Table 1 were estimated
to cost upward of $21 million (Menn 1983). These costs
have escalated to $30 to $50 million in 1990 (Flattum,
personal communication, 1990). In addition, the cost of a
manufacturing plant would also range from $20 to $40
million. Therefore, the total investment in a new insecti¬
cide could reach upward of $70 million with an average
elapsed time to market of 96 months. Because of these high

costs and long time span, development of a pesticide is a
high-risk venture, and only the large multinational compa¬
nies can afford to be in this enterprise today.

Due to societal, regulatory, and environmental consider¬
ations and pressures, the industry has been emphasizing
development of more selective insecticides that harmonize
control with relative safety to beneficial insects and other
non target organisms. Due to these constraints, very few
insecticides have appeared on the market in recent years. In
fact, in 1990, there is not a single new insecticide undergo¬
ing registration review for cotton in the United States.

A cardinal requirement for continued use of existing
insecticides or introduction of new compounds is that their
direct and/or indirect effect on beneficial insects and mites
be determined (Croft and Brown 1975; King 1976; Croft
1990; and Elzen and King 1990).

Laboratory and field evaluations are important for provid¬
ing information on the overt and latent capability of
insecticides to reduce populations of beneficial insects and
mites. Metabolic studies are also important, because they
can provide information for predicting the potential hazard
of an insecticide and/or its metabolites to these beneficial
organisms. Krieger et al. (1971) have shown that certain
beneficial insects have more limited detoxifying capacity
than their lepidopterous hosts. The latter have evolved a
greater variety of oxidative, reductive and hydrolytic
enzymes capable of processing a much broader array of
plant constituents than their parasites and/or predators.

Generally, OP insecticides have taken a heavier toll of
beneficial insects than SP's and formamidines. SP's if
properly used in late season spare beneficials, since they
occur in reduced numbers in late season. Beneficials are of
prime importance in the early season, when host popula¬
tions are still relatively low. Chlordimeform (Galecron)
was a particularly useful insecticide and miticide in IPM
programs since it is a behavior modifying agent that has
an indirect ovicidal and larvicidal action (Hollingworth and
Lund 1982). However, chlordimeform was recently
deregistered in the United States. Possibly a related
compound, amitraz (Mitac, Ovasyn), may replace some of
the uses lost to chlordimeform. Amitraz is active against a
broad range of spider mites, and against the whitefly
[Bemisia tabaci (Genn.)]; it is a first-instar larvicide and
ovicide against the "Heliothis" complex (Peregrine 1989).

106

Approaches to Discovery of Selectivity

Most synthetic chemical pesticides were discovered through
empirical synthesis coupled with broad-spectrum short¬
term- exposure screening tests on target insects. Screening
tests that include nontarget, beneficial insects and mites are
conducted only sparingly in industry, and most of the
useful information on the response of beneficials to new
insecticides has come from universities and the USDA,
Agricultural Research Service (ARS). An example of such
information was reported by Plapp and Bull (1978), among
others, who showed that SP's were reasonably well
tolerated by an important predator of the "Heliothis"
complex in cotton. This information was most helpful in
developing a midseason to late-season IPM strategy for
applying SP's on cotton.

The growing emphasis on selective insect control that is
compatible with IPM practices has had a marked impact on
the chemical industry. It has strongly influenced research
thinking and redirection of resources. As a result, more
directed approaches for discovering selective insecticides
are being used.

These approaches include (1) the synthesis of highly active
analogs of biologically active compounds, the direction of
the synthesis being guided by the results of quantitative
structure-activity relationship (QSAR) analyses, (2) the
discovery of natural-product insecticides and also the
synthesis of their highly active analogs, and (3) use of a
biorational approach to design and synthesize insecticides.
These approaches have been described in detail by Morrod
(1981) and Magee et al. (1985).

Directed Synthesis

Once a highly active new class of insecticides appears in the
patent literature, intensive efforts usually follow in compet¬
ing industrial laboratories to synthesize analogs with even
greater bioactivity through directed research. A pertinent
example is the synthesis of selective chitin-synthesis
inhibitors that are analogs of diflubenzuron and other
benzoylphenylureas (Verloop and Ferrell 1977) (Figure 2).

Diflubenzuron, an insect growth regulator that interferes
with chitin biosynthesis in insects, was introduced by
Duphar B. V. in 1973. It eventually found a practical
niche in cotton production by controlling the boll weevil

and armyworms. In subsequent years, thousands of analogs
were synthesized in industry with varying degrees of
efficacy and selectivity. In the 1980s, a new compound
that retains the benzoylphenylurea core was introduced by
Duphar B. V. under the name Andalin (Figure 2). This
compound is also targeted as a cotton insecticide and
miticide that is compatible with IPM. It is applied at the
rate of 40-50 grams ai/ha.

Another example of directed synthesis involves a new
thiourea- derived insecticide and miticide that is active
against hemipterous insects in cotton and compatible with
IPM. This compound, Diafenthiuron (Figure 3), was
introduced by Ciba-Geigy AG. Interestingly, this com¬
pound exerts its action via photoconversion in sunlight and
metabolism in insects and mites to the more reactive
carbodiimide (DFCD). DFCD is more toxic than the
parent compound to bulb mites [Rhizoglyphus echinopus
(Fumouze and Robin)] and two spotted spider mites
[Tetranychus urticae Koch]. Kadir and Knowles (1990)
reported that DFCD was a potent agonist of octopamine-
activated adenylate cyclase. Thus, they concluded that
DFCD is an octopamine agonist.

Undoubtedly, other new candidate insecticides are being
tested worldwide on various crops, including cotton.
However, if we exclude SP's, only very few representatives
of novel classes of insecticides have been described. Most
likely, the paucity in new insecticide discovery is due to
increased confidentiality because of patent competition and
difficulty in discovering selective, cost-effective insecti¬
cides that harmonize with IPM.

Natural Products

Natural products derived primarily from plants and micro¬
organisms have provided a vast storehouse of bioactive
substances. However, only a few of these have been
developed as agricultural insecticides to date.

Synthetic Pyrethroids

The best known and economically most important com¬
pounds based on natural products models are the Synthetic
Pyrethroids (SP's): they had their genesis in pyrethrins I
and II (Elliott 1989). Elliott and co-workers in England

107

have shown that the photolabile center of pyrethroids could
be chemically modified to form compounds with much
greater stability in air and light, but with the same or
enhanced high insecticidal activity, especially against
lepidopterous larvae. The SP's also displayed low mamma¬
lian toxicity and low to moderate toxicity against
hemipteran predators, but relatively high toxicity to
phytoseiid mites (Croft and Whalon 1982).

Currently, the SP's are the premier cotton insecticides in
the world. In 1987, the value of the agricultural SP market
reached $1.5 billion, which represented approximately 25%
of the global sales of insecticides (Morton and Collins
1989). In the United States, the pyrethroids are the
dominant group of insecticides registered and used against
lepidopterous larvae in cotton, primarily to control the
"Heliothis" complex. Currently, ten SP's [bifenthrin,
cyfluthrin, cyhalothrin, cypermethrin, esfenvalerate,
fenvalerate, flucythrinate, karate (lambda-cyhalothrin),
permethrin, and tralomethrin] are recommended for
application in midseason to late season. If used according
to this schedule, there is a better likelihood to delay the
development of resistance to these valuable insecticides and
protect beneficial insects in the early season (Bull and Menn
1990; Elzen and King 1990).

Avermectin Insecticides and Miticides

The avermectins are a novel class of lactones that are
natural products derived by fermentation of the soil
microorganism Streptomyces avermitilis. The discovery of
these microbial products was reported by Campbell et al.
(1984). Members of these macrocyclic lactones were
shown to possess outstanding anthelmintic (ivermectin) and
insecticidal and miticidal activities (abamectin).

The mode of action of the avermectins is different from that
of most other insecticides, including the OP's, carbamates,
and the SP's. The avermectins are gamma-amino-butyric
acid (GABA) agonists that cause paralysis by inhibiting
signal transmission at the neuromuscular junction and
potentiating a GABA-like action on the GABA receptor/
chloride ion channel complex (Campbell 1981). Interaction
with the GABA receptor is thought to be a major compo¬
nent of the mode of action of these compounds. However,
other potential sites of action are currently being investi¬
gated. The avermectins are biologically active in all
organisms having a gabaergic system. Selectivity and

differential toxicity of these compounds are a function of
uptake, differential metabolism, and mode of delivery. The
structures of avermectin B1 (abamectin) and an analog,

EMA (MK-243) are shown in Figure 4. Abamectin
consists of approximately 80 % of component a, where R =
C2H5, and 20% of component b, where R = CHr
Abamectin is primarily a broad-spectrum miticide, provid¬
ing long residual control of spider mites on cotton. The
miticide is active through translaminar foliar uptake. It
disappears rapidly from the surface of cotton leaves but is
retained in the epidermal cells, the feeding site of spider
mites. Because of this translocation, abamectin is rela¬
tively nontoxic to predators and parasites and fits well in
IPM programs. In field trials, mite control and yields were
considerably better in cotton fields treated with abamectin
than with dicofol or comite.

In the course of chemical modification studies on abamectin
to optimize activity against lepidopterous larvae, it was
discovered that substituting an amino group for an OH
moiety at the 4"-position in the terminal disaccharide
moiety of abamectin (Figure 4) produced a remarkable
increase in selective toxicity against major lepidopterous
larval pests of cotton (Dybas et al. 1989).

Dybas et al. (1989) determined the toxicities of EMA (MK-
243), abamectin, two carbamates, and two SP's in tests
with the larvae of three major lepidopteran cotton insect
pests (Table 2). The LC^ values showed that EMA was
significantly more toxic to the three lepidopteran species
than the other pesticides. It is remarkable that against
CBW, cypermethrin was less toxic than EMA by a factor of
0.005; that is, EMA was 205 times as toxic as
cypermethrin. As a contact insecticide, however, EMA
was only slightly more toxic than the SP's. Nevertheless,
due to its unique mode of action, EMA should be exten¬
sively tested under field conditions and in various formula¬
tions for selectively controlling lepidopterous larvae in
cotton.

Azadirachtin

Azadirachtin (AZ) is a tetranortriterpenoid (limonoid). Its
structure was fully elucidated by Bilton et al. (1987) and is
shown in Figure 5. AZ is the most important insect-active
component of the seeds of the neem tree, or Indian lilac
[. Azadirachta indica ], belonging to the Meliaceae (ma¬
hogany) family. There is a long history of folklore

108

medicine and insect control associated with the neem tree.
The properties and potential of neem in insect management
was recently reviewed by Schmutterer (1990).

AZ acts as an insect growth regulator apparently by
blocking the action of the moulting hormone ecdysone. It
also has a phagorepellent action in locusts and other insects.
In addition to AZ, other neem seed kernel extracts (NSKE)
and extracts of neem leaves have been shown to produce a
variety of repellent, oviposition inhibition, feeding deter¬
rent and morphogenetic effects in various insects. Pres¬
ently, only the most active insecticidal ingredient, AZ, has
been registered for nonfood use on ornamental crops.

Under the tradename Margosan-O, AZ is used against the
greenhouse whitefly [Trialeurodes vaporariorum
(Westwood)], the sweetpotato whitefly, mealybugs,
leafminers, and several lepidopterans at rates of 20 to 50 g
AZ/ha. A strain of the sweetpotato whitefly has become a
major pest of cotton in the western United States and
appears to tolerate most insecticides. There is a possibility
that it was introduced from the Middle East, where it is
already the major pest of cotton.

Price and Schuster (1990) reported promising control of
second- stage nymphs of the sweetpotato whitefly with
single foliar spray applications of 220 ppm AZ Margosan-O
to poinsettias. Promising control was also reported on
tomatoes by Schmutterer (1990). These results suggest that
AZ should be extensively tested for whitefly control on
cotton.

Neem-based pesticides have proved to be relatively harm¬
less to many beneficial insect parasites and predators and
also to honey bees. The weak contact toxicity and selective
mode of action are most likely significant contributors to
this bioactivity profile (Schmutterer 1990). Neem-based
pesticides also have a favorable acute toxicity profile in
warm-blooded animals and fish, and a rapid disappearance
rate in the environment.

Bacillus thuringiensis ( B . t. )

Great hopes were generated in past years to control the
"Heliothis" complex in cotton with spray formulations of
spores of Bacillus thuringiensis Berliner var. Kurstaki.

Field trials conducted over several seasons with strains HD-
1 and HD-263 failed to provide economic control (Menn,
King, and Coleman 1989). Although B.t. is a relatively

effective stomach poison, a residual contact toxicant is
required to effectively control the ' 'Heliothis" complex.
Heliothis larvae feed sparingly on foliage and rapidly
invade squares and bolls, thus escaping surface residues.

Renewed interest in B. t. has been generated recently
spurred by advances in biotechnology. Through genetic
engineering, the B.t. delta endotoxin gene has been success¬
fully expressed in cotton plants, producing transgenic plants
that are toxic to lepidopterous larvae feeding on cotton.
Currently field trials are under way in Mississippi to
evaluate protection of transgenic cotton plants from attack
by first instar larvae of Heliothis virescens. These trials are
being conducted cooperatively in the United States by the
Monsanto Agricultural Company, St. Louis, Missouri, and
the USDA, ARS, Crop Science Research Laboratory,
Mississippi State, Mississippi (J.N. Jenkins, personal
communication).

Biochemical Design

Most insecticides were discovered by the empirical or
random synthesis-and-screen approach. However, as
discussed earlier, more and more resources and attention
are being directed toward biochemistry and physiology in
designing novel pesticidal molecules. It was this cogent
biorational approach that led to the design and synthesis of
juvenoids that disrupt insect development (compounds
mimicking the action of juvenile hormones) (Menn and
Henrick 1981). Unfortunately, exposure to sunlight caused
many of the juvenoids to become unstable and inactive
against insects attacking cotton.

A recent breakthrough was the discovery of the first potent
nonsteroidal agonist of ecdysone (Wing et al. 1988). This
compound, RH-5849 [1,2-dibenzoyl-l-tert-butylhydrazine]
(Figure 6), was discovered as a result of a long-term
biorational program to discover antijuvenile or moulting
hormone inhibitors at the Rohm and Haas research labora¬
tories.

When injected into larvae of Manduca sexta (L.), RH-5849
was 50 times more active than 20-hydroxyecdysone in
initiating premature moulting. It was even more effective
in larval diets where it was 670 times more active than 20-
hydroxyecdysone in initiating premature moulting. Similar

109

activity was determined in members of three other lepidopt-
eran families; Noctuidae, Pyralidae, and Pieridae (J.A.
Svoboda, personal communication, 1989).

Biochemical studies (Wing 1988) established that RH-5849
is a true ecdysone agonist and does not compete directly
with ecdysone for target tissues. Although similar in
chemical structure to the benzoylphenylureas that interfere
with chitin biosynthesis, RH-5849 has a markedly different
mode of action.

The novel and selective mode of action of RH-5849 which
results in rapid cessation of feeding regardless of age or
instar suggests that this compound, and possibly more
active analogs, should be field tested in IPM programs for
cotton. However, like the avermectins, this class of
ecdysone agonists may suffer from the drawback that they
are primarily active via ingestion. Experience has shown
that the most successful insecticides against lepidopterous
pests of cotton exert their primary action via contact.

Nevertheless, since seldom do such exquisitely selective
compounds come along, every opportunity should be
afforded to exploit their unique action through further
synthesis of analogs and extensive field testing in cotton.

Semiochemicals

Semiochemicals, including pheromones, attractants, feeding
stimulants and arrestants have been used in cotton and other
crops for a number of years primarily for monitoring
population densities of several pest species.

Only very recently has the technology of using
semiochemicals to disrupt mating and to mass-trap insect
pests become economically feasible for controlling two
major North American pests of cotton— the boll weevil and
the pink bollworm.

In the Mississippi Delta cotton region, a mass-trapping
system has proven economically feasible for controlling
boll weevils in the early season. McKibben and co-workers
(McKibben et al. 1990) developed an attracticide trap. The
ingredients consist of 40 ng grandlure, feeding stimulants,
cottonseed oil, and 1 g cyfluthrin formulated in polyvinyl
chloride. (Grandlure, the boll weevil pheromone, is made
of four unique alicyclic compounds— two alcohols and two
aldehydes.) To this mixture are added a natural shellac, as

a binder, and a green pigment. This mixture is applied to
wooden stakes placed vertically in the field at the rate of
two stakes per ha. This trapping system attracts both sexes
and has proved highly efficient in mass reduction of weevil
populations. The traps need to be recharged every 4 weeks
and are employed during a 2-month period. This selective
population management system has the potential to become
important for controlling boll weevils in the early season
and may significantly reduce the need for large-scale
treatments with
insecticides.

In recent years, the cotton acreage in Arizona, the Imperial
Valley of California, and the northeastern valleys of
Mexico have drastically decreased or disappeared altogether
because of the ravages of the pink bollworm. A new
technology based on mating disruption that had its genesis
in the pioneering research of Shorey and co-workers
(Shorey et al. 1976) is now coming to practical realization.
Large-scale field trials are currently under way in the
Parker Valley of Arizona to evaluate the commercial
potential of several formulations of gossyplure to disrupt
early-season mating and, thus, preclude the need for
treating the cotton fields with insecticides early in the
season (Brosten and Simmonds 1990).

Staten and co-workers (1987) obtained promising field
results in the Imperial Valley of California and in the
Mexicali Valley, Mexico, with hand-applied PBW-ROPE
dispensers containing 78 mg of 96% (Z,Z) and (Z,E)
isomers of gossyplure in a ratio of 49:51 respectively.

Field laborers placed the dispensers around the stem of
individual plants at the first-flower-bud stage and at a rate
of 1000/ha. The gossyplure remained active for approxi¬
mately 60 days. Use of the dispensers reduced the need for
insecticide applications byu 41 % and resulted in a 97 %
reduction in trap catches of the moths in August.

It is evident from the foregoing that semiochemical technol¬
ogy may have an economic future in IPM programs. It
may help us improve the quality of the environment by
enabling us to greatly lower our use of chemical pesticides.

The increased emphasis on IPM and environmental,
economic and social pressures and demands are causing
dynamic changes in the development of new chemicals for
insect and mite control. The most obvious outcome is a
drastically reduced number of new chemical compounds

110

appearing in the marketplace. Several of the new chemical
insecticides and miticides presented here would seem to be
compatible with biological control agents and
semiochemicals in improved pest management systems for
cotton. However, extensive field experience will be
required to verify this.

References Cited

Bilton, J. N., H. B. Broughton, P. S. Jones et al. 1987.
An x-ray crystallographic, mass spectroscopic and NMR
study of the limonoid insect antifeedant azadirachtin and
related derivatives. Tetrahedron 43: 2805-2815. '

Braunholtz, J. T. 1981. Crop Protection: The role of the
chemical industry in an uncertain future. Philosophical
Transactions Royal Society London B. 295: 19-34.

Brosten, D., and B. Simmonds. 1990. Cotton enters the
environmental decade. Agrichemical Age 34: 30-31.

Bull, D. L., and J. J. Menn. 1990. Strategies for manag¬
ing resistance to insecticides in Heliothis pests of cotton.

In M. B. Green, H. M. LeBaron, W. K. Moberg [eds.],
Managing Resistance to Agrochemicals. American Chemi¬
cal Society Symposium Series 421: 118-113, American
Chemical Society, Washington, D.C.

Campbell, W. C. 1981. An introduction to the
avermectins. New Zealand Veterinary Journal 29: 174-
178.

Campbell, W. C., R. W. Burg, M. H. Fisher, and R. A.
Dybas. 1984. The discovery of ivermectin and other
avermectins. In P. S. Magee, G. K. Kohn, J. J. Menn
[eds.], Pesticide Synthesis Through Rational Approaches.
American Chemical Society Symposium Series 255: 5-20,
American Chemical Society, Washington, D.C.

Croft, B. A. 1990. Arthropod Biological Control Agents
and Pesticides. 723 pp. John Wiley & Sons, New York.

Croft, B. A., and A. W. A. Brown. 1975. Responses of
arthropod natural enemies to insecticides. Annual Revue
Entomology 20: 285-325.

Croft, B. A., and M. E. Whalon. 1982. Selective toxicity
of pyrethroid insecticides to arthropod natural enemies and
pests of agricultural crops. Entomophaga 27: 3-21.

Dybas, R. A., N. J. Hilton, J. R. Babu, and others. 1989.
Novel second-generation avermectin insecticides and
miticides for crop protection. In A. L. Smith, G. A.
Somkuti, J. C. Hunter-Cevera, and H. W. Rossmore [eds.],
Novel Microbial Products for Medicine and Agriculture,
pp. 203-212. Elsevier/Society for Industrial Microbiology,
New York.

Economic Research Service, USDA. 1990. Agricultural
Resources Inputs Situation Outlook. Report Economic
Research Service AR-17, 62 pp.

Elliott, M. 1989. The Pyrethroids: Early discovery,
recent advances and the future. Pesticide Science 27: 337-
351.

Elzen, G. W. and E. G. King. 1990. Periodic release and
manipulation of natural enemies. In T. Fisher, T. Bellows,
L. Caltagirone, D. Dahlsten, G. Gordh, C. Huffaker
[eds.], Principles and Application of Biological Control.
University of California Press, Berkeley. In Press.

Flattum, R. F. 1990. Personal Communication.

Frisbie, R. E. 1985. Regional implementation of cotton
IPM. In R. E. Frisbie, and P. L. Adkisson [eds.], Inte¬
grated Pest Management on Major Agricultural Systems,
pp. 638-651. Texas A&M University, MP-1616.

Giedraitis, C. 1990. Personal Communication.

Jenkins, J. N. 1990. Personal Communication.

Kadir, H. A., and C. O. Knowles. 1990. Octopaminergic
action of the carbodiimide metabolite of Diafenthiuron
acaricide/insecticide. Abstr. 200th American Chemical
Society National Meeting, Washington, D.C., August 26-
31, 1990.

Hollingworth, R. M., and A. E. Lund. 1982. Biological
and neurotoxic effects of amidine pesticides. In J.R. Coats,
ed., Insecticide Mode of Action, pp. 189-227. Academic
Press, New York.

Ill

King, E. G. 1986. Insecticide use in cotton and the value
of predators and parasites for managing Heliothis. Pro¬
ceedings Beltwide Cotton Production Research Conference,
Las Vegas, Nevada, 1985, pp. 155-162.

Krieger, R. L., P. P. Feeny, and C. F. Wilkinson. 1971.
Detoxication enzymes in the guts of caterpillars: An
evolutionary answer to plant defenses. Science 172: 579-
581.

Magee, P. S., G. K. Kohn, and J. J. Menn. 1985. Ap¬
proaches to rational synthesis of pesticides. American
Chemical Society Symposium Series 255: 1-207. American
Chemical Society, Washington, D.C.

McKibben, G. H., J. W. Smith, and W. L. McGovern.
1990. Development of an attract-and-kill device for the
boll weevil (Coleoptera: Curculionidae). Journal Entomo¬
logical Sciences (In press).

Menn, J. J. 1983. Present insecticides and approaches to
discovery of environmentally acceptable chemicals for pest
management. In D. L. Whitehead, ed., Natural Products
for Innovative Pest Management, pp. 5-31. Pergamon
Press, Oxford and New York.

Menn, J. J., and C. A. Henrick. 1981. Rational and
biorational design of pesticides. Philosophical Transactions
Royal Society London B 295: 57-71.

Menn, J. J., and C. A. Henrick. 1985. Newer chemicals
for insect control. In J. L. Hilton, ed., Agricultural
Chemicals of the Future, pp. 247-265. Rowman and
Allanheld, Totowa.

Menn, J. J., E. G. King, and R. J. Coleman. 1989.

Future control strategies for Heliothis in cotton. In M.

B. Green and D. J. de B. Lyon [eds.], Pest Management in
Cotton, pp. 101-121. Ellis Horwood Ltd., Chichester.

Morrod, R. S. 1981. Lead generation: Designing the
right approach. Philosophical Transactions Royal Society,
London B 295: 35-44.

Morton, N., and M. D. Collins. 1989. Managing the
pyrethroid revolution in cotton. /«M. B. Green, D. J. de
B. Lyon [eds.], Pest Management in Cotton, pp. 151-165.
Ellis Horwood Ltd., Chichester.

Peregrine, D. J. 1989. The role of amitraz in cotton pest
control. In M. B. Green and D. J. de B. Lyon [eds.], Pest
Management in Cotton, pp. 191-199. Ellis Horwood Ltd.,
Chichester.

Plapp, F. W., and D. L. Bull. 1978. Toxicity and
selectivity of some insecticides to Chrysopa cornea , a
predator of the tobacco budworm. Environmental Ento¬
mology 7: 431-434.

Price, J. F., and D. J. Schuster. 1990. Responses of sweet
potato whitefly to extracts of the neem tree ( Azadirachta
indica) seeds. In Neem's Potential in Pest Management
Programs. USDA Neem Workshop, April 16-17, 1990,
Beltsville, Maryland.

Schmutterer, H. 1990. Properties and potential of natural
pesticides from the neem tree, Azadirachta indica. Annual
Revue Entomology 35: 271-297.

Shorey, H. H., L. K. Gaston, and R. S. Kaae. 1976. Air
permeation with gossyplure for control of the pink boll-
worm. In M. Beroza, ed., Pest Management with Insect
Sex Attractants. American Chemical Society Symposium
Series 23: 67-64. American Chemical Society, Washing¬
ton, D.C.

Staten, R. T., H. M. Flint, R. C. Weddle, and others.

1987. Pink bollworm (Lepidoptera: Gelechiidae) large-
scale field trials with a high rate gossyplure formulation.
Journal Economic Entomology 80: 1267-1271.

Suguiyama, L., and C. Osteen. 1988. The economic
importance of cotton insects and mites. U.S. Department
of Agriculture, Economic Research Service, AER-599.

Svoboda, J. A. 1989. Personal Communication.

Verloop, A., and D. C. Ferrell. 1977.

Benzoylphenylureas - A new group of larvicides interfering
with chitin deposition. In J. R. Plimmer et al. [eds.],
Pesticide Chemistry in the 20th Century. American
Chemical Society Symposium Series 37: 237-270. Ameri¬
can Chemical Society, Washington, D.C.

Vinson, S. B. 1977. Behavioral chemicals in the augmen¬
tation of natural enemies. In R. L. Ridgway and S. B.
Vinson [eds.], Biological Control by Augmentation of
Natural Enemies, pp. 237-279. Plenum Press, New York.

112

Wing, K. D. 1988. RH-5849, a nonsteroidal ecdysone
agonist: Effects on Drosophila cell line. Science 241:
467-469.

Wing, K. D., R. A. Slawecki, and G. R. Carlson. 1988.
A nonsteroidal ecdysone agonist: Effects on larval lepi-
doptera. Science 241: 470-472.

Table 1 . Primary stages involved in discovery and development of a pesticide

Activity

Time Frame (Months)

Synthesis and initial bioscreen

18

Advanced bioevaluation

24

Development

Field trials

Formulations

Metabolism and environmental studies
Toxicology (acute and chronic)

40

Test marketing and sales

12

Total

96

Table 2. Toxicities of EMA and abamectin, two carbamates, and two synthetic pyrethroids when tested on
neonate larvae of He/iothis virescens 5TBW)1., H. zea (CBW),1 and Spodoptera eridania (SAW)1

Compound

IC^ (ppm) at 36 h2

Relative potency

EMA

SAW

TBW

CBW

CBW

EMA (MK-243)

0.005

0.003

0.002

1.0

Abamectin

5.83

0.128

0.21

0.01

Methomyl

8.6

10.0

8.1

0.00025

Thiodicarb

4.42

5.0

6.6

0.0003

Cypermethrin

0.41

0.005

Fenvalerate

1.34

1.5

‘TBW - tobacco bud worm; CBW - cotton bollworm; SAW - southern army worm.
2 Bioassays based on exposure of larvae to treated foliage in enclosed containers.

Adapted from: Dybas et al. (1989).

113

PESTICIDE

ACEPHATE

ALDICARB

azinphos-methyl-E

CHLORDIMEFORM
CYFLUTHRIN
CYPERMETHRIN
DICROTOPHOS
DIMETHOATE
ESFENVALERATE
FENVALERATE-
KARATE"

M ETHYL PARATHION -
PARATHION -I
PROFENOFOS
THIODICARB
TRALOMETHRIN

2.64%

3 4.48%

^ 4.61%

8.96%

2.53%

1 1

3.66%

6.41%

10.73%

1 22.65%

2000 4000 6000 8000 10000 12000

ACRES TREATED (XI 000)

Fig. 1. Insecticide use on U.S. cotton, 1988.

114

Diflubenzuron (PH 60-40)

F

Andalin (Duphar)

Fig. 2. Structures of benzoylphenylurea insecticides: diflubenzuron and andalin.

115

H — C S — N H — C (CH 3 ) 3

insects

sunlight

t

- N =C=N — C (CH 3 )

Fig. 3. Structure of the thiourea insecticide and miticide diafenthiuron (Ciba Geigy) and its photoconversion
product and metabolite-carbodiimide.

Fig. 4. Structures of avermectin B! (abamectin) and its amino analog, 4"-deoxy-4"-methylamino-4"-
epiavermectin B1 (EMA, MK-243). Adapted with permission from Dybas et al. (1989).

116

Azadirachtin

Fig. 6. Chemical structure of the ecdysone agonist, RH-5849 (Rohm & Haas)

117

Systems Analysis and Modeling for
Decision Making in Pest Management

Terence L. Wagner1

Abstract

Models describing agricultural systems have changed
dramatically over the last two decades. They have changed
in composition, form, and how they are applied. Early
cotton models described the growth and development of an
average plant. Models developed for integrated pest
management described populations of plants and insects.
From the outset, they were used as research tools, but
recent advancements in computer science and artificial
intelligence have made them viable management tools. As
management aids, models have progressed from decision
support (simply providing information on which decisions
can be based) to actually making decision recommenda¬
tions. Exciting innovations are on the horizon that will
further expand the form and function of agricultural
models.

Introduction

Imagine you are responsible for managing 500 hectares of
cotton, and your livelihood depends on your success. The
cotton is planted on fairly productive land within a farming
community of mixed private ownership, part of a larger
landscape containing an assortment of crops and woodlots.
This situation is fairly common in the Cotton Belt of the
United States.

Your job is to select the best cotton variety for the growing
conditions, determine the pre-plant fertilizer requirements
and the best planting date. You will have to decide if,
when, how much, how to apply, and what type of post-
emergent fertilizer, plant growth regulator, and irrigation is
needed; if, when, and how to control the numerous arthro¬
pod, disease, and weed species affecting the crop; when and
how to terminate the crop, and when to harvest. Along the
way, you will need to schedule conflicting management
activities and make sundry other decisions that affect the
future outcome of the crop. Assuming you are interested in
making a profit, you will have to decide if, when, and how
much pre-harvest crop to sell (playing the futures market).

1 Crop Simulation Research Unit, USDA, ARS, Mississippi State,
Mississippi

I

Now imagine the type of expertise you will need to
successfully accomplish these tasks. You will certainly
have to call on knowledge from agronomy, plant physiol¬
ogy, soil chemistry, soil physics, entomology, pathology,
weeds science, accounting, and marketing, to name a few.
Considerable attention will have to be given to the uncer¬
tainties associated with the weather, equipment, scouting
information, personnel, the market place, etc. These
uncertainties will directly affect your choice of decisions,
altering the risks associated with the production and sale of
the crop. In addition, the risks of borrowing large sums of
money to initiate and maintain production, without the
assurance of a profitable return, will be worrisome.

I

This imaginary job assignment serves to demonstrate the
nature of farming. It probably arouses similar reactions in
all of us. Farming is anything but trivial, especially in
today's highly competitive world marketplace. Farm
managers must deal with enormous complexities, uncertain¬
ties, and risks in making decisions that affect their lives and
livelihoods, and the returns are often marginal at best.

Very few professions have these disincentives and pres¬
sures.

With these facts in mind, a relevant question to ask is: how
can science help farmers improve their understanding of the
systems with which they must deal, thereby reducing the
complexities, minimizing the uncertainties, and lowering
the risks of their business? This paper describes some of
the general research directed at these issues, especially as it
relates to systems analysis and modeling in agricultural
management. In the process, I hope to convey the useful¬
ness and some of the history of these approaches, using
cotton as an example.

Foundations of Systems Research

We live in an extraordinary time of technological develop¬
ment, a prominent example is the computer. Over the last
three decades computers have advanced from large ma¬
chines that process data relatively slowly to small machines
with fast processing speeds. In the United States, computer
costs have decreased exponentially during this period,
resulting in their emergence from obscure (yet exceptional)
research facilities to the commonplace of the desktop.
Concurrent developments in software systems have trans¬
formed computers into powerful, "user-friendly" tools with
hundreds of applications. The applications have virtually

118

eliminated the need to understand these complex machines,
and in the process, have facilitated remarkable changes in
many aspects of American life. Agriculture is no exception
(Holt 1985).

Systems Science and Modeling

Systems science is the study of systems. Watt (1966)
defined a system as "a group of interacting and interdepen¬
dent components forming a unified whole" [from Getz and
Gutierrez (1982)]. Thus, a systems approach to problem
solving attempts to identify and understand the cause and
effect relationships between interacting components of a
system. It follows that systems analysis uses a systems
approach to problem solving. Specifically, systems
analysis is the application of quantitative and qualitative
techniques that enhance the understanding of the interac¬
tions between system components (Getz and Gutierrez
1982). It is a holistic approach that explains the behavior
of the entire system based on an understanding and descrip¬
tion of its interacting parts.

The relationship of models to systems analysis is clear —
models represent and describe the nature of objects and
events. If the problem calls for a systems approach (i.e.,
its goal is to understand and describe the relationships
between system components) then models are an appropri¬
ate strategy for solving the problem. In fact, analytical
models are an eloquent way of addressing these problems
because of their simplicity, precision, and ease of integrat¬
ing and describing interactions among objects and events.

In this paper I will examine some of the ways in which
analytical models (quantitative and qualitative) have been
applied to agricultural problems.

To summarize the discussion on systems analysis, it can be
thought of as a process that encompasses problem defini¬
tion, system description (including the identification of
system boundaries, component parts, and interactions),
inquiry into the cause and effect relationships between
components, synthesis of resulting knowledge into a
working model(s), and testing the behavior of the model for
realism. During analysis, knowledge gaps will be identi¬
fied leading to additional inquiry, synthesis, and testing;
and so the process continues.

An Argument For a Systems Approach In
Agriculture - Implementing More Effective
Management

A systems approach is relevant for dealing with agricultural
problems. The biotic, edaphic, and climatic elements of
agroecosystems are extremely complex. They are com¬
posed of numerous interacting and interdependent parts that
are dynamic in time and space. Because of these dynamic
properties, agroecosystems can produce unexpected results.
Kenneth Watt (1970) argued this point regarding pest
management. He provided examples of counterintuitive
situations relating to chemical control, citing cases where
pesticides promoted and prolonged outbreaks of the very
insects they were meant to regulate. He also cited an
example meaningful to biological control, where a smaller
number of parasite and predator species exerted better
control than a larger number of species (apparently due to
competition among the larger number of interacting
species).

The information used to describe the status of the cropping
system and upon which decisions are based is often incom¬
plete, inaccurate, or dated. This condition, combined with
the characteristics described above, make it very difficult to
evaluate and decipher the singular and composite behaviors
of the system through time. Thus, understanding how the
system will respond to changing environmental conditions
and to imposed alternative management practices is not
always clear, not even to the experts who are responsible
for its management.

Given man's limited ability to comprehend agroecosystem
events as they unfold, there is a great potential for manag¬
ers to take action when no action is needed, not take action
when it is needed, or take the wrong action in a given
situation. These actions may be ineffective, wasting time
and money, or worse — they may produce the opposite
effect desired. Models have the power to integrate,
analyze, and interpret complex information on multiple
interacting variables in a manner not possible by man's
reasoning power alone, not even that of the experts. They
can be used to predict future events, study system behavior,
test alternative hypotheses, and deliver management advice
at the time when it is needed — as events unfold
Agroecosystem models have the capacity to provide
counterintuitive results not anticipated by the experts.

119

The Changing Status of Systems Research
in Agriculture

The body of techniques and theories associated with
systems analysis and modeling has changed considerably
over the last three decades. Some of these changes are
relevant to a discussion on computer-based management
systems in agriculture. They serve as an excellent means
for describing the transformation of models in agriculture,
and because of the extensive modeling efforts on cotton,
this cropping system will be emphasized here. Two types
of changes will be discussed: those involving the descrip¬
tion of the system under study and those dealing with the
methods or tools of analysis used in modeling this system.
Because these changes did not occur independently, they
will be considered together.

Agroecosystem Structure

Biological systems can be classified hierarchically. For
example, five levels of organization might be considered in
studying and modeling agricultural systems: the organism,
population, community, ecosystem, or landscape. Each
level is a system in its own right, somehow subject to or
dependent upon the level(s) above it. The individual plants
and insects in a particular field can each be considered a
system at the organism level. Groups of interacting plants
and insects of the same species are systems at the popula¬
tion level, and groups of populations are systems at the
community level. Higher up the hierarchy, multiple
interacting communities in association with their abiotic
environment form an ecosystem, and multiple interacting
ecosystems form a landscape.

Selecting the proper level of organization is important in
modeling biological systems, and is part of the problem
definition that must be addressed prior to initiating study.

As pointed out by Getz and Gutierrez (1982), this process
is not always easy, even though the structure of each system
is fairly well defined. The proper level of system organiza¬
tion depends on the type of questions being asked; and in
modeling agricultural systems, the questions (or objectives)
have changed over the years as has the technology support¬
ing these alternate perspectives. The system of interest to
agricultural scientists has expanded toward higher levels of
ecological organization and greater levels of system
complexity.

The Past

Early cotton modelers were interested in describing the
growth and development of a single, average plant (Duncan

1971, Baker et al. 1972, Stapleton et al. 1973, McKinion et
al. 1975). In this case, the system under study was the
organism, and the components of interest were plant organs
and the factors controlling their growth and development.
The primary objectives were to understand and describe the
theoretical yield limits of cotton. A good example of this
work is SIMCOT II, the first simulation model to realisti¬
cally describe the fruiting curves of cotton. This model
quickly evolved into GOSSYM which, interestingly, made
its public debut at a Russian symposium in 1976 (Baker et
al. 1976, 1983).

The use of systems science techniques to study insect pest
management had its origins in the 1960s (Watt 1961,

1966), but not until the following decade was there a rapid
expansion of systems applications to these problems
(Ruesink 1976, Getz and Gutierrez 1982). Several impor¬
tant events made this possible, including advances in
computer technology. Just as important, however, was the
changing opinion regarding the environment and its
safeguard.

The use of chemicals had become the principal choice of
pest control in the United States beginning in the 1950s.
Their application became so widespread and heavy that
some investigators labeled their use addictive (see Getz and
Gutierrez 1982). By the 1960s, the hazards of pesticides
were being widely documented and publicized (Carson
1962, Rudd 1964, Smith 1970), prompting calls for
improved procedures and standards for their use (Pimentel
1970). These events acted as a catalyst for change, and in

1972, the government sponsored several large research
projects directed at developing alternative pest management
strategies within an integrated pest management (IPM)
framework. Several projects in forest and agricultural
systems were funded between 1972 and 1985. For ex¬
ample, the "Huffaker" and ensuing "CIPM" Projects
applied systems science methodology to pest problems in
alfalfa, apple, citrus (Huffaker Project only), cotton, and
soybeans (Frisbie and Adkisson 1986). These accelerated
research and development projects used simulation models
to synthesize and unify knowledge about the cropping
systems.

120

In contrast to the agronomists who were interested in
modeling the theoretical yield limits of crop production, the
IPM investigators were interested in a much broader
problem — understanding the impact of arthropod pests on
the crop and developing new strategies of control. This
problem required a higher level of system organization, and
resulted in the selection of the population as the primary
unit of investigation. The "life system" became a standard
approach for organizing and studying the important
components of the system. Clark et al. (1967) defined this
term as "the part of an ecosystem which determines the
existence of a particular population — composed of the
subject species and its effective environment. " The concept
even influenced our views of pest management, defined by
Rabb (1970) as "the reduction of pest problems by actions
selected after the life system of the pests are understood and
the ecologic as well as economic consequences of these
actions have been predicted, as accurately as possible, to be
in the best interest of mankind. "

So successful were the IPM projects that literally hundreds
of crop, pest, beneficial, and economic assessment models
were developed in the 15 years from 1972. These simula¬
tion models contributed to the understanding of process-
level biology, population dynamics, and insect/host
interactions, the development and testing of new manage¬
ment practices, and the evaluation of the pest status of
certain species.

Wagner et al. (1991) recently reviewed the primary IPM
models associated with cotton in the USA. Most prominent
among the pest population models are five describing
Anthonomus grandis grandis Boheman (Fye and Bonham
1972, Jones et al. 1977, Wang et al. 1977, Curry et al.
1980, and Gutierrez et al. 1990a), three describing
Heliothis virescens (F.) and Helicoverpa zea (Boddie)
(Stinner et al. 1974, Brown et al. 1983, Hartstack and Witz
1983), one of Lygus hesperus (Knight) (Gutierrez et al.
1979), three of Pectinophora gossypiella (Saunders)
(Gutierrez et al. 1977, Stone and Gutierrez 1986, and
Hutchison 1991), and one of Pseudatomoscelis seriatus
(Reuter) (Hartstack and Sterling 1986). Most of these
models were linked to a cotton model (Gutierrez et al.

1975, 1984, 1990b; Curry et al. 1980; Jones et al. 1980;
Hartstack and Witz 1983; Brown et al. 1983), and some
were combined with models of other pest or beneficial
species (Brown et al. 1983, Hartstack et al. 1990).

Notwithstanding these accomplishments, very few of the
IPM models have been used on the farm for management
purposes. There are several reasons for this lack of use,
not the least of which were difficulties of novice users
interpreting model output (see Coulson et al. 1990a,
Wagner et al. 1991). TEXCIM, a simulation model
containing multiple pest species of cotton (Hartstack et al.
1990), is somewhat of an exception. Unlike other IPM
models, components of TEXCIM have been under develop¬
ment for almost two decades, leading to farm application in
1988. Model application is limited in range and function,
however, serving few users in the state of Texas for
decision support only (Sterling and Hartstack 1987,

Sterling et al. 1990). In this role, the model assists in
problem solving by furnishing simulation results to the
decision maker. Other early efforts attempting to apply
simulation models to problem solving led to the develop¬
ment of well-defined computer aids called decision support
systems (Coulson and Saunders 1987). Because the IPM
models were not adapted to decision making per se, the
consensus was that they fell short of their original
expectations.

The Present

During the 1980s, research arising from computer science
and artificial intelligence (Al) provided new techniques for
encoding, analyzing, and adapting quantitative and qualita¬
tive information. These applications have had a remarkable
impact on agricultural modeling in the last 5 years, particu¬
larly in the decision making and problem-solving domains.
Again, the new technology and techniques combined with
inadequacies of the past helped promote this redirection in
research, this time to modeling the management process
explicitly.

A prominent example of the new technology is the expert
system (ES). These computer programs are designed to
solve complex, but narrowly defined, problems by mimick¬
ing human reasoning. They apply information assembled
from a domain expert, encoding information (generally) as
a set of rules residing in a knowledge base. For this
reason, ES's are also called knowledge-based systems.
Initially, ES's were written in specialized languages such as
LISP and PROLOG on expensive and dedicated computers
that optimized symbolic computation. Symbolic computa¬
tion involves the processing of symbols and relationships
rather than numbers and characters.

121

Early development costs of ES's prohibited the application
of this technology to most problems domains (McKinion
and Lemmon 1985); but as knowledge-systems software
became available in the mid-1980s, the applications
expanded to other areas including agriculture. This
software (called ES shells) provided a set of methods that
empowered personal computers and workstations in the
development of ES's, dramatically reducing their costs.
Presently, several cotton pest models use ES's. CIC-EM
(Cotton Insect Consultant for Expert Management) is an
excellent example of a classic rule-based ES (Bowden et al.
1990)

Because the knowledge base of an ES is a collection of
heuristic knowledge (containing opinions, judgements,
generalities, and facts), these systems can take advantage of
qualitative information that simulation models cannot. This
feature makes ES's more tractable in dealing with manage¬
ment problems. However, ES's do not have the mathemati¬
cal, and typically, the predictive powers of simulation
models. The individual weaknesses of the two modeling
approaches were offset by their merger, which was an
extremely important step in adapting computer-based
systems to problems of agricultural management. Com¬
bined ES's and simulation models are often called model-
based reasoning systems (MBRS) (Intellicorp, 1986).

In 1985, the expert system COMAX (COtton MAnagement
eXpert) was linked to the cotton model GOSSYM (Lemmon
1986, McKinion et al. 1989). This was the first MBRS in
agriculture. Initial versions of GOSSYM/COMAX
optimized the timing and amount of irrigation water and
nitrogen fertilizer needed to develop the crop. The model
also provided information that could be used to determined
the timing of crop termination through the use of chemical
defoliants and boll openers. During the pilot test (1984-
1988), GOSSYM/COMAX accurately predicted crop
yields, and through its expert advice, routinely increased
farmer profits. These successes promoted its widespread
use across the US Cotton Belt, resulting in 1988 in the
establishment of the first federally-funded technology
transfer unit to support an agricultural model. This group,
called the GOSSYM/COMAX Information Unit, trains
system users (e.g., cotton producers, extension service
personnel, consultants, etc.), transfers technical knowledge
to user groups, resolves user problems, and refers difficult
problems back to the research group for additional work.

In 1990, GOSSYM/COMAX assisted in the management of
about 750,000 hectares of cotton on over 300 sites across
the Cotton Belt.

As an ongoing federal project, GOSSYM/COMAX is under
continual development. Recently, a mouse-driven
windowing environment was added to the model providing
a "user-friendly" interface. When COMAX was written,
commercial ES shells were not available. The knowledge
base and inference engine were written in LISP on a
Symbolics Corporation computer. COMAX is now written
in C using ART-IM (a shell from Inference Corporation).

It is also being redesigned, employing a tool-kit (modular)
structure with advisors for irrigation, fertilization, and crop
termination. An advisor for a cotton growth regulator
(PIX) will soon be incorporated, and a large pest manage¬
ment advisor is in the prototype stage. This latter advisor,
called WHIMS (Holistic Insect Management System),
presently consists of a comprehensive rule-base for 10
arthropod pests. The development of an affiliated simula¬
tion model is underway using an object-oriented program¬
ming (OOP) approach (Olson et al. 1990a, b, c).

OOP is a programming paradigm that has several advan¬
tages over the approach used in conventional computer
languages (Thomas, 1989). For example, it formalizes the
concept of an object as the unit of modularity. Objects
consist of data and a set of methods (e.g., mathematical
functions). The methods manipulate the data and describe
the behavior of the object. The state of the object (its data)
is sheltered in private memory that is only accessed when
the method is implemented (through a specific message).

In this manner, the data is hidden from all but its intended
use, a feature known as encapsulation.

Another feature of OOP is polymorphism, which describes
the condition where multiple objects use the same message.
Because each object has a different method, one message
directed to several objects invokes different responses.
There is no problem with multiple objects responding to the
same message because a method is part of the object and
not a global entity. A third feature of OOP is inheritance,
or the ability of an object to acquire all the data and
methods of other objects, in addition to those of it own.
This feature provides structure to a program (similar to our
view of ecological hierarchies), thus making multi-level
modeling possible.

122

These combined attributes of OOP promote the reuse and
understandability of code. They foster code sharing and
co-development, factors that are important in large projects,
especially if members of the development team are co¬
located. OOP creates a plug-and-play environment (for
objects), making it easy to prototype and test new ideas.
These benefits speed software development and mainte¬
nance.

Two additional MBRS's on cotton also employ elements of
OOP, but these systems are interesting for other reasons as
well. COTFLEX (Stone and Toman 1989) and CALEX
(Plant 1989a, b) have adopted several programming and
analytical features that have relevance to other agricultural
models. The application of rule-based ES's in agriculture
has presented several challenges to researchers, one
involving the scope of the problem itself. As discussed,
agricultural decision making involves the application of
dynamic knowledge from multiple, interacting disciplines
and sources. To be reliable, however, the problem domain
of an ES must be narrowly defined.

To overcome this problem, recent ES's have employed a
modular design. For example, COTFLEX employs three
advisors that describe agronomic practices for farm man¬
agement, pest management, and farm policy (Stone and
Toman 1989). Each advisor consists of sets of ES's that
are rigidly defined. When integrated, the advisors allow
the overall system to address the broad problems of total
crop management. This system is presently being tested in
Texas.

Another challenging problem arising from the application
of ES's in agriculture is that of scheduling crop production
and pest control activities. These activities may tempo¬
rarily conflict with one another. For example, it is difficult
or even impossible to cultivate, apply fertilizers, or scout
for pests immediately after irrigating. Activities may also
interact in more subtle ways. For instance, agronomic
practices such as irrigation and fertilization may predispose
the crop to pest problems.

To deal with these issues, CALEX applies the concept of a
blackboard within an object-oriented framework (Plant
1989b). Within the program, multiple "experts" individu¬
ally analyze the state of the crop relative to their own
knowledge (e.g., pest infestation level, nutrient demand,
etc.). Each expert determines an appropriate response for a
given field situation, writing a schedule of recommended

actions and other relevant information on the blackboard.
The expert iteratively examines the responses of the others,
and revises its schedule to account for conflicts or multiple
influences. CALEX then synthesizes the management
recommendations of all the experts in a clear and concise
manner.

The issue of uncertainty is an active area of research in
agricultural modeling. Agricultural systems are inherently
complex and dynamic, creating uncertainty that directly
affects the confidence of decisions. Several methods have
been developed to deal with uncertainty in ES inference
(Spielgelhalter 1986). For example, CALEX (Plant 1989a)
applies certainty factors (c) as antecedents to rules. Con¬
clusions drawn from these rules are displayed as categories:
most likely to occur (c= 1.0), very likely (0.75 c< 1.0),
reasonably likely (0.5 c<0.75), and possible (0.25
c < 0.5). Conclusions with c < 0.25 are not displayed. In
this way, the system provides the user with knowledge of
all reasonable conclusions.

Bayesian Belief Networks (BBN) (Pearl 1988) are another
inferencing formalism that deal with uncertainty. This
approach has been selected for use in WHIMS, the pest
management component of GOSSYM/COMAX (Olson et
al. 1990b, c). It consists of a network of nodes and
connecting arrows. The nodes represent the propositions
that are modeled, and the arrows represent the causal
relationships between nodes. For example, if an arrow
points from node A to B, then A causes (influences) B.
BBN's use Bayesian probabilities to represent and update
the belief about the nodes in the system. By changing the
belief at one node, change is propagated in both directions
along the causal chain. This method has an advantage over
typical rule-based ES's, which cannot easily accomplish bi¬
directional reasoning. It is also an example of causal
modeling, which is a new approach from AI research that
promises to further expand the flexibility and power of
ES's.

The Future

The last 2 decades have brought extraordinary changes in
agricultural models, both in their structure and function.
Initially, models described the biological system only.
Their complexity increased over the years, moving through
the ecological hierarchy from levels of the organism,
population, and presently to describing ecosystem interac-

123

tions. Models are now beginning to consider the next
level, the landscape, to account for the management of
mixed cropping systems and the utilization of multiple
ecosystems by arthropod pests.

The use of geographical information systems (GIS) will
promote this trend. A GIS is a computerized mapping
system for capturing, storing, retrieving, and analyzing
spatial and descriptive data (Coulson et al. 1989, 1990b).
When integrated with an ES, the GIS can interpret relation¬
ships within and among landscape data themes, select
appropriate solutions for management, and guide in the use
of the system. This powerful new tool has important
implications to ecosystem and landscape management and
planning, and it will certainly be incorporated in agricul¬
tural models in the future. In fact, a GIS is presently being
investigated for use in WHIMS.

Agricultural models no longer describe just aspects of the
biological system. Whereas yield prediction and pest
population dynamics were once the dominant themes,
models now consider management, economic, marketing,
and farm policy issues as well.

New programming languages and software techniques are
changing the development, maintenance, operation, and
look-and-feel of software systems. Development and
maintenance are becoming easier, and the type of informa¬
tion going into the models is becoming more generalized.
These types of developments will continue, making systems
far more powerful and flexible.

These trends will allow greater numbers of persons with
various backgrounds to contribute to the modeling process.
They are already creating an environment where scientists
at different locations can realistically contribute to the
development of a unified model. For example, efforts are
presently underway to coordinate all cotton modeling work
within the USDA, Agricultural Research Service.

The role of models in agriculture has also changed from
mainly research tools, to decision support, and presently
decision making. The next step will probably involve
decision implementation. For example, models will not
only decide when to irrigate but will send messages that
turn water valves on and off.

Conclusion

There is a relationship between the complex and dynamic
nature of the agroecosystem, the uncertainty associated with
it management, and the risks of decision making. By
nature, complexity and change foster conditions of uncer¬
tainty, and uncertainty imposes risks on decision making.
Many of the factors that promote uncertainty in agriculture
are beyond the farmer's control. For example, he can do
little about the weather, the general state of the economy,
resource supplies, government policy toward agricultural
products, the cost of borrowing money, etc. He can,
however, adopt innovations that reduce the uncertainties
and risks of doing business. These innovations should alter
the methods of management, production, or distribution to
favorably affect costs and revenues. Computer models are
rapidly becoming a viable means of accomplishing these
objectives.

The use of models in agriculture will not eliminate uncer¬
tainty and risk; but they will provide increased understand¬
ing of the ecosystem, thereby reducing risks to a point
where alternative management practices can be adopted.
Like all new innovations, there are risks associated with
initial use. Just as the expectation of profit provides the
incentives to adopting new technology, those who are first
to accept these risks will reap the greatest benefits.

I believe it is not the direct risks associated with the use of
this technology that will dictate the acceptance of models in
the near term. These risks are minimal, as there are always
checks on model behavior. It is the changes that the models
will bring in management and production that are resisted,
perhaps even feared. The adoption of pest management
models will require reformed attitudes and approaches to
pest control and will affect, most dramatically, our present
use of pesticides. There are entire industries built around
this 50 year-old technology that superimpose well estab¬
lished protocols on agro-business, especially in cotton.
Herein lies the immediate problem — gaining acceptance (or
at least cooperation) from those who will be most affected
by the impending changes.

While farmers tend to be conservative and risk averse (for
good reason), they are also realistic business people. They
must be in this highly competitive environment. Ulti¬
mately, they will decide the value and role of agricultural
models. Given the rapid and exciting changes occurring in
this area, I believe that models will play a vital role in

124

agricultural management — from testing new ideas on the
farm, to recommending actions that prevent or resolve
problems, and implementing decisions in the future.

References Cited

Baker, D. N., J. D. Hesketh, and W. G. Duncan. 1972.
Simulation of growth and yield in cotton. I. Gross
photosynthesis, respiration, and growth. Crop Sci. 12:
431-435.

Baker, D. N., J. R. Lambert, C. J. Phene, and J. M.
McKinion. 1976. GOSSYM: A simulator of cotton crop
dynamics, pp. 100-133. In Computers Applied to Man¬
agement of Large Scale Agricultural Enterprises. Proc.
US-USSR Seminar, Moscow, Riga, Rischinev.

Baker, D. N., J. R. Lambert, and J. M. McKinion. 1983.
GOSSYM: A simulator of cotton crop growth and yield.

S. C. Agric. Exp. Stn. Tech. Bull. 1089.

Bowden, R. O., R. G. Luttrell, and L. G. Brown. 1990.
Cotton insect consultant for expert management (CIC-EM):
An expert system for managing cotton insects. Proc.
Beltwide Cotton Prod. Res. Conf., Las Vegas, NV, in
press. National Cotton Council of America, Memphis,

TN.

Brown, L. G., R. W. McClendon, and J. W. Jones. 1983.
A cotton insect management simulation model, pp. 437-
479. In R. L. Ridgway, E. P. Lloyd, and W. H. Cross,
[eds.], Cotton Insect Management. U.S. Dept. Agric.,
Agric. Res. Serv., Agric. Handbook No. 589.

Carson, R. L. 1962. Silent spring. Houghton Mifflin.

Clark, L. R., P. W. Geier, R. D. Hughes, and R. F.
Morris. 1967. The ecology of insect populates in theory
and practice. Methuen & Co., London.

Coulson, R. N., and M. C. Saunders. 1987. Computer-
assisted decision-making as applied to entomology. Ann.
Rev. Entomol. 32: 417-437.

Coulson, R. N., M. C. Saunders, D. K. Loh et al. 1989.
Knowledge system environment for integrated pest manage¬
ment in forest landscapes: The southern pine beetle

(Coleoptera: Scolytidae). Bull. Entomol. Soc. Am. 35:
26-32.

Coulson, R. N., R. O. Flamm, F. L. Oliveria et al.

1990a. Implementation of emerging technologies: For¬
estry. In Entomol. Soc. Am. Cent. Symp., Entomol. Soc.
Am. Natl. Meeting, San Antonio, Texas, December 11-14,
in press.

Coulson, R. N., C. N. Lovelady, R. O. Flamm et al.
1990b. Intelligent geographic information system for
natural resource management. In M. G. Turner, and R. G.
Gardner, [eds.]. Quantitative Methods in Landscape
Ecology. Spinger-Verlag, in press.

Curry, G. L., P. J. H. Sharpe, D. W. DeMichele, and J.

R. Cate. 1980. Towards a management model of the
cotton-boll weevil ecosystem. J. Environ. Manag. 11:
187-223.

Duncan, W. G. 1971. SIMCOT: A simulator of cotton
growth and yield. Proc. of Workshop on Tree Growth
Dynamics and Modelling, Duke Univ., Durham, NC, Oct.
11-12, 1971, pp. 115-118.

Frisbie, R. E., and P. L. Adkisson, [eds.] 1986. Inte¬
grated pest management on major agricultural systems.
Texas Agric. Exp. Stn. MP-1616.

Fye, R. E., and C. D. Bonham. 1972. Relationship of
temperatures to boll weevil complex populations in Ari¬
zona. U.S. Dept. Agric., Agric. Res. Serv., Prod. Res.
Rep. No. 136.

Getz, W. M., and A. P. Gutierrez. 1982. A perspective
on systems analysis in crop production and insect pest
management. Ann. Rev. Entomol. 27: 447-466.

Gutierrez, A. P., L. A. Falcon, W. Loew et al. 1975. An
analysis of cotton production in California: A model for
Acala cotton and the effects of defoliators on its yields.
Environ. Entomol. 4: 125-136.

Gutierrez, A. P., G. D. Butler, Jr., Y. Wang, and D.
Westphal. 1977. The interaction of pink boll worm
(Lepidoptera: Gelechiidae), cotton, and weather: A
detailed model. Can. Entomol. 109: 1457-1468.

125

Gutierrez, A. P., Y. Wang, and U. Regev. 1979. An
optimization model for Lygus hesperus (Heteroptera:
Miridae) damage in cotton: The economic threshold
revisit[ed.] Can. Entomol. 111:41-54.

Gutierrez, A. P., M. A. Pizzamiglio, W. J. Dos Santos et
al. 1984. A distributed delay time varying life table plant
population model: Cotton ( Gossypium hirsutum L.) as an
example. Ecol. Modelling 26: 231-249.

Gutierrez, A. P., W. J. Dos Santos, M. A. Pizzamiglio et
al. 1990a. Modelling the interaction of cotton and the
cotton boll weevil. II. Boll weevil ( Anthonomus grandis
Boh.) in Brazil. J. Applied Ecol., in press.

Gutierrez, A. P., W. J. Dos Santos, A. M. Villacorta et al.
1990b. Modelling the interaction of cotton and the cotton
boll weevil. I. A comparison of growth and development
of cotton varieties. J. Applied Ecol., in press.

Hartstack, A. W., and J. A. Witz. 1983. Models for
cotton insect pest management, pp. 359-384. In R. L.
Ridgway, E. P. Lloyd, and W. H. Cross [eds.], Cotton
Insect Management with Special Reference to the Boll
Weevil. U.S. Dept. Agric., Agric. Res. Serv., Agric.
Handbook 589.

Hartstack, A. W., and W. L. Sterling. 1986. Texas cotton
fleahopper model users guide, version 2: Basic. Texas
Agric. Exp. Stn. Misc. Publ. MP-1595.

Hartstack, A. W., W. L. Sterling, and D. A. Dean. 1990.
TEXCIM40: The Texas cotton-insect model. Texas Agric.
Exp. Stn. Misc. Publ. MP-1646 (revised).

Holt, D. A. 1985. Computers in agriculture. Science.
228: 422-427.

Hutchison, W. D. 1991. A management model of pink
bollworm population dynamics: development, validation,
and insecticide timing, in preparation.

IntelliCorp. 1986. Model-based reasoning in the KEE and
SimKit systems. IntelliNew 2(2), 1 1 pp.

Jones, J. W., H. D. Bowen, R. E. Stinner et al. 1977.
Simulation of boll weevil population as influenced by
weather, crop status and management practices. Trans.

Am. Soc. Agric. Engin, 20: 121-125, 131.

Jones, J. W., L. G. Brown, and J. D. Hesketh. 1980.
COTCROP: A computer model for cotton growth and
yield, pp. 209-241. In J.D. Hesketh and J.W. Jones [eds.],
Predicting Photosynthesis for Ecosystems Models, Vol. 3.
CRC Press, Inc., Boca Raton, Florida.

Lemmon, H. E. 1986. COMAX: An expert system for
cotton crop management. Science 233: 29-33.

McKinion, J. M., and H. E. Lemmon. 1985. Expert
systems for agriculture. Comput. Electron. Agric. 1:31-
40.

McKinion, J. M., D. N. Baker, J. D. Hesketh, and J. W.
Jones. 1975. SIMCOT II: A simulation of cotton growth
and yield, pp. 27-82. In Computer Simulation of a Cotton
Production System: Users Manual. U.S. Dept. Agric.,
Agric. Res. Serv. ARS-S-52.

McKinion, J. M., D. N. Baker, F. D. Whisler, and J. R.
Lambert. 1989. Application of the GOSSYM/COMAX
system to cotton crop management. Agric. Systems 31:
55-65.

Olson, R. L., T. L. Wagner, M. R. Williams, and J. L.
Willers. 1990a. The development of WHIMS, a knowl¬
edge-based insect management system. Proc. Beltwide
Cotton Prod. Res. Conf., Las Vegas, NV, (in press).
National Cotton Council of America, Memphis, TN.

Olson, R. L., T. L. Wagner, and J. L. Willers. 1990b. A
framework for modeling uncertain reasoning in ecosystem
management. I. Background and theoretical consider¬
ations. Al Appl. Nat. Res. Man. (in press).

Olson, R. L., J. L. Willers, and T. L. Wagner. 1990c. A
framework for modeling uncertain reasoning in ecosystem
management. II. Bayesian belief networks. Al Appl. Nat.
Res. Man. (in press).

Pearl, J. 1988. Probabilistic reasoning in intelligent
systems: Networks of plausible inference. Morgan
Kaufmann, San Mateo, California.

Pimentel, D. 1970. Training in pest management and the
"systems approach" to control, pp. 209-219. In R. L.
Rabb, and F. E. Guthrie [eds.]. Concepts of Pest Manage¬
ment: Proceedings. North Carolina State Univ. Press,
Raleigh, NC.

126

Plant, R. E. 1989a. An integrated decision support system
for agricultural management. Agric. Systems 29: 49-66.

Plant, R. E. 1989b. An artificial intelligence based
method for scheduling crop management actions. Agric.
Systems 31: 127-155.

Rabb, R. L. 1970. Introduction to the conference, pp. 1-
5. In R.L. Rabb, and F.E. Guthrie [eds.], Concepts of
Pest Management: Proceedings. North Carolina State
Univ. Press, Raleigh, NC.

Ruesink, W. G. 1976. Status of the systems approach to
pest management. Ann. Rev. Entomol. 21: 27-44.

Rudd, R. L. 1964. Pesticides and the living landscape.
Univ. of Wisconsin Press.

Smith, R. F. 1970. Pesticides: Their use and limitations
in pest management, pp. 103-113. In R. L. Rabb and F. E.
Guthrie [eds.], Concepts of Pest Management: Proceed¬
ings. North Carolina State Univ. Press, Raleigh, NC.

Spiegelhalter, D. J. 1986. A statistical view of uncertainty
in expert systems, pp. 17-56. In Gale, W. A. [ed.].
Artificial Intelligence and Statistics, Addison-Wesley,
Reading, MA.

Stapleton, H. N., D. R. Buxton, F. L. Watson et al. 1973.
COTTON: A computer simulation of cotton growth. Ariz.
Agric. Exp. Stn. Tech. Bull. 206.

Sterling, W. L., and A. Hartstack. 1987. TEXCIM -- A
fleahopper-f/e/iotfiw-plant model for on farm use. Proc.
Beltwide Cotton Prod. Res. Conf., Dallas, TX, pp. 258-
259. National Cotton Council of America, Memphis, TN.

Sterling, W. L., A. W. Hartstack, and D. A. Dean. 1990.
TEXCIM40: Testing and use on the farm. Proc. Beltwide
Cotton Prod. Res. Conf., Las Vegas, NV, (in press).
National Cotton Council of America, Memphis, TN.

Stinner, R. E., R. L. Rabb, and J. R. Bradley. 1974.
Population dynamics of Heliothis zea (Boddie) and H.
virescens (F.) in North Carolina: A simulation model.
Environ. Entomol. 3: 163-168.

Stone, N. D., and A. P. Gutierrez. 1986. Pink bollworm
control in Southwestern desert cotton. I. A field-oriented
simulation model. Hilgardia 54: 1-24.

Stone, N. D., and T. W. Toman. 1989. COTFLEX: An
integrated expert and database system for decision support
in Texas cotton production. Proc. AAAI/ASAE Work¬
shop: Integration of Expert Systems with Conventional
Problem Solving Techniques in Agriculture, August 10-12,
San Antonio, TX.

Thomas, D. 1989. What's in an object? Byte 14: 231-
240.

Wagner, T. L., R. L. Olson, J. L. Willers, and M. R.
Williams. 1991. Modeling and computerized decision
aids. In E.G. King, and J.R. Phillips, [eds.], Cotton
Insects and Mites: Characterization and Management.
Cotton Foundation Ref. Book Series No. 3, National
Cotton Council of America, Memphis, TN. (in press).

Wang, Y., A. P. Gutierrez, G. Oster, and R. Daxl. 1977.
A population model for plant growth and development:
Coupling cotton-herbivore interaction. Can. Entomol.

109: 1359-1374.

Watt, K. E. F. 1961. Mathematical models for use in
insect pest control. Can. Entomol. 93 (Suppl. 19): 1-62.

Watt, K. E. F. 1966. Systems analysis in ecology.
Academic Press, Inc., NY.

Watt, K. E. F. 1970. The systems point of view in pest
management, pp. 71-79. In R. L. Rabb, and F. E. Guthrie
[eds.], Concepts of Pest Management: Proceedings. North
Carolina State Univ. Press, Raleigh, NC.

127

Boll Weevil Eradication in the
Western United States and
Northwestern Mexico, a Cooperative
Program

L. C. Antilla1 2, F. D. Myers1 and R. T. Staten3

Introduction

Anthonomus grandis Boheman, the boll weevil, has been
found only sporadically as an economic pest in Arizona
before 1978, and was usually associated with stub cotton
(Fye 1968). Stub cotton is cotton grown as a perennial
crop which, in its second year after planting, is grown from
plant stubs remaining after plant stalks are shredded in the
fall. It is a practice avoided by the majority of growers and
usually prevented by regulatory law within each state
because it enhances the biotic potential of the boll weevil
and the pink bo 11 worm, Pectinophora gossypiella (Sand¬
ers). In 1978, cotton growers in central Arizona, where
climatic conditions favor stub cotton, sought and obtained
regulatory changes to allow this practice. The resultant
change was immediately followed by detection of economic
populations of boll weevil for the first time since the 1960s
(Bergman 1983). As reported by Bergman, populations
were detected in 1978 only in three fields. By 1981,
populations were found in 25 locations. Within three
years, population centers were found across Arizona in such
diverse places as Parker, on the western border of Arizona,
Laveen in central Arizona, and in the San Louis/Mexicalli
valley of northwest Mexico. The weevil populations were
no longer restricted to stub cotton and were now obviously
well adapted to our desert climate. Economic damage was
extensive and increasing in Arizona and threatening those
desert valleys of California previously unaffected. By the
end of 1984, a cooperative program in California, Arizona,
and northwest Mexico was organized for initiation in 1985.
This program was to include the Colorado River Basin,
with parts of the Gila River Basin as a buffer (Figure 1). In
1987, after viewing the success of this program, the
Arizona cotton growers agreed (with major funding) to
conduct an expanded Arizona-U.S. Department of Agricul¬
ture Program starting in 1988 for the remainder of Arizona.
This program is now in its final stages. This report will
cover the activity of these two related projects which
demonstrate the feasibility of eradicating a recently estab¬
lished boll weevil population that has adapted to irrigated
cotton in desert climates.

1 Southwest Boll Weevil Eradication Program, Phoenix, Arizona

2 Science 8l Technology, USDA, APHIS, Phoenix, Arizona

i

i

Program Operations

The boll weevil eradication program, as operated in the
western U.S., has followed modifications (minimal) of
eastern programs. It has three operational aspects: 1) a
required host free period of at least 60 days, 2) extensive
population monitoring using grandlure baited pheromone
traps, 3) treatment of infested cotton with 16 oz./acre ultra
low volume malathion initiated at a low threshold, based on
pheromone traps. It is important to note that all three
phases of this program are critical and form an integrated
complete program. Though each aspect will be separated in
our discussion of methods employed, this integration of
technology should not be minimized. Although the
complete eradication program was carried out in two
phases, operations have occurred with minimal managerial
discretion, the Colorado River Basin program operated as a
full eradication project in 1985 and continued through
1988. Population monitoring continues in this area with no
significant treatments or weevil detection. The central
Arizona program started in 1988, and is expected to have
accomplished its goal in 1991. In all cases, the program
technology used was the same from 1985 through the
present time.

Cultural control is implemented on a mandatory basis (by
law). It is not a voluntary program. Growers are encour¬
aged to exceed its limits in terms of providing longer host
free periods where operationally feasible. A 60 day host
free period is minimal and required. Statutory regulations
specify that all cotton residue (stalks and other crop debris)
will be shredded and tilled into the earth by a specific plow j
down date. No cotton stalks will be permitted to re-grow.
No cotton may be planted until a date 60 days after the
mandated plow down date from the preceding season.
Growers who do not comply are penalized by monetary
fines. In effect, a 60 day host free period as prescribed by
the above restrictions produces a longer biological host free
period. This is due to the fact that weevils must have
fruiting forms to reproduce. Once cotton is picked,
shredded and tilled, it is more than 60 days before any
hostable squares are available. Boll weevil will feed on a
square about 3/16 of an inch in diameter. Thus, the
minimum biological host free period is approximately 105
days, depending on the specific growing conditions for that
season.

128

Although cotton fruiting forms may be visually examined
for reproducing weevil populations, the principle program
monitoring activity is dependent on pheromone traps baited
with grandlure. This represents 99.9% of all populations
monitoring activities within a program. In our programs,
the boll weevil trap used has been described by Dickerson
et al. (1981). This trap uses a controlled release formula¬
tion of grandlure designed to emit an average of 0.5 mg per
day for 14 days. Grandlure is a very potent aggregation
pheromone capable of attracting large numbers of weevils
in the absence of fruiting cotton in heavy populations as
well as detecting extremely low level populations where
only a few weevils are present in a field. Improvements in
traps and lure formulations allowing this low level of
population detection are among the single most important
components of this program (Brazzel, unpublished). In this
program, all applications of insecticide are based on
population monitoring by this trap, except on very rare
occasions, where incipient infield reproducing populations
are detected by examination of fruiting forms.

The insecticide of choice in the western program has been
malathion sprayed in ultra low volume (ULV) formula¬
tions. This material at over 91 % active ingredient is
sprayed without dilution at 16 oz. per acre. It is usually
sprayed with a fixed wing aircraft at 120 miles per hour
with 40 lbs. of pressure and twelve 80002 tee jet flat fan
nozzles directed perpendicular to the path of the aircraft. In
addition to fixed wing aircraft, the program also used
helicopter applications and mist-blower ground equipment
to treat with this material. These two alternatives are used
where fixed wing aircraft cannot reach the target with spray
or operate without creating environmental problems. These
treatments are initiated throughout the program based on
specific guidelines. Although these guidelines have been
altered slightly to cover local needs, the program through¬
out has principally treated when two or more weevils have
been captured in the spring as cotton reaches early bud
formation. Treatment is initiated only after hostable flower
buds or squares are present. This threshold continues
through June. In July control activities are only initiated
after 5 weevils per week have been captured in a given
field. As the season progresses the threshold is again raised
to 10 per week. This is done in October during the time
when treatments are designed to reduce numbers which can
successfully reach overwintering quarters.

Results

The Colorado River Basin program has now been com¬
pleted as an eradication operation. All program statistics
are shown in Table 1. With the exception of 1990, this
program covered a minimum area of 181,926 acres in 1986,
to a maximum of 305,561 acres in 1988. Traps were
maintained between 1 trap for each 6.78 acres in 1986, to 1
trap for 9.48 acres in 1989. In 1990, when all evidence of
boll weevils was eliminated, trap ratios were reduced. This
elimination of weevil populations may be noted by examin¬
ing the number of total weevils captured. In 1985, during
the first year of operation with 9.8 acres covered by a
single trap, the program operated a total of 25,252 traps
weekly during the entire growing season. These traps
captured 302,758 weevils (derived by summarizing all trap
data as provided by field crews). As of 1989, a comparable
number of traps, 26,816, captured a total of 751 weevils.

It is important to note that only 20 of these weevils were
captured outside the Gila river area, which served as a
buffer zone between the Colorado River Basin area and the
generally infested central Arizona region. These weevils
were captured only when weevils are known to be dispers¬
ing out of or into cotton and were not considered to be part
of a reproducing population. This has been verified in part
by the total lack of capture in 1990. The areas requiring
treatment also were comparably reduced as would be
expected, and the cumulative acres treated were highest in
1985 at 307,134. It should be noted that this is an average
of less than 2 treatments for each program acre. Treatments
within the program area varied from areas with no treat¬
ment to fields requiring extensive treatment. The latter was
restricted to only those supported by the trap detection of
weevils. In 1990, no insecticides were applied in the
Colorado River Basin eradication zone for boll weevil. It is
interesting to note that California and Mexico were without
treatment as early as 1988 within the non-expanded
eradication zone. Even in 1988, of the 3,029 weevils
captured, only 35 were captured outside the buffer zone in
Arizona.

The central Arizona or expanded program began in 1988.
This program started when weevil populations were
expanding drastically toward the eastern edge of Arizona
and into neighboring New Mexico. Population densities
also were escalating drastically. This extreme population
pressure may be noted by examination of numbers of
weevil captured in 1988, the first year of program opera¬
tion. This data is in Table 2. This program is now in its

129

third year. It will not be completed, i.e. at 0 weevils
detected, before 1991 or 1992. It is, however, well on its
way, as is indicated by the major decline in total weevils
captured in the entire program area. Although we have
summarized data in this paper through August 1, 1990, our
populations have declined drastically from the 2,724,593
weevils captured annually in 1988. As of August 1, 1990,
our total capture level has been at 2,853. Even if fall
captures double or triple, there will be more than a 99 %
reduction in total population.

Summary

The boll weevil in the western irrigated areas is eradicable
and will be eradicated by 1992 at the latest. If eradication
had not been the chosen route, growers would have been in
a constant expensive control program. This management
program, being heavily dependent on pesticides, would
have been expensive to the environment and to the econom¬
ics of each grower. Benefits could be expected to be as
good as those achieved in the southeast (Carlson et al.

1989) where a 97 % return per total dollar spent was
expected each year after the program was completed.

References Cited

Bergman, D., T. J. Henneberry, L. A. Bariola, and J. M.
Gillespie. 1983. Studies of Pest and Beneficial Insects in
Arizona Stub and Planted Cotton. U.S. Department of
Agriculture. ARR-W-32. 25pp.

Carlson, G. A., G. Sappie, and M. Hammig. 1989.
Resources and Technology Division, Economic Research
Service, U.S. Department of Agriculture. Agricultural
Economics report. No. 621. 31 pp.

Dickerson, W. A., G. H. McKibben, E. P. Lloyd, J. F.
Keamey, J. J., Jr. Lam, and W. H. Cross. 1981. Field
Evaluation of a Modified In-field Boll Weevil Trap. J.
Econ. Entomol. 74: 280-282. 968 pp.

Fye, R. E. 1968. Populations of Boll Weevil in Selected
Fields in Arizona in 1965 and 1966. J. Econ. Entomol.
61: 377-380.

130

Table 1. Colorado River Basin Boll Weevil Eradication Program Statistics

Year

Acres including
Buffer Zone

Average

acres/trap

Cumulative
acres treated

Total weevils
captured

1985

233,357

9.18

307,134

302,758

1986

181,926

6.78

109,385

53,672

1987

224,043

8.01

52,578

27,245

1988

305,561

9.57

35,373

28,433

1989

211,727

9.48

3,029

751

1990

(As of August 1)

67,000 1

23.23 2

O2

O2

1 Buffer zone only

3 Ratio in non-buffer zone (central eradication zone)

Table 2. Central Arizona Boll Weevil Eradication Program Statistics

Year

Acreage

Acres
per trap

Cumulative
acres treated

Total Weevils
captured

1988

426,466

8.59

859,429

2,724,593

1989

444,540

6.70

545,645

586,961

1990'

431,776

6.26

24,829

2,853

All data summarized through August 1 .

Use of Fungal Pathogens for
Biological Control of Insect Pests in
Alfalfa

Raymond /. Can others ' and Kirsten HuraP

Abstract

The U.S. Department of Agriculture (USD A) has been
involved in the development of biological control for over
100 years. Most of these efforts have employed predators
or parasites of insect or weed pests and only recently has
much emphasis been placed on the use of pathogens
(bacteria, viruses, fungi, etc.) as biological control agents
of insect pests. Entomopathogenic fungi are important
natural control agents of many insects including several
pests of alfalfa. Basic fungal pathogen biology, life cycles
and environmental factors regulating disease development
and spread are discussed from ecological and pest manage¬
ment perspectives. Several different approaches to biologi¬
cal control are discussed including: awareness and conser¬
vation of natural biological control , enhancement of
biological control impacts through manipulations, augmen¬
tation of natural enemy numbers, and introduction of exotic
natural enemies. Future interactions between IPM,
biological control, insect pathology and biotechnology
specialists are discussed.

Introduction

Biological control, the use of natural enemies to help
regulate pest populations, has been an important component
of many Integrated Pest Management (IPM) programs. In
most cases, biological control involves the use of parasites,
predators or pathogens which attack, injure or kill the
target pest. To that end, the U.S. Department of Agricul¬
ture has been involved in the development and implementa¬
tion of biological control programs for insect and weed
pests since 1888 when C. V. Riley helped to introduce the
vedalia beetle for biological control of cottony cushion
scale in California (Coppel and Mertins 1977, King and
Coulson 1988). Most of these efforts have focused on the
collection and use of exotic parasites and predators for
control of pests of forests and agriculture. Several classical
examples of biological control successes have been docu¬
mented by Huffaker (1971) and by Coppel and Mertins
(1977). Although some early attempts were made to use
insect pathogens as biological control agents,

'Subtropical Agricultural Research Laboratory, USDA, ARS, Weslaco,
Texas

2 Department of Entomology, Cornell University, Ithaca, New York

much less emphasis has been placed on microbial control
than on other areas of biological control. Recently,
however, interest has increased in the use of insect patho¬
gens (bacteria, fungi, viruses, etc.) as biological control
agents.

Some pathogens are known to produce widespread epizoot¬
ics in nature, often decimating their insect host populations,
yet scientists have been relatively unsuccessful in producing
insect disease outbreaks at will (Carruthers and Soper
1987). Most research efforts have focused on isolating and
culturing highly pathogenic microbes with characteristics
favoring long-term storage and application, as if they were
chemical insecticides. Although this approach has been
successful with some bacterial toxins, the use of living
microbes as replacements for chemical insecticides has
usually resulted in unpredictable and inadequate responses
under field conditions (Fuxa 1987). One of the primary
reasons for these failures is our lack of understanding of the
natural dynamics of host/pathogen life systems. Diseases
develop in complex ways based on biological and physical
associations between the host, the pathogen and the
environment. Understanding these associations in general
and more specifically on a system-by-system basis should
significantly aid our ability to manipulate insect pathogens
for IPM purposes.

In this paper, we will discuss important aspects of insect
disease development in hopes of increasing research interest
and use of fungal pathogens as biological control agents.
Topics of discussion include; basic fungal pathogen
biology, disease epizootiology, and the use of fungal
pathogens for biological control of insect pests. Emphasis
will be placed on use of these biological control agents in
IPM systems and several examples from alfalfa production
will be discussed.

Fungal Pathogen Life Cycles and Biology

Fungal pathogens, in particular, are important natural
biological control agents of many insects and other
arthropods and frequently cause epizootics which may
significantly reduce host populations (MacLeod 1963,
Burges 1981, Carruthers and Soper 1987, McCoy et al.
1988). They are known to cause mycosis in many different
taxa of arthropods and in almost every order of the Insecta
(Bell 1974, Gillespie and Morehouse 1989). They are
known from all life stages of insects (eggs, nymphs, larvae

132

and adults) and are commonly found in aquatic, terrestrial
and subterranean habitats (Ferron 1978).

Although fungal pathogens have much in common with
viruses, bacteria and other insect-pathogenic microbes, they
are unique in many ways (Ferron 1978). Perhaps the most
significant difference lies in the mode of infection; whereas
most entomopathogens infect their hosts through the gut
following consumption, fungi typically infect their insect
hosts by direct penetration of the cuticle (Roberts and
Humber 1981). Although adhesive processes have not yet
been intensively studied, spores of entomopathogens are
thought to adhere to the insect cuticle through a combina¬
tion of physical and chemical interactions (Fargues 1984).

Spore germination and thus infection, is highly dependent
upon environmental moisture and is thought to require free
water (Newman and Carner 1975, Roberts and Campbell
1977, Shimazu 1977, Kramer 1980), but this requirement
may be met by moisture conditions of the microclimate in
the absence of measurable precipitation (Hall and Dunn
1957, Tanada 1963, Ben-Zeev and Kenneth 1980, Kramer
1980, Mullens et al. 1987). Penetration of the cuticle is
accomplished by a germination tube originating from the
spore or by a specialized appressorium which attaches to
the cuticle and gives rise to a narrow penetration peg
(Zacharuk 1973, Roberts and Humber 1981, Boucias and
Pendland 1982; Wraight et al. 1990). In either case,
penetration is both a mechanical and an enzymatic process
(Chamley 1984, St Leger et al. 1987, 1988b; McCoy et al.
1988).

Once the pathogen has passed through the insect cuticle, it
begins its vegetative phase of development in the host's
hemocoel. Different fungal pathogens may produce either
filamentous mycela, discrete yeast-like structures (hyphal
bodies), protoplasts (cells without cell walls) or combina¬
tions of the structures within the hemocoel of the host
(Tyrrell and MacLeod 1972, Roberts and Humber 1981,
Latge et al. 1988). These vegetative stages rapidly colonize
the insect's circulatory system and affect the host by
emitting exoenzymes into the hemocoel that allow the
pathogen to digest and assimilate nutrient material from the
host. The length of the incubation period (time from
infection to host death) varies among species, however
disease development during the vegetative stage is typically
temperature dependent ( Hall 1981, Carruthers et al. 1985a,
Carruthers and Soper 1987).

Pathogens may simply overcome their hosts by consump¬
tion of the available nutrients in the hemocoel (Roberts
1981) or by digestion of host tissues and organs (Brobyn
and Wilding 1977). In cases where fungi overcome their
hosts after a relatively short period of vegetative growth,
toxins produced by the pathogen are presumed to be the
cause of death (Roberts 1981). Compounds toxic to insects
are produced by several entompathogenic fungi both in
vitro and in vivo (Roberts 1981). However, their role in
pathogen-induced mortality in nature is not well under¬
stood.

Shortly after host death, the fungus re-penetrates the host
cuticle from within and terminates in the formation of
sporophores (usually conidiophores) that yield asexual
spores (conidia) which function as dispersive and infective
units. In many species of fungi, the production of conidia
is highly dependent upon moisture (Wilding 1969,

Millstein et al. 1983). Conidia are the infective propagules
of secondary infection and are important in determining the
rate of disease development and spread within a season.
Environmental factors which control conidial production,
survival and germination are critical to the rate of epizootic
development (Carruthers and Soper 1987).

Entomopathogenic fungi survive adverse environmental
conditions or the absence of their host by producing resting
spores or other resistant structures, or as dormant mycelia
in dried insect mummies (Kenneth et al. 1972, Wilding
1973). When dormancy requirements have been met and
other environmental conditions are correct, resting spores
germinate to produce germ conidia which then initiate
mycoses in susceptible hosts.

Disease Epizootiology

Under natural field conditions, disease development and
spread are governed by characteristics of both the host and
pathogen populations and by the environment in which their
interactions occur. In managed ecosystems this scheme is
further complicated as human intervention adds a fourth
component. These four components (host, pathogen,
environment and human intervention) and their interactions
have been represented by the disease tetrahedron (Zadoks
and Schein 1979, Carruthers and Soper 1987). The factors
encompassed by this model are necessarily interrelated and

133

although they may be discussed individually, it must
remembered that their interactions are critical to overall
system behavior.

The Pathogen Population

Properties of the pathogen population which are important
in epizootiology include virulence and pathogenicity,
dispersal and survival in the host's environment, and
inoculum density and spatial distribution (Tanada and Fuxa
1987). Most fungal pathogens are considered highly
virulent relative to other pathogens because they have short
incubation periods, produce copious amounts of secondary
inoculum and can cause a rapid increase in disease preva¬
lence. A fungal species that is pathogenic on a wide range
of hosts may be more likely to persist in an environment
because of the availability of alternate hosts. With a greater
number of susceptible hosts there may be a greater reservoir
of inoculum available to produce an epizootic (Tanada
1963). Pathogen survival in the host's environment is
necessary for the long-term persistence of disease in the
population; as previously mentioned, fungi accomplish this
by producing various types of resting spores or structures.
The ability to survive adverse conditions or periods of host
absence may determine the frequency with which epizootics
occur, because without some means of survival, infection is
dependent upon the movement of inoculum into the host's
habitat (Tanada 1963).

Dispersal is necessary for the rapid spread of disease.
Abiotic factors such as rain or wind may carry spores, or
the movement of infected and uninfected hosts may trans¬
port inoculum from one place to another (Hall and Dunn
1957). Some fungal pathogens cause their hosts to climb to
aerial locations just prior to death (MacLeod 1963); such
behavior may aid dispersal of infective conidia which may
shower down upon or blow to nearby hosts. Pathogen
population density and spatial distribution are key factors in
the development of an epizootic because they affect the
likelihood of contact with viable hosts (Tanada and Fuxa
1987). There has been substantial work on the relationship
between pathogen density and mortality in the laboratory,
generally concluding that mortality increases with increas¬
ing inoculum (Pinnock and Brand 1981). Although this
relationship has rarely been studied under realistic condi¬
tions, field infection studies reveal that natural infection
may follow a logistic response curve (Feng et al. 1990).

The Host Population

Important host population factors that must be considered
are susceptibility, density, movement, and spatial distribu¬
tion of individuals. Although host susceptibility may vary
between geographic regions, there are very few reports of
genetic variability in susceptibility of insects to fungal
disease within local host populations. Paperiok and
Wilding (1979) reported that one of two clones of the pea
aphid ( Acyrthosiphon pisum) from alfalfa was resistant to
infection by Conidiobolus obscurus (Hall and Dunn)
Remaudiere and Keller, compared to the other highly
susceptible clone. Milner (1982, 1985) was able to
characterize field-derived clones of pea aphids, also from
alfalfa, as resistant or susceptible based on percent mortal¬
ity when exposed to a particular isolate of Pandora
neoaphidis (Remaudiere and Hennebart) Humber
( — Entomophthora aphidis, Erynia neoaphidis). Undoubt¬
edly the population dynamics of variability in host resis¬
tance and pathogen virulence would have an impact on
epizootiology and the long-term stability of disease inci¬
dence in the population; however, this is an area of research
which lacks sufficient study.

Host density may directly influence the rate of disease
build-up by increasing host-to-host contact or by increasing
the levels of secondary inoculum (Carruthers et al. 1985b,
Benz 1987, Watanabe 1987). In some situations the spatial
arrangement of hosts may be as important as the actual
density of the host population. If hosts are highly aggre¬
gated it may be difficult for the pathogen to disperse
between aggregates, unless infected hosts or pathogen
propagules are highly mobile.

The Abiotic Environment

Abiotic environmental factors such as moisture, tempera¬
ture, and solar radiation affect the rates of many biotic
processes important for disease development; moreover,
they may determine whether or not infection can occur at
all. Atmospheric moisture is often considered the most
important abiotic factor in the epizootiology of fungal
diseases (Nordin et al. 1983, Fuxa and Tanada 1987). As
mentioned in the previous section, germination and sporula-
tion of most fungi are highly dependent upon moisture.
However, fungi are able to acquire moisture in forms other
than just precipitation, such as from dew set on plant
surfaces or from the boundary layer of the host itself. The

134

humidity of the microclimate, such as that associated with a
dense plant canopy, may be much higher than that of the
ambient air, even under dry conditions (Tanada 1963,
Kramer 1980, Fuxa and Tanada 1987).

Temperature-dependent processes which directly affect
disease progress are the rate at which insects develop, the
rate of fungal development within the insect, and the rate
and quantity of spore production (Benz 1987). Most fungal
pathogens, however, do well at all temperatures suitable for
insect growth (Benz 1987). Fungal spores, particularly
conidia, may be very sensitive to solar radiation and
desiccation (Ignoffo et al. 1977, Carruthers et al. 1988).
Spore longevity, germination potential and secondary
infection levels may be increased if the microhabitats, such
as a dense plant canopy, can protect the spores from
harmful environmental conditions.

Interactions

Secondary infection is crucial in the epizootic development
of fungal diseases, in that repetitive cycles of infection
result in a rapid increase in disease prevalence (Zadoks and
Schein 1979). Pathogens which have many secondary
cycles during one or a few generations of the host are most
likely to cause dramatic epizootics. This is particularly true
if this process occurs in an environment that is conducive to
pathogen survival and host infection. Transmission effi¬
ciency has been shown theoretically to be an important
parameter related to the rate at which secondary infections
occur and thus of epizootic progress (Anderson and May
1980). Transmission of fungal pathogens occurs primarily
through the insect integument; this may be accomplished
through direct host-to-host contact or by host contact with
infective spores in the environment, such as conidia
deposited on plant surfaces. Disease transmission is thus
influenced by host and inoculum density and spatial
distribution.

Theoretical models have demonstrated that a pathogen can
only be maintained within a host population if host density
exceeds a threshold value, and this value has been defined
as inversely proportional to transmission efficiency (Ander¬
son and May 1980). Brown and Nordin (1982) determined
that the threshold density for Zoophthora sp. epizootics of
alfalfa weevil populations is 1.7 weevils per stem. Subse¬
quent field studies showed that larval densities below this
threshold could not support a Zoophthora epizootic.

Nordin et al. (1983) concluded that the initiation of disease
was best correlated with degree-day accumulations, and that
epizootic dynamics were controlled by atmospheric mois¬
ture levels as long as the host population remained above
the threshold density. Further methods of enhancing the
development and spread of Zoophthora sp. are discussed in
a subsequent section.

In recent years, systems analysis and modeling have been
recognized as useful aids to understanding the complex
dynamics of fungal epizootics (Anderson and May 1980,
Carruthers 1985, Brown 1987, Carruthers and Soper 1987,
Carruthers et al. 1988, Onstad and Carruthers 1990).
Relatively simple models have been used to explore basic
questions in epizootiology (Anderson and May 1980,
Anderson 1982, Brown and Nordin 1982, Brown 1984,
Regniere 1984, Hochberg 1989). More detailed simulation
models of specific host/pathogen systems have also been
developed for a few fungal pathogens (Carruthers et al.
1986, Carruthers et al. 1988). These models and the
techniques of systems science have proven to be very useful
in guiding the collection, analysis and synthesis of informa¬
tion about the dynamics of host and pathogen life systems.
In addition to providing basic understanding, models can
assist in pest management decision making and can be used
to help design and evaluate strategies for employing fungal
pathogens in biological control programs. Expert systems
(Logan 1988) and intelligent modeling systems (Larkin et
al. 1988, Larkin and Carruthers 1990) are becoming
available to make the analytical power of systems tech¬
niques more accessible to scientists and pest management
specialists who may have little experience with modeling.

Although epizootics of fungal disease are known to cause
major declines in pest populations, very little information is
available on the long-term regulatory ability of these
pathogens. This is particularly true during periods of
enzootic activity when disease prevalence is low and signs
of host infection are difficult to detect. Disease assessment
is complicated further if pathogen-induced mortality occurs
in habitats difficult to observe, as is the case in soil or
aquatic ecosystems. Over the past two decades, however,
several significant descriptive and experimental studies on
disease dynamics have provided new insights into the
population biology and possible use of fungi as biological
control agents (MacLeod et al. 1966, Soper et al. 1976,
Kish and Allen 1978, Soper and MacLeod 1981, Nordin et

135

al. 1983, Carruthers et al. 1985b). Their effects in regulat¬
ing insect pests of alfalfa are particularly noteworthy and
will be discussed in detail below.

Biological Control of Alfalfa Pests Using
Fungi

Naturally occurring epizootics caused by fungal pathogens,
particularly those caused by fungi in the Order
Entomophthorales, are frequently noted in both natural and
managed ecosystems (including alfalfa) (Pickford and
Riegert 1964, Wilding 1975, Soper et al. 1976, Mohamed
et al. 1977, Wilding 1981, Nordin et al. 1983, Carruthers
et al. 1985b, Lewis and Cossentine 1986, Carruthers and
Soper 1987, McCoy et al. 1988). Due to the catastrophic
impacts these pathogens have on their host populations,
they hold significant potential for biological control of
many pest species (Ferron 1978, McCoy 1981, Wilding
1981, Carruthers and Soper 1987).

The use of entomopathogenic fungi for biological control in
managed ecosystems has followed four basic strategies; 1)
Awareness and conservation of natural biological control,

2) Enhancement of biological control through active
manipulations, 3) Augmentation of natural enemies, and 4)
Introduction of exotic natural enemies (Ignoffo 1985,
McCoy et al. 1988). We will provide examples of research
that demonstrates current levels of expertise and important
aspects of these areas of biological control of insect pests
on alfalfa.

Alfalfa, Medicago saliva, is the world's most valuable
forage crop, providing the best food value for all classes of
livestock. Forages provide in excess of half the feed units
for livestock, which in turn provides half of the nutrients
used by American consumers. In addition to its nutritional
value, alfalfa is important since it has high nitrogen fixing
capabilities and because it is a soil-conserving perennial
crop. Alfalfa is grown on over 26 million acres in the
United States and is exceeded in U.S. production acreage
only by com, soybeans and wheat. Alfalfa, however,
exceeds each of these crops in protein output per unit area,
producing an overall U.S. cash value estimated in excess of
$6 billion per year (USDA 1989). Alfalfa provides food
and shelter for a diverse group of insects, several of which
are major economic pests. Although the pest complex varies
regionally, six major groups (weevils, plant bugs, aphids,
grasshoppers, leafhoppers and leafminers) cause damage to

alfalfa across the United States. Although other insects can
periodically become pests, the main pest species in the
northeastern and north central United States are the alfalfa
weevil ( Hypera postica), spotted alfalfa and pea aphids
( Iherioaphis maculata and Acyrthosiphon pisum, respec¬
tively), the alfalfa blotch leafminer ( Agromyza frontella)
and the potato leafhopper ( Empoasca fabae). Biological
control has been researched and/or implemented for most of
these pests, and control has been successful in several cases
(for parasitoid and predator examples see article by G.
Cunningham, USDA-APHIS). The alfalfa weevil, alfalfa
blotch leafminer and most aphid species are currently under
biological control throughout most of the northeastern and
north central United States.

Of the five major pests of alfalfa listed in the previous
paragraph, four are known to have fungal pathogens that
exert high levels of mortality under some field situations.
Some of these pathogens were found to occur naturally
while others were introduced as exotic biological control
agents.

Although several host/pathogen associations from the
alfalfa production system could be discussed, the following
host and pathogens were selected as they provided examples
of the four general strategies of biological control now used
for fungi. These examples also illustrate the potential
impacts of fungal diseases on four of the most important
alfalfa pests of the northeastern and north central United
States. While some of these pathogens being discussed are
currently operating as fully-functional biological control
agents, others are still in the research and developmental
stages. We hope to provide an up-to-date status report on
each of the pathogens discussed, although limited space
required exclusion of many specific details.

Awareness and Conservation of Naturally Occurring
Pathogens

As previously mentioned, dramatic epizootics which cause
high rates of mortality have been reported from many
different cropping situations, including several examples
from alfalfa (Arthor 1886ab, Hall and Dunn 1957,
Harcourt et al. 1974, Harper 1987). Rapid disease spread
and high rates of mortality frequently attract attention and
generate interest in fungal pathogens for use in pest control
programs. The more typical situation, however, is an
enzootic level of disease that often escapes our notice yet

136

causes significant pest mortality. The endemic occurrence
of a pathogens at low-moderate levels may play an ex¬
tremely important role in maintaining pest levels below
economic thresholds. It is thus important to study and to
evaluate the relative impact of all natural mortality factors
(including fungi) on the population dynamics and regulation
of any particular pest.

Recognition of natural pathogen-induced mortality, of any
magnitude, is the first step to understanding and controlling
the biotic and abiotic factors which influence infection
rates. This understanding may allow producers to avoid
pest related concerns or unneeded pesticide applications in
some circumstances. Gaining insight into the operation of a
naturally occurring disease is also the first step toward
manipulating biological control agents for pest management
purposes. Occasionally, insignificant changes in cultural
practices may allow us to take advantage of natural sources
of pest control that might otherwise be unavailable (see
Brown and Nordin 1986, Carruthers et al. 1986, Carruthers
and Soper 1987, Harper 1987). The first insect/ pathogen
life system from alfalfa that we will discuss provides such
an example.

The pea aphid, Acyrthosiphon pisum (Harris), is a cosmo¬
politan pest of many legumes which at high densities causes
yellowing and stunting of alfalfa. In most areas of the
United States pea aphid densities do not usually reach
economic threshold levels. This natural control is presum¬
ably the result of mortality caused by introduced parasites
and predators (Anonymous 1979, Edwards et al. 1979).

The pea aphid has been found to be highly susceptible to
several species of fungal pathogens which may cause
dramatic epizootics, yet fungal pathogens have not often
been considered as important biological control agents in
alfalfa pest management programs.

Pandora neoaphidis (= Entomophthora aphidis, Erynia
neoaphidis Remaudiere and Hennebert) Humber is the most
common pathogen of pea aphids in the northern United
States (Hutchison and Hogg 1985, Hural unpublished data),
Australia (Cameron and Milner 1981), and Europe (Wild¬
ing 1975). Pandora neoaphidis is apparently obligately
asexual; no sexual or resting structures have been found in
nature (Milner et al. 1980, Hural unpublished data).
Pandora neoaphidis has a wide host range among aphids
and has caused devastating epizootics in populations of
Myzus persicae, Aphis fabae, Capitophorus horni
(Missonier et al. 1970) and others (Milner et al. 1980).

Warm, wet conditions are generally necessary to cause
epizootics (Rockwood 1950, Missonier et al. 1970, Milner
et al. 1980, Milner and Bourne 1983).

Pandora neoaphidis infections begin to appear in early
spring, soon after the aphids have emerged. Infections
occur throughout the growing season and are greatly
influenced by environmental conditions, particularly
atmospheric humidity (Wilding 1975). Saturated condi¬
tions are required for conidial germination (Yendol 1968)
and sporulation. It is unknown how this fungus overwin¬
ters, as resting structures have not been found. It has been
suggested that it survives vegetatively in aphid cadavers
(Wilding 1973, Courtois and Latteur 1984). Latteur (1977)
suggested that conidia might be capable of surviving in
damp soil but his data was not conclusive. Conidia do not
survive long in air that is not saturated (Brobyn et al. 1985,
1987).

Recent field studies in Wisconsin and New York
(Hutchinson and Hogg 1985; Hural, unpublished data)
showed that P. neoaphidis infection is one of the most
important mortality factors in pea aphid populations in
alfalfa, typically outweighing parasites and predators. In
the Wisconsin study, the mean number of embryos per
adult decreased with the infection rate, which Hutchinson
and Hogg (1985) assumed would result in a proportional
reduction in birth rates. Time-specific life tables con¬
structed for these populations indicated that the impact of
P. neoaphidis on pea aphid density was mostly due to this
reduction in adult fecundity.

In New York, the prevalence of P. neoaphidis infection is
highest during spring and fall, and decreases during the
warmest period of mid-summer. Because high temperatures
also inhibit aphid development, pea aphid densities exhibit
a similar pattern. Pandora neoaphidis infection is usually
present throughout the entire growing season but only
causes epizootics in the fall. The relative importance of P.
neoaphidis and other sources of natural mortality such as
parasitism by Aphidius and Praon species appeared to
change throughout the season but more research is needed
to determine exactly how these mortality factors interact in
the field. It may, however, be possible to enhance or
manipulate infection in this system by altering cultural
practices or by providing inoculum earlier in the alfalfa
production season.

137

Enhancement of Naturally Occurring Fungal
Pathogens

Depending on specific details of the host and pathogen
population and the associated ecosystem, enhancement of
disease development and spread has been accomplished by a
number of different methods (Hostetter and Ignoffo 1978,
McCoy et al. 1988). Some specific methods include habitat
manipulation, altering cultural practices such as planting or
harvesting (Brown and Nordin 1986), and managing
controlled inputs such as irrigation or pesticides (McCoy et
al. 1976, Kish and Allen 1978, Sprenkel et al. 1979;

Hamm and Hare 1982). The following discussion of
Zoophthora sp. mycosis of the alfalfa weevil is an example
of how early harvesting has been used to enhance the
effects of naturally occurring populations of this pathogen
in an IPM program.

Zoophthora sp. mycosis of the alfalfa weevil: The fungal
pathogen Zoophthora phytonomi (Arthur) Batko is known
for its natural regulatory effects on populations of the
clover leaf weevil, Hyper a punctata in North America
(Arthor 1886a,b; USDA 1956) and a related weevil species,
H. variabilis, in Israel (Ben-Zeev and Kenneth 1980).
Following the introduction of the alfalfa weevil (H. postica)
into North America, 17 years of extensive surveys revealed
no infections in H. postica populations (Puttier et al. 1978)
until 1973, when a pathogen similar to Zoophthora
phytonomi was first observed causing significant epizootics
in alfalfa weevil populations throughout southern Ontario,
Canada (Harcourt et al. 1974). Harcourt et al. (1984)
suggest that this pathogen was one of the primary reasons
for the major alfalfa weevil decline seen in Ontario during
the mid 1970's. Subsequently, this pathogen spread into
other North American alfalfa production areas where it also
caused significant mortality of H. postica (Puttier et al.
1978, Gardner 1982, Nordin et al. 1983, Pienkowski and
Mehring 1983). Taxonomic questions associated with this
pathogen still exist, some evidence suggests that the
pathogen that infects alfalfa weevils is a different species of
Zoophthora than is found infecting the clover leaf weevil
(Harcourt et al. 1981), while more recent finding suggest
that the pathogen is actually the same pathogen (Z.
phytonomi ) that is found infecting the clover leaf weevil (R.
A. Humber, unpublished data).

Due to uncertainty associated with the origin of this
pathogen, it is unclear whether the pathogen causing
mycosis in the alfalfa weevil was introduced into North

America in conjunction with weevil introductions or exotic
parasite releases, or if the pathogen switched from another
North American host species, possibly H. punctata, onto
H. postica. Although the origin of this pathogen is still
unknown (Puttier et al. 1978), the fungus has become well
established as a major natural biological control agent of
the alfalfa weevil throughout much of its range. Epizootio-
logical studies have shown that disease prevalence varies
between sites and seasons, but disease levels from 30-70%
are not uncommon at the time of peak larval occurrence
(Puttier et al. 1980), and have approached 100% late in the
host's developmental cycle, particularly when densities are
high (Harcourt et al. 1974, Nordin et al. 1983). Disease
levels of this magnitude produce larval mortality between
65-90% with an additional 40-50% expressed in the less-
visible pupal stage (Harcourt et al. 1974).

Brown and Nordin (1982) determined that a threshold host
density (1.7 weevils/ alfalfa stem) was required for
Zoophthora sp. to induce epizootics. Although this
threshold varied with environmental moisture, low host
densities were thought to limit early-season spread of
mycosis. Using laboratory, field and simulation modeling
experiments, Brown and Nordin (1986) developed early
harvesting strategies that maximized the development and
spread of Zoophthora sp. even when weevil populations
were lower than critical levels for epizootic development.
This was accomplished by harvesting the alfalfa when
Zoophthora sp. mycosis of weevil larvae was first ob¬
served. Early harvesting is thought to increase disease
prevalence because it: 1) concentrates larvae in windrows;
2) damages or stresses larvae making them more suscep¬
tible; and 3) alters micro-environmental conditions (in¬
creases moisture) to enhance sporulation, germination and
thus infection. Research fields managed using this tech¬
nique showed higher disease prevalence, which demon¬
strates that enhancement of fungal diseases is not only
possible but is practical and economical as it is based on
accepted production practices and equipment (Brown and
Nordin 1986). However, a better understanding of the
spatial distribution and dynamics of this pathogen is
necessary to effectively implement this strategy in a
statewide IPM program (Brown and Nordin 1986).

138

Augmentation of Fungai Pathogens for Biological
Control

Augmentation (increasing pathogen inoculum density)
through the development of microbial insecticides has
received substantial attention and will only briefly be
discussed here as it is the focus of several other articles
(Burges and Hussey 1971, Burges 1981, McCoy 1990). It
must be said, however, that augmentation has not only
taken the form of microbials applied like pesticides with the
goal of high acute host mortality, but has also been used to
initiate epizootics prematurely (Ignoffo et al. 1976, Ferron
1981) or in situations where epizootics would not naturally
develop (McCoy 1981, Riba 1984). In these cases, disease
development is dependent not only on the efficacy of the
fungal material originally applied to the target host but also
on the pathogen's ability to become established in the
environment and produce secondary inoculum capable of
polycyclic infection. These methods of pathogen augmen¬
tation require detailed information about host, pathogen and
disease dynamics which have only recently been researched
(Allen et al. 1978, Carruthers and Soper 1987, McCoy et
al. 1988).

Use of Zoophthora radicans (Brefeld) Batko for control of
Empoasca fabae : Although several insect species attack
alfalfa throughout the United States, the potato leafhopper,
Empoasca fabae, is currently considered to be the most
serious economic threat to alfalfa production (Gyrisco et al.
1978). Feeding by the potato leafhopper causes severe
yellowing and stunting both in seedling and established
alfalfa stands. Losses combined with control expenditures
cost US alfalfa producers millions of dollars annually.
Entomologists in Illinois and Wisconsin estimate that in the
average season, yields are decreased approximately 20% in
terms of quantity with additional losses of approximately
20% through reduced protein content. In addition, E.
fabae is a major pest and virus vector in numerous crops
(potatoes, soy and other beans, clovers and 100+ other
cultivated plants), causing significant losses (Gyrisco et al.
1978).

Successful biological control agents, both insect parasitoids
and microbial pathogens, are currently helping to reduce
populations of other major alfalfa pests (e.g., the alfalfa
weevil, the pea aphid and the alfalfa blotch leafminer) well
below economic thresholds in many areas of the country
(Dunbar and Hower 1976, Brown and Nordin 1982, Hower
and Davis 1984). No such biological control organisms are

being used to manage potato leafhoppers. After five years
of effort, the ARS-European Parasite Laboratory has
terminated its program aimed at locating beneficial insects
to control the potato leafhopper. Currently the only control
option for leafhopper management is the application of
chemical insecticides, which is not only costly but inter¬
feres with other highly successful biological control
programs in alfalfa. More insecticides are now being
applied to alfalfa in the central and eastern US to control E.
fabae than all other insect pests combined. Microbial
control of E. fabae may provide an alternate management
tool in the alfalfa production system that would be more
compatible with existing alfalfa pest management programs,
provide less hazard to livestock and in general, would be
more environmentally sound.

Initial research on the effects of Zoophthora radicans on
Empoasca fabae was conducted through cooperative efforts
of various agencies including USDA-Agricultural Research
Service, USDA-Cooperative States Research Service and
US-Agency for International Development. The USDA-
ARS, Plant Protection Research Unit joined in an effort
with the Boyce Thompson Institute for Plant Research, the
Illinois Natural History Survey and Cornell University to
explore the possibility of microbial control of the potato
leafhopper using Z. radicans isolates collected from a
related leafhopper species in Brazil. Through efforts made
in this program, it became clear that Z. radicans had
significant potential for biological control of E. fabae
(Wraight et al. 1985, 1986; McGuire et al. 1987a, b;
Galaini-Wraight et al. 1990).

Epizootic levels of Z. radicans are commonly seen in
Brazilian leafhopper populations and are known to cause
significant population reductions in both bean and cowpea
crops (Wraight et al. 1985, Galaini-Wraight et al. 1990).
Pathogen isolates from these epizootics were imported to
the US and tested against the potato leafhopper under
laboratory conditions at the Boyce Thompson Institute.
These studies show that pathogen levels as low as 3.3
spores/mm2 cause significant levels of infection (>70%) in
late-instar nymphs. A ten-fold increase in dose (33.0
spores/mm2) repeatedly caused 100% mortality. Time from
infection to host death and conidial production (pathogen
multiplication) is very rapid, 2-4 days depending on host
age and environmental conditions.

Zoophthora radicans is a very attractive pathogen for use in
biological control since it may either be introduced in an

139

inoculative release or used more as a microbial pesticide
through inundative releases (here we will discuss this
second aspect). This pathogen can be easily produced in
large quantities and stored for relatively long periods using
the marcescence technology developed by the ARS-Plant
Protection Research Unit (Soper and McCabe US-patent
4,530,834). Fungal mycelium can be grown in liquid
media, dried, milled, stored and then applied in the field
where it will rehydrate and produce infective conidia in
patterns very similar to fungi growing directly from a
newly killed host.

This material has been applied to small field plots and
evaluated for its sporulation potential and pathogenicity to
caged populations of E. fabae. These studies not only
showed primary infection levels as high as 70% in caged
leafhoppers within 24 hours, but also high infection levels
in natural populations through mycosis of leafhoppers
outside the test cages (Wraight et al. 1986). Inoculation
densities of approximately 5.0 g/m2 of dried hyphae were
applied to a total of 12 m2, inducing infection levels as high
as 92% across a 0.75 acre test field over a three week
sample period. The epizootic reduced leafhopper popula¬
tions from 1.7 to 0.08 leafhoppers per stem. A second
study conducted in an adjacent field gave similar results.

In Illinois, a similar release (using infected cadavers in a
potato field) produced the same general results (McGuire et
al. 1987b). Although not clearly documented in the year of
release, Z. radicans became established in central Illinois
and has caused substantial epizootics in leafhopper popula¬
tions over the last few seasons. Infection levels as high as
90 % have been recorded among leafhoppers in beans where
the epizootics have been most evident. Since its introduc¬
tion into Illinois, Z. radicans has spread approximately 20
miles from the original release sites and has caused major
leafhopper declines in bean, potato and alfalfa crops.
Additional introductions in Illinois and New York have
continued to cause similar epizootics (Carruthers unpub¬
lished data).

Although a single isolate of Z. radicans has shown signifi¬
cant potential to control E. fabae under natural field
conditions, several additional isolates have been collected
from South America and Europe. A number of these
isolates have shown equal or higher virulence to E. fabae in
laboratory screenings than the material originally tested and
released in the field. These isolates may also possess other
beneficial characteristics, such as a faster speed of kill or

better overwintering ability, but are yet to be evaluated.
Screening of this germplasm in the laboratory is currently
underway and further evaluation of selected pathogen
isolates will be conducted in the field following USDA-
Animal, Plant Health Inspection Service and US Environ¬
mental Protection Agency approvals (summer 1990).
Hopefully, these studies will allow us to use Z. radicans
either as a myco-insecticide or as an introduced biological
control agent throughout the full geographic range of the
potato leafhopper (Carruthers and Soper 1987).

Introduction of Exotic Fungal Pathogens

Classical biological control (the introduction of exotic
fungi, either new species or more virulent strains of a
species that already exists) has received very little attention
in the field of insect pathology (Roberts 1978). Some
reports of attempts to establish exotic fungi include the use
of Coelomomyces stegomyiae Keilin for mosquito control
(Laird 1967), Entomophaga aulicae (Reichardt in Bail)
Humber for control of the browntail moth (Speare and
Colley 1912), Entomophthora erupta for control of the
green apple bug (Dustan 1923) and Zoophthora radicans
for control of the spotted alfalfa aphid (Hall and Dunn
1957, Milner et al. 1982). Although releases were made in
each of these cases, detailed population assessments
documenting the effects of these fungi were rarely con¬
ducted. The introduction of Zoophthora radicans into
Australia provides an excellent example of an introduction
of a fungal pathogen and its impact on the host population.

Introduction of Zoophthora radicans to control the spotted
alfalfa aphid: In 1979, different isolates of Z. radicans
were introduced into New South Wales (NSW), Australia in
a cooperative effort between the USDA-ARS Plant Protec¬
tion Research Unit and CSIRO for biological control of the
spotted alfalfa aphid (Milner and Soper 1981). Zoophthora
radicans quickly became established in NSW, caused
epizootics in the pest population and spread over 200 miles
from the original release site. In the third field season, Z.
radicans was again found to induce epizootics (up to 74%
prevalence) in sampled spotted alfalfa aphid populations
(Milner and Lutton 1986) and has spread over 300 km from
the original release sites (Milner and Soper 1981, Milner et
al. 1980, 1982; Milner and Lutton 1983, Milner personal
communication). This example demonstrates the potential
for an exotic fungal pathogen to be successfully introduced,
become established and provide ongoing regulation of an

140

insect pest of alfalfa. Further exploration and collection of
fungal isolates from other important pest species could
reveal additional untapped pathogens with potential for
insect pest management.

Summary of Insect/Fungal Biological Control in
Alfalfa

The four examples discussed above, demonstrate how
fungal pathogens may be used in a number of ways to help
manage insect numbers in alfalfa. Alfalfa is a particularly
good candidate crop for a multifaceted biological control
program using fungal pathogens. It is a low-value crop •
(compared to fresh market fruit and vegetables) so produc¬
ers seek to minimize cash inputs for its management; the
crop is primarily consumed by livestock, thus substantial
cosmetic damage may be tolerated; and the structure of the
crop canopy provided a moist and protected environment
where disease transmission readily occurs. With additional
research, pesticide application in alfalfa production may be
further reduced while increasing the level of insect control.

Conclusions: Research Needs and Future
Application of Biotechnology

Entomopathogenic fungi are important natural regulators of
many arthropod populations, including several pest species.
A variety of strategies have been used successfully to
manipulate fungi in biological control programs. While
there have been fewer documented biological control
successes with fungal pathogens than with parasitoids and
predators, much less effort and capital have been expended
to understand and manage them. Most of the research
conducted on fungal pathogens of insects has emphasized
the use of these organisms as microbial insecticides.
Application of pathogens by inundative release may indeed
prove to be a useful tactic, but fungi should not be consid¬
ered as direct replacements for chemical insecticides.

Fungi are complex organisms that interact with their hosts
and the environment in intricate ways. Intelligent use of
fungi as biological control agents will require detailed
knowledge of their pathogenesis, epizootiology and
interactions with other components of the ecosystems in
which they are to be used. This will become increasingly
important as scientists begin to genetically alter fungal
pathogens to improve their efficacy as biological control
agents.

Some researchers feel that fungus-induced epizootics are
too dependent on high host densities and environmental
conditions, particularly moisture, to be effective biological
control agents (Bucher 1964), however, the majority of
researchers who study insect pathogens believe that fungi
will play a vital role in IPM systems in the near future
(Allen et al. 1978, Carruthers and Soper 1987, Fuxa 1987,
McCoy et al. 1988).

Admittedly, fungi are limited in their ability to control
insect pests. For example, not all pest species are suscep¬
tible to fungal pathogens, and even if susceptible, the target
hosts may live in an environment that is not conducive to
fungal infection and transmission. As mentioned previ¬
ously, fungal pathogens are highly dependent on moisture
for spore germination and infection. They are also ad¬
versely affected by high temperatures. Pest control using
any biological control agent will not be as certain or
repeatable as the action of chemical insecticides. For this
reason, we need to alter our expectations, develop inte¬
grated control programs using multiple tactics and continue
to improve our efficiency in using fungi and other patho¬
gens as biological control agents. Most importantly, we
should develop a research agenda aimed at solving specific
problems associated with the development and application
of microbial agents for biological control.

Although the list of needed research on fungal pathogens
and their use in biological control is long, we would like to
highlight five specific areas needing additional attention.
First, it is important that fungal pathogen germplasm be
collected now and preserved for future evaluation and use.
Second, continued expansion of research on fungal
epizootiology and ecology is needed so that we may
understand disease dynamics, the impact of pathogens on
host populations and the factors that limit disease develop¬
ment and spread under field conditions. Third, we need to
study genetic aspects of host and pathogen populations that
affect the establishment, spread and maintenance of disease
in insect populations. We must recognize that natural
populations of fungal pathogens contain high levels of
genetic variation which could contribute to their ability to
regulate host populations. Furthermore, we must explore
the potential for the evolution of resistance in the host
population before it becomes a problem in the field.

Fourth, we must increase our efforts to integrate the use of
pathogens with other control tactics. Until control efforts
use multiple tactics to manage a complex of pest species
IPM will remain in its infancy. Last, we need to expand

141

the use of innovative techniques in addressing problems
associated with biological control agents.

New biotechnologies provide us with tools to address
questions that we have never been able to ask before. We
expect innovative biological techniques to be used to
manipulate desirable traits, and thus improve the effective¬
ness of some fungal pathogens. Recombinant DNA
techniques are being used to study the mechanisms of
pathogenicity and virulence at the molecular level. For
example, fungal enzymes and associated genes involved in
penetration of the insect cuticle have now been identified
(St. Leger et al. 1987, 1988a). Knowledge of these genes
and gene products may eventually lead to genetic alteration
of fungal pathogens. Using transformation systems
developed for other types of fungi (Shimazu 1987, Yoder et
al. 1987), it may be possible to clone the genes responsible
for these products and transfer them to fungi that may have
poor penetration ability but are well adapted to a particular
environment. The genes involved in toxin production are
also candidates for cloning.

Biotechnological methods also are being applied to improve
methods for field monitoring of insect fungal diseases.
Enzyme linked immunosorbent assay (ELISA) is used to
detect a variety of fungi in plant tissues (Mendgen 1986)
and has shown promise for detecting fungal cells of
Entomophaga maimaiga in gypsy-moth larval hemolymph
(Hajek et al. 1989). Other techniques such as isoenzyme
polymorphisms (Boucias et al. 1982, Micales et al. 1986)
and restriction fragment length polymorphisms (RFLP's)
will become increasingly useful in detecting different fungal
pathogen strains (Hajek et al. 1990).

It is important however, that we do not get lost in the
techniques and lose sight of the important question: How
can we improve our ability to manage pest problems while
improving both our economic and environmental situation?
Despite the promise of new technologies, successful
biological control depends on a fundamental knowledge of
host and pathogen biology, not only at the molecular level,
but at the cellular, organismal, population, and ecosystem
levels as well. It is evident that specialists from many
subdisciplines of entomology, genetics, mycology, systems
science and other fields of study will be required to
successfully identify, manipulate and use fungi for biologi¬
cal control purposes in the future.

References Cited

Allen, G. E., C. M. Ignoffo and R. P. Jacques [eds.].

1978. Microbial control of insect pests: Future strategies
in pest management systems. NSF-USDA-Univ. of
Florida, Gainesville.

Anderson, R. M. 1982. Theoretical basis for the use of
pathogens as biological control agents of pest species.
Parasitology 84: 3-33.

Anderson, R. M., and R. M. May. 1980. The population
dynamics of micro-parasites and their invertebrate hosts.
Philosophical Trans. Roy. Soc. Brit. 291: 451.

Anonymous. 1979. Alfalfa Seed Insect Pest Management.
Western Regional Extension Publication 0012.

Arthor, J. C. 1886a. Disease of clover-leaf weevil.
Entomophthora phytonomi Arthor. 4th Annual Rep NY
Agric Exp. Sta. 1885: 258: 120.

Arthor, J. C. 1886b. A new larval Entomophthora. Bot.
Gaz. 11: 14.

Bell, J. V. 1974. Mycoses. In G. Cantwell [ed.], Insect
Diseases, Vol. I. Marcel Dekker Inc, New York.

Ben-Zeev, I., and R. G. Kenneth. 1980. Zoophthora
phytonomi and Conidiobolus osmodes, two pathogens of
Hypera species coincidental in time and place.
Entomophaga 25 : 171.

Benz, G. 1987. Environment. In J. R. Fuxa and Y.
Tanada [eds.], Epizootiology of Insect Diseases. John
Wiley, New York.

Boucias, D. G., C. W. McCoy and D. J. Joslyn. 1982.
Isozyme differentiation among 17 geographical isolates of
Hirsutella thompsonii. J. Invertebr. Pathol. 39: 329.

Boucias, D. G., and J. C. Pendland. 1982. Ultrastructural
studies on the fungus, Nomuraea rileyi, infecting the
velvet-bean caterpillar, Anticarsia gemmatalis. J.

Invertebr. Pathol. 39: 338.

142

Brobyn, P. J., and N. Wilding. 1977. Invasive and
developmental processes of Entomophthora species infect¬
ing aphids. Trans. Brit. Mycol. Soc. 69: 349.

Brobyn, P. J., N. Wilding and S. J. Clark. 1985. The
persistence of infectivity of conidia of the aphid pathogen
Erynia neoaphidis on leaves in the field. Ann. Appl. Biol.
107: 365.

Brobyn, P. J., N. Wilding, and S. J. Clark. 1987.
Laboratory observations on the effect of humidity on the
persistence of infectivity of conidia of the aphid pathogen
Erynia neoaphidis. Ann. Appl. Biol. 110: 579-584.

Brown, G. C. 1984. Stability in an epizootiological model
with age-dependent immunity and seasonal host reproduc¬
tion. Bull. Math. Biol. 46: 139.

Brown, G. C. 1987. Modelling. In J. R. Fuxa and Y.
Tanada [eds.], Epizootiology of Insect Diseases Wiley,

New York.

Brown, G. C., and G. L. Nordin. 1982. An epizootic
model of an insect-fungal pathogen system. Bull. Math.
Biol. 44: 731.

Brown, G. C.and G. L. Nordin. 1986. Evaluation of an
early harvest approach for induction of Erynia epizootics in
alfalfa weevil populations. J. Kan. Entomol. Soc. 59: 446.

Bucher, G. E. 1964. The regulation and control of insects
by fungi. Ann. Soc. Entomol. Quebec 9: 30.

Burges, H. D. [ed.]. 1981. Microbial control of pests and
plant diseases: 1970-1980. Academic Press, London.

Burges, H. D., and N. W. Hussey [eds.], 1971. Microbial
control of insects and mites. Academic Press, New York.

Cameron, P. J., and R. J. Milner. 1981. Incidence of
Entomophthora spp. in sympatric populations of
Acyrthosiphon kondoi and A. pisum. New Zeal. J. Zool. 8:
441-446.

Carruthers, R. I. 1985. The use of simulation modeling in
insect-fungal disease research. Proc. Summer Computer
Simulation Conf, July 22-24 Chicago, Ill.

Carruthers, R. I., Z. Feng, M. E. Ramos and R. S. Soper.
1988. The effect of solar radiation on the survival of
Entomophaga grylli conidia. J. Invertebr. Pathol. 52: 154.

Carruthers, R. I., Z. Feng, D. S. Robson and D. W.
Roberts. 1985a. In vivo temperature-dependent develop¬
ment of Beauveria bassiana mycosis of the European com
borer. J. Invertebr. Pathol. 46: 305.

Carruthers, R. I., D. L. Haynes and D. M. MacLeod.
1985b. Entomophthora muscae mycosis of the onion fly,
Delia antiqua. J. Invertebr. Pathol. 45: 81.

Carruthers, R. I., T. L. Larkin and R. S. Soper. 1988.
Simulation of insect disease dynamics: an application of
SERB to a rangeland ecosystem. Simulation 51: 101-109.

Carruthers, R. I., and R. S. Soper. 1987. Fungal diseases.
In J. R. Fuxa and Y. Tanada. eds., Epizootiology of Insect
Diseases. Wiley, New York.

Carruthers, R. I., G. W. Whitfield, R. L. Tummala and D.
L. Haynes. 1986. A systems approach to research and
simulation in of insect pest dynamics in the onion
agroecosystem. Ecol. Modelling 33: 101-121.

Chamley, A. K. 1984. Physiological aspects of destruc¬
tive pathogenesis in insects by fungi: A speculative review.
In J. M., anderson, A. D. M. Raynor and D. W. H.

Walton [eds.], Invertebrate-Microbial Interactions, Cam¬
bridge Univ Press, Cambridge.

Coppel, H. C. and J. W. Mertins. 1977. Biological insect
pest suppression. Springer-Verlag, New York.

Courtois, P. and G. Latteur. 1984. Etude quantitative de
la survie des corps hyphaux d 'Erynia neoaphidis en
fonction de la temperature et de l'humidit£ relative.
Parasitica 40: 211-220.

Dunbar, R. B., and A. A. Hower. 1976. Relative toxicity
of insecticides to the alfalfa weevil parasite, Microctonus
aethiopoides and the influence of parasitism on host
susceptibility. Environ. Entomol. 5: 311-315.

Dustan, A. G. 1923. Studies on a new species of Empusa
parasitic on the green apple bug ( Lygus communis var.
novascotiensis Knight) in the Annapolis Valley. Proc.
Acadian Entomol. Soc. 9:14.

143

Edwards, C.R. ed., 1979. Alfalfa, A guide to production
and integrated pest management in the midwest. North
Central Regional Extension Publication 113.

Fargues, J. 1984. Adhesion of the fungal spore on the
insect cuticle in relation to pathogenicity. /« D. W.

Roberts and J. R. Aist [eds.], Infection processes of fungi.
A Bellagio Conference, The Rockefeller Foundation.

Feng, Z, R.I. Carruthers and M. E. Ramos. 1990.
Laboratory and field dose-mortality responses of
Entomophaga grylli mycosis of the clearwinged grasshop¬
per, Camnula pellucida. Environ. Entomol. (in review).

Ferron, P. 1978. Biological control of insect pests by
entomogenous fungi. Ann. Rev. Entomol. 23: 409.

Ferron, P. 1981. Pest control by the fungi Beauveria and
Metarhizium. In H. D. Burges [ed.], Microbial control of
pests and plant diseases, 1970-1980. Academic Press,
London.

Fuxa, J.R. 1987. Ecological considerations for the use of
entomopathogens in IPM. Ann. Rev. Entomol. 32:225.

Fuxa, J. R., and Y. Tanada [eds.], 1987. Epizootiology of
Insect Diseases. John Wiley and Sons, New York.

Galaini-Wraight, S., S. P. Wraight, R. I. Carruthers, B. P.
Magalhaes and D. W. Roberts. 1990. Description of a
Zoophthora radicans epizootic in a population of Empoasca
krameri on beans in central Brazil. J. Invertebr. Pathol, (in
press).

Gardener, W. A. 1982. Occurrence of Erynia sp. in
Hyper a postica in central Georgia. J. Invertebr. Pathol.

40: 146.

Gillespie, A. T., and E. R. Moorhouse. 1989. The use of
fungi to control pest of agricultural and horticultural
importance. In Biotechnology of fungi for improvement of
plant growth, J. M. Whipps and R. D. Lumsden [eds.],
Cambridge Univ. Press, Cambridge.

Gyrisco, G, G., D. Landman, A. C. York, B. J. Irwin and
E. J. Armbrust. 1978. The literature of arthropods
associated with alfalfa. Part IV. A bibliography of the
potato leafhopper, Empoasca fabae. Agr. Exp. Sta. Spec
Publ. 51.

Hajek, A., T. M. Butt and L. I. Strelow. 1989. Develop¬
ment of an enzyme-linked immunosorbent assay for
detection of Entomophaga maimaiga in Gypsy Moth,
Lymantria dispar, larvae. Proceedings, New directions in
biological control, Frisco, CO (poster).

Hajek, A. E., R. A. Humber, J. S. Elkinton, B. P. May,

S. R. A. Walsh and J. C. Silver. 1990. Allozyme and
RFLP analysis confirms Entomophaga maimaiga respon¬
sible for 1989 epizootics in North American gypsy moth
populations. Proc. Nat Acad Sci (in press).

Hall, R. A. 1981. The fungus Verticillium lecanii as a
microbial insecticide against aphids and scales, In H. D.
Burges [ed.], Microbial control of pests and plant diseases,
1970-1980. Academic Press, London.

Hall, I. M., and P. H. Dunn. 1957. Fungi on spotted
alfalfa aphid. Calif. Agr. 11: 5.

Hamm, J. J., and W. W. Hare. 1982. Applications of
entomopathogens in irrigation water for control of fall
armyworms and com earworms. J. Econ. Entomol. 75:
1074.

Harcourt, D. G., J. C. Guppy, D. M. MacLeod and D.
Tyrrell. 1974. The fungus Entomophthora phytonomi
pathogenic to the alfalfa weevil, Hypera postica. Can.
Entomol. 106: 1295.

Harcourt, D. G., J. C. Gupp, D. M. MacLeod and D.
Tyrrell. 1981. Two Entomophthora species associated with
disease epizootics of the alfalfa weevil, Hypera postica in
Ontario. Great Lakes Entomol. 14: 55.

Harcourt, D. G., J. C. Guppy and M. R. Binns. 1984.
Analysis of numerical change in subeconomic populations
of the alfalfa weevil, Hypera postica, in eastern Ontario.
Environ. Entomol. 13: 1627.

Harper J. D. 1987. Applied Epizootiology: Microbial
Control of Insects. In Epizootiology of Insect Diseases, J.
R. Fuxa and Y. Tanada [eds.] Wiley, New York.

Hochberg, M. E. 1989. The potential role of pathogens in
biological control. Nature 337: 262-265.

144

Hostetter, D. L., and C. M. Ignoffo. 1978. Induced
epizootics: fungi. In G. E. Allen, C. M. Ignoffo and R. P.
Jacques [eds.], Microbial control of insect pests: Future
strategies in pest management systems. NSF-USDA-Univ
of Florida, Gainesville.

Hower, A. A., and G. A. Davis. 1984. Selectivity of
insecticides that kill the potato leafhopper and alfalfa weevil
and protect their parasite, Microctonus aethiopoides . J.
Econ. Entomol. 77: 1601-1607.

Huffaker, C. B. [ed.], 1971. Biological control. Plenum/
Rosetta, New York.

Hutchison, W.D., and D.B. Hogg. 1985. Time-specific
life tables for the pea aphid, Acyrthosiphon pisum (Harris)
on alfalfa. Researches on Population Ecology 27: 231-253.

Ignoffo, C. M. 1985. Manipulating enzootic-epizootic
diseases of arthropods. In M. A. Hoy and D. C. Herzog
[eds.], Biological control in agricultural IPM systems.
Academic Press, Orlando.

Ignoffo, C. M., D. L. Hostetter, P. P. Sikorowski, G.
Sutter and W. M. Brooks. 1977. Inactivation of represen¬
tative species of entomopathogenic viruses, a bacterium,
fungus, and protozoan by an ultraviolet light source.
Environ. Entomol. 6: 411.

Ignoffo, C. M., N. Marston, D. L. Hostetter, B. Puttier
and J. V. Bell. 1976. Natural and induced epizootics of
Nomuraea rileyi in soybean caterpillars. J. Invertebr.
Pathol. 27: 191.

Kenneth, R., G. Wallis, U. Gerson and H. N. Plant.

1972. Observation and experiments on Triplosporium
floridanum attacking spider mites in Israel. J. Invertebr.
Pathol. 19: 366.

King, E. G., and J. R. Coulson. 1988. ARS national
biological control program, proceedings of workshop on
research priorities. USDA-ARS, Beltsville.

Kish, L. P., and G. E. Allen. 1978. The biology and
ecology of Nomuraea rileyi and a program for predicting its
incidence on Anticarsia gemmatalis. Bull. Fla Agr. Exp.
Sta. No. 795.

Kramer, J. P. 1980. The House-fly mycosis caused by
Entomophthora muscae: Influence of relative humidity on
infectivity and conidial germination. NY Entomol. Soc.

88: 236.

Laird, M. 1967. A coral island experiment: a new
approach to mosquito control. World Health Org Chron
21:18.

Larkin, T. S., and R. I. Carruthers. 1990. Development
of a hierarchical simulation environment for research
biologists. In A. Guasch (ed), Object oriented simulation.
Soc. for Computer Simulation, San Diego.

Larkin, T. S., R. I. Carruthers, and R. S. Soper. 1988.
Simulation and object-oriented programming: the develop¬
ment of SERB. Simulation 51: 93-100.

Latge, J. P., J. Eilenberg, A. Beauvais and M. Prevost.
1988. Morphology of Entomophthora muscae protoplasts
grown in vitro. Protoplasma 146: 166.

Latteur, G. 1977. Sur la possibilite d' infection directe
d'aphides par Entomophthora a partir de sols hebergeant un
inoculum naturel. C.R. Acad. Ac. Paris 284: 2253-2256.

Lewis, L. C., and J. E. Cossentine. 1986. Season long
intraplant epizootics of entomopathogens, Beauveria
bassiana and Nosema pyrausta, in a com agroecosystem.
Entomophaga 3 1 : 363.

Logan, J. A. 1988. Toward an expert system for develop¬
ment of pest simulation models. Environ. Entomol. 17:
359-376.

MacLeod, D. M. 1963. Entomophthorales infections. In
Steinhaus E. A. [ed.], Insect pathology: An advanced
treatise, Vol. II. Academic Press, New York.

MacLeod, D. M., J. W. Cameron, R. S. Soper. 1966.

The influence of environmental conditions on epizootics
caused by entomogenous fungi. Revue Roumaine de
Biologie 11: 125.

McCoy, C. W. 1981. Pest control by the fungus
Hirsutella thompsonii. In H. D. Burges [ed.], Microbial
control of pests and plant diseases, 1970-1980. Academic
Press, London.

145

McCoy, C. W. 1990. Entomogenous fungi as microbial
pesticides. In T. Baker and P. Dunn [eds.], New Direc¬
tions in biological control: Alternatives for suppressing
Agricultural pests and diseases. A. R. Liss, Inc., New
York.

McCoy, C. W., R. F. Brooks, J. Allen and A.G. Selhime.
1976. Management of arthropod pests and plant diseases in
the citrus agroecosystem. Proc. Tall Timbers Conf. Ecol.
Anim. Control Habitat Manag. 6: 10.

McCoy C. W., R. A. Samson, D. G. Boucias. 1988.
Entomogenous fungi. In C. M. Ignoffo and N. B.

Mandava [eds.], Handbook of natural pesticides, Vol. V -
Microbial Insecticides, Part A - Entomogenous protozoa
and fungi. CRC Press, Boca Raton.

McGuire, M. R., J. V. Maddox and E. J. Armbrust.

1987a. Host range studies of an Erynia radicans strain
isolated from Empoasca fabae. J. Invertebr. Pathol. 50:
75-77.

McGuire, M. R., M. J. Morris, E. J. Armbrust and J. V.
Maddox. 1987b. An epizootic caused by Erynia radicans
in an Illinois Empoasca fabae population. J. Invertebr.
Pathol. 50: 78-80.

Mendgen, K. 1986. Quantitative serological estimations
of fungal colonization. In N. J. Fokkema and J. van den
Heuvel [eds.]. Microbiology of the Phyllosphere. Cam¬
bridge University Press, Cambridge.

Micales, J. A., M. R. Bonde and G. L. Peterson. 1986.
The use of isozyme analysis in fungal taxonomy and
genetics. Mycotaxon 27: 405.

Millstein, J. A., G. C. Brown and G. L. Nordin. 1983.
Microclimatic moisture and conidial production in Erynia
sp.: In vivo moisture balance and conidiation phenology.
Environ. Entomol. 12: 1339.

Milner, R. J. 1982. On the occurrence of pea aphids,
Acyrthosiphon pisum, resistant to isolates of the fungal
pathogen Erynia neoaphidis. Entomol. exp. appl. 32: 23.

Milner, R. J. 1985. Distribution in time and space of
resistance to the pathogenic fungus Erynia neoaphidis.
Entomol. exp. appl. 37:235.

Milner, R. J., and J. Bourne. 1983. Influence of tempera¬
ture and duration of leaf wetness on infection of
Acyrthosiphon kondoi with Erynia neoaphidis. Ann. Appl.
Biol. 102: 19-27.

Milner, R. J., and G. G. Lutton. 1983. Effect of tempera¬
ture on Zoophthora radicans (Brefeld) Batko: An intro¬
duced microbial control agent of the spotted alfalfa aphid
Iherioaphis trifolii (Monell) f. maculata. J. Aust.

Entomol. Soc. 22: 167.

Milner, R. J., and G. G. Lutton. 1986. Microbial control
of insects: Biological control of spotted alfalfa aphid.
CSIRO Biennial Report 1983-85. Canberra.

Milner, R. J., and R. S. Soper. 1981. Bioassay of
Entomophthora against the spotted alfalfa aphid
Therioaphis trifolii (Monell) f. maculata. J. Invertebr.
Pathol. 37: 168.

Milner, R. J., R. S. Soper and G. G. Lutton. 1982. Field
release of an Israeli strain of the fungus Zoophthora
radicans (Brefeld) Batko for biological control of
Therioaphis trifolii (Monell) f. maculata. J. Aust Entomol.
Soc. 21: 113.

Milner, R. J., R. E. Teakle, G. G. Lutton and F. M. Dare.
1980. Pathogens of the blue-green aphid Acyrthosiphon
kondoi Shinji and other aphids in Australia. Aust. J. Bot.
28: 601.

Missonnier, J., Y. Robert and G. Thoizon. 1970.
Circonstances epid^miologiques semblant favoriser le
dSveloppement des mycoses a entomophthorales chez trois
aphides, Aphis fabae, Capitophorus horni, et Myzus
persicae. Entomophaga 15: 169-190.

Mohamed, A. K. A., P.P. Sikorowski and J. V. Bell.

1977. Susceptibility of Heliothis zea larvae to Nomuraea
rileyi at various temperatures. J. Invertebr. Pathol. 30:

414.

Mullens, B. A., J. L. Rodriguez and J. A. Meyer. 1987.
An epizootiological study of Entomophthora muscae in
muscoid fly populations on southern California poultry
facilities, with emphasis on Musca domestica. Hilgardia
55: 1.

146

Newman, G. G., and G. R. Camer. 1975. Environmental
factors affecting conidial sporulation and germination of
Entomophthora gammae. Environ. Entomol. 4: 615.

Nordin, G. L., G. C. Brown and J. A. Millstein. 1983.
Epizootic phenology of Eryttia disease of the alfalfa weevil,
Hypera postica, in central Kentucky. Environ. Entomol.

12: 1350.

Onstad, D., and R. I. Carruthers. 1990. Epizootiological
models of insect diseases. Ann. Rev. Entomol. 35: 399-
419.

Papierok, B., and N. Wilding. 1979. Mise en evidence
d'une difference de sensibilite entre 2 clones du puceron du
Pois, Acyrthosiphon pisum, exposes a 2 souches du
champignon Phycomycete: Entomophthora obscura. CR
Acad Sci Paris, Ser D 288: 93.

Pickford, R., and P. W. Riegert. 1964. The fungus
disease caused by Entomophthora grylli Fres. , and its effect
on grasshopper populations in Saskatchewan in 1963. Can.
Entomol. 96: 1158.

Pienkowski, R. L., and P. R. Mehring. 1983. Influence
of Avermectrin B and carbofuran on feeding by alfalfa
weevil larvae. J. Econ. Entomol. 76: 1167.

Pinnock, D. E., and R. J. Brand. 1981. A Quantitative
approach to the ecology of the use of pathogens for insect
control. In H. D. Burges [ed.], Microbial Control of Pests
and Plant Diseases 1970-1980. Academic Press, London.

Puttier, B., D. L. Hostetter and S. H. Long. 1978.
Entomophthora phytonomi, a fungal pathogen of the alfalfa
weevil in the Mid-Great Plains. Environ. Entomol. 7: 670.

Puttier, B., D. L. Hostetter, S. H. Long and A. A. Borski.
1980. Seasonal incidence of the fungus Entomophthora
phytonomi infecting Hypera postica larvae in central
Missouri. J. Invertebr. Pathol. 35: 99.

Regniere, J. 1984. Vertical transmission of diseases and
population dynamics of insects with discrete generations: a
model. J. Theor. Biol. 107: 287-301.

Riba, G. 1984. Application en essais parcellaires de plein
champ d'un mutant artificial du champignan

entomopathogene Beauveria bassiana contre la pyrale du
mais, Ostrinia nubilalis. Entomophaga 29: 41.

Roberts, D. W. 1978. Introduction and colonization:
Fungi. In G. E. Allen, C. M. Ignoffo and R. P. Jacques
[eds.], Microbial control of insect pests: Future strategies a
in pest management systems. NSF-USDA-Univ Florida,
Gainesville.

Roberts D. W. 1981. Toxins of entomopathogenic fungi.
In H. D. Burges [ed.]. Microbial Control of Pests and Plant
Diseases 1979-1980. Academic Press, London.

Roberts, D. W., and A. S. Campbell. 1977. Stability of
entomopathogenic fungi. Misc. Publ. Entomol. Soc.

Amer. 10: 19.

Roberts, D. W., and R. A. Humber. 1981.

Entomogenous Fungi. In G. T. Cole, B. Kendrick [eds.],
Biology of conidial fungi. Academic Press, New York.

Rockwood, L. P. 1950. Entomogenous fungi of the
family Entomophthoraceae in the Pacific Northwest. J.
Econ. Entomol. 43: 704-707.

Shunazu, M. 1977. Factors affecting conidial germination
of Entomophthora delphacis Hori. Appl. Entomol. Zool
12: 260.

Shimazu, M. 1987. Protoplast fusion of insect pathogenic
fungi. In K. Maramorosch (ed), Biotechnology in inverte¬
brate pathology and cell culture. Academic Press, New
York.

Soper, R. S., and D. M. MacLeod. 1981. Descriptive
epizootiology of an aphid mycosis. USDA Tech. Bull. No
1632.

Soper, R. S., L. F. R. Smith and A. J. Delyzer. 1976.
Epizootiology of Massopora levispora in an isolated
population of Okanagana rimosa. Ann. Entomol. Soc.
Amer. 69: 275.

Speare, A. T., and R. H. Colley. 1912. The artificial use
of the brown-tail fungus in Massachusetts with practical
suggestions for private experiment, and a brief note on a
fungous disease of the gypsy caterpillar. Wright and Potter
Print Co., Boston.

147

Sprenkel, R. K., W. M. Brooks, J. W. Van Duyn and L.
Deitz. 1979. The effects of three cultural variables on the
incidence of Nomuraea rileyi, phytophagous Lepidoptera,
and their predators on soybeans. Environ. Entomol. 8:

334.

St Leger, R. J., R. M. Cooper and A. K. Chamley. 1987.
Production of cuticle-degrading enzymes by the
entomopathogen Metarhizium anisopliae during infection of
cuticles from Calliphora vomitoria and Manduca sexta. J.
GenMicr 133: 1371.

St Leger, R. J.,R. M. Cooper, A. K. Chamley. 1988a.

The effect of melanization of Manduca sexta cuticle on
growth and infection by Metarhizium anisopliae. J.
Invertebr. Pathol. 52: 459.

St Leger, R. J., P. K. Durrands, A. K. Chamley and R.

M. Cooper. 1988b. Role of extracellular chymoelastase in
the virulence of Metarhizium anisopliae for Manduca sexta.
J. Invertebr. Path 52: 285.

Tanada, Y. 1963. Epizootiology of infectious diseases, In
E.A. Steinhaus [ed.], Insect pathology: An advanced
treatise, Vol. II. Academic Press, New York.

Tanada, Y., and J. R. Fuxa. 1987. The Pathogen Popula¬
tion. In J. R. Fuxa and Y. Tanada [eds.], Epizootiology of
Insect Diseases. Wiley, New York.

Tyrrell, D., and D. M. MacLeod. 1972. Spontaneous
formation of protoplasts by a species of Entomophthora. J.
Invertebr. Pathol. 19: 354.

US Dept. Agric. 1956. The clover leaf-weevil and its
control. Fanner's Bull. 1484.

US Dept Agric. 1989. Agricultural Statistics. U.S.
G.P.O., Washington DC.

Watanahe, H. 1987. The Host Population. In J. R. Fuxa
and Y. Tanada [eds.], Epizootiology of Insect Diseases.
John Wiley, New York.

Wilding, N. 1969. Effect of humidity on the sporulation
of Entomophthora aphidis and Entomophthora thaxteriana.
Trans. British Mycol Soc. 53:126,

Wilding, N. 1973. The survival of Entomophthora sp. in
mummified aphids at different temperatures and humidities.
J. Invertebr. Pathol. 21: 309.

Wilding, N. 1975. Entomophthora species infecting pea
aphids. Trans. Roy. Entomol. Soc. Lond 127: 171.

Wilding, N. 1981. Pest control by Entomophthorales. In
Burges HD (ed). Microbial control of pests and plant
diseases, 1970-1980. Academic Press, London.

Wraight, S. P., S. Galaini-Wraight, D. W. Roberts and R.
I. Carruthers. 1985. Epizootiological investigations of
Erynia radicans in an Empoasca kraemeri population in
central Brazil. Proceedings of the Annual Meeting of the
Society of Invertebrate Pathology, Davis CA.

Wraight, S. P., S. Galaini-Wraight, R. I. Carruthers and
D. W. Roberts. 1986. Field transmission of Erynia
radicans to Empoasca fabae leafhoppers in alfalfa follow¬
ing application of a dry, mycelial preparation. In Funda¬
mental and Applied Aspects of Invertebrate Pathology, R.
A. Samson, J. M. Vlak, and D. Peters [eds.], Foundation
of the 4th International Colloquium of Invertebrate Pathol-
ogy, Wageningen.

Wraight, S. P., T. M. Butt, S. Galaini-Wraight, L. L.
Allee and D. W. Roberts. 1990. Germination and infec¬
tion processes of the entomophthoralean fungus Erynia
radicans on the potato leafhopper, Empoasca fabae. J.
Invertebr. Pathol. 56: 157-174.

Yendol, W. G. 1968. Factors affecting germination of
Entomophthora conidia. J. Invertebr. Pathol. 10: 116-121.

Yoder, O. C., K. Weltring, B. G. Turgeon, R. C. Garber
and H. D. VanEtten. 1987. Prospects for development of
molecular technology for fungal insect pathogens. In K.
Maramorosch [ed.], Biotechnology in Invertebrate Pathol¬
ogy and Cell Culture. Academic Press, New York.

Zacharuk, R. Y. 1973. Electron-microscope studies of the
histo-pathology of fungal infections by Metarhizium
anisopliae. Misc. Publ. Entomol. Soc. Amer. 9: 112.

Zadoks, J. C., and R. D. Schein. 1979. Epidemiology
and Plant Disease Management. Oxford University Press,
New York.

148

Use of Beauveria bassiana and
Bacillus thuringiensis for Suppression
of the European Corn Borer, Ostrinia
nubilalis* 1

Leslie C. Lewis2

Abstract

Bacillus thuringiensis and Beauveria bassiana, a microbial
insecticide, and an entomopathogenic fungus, respectively,
are efficacious alternatives to chemical insecticides to
suppress Ostrinia nubilalis. Past, present, and future of
these microbes are presented. Beauveria bassiana was first
used to suppress O. nubilalis on maize in 1934. Early
researchers determined the mode of action of Beauveria
bassiana and utilized this fungus for suppression of O.
nubilalis. Commercialization of this organism is in its
infancy. B. bassiana does, however, occur naturally and
produces epizootics. Studies initiated to conserve and
enhance B. bassiana for suppression of O. nubilalis led to
the discovery of a unique endophytic relationship between
B. bassiana and the maize plant; this relationship will most
likely be exploited in pest management programs. Bacillus
thuringiensis was utilized in control efforts against O.
nubilalis as early as 1928. The techniques utilized were
primitive, and researchers had little understanding how this
bacterium killed O. nubilalis. Conceptualized systematic
research has been conducted for many years to elucidate the
histopathology of this bacterium against O. nubilalis , to
understand the impact of environmental factors on B.
thuringiensis and to develop formulations to deliver it to
the maize plant to control this serious pest of maize. This
research has been the impetus for the commercialization of
B. thuringiensis. Future management programs will
definitely rely greatly on an increased use of these two
organisms by themselves, together, and in combination
with other entomotoxins.

Introduction

The European com borer, Ostrinia nubilalis, is a primary
pest of maize in many parts of the world. It damages maize
by feeding on the unfurled leaves in the whorl of the plant
during early vegetative stages (whorl stage), by feeding on
the leaf sheath tissue of maize during anthesis
(pollen-shedding), and, in some instances, by feeding on

'Joint contribution: USDA, ARS, and the Iowa Agriculture and Home
Economics Experiment Station (Journal Paper No. J-14195), Ames, Iowa.
Project No. 2513.

1 Com Insec ts Research Unit, USDA, ARS, Ankeny, Iowa

the ear. These types of feeding cause economic loss by
either direct or by physiological damage (reduction in the
nutrient transport systems).

Over the years, many approaches have been taken to
suppress damage to the maize plant by O. nubilalis. These
attempts include all recognized methods; i.e., mechanical,
chemical, biological, host plant resistance, and an integra¬
tion of all methods. In this paper, I will address only a
narrow segment within the biological control component,
specifically the use of Bacillus thuringiensis and Beauveria
bassiana to suppress O. nubilalis.

Bacillus thuringiensis

Bacillus thuringiensis is an aerobic, gram-positive, spore
and crystal-forming bacterium with 34 known subspecies
categorized into 27 serotypes (de Baijac and Frachon
1990). Many entomotoxins are produced by different
subspecies of B. thuringiensis ; however, delta-endotoxin
and beta-exotoxin are toxic to the widest range of insects.

The delta-endotoxin is a crystalline protein produced during
sporulation of the bacterium and was first noted by Berliner
(1915). The role of the crystal in the entomopathogenicity
of this bacterium towards insects was suggested by Hannay
(1953) and confirmed by Angus (1954, 1956). Lepidopter-
ous insects, including O. nubilalis, vary in their response to
the toxin. Growth of O. nubilalis larvae is slowed when
spores or crystals are fed to the insects (Sutter and Raun
1966), and some mortality occurs when early instars are fed
only crystals (Mohd-Salleh and Lewis 1981). The latter
researchers, however, reported a substantial increase in
mortality of early instars with the addition of spores and
found that both spores and crystals were necessary for
maximum kill of later instars. Bacillus thuringiensis kills
O. nubilalis by causing a sloughing of the midgut epithelial
cells into the gut lumen, exposing the basement membrane
to attack by vegetative rods from germinating spores (Sutter
and Raun 1967).

The beta exotoxin is a thermostable nucleotide produced at
the time of sporulation (Cantwell et al. 1964). It is toxic to
several species of insects (de Baijac 1960) including O.
nubilalis (Mohd-Salleh and Lewis 1982, 1983). It seem-

149

ingly kills insects by inhibiting enzymes involved with
nucleotide substrates such as nucleotidases and DNA-
dependent RNA polymerases (Sebasta and Horska 1968).

Bacillus thuringiensis is rarely isolated from insects other
than in an insectary environment. An extensive survey of
microorganisms associated with O. nubilalis was made in
Iowa; however, B. thuringiensis was never isolated (Raun
et al. 1959, Raun and Brooks 1963). In a more recent
study, Bacillus thuringiensis subspecies kurstaki was
isolated from field-collected egg masses and young larvae
of O. nubilalis (Lynch et al. 1976a). The bacterium was
isolated three times from egg masses and two times from
larvae in a total of 221 specimens. These five isolates were
pathogenic when tested against laboratory-reared O.
nubilalis and represents the first report of B. thuringiensis
isolation from an insect egg (Lynch et al. 1976b).

The use of B. thuringiensis to suppress populations of O.
nubilalis has a long history. Unsuccessful attempts were
made to isolate a bacterium from O. nubilalis that could be
cultured and applied to suppress this pest of maize in
Hungary, but researchers did successfully kill O. nubilalis
larvae with an isolate of B. thuringiensis isolated from the
mediterranean flour moth, Anagasta kuhniella (Husz 1928;
Chorine 1929). In other early work in Europe, an isolate
of B. thuringiensis from Galleria mellonella was applied to
maize as a spore powder. This preparation killed 90% of
the larvae (Metalnikov and Chorine 1929a,b; Metalnikov et
al. 1930a, b). In other work, B. thuringiensis was applied
as a dust or spray, to maize in field experiments suppress¬
ing O. nubilalis larvae by 96.8 to 99.2%. This work was
continued with spray and dust formulations that suppressed
O. nubilalis up to 2 weeks after application of the bacte¬
rium (Chorine 1930). Multiple applications of B.
thuringiensis 7 days apart also were very effective in
reducing plant damage caused by O. nubilalis (Husz 1930,
1931).

The aforementioned research with dusts and sprays to apply
B. thuringiensis involved the unsophisticated combination
of the bacterium and a carrier, usually water or talcum
powder. The event of commercial production of B.
thuringiensis triggered widespread efficacy evaluation of
this bacterium against many species of lepidopterans. This,
in turn, ushered in the era of formulation research with B.
thuringiensis. Even so, for many years, researchers merely

adopted techniques from the chemical industry to formulate
B. thuringiensis. These formulations were used in efficacy
experiments against O. nubilalis in many countries.

In France, granular formulations of B. thuringiensis were
superior to sprays and equal in some tests to DDT in
suppressing damage by O. nubilalis (Cangardel and
Anglade 1969; Martouret and Anglade 1971). The
granular formations, BactucideG, BactospeineG, and
EcatoxG all effectively controlled O. nubilalis up to 30
days after application in experiments in Hungary (Injac
1979). An extensive comparison between a B.
thuringiensis spray, Bactucide P, a granule, Bactucide G,
and the chemical Diazinon to suppress O. nubilalis were
made at many locations in Italy. Granules were always
superior to sprays, but neither was as efficacious as
Diazinon (Coppolino et al. 1984). In the U.S., attapulgite
clay granules, with B. thuringiensis adequately suppressed
O. nubilalis in whorl-stage maize but did not in pollen-
shedding maize (Raun 1963). Bacillus thuringiensis
subspecies thuringiensis and kurstaki were applied as spray
and granule formulations. Granules were superior to spray
formulations and equal to DDT and Diazinon in efficacy
against O. nubilalis larvae (McWhorter et al. 1972). In
only one study has a spray formulation been superior to a
granular formulation. That was in the Federal Republic of
Germany, where a spray formulation of Dipel suppressed
O. nubilalis by 70%, whereas there was no difference in
suppression between the granular plot and the untreated
check (Langenbruch 1976).

The previously discussed research was conducted by
applying B. thuringiensis with over-the-row application
equipment. In the west central U.S. com belt, a large area
of maize is grown under irrigation. Researchers attempted
to suppress O. nubilalis larvae in this area by applying B.
thuringiensis through center-pivot sprinkler irrigation
systems. It was concluded that, even though larval densi¬
ties were reduced, the magnitude of reduction was too small
to recommend B. thuringiensis as an economically sound
substitute for chemical insecticides (Nolting and Poston
1982).

Occasionally, B. thuringiensis is applied to plants other
than field maize to suppress larval populations of O.
nubilalis. When applied to sorghum foliage more than
50% of the original insecticidal activity was lost 24 h after
application, making it ineffective (Gardner and Hornby
1987). Likewise, in Israel, Bactospiene III sprayed on

150

sweet com leaves lost activity very readily, with no activity
remaining 8 days after application (Navon and Melamed-
Madjar 1986). In other experiments (Hudon 1962, 1963),
a wettable powder of B. thuringiensis was applied to sweet
com to control O. nubilalis. Efficacy was unsatisfactory
compared with recommended chemical insecticides.

It is obvious from many studies that B. thuringiensis is
efficacious against O. nubilalis larvae, but results within
and between experiments are highly variable. Research has
been conducted to better understand this variability and
make B. thuringiensis more competitive with chemical
products. Also, it was obvious that the placement of B.
thuringiensis on the plant relative to feeding sites of the
larvae and protection of B. thuringiensis from the environ¬
ment were extremely important. Additional work to
control maize insects demonstrated the susceptibility of B.
thuringiensis to degradation from exposure to ultraviolet
light (Raun et al. 1966).

The initial work to produce a formulation that would
protect B. thuringiensis from ultraviolet light involved an
encapsulation process (Raun and Jackson 1966). This
technique produced a "clump" of bacterial powder sur¬
rounded by a capsular wall. The powder was then prepared
as a spray, a granule, or an emulsifiable concentrate and
applied to maize. These formulations were effective in
suppressing larval populations of O. nubilalis ; reduction
ranged from 45 % with sprays to 90 % with granules
compared with untreated checks. DDT granules (5 %)
reduced larval populations by 90% also. More recently,
granule volume and concentration of pathogen in the
granules were evaluated. Clay granules differing in
thuricide HPC concentration (5.6, 11.3, and 22.5% wt ./
wt.) were applied to maize during the whorl- and pollen-
shedding stages of plant development. Granule volume per
unit area did not significantly influence the degree of
control, but the 22.5% concentration was significantly
more effective than the 11.3% concentration (Lynch et al.
1977b). In a continued effort to understand the lack of
consistency with which O. nubilalis larvae can be sup¬
pressed, two granular formulated products, Dipel and
Thuricide HPC, were applied to maize on an equal basis;
i.e., equal number of international units. This approach
reduced variability between commercial products (Lynch et
al. 1977a).

Efficacy of B. thuringiensis is routinely measured by
counting the number of larvae remaining in a plant or by

measuring physical damage to the maize plant caused by O.
nubilalis. In an attempt to delineate additional factors other
than the environment that produce variability in efficacy of
B. thuringiensis, products research was conducted to
quantify the amount of B. thuringiensis deposited in the
target area of the plant and the persistence of this material
when applied with granular, foam, or spray formulations.

These researchers (Lynch et al. 1980) found that granular
and foam formulations applied to maize were more effective
than a spray formulation. Also the granular and foam
formulations were superior to the target-area deposits.
Bacillus thuringiensis applied with these formulations
persisted longer than that delivered by a spray formulation.
The latter work again indicated that, when B. thuringiensis
is placed where O. nubilalis feeds in the plant (i.e., within
the whorl or behind the leaf-sheath collar) its efficacy is
greatly increased. This was instrumental in initiating
research to develop a granule formulation that would both
deliver B. thuringiensis to the target area and protect it
from ultraviolet (UV) light.

A simple and potentially economically feasible
macroencapsulation system involving the use of starch to
encapsulate B. thuringiensis was developed (Dunkle and
Shasha 1988). This unique system allows for the easy
addition of phagostimulants, feeding attractants, and UV
protectants such as congo red. The addition of congo red
greatly extended the life of B. thuringiensis. More than
50% of the original spore activity remained 12 d after
exposure to unobstructed sunlight in a plastic arena (Dunkle
and Shasha 1989). Granules made with the starch-matrix
process have been evaluated for O. nubilalis suppression in
the field. The UV screen, congo red, and a feeding
stimulant. Coax, were used as additives. Congo red did not
increase efficacy of B. thuringiensis ; however, when Coax
was added to the smallest dosage of B. thuringiensis, larval
suppression was equal to that from the treatment that
contained four times as much B. thuringiensis (McGuire et
al., in press). Research with starch encapsulation is
continuing and will most likely add greatly to the effort to
make B. thuringiensis an economical as well as efficacious
alternative to chemical insecticides.

The future of B. thuringiensis for suppression of O.
nubilalis is very promising. The public is demanding and
many producers, especially hybrid maize seed producers,
are looking for alternatives to chemical insecticides. The
starch matrix granule has shown great promise as a superior

151

delivery system for this bacterium. Furthermore, genetic
engineering has been utilized to develop new B.
thuringiensis products specifically for suppressing 0.
nubilalis. Genes for toxin production in B. thuringiensis
are carried on plasmids. These plasmids can be taken from
B. thuringiensis isolates and inserted into another (Carlton
1988). This technique has been used to develop Foil, a B.
thuringiensis product with both coleopteran and lepidopt-
eran activity to control O. nubilalis and Leptinotarsa
decemlineata on potatoes. MCap was developed by using
recombinant DNA techniques in which the gene that
produces B. thuringiensis toxin was inserted into Pseudo¬
monas fluorescens , also a naturally occurring bacterium.
The P. fluorescens carrying the B. thuringiensis toxin gene
are then killed, forming a biocapsule around the protein
crystal protecting it from harsh environmental factors.

Little experimentation has been reported in which efforts
have been made to develop transgenic plants specifically for
use against O. nubilalis. The possibility for success is
unlimited. Investigations are under way to introduce the
toxin of B. thuringiensis into maize by incorporating the
gene controlling toxin production into an endophytic
bacterium (Clavibacter xyli subsp. cynodontis). This
transgenic bacterial endophyte (InCide) is forced into the
maize seed, resulting in a plant containing B. thuringiensis
toxin. Presently, toxin production within the plant is
insufficient to kill insects. The development of transgenic
plants or the use of transgenic endophytes does, however,
create a dilemma. The problem is the development in O.
nubilalis of resistance to B. thuringiensis. Resistance was
not considered a possibility until the discovery of resistance
to B. thuringiensis in Plodia interpunctella (McGaughey
1985). We, as crop protection specialists, must be cogni¬
zant of this possibility and develop methodologies to
manage resistance and, in turn, maximize the products of
maize research.

Beau vena bassiana

Beauveria bassiana is an ubiquitous entomopathogenic
fungus efficacious to many species of insects, including O.
nubilalis. Beauveria bassiana is readily isolated from the
soil and plant debris of a maize agroecosystem, and it can
cause an epizootic in O. nubilalis on a yearly basis (Lewis
1985). Beauveria bassiana kills insects when a conidium
attaches to larval epicuticle, germinates, penetrates the
cuticle, develops hyphal bodies, and produces toxic

metabolites (Lefebvre 1934; Roberts 1981). The specific
histopathology of B. bassiana invasion of O. nubilalis has
been described (Lefebvre 1934).

Earlier work with B. bassiana for maize plant protection
was similar to that with B. thuringiensis in technique.
Conidia were combined with cornstarch or wheat flour and
dusted on plants. Larval mortality of 1 to 79 % was
obtained (Bartlett and Lefebvre 1934). Other workers
applied conidia in a flour carrier at 1 1-d intervals to pollen-
shedding maize and reduced the O. nubilalis by 33 to 63%
(Stirret et al. 1937). These researchers reported the
importance of placement of B. bassiana on the plant,
timing of application relative to insect infestation, and
microhabitat. They also reported that plant damage
occurred before the larvae were killed; e.g., there was a 60-
70% reduction in larval population but little reduction in
plant damage (Beall et al. 1939).

Little additional data were reported on the use of B.
bassiana for O. nubilalis suppression for 20 years. In a
renewed effort, sprays, dusts, and granulars were utilized to
deliver B. bassiana to whorl-stage maize. More than over
90% of the larvae were killed; however, these formulations
of B. bassiana were ineffective against O. nubilalis on
mature maize (York 1958). Over 95% reduction of O.
nubilalis larvae has been reported in the People's Republic
of China by applying 1.5 x 108 conidia per plant (Hsiu et
al. 1973), and larval populations were reduced by 91. 1 % in
France with 1. x 1012 conidia per ha (Riba 1984). Dosage
of conidia, isolate of the fungus, and stage of O. nubilalis
larvae are very important when attempting to maximize the
response to B. bassiana. First instars are reported to be
most susceptible (Feng et al. 1985). Also fourth instars are
the least susceptible. Isolates of B. bassiana, however,
differed greatly in pathogenicity (Feng et al. 1985).

In efforts to better understand the high variability and many
times low efficacy of B. bassiana against O. nubilalis,
recent research has defined the microhabitat necessary for
this fungus to express its pathogenic properties. Tempera¬
ture is the dominant factor affecting in vivo mycoses in O.
nubilalis (Carruthers et al. 1985). Relative humidity (RH)
is very important, not only for survival of the fungus, but
also for invasion. Conidia of B. bassiana can invade the
cuticle of O. nubilalis at 30% or 70% RH; however, the
lethal time was increased at the lower RH (Riba and
Marcandier 1984). Beauveria bassiana is readily found in
nature in the soil, on decaying organic material, and on

152

insect cadavers and insect eggs (Raun et al. 1959; Brooks
and Raun 1965; Lynch and Lewis 1978; McCoy et al.

1988). This fungus is readily isolated from field-collected
larvae. Ostrinia nubilalis collected for laboratory coloniza¬
tion must be surface-sterilized B. bassiana conidia (Lewis,
unpubl. results). Diapausing O. nubilalis larvae have been
collected from several locations in France and later died
from B. bassiana infection within 60 d (Marcandier and
Riba 1986).

Research has been conducted recently to better understand
the temporal relationships between B. bassiana and O.
nubilalis within a maize ecosystem. Researchers developed
a model predicting the timing of larval mortality following
exposure to B. bassiana (Feng et al. 1988). When B.
bassiana conidia and O. nubilalis larvae were placed on the
plant simultaneously, more than 60% of the larvae were
killed. When conidial application preceded larval applica¬
tion, mortality decreased linearly as the time between
applications increased. These researchers concluded that
maximum mortality occurs when O. nubilalis larvae and
conidia come in contact immediately. Anything delaying
this meeting decreases mortality. Possible explanations are
that larvae move to inner areas of the plant and do not
readily contact the fungus or that the fungus precedes
arrival of the larvae, and viability of the conidia is lost.

In a different, but related, temporal experiment, larvae and
conidia were applied to whorl-stage maize, and 3 d later, B.
bassiana conidia were applied. Mortality in excess of 70%
was recorded (Lewis and Cossentine 1986). This experi¬
mentation was continued by infesting these plants in the
pollen-shedding stage with additional O. nubilalis larvae.
Tunneling by the latter larvae was significantly reduced by
B. bassiana placed on the plant 4 weeks earlier. Research¬
ers hypothesized that conidia were produced from insect
cadavers from the first larval infestation or that conidia
from the original inoculum remained viable on the plant.
Recent work has suggested another explanation for this
extended conidial viability. Beauveria bassiana was
applied to whorl-stage maize by using a granular formula¬
tion (Lewis and Bing, in press). This inoculum caused a
significant reduction of larvae during this stage and also
significantly reduced larvae from the filial generation of O.
nubilalis. This reduction in larval damage was significantly
correlated with the incidence of B. bassiana in the pith of
the plant.

Beauveria bassiana is known to colonize certain genotypes
of maize (Vakili 1990). It has been isolated from the pith
of several inbred lines. It is now evident that the season-
long suppression of O. nubilalis larvae reported by Lewis
and Cossentine (1986) and Lewis and Bing (submitted) was
probably due to colonization of the maize plant by B.
bassiana. A hybrid maize was utilized in both of these
studies; thus, the phenomenon is not unique to inbred maize
lines.

Evidence is very strong that an endophytic relationship
exists between B. bassiana and the maize. Subsequent
research confirms this and provides additional insight into
this phenomenon (Bing 1990). Beauveria bassiana was
applied to whorl-stage maize plants by a granular formula¬
tion or injected into the plant above the growing tip by
using a medical syringe and hypodermic needle. Pith
samples were removed from plants at senescence and
assayed for B. bassiana. The fungus persisted in the pith
of the plants and significantly suppressed O. nubilalis
larvae throughout the season. In additional studies, B.
bassiana was again applied by granular formulation or
injection to whorl-stage maize. O. nubilalis larvae were
then placed on the plants at whorl stage, late-whorl stage,
and pretassel stage of plant development. At senescence,
suppression of O. nubilalis was determined by measuring
larval tunneling. Larvae applied at the two earlier stages of
plant development caused significantly more tunneling than
those applied at pretassel stage of development. Most
likely, this reduction was due to B. bassiana colonizing
more of the plant by the late developmental stage. In a
final study in which plants were injected with B. bassiana
at anthesis, it was determined that the inoculum moved
upward in the plant perhaps with plant photosynthesis.
Furthermore the presence or absence of O. nubilalis larvae
did not influence the endophytic relationship between the
fungus and the plant.

Combinations of Entomotoxins

Reports of the simultaneous use of more than one microbe
to suppress populations of O. nubilalis are infrequent.
Theoretically, two pathogenic organisms would stress the
insect and increase the efficacy of each. Extensive research
was conducted in which O. nubilalis larvae were placed on
whorl-stage and pollen-shedding-stage plants. Subse¬
quently, these plants were sprayed with a suspension of
Nosema pyrausta, a mjcrosporidium. Then a granular

153

formulation of B. thuringiensis was applied to the same
plants. B. thuringiensis and N. pyrausta were additive in
terms of reducing larval populations (Lublinkhof et al.
1979). In a similar study in which the larvae were
transovarially infected with N. pyrausta, B. thuringiensis,
and N. pyrausta acted independently in suppressing the
larvae (Lewis et al. 1982). There were, however, several
interactions indicating a more severe impact from N.
pyrausta in a transovarial infection than that from a per os
infection.

In additional research using a split-plot experimental
design, (Lewis, unpublished) B. thuringiensis and Furadan,
a carbamate insecticide, were combined with B. bassiana to
suppress O. nubilalis larvae. B. bassiana alone and in
combination with the microbial and chemical insecticides
significantly reduced damage by O. nubilalis larvae. There
were no significant interactions between whole-plot and
split-plot treatments. In a study to provide immediate and
season-long suppression of O. nubilalis, B. thuringiensis
and B. bassiana were formulated on granules and applied
separately and in combination to field maize. Each patho¬
gen and combination of pathogens caused a significant
reduction in tunnelling compared with the check.

Beauveria thuringiensis and B. bassiana were independent
of each other in their impact on insect suppression (Lewis
and Bing, submitted).

The impact of two indigenous insect pathogens, B.
bassiana and N. pyrausta, in suppressing O. nubilalis was
investigated. B. bassiana alone reduced larval damage by
63 % compared with that of the check; N. pyrausta alone
reduced larval damage by 68 % , and the combination of
pathogens reduced plant damage by 78% (Lewis, unpub¬
lished data).

In general the efficacy of Bacillus thuringiensis and
Beauveria bassiana against O. nubilalis is not compromised
when either is applied to maize in combination with another
toxicant. The response is usually independent and often
additive. Future research should build on these data in
developing ecologically sound management systems for O.
nubilalis.

Acknowledgment

The author acknowledges Lori Anderson Bing and John J.
Obrycki for their critical review of this manuscript.

References Cited

Angus, T. A. 1954. A bacterial toxin paralyzing silkworm
larvae. Nature (Lond.) 173: 545-546.

Angus, T. A. 1956. Association of toxicity with protein-
crystalline inclusions of Bacillus soto Ishiwata. Can. J.
Microbiol. 2: 122-131.

Bartlett, K., and C. Lefebvre. 1934. Field experiments
with Beauveria bassiana (Bals.) Vuill. a fungus attacking
the European com borer. J. Econ. Entomol. 27: 1147-
1157.

Beall, G., G. M. Stirrett, and I. L. Conners. 1939. A
field experiment on the control of the European com borer,
Pyrausta nubilalis Hubner, by Beauveria bassiana Vuill.

II. Sci. Agric. 19: 531-534.

Berliner, E. 1915. Uber die Schlaffsucht der
Mehlmattenraupe ( Ephestia kuhniella Zell.) und ihren
Erreger Bacillus thuringiensis n. sp. Z. Angew. Entomol.

2: 29-56.

Bing, L. A. 1990. The entomopathogen, Beauveria
bassiana (Balsamo) Vuillemin: Colonization and move¬
ment in Zea mays L. and potential for Ostrinia nubilalis
(Hubner) suppression. M.S. thesis. 144 pp. Parks
Library. Iowa State Univ. Ames.

Brooks, D. L., and E. S. Raun. 1965. Entomogenous
fungi from com insects in Iowa. J. Invertebr. Pathol. 7:
79-81.

Burgeijon, A., and H. de Baijac. 1960. Nouvelles
donnees sur le role de la toxine soluble thermostable
produite par Bacillus thuringiensis Berliner. C.R. Acad.

Sci. Paris. 251: 911-912.

Cangardel, H., and P. Anglade. 1969. Recherches sur
l'activite' insecticide de Bacillus thuringiensis contre la
pyrale du mais ( Ostrinia nubilalis (Hubner)). I. Micro-
essais sous infestation artificielle. Rev. Zool. Agric. 68:
113-122.

Cantwell, G. E., A. M. Heimpel, and M. J. Thompson.
1964. The production of an exotoxin by various crystal
forming bacteria related to Bacillus thuringiensis var.
thuringiensis Berliner. J. Invertebr. Pathol. 6: 466-480.

154

Carlton, B. C. 1988. Development of genetically im¬
proved strains of Bacillus thuringiensis - a biological
insecticide. Amer. Chem. Soc. Symp. Ser. pp. 260-277 .
Snowbird, Utah.

Camithers, R. I., Z. Feng, D. S. Robson et al. 1985. In
vivo temperature-dependent development of Beauveria
bassiana (Deuteromycotina: Hyphomycetes) mycosis of the
European com borer, Ostrinia nubilalis (Lepidoptera:
Pyralidae). J. Invertebr. Pathol. 46: 305-311.

Chorine, V. 1929. New bacteria pathogenic to the larvae
of Pyrausta nubilalis Hb. Int. Com Borer Invest. 29: 39-
53.

Chorine, V. 1930. On the use of bacteria in the fight
against the com borer. Int. Com Borer Invest. 3: 94-98.

Coppolino, F., B. Giuliano, and G. Fontana. 1984. An
experiment on the biological control of Ostrinia nubilalis
Hb. by Bacillus thuringiensis Berliner. Estratto da Redia
67: 235-246.

de Baijac, H., and E. Frachon. 1990. Classification of
Bacillus thuringiensis strains. Entomophaga 35: 233-240.

Dunkle, R. L., and B. S. Shasha. 1988. Starch-encapsu¬
lated Bacillus thuringiensis: A potential new method for
increasing environmental stability of entomopathogens.
Environ. Entomol. 17: 120-126.

Dunkle, R. L., and B. S. Shasha. 1989. Response of
starch-encapsulated Bacillus thuringiensis containing
ultraviolet screens to sunlight. Environ. Entomol. 18:
1035-1041.

Feng, Z., R. I. Carruthers, and T. S. Larkin. 1988. A
phenology model and field evaluation of Beauveria
bassiana (Bals.) Vuillemin (Deuteromycotina:
Hyphomycetes) mycosis of the European com borer,
Ostrinia nubilalis (Hbn.) (Lepidoptera: Pyralidae). Can.
Entomol. 120: 133-144.

Feng, Z., R. I. Carruthers, D. W. Roberts et al. 1985.
Age-specific dose-mortality effects of Beauveria bassiana
(Deuteromycotina: Hyphomycetes) on the European com
borer, Ostrinia nubilalis (Lepidoptera: Pyralidae). J.
Invertebr. Pathol. 46: 259-264.

Gardner, W. A., and J. A. Homby. 1987. Bioassay of the
persistence of Bacillus thuringiensis on sorghum panicles.

J. Entomol. Sci. 22: 210-211.

Hannay, C. L. 1953. Crystalline inclusions in aerobic
sporeforming bacteria. Nature (Lond.) 172: 1004.

Hsiu, C. F., Y. Chang, C. M. Kwei, Y. M. Han et al.
1973. Field application with Beauveria bassiana (Bals.)
Vuill. for European com borer control. Acta Entomol.

Sin. 16: 203-206.

Hudon, M. 1962. Field experiments with Bacillus
thuringiensis and chemical insecticides for the control of
the European com borer on sweet com in southwestern
Quebec. J. Econ. Entomol. 55: 115-117.

Hudon, M. 1963. Further field experiments on the use of
Bacillus thuringiensis and chemical insecticides for the
control of the European com borer Ostrinia nubilalis on
sweet com in south-western Quebec. J. Econ. Entomol.

56: 804-808.

Husz, B. 1928. Bacillus thuringiensis Berl., a bacterium
pathogenic to com borer larvae. Int. Com Borer Invest. 1:
190-193.

Husz, B. 1930. Field experiments on the application of
Bacillus thuringiensis against the com borer. Int. Com
Borer Invest. 3: 91-93.

Husz, B. 1931. Experiments during 1931 on the use of
Bacillus thuringiensis Berliner in controlling the com
borer. Int. Com Borer Invest. 4: 22-23.

Injac, M. 1979. Results of the investigation of efficiency
of Bacillus thuringiensis Berliner, of the granulates
Bactuicide G., and Bactospein G in control of the com
borer ( Ostrinia nubilalis Hbn.). Zast. Bilja 30: 105-112.

Langenbruch, G. A. 1976. Maiszunslerbekampfung mit
Bacillus thuringiensis. Nachrichtenbl. Dtsch.
Pflanzenschutzdienstas. (Braunschw.) 28: 148-155.

Lefebvre, C. L. 1934. Penetration and development of the
fungus Beauveria bassiana, in the tissues of the com borer.
Ann. Bot. 48: 441-452.

155

Lewis, L. C. 1985. Epizootiology of entomopathogens in
an Iowa agroecosystem. Abstracts, Society for Invertebrate
Pathology, XVIII Annual Meeting, Sault Saint Marie,
Ontario, Canada, August 4-8.

Lewis, L. C. , and L. A. Bing. Bacillus thuringiensis and
Beauveria bassiana for European com borer control:
Program for immediate and season-long suppression. Can.
Entomol. submitted.

Lewis, L. C., and J. E. Cossentine. 1986. Season long
intraplant epizootics of entomopathogens, Beauveria
bassiana and Nosema pyrausta in a com agroecosystem.
Entomophaga 31: 363-369.

Lewis, L. C., J. Lublinkhof, E. C. Berry et al. 1982.
Response of Ostrinia nubilalis [Lepidoptera: Pyralidae]
infected with Nosema pyrausta [Microsporida:

Nosematidae] to insecticides. Entomophaga 27: 211-218.

Lublinkhof, J., L. C. Lewis, and E. C. Berry. 1979.
Effectiveness of integrating insecticides with Nosema
pyrausta for suppressing populations of the European com
borer. J. Econ. Entomol. 72: 880-883.

Lynch, R. E., and L. C. Lewis. 1978. Fungi associated
with eggs and first-instar larvae of the European com
borer. J. Invertebr. Pathol. 32: 6-11.

Lynch, R. E., L. C. Lewis, and E. C. Berry. 1980.
Application efficacy and field persistence of Bacillus
thuringiensis when applied to com for European com borer
control. J. Econ. Entomol. 73: 4-7.

Lynch, R. E., L. C. Lewis, E. C. Berry et al. 1977a.
European com borer control with Bacillus thuringiensis
standardized as com borer international units. J. Invertebr.

Pathol. 30: 169-174.

Lynch, R. E., L. C. Lewis, E. C. Berry et al. 1977b.
European com borer: Granular formulations of Bacillus
thuringiensis for control. J. Econ. Entomol. 70: 389-391.

Lynch, R. E., L. C. Lewis, and T. A. Brindley. 1976a.
Bacteria associated with eggs and first-instar larvae of the
European com borer: Identification and frequency of
occurrence. J. Invertebr. Pathol. 27: 229-237.

Lynch, R. E., L. C. Lewis, and T. A. Brindley. 1976b.
Bacteria associated with eggs and first-instar larvae of the
European com borer: Isolation techniques and pathogenic¬
ity. J. Invertebr. Pathol. 27: 325-331.

Marcandier, S., and G. Riba. 1986. Endemisme de la
mycose a Beauveria bassiana (Bals.) Vuillemin dans les
populations geographiques de la pyrale du mais, Ostrinia
nubilalis

(Hubner). Acta Ecol. (Ecol. Appl.) 7: 39-46.

Martouret, D. and P. Anglade. 1971. Bacillus
thuringiensis dans la lutte contre la pyrale du mais, Ostrinia
nubilalis Hubn. (Lepidopteres, Pyralides) Ann. Zool. Ecol.
Anim. 3: 57-68.

McCoy, C. W., R. A. Samson, and D. G. Boucias. 1988.
Entomogenous Fungi. In C.M. Ignoffo and N.B. Madava
[eds.], Handbook of Natural Pesticides, Volume 5, pp.
151-236. CRC Press, Inc., Boca Raton, Florida.

McGaughey, W. H. 1985. Insect resistance to the biologi¬
cal insecticide Bacillus thuringiensis. Science 229: 193-
195.

McGuire, M. R., B. S. Shasha, L. C. Lewis et al. 1990.
Field evaluation of granular starch formulations of Bacillus
thuringiensis against Ostrinia nubilalis (Lepidoptera:
Pyralidae). J. Econ. Entomol. (in press).

McWhorter, G. M., E. C. Berry, and L. C. Lewis. 1972.
Control of the European com borer with two varieties of
Bacillus thuringiensis. J. Econ. Entomol. 65: 1414-1417.

Metalnikov, S. and V. Chorine. 1929a. Experiments on
the use of bacteria to destroy the com borer. Int. Com
Borer Invest. 29: 54-59.

Metalnikov, S., and V. Chorine. 1929b. On the natural
and acquired immunity of Pyrausta nubilalis HB. Int. Com
Borer Invest. 29: 22-38.

Metalnikov, S., J. Ermolev, and V. Skobaltzyn. 1930a.
New bacteria pathogenic to the larvae of Pyrausta nubilalis
HB. Int. Com Borer Invest. 30: 28-36.

Metalnikov, S., B. Hergula, and D. M. Strail. 1930b.
Experiments on the application of bacteria against the com
borer. Int. Com Borer Invest. 30: 148-151.

156

Mohd-Salleh, M. B., and L. C. Lewis. 1981. Toxic
effects of spore/crystal ratios of Bacillus thuringiensis on
European com borer larvae. J. Invertebr. Pathol. 39: 290-
297.

Mohd-Salleh, M. B., and L. C. Lewis. 1982. Feeding
deterrent response of com insects to B-exotoxin of Bacillus
thuringiensis. J. Invertebr. Pathol. 39: 323-328.

Mohd-Salleh, M. B. and L. C. Lewis. 1983. Comparative
effects of spore-crystal complexes and thermostable
entomotoxins of six subspecies of Bacillus thuringiensis on
Ostrinia nubilalis (Lepidoptera: Pyralidae). J. Invertebr.
Pathol. 41: 336-340.

Navon, A., and V. Melamed-Madjar. 1986. Screening of
Bacillus thuringiensis preparations for microbial control of
Ostrinia nubilalis in sweet com. Phytoparasitica 14: 111-
117.

Nolting, S. P., and F. L. Poston. 1982. Application of
Bacillus thuringiensis through center-pivot irrigation
systems for control of the southwestern com borer and
European com borer (Lepidoptera: Pyralidae). J. Econ.
Entomol. 75: 1069-1073.

Raun, E. S., 1963. Com borer control with Bacillus
thuringiensis Berliner. Iowa State J. Sci. 38: 141-150.

Raun, E. S., and D. L. Brooks. 1963. Bacterial pathogens
in Iowa com insects. J. Invertebr. Pathol. 5: 66-71.

Raun, E. S., and R. D. Jackson. 1966. Encapsulation as a
technique for formulating microbial and chemical insecti¬
cides. J. Econ. Entomol. 59: 620-623.

Raun, E. S., W. R. Lockhart, and R.J. Beers. 1959.
Microorganisms found in field specimens of diseased com
borer larvae. Proc. Iowa Acad. Sci. 66: 508-512.

Raun, E. S., G. R. Sutter, and M. A. Revelo. 1966.
Ecological factors affecting the pathogenicity of Bacillus
thuringiensis var. thuringiensis to the European com borer
and fall armyworm. J. Invertebr. Pathol. 8: 365-375.

Riba, G. 1984. Application en essais parcellaires de plein
champ d'un mutant artificiel du champignon
entomopathogene Beauveria bassiana [Hyphomycete]

contre la pyrale du mais, Ostrinia nubilalis [Lep.:
Pyralidae]. Entomophaga 29: 41-48.

Riba, G., and S. Marcandier. 1984. Influence de
rhumidite' relative sur l'agressivite' et la viabilite' des
souches de Beauveria bassiana (Bals.) Vuillemin et de
Metarhizium anisopliae (Metsch.) Sorokin, hyphomycetes
pathogenes de la pyrale du mais, Ostrinia nubilalis Hubn.
Agronomie 4: 189-194.

Roberts, D. W. 1981. Toxins of entomopathogenic fungi.
In H.D. Burges, ed., Microbial Control of Pests of Plant
Disease 1970-1980, pp. 441-464. Academic Press, New
York.

Sebasta, K., and K. Horska. 1968. Inhibition of DNA-
dependent RNA polymerase by the exotoxin of Bacillus
thuringiensis var. gelechiae. Bioch. Biophys. Acta 169:
281-282.

Stirrett, G. M., G. Beall, and M. Timonin. 1937. Afield
experiment on the control of the European com borer,
Pyrausta nubilalis Hubn., by Beauveria bassiana Vuill.
Sci. Agric. 17: 587-591.

Sutter, G. R., and E. S. Raun. 1966. The effect of
Bacillus thuringiensis components on the development of
the European com borer. J. Invertebr. Pathol. 8: 457-460.

Sutter, G. R., and E. S. Raun. 1967. Histopathology of
European com borer larvae treated with Bacillus
thuringiensis. J. Invertebr. Pathol. 9: 90-103.

Vakili, N. 1990. Biocontrol of stalk rot in com. Proc.
Annu. Com Sorghum Res. Conf. 44: 87-105.

York, G. T. 1958. Field tests with the fungus Beauveria
sp. for control of the European com borer. Iowa State J.
Sci. 33: 123-129.

157

The Development of the National
Biological Control Institute

Dale E. Meyer dirk1

Abstract

A National Biological Control Institute has been established
by the United States Department of Agriculture, Animal
and Plant Health Inspection Service in Hyattsville, Mary¬
land. Its mission is to promote, facilitate and expedite
biological control and provide national leadership.

Introduction

The United States Department of Agriculture (USDA), in
addition to other national and international agencies, is
vitally concerned about the misuse of pesticides. A shift in
emphasis toward non-chemical pest control techniques is
being implemented by a multitude of federal agencies. This
can directly benefit the farmers in terms of increased safety
and reduced production costs in comparison to use of
dangerous and expensive pesticides. There is growing
support in Congress for information and support for
projects emphasizing non-chemical approaches to combat
such pests as grasshoppers, wheat aphids, hydrilla, leafy
spurge, and others. Biological control is an effective non¬
chemical technology now available for the control of a wide
variety of agricultural pests.

Justification for Biological Control

The justification for biological control technology is to
assure food safety and affordability, enhance water quality,
prevent disease transmission, replace obsolete pesticides,
protect biodiversity, reduce petroleum depletion, and
increase profitability of farming. In addition, it protects
endangered species, and overcomes pesticide resistance in
target pests.

Definition of Biological Control

The definition of biological control adopted for this
presentation is "the action of parasites, predators or
pathogens in maintaining another organism's population
density at a lower average than would occur in their
absence" (DeBach 1964).

Historical Accounts

Through the "cooperation" of numerous research institu¬
tions such as the USDA, Agricultural Research Service
(ARS), Cooperative State Research Service (CSRS),
Extension Service (ES), and operational agencies such as
state departments of agriculture and the USDA, Animal and
Plant Health Inspection Service (APHIS), many classical
biological control programs have been completely success¬
ful in the United States. The cottony cushion scale, Icerya
purchasi Masked, was found attacking citrus in California
as early as 1868, but was successfully controlled in 1889 by
the vedalia beetle, Rodolia cardinalis (Mulsant) (Doutt
1958). Other complete successes include the cereal leaf
beetle, Oulema melanopus (Linnaeus), which was attacking
oat, wheat and barley (Pierce 1982); the citrus blackfly,
Aleurocanthus woglumi Ashby, attacking citrus (Dowell et
al. 1979); the Comstock mealybug, Pseudococcus
comstocki (Kuwana), attacking citrus and ornamentals
(Meyerdirk et al. 1981); and the alfalfa weevil, Hypera
postica (Gyllenhal) on alfalfa (Day 1981).

USDA, APHIS Biological Control Operational
Programs

APHIS has established two strategic biological control
operational laboratories within the agency. One is located at
Niles, Michigan and the other at Mission, Texas. A
satellite laboratory facility also has been established in
Bozeman, Montana to accommodate the biological control
of weeds programs. Operational programs are being
conducted against the alfalfa weevil, Russian wheat aphid,
Diuraphis noxia (Mordvilko), the Colorado potato beetle,
Leptinotarsa decemlineata (Say), the European com borer,
Ostrinia nubilalis (Hubner), diffuse and spotted knap¬
weeds, Centaurea diffusa Lam. and C. maculosa Lam., and
leafy spurge, Euphorbia esula Linnaeus.

As an action Agency, APHIS plans to develop a greater
commitment to enhance our biological control activities
within the United States.

'Plant Protection and Quarantine, USDA, APHIS, Hyattsville, Maryland

158

National Biological Control Institute

APHIS has recently established the National Biological
Control Institute (NBCI) to serve as a focal point for
providing national leadership in coordinating and imple¬
menting the regional and national use of natural enemies in
suppressing agricultural pests. NBCI will serve as a
national resource for the collection, analysis, and dissemi¬
nation of general information and scientific findings
regarding biological control.

Reason for Establishment by USDA, APHIS

The reason for the establishment of the NBCI by USDA,
APHIS was to assure success in the development and
implementation of biological control programs. There was
a need for an action agency at the national level to provide
leadership in the implementation of regional or national
programs for biological control of agricultural pests. State
and local governments are not in a position to support such
efforts. The increased commitment by APHIS to applied
biological control fills a void not formerly supported by
industry. As a result of the establishment of the NBCI, the
implementation of biological control in the United States
will be able to excel to its maximum potential.

Mission

The mission of the NBCI is to promote, facilitate, and
expedite biological control and provide national leadership.
The NBCI will serve as a national resource for the collec¬
tion, analysis and dissemination of biological control data,
provide training and technical information, service State
cooperators, and assist APHIS and other Federal Agencies
in expediting state, regional or national biological control
programs. The Institute will enlist involvement of research
institutions, identify and support the needs of cooperators
and other interest groups, facilitate and coordinate technical
education and training in biological control, and interface
with other integrated pest management technologies.

Linkage to Other Organizations

Research, development and implementation efforts in
biological control involves collective support from federal
and state agencies, universities, and the private sector. An

improved communication process is being developed to
assure close cooperation between APHIS, ARS, CSRS and
other federal agencies, plus state cooperators including
universities, agricultural experiment stations and depart¬
ments of agriculture, plus the private sector including
industry, special interest groups and other cooperators.
Biological Control Working Groups already established in
ARS include: classical, augmentation and conservation,
microbial, ecology and behavior, and natural products. A
USDA Interagency Biological Control Coordinating
Committee (IBC3) also has been developed within USDA.
The NBCI will develop close linkages with these groups
and others. In addition, the NBCI will develop a manage¬
ment team and establish a User-Advisory Committee
comprised of a variety of Federal agencies which will
advise, identify and prioritize national biological control
programs within NBCI.

Location

The NBCI is located in the Federal Building on the fifth
floor in Hyattsville, Maryland and houses the Information
and Documentation Center and administrative staff.

Organization

The NBCI organization consists basically of a Director,
Technical Consultant, Database Manager and Technical
Coordinator.

Director

The director will provide active and supportive leadership
for the NBCI. He or she will manage the institute's
resources, establish a user-advisory committee, develop
initiatives, develop strategic and operational plans; imple¬
ment marketing programs, maintain financial support, and
assure the quality of all services performed by the NBCI.

Technical coordinator

The technical coordinator will foster APHIS and
interagency liaison, integrate biologically based technolo¬
gies, coordinate biological control method development
needs, communicate and publicize biological control,
monitor APHIS extramural projects, facilitate training and
implementation, and build a biological control network.

159

Technical consultant

The technical consultant will provide technical advice to a
variety of user groups, provide technical information, build
informational linkages, and will also be responsible for
communicating and publicizing biological control.

Database manager

The database manager will supervise the Documentation
Center, develop and maintains all databases, develop and
maintain computer programs; and provide reports for
specific applications. Data will include tracking all foreign
collection information, importation, quarantine screening,
release and establishment of exotic natural enemies.

Organizational Interface Within APHIS

APHIS is comprised of several units under the Administra¬
tor including Plant Protection and Quarantine (PPQ). The
biological control operational program and laboratories are
in PPQ. The NBCI will provide a strong linkage and
support of PPQ operational programs to the Methods
Development Centers within APHIS and other Federal
Agencies.

References Cited

Day, W. H. 1981. Biological control of the alfalfa weevil
in the northeastern United States, pp. 361-374. In G. C.
Papavizas. Biological control in crop production.
Allanheld, Osmun, Totowa, N.J.

DeBach, P. 1964. Biological Control of Insect Pests and
Weeds. Chapman and Hall Ltd. , London. 844 pp.

Doutt, R. L. 1958. Vice, virtue and the vedalia.

Entomol. Soc. Amer. Bull. 4: 119-123.

Dowell, R. V., G. E. Fitzpatrick and J. A. Reinert. 1979.
Biological control of citrus blackfly in southern Florida.
Environ. Entomol. 8: 595-597.

Meyerdirk, D. E., I. M. Newell and R. W. Warkentin.
1981. Biological control of Comstock mealybug. J. Econ.
Entomol. 74: 79-84.

Pierce, R. 1982. Saga of the cereal leaf beetle war.

Agric. Res. Nov. 8.

Conclusion

Never before in the history of organized agriculture has
there been a period when so much interest and need for
biological control could be coupled with the technology for
its implementation. Biological control has moved to the
forefront as a viable means of controlling animal and plant
pests. We now have the opportunity to develop and
implement biological control programs that will signifi¬
cantly improve food safety and affordability, prevent
transmission of animal and plant diseases, provide alterna¬
tive pest control technologies, protect biodiversity, enhance
water quality, and preserve the overall environment. The
NBCI will enhance our opportunities of developing sound
biological control programs in the United States and
potentially international programs.

•£U.S. Government Printing Office : 1992

160

311-368/60524