Biopesticides: Use and Delivery (Methods in Biotechnology)

Biopesticides

METHODS

IN

BIOTECHNOLOGY’”

John M. Walker, SERIES EDITOR 12 EnvironmentalMonitoring of Bacteria,edited by CloveEdwards, 1999 11 AqueousTwo-PhaseSystems, edited by Rap Hattl-Kuul, 1999 10. CarbohydrateBiotechnologyProtocols,edited by Chrrstopher Bucke, 1999 9. DownstreamProcessingMethods,edited by Mohamed LIesal, 1999 8 Animal Cell Biotechnology,edlted by Nlgel Jenkms, 1999 7. Affinity Biosensors:Techniques and Protocols, edlted by Km R Rogers and Ashok Mulchandanr, 1998

6. Enzymeand Microbial Biosensors:Techniques and Protocols, edlted by 5. 4. 3. 2. 1

Ashok Mulchandam and Kern R Rogers, 1998 Biopesticides:Use and Delivery, edited by Frankhn R. Hall and Julius J. Menn, 1999 Natural ProductsIsolation, edited by Rxhard J P Cunnell, 1998 RecombinantProteinsfrom Plants:Prod&on and Isolation of Clinically Useful Compounds, edited by Charles Cunmngham and AndrewJ R Porter, 1998 BioremediationProtocols,edited by David Sheehan, 1997 Immobilizationof Enzymesand Cells, edited by Gordon F Blckerstafi 1997

Biopesticides Use and Delivery

Edited by

Franklin R. Hall Ohio State Unwersity, Wooster, OH

and

Julius J. Menn Excipula, Inc , Highland, MD

Humana

Press

Totowa, New Jersey

0 1999 Humana Press Inc 999 RIvervIew Drive, Suite 208 Totowa, New Jersey 075 I2 All rrghts reserved No part of this book may be reproduced, stored m a retrieval system, or transmltted m any form or by any means, electromc, mechamcal, photocopymg, mlcrofilmmg, recording, or otherwlse wlthout wrttten permIssIon from the Publisher Methods m Blotechnologyc\* IS a trademark of The Humana Press Inc All authored papers, comments, opmlons, conclusions, do not necessarily reflect the views of the publisher This publlcatlon IS printed on acid-free ANSI 239 48-1984 (American Standards Permanence of Paper for Prmted Library Covet Brian

lllustratlon A FederIc

Cover

design

Fig

by Patricia

For additional copies, Humana at the above E-mall

I A m Chapter

or recommendations

are those

of the author(s),

and

paper @ Institute) Materials I6 “Naturally

Occurrmg

Baculovlruses

for Insect

Pest Control,”

by

F Cleary pricing address

humanaahumanapr

for bulk purchases, and/or mformatlon or at any of the followmg numbers Tel corn,

or visit

our Webslte

about other 973-256-1699, http //humanapress corn

Humana Fax

titles, contact 973-256-834 1,

Photocopy Authorization Policy: Authorlzatlon to photocopy items for internal or personal use, or the Internal or personal use of specific clients, IS granted by Humana Press Inc , provtded that the base fee of US $8 00 per copy, plus US $00 25 per page, IS paid directly to the CopyrIght Clearance Center at 222 Rosewood Drive, Danvers, MA 01923 For those orgamzattons that have been granted a photocopy license from the CCC, a separate system of payment has been arranged and IS acceptable to Humana Press Inc The fee code for usersofthe TransactIonal Reporting Service IS [O-89603-5 15-8198 $8 00 + $00 251 Printed

m the Umted

Library

of Congress

Mam

entry

Methods

under

States Cataloging

of America m Pubhcatlon

IO 9 8 7 6 5 4 3 2

I

Data

title

In biotechnology’*

Blopesttcldes

use and dellvery/edlted by Franklm R Hall and Juhus J Menn cm --(Methods In biotechnology, 5) P Includes Index ISBN O-89603-5 15-8 (alk paper) I Natural pestlctdes 2 BIologIcal pest control agents 3 Agrrcultural 1 Hall, Franklin R II Menn, Juhus J III Series SB95l I45 M37B56 1998 632’ 954~2 I 98-25118 CIP

pests-Blologtcal

control

Preface It was our intention and goal to bring together m Biopestzcides Use and the latest advances in the science and technology of the evolving field of biopesticides In the context of crop protectton, btopesttcides are a key component of integrated pest management (IPM) programs, m which biopesticides are delivered to crops m inundative quantities, vs the moculative approach, which is charactertstic of classical biological control. Although there are several definitions of biopesttcides m the literature, we chose to define them as either microbials themselves or products derived from microbials, plants, and other biological entities. In the developed, industrial countries, primarily in Western Europe and the United States, biopesticides are receiving more practical attention, smce they are viewed as a means to reduce the load of synthetic chemical pestttides m an effort to provide for safer foods and a cleaner envtronment. In the developing countries, biopestictdes are viewed as having the potential to exploit nattve resources to produce crop protection agents that would replace imported chemical pesticides and conserve much-needed hard currencies These trends are well represented by the dynamic growth of engineered crops expressing the delta-endotoxm insecticidal protem crystals of Bacillus thuringzenszs (B.t ) m corn, cotton, and potatoes and the development of recombinant B.t.s and biopesticides as key crop protection agents against such pests as the soybean caterpillar, which is effectively controlled by a nuclear polyhedrous virus m Brazil, the use of neem extracts m East Africa and India, and various botamcals m East Africa and South America. The btopesticide market is expected to grow at a rate of 10% per year vs l--2% for chemical pesticides. The current biopesticide market IS estimated at $500 million worldwtde At the projected rate of growth, the sales volume ~111 double by the year 2007. It is likely that major breakthroughs m biopesticide technology ~111further increase the rate of growth of biopesttcide usage A pertinent example mvolves the efforts by several multmattonal companies to produce baculoviruses m deep fermentation Success m this area could provide a major boost to the mcorporation of baculoviruses as major crop protection agents m the biopesticide armamentarmm. Delavery

V

Preface

VI

We do not view biopesticides as replacements for chemtcal pesticides on major crops Biopesticides, viewed realistically, will most likely find uses as supplements to chemical pesticrdes and as rotation agents m early season on major crops to retard the onset of resistance. Other uses that will increase the acceptance of biopesticides ~111be m IPM programs on minor crops and niche markets We have mvrted leading experts m the btopesticide field to contribute comprehensive chapters on mode of action, development, productron, dehvery systems (formulations), and market prospects for the future. In addition, we invited registration experts from both government and industry to review current registration requirements, time frames, and costs of registration and compare them with registration requirements of conventtonal pesticides. We also have contributions describing momtormg procedures and management of resistance of pests to biopesticides. It is our goal that this volume will serve as the current, most comprehensive treatise on the rapidly emerging field of btopesttcides and a useful resource for practitioners, students, regulators, and mdustrial planners and marketers. Franklin Julius

R. Hall J. Menn

Contents .. .- . . .. . .. . .. . . . . . ,.,,... . . . .. . .. .. . .. . . .. . . .. . . , V Preface .. . . .. . . . . .. . . . . .., XI Lrst of Contributors . . .. . . .. . . . . , . .. . .. .. . . . .. .. . .. . 1 Biopesticides. Present Status and Future Prospects .. . . . . . . I Julius J. Menn and Franklin R. Hall.. . .. ... . PART I

PROJECTIONS ON OPPORTUNITIES FOR BIOPESTICIDES IN CROP PROTECTION . . . .. . . . .. .. . .. . . . . . . . . ..

..

I1

2 The North American Scenario . .. .. . . . . . .. 13 . . . . *. . .. ., , .. . Jerry Caulder 3 Microbial Bropesticides: The European Scene Tariq M. Butt, John G. Harris, and Keith A. Powell. . . . .. ...*.... . . . 23 4 Developing Countries 45 Balasobramanyan Sugavanam and Xie Tianjian . . . . .. . .. . .. . , . . 5 Pesticide Policy influences on Biopesticide Technologies 55 . . . .,.. . . .. . .. . .. .. . .. . .. . . . . . . . .. . . .. . Noel D. llri . . . . . .. . . 75 . .. . .. . . .. , . .. . .. . .. . . . . . PART I I. BIOFUNGICIDES . . 6 Commercral Development of Brofungicides 77 ...* ,. . Rafael Hofstein and Andrew C. Chapple . 7 Biological Control of Seedlrng Diseases . . . . 103 K. Prakesh Hebbar and Robert D. Lumsden . 8 Joint Action of Microbials for Disease Control .. . . . . 117 Claude Alabouvette and Philippe Lemanceau .. 137 . .. .. . .. . .. . . . . . . .. .. . . . .. . . . .. . . . .. .. . . PART II I BIOINSECTICIDES 9 Neem and Related Natural Products . . . . . . , . . . .. . ,. . . . . .. . . . .. 139 Murray B. lsman 10 Commercial Experience with Neem Products 155 . . .. . .. ... . ,. . James F. Walter 11 Fermentation-Derived Insecticide Control Agents The Spinosyns Thomas C. Sparks, Gary D. Thompson, Herbert A. Kirst, Mark B. Hertlein, Jon S. Mynderse, Jan R. Turner, . .. . . . 171 and Thomas V. Worden . . .. . . . . . . . . . . 12 Baa//us thunng/ensis. Natural and Recombinant Bllomsecticide Products 189 James A. Baum, T. B. Johnson, and Bruce C. Carlton . . .. . . . . vii

*., VIII

Contents

13 Transgenrc Plants Expressing Toxms from Bacillus thuringrems . , . .. . .. Jonnie N. Jenkins 14 Production, Delivery, and Use of Mycoinsectrcides for Control of Insect Pests on Field Crops . Steven P. Wraigt and Raymond 1. Carruthers . 15 Entomopathogemc Nematodes .. . .. , Parwinder Grewal and Ramon Georgis 16 Naturally Occurring Baculovrruses for Insect Pest Control . . . . Brian A. Federici . . . .. . . . 17 Recombinant Baculoiviruses . . .. , . .*.. . . .,... . . .. .. ... Michael F. Treaty 18 Joint Actron of Baculoviruses and Other Control Agents ,.. .. .. William F. McCutchen and Lindsey Flexner . . . . . .. . PART IV BIOHERBICIDES . . . .. . .. . .. . . . . . . .. . .. . .. . .. . .. . . . . . . .. . .. . . .. . , .. . .. .,.,. 19 Mycoherbrcrdes . .. Alice L. Pilgeram and David C. Sands . . . 20 Formulation and Application of Plant Pathogens for Brologrcal Weed Control . . .. . .. .. . Nina K. Zidack and Paul C. Quimby .. . . . PART V OTHER BIORATIONAL TECHNOLOGIES 21 Phereomones for Insect Control. Strategies and Successes . . .. . . .. . . . . D. R. Thomson, L. J. Gut, and J. W. Jenkins . . . . ., PART VI REGISTRATION OF B!OPESTICIDES ,..... ..,. . , , . 22 The Federal Registratron Process and Requirements for the United States .. ... . . .. . ., J. Thomas McClintock , .. . . 23 IR-4 Bropestrcrde Program for Minor Crops . , Christina L. Hartman and George M. Markle 24 RegistratrorYRegulatory Requirements In Europe Mike Neale and Phil Newton .. 25 Environmental and Regulatory Aspects, industry Wew and Approach .. Joseph D. Panetta , .. PART VII MANAGEMENT PROTOCOLS . . .. . 26 Formulatrons of Biopesticides Susan M. Boyetchko, Eric Pedersen, Zamir K. Punja, . . ., , . .. . . and Munagala S. Reddy .. . .. . . . . . .. . . 27 Delivery Systems and Protocols for Biopesticides ,.. ,. .. Roy Bateman ... . .,.* . . *. . . . .. . .* .. .

211

233 271 301

321 34 1 357 359

371

383 385 473

415

443 453 473 485

487 509

ix

Contents 28

Analysis, Momtormg, and Some Regulatory Implications *.*,... . . . ., .. . . . . .. . .. . . Jack R. Piimmer 29 Principles of Dose Aquisitlon for Bioinsecticides . .. . .. . . . .. .. . .* * Hugh F. Evans, .., ..*... . . ....* .. . 30 Strategies for Resistance Management .. .. , . ,. . ., . ..., . . . . .. . .. .. . Richard T. Roush 31 Field Management’ Delivery of New Technologies to Growers .. . Mark E. Whalon and Deborah L. Norris . . . .. .. . . . .*... ..* . . . Index . . .. . . . , . .

529 553 575 .

595 609

Contributors Laboratoire de Recherches SUPla Flare pathogene duns le Sol, Doon, France ROY BATEMAN International Institute of Biologtcal Control, Ascot, UK JAMES A. BAUM Ecogen, Inc., Langhorne, PA SUSAN M. BOYETCHKO Agriculture and Agri-Food Canada, Saskatoon, Canada TARIQ M. BUTT IACR-Rothamsted, Harpenden, UK BRUCE C. CARLTON New Jersey Agrtcultural Experimental Statton, Cook College, Rutgers Untversity, New Brunswick, NJ RAYMOND I. CARRUTHERS National Program Stag USDA/ARS, Beltsvtlle, MD JERRY CAULDER Mycogen Corporation, San Dtego, CA ANDREW CHAPPLE Ecogen, Inc., Langhorne, PA HUGH F EVANS Forestry Commtsston Research Agency, Wreccleshan, UK BRIAN A. FEDERICI Department of Entomology and Interdepartmental Graduate Program tn Genetics, University of California, Riverside, CA LINDSEY FLEXNER DuPont Agricultural Research Center, Newark, DE RAMON GEORGIS Therm0 Trtlogy Corp., Columbta, MD PARWINDER GREWAL Department of Entomology, Ohio State Untverstty, Wooster, OH L. J. GUT Department of Entomology, Michigan State University, East Lanstng, MI FRANKLIN R. HALL Ohto State University. Wooster, OH JOHN G. HARRIS Zeneca Agrochemicals, Bracknell, UK CHRISTINA L. HARTMAN IR-4 Biopesticide Coordinator, Rutgers, The State Untverstty of New Jersey, New Brunswick, NJ Present Address. Agricultural Products Research Division, American Cyanamid Co , Princeton, NJ K. PRAKESH HEBBAR USDA/ARS, Beltsvtlle, MD MARK B. HERTLEIN DowElanco Discovery Research, Indianapolis, IN RAPHAEL HOFSTEIN Ecogen, Inc., Langhorne, PA MURRAY B. ISMAN Department of Plant Science, University of Brtttsh Columbia, Vancouver, Canada CLAUDE ALABOUVETTE

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

xi

xii

Contributors

JOHNIE N. JENKINS

Crop Sctence Research Laboratory, USDA/ARS, State, MS J. W. JENKINS Pactfic Btocontrol Corp , LitchJield, AZ TIMOTHY B. JOHNSON Ecogen, Inc , Langhorne, PA HERBERT A. KIRST Elanco Animal Health Research and Development, Greenfield, IN PHILIPPE LEMANCEAU Laboratotre de Recherches sur la Flore pathogene darts le Sol, Dyon, France ROBERT D. LUMSDEN USDA/ARS, Beltsvtlle, MD GEORGE M MARKLE . IR-4 Associate Dtrector/Professor, Rutgers, The State Untverstty of New Jersey, New Brunswtck, NJ J THOMAS MCCLINTOCK USEPA, Washtngton, DC WILLIAM F. MCCUTCHEN DuPont Agricultural Research Center, Newark, DE JULIUS J. MENN 9 Exctpula, Inc , Highland, MD JON S MYNDERSE Lilly Research Laboratories, Elt Lilly and Co, Indtanapolts, IN MIKE NEALE Novartts Crop Protectton, Basel, Switzerland PHIL NEWTON Novartts Crop Protectton, Basel, Swttzerland DEBORAH L. NORRIS Michigan State Unzversity, East Lansing, MI JOSEPH D PANETTA 9 Regulatory and Envtronmental Affairs, Mycogen Corp , San Diego, CA ERIC PEDERSEN Agrtum Inc., Saskatoon, Canada ALICE L. PILGERAM 9 Department of Plant Pathology, Montana State Untverstty, Bozeman, MT JACK R. PLIMMER Tampa, FL KEITH A. POWELL Zeneca Agrochemtcals, Bracknell, UK ZAMIR K PUNJA Centre for Pest Management, Simon Fraser Untverstty, Burnaby, Canada PAUL C. QUIMBY, JR Northern Plains Agrtcultural Research Laboratory, USDA/ARS, Stdney, MT MUNAGALA S. REDDY Saskatoon, Canada RICHARD T. Rousn Department of Crop Protectton, Untverstty of Adelatde, Glen Osmond, Australia DAVID C. SANDS Department of Plant Pathology, Montana State Untverstty, Bozeman, MT THOMAS C. SPARKS DowElanco Dtscovery Research, Indtanapolts, IN BALASUBRAMANYAN SUGAVANAM UNIDO, China GARY D. THOMPSON DowElanco Dtscovery Research, Indtanapolts, IN D. R. THOMSON Pact@ Btocontrol Corp , Vancouver, WA l

Mississippt

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

Contributors

xii

XIE TIANJIAN UNIDO, China MICHAEL F TREACY Amerctan Cyanamid Co., Prtnceton, NJ JAN R TURNER Ltlly Research Laboratones, Elt Lilly and Co , l

l

l

Indtanapolts, IN NOEL D. URI Economtc Research Service, US Department of Agriculture, Washington, DC JAMES F. WALTER Rohm and Haas Co., Philadelphia, PA MARK E. WHALON 9 Michigan State Untverstty, East Lansing, MI THOMAS V. WORDEN DowElanco Discovery Research, Indtanapolts, IN STEPHEN P WRAIGH~ Plant Protectron Research Unit, US Plant, Soil, and Nutrition Laboratory, USDA/ARS, Ithaca, NY NINA K. ZIDACK Department of Plant Pathology, Montana State Untverstty, Bozeman, MT l

l

l

l

l

1 Biopesticides Present Status and Future Prospects Julius J. Menn and Franklin R. Hall

1. Introduction Bropestlcrdes, mcludmg mrcrobral pestrctdes,entomopathogenic nematodes, baculoviruses, plant derrved pestrcides, and Insect pheromones, the latter when used as mating drsruptlon agents, are recetvmg increased exposure m scientific annals and the lay press, as alternatrves to chemical pestrctdesand as key components of integrated pest management (IPM) systems(2,2). The reality, however, IS that biopesticrdes currently represent only a small fractron of the world pestrcide market. At the present time, various economrc forecasting servtces estrmated the world market for pesticides m 1995 at approx $29 btlhon 0. The bropestrcrde share of the market was estimated to be around $380 million in 1995 (1). Although representing only 1.3% of the total, and since the majority of bropesticides are currently marketed for insect control, blopesttcides represent approx 4.5% of the world insecticide sales. However, the growth rate for bropestrcrdes over the next 10 years has been forecast at 1O-l 5% per annum m contrast to 2% for chemical pesticides (2) The foregoing IS based on several assumpttons that are predicated on the successor failure of several technologies; notably, transgemc crops expressing Baczllus thuringzensis (Bt) insect protein toxins. Should the latter be successful and allowmg that Bt engineered crops are m the biopesttcide arena, the growth rate of biopestrcides will be much greater than forecast The potential for resistance developmg to B&engineered crops m lepldopterans may alter the equatron stgmficantly. Furthermore, the agrochemlcal mdustry has been introducing recently hrghly Insect-pest-specific insecticrdes wrth modes of actlon From

Methods F R Hall

m B/otechno/ogy, and

J J Menn.

vol 5 B/opestmdes eds

0 Humana

1

Press,

Use and De//very Totowa,

NJ

2

Menn and Hall

that are targeted to pest species.Such selective msectlcldes may also slow down the rate of growth of blopestlcldes. However, the science, technology, and marketing of blopesticldes are forging ahead as key components of IPM and sustamable agriculture. In this volume we have brought together a select group of biopesticlde experts reporting on the current status and prospects for blopestlcldes, their role m crop protection, delivery systems,monitormg resistance management, marketing, reglstratlon and pohcles influencing their future prospects. It IS our mtent to provide the readers of this volume with the state of knowledge, current status, and future prospects for blopestlcldes as key components of IPM programs in contemporary agriculture. 2. Opportunities for Biopesticides The followmg discussion highlights chapters included m this volume. The prospects and economic projections for bropesttcldes m the United States, Europe, and developing countries are discussed m Chapters 2, 3, and 4 The current state of blopestlclde development for the Unlted States (Chapter 2) 1s presented from an industry standpoint and although optlmlstlc, expectations for future commercializations are tempered by a plea for increased coordmatlon by industry, academics, growers, and others, because of the need to more effectively utilize these tools within the context of IPM, transgemc plants, and so forth. The European scene (Chapter 3) 1sa detailed descrlptlon of some of the microbial experiences and lrkewlse calls for added support from Extension/industry m the “implementation” phase. In the developing countries (Chapter 4), experiences (mainly with Bt) m China, Thailand, and other countries are identified wherein quality control, education, and a request for more activity from international organizations will aid the transfer of blopestlcide technology to active usage as complementary technologies to chemical pestlcldes, rather than as “mere replacements ” As a final component to the sectlon on opportumtles, pesticide pohcy issues, constraints and incentives for new technologies, and the recent impact of US legislation are outlined m Chapter 5 Although the outlook IS comphcated, the key factors of pesticide regulation, the Fan Act and Food Quality Protection Act (FQPA), government subsidies affecting crop dlverslty, management expertise with ecologlcally based systems, and “demand” for green produce will influence technology identification and lmplementatlon 3. Biofungicides In a section focusing on Blofunglcldes, it 1s clear that relatively few blofunglcldes have reached marketing status to date. AQ 1O@1sa recent mtroduction as a control agent against powdery mildew of fruit, primarily grapes

Blopes trades

3

(Chapter 6; ref. 3). These authors provide a well-documented roadmap of laboratory and field experiences with AQlO@ that should serve as a clear guide to “what works and what doesn’t” m developing a better understandmg of these btoagents n-rthe drffcult “real world ” As the authors conclude, dtssectmg the molecular elements contributing to fungal pathogenrcity will indeed aid progress m thts arena as we attempt to insure that all possrble attributes are expressed m targeting thts brofungtctde. It will be interesting to follow the fortunes of thts proneermg mtroductron m competttton wtth an old establtshed product, such as sulfur, and newer selective chemrcal fungtcrdes. Brofungrcrdes for control of seedling diseases, such as pythmm, fusarmm, rhizoctoma, and vertrcillmm, have received mcreasmg attention m recent years (41, primarily as nontoxrc, nonresidue producmg control agents. The utrhty and development of Glzocladzumvzrens for control of sorlborne pathogens IS described m Chapter 7, and the authors emphasize the necessity of a logical, well-documented series of studies from dtscovery through ecologtcal understanding and final usage studies m order to fully develop opportunmes for seedlmg brocontrol agents. Joint actton of brofungtctdes IS a relatively new area of mvestrgatton. Joint actton, or posstbly synergy, may increase the utrhty of brofungicrdes, as was demonstrated by combmmg a nonpathogemc strain of Fusarzum oxysporum wtth a strain of Pseudomonasfluorescens to control Fusarzum wilts (5) More recent advances m this mterestmg arena are described m Chapter 8. In thts complex array of poorly understood interacttons, It IS clear that good knowledge of antagomsts and then behavior are key constraints to more raprd commerctalrzatron It IS likely that niche markets may well prove to be useful starting points for marketing these btocontrol agents 4. Bioinsecticides Broinsecttcrdes represent the maJor segment of bropesticldes and comprrse the largest array of diverse mtcrobrals and natural products m the bropestrcrde armamentarium. Chapters 9 and 10 deal primarily with neem seed derived products and neem or1 as insect growth regulators, antrfeedants, msectrcrdes, and fungicides, and provide a clear outline of the hrstory and the current and future status of neem-related products. The potential of Azadnachtm, the prmctpal Insect-active macrocyclic lactone component of neem seed 011, has broad msectrcrdal actrvrty that makes rt an attractive candidate insecticide for specralty and niche crops (6). To date, the most economrcally significant bromsectrcrdes were the avermectms, derived by fermentation of a streptomyces species and chemrtally defined as macrocylic lactones The most notable are abamectrn (avermectm Bl) and Its chemrcally modified analog, Emamectm benzoate

4

Menn and Hall

(MK-244). These blomsectlcldes were described m numerous publlcatlons, including ref. 7. More recently, a chemically related class of fermentation-derived msectlcldes, the spmosyns from Succharopolyspora species, was introduced by DowAgroSclences (8). The lead member IS spmosad, with excellent msectlcldal activity against lepldopterous larvae and other msects (Chapter 1I). An Interesting story describing Its development 1s hlghhghted by the authors’ awareness of resistance management and the opportunities to develop a unique set of molecules possessingsafe envlronmental and mammalian profiles, while at the same time proving to be selective, and thus somewhat market restricted. Clearly, there would seem to be “more to come” from this group of interesting bioactive naturally derived products. The best known group of blopestlcides, the Bts, are also produced by fermentation but yield insectlcldal crystal proteins rather than discrete chemicals, such as the spmosyns and avermectins By selection of natural Isolates from the thousands known, as well as those derived from plasmtd conJugatlon and recombinant DNA technology, it has become feasible to tailor make Bts that are targeted against specific insect pests (9,ZO; Chapter 12). Although identifying the limitations of field persistence and mcomplete coverage of target surfaces, the authors project a positive outlook for this vast array of toxins as attractive alternatives to existing products. The successwill depend on developments m improved formulations and an improved understanding of dosetransfer processes (as is the case with almost every blopestlcldal agent). The Bt technology has been taken a step further by engineering cotton, corn, and other crops to express truncated versions of the cryIA genes from Bt vanety kurstaki Engineered plants provide protection against certain lepidopterous larvae, notably Heliothis vzrexens, a major pest of cotton and corn (IO; Chapter 13). Although the picture looks bright as indicated by the recent flurry of merger activity among seed, chemical, and biotechnology companies, such as Monsanto, DuPont, AgrEvo, and others, rt IS tempered with a divergence of opmrons about llmltatlons resulting from pest resistance. As noted m Chapters 14 and 30 (under management strategies), one of the key concerns mvolvmg transgemc crops expressing Bt toxins IS the potential for resistance development m target insects. Contmuous exposure to the toxin challenge has a high probablllty for selection of reslstant mdlvlduals. Strategies have been developed to delay and/or ameliorate the onset of resistance mvolvmg establishment of refugia and high expression of Bt protein toxins in engineered plants (12,13) Entomopathogenic fungi (mycoinsecticldes) are gaming increased attention as environmentally friendly insect control agents. Although over 750 species were reported to infect insects, few have received serious conslderatlon as potential commercial candidates. Beauverza bassiana appears to have the

Biopestmdes

5

broadest potential as a viable msect control agent (14). The status of this area of insect control IS identified m Chapter 14. Technological advances m production, formulation, and shelf life have contributed substantially to the viabtlrty of mycoinsectictdes as practtcal insect management agents that can compete economically with chemical insecticides in certain situations, such as fruit and specialty crops. A comprehensive review of the use of entomopathogemc nematodes is covered m Chapter 15 Summarizing the rapid commercialization (prmctpally in the 199Os), the authors relate these successesto large-scale productton technology and innovative formulations allowing shelf stability that has allowed a cottage-type industry to emerge in both the United States and Europe. An ongoing plea IS made for more integration of crop protection technologies and technology transfer trainmg; both are emphasized as constraints to continued progress. Baculoviruses are summarized in three chapters dtscussmg natural (Chapter 16) and recombinant (Chapter 17) baculovnuses, as well as experiences with jomt action of baculovu-uses with other control agents (Chapter 18) Baculoviruses have already been employed as biopesticides m the United States, m the 1970s for control of the cotton bollworm, Helicoverpa zea, m cotton and the gypsy moth, Lymantrza dispar, m forests. However, the product, Elcar, for control of H. zea failed as a commercial entity, largely resultmg from inability to compete with the then newly introduced pyrethroids, mstabthty in the field, slow action, and narrow spectrum of bioactivity (15). In contrast, considerable success m control of the velvetbean caterpillar, Anticawa gemmatalis, was achieved m Brazil on large soybean acreage (16, Chapter 16). However, these represent isolated cases of successful field use of baculovnuses. Broader use was precluded as a result of the already listed shortcommgs. With the advent of recombmant DNA technology, baculoviruses have received renewed attentton because of their small genome and uncomplicated molecular organization, allowmg for ready mtroductton of quick-acting msect toxic genes, such as an engineered portion of the scorpion toxin, AaIT, which greatly improves the field performance of the engineered baculovuus by killing the pest insects more rapidly (17,28; Chapters 17 and 18). An improved understanding of the toxtctty processes involved with the joint action of baculovirus with other toxins and chemical pesticides may jump start this group of viruses to greater market potentials. 5. Bioherbicides Btologtcal weed control (bioherbicides) has been actively pursued for several years, primarily m academic and government research institutions. To date, over 100 weed pathogens have been reported (19). However, only a handful of

6

Menn and Hall

bioherbictdes have entered commercial channels with very limited success, primarily because of their mabihty to compete with chemical herbicides that are specific, economic, and pose no environmental and residue hazard. However, bioherbicides are ideally suited for control of undesirable vegetation m pasture and rangeland, where the use of chemical herbtctdes would be prohibitively expensive (Chapters 19 and 20) 6. Other Technologies The successful development of pheromone based insect mating disruption m fruit trees and vineyards m North America and Western Europe has given pheromones a new dtmension as biopestictdes and provides promise that this technology, using pheromones as biocontrol agents, ~111reduce dependence on chemical pesticides for insect control m orchards and vmeyards (Chapter 2 1). 7. Registration of Biopesticides In recogrntton of the White House polictes concernmg environmental quality improvements and reduced dependence on chemical pesticides, wtth stated but unlikely attainable goals that 75% of chemical pesticides will be replaced with biopesticides by the year 2000, the USEPA established a Biopestictdes Pollution and Prevention Divtsion (BPPD) to manage accelerated registration and registration of biopesticides. The BPPD approved for registration 14 new biopestlctdes m 1995 and 10 m 1996, representing 35-40% of all new pesticide registrations. The average duration for registration of a btopesticide has been 12 mo vs 36-45 mo for a conventional chemtcal pesttctde. Furthermore, the agency required stgmficantly fewer data for a biopesticide m support of fmdmg no significant adverse effect to humans and the environment. The agency provides for an expedited review of the registrant’s application, the tolerance fee 1s waived, and requirements for an emergency use permit (EUP) may also be waived. The registration and associated policies m the US and Europe are described m detail m Chapters 22 and 24. The enlightened pohcuesand regulations governing btopesticides are especially sigmficant with regard to the approval process for use of biopestictdes on minor crops involvmg the IR-4 program (Chapter 23). The views of the industry relating to recent “safer” policies, by USEPA and other agenctes, are described m Chapter 25 Future expectations are tempered by the expressed need (by industry) for greater incentives provided by the USEPA, a renewed focus on clarifymg the benefits of such agents, greater coordmation/consolidation of databases, and greater flextbility m registration processes to provide recognition and encouragement of these more selective molecules offering greater environmental and human safety.

Biopes ticides

7

8. Management Protocols Completmg the “hfe cycle of development” of crop protection biopestictdes IS that area that seemsto be mcreasrngly important in successfultmplementation of more complex molecules/strategres,i.e., user management.Biopestrcides,posstbly even more so than chemical pesticides,are very dependenton proper formulatton and delivery systemsas key elements in performance. Furthermore, shelf life and ratesof apphcation are also largely a function of appropnate formulations (Chapters 26 and 27). Momtonng the fate of biopesticidespost application 1salso critical m determimng efficacy and survival of btocontrol microbials. Much of the momtormg scienceand technology ISrelatively new and mvolves growing emphasisand attention to this area(Chapter 28). The doseacquisttion(including pest encounter)processes(and realitiesof poor use efficiencies)arejust now being betterunderstoodandtranslatedto improved placement cnterra for thesemolecules.Chapter29 provides a somewhatspectficbut unique look into the array of detailsneededfor anoptimizeduse.It provides an interestingcorollary to the storyrelatedin Chapter6, I.e.,the researchexperiences,researchma1successand failures with AQ 10 However, whethercompaniesarereadyto assumeagreaterrole in more basicstudiesmvolvmg this technologyis yet unknown. Resistance management 1sa recurring theme throughout, with recogmtion bemg given to this Important phase of developing any crop protection agent. Chapter 30 provides some interesting points about the practicality of managing resrstance, which is mcreasmgly receiving much attention by the industry, A final thrust to the story of bropesticide development covers the user acceptance of increasingly complex crop protection strategies (Chapter 3 1) These new crop protection agents (biopesticldes) are more specific, lessrobust m environmental persistence, require more precise timing, offer significant opportumtres as alternatives to conventional pesticides, yet clearly are not quite as reliable. Management of information, monitormg, and use profiles will help m achieving success,but this will require additional grower education by all parties concerned with education and technology transfer of new crop protectton strategies. Continued emphasis by the industry on conventional chemistries as the “development model” for pesticides, the lack of new visionary thmkmg about long-term economic and ecological benefits of biopesticides, and needed educational thrusts about cosmetic standards, performance expectations, and detailed use requirements remam serious constraints hmitmg this technology (20). 9. Conclusions In summary, bropesttcides are achieving a modicum of growth as alternatives to conventional pesticides. However, their full potential has yet to be reached. The key factors/trends envisioned to significantly impact future developments mclude the following

Menn and Hall Regulatory/legislative and economic pohcles, mcludmg the new FQPA, may encourage further growth via fast trackmg/reduced reglstratlon reqmrements for bloratlonals. Developed country policies (globally) will continue to press for safer materials and alternatives to chemical pestlcldes, mcludmg more complex crop protectlon strategies, such as IPM/ICM, and others Developing countries have already mltlated thrusts to more efficiently utlhze such materials User education m both cases remains a serious constraint for Increased growth of blopesticldes Greater attention by industry/academra to reduced risk agendas (human and envlronment), as well as an Improved identlficatlon of the economic benefits profile of a crop protection strategy will greatly aid a sustainable growth of bropestlcldes Industry experiences such as with AQ 10, further illustrate the need for a comprehensive review of the plant/disease ecosystem complexity regarding the reduced reliance on the sole “conventional wisdom” approach of tradltlonal pestlcldes Blopestlcldes are more specific, more complex, less robust m the environment, and require greater knowledge about the ecosystem mteractlons. Partnermg of expertise will engender greater successwith these more complex agents, 1e , mterdlsclphnary teams must be the accepted norm m the development strategy There must be continued recogmtlon of the “power” of the transgetuc technology to influence a sustainable agriculture and plans made for such options as reslstance management, pestlclde pohcy mfluences on crop diversity, density, and other variables, frequency of planting, and so forth. Recogmtlon by the industry that “delivery” Is an essential part of the crop protection use pattern ~111 aid the speed of user acceptance of blopestlcldes Contmuatlon of the emphasis on environmental and ecotoxiclty requirements should aid increased acceptance and registration of btopestlcldes, whrch will probably remam in niche crops. The ldentificatlon and utrllzatlon of natural plant defense mechamsms interfaced with temporal/spatial phenomena may reveal an mterestmg array of new product opportunities. Industry willingness to mvest more than the traditional knowledge/ marketing skills m this approach wdl dictate ultimate success. Recent industry mergers and amalgamations accompamed by developments m transgemc technologies and value-added products would suggest not only unique crop protection opportumtles, but also reveal a powerful agenda m life science strategies for the new millennium agriculture.

References 1 Menn, J J (1997) Blopestlcldes-are they relevant7 m Focus on Blopestzcrdes, The Royal Society of Chemistry, pp, 1,2. 2 Menn, J J (1996) Biopestlcldes Has then time come? J Envrron ,%I Health B 31(3), 383-389.

3 Daust, R A and Hofstem, R. (1996) Ampelomyces quuqualu, a new blofunglclde to control powdery mildew rn grapes Brrghton Crop Protection Conference on Pests and Diseases-November 18-2 1, 1996, BrIghton, UK, 1, 33-40.

Biopes ticides

9

Papavizas, G. C. (1985) Trichoderma and Ghocladmm: biology, ecology, and potential for blocontrol. Ann Rev Phytopathol 23,23-54. Alabouvette, C. (1996) Blological control of Fusarium wilts IOBC wprs Bulletln/Bulletln OILB srop 19(8), 58

Leskovar, D. I. and Boales, A K (1996) Azadlrachtm: potential use for controlling lepidopterous insects and increasing marketability of cabbage Hort Scr 31(3),405-409

Mrozlk, H (1994) Advances m research and development of Avermectms, m Natural and Engineered Pest Management Agents (Hedm, P A , Menn, J. J , and Hollingworth, R. M , eds.), ACS Symposium Series, vol 551, no 5, American Chemical Society, Washington, DC, pp. 54-73. 8 Larson, L L , Sparks, T C , Worden, T V., Winkle, J. R., Thompson, G D., Klrst, H A., and Mynderse, J. S. (1996) Spmosad, the first members of a new class of Insect control products, the Naturalytes Proc. XX Znternatlonal Congress ofEntomology 19-009, 9

10.

11

12 13.

587

Carlton, B. C. (1992) Development of improved bloinsectlcldes based on Bacdlus thuringzensis,m PestControlwzthEnhancedEnvironmentalSafe@(Duke, S O., Menn, J J , and Phmmer, J. R , eds.), ACS Symposmm Senes, vol. 524, no. 18, American Chemical Society, Washington DC, pp 258-266. Carlton, B C (1993) Genetics of BT insecticidal crystal proteins and strategies for the construction of improved strains, in Proceedings of the X&W Be&v&e Symposzum(Lumsden, R. D. and Vaughn, J, L , eds.), ARS, USDA, Beltsvdle, MD, pp 326-337 Jenkms, J. N. (1993) Use of Bacrllus thurlnglensls genes m transgenlc cotton to control lepldopterous insects (Duke, S 0 , Menn, J J , and Pllmmer, J. R., eds ), ACS Symposium Series, vol. 524, no 19, American Chemical Society, Washington, DC, pp 267-280 Tabashnik, B E (1995) Resistance to msectlctdes, Bacillus, and transgemc plants Pestle. Outlook 6(4), 24-27 Tabashmk, B. E , Malvar, T , Lm, Y-B., Finson, N., Borthakur, D., Shin, BS , Park, S-H , Masson, L., Maagd, R. A., and Bosch, D. (1996) Crossresistance of the DIamondback moth indicates altered interactions with domam II of Bacillus thurlngienszs toxins. Appl Environ Mlcrobzol 62(8), 2839-2844,

14 Feng, M. G., Poprawskl, T. J., and Khachatounans, G. G (1994) Production, formulation and application of the entomopathogemc fungus Beauverla basslana for Insect control. current status. Blocontrol SczenceTechnol 4, 3-34 15 Huber, J (1986) Use of baculovlruses m pest management programs, m The Bzology ofBaculovtruses, vol. 2 (Granados, R R and Fedenci, B. A , eds.), CRC, Boca Raton, FL, pp I8 l-202 16. Moscardl, F. (1988) Production and use of entomopathogens m Brazil, m Bzotechnology, Blologlcal Pesticidesand Novel Plant-Pest Resistancefor Insect Pest Management (Roberts, D. W. and Granados, R. R., eds.), Cornell University

Press, Ithaca, NY, pp 53-60

10

Menn and Hall

17 Bonnmg, B C and Hammock, B D. (1996) Development of recombmant baculovuuses for insect control Ann Rev Entomol 41, 191-2 10 18 Kmg, L A , Possee, R D , Hughes, D S., Atkmson, A E , Palmer, C P , Marlow, S A , Ptckermg, J M , Joyce, K A , Lawrte, A M , Mtller, D P , and Beadle, D. J (1994) Advances m Insect vtrology Adv. Insect Physlol 25, l-73 19 Zorner, P S , Evans, S L , and Savage, S D (1993) Perspecttves on provtdmg a reallsttc techmcal foundanon for the commerctahzatton of btoherbtcrdes, in Pest control with Enhanced Envwonmental Safety (Duke, S 0, Menn, J J , and Plrmmer, J. R., eds ), ACS Symposmm Series, vol 524, no 6, American Chemtcal Society, Washmgton, DC, pp 79-86 20 Gaugler, R (1997) Alternative paradigms for commercializing btopesttctdes Phytoparasltzca 25(3), 179-l 82

I

PROJECTIONS ONOPPORTUNITIES FOR BIOPESTICIDES IN CROPPROTECTION

2 The North American Scenario Jerry Caulder 1. Historical Trends and Highlights of Significant Advances The first glimmermg that microbes could be used to control msects is generally traced back to 1834 when Aogostmo Bassi discovered that a fungus, Beauveria basslana, caused an mfectious disease m the stlkworm. However, it was not unttl some 40 yr later that the first attempts were made to use insect pathogens to control pest populations, when Metchnikoff m Russia expertmented with A4etarhizium anzsoplzae for control of a beetle attacking wheat. Large scale attempts to use Insect pathogens took place in the Umted States toward the end of the 19th century, when a variety of pathogens were evaluated for control of the chmch bug (I). The tdentlflcatlon of bactertal pathogens of insectswas not far behind A key dtscovery took place in 190 1 when Ishiwata discovered a Bacdlus spp attackmg the silkworm Berliner discovered the same species mfectmg flour moths m Germany in 1915, and named the organism Bacrllus thurrng~ensu (Bt), after the German province of Thueringen (2), At the same time, others were looking elsewhere for novel ways to control msect pests and decrease the substantial crop losses they caused each year. Insecticidal chemicals were identified from a variety of sources,mcludmg such compounds as lead arsenate, as well as natural plant compounds, such as the pyrethrins and mcotme. In 1939, the era of the synthetic organic msecttctdes was ushered m with the discovery by Mueller of the msectictdal properties of DDT. The excellent efficacy and low cost of petrochemrcal-based compounds like DDT, coupled with then quack kill and broad spectrum of actrvtty, quickly made them the universal standards for controllmg Insect and mite pests. Research with biopesticrdes contmued to advance, with tdentification of new strains and species of pathogens and their metabolites. But it was not until the From Methods UI Biotechnology, vol 5 Btopesbodes Use and Dellvery Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

13

14

Caulder

1940s that the first btopesttctde product was commerctahzed m the Umted States, when a bacterial product based on Bacllluspoplllae was mtroduced for control of the Japanese beetle. In the 1950s the first Bt products were commercrahzed m the Umted States for control of a vartety of caterpillar pests attackmg crops and forests The greatest advances m the area of btopestrctdes have taken place wtth Bt products. These advances involved isolation of novel strams with htgher potency, as well as tdenttficatton of new strains producmg entirely novel toxms active on different pest species The tdenttficatton of Bt variety tsraelensts (3) as a potent mosqutto larvtctde suggested that Bt toxms could be active on a variety of targets. The 20 yr of research and product development that followed that key discovery have more than borne out those expectations At the same time, the revolution m molecular biology and genetic engmeermg allowed the unique and highly potent 6-endotoxms produced by Bt to be manipulated u-t a variety of ways that substanttally enhance then utthty and performance Some of these key advances are discussed m more detail below.

2. The Biopesticide

Conversion

As we look at the growing interest m bropesttctde-based products, tt 1s easy to see that a maJor trend m agrtculture today 1s a converston of pest management practtces from tradtttonal broad spectrum chemical pesticides to highly specttic btologtcally based products What has created and driven this btologtcal converston, and what are the prospects for btopesttctdes in North America? We can start by notmg the very different approaches involved m tdenttfymg new chemical pesticide leads vs new btopesttctde leads. On the surface, they appear quite stmtlar, because a great deal of drscovery and screening IS mvolved m both types of crop-protectron agents. However, the underlymg concepts are fundamentally different The chemtcal screening approach has htstortcally been very much an Edtsoruan search, i.e., a random screening of thousands of synthesized compounds for pesttctdal acttvtty. This has led to the development of many new compounds with excellent activity at a low cost to farmers Bropesttctdes, on the other hand, are based on btologlcal relationshrps, with mtttal tdenttficatton of a new class of actlves, typically resulting from tsolatton of a pathogen from the target pest or tts envtronment. Once a new active ingredient IS rdenttfied, then the development effort becomes very similar to the chemical approach, m that random screenings agamst insects are mtttated. However, even then, the screening 1s generally more targeted than for a chemtcal screenmg program Thus, the base concept of beginning with biological relattonshtps, such as pathogenests or antagonism, as a strategy for discovermg new acttves, means that the agents identtfted will generally be much more selective m their actton.

North American

Scenario

15

Issues of environmental contammatton, worker and food safety, and nontarget effects from the use of chemicals are all derived from then nontarget actrvrty, and it 1sthese problems that are driving the development of alternatives such as the btopesttcldes. If biopesticides provide a safer, more selective form of insect control, then why do they account for only a fraction of the total pesticide market m North America and elsewhere’ Several factors limit the size and growth rate of btopestictdes Btopesttcidesgenerally lack the broad-spectrum activity, speedof control, residual life, and low cost of their chemical counterparts In the context of theseperceived limitations, biopesticides also tend to be more dtfficult to use, have shorter shelf lives, and tend to be more costly than traditional chemicals As a result of these challenges, penetration mto major insectictde markets has been limited There are at least two factors that will drive the growth of btopestrctdes. First, as society increasingly looks at the total value of a particular technology for pest control, this means lookmg beyond acute control and assessinghow a pest control product or group of products fits into the overall crop ecosystem, Another factor that will drive the growth of btopesttctdes 1sthe enthusiasm that 1scentered around biotechnology and the realization that the tools are available to improve effectiveness of btopesttcides while retammg then qualities of low environmental impact. 3. Use of Genetic Engineering to Enhance Biopesticides Has Become a Dominant Trend Begmnmg m the 1980s and to the present, a variety of molecular approaches has been used to improve market acceptance of bropestrcides. Early on, most of these efforts were directed at improving microbial msecticides, such as Bt, which has been in commercial use for over 40 yr (4). However, its use has been largely restricted to niche markets. Specific charactertstics that limit the wider use of Bt include hmtted host-range specificrty, inability to target cryptic feedmg pests, slow action compared to chemical msecticides, and lack of residual activity (4). Two general approaches were used to improve these charactertstics. One approach has been to transfer Bt toxins (genes) mto alternate mtcrobtal hosts, to address the issues associated with insect feeding or residual actrvtty. In 199 1, Mycogen (San Diego, CA) received EPA registration for the first two genetically engineered mtcrobial bioinsecticides, MVP@ and M-Trak@. These products consisted of Bt toxins that were produced and encapsulated m the host bacterium, Pseudomonas jluorescens. Recently, Ecogen has developed methods for homologous recombinatton with Bt This technology was used m the recent development of CryMax@ bioinsectrctde. These engineered microbial msecttcides do address such issues as spectrum of activity, residual hfe, and cost of product.

16

Caulder

4. Plant Expression of Biologically Derived Toxins Will Allow Penetration of Biopesticides Into Key Markets In North America, the largest msectlclde markets are associated with the control of insect pests m corn and cotton. The difficulty of blopestlcldes penetratmg these markets has to do with the biology of the primary insect pests.In corn, the primary insect pest 1s the corn rootworm complex, which IS comprised of several species wlthm the genera Diabrotzca. These larvae hve m the so11and cause economtc damage by feedmg on corn roots. Because of the solldwelling habltat, delivery of a mlcroblal product mto the feeding zone has been impractical or cost prohibitive. In cotton, the main pest 1sthe Heliothine complex. In North America, this complex consists of the tobacco budworm and bollworm. These larvae feed on parts of the plant, such as terminals and flowers, that rapidly outgrow msecticlde sprays. Larger larvae burrow mto bolls and become inaccessible to sprayed msectlctdes. For these insects, expression of the Bt toxin m the plant tissue may be the optlmal solution. Thts approach provides the ideal solution to the pesticide delivery issue and lack of residual activity The commercial impact of plants expressing Bt is just now bemg felt. 1996 marked the first sales of msect-reslstant transgemc corn by Mycogen and Cuba (Novartls), and insect resistant cotton and potatoes by Delta and Pme Land and Monsanto 5. Plant Expression Will Enhance the importance of Bt as a Source of Insect-Resistant Proteins The first insect-resistant transgemc crops entering the marketplace contam CryIA Bt toxms, which are active against a variety of lepldoptera Based on the initial demand for these crops, there IS every indication that the marketplace will demand crops that are resistant to other major insect pests. To satisfy this demand, additional insect active proteins will have to be discovered. Although blologlcally active proteins can and will be identified from many sources, it 1s likely that Bt will continue to serve as the primary source of insect active proteins for the near future. Strong current demand for Bt toxins is, in part, a result of their reputation for safety and efficacy. This demand will be met by the incredible diversity of toxin proteins that 1sproduced by Bt The pace at which new Bt proteins are being discovered 1s Illustrated by the fact that, m 1989, 14 toxin proteins were discussed m the classic Hofte and Whitely review (5); the latest Bt review, refers to over 100 Bt proteins (6). The discovery of this vast array of Bt proteins has important lmpllcatlons. One 1sthe ability to manage or delay resistance development. Resistance has not been a major problem with Bt msectlcldes, probably because of the limited use of these products Exceptions do occur when Bt usage IS particularly intense. A good example IS the use of Bt m cole crops to control diamondback

North American

Scenario

17

moth. In tropical or semitropical areas, where growing seasonsare long, diamondback moth populations have developed resistance to CryIA toxins. Resistant populations have been isolated from insects m Florida, Hawaii, the Phihppmes, and Southeast Asia. Even though these insects are resistant to CryIA toxins, they still are susceptible to other Bt toxins, such as CryIC. Given the structural dtversity of Bt toxins, the author is optimrstic that additional protein families with no cross resistance ~111be identified. With an adequate number of noncross-resistant toxins, rotation and other use schemes can be developed to delay the onset of resistance. The challenge facing both industry and academia will be to reach an agreement over which management schemes are most appropriate. As more chemically diverse &endotoxms are discovered, the number of insects that are susceptible to these new Bt toxins is expanding. A case in point is the identtfication of Bt toxins that are active against black cutworm Black cutworm, Agrotis zpszlon, has traditionally been difficult to control with Bt toxins. Recently, the activity of two distmct Bt toxins has been repoited agamst cutworm (7,s). Similar progress is being made agamst recalcitrant Coleoptera, such as the corn rootworm (9). One can envision the impact that controllmg these two soil pests of corn ~111have on the overall msectrcide market, since over $300 milhon IS spent on chemical pesticides to control corn soil pests 6. Field Performance: Examples and Role in IPM Implementation Another key problem associated with the use of broad-spectrum chemicals has been then- impact on the crop ecosystem. Dependence on these materials has often created an unstable situation in agricultural pest management. Insecticide resistance, combmed with lethal effects on natural enemies, has resulted in pest resurgence and secondary pest outbreaks. The concept of integrated pest management (IPM) was developed in the 1950s and 1960s to address these agronomic problems, as resistance to DDT and organophosphate msecttcides began to cause crop failures m cotton and other crops m a number of countries IPM is an ecological approach to pest control that requires selective agents that are effective on target-pest species, yet preserves natural enemies and retains their contribution to overall control. Btopesticides are the most selective of currently available pest-control agents. As we commercialize a greater diversity and number of blopesticrde products, we Increase our potential to develop effective and comprehensive ecologically based IPM programs that will provide farmers with the greatest value, efficacy, and sustainability. Biopesticides have delivered great benefits to growers when used m IPM systems in a variety of crops. In fresh market tomatoes in Caltforma, for example, Trumble et al. (10) compared total input costs and yields achieved, under a standard conventional treatment regimen that used 7-8 applications of

18

Caulder

methomyl and permethrm, to an IPM program that used 3-4 applications of a Bt product. In these trtals, conducted m 1992 and 1993, the IPM program provided net profits equal to or better than those obtained with the chemical standard treatment. In replicated commerctal trials conducted in three valleys of Smaloa, Mexico, m processed tomatoes, Trumble and Alvarado-Rodriguez (21) evaluated an IPM program based on regular sampling and action thresholds The IPM program used a Bt product, a mating disruption pheromone product, abamectm, and releases of egg parasites. The IPM program was compared to a conventional program using chemical msecticldes applied 35 times m what was designed to be the best approximation to conventional grower practices. The results of these season-long trials showed that m the autumn planting the IPM program yielded net per-acre profits that were $304-579 higher than the conventional program. During the winter and spring plantings, only the IPM program was profitable Subsequent to these trtals, Alvarado-Rodriguez began implementmg this IPM program on a number of farms m Smaloa, Mexico, and m Florida, Honduras, and Nicaragua. On these farms, sigmficant yield increases have also been realized, compared to conventional practices. IPM programs like this one are showing that biopesticides can deliver excellent value, not only m addressing environmental and safety concerns, but m providing farmers with the agronomic and economic benefits of lowered input costs and higher yields. Thus, one can see that the bioconversion of agriculture IS being driven by a fundamental shift, not Just m the types of agents used, with a strong movement to biopesttcides and pest-resistant transgemc plants, but also m the systems m which these innovattve biotechnology products are being employed. As mentioned, the biopesticide industry has completed its first season sellmg Bt corn and one season sellmg Bt cotton. Bt corn, which expresses toxins to control European corn borer, has performed quite well. Likewise, growers using Bt cotton needed fewer msecticide applications than growers that planted conventional cotton. However, during the 1996 growing season, there was some concern over the efficacy of Bt cotton. The problem was greater than normal bollworm populations. The reduced efficacy of Bt cotton against high populations of bollworm was demonstrated prior to the commercial launch (12). Nonetheless, the need for additional insecticide use took many people (farmers, researchers, and the press) by surprise. This situation illustrates the importance of an IPM mmdset as opposed to the mmdset that accompanied the mtroduction of synthettc msecticides in the 1940s and 195Os,when application of an msecticide was thought to be the complete solution to a pest problem.

North American

Scenario

19

7. The Future Looking forward, tt seemsinevitable that the transition to btologtcally based pest control practices wtll contmue. A major driving force impactmg agriculture IS the exponential rate of population growth that will produce a population of over 6 billion people by 2000. The impact that this growth will have on entomology and agriculture IS multifold (13). As the populatton increases by 1 billion every 10 yr, there will be extraordinary demands on our ability to produce food and fiber Most of the population growth is taking place m less developed countrtes. Although populations are more stable m the more developed countrtes, acreage devoted to agriculture IS decreasing. The challenge to both private and pubhc sector researchers must be to find addmonal ways to Increase agricultural productivtty m the face of an increasing populatton and decreasing acreage. The rate of population growth necessitates mcreasmgly efficiency m moving research ideas into the marketplace. An important trend IS the formatton of consortta and alliances between research efforts m public and private sectors, These alliances allow groups to focus on key research goals and make progress in a timely manner This trend must continue. Pesticide safety has been an important consideration for many years and will continue to drive btoinsecttcide growth. People have a reasonable concern about pesticides in then diet and environment. The issue of environmental quality 1swell established m more-developed countries. Stated simply, people do not want synthetic chemical residues m their foods. This issue 1s as much a perception as it IS a true toxtcology issue. As the population density increases, environmental quality ISlikely to become a more important issue m more countries, regardless of their state of economic development. Absent significant changes m current practices, nontarget effects of pesticides have an mcreasingly significant impact on environmental quality. One consequence of the increasing populatton and urbanization of farmlands is the increased contact between people and agricultural production. As this increase contmues, the safety of synthetic insecticides will become a greater issue. Biopestlcides effectively address the issues of environmental quality and safety. Since the active ingredients m biopesttcides exist in the natural environment, there 1sa greater level of comfort or a greater perception of safety. As progress m brotechnology and the discovery of new Bt toxins continue, people can look forward to increased use of biopestrcides and btologlcal toxins expressed m transgemc plants. People will have to ask how best to use these tools For example, should biopestrcides be deployed as synthetic msectrcrdes? Are pest-resistant transgemc plants an example of scheduled spraying taken to an extreme position? Alternatively, should these tools be used as part of an overall management strategy that includes parasltolds, predators, and pathogens, as outlined m the early IPM concepts proposed by Smith and van den

20

Caulder

Bosch (14) and others. The selecttvtty of bropestrcide and pest-resistant transgenie plants are complimentary for preservation of predatory msects.Moreover, the sublethal effects of biopesticides may actually mamtam predator populations (15). The integration of transgenic plants mto IPM is also an important Issue. On the surface, the scheduled spraying model for both bropesticides and transgemc plants may seem lakea good Idea. However, there are undesirable consequences, such as resistance development. Would it be better to use these plants m the context of IPM, in which the selectron pressures of a toxin expressed m plants are made less intense? These biopestrcide and transgemcs plants will play an important role. Their rmplementatton will require the coordmated actions of manufacturers, academics, growers, and private consultants.

References I Madehn, M. F. (1963) Diseasescaused by hyphomycetous fungi, tn Insect Pathology, vol 2 (Stemhaus, E A , ed ), Academic, New York, pp 233-272 2. Luthy, P , Jaquet, F., Huber-Lukac, H. E , and Huber-Lukac, M. (1982) Physiology of the delta-endotoxm of Bacrflus thurmgzenszs mcludmg the ultrastructure and histopathologrcal studies,in Basic Bzology of Mlcroblal Larvzczdes of Vectors of Human Dzseases (Michal, F , ed ), UNDP/World Bank/WHO, Geneva, Switzerland, pp 29-36 3 Goldberg, L H and Margaht, J (1977) A bacterral spore demonstratmg rapid larvrcrdal activity against Anopheles sergentu, Uranotaenla unguzculata, Culex unlvattatus, Aedes aegyptz and Culexprplens Mosq News 37,355-358. 4. Gelernter, W. and Schwab, G E (I 993) Transgemc bacteria, vu-uses, algae and other microorganisms as Bacillus thurznglensls toxin delivery systems, in Bacillus thurmgrenszs, An Environmental Blopesttclde Theory and Practice (Entwistle, P F ,

Cory, J. S.,Bailey,M. J., andHrggs,S , eds), Wiley, Chrchester,UK, pp. 89-104 5 Hofte, J and Whiteley, E B. (1989) Insectrcidal crystal protems of Bacillus thurmglensu. Microblol Rev 53,242-255 6 Schnepf, E , Crickmore, N , Van Rie, J , Lereclus, R , Baum, J , Feitelson, J , Zergler, D. R , and Dean, D. H (1997), manuscript m preparation. 7. Estruch, J J , Warren, G. W , Mullms, M A , Nye, G J , Craig, J. A , and Kozrel, M G (1996) V1p3A, a novel Bacillus thurmgiensis vegetative msecticidal protein with a wide spectrum of actrvities against lepidopteran msects Proc Nat1 Acad Scl 93,5389-5394

8 Lambert, B., Buysse,L., Decock,C , Jansens,S , Piens,C , Saey,B., et al (1996) A Baccllus thurlngzensls msectrcrdal crystal protem wrth a hrgh activity against members of the family Noctmdae. Appl Environ Microbzol 62, 8&86 9 Fertelson, J. S , Payne, J., and Kim, L (1996) Baczllus thurzngzensls insectsand beyond. Bzo/Technology lo,27 1-275. 10 Trumble, J T , Carson, W G., and White, K K (1994) Economic analysis of a Baczllus thurzngzensu-based Integrated pest-management program m fresh-market tomatoes. J Econ Entomol 87, 1463-1469

North Amencan

Scenario

21

11 Trumble, J T and Alvarado-Rodriguez, B (1993) Development and economtc evaluation of an IPM program for fresh market tomato productlon m Mexico Agriculture Ecosystems Environ 43,267-284 12 Mahaffey, J. S , Bacheler, J S., Bradley, J. R., and Van Duyn, J W (1994) Per-

formance of Monstanto’s transgemc B t. cotton agamst hrgh populations of Lepropterous pests m North Carolma, m Proceedmgs Beltwzde Cotton Conferences, pp 1061-1063. 13 Metcalf, R L (1996) Applted entomology m the twenty-first century Am Entomol 42,2 16-227 14 Smtth, R F and van den Bosch, R (1967) Integrated control, m Pest Control Blologlcal, Physical and Selected Chemzcal Methods (Kllgore, W. W and Doutt,

R L , eds ), Academx, New York, pp. 295-340 1.5 Soares, G G., Lewts, W J , Strong-Gunderson, J. M., Waters, D. J , and Hamm, J J.

(1993) Integratmg the use of MVPB bromsectxlde, a umque Bt-based product, with natural enemies of Noctmd pests. a novel concept m cotton IPM, m Proceedzngs of the 2nd Canberra Bacdlus thurmgzenszs Meeting, pp 133-145

3 Microbial

Biopesticides

The EuropeanScene Tariq M. Butt, John G. Harris, and Keith A. Powell 1. Introduction There 1s growmg interest in the explottatton of naturally occurrmg mtcroorganisms for the control of crop pests, weeds, and diseases.Btologtcal control agents (BCAs) may offer more environmentally friendly alternatives to chemical pesticides.They could also be used where pestshave developed resistanceto conventional pesticides.Unfortunately, there 1scomparattvely httle investment m the research and development of these organisms compared with that spent on the discovery of chemical pesttctdes.Two reasonsfor this are that microbial pesticides usually have a narrow host range, and that they often give mconststentand poor control in field trials. Consequently, more attentton 1sbeing given to the selectton of broad-spectrum blopesttcidesand improvements m production, formulation, and apphcatton technologtes. Efforts are also being made to opttmtze the impact of these agentsby mtegrating them with other novel crop protection strategies(I). One factor that cannot be ignored is the market potential of btopestictdes.Currently, only speciahzed,niche markets exist. Their full potential has not been realized because of the absenceof strong mcentrves to develop these agents and/or discourage chemical pesticides;availabrlity of new, biodegradable chemical pestlcrdes; absence/breakdownof the mfrastructure that facilitates transfer of new technologles and research knowledge to the end user (i.e., grower); absence of a universally acceptable regrstratron procedure; restrtcttons in the use of exottc BCAs; and lack of robust and reliable field effects Progress is also slow because the chtef producers are often small-medium-size enterprises(SMEs) that have hmtted resources for effective development and marketing of products. From Edited

Methods m B~olechnology, vol 5 Blopeshodes by F R Hall and J J Menn 0 Humana Press

23

Use and Dehvery Inc , Totowa, NJ

Butt, Harris,

24

and Powell

Fungicide 20% Herbickie 49%

hectic

Fig. 1. The share of crop protection in each of the major sectors of the world market.

Europe

Fungicide 30% Herbii5e 48%

Insecticide 22%

Fig. 2. The European Market share for fungicides, insecticides, and herbicides.

Figure 1 showsthe current shareof the agrochemicalmarket worldwide for insect, fungal, and weed control. For convenience, insecticides also include nematicides and

molluscicides.The Europeanmarkethasa different shape,with plant pathogensbeing more important and insectcontrol being a more minor component (Fig. 2). Clearly, from thesefigures, one might expectthe development of biological control for plant pathogensto have beencenteredin Europe. There is, however, little evidence of this. The useofBacillus thuringiensis (Bt) for control of insectshasbeendeveloped mostly in the United Statesand Canada.It is in the key cotton crop that B&derived genes have beenused to provide crop resistanceto insects. A wide range of BCAs have or arebeing developed ascommercial biopesticides produced in Europe, often with global markets in mind. In order to survive, many SMEs market products of other companies or produce BCAs under license. Presumably, this mutualism will decline as the use of BCAs increases(i.e., the market expands) and it becomes more lucrative for individual companies to develop their own agents. 2. Formulation and Delivery For any crop-protection agent, an efficient formulation is a necessityto translate laboratory activity into adequate field performance. The formulation must

Microbial

Biopesticides:

European

Scene

25

be mtrmsrcally compattble with the BCA, and, ideally, the formulated maternal should have superior performance compared wtth the unformulated material (2) For BCAs, there are particular challenges to be faced, because the active ingredient 1sfrequently a living organism that must be kept relatively tmmobile and inactive while in storage, but quickly resume its normal metabohc processes once applied to the target site To achieve this, some form of drying process ts usually done, such as an-drying, freeze-drying, or lyophihzation. A preservative, such as Proxel (1,2-benzisothtazohn-3-one), will often be mcorporated to prevent microbial contammatton Subsequent formulation IS then usually as a wettable powder, water-dispersible granule, or dust. Another approach is suspension in oil, m which the purpose IS to exclude oxygen from the organism, thus preventing respiration. An example is the Bt subspp kurstuki insecticide formulation Dipel ESNT, in which the actrve ingredient is encapsulated and then suspended in an 011base Moore et al. (3) found that dried comdra stored m 011formulatrons remamed viable longer than those stored as a dried powder, especially If stored at relatively low temperatures (IO-14 vs 28-32OC). Additton of sihca gel to oil-formulated conidia prolongs then shelf life. Undried conrdra of Metarhzzzum jluvovinde, without sthca gel, lose viability rapidly, with germmatron droppmg below 40% after 9 and 32 wk at 17 and 8°C respecttvely. After 127 wk in storage, germination remained at over 60 and 80% for the drred formulations at 17 and 8°C respecttvely (4). These comdra were found to have retained full vtrulence, compared with freshly prepared formulattons. Furthermore, conidta dried to 4-5% morsture content showed greater temperature tolerance than conidia with higher motsture content. McClatchie et al. (5) report that htghtemperature treatments caused delay in germmatton, as well as death of A4 flavovzrzde comdia. However, drymg comdta by adding sthca gel to 011 formulatrons greatly increased temperature tolerance Desprte thus, formulatrons composed of living cells still suffer sigmficant degradatron over trme, and this IS a problem that still needs to be solved. This IS a particular issue for the nonspore-forming bacterta and fungt The situation is rather easier wrth spore-forming organisms. Restmg stages, such as spores, are desrgned to retam water, be robust, and survive m a viable state, even when subjected to harsh envtronments. Nucleopolyhedrovrruses also produce propagules that are encapsulated with polyhedrin, giving them a tough, reststant coatmg that facthtates survival These types of organisms are, therefore, more easily formulated, and ltqurd products, such as suspension concentrates, are quite feastble. Equally important, the formulation type and packaging materials must be broadly similar to those wtth which the grower 1s already famthar. The products should be capable of application through the standard hydraulrc sprayer or

26

Butt, Hams, and Powell

applicatton equipment that 1scommon to a partrcular market, and have as few unique requirements as posstble A grower IS unhkely to mvest m new spray equtpment solely to treat a BCA, nor 1s he going to accept a very different spray regime or more frequent apphcattons than 1snormal practice. The grower will also want to purchase his BCAs through the same dtstributton chain as hts agrochemtcals. Yet, a distributor is not going to be happy to have to handle BCA formulations differently from his normal chemical stock. The dtstributor will expect the product to be packaged m standard sizesand types of contamers, as used throughout the agrochemtcals mdustry Storage stabthty must be such that product purchased at the start of one seasonIS good for the whole of that seasonand the next, without any special storage requirements. If the shelf hfe of a BCA forrnulatton ISvery limited, then a distributor may only be prepared to buy small quanttttes,thus hmitmg avatlabthty, or will only stockproduct on a consignment basts,which is a major mconvemenceto the BCA supplier Also, some BCAs have a specific need for refrigeratton, but very few dtstrtbutors m Europe have such facthttes, and even fewer would be prepared to invest m them. Finally, the BCA formulation must be composed of safe materials One of the mam features of BCAs is their perceived safety benefits to the envtronment, beneficials, nontargets, and, of course, the applicators. These benefits would be negated tf toxic formulatton components were used m the product. If the product 1stargeted at a fohar pest or disease, then addtttonal problems are subsequently encountered. All the usual requtrements for successful fohar treatment need to be met, mcludmg good sprayabtlity, no nozzle blockage, good leaf deposttton and dtstributton, and adequate rainfastness. In order to achieve these attributes, surfactants, stickers, and wetters must be mcorporated mto the formulation or applied m tank mixture. One of the chief causes of macttvation of many BCAs on the leaf surface 1sthe effect of ultravtolet (UV) radiation. High levels of UV can lead to rapid degradation of the material, perhaps wtthm a single day. Mtcroencapsulatton of the BCA has been one approach to overcome this problem, although this IS technically difficult and quite expensive to manufacture. Another technique that has been used, particularly with baculovnuses, is to incorporate mto the formulation or tank mtx an optical brightener. These brighteners may reflect UV or reduce its impact by disstpatmg the energy as fluorescent light and heat. But they can only protect the BCA when they are m the immediate proximity of the organism Because this is hard to achieve when simply admtxed mto the formulation, results with this approach have been highly variable. To exploit this technology effectively, ways have to be found to keep the brightener or other UV protectant materials m mtimate contact with the BCA. Also reported are the usage of otlbased formulations, which have proved to be beneficial to UV protection of comdta of M flavovmde (4,6,7)

Microbial

Biopesticides:

European

Scene

27

In the case of mrcrobtal msectlcides, good drstrtbutron over the leaf 1sessential, because they are nonsystemtc materials, and must come m contact with, or be consumed by, the target insect, m order to deliver a toxrc dose Feedmg attractants incorporated mto the formuiatron may be useful to encourage Insects to feed on the BCA (8), and Bt products are frequently tank-mixed wrth 2 kg/ha of sugar for fruit and vme pests in Europe, with good reported effect In other cases,some very specrfic additives can have a positive effect For example, the LD,, for some formulatrons of Beauveria basslana was reduced by 97% by the addrtron of coconut 011.It was suggested that the cutmophtlic propertres of the or1 could allow a greater number of fungal comdta to penetrate the mouth parts of the insect (9) 011 carriers can also distribute the moculum over the insect cutrcle, often carrying the comdta to the thm mtersegmental membranes, which are more readily penetrated by entomogenous fungr (Butt, personal observation). A different series of prtortttes come mto play when devrsmg formulattons for the control of soilborne pests and diseases Placement of the BCA IS of prime importance, to ensure that dlstrtbution through the soil 1s even, and, therefore, there IS a good chance that the BCA and the pest or pathogen will come in contact with one another. Preplantmg, BCAs can be applred as granules m-furrow, and by drench apphcatton On a small scale, such as m glasshouse sttuattons, material can be thoroughly incorporated Into the so11 Postplantmg, a drench applicatton or irrrgatron can be employed to take the BCA mto the son. Seed treatment may also be an option, although most seed tend to recetve a chemtcal fungrcide to control seedborne drseases, and these treatments can often be antagomstrc to BCAs. 3. Biological Control of Plant Pathogens Biological control of plant pathogenic fungi has long been an ambttton for academic and industrial researchers. However, there IS little evidence of major breakthroughs m the marketplace. The issues and problems have been dealt with m many publlcattons (IO). There are products that are avatlable for both soilborne and foltar pathogens, as well as considerable research on postharvest disease(11,12). Several products are or have been registered m Europe (Table 1) Current products are shown in Table 2. Although the products listed m Table 2 are commercially avarlable, the total sales of btological agents for crop protectron agamst plant pathogens amounts to much less than 1% of the fungrcrde market m Europe. The problem faced by developers of brologrcal agents for control of disease are complex and drfficult. Crops are grown under a variety of chmattc and environmental conditions, and temperature, rainfall, sot1type, crop variety, and

2

Fmland, Sweden, Norway

Adapted with permIssIon from ref. 33.

(= Penzophora) glgantea

Phlebropsrs

Country

for Biological

Netherlands Sweden, France United Kingdom

in Europe

Vertlc&um dahliae Truzhoderma harzlanum Trrchoderma wnde

Registered Finland, Hungary, Norway,

Agents

Streptomyces grzseovwzdzs

Agent

Table 1 Some Microbial Switzerland

Control

Target pathogen Fusarwm spp, Alternana, Pythlum, Botrytq and other so11 pathogens Agamst Dutch elm disease Sell-borne fungal diseases Antagonist to silverleaf fungus (Stereum purpureum) Fomes annosus

of Plant Diseases

Kemrra Agro Oy Kemu-a Agro Oy Gustafson Various Ecogen Ecogen Natural Plant Protectron (NW Various

GlioMix Mycostop Kodiak Vanous AQlO Aspire Fusaclean Various

Truzhoderma viride

Supplier

Control

Kemrra Agro Oy

Commercial name

Sold in Europe for Biological Rotstop

Products

Phleblopsls (= Penrophora) gigantea Ghocladmm sp Streptomyces grrseovwldls Bacillus subtills Agrobacterlum radlobacter Ampelomyces quzsqualls Candida oleophda Fusarlum oxysporum FO 47

Product

Table 2 Some Commercial

Wood treatment

Promotes plant growth, competes wrth soil microbes Fusarwm spp and other sod pathogens Soil pathogens Agrobacterium tumefaclens Powdery mildew Postharvest decay of citrus and apple Fusanum oxysporum, F momliforme

Fomes annosus

Target pathogen

of Plant Pathogens

30

Butt, Hams, and Powell

pathogen can change from farm to farm, or even within one field The producer of a crop protection product has to be able to give some assurance to the farmer that the product ~111 be robust, m order for the product to be used. The avallablllty of effective chemical controls for fohar pathogens has made it unlikely that a biological agent will compete effectively It is, therefore, not surprlsmg that the maJorrty of efforts m research has been concentrated on sotlborne or postharvest diseases Even m these situations, the lack of robustness has hmlted the penetration of such products. When success has been achteved, it has been m limlted crop-pathogen mteractlons, and has often been specific to crop, pathogen, and growmg sltuatlon The control of Agrobacterzum tumefaclens by the closely related bacterium Agrobactenum radzobacter (13) IS a classic example. The bIologIca agent IS apphed to unmfected roots of cuttings, and 1s then allowed to colomze the niche that would normally be at risk from the pathogemc species The success of this treatment may be enhanced by the productlon of a specific antlblotlc that mhlblts growth of the pathogen An lllustratlon of the problems faced m control of disease under more varied condltlons was provided by the excellent work of Defago et al. (14). This group showed that the sol1 environment played a key part m the ablhty of Pseudomonasfluorescens to control black root rot of tobacco One so11 type showed excellent potential for disease control, but a second negated the effect of the bacterium The application of nonpathogenic strains to control pathogenic fungi has been demonstrated by the work of Alabouvette et al. (15) Nonpathogemc Fusarzum oxysporum llmlted disease caused by pathogemc Fusarzum. the suppression was linked to the ratio of density of the nonpathogemc population m relation to the pathogenic species. The nonpathogemc strams compete with other mlcroorgamsms for nutrients and elicit a defense response m the plant, which protects It agamst more aggressive strains of Fusarzum Trlchoderma harzlanum and T wide have been proposed by various groups as potential BCAs. Trzchoderma sp appear to have the ablhty to rapidly colonize clean surfaces, such as compost or freshly cut timber Several products have been available, but the market share has been mmlmal. One problem may be the lack of competltlve ablhty of these species when apphed to so11or other media with an active mlcroblal population. One exception 1s GhoMlx, which IS a formulation of the fungus Glzocladzum. This product promotes plant growth and, presumably because of its rapld growth, the fungus prevents establlshment of potential disease-causing microbes. Antagonists of postharvest diseases share some of the attrlbutes of the above For example, the success of Epicoccum nlgrum, Penlcllllum oxalicum, and Candzda sake, which are bemg developed to control Monzlwzza laxa (brown rot

Microbial

Blopes ticldes:

European

Table 3 Factors That May Affect the Success

Scene

31

of BCAs for Plant Pathogens

Factors affectmg long-term survival Physlcal Soil type Matrtc potential Temperature PH Blottc Ablllty to colonize substrate Competition Survival structures Host genome Resistance of pathogen to BCA

Factors affectmg speed of effect BCA Germmatlon and recovery time Release from formulation Speed of growth Movement to site of actlon Pathogen Growth rate Degree of mche protectton Productlon of toxins

Adapted with permlssron from ref. IO

of peaches and other fruit), Fusarzum oxysporum fsp lycopersicz (tomato wilt), and Penzczllium spp, respectively, IS dependent upon their ability colonize the

fruit surface raptdly and displace the disease. Antagomstlc yeastsalso produce extracellular materials (mostly polysacchandes) that not only enhance their survival, but also restrict colomzation sites and the flow of germmatlon cues to other fungal propagules. Bacterial antagonists like Badus subtzh produce antlblotlcs (e.g , Iturm), which mhlblt diseases hke the brown rot pathogen Monzlznza fructzcola. The exact mechanisms of antagonism are poorly understood, but involve a complex of attributes, including nutrient competltlon, site exclusion, attachment of the antagonist to the pathogen, secretion of pathogemcity-related enzymes, induced resistance, and antlblosls caused by the actlon of bloactlve compounds Table 3 illustrates the issuesfaced when attempting to develop a BCA for so11 diseases,and shows quite clearly the need to take Into account many factors when considering the potential for a new project in this area. Given the difficulties listed above, the lack of commercial successfor blologlcal control IS perhaps not surpnsmg. It 1snevertheless possible, as demonstrated by Rlshbeth (16) m the control of Fomes anrzosus by Phlebzopszs (= Penzophora) gzgantea In this sltuatlon, the cut tree stump is inoculated with the blologlcal agent immediately after cutting. The control ISachieved becausethe substrateIS rapldly colonized by the control agent and the pathogen 1sunable to enter and thus Infect the nearby trees via the root system.This example shows clearly that specificity, competition, and growth rate are all important This lessonshould be heededby all those tempted to embark on a new project for biologtcal

control

32

Butt, Hams, and Powell

4. Biological Insecticides: Field Performance and Role in IPM Implementation 4. I. Mycoinsecticides Some of the pathogens that have been or are being developed for the control of insect pests belong to the fungal division Deuteromycotma, class Hyphomycetes (Table 4). The most common species are Metarhlzzum anzsopliae, Beauverla basslana, Paecdomyces fumosoroseus, and Vertdlwm lecarw These can be readily isolated from soils from most parts of the world, and are known to have a wide host range, but strains can doffer m then- specificity and virulence (I 7,18). Fungi, unlike bacteria and vnuses, do not have to be ingested to cause infection, but can penetrate the insect cuticle directly, using a combtnation of enzymes and mechanical (18-20). Once the pathogen has gained accessto the nutrient-rich hemocoel, the fungus may grow as thin-walled blastospores or hyphae. Most strains secrete htstolyttc enzymes and msecttctdal metabohtes. The latter can disorient the host, stop It from feeding, and cause death before mycelial colomzatton of the hemocoel 4.1.1. Verticdlium lecanll The fungus, Vertzczlliumlecanu, has been developed as a btomsectlctde for the control of aphids, whiteflies, thrtps, and red spider mites Christian Hansen’s BIO Systems were particularly acttve developmg the product for usage m Scandinavta, as MtcroGermin (21). The fungal spores are dried to form a wettable-powder formulatton, whtch can be mixed with water to produce a suspension suitable for immersion of plant cuttings. Shelf life ISclaimed to be 6 mo at 5°C. The chief market has proven to be the protection of glasshouse cuttings by dipping, prior to potting. Aphids and whttefltes can be a major problem m glasshouse plants, and, m such a closed envtronment, reststance to chemical msectictdes tends to develop quite qutckly. MtcroGermm offers an effective alternattve to chemicals m what 1s a high-value market, Potential outside of a glasshouse environment is limited, however, because the fungus reqmres a relative humidity of between 95 and 100% for at least 10-12 h to germinate effectively and colonize the insects. Late in 1995, the Dutch company, Koppert, purchased Christian Hansen’s Bio Systems. Koppert already had a similar product m then range, called Mycotal, and so it IS expected that efforts will be made to market a single product for other northern European glasshouse markets alongside then- extensive product range of predators and biocontrol agents. 4.7.2. Beauvena and Metarhizium Wlthm each genus, there are two species that have been examined for biological control potential. Beauveria bassiana and B. brongmartiz have been

Fungus

of Mycoinsecticides Pest insect Whltefly and thrlps Aphids Vine weevil Coffee berry borer Sugar cane white grub Corn borer Colorado beetle Colorado beetle Colorado beetle Cockchafer Cockchafer Cockchafer Locusts, grasshoppers Whitefly

in Europe

aReglstered, but not on sale, other products under development ‘Some commentators suggest quite large scale use (>lO,OOO ha), but this has been disputed (34). =From ref. 34 dUnder hcense from ThermoEcotek (now Therm0 Tnlogy)

Vertlcdlwn lecanu V. Iecann Metarhizlum anlsopllae Beauvena basslana Beauverla brongnlartn Beauveria basslana B. bassiana B. bassiana B basslana B. brongnlartii B. brongnlartti B brongnlartu Metarhlzlum flavovirlde Paecllomyces fumosoroseus

Scale Production

Mycotal Vertalec BIO 1020” Comdla Betel Ostrinil Bovenn Boverol Boverosil Engerlingspllz Schweizer-Beauvena Melocont Green Muscle PreFeRal

Product

Table 4 Commercial

Koppert, Holland Koppert, Holland Bayer, Germany AgrEvo, Germany NPP (Calliope), France NPP (Calliope), France Former USSRb Czech Republic/Slovaklac Czech RepublicYSlovakla” Andermatt, Switzerland Eric Scwelzer, Switzerland Kwlzda, Austria CABI, UK Biobest, Belglumd

Producer

34

Butt, Harris,

and Powell

the focus in this genus, and these species are known to produce a cyclodepsipeptide called beauvericin, which is toxic to insects. Similarly, Metarhizium anisopliae and M. flavoviride produce related insecticidal metabolites called destruxins. These toxins play an important role in the pathogenicity of the fungus, and probably help to disable the insect’s self-defense mechanisms while the fungus invades and colonizes the hemocoel. In Switzerland, a biological method to control the subterranean pasture pest, Melolontha melolontha, with the insect-pathogenic fungus B. brongniartii was developed during the 1980sand the pathogen was registered in 1990. Since then, approx 5000 kg are applied per annum to 100-I 50 ha of mostly pasture. Another strain of this fungus has been developed by Natural Plant Protection (NPP; France), and is sold under the commercial name Betel for the control of the sugar cane white grub (Hoplochelus marginalis, Melolonthinae) in the tropics. NPP also developed a strain of B. bassiana for the control of European corn borer, Ostrinia nubilalis (Table 4). Both pathogens are produced using solid fermentation technology. The pathogenis formulated in clay granules and applied at 25 kg/ha (Ostrinil) or 50 kg/ha (Betel). According to the manufacturers, the granules are applied in the sameway as chemical granules. Both products are stable for 1 mo at 35”C, but the shelf life is increased at lower temperatures. In Germany, Bayer AG developed BIO 1020, a strain of M. anisopliae for the control of Otiorhynchus sulcatus (black vine weevil) (22). This pest is a major problem on several ornamental crops in glasshouses and nursery stock. The product consists of dry granules with a claimed shelf life of 6 mo if kept in cool conditions. The granules are admixed with soil, and, following water uptake by the granules, the fungus produces large numbers of conidia that are viable for many months. Insects become infected when they come into contact with these conidia. This does mean, however, that the application must be of a preventative nature, but, because of the long survivorship of the conidia, satisfactory protection should be achieved for the whole duration of the crop. The recommended rate of the formulated product is 1 g/L soil, at which good levels of control of 0. sulcatus have been seenon a number of glasshousecrops (Fig. 3). Such a product has an excellent fit with other predators, parasites, and microbial agents for glasshouse crop protection. Recently, the use of M anisopliae has been reported for the control of western flower thrips, Frankliniella occidentalis (23). Strains of this fungus have also been shown to be highly pathogenic to several crucifer pests, yet is harmless to honey bees (17). In both instances, conidia were formulated in either aqueous solutions or oil, and were applied to leaf surfaces using conventional and/or electrostatic sprayers. Poor control can be attributed mostly to low temperatures and humidities, which will prevent spore germination and infection. In addition, UV light will

Microbial Biopesticides: European Scene

Azalea

I

Fuchsia

Chrysanthemum

I

Cyclamen

Begonia

Fig. 3. Effectiveness of BIO 1020 against Otiorhynchus sulcatus eggs and larvae in ornamentals under glasshouse conditions 28 d after treatment. (Adapted with permission from ref. 22.)

quickly inactivate spores, and leaf expansion and rain will dilute the inoculum on the leaf surface. Overcoming these factors will greatly improve fungal efficacy. Equally important is maximizing spore viability and ensuring better contact of the inoculum with the insect surface, because mortality is dose-related. The speed of kill (LT,,) is also dose-dependent (17,241. The more inoculum contacting the pest, the shorter the time to death. Selection of virulent, ecologically competent strains, combined with improved formulations and more effective targeting of the pathogen, will lead to more efficacious pest control. At IACR-Rothamsted, a push-pull strategy, based on the use of semiochemicals, is being developed in which pests are encouraged into trap crops or discard areas, where they are inundated with fungal pathogens. When developed, this strategy should greatly reduce the use of chemical pesticides (I). 4.2. Bacteria Only two species of bacteria have been developed for control of insect pests: Bt and B. sphaericus (Table 5). Products based on the different strains of Bt are the most widely used biological control agents in Europe. Bt subspp kurstaki (Btk) is used to control lepidopterous pests in vegetables, tomatoes, top fruit, vines, olives, and forestry. An example of the latter, in a major Polish forestry project, is described below.

Table 5 Insect-Pathogenic Active ingredient Bt serotype 3 Bt subsp azzawaz Bt subsp kurstakz Delta endotoxm of Bt subsp kurstakz Bt subsp tenebrionzs Bt subsp zsraelenszs Baczllus sphaerzcus

Bacteria

and Toxins

Registered

Registered uses (pest/crop) Wide range of crops and orchards Lepidopteran larvae Lymantria,

Ostrznza

Lepidopteran pests Coleopteran larvae Dipteran pests Mosquito larvae

in Europe Country

France France, Germany France, Germany, Hungary, Italy Austria Austria, Germany, Hungary Finland, France, Germany, Hungary Italy, Netherlands, Sweden France

Adapted with permtsslon from ref. 33

Poland IS a major softwood producer. Forests cover about 8.8 mullion ha, approx 30% of the land mass, and are a major earner of foreign currency. Protecting this resource is clearly a very high priority, especially because there are at least five sertous lepidopteran pests that can cause significant losses. Nun

moth (Lymantrza monacha) IS the most serious of these pests m Polish comferous forests. The larvae are major defoliators, feeding particularly off young leaves. Infestations are usually cyclical, with serious outbreaks occurring about every 50-60 yr. Unexpectedly, though, in the spring of 1994, over 600,000 ha of Polish forestry became seriously infested. This had not been predicted, since the last serious outbreak had only been 12 yr earlier. Polish scientistswere of the opmlon that the usage of broad-spectrum chemical insecticides to control the previous outbreak had made the forest more vulnerable to attack (25). It was postulated that the msecticide treatments had also elrmmated key parasites and predators of the nun moth, thus enabling the pest to increase m numbers more qutckly, and effectively shortenmg the outbreak cycle. A deciston was taken to spray the forests to control the latest nun moth

outbreak, but this time using more selective insecticides. An emergency program was put together in May 1994, with funding from the World Bank, European Union’s FAIR program, the Danish Environmental Protection Agency, and the British Know-how fund, totaling m excessof $9 mrlhon. Together with funding provided by the Polish government’s National Fund for Environmental Protectron, the total project cost came to more than $22 million (26). A special forestry formulation of Btk called Foray 48B, from Novo Nordisk accounts for almost 25% of the total insecticide used m forest systems. This

M/crob,al Biopestm-ies. European Scene

37

material is suitable for spraying from aircraft or helicopters at volumes as low as 4 L/ha when applied through ultra-low volume (ULV) spray equipment. As part of the project, Micronair ULV equipment was provided, to completely re-equip the Polish aerial applicator fleet, and pilot training was provided to enable the best results to be obtained. Usage of Btk was especially concentrated m environmentally sensitive areas, or where forest sorls dramed into rivers or lakes. Brological and economic impact studies were put mto place, to examme the implications of the spray program. The results were excellent, with Btk providmg about 95% control of the pest on average, while having minimal impact on beneticials and nontarget organtsms. Smaller, follow-up spray programs m 1995 and 1996 have effectively finished the job, and one would hope that the pest cycle in Poland has now been returned to its more natural 50-yr pattern. The nun moth causes similar problems m adjacent countries, m particular, Germany, Belarus, Ukraine, the Baltic Republics, and the Czech Republic. The Polish project has acted as a model for these other countries, and similar programs utrhzmg Btk are now m place, for example, m Belarus. For some years now, products based on Btk have become the first choice for lepidopteran pest control m North American forestry, especially for gypsy moth (Lymantria &spar) outbreaks. This trend is quickly spreading to Europe, where Btk IS mcreasmgly perceived as a highly effective and environmentally benign forestry msecticide Bt subspp zsraelenszs (Bti) is a quite different type of microbial msecticide. This strain is active against the larval stagesof Diptera, m particular, mosquito larvae. Mosqmtoes are, by nature, only regarded as pests when they are m the vicmity of people. Traditionally, mosquitoes have tended to be controlled by spraying or fogging the adults with broad-spectrum chemical insecticides, but this also exposes the human population to chemical residues of these pesticides. In these more environmentally conscious times, alternatives are being sought to reduce exposure risk. Although habitat alteration, such as marshland draining, has hrstorically been one of the best and most long-lived techniques for mosquito control, the opportumties to adopt these techniques are rapidly drmmishmg. Indeed, within Europe, there is a trend to protect and increase wetland areas, because they are regarded as being extremely valuable to wrldltfe, especially birds, but at the same time they present mosquito species with new opportunmes for colomzation and breeding (27). Control efforts have mcreasmgly moved toward tackling the problem at source, by applying larvrcrde to the waterbody utmzed by the mosqurtoes for breeding. However, because many breeding sites are m environmentally sensi-

38

Butt, Harris, and Powell

Fig. 4. Bti Europeanmarket, 1992-1993. tive areas, chemical larvicides (especially organophosphates) are inappropriate. A viable alternative has proven to be mosquito larvicides based on Bti. These products, such as VectoBac or Bactimos, contain a mixture of dipteran-active crystal proteins and spores. Various formulations are available, including liquids, wettable powders, granules, and slow-release briquettes. When applied to the water surface, the mosquito larvae filter out the Bti particles (together with other particulate organic matter), acquiring a toxic dose. Younger instars are more susceptible that older instars, and pupae and adults are not affected. Thorough, regular, and accurate scouting of the breeding site is therefore essential, to time the larvicide application correctly to target predominantly young instar stages. European countries have taken a number of different approaches to mosquito control and employ Bti products to varying degrees (Fig. 4). In France, mosquito control is predominantly in tourist areas, especially along the coast of southern France. Here, more that 80% of the Bti used is in the form of locally made sand granules. These are produced by mixing Bti primary powder (active ingredient) or wettable-powder formulations with high-grade sand, plus a small quantity of vegetable oil. Specially converted helicopters apply these granules, especially to forested or marsh areas, for the control of Aedes caspius. In Germany, a major mosquito control program is conducted in the Upper Rhine Valley, where the river banks tend to become flooded, especially in the

Microbial

Biopesticides:

European

Scene

39

spring, as the river rises because meltwater from Switzerland swells the volumes carried. In the valley itself, the major problem IS Aedes vexans, whtch tends to lay eggs on dry ground that wdl subsequently become temporarrly flooded. The primary control method IS using Bti larvtctde. Scouting of the breeding sites allows targeted applications to be made either by knapsack sprayer or helicopter-applied granular treatments over larger water bodies and in thick woodland. Another problem faced by this team IS that 50% of the houses m the valley villages collect rainwater in barrels for home/garden usage, and A vexans and Culex pzplens are found to proliferate in these To combat these pests, Btt tablets (Culmex) are manufactured locally and distributed freeof-charge to the householders to treat these contamers (28) (N Becker, personal communicatton) Hungary has a large number of government-sponsored tenders covering a major proportion of the country Key areas are Lake Balaton, Danube, and Ttsza River valleys. A wide range of mosquito pestsare found, includmgAea%s, Anopheles, Culex, and Mansonza A number of prtvate apphcatton companies compete for the spray contracts, using mostly Btl liquid formulations applied from helicopters. Although Btt usage patterns vary constderably across Europe, the products are viewed as possessmg many advantages, including being envnonmentally benign and cost-effective m practical usage. 5. New Developments 5.1. Baculoviruses

Nucleopolyhedrovnuses (NPVs) are increasingly being considered for use as insect control agents. Such vu-uses are almost exclusrvely pathogenic to Lepidoptera. These vtruses produce restmg structures called polyhedra, contaming one or more mfecttve vrrions. When the insect larva consumes contaminated foliage, the polyhedra dissolve in the midgut, releasing the virtons, which then invade midgut epithehal cells. The virus replicates wlthm these cells, uttltzmg them as a stepping stone to build up viral numbers, and subsequently, to invade the hemocoel and colonize the remainder of the Insect. The life cycle IS completed when the insect dies, liquefies, and releases large quantitles of new polyhedra onto the leaf surface. A small number of wild-type baculoviruses have been commerctalized m Europe A typtcal example IS Mamestra brassicae NPV, sold as Mamestrm by Calllope SA, which IS used to control early mstars of Helicoverpa armigera, Mamestra brasslcae, and Plutella xylostella. The product IS formulated as a suspension concentrate and used at 5 L/ha It is registered for use on vegetables m France, Germany, Belgmm, Switzerland, Italy, and United Kingdom, and has been trtaled against cotton pests in Africa.

Butt, Harris, and Powell

40 Table 6 Virus Yield from Wild-Type Stram

AcNPV-wild-type AcNPV + AaHIT (AcST3)

and Transgenic

AcNPV

Virus yield/larva (PIBs x lo-*) 96 1.0

The biggest disadvantage m using wild-type baculoviruses for insect control is the slow speed of kill. Insects can often take longer than 100 h to die, and are stimulated by the mfection to increase their feeding rate above normal in the interim. The consequence of this can be that unacceptable crop damage can occur before the population can be controlled. Researchers have sought to tackle this issue by arming baculoviruses with foreign genes expressmg msectspecific toxins, using genetic engineering techniques. It 1s believed that, by expressing a fast-acting toxin within the insect cell, feeding mhibmon and paralysis could occur quite quickly followmg ingestion, and that acceptable crop protection could be achieved. The Institute of Virology at Oxford, a facility of the National Environment Research Council (NERC), has been one of the pathfinders m relation to the research, and subsequent field testing, of genetically modified viruses m Europe Under extremely strict conditions of containment, modified Autographa calzfornica nucleopolyhedrovuuses (AcNPV), expressing a marker gene, were first trtaled m 1986, to determine the spread and survival of the virus. This was followed up with viruses that contained suicide genes, and then examples that were polyhedrm-negative, that is, deficient m then- protective coat protems, significantly reducing their persistence in the environment (29). More recently, work has concentrated on AcNPV containmg the Androctonus australis insectspecific scorpion toxin gene (coded AcST3). This modtfied vu-us was demonstrated to give significantly faster speed-of-kill m larvae of Trichoplusza ni under field conditions. Furthermore, because a baculovn-us is only capable of reproducing while the host is still alive, polyhedra yields from the msects infected with recombinant vnus were substantially lower than those infected by the equivalent wild-type (29) (Table 6) This suggests that, under practical use conditions, recombinant baculovnuses are competitively disadvantaged compared to wild-types, and, as such, are likely not to predominate m the environment. Other toxin genes can also be considered, as demonstrated m Fig. 5. Two constructs are compared with the wild-type AcNPV. The first construct has received a spider-derived gene encoding for a-laterodectus insect toxin (30). The second construct has received a gene described as Tox 34#4, which is an RT-PCR-generated cDNA clone encoding a protein with a high level of

Microbial Biopesticides: European Scene

41

Dose

Activity

1 x 10E6

PIB’slml

days after treatment

Fig. 5. Activity of wild-type AcMNPV vs constructsagainst Heliothis virescens (tobaccobudworm). (Adapted with permissionfrom ref. 32.) homology (94% identity) to the original TxP-1 itch mite (@emotes tritici) toxin (31). Both of these constructs have the toxin gene driven off the PlO viral promoter, which commences expression about 36 h postinfection. As can be seen,there is a substantial benefit in terms of speed-of-kill, especially with AcNPV + Tox 34#4 (32). This virus is used solely as a research standard within Zeneca. Arguably, highly effective recombinant baculoviruses offer the best likelihood of microbial insecticides achieving chemical-like levels of efficacy and crop protection effect, and have excellent potential for usage in key European markets, such as vegetables, fruit, and tomatoes. 6. Future Prospects The momentum for the development of natural agents for the control of pests, weeds, and diseaseswill be sustained. This will, in part, be the result of increasing awareness and sensitivity of the general public to health and environmental risks associated with chemical pesticides. However, as knowledge and experience in the harnessing of BCAs increases, they will be easier to deploy, and, therefore, will be used more extensively than at present. This momentum can only be maintained if there is investment in the research and development of BCAs, combined with support from extension services and industry, to optimize the impact of these agents. Only then can the technology

42

Butt, Harris, and Powell

be transferred to the end user, and a sustainable, envn-onmentally protection program established.

friendly

crop

References 1 Pickett, J A , Butt, T M , Doughty, K J., Wallsgrove, R M , and Wtlltams, I H (1995) Mmimismg pesticide input m otlseed rape by explomng natural regulatory processes Plenary lecture, m Proceedzngs of the GCIRC 9th Internatzonal Rapeseed Congress, Cambridge, UK, &7 July, 2,565-57 1 2 Rhodes, D J (1990) Formulatton requirements for btologlcal control agents Aspects Appl Bzol. 24, 145-153 3 Moore, D , Douro-Kpmdou, 0 K., Jenkins, N E , and Lomer, C J (1996) Effects of moisture content and temperature on storage of Metarhzzzum flavovzrzde comdta Bzocontrol Scz Technol 6, 5 16 1 4 Moore, D., Bateman, R P , Carey, M., and Prior, C (1995) Long-term storage of Metarhzzzumflavovzrzde comdra m 011formulattons for the control of locusts and grasshoppers. Bzocontrol Scz Technol 5, 193-199 5 McLatchte, G V , Moore, D , Bateman, R P , and Prior, C (1994) Effects of temperature on the viabtlny of the comdta of Metarhzzzum flavovzrzde m oil formulations Mycologzcal Res 98, 749-756 6 Moore, D , Bndge, P. D , Htggms, P. M , Bateman, R P , and Prior, C (1993) Ultravtolet radiation damage to Metarhzzzum flavovzrzde conrdta and the protectton given by vegetable and mineral oils and chemical sunscreens Ann Appl Bzol 122,605-616 7 Jenkins, N E and Thomas, M B (1996) Effects of formulation and appltcatton method on the efficacy of aerial and submerged comdta of Metarhzzzum flavovzrzde for locust and grasshopper control Pestzczde Scz 46,299-306 8 Smith, D B., Hostetter, D. L , and Pmnell, R E. (1980) Laboratory formulation compartsons for a bacterial (Bacillus thurzngzenszs) and vtral (Baculovrrus Helzothzs) msecttctde J Econ Entomol 73, 18-21 9 Lisansky, S. (1989) Btopestmides fall short of market proJections Performance Chem l&387-396 10 Powell, K A , Faull, J L , and Renwrck, A. (1990) Commercial and regulatory challenge, m Bzologzcal Control of Sozl-Borne Plant Pathogens CAB lnternational, Wallmgford, UK 11 Pusey, P L , Wtlson, C L , and Wtsmewski, M E (1993) Management of postharvest diseases of fruits and vegetables strategies to replace vamshmg fungicides, m Pestzczde Interactzons zn Crop Productzon Benejczal and Deleterzous Effects (Altman, J , ed ), CRC, Boca Raton, FL, pp 477-492 12 Wilson, C L. and Wismewskt, M E (1989) Biological control of postharvest diseases of fruits and vegetables an emerging technology Annu Rev Phytopathol 27,425-441 13 Ryder, M H and Jones, D A (1990) Biological control of crown gall, m Bzologzcal Control of So&Borne Plant Pathogens CAB International, Wallmgford, UK 14 Defago, G. and Haas, D (1990) Pseudomonads as antagomsts of sotlborne plant pathogens. mode of action and genetic analysts Sozls Bzochem 6, 249-29 1

Microbial

Biopesticides:

European

43

Scene

15 Alabouvette, C (1990) Biological control of Fusarzum welts m suppresstve solIs, m Btologzcal Control of Sod-Borne Plant Pathogens CAB lnternattonal, Wallingford, UK 16 Rtshbeth, J. (1963) Stump protection against Fomes annosu~ III. Inoculatton with Pentophora gtgantea. Ann Appl Btol 52,63-77

17 Butt, T M , Ibrahtm, L , Ball, B. V , and Clark, S J (1994) Pathogemctty of the entomogenous fungt Metarhtztum antsopltae and Beauverta basstana agamst cructfer pests and the honey bee Bzocontrol Set Technol 4,207-2 14 18 Butt, T M , Ibrahtm, L , Clark, S J., and Beckett, A (1995) The germmation behavtour of Metarhtztum anzsopltae on the surface of aphid and flea beetle cuticles Mycologzcal Res 99, 945-950. 19 St. Leger, R , Butt, T M., Staples, R , and Roberts, D W (1989) Syntheses of proteins mcludmg a cuticle-degrading protease during dtfferenttatton of the entomopathogemc fungus Meturhtztum antsopltae Exp Mycol 13,253-262 20 St Leger, R , Butt, T M , Goettel, M S , Staples, R , and Roberts, D W (1989) Productton in vrtro of appressorta by the entomopathogemc fungus Metarhtztum antsopltae Exp Mycol 13,274-288 21 Anon (1989) Agrow Btologtcal Crop Protectton, PJB, Surrey, UK. 22. Remecke, P , Andersch, W , Stenzel, K , and Hartwtg, J (1990) BIO 1020, a new mtcrobtal msecttctde for use m horttcultural crops, m Proceedtngs ofthe Brtghtun Crop Protectton Conference, Pests and Diseases, vol 1, pp 49-54 Vestergaard, S , Gtllespte, A. T , Butt, T M., Schretter, G , and Etlenberg, J. 23 (1995) Pathogentctty of the hyphomycete fungt Verttctlltum lecantt and Metarhtztum anzsoplrae to the western flower thrtps, Frankitntella occtdentalts Btocontrol SCZ Technol 5, 185-192 24. Butt, T M., Barrtsever, M , Drummond, J, Schuler, T H , TIRemans, F T , and Wtldmg, N (1992) Pathogemcity of the entomogenous, hyphomycete fungus, Metarhtzzum antsoplzae agamst the chrysomeltd beetles Psylltodes chrysocephala and Phaedon cochleartae Btocontrol Set Technol 2,325-332 25. Glowacka B and Malmowskt, H (1994) Department of Forest Protectton, Warsaw, personal communicatton 26 Anon (1994) A threat to a thud of Poland’s forest Bzotzmes (Novo Nordtsk) 27. Harris, J. G (1994) Expertence of B t z usage m U.S A , Europe, and Afrtca, m Proceedtngs of the Semtnar on Mtcrobtal Control ofMosquttoes, Untversttt Sams Malaysia, Penang, Malaysta 28. Becker, N and Margaltt, J. (1993) Use of Baczllus thurzngzenszs zsraelenszs agamst mosquttoes and blacklies, m Bactllus thunngienszs, An Envtronmental Btopestzczde Theory and Practzce (Entwtstle, P. F , Cory, J S., Bailey, M J , and Htggs, S , eds ), Wtley, Chtchester, UK, pp 147-170 29 Cory, J (1995) Open Baculovirus Publtc Meetmg, NERC Oxford, November 30. Watkms, M , Ktyatkm, N , Hughes, D , Beadle, D , and Ktng, L (1997) Study on the btologtcal properties of a novel recombinant baculovirus, tn Proceedtng of the BCPC Necesstty? pp. 279-284

Conference,

Microbtal

Insecttctdes,

Novelty

or

44

Butt, Harm, and Powell

31 Tomalskr, M D and Mtller, L. K (1991) Insect paralysis by baculovn-us medrated expresston of a mtte neurotoxm gene. Nature (Lond) 352,82-85. 32 Harrts, J G (1997) Mrcrobtal msecttcldes an industry perspective. How the mdustry sees the future, the opportumttes and the tools to solve the problem, m Proceedmgs of Mcroblal Insectudes Novelty or Necewty? BCPC Symposium 68,4 I-50. 33. OECD (1996) Data requirements for registration of bropesttcrdes in OECD member countries survey results. Envnonment monograph No 106, OECD, Pans, France, p 12 1 34. Feng, M G , Poprawskt, T. J , and Khachatourians, G. G (1994) Productton, formulation and appltcatton of the entomopathogemc fungus Beauverla basslana for insect control. current status Bzocontrol Scl Technol 4,3-34

4 Developing Countries Balasubramanyan

Sugavanam

and Xie Tianjian

1. Introduction Blo- and botanical pesticides are often grouped together m developing countries as possible alternatives to chemical pesticides. In reality, botanic pestlc!des are no different from chemical pesticides, but blopestlcides are all far removed from them. Botanical pesticides derived from tobacco (Nicotiana tabacum), pyrethrum (Chrysanthemum cznerariaefolium), derrls roots, neem, and so on, have been known for many decades, but occupy only a very small fraction of the overall pesticide market, which IS now worth almost $28 bllhon, and 1slikely to grow to $34 bilhon by the year 1998. Some of the well-known botanical pesticides will fail today’s strict and exhaustive reglstratlon requirements. However, the botanical pesticides gave ideal models for sclentlsts to modify structure and optimize blologlcal activity The synthetic pyrethroids revolutionized the pesticide industry m the 197Os,and today share more than a $2 bllllon market. In the 1980s and 199Os,based on mcotme and strobllurm, major inventions were made m bringing to the market compounds, such as Bayer’s (Germany) tmidacloprid (11 and ICI (UK) A5504 (azoxystrobm) (2), shown in Fig. 1. In addition, BASF also has invented a strobllurin analog called kresoxlme methyl (BAS490F). These will have a big impact In the plant funglclde market. In these mventlons, traditional structure-activity relationship, partltlon coefficients, mode-of-action studies, and computer graphics have been utilized by synthetic chemists to invent key molecules that could fit the relevant enzyme surface, like lock and key, to invoke the needed blologlcal actlvlty and at the same time not interfere with mammalian and aquatic species and not accumulate in the environment. In the case of blopesticldes, there was a great hope for these products during the early 1970s. However, because of major breakthroughs m conventronal From Methods m Botechnology, vol 5 Elopesfudes Use andDe/wery E&ted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

45

46

Sugavanam and Xie

Imidadoprid

Fig 1 Chemicalsrecently synthesizedusmgnatural products as lead compounds chemistry durmg the 1970s and 1980s in the invention of new synthetic chemlcals, the resources available to blopestlcldes became less and less At the same time, with the soarmg cost of inventing new pestlcldes, the long and expensive time required for reglstermg pesticides, and also problems encountered m developing countries caused by misuse of pesticides and the unnecessary exposure of workers and the environment to highly toxic pesticides, there was a great necessity and urgency to move to user- and environment-friendly pestlcldes. Blopestlcldes came at an ideal time and also, many countries agreed to relax registration requirements and reduce the time required to register blopestlcldes. This started a flurry of research actlvltles by many companies to Invest more m research and development to dlscover blopestlcldes. A large number of pheromones, and blologlcal agents based on fungi and viruses, have been successfully tested m the laboratorles and the field, but there has been very little impact m the market Still, the Baczllus thurzngzenszs (Bt) registered during the 1970s for insect control, and the various genetic mampulatlons of It, dominate the biopestlclde market 2. The Situation in Developing Countries With great importance given to integrated pest management (IPM) worldwide, the role of biopestlcldes 1scrucial as one of Its components. In developmg countries, excessive use of pesticides has caused development of resistance by insects and fungi to many conventional pesticides, and also resurgence of pests caused by destruction of natural enemies. Classic examples, including the brown plant hopper on rice, and bollworms and white fly on cotton, are all problems emanating from human use or misuse of synthetic pestlcldes. Barbosa (3) lists hundreds of pests of mternatlonal importance m more than eight major crops to which Bt could play a role. Although Bt and other blopestlcldes have good potential, both on their own or within the definition of IPM, there are many barriers to cross, because the new synthetic pestlcldes are becoming more and more effective, the formulations are becommg more and more userand environment-friendly, while blopestlcldes lack consistency m their field performance, and suffer quality variations and shelf-life mstablhty The luxury

Developing

Countries

47

of usmg synthetic pesticides m various types of formulattons and delivery systems are hmited m the case of biopesttcides Again, big companies getting fully mvolved m developmg biopesttcides are also lrmited because of lower profit margms, hence, only companies addressmg mche markets, or when organic farms supplemented with attractive subsidies to compensate for addltional costs m terms of using biopesticides and biological agents, are taking up biopesticrdes. In developing countries, lower labor costs would make biopesttcides or biological agents, such as viral msecticides, more attractive, provided there are enough mcentives and technical assistancegiven by the governments and international agencies, such as FAO, WHO, and UNIDO. Such countries such as Indonesia, the Philippines, and Egypt have been phasmg out a number of msecticides, and here the biopesticides could replace some of them and fit wtthin the prmctples of IPM. 3. Production of Biopesticides in Developing Countries Developtng countrtes began looking into the local production of biopesticides, focusing mostly on Bt, almost 30 years ago. In 1964, China estabhshed a research and development laboratory to explore possible local production. A group of researchers at Campmas Umversity m Braztl studied all aspectsof Bt fermentation in the 197Os,and filed two patents In Egypt, Bt production was carried out in a sugar and dtsttllery company using a 5000 L fermenter equipped with centrifugatton and drymg facihties Countries such as India, Mexico, and Thailand tried to look mto low cost production of Bt Unfortunately, exceptmg China, most of these attempts never became a commercial reality, because of lack of experience m quality control, formulation, and application of btopesttcide at that time. In Chma, Beauverza basszsana was the most important fungal msectitide, and was developed more than 10 yr ago A vast amount of vtgorous mycelium was harvested from fermentation (liquid stage) and transferred mto semisolid media (solid stage), consisting of bran, sugar, agar, and minerals. The final product contained 1.8 x IO” spores/g. Many village manufacturers adopted this procedure to produce 5. basszana for control of pine caterpillar and corn borer. Chma’s Ministry of Forestry is m charge of formulatmg the product standards, and will be publishing them soon. For viral msecticrde, the first factory was established m 1985 to produce Hellothzs armzgera NPV. The production flow chart is given tn Fig. 2 Artificial diet for rearing msects was composed mostly of soybean, barley, and yeast Recently, most HaNPV preparations have been processed mto cheaper liquid formulattons. In China, three HaNPV firms have been registered, and about 100 t of this product, contammg 3 billion PIB/mL, are now sold to farmers for control of cotton bollworm

Sugavanam and Xie

48

new hatched

the fourth

tnster

lnrvae

larvae’

Fig. 2 Flow chart of HaNPV production.

4. Bt Production in China Bt production m China has a long history. In the early stages, problems m fermentation and product quality control caused fluctuation m the performance of the product However, since the 198Os,great improvements have been made m screening strains and overcoming phage contammation, changing downstream processmg, formulation, and bioassay procedures. The flow chart of production of Bt m China is given below: Llquld Liquid stock seed + slant seed -+ fermentation + sieving -+ centnfugatlon + checking quality -+ packaging liquid formulation Solid Spray drying of the centrifugal hquld + mlcromlllmg + blendmg + checking quality + packaging powder formulation

In China, use of local Bt strains is encouraged. Several hundred Bt strains are screened every year. One strain, MP-342(H3ab), screened by Hubet Bt Research Centre (Bt RDC), containing cryIA(a), cryIA(c), cryIIA, cryIIB, and cryV genes, shows high toxicity to cotton bollworm. It is the most important Bt strain for commerctal production. Thousands of local Bt strains have been collected in several institutes and universities in China. Primarily defatted soybean flour and cottonseed flour are used as raw material.

During the early days

Developing

Countries

49

of commerctal productton, the concentration of medta was between 3 and 4%, whrch induced low potency of broth. Based on Bt metabolism studies, optrmtzmg fermentation media composition and fermentation parameters, the potency now can reach the level of 3000-4000 IU/pL Phage contammatton, which once threatened Bt fermentation, 1s now completely removed, and the failure rate Induced by phage has been kept to below 1% since 1986. Two kinds of Bt formulattons are produced. The liquid formulatron, usmg only centrifugatton, 1snot as good as the wettable powder formulatton, but is very popular, and today makes up 70-80% of the total Bt market m China. 5. Quality Control Quality control plays a key role m Bt productron. Spore count has been totally replaced by bioassay. The standard sample developed by BtRDC and the Central Chma Agricultural University have been adopted by the scientific community and manufacturers. Two testing insects, cotton bollworm (Heliothzs armzgera) and diamondback moth (Plutellu xylostella), are used in the laboratory The product quality standards are as follows: liquid formulation: 20004000 IUIyL, 1 yr stabthty period, powder formulatton: 16,000-32,000 IU/pL, 2 yr stab&y period. Good-quality Bt products, reasonably priced, are becoming popular m China. Over 40 factories have been registered; all equipment used m the plants are made locally, and give satisfactory performance. BtRDC is the biggest Bt manufacturer in China, and has increased its capacity by moving from 7000-L fermentor in the early 1980s to 40,000-L fermentor now. The production of ltqutd formulation has increased from 1200 Mt in 1993 to 1400 Mt in 1994 to 2500 Mt in 1995 to 5500 Mt in 1996. Most of the wettable-powder formulations are exported. In China, more than 1 mtllton ha are treated with Bt, and local productron will keep increasing, and will complement synthetic pestrctdes in selected important outlets. 6. Registration Requirements Registration requirements of btopesttcides vary from country to country. In China, the Instrtute for the Control of Agrochemlcals, Ministry of Agrtculture (ICAMA), 1sthe authority for pesticide registration. Based on the fact that there are no accidents reported in large-scale Bt production and application, and no adverse effect on nontarget species, such as honey bees, buds, and fresh water fish for decades,and that all Bt manufacturers use native strams for local production, ICAMA makes registration requirements for Bt simple and effective For finished products, the guidelines of data required are given in Table 1. The potency of Bt products is determined by bioassay with newly hatched cotton bollworm (Helzothzs armigera) or dlamondback moth (Plutella

50

Sugavanam Table 1 Guidelines

for Data Required

for China’s

and Xie

ICAMA Registration

Chemical and physical properties Name and type of formulation Quantity of active ingredient Content and identity of nonactive ingredients, such as UV protectors Water-retaining agents, and so on Content of extraneous organisms Chemical and physical properties Stability of product and effect of temperature and storage conditions on biological activity Method of analysis Toxicology Acute oral toxicity or pathogenicity (to mice) Acute dermal toxicity (to Oryctolugus cuniculus) Efficacy Crops to be protected Target pests Degree of specificity for target pests Effective dose level and mode of action Rate, frequency, and method of application

xylostella) in different factories, and by comparing this to a standard sample supplied by ICAMA. According to the guidelines and analysis method mentioned above, most manufacturers do not have difficulties in obtaining registration for Bt products. It is clear that ICAMA makes a great contribution to promoting Bt production and development. Registration of viral and fungal insecticides is just in the early stages. 7. Biopesticides

in Thailand

Thailand has played an important role in promoting biopesticides. Research and development carried out at Mahidol University and the Ministry of Agriculture at Keserat University have promoted both Bt and viral insecticides that have been taken up by private entrepreneurs. Thailand has also played the role of focal point in UNIDO’s program (4) Regional Network on Pesticides for Asia and the Pacific (RENPAP) (see also Subheading 9.). India has been a long supporter of biopesticides, but most of the work was research-oriented, with very little applied research and promotion of production. Because of a highly subsidized malaria eradication program using DDT, there was no incentive

to use biopesticides

in vector control.

Resistance

to

Developing

Countries

51

DDT and its illegal use in agriculture might be a major incentive for moving to biopesticides as environmentally friendly alternatives in malaria vector control. Export barriers because of pesticide residues might also provide outlets for the application of biopesticides in vegetables, cotton, and so on. 8. Other Countries In 1995 UNIDO, in collaboration with the RENPAP, organized a workshop on production and quality control of biopesticides in China (5), in which many Asian countries participated. The main objective of the workshop was to assist countries in Asia to develop capabilities in production and use of biopesticides. Many papers presented in the workshop (5) revealed the situation in many countries of the region. In Pakistan, the use of microbial insecticide has been adopted as part of IPM approach to provide an environment-friendly alternative to generally hazardous broad-spectrum insecticides used against Heliothis armigera. Laboratory bioassays using spore crystal preparations of Bt kurstaki indicated high mortalities of the first instar larvae of H. armigera. Potted chickpea (Cicer arietinum L) plant testsrevealed that the biopesticide Dipel2x and Dipel ES (Bt kurstaki), at rates of 1.6 kg/ha and 2.0 L/ha, caused 81.48 and 84% larval mortality, respectively. Field tests of Bt on chickpea crops (three consecutive seasons)indicated that Dipel2x and Dipel ES (Bt kurstaki; Abbot, USA), at the rates of 1.6 kg/ha and 1.5 L/ha (with and without molasses), respectively, caused significant increase in grain yield, compared to control plots. In Pakistan, where the major portion of insecticides are used on cotton, use of biopesticides and biological control agents would be economically and environmentally appealing, but commercial production and use of these is still not seriously taken up in that country. In the Philippines, the use of biopesticides is becoming increasingly important. The total market share of Bt in the insecticide market increased from 4% in 1992 to over 9% in 1995. The use of Btk is primarily in cabbage, against diamondback moth. In the Philippines, lack of facilities and financial support for Bt research and development, and lack of technology on bioassay and commercial production of Bt, are the major constraints in the development of Bts and their formulations. In South Korea, in order to reduce adverse impact on nontarget organisms and the environment, the government has put the emphasis on development of alternative pest control agents based on low input and sustainable agriculture. Biopesticides obviously appear as good alternatives. Of the 568 pesticides registered in the country, biopesticides account for 3% (15 products). Production of biopesticides increased almost 200-fold in the last 10-yr period (1984-l 994). Quality control of pesticides is regulated under the Pesticide Management Law through the National Agricultural Science and Technology Institute (NASTI). The quality of Bt-based pesticides is switched to Diamondback Moth Unit

52

Sugavanam

and Xie

(DBMU) from Biologtcal International Unit (BIUL). In South Korea, signif? cant research work 1sbemg carried out on btopesttctdes, but needs integrated gutdelmes, tncludmg btologtcal testmg methods, requu-ed for quality assurance of btopesticides. In Vietnam, research work on the utthzatton of btopestictdes was mmated m 1970, m order to develop domestic production of Bt. Bt application IS quite effective and popular m controllmg pests, such as Plutella xylostella in vegetables, and other leptdopteran pests. In collaboratton with the Food Industries Instttute, 300-L batches of Bt kurstaki, with a shelf life of 7 mo under low temperature, are bemg produced, and the product has been commerctahzed Though the locally produced Bt is cheaper than imported Bt, product standardization and contammations are some of the problems yet to be solved. In Africa, major research and development work has been cart-ted out at the International Center of Insect Physiology and Ecology (ICIPE), m Nairobi, Kenya Independently, and m collaboratton wtth mternattonal orgamzatton and bilateral agencies, the center has developed Bts and carried extensive field studies. Recently, they have been testing a Bt product from Finland called Dudstop for the control of filth flies, whtch are known to be potenttal agents for transmtsston of entertc diseases,such as dysentery, mfanttle dtarrhea, typhoid, and trachoma, among others. Most of the tests have been carrted out m refugee camps m Kenya, Ethiopia, Tanzania, and Et-urea. In collaboratton with the Hebrew Umversity of Israel, the Institute ts testing Bt uraelensu and Bacdlus sphaerzcus for larvicidal effect on different mosqutto spectes and then perststence. The Institute is eager to establish pilot plant factlittes for productton of some Bts, but, again, lack of funds and support make these goals unattainable. 9. Activities of International Organizations International orgamzations have been acttve for many years m promotmg btopesttctdes, on their own, or as part of an IPM strategy With the development of resistance to many pesticides and the problems faced with the contmued use of DDT m malaria vector control, and the recent concern over perststent orgamc pollutants, Bt lsraelenszs offers an excellent opportunity to interfere wtth the mosqutto cycle at tts cructal stage as a larvtctde. As early as 1982, the World Health Organization, supported by the World Bank and the United Nations Development Programs (UNDP), developed gutdelmes (6) for productton of Bt H-14 for biological control of vectors. The guidelines supported production of Bt m developmg countrtes for both public health and agricultural requirements The FAO and UNIDO were requested to assist in early-stage evaluatton and plannmg. UNIDO has been active m orgamzmg workshops m productton, quality control, and bioassay in developmg countries. The FAO also wtll extend its code of conduct for btopesttctdes. Most of the work so far

Developmg Countr/es

53

has been concentrated on conductmg workshops and expert group meetings, with many recommendations for commercial production of biopesticides to complement synthetic pesticides. As always, most of these recommendations were not followed by financial and policy incentives and commitments by donor or recipient countries. 10. Opportunities for Developing Countries With developmg countries giving great emphasis to IPM, biopesticides offer an excellent opportunity for developing coutries. Some of the reasons for this, as could be seen from the experience of China, are summarized below. 1 Raw materials are locally available, and could be made from locallyavailableBt strams 2 Preparation of vuus msecttctdes needs manual labor and, hence, 1s highly suttable to developmg countrtes 3. Technology transfer and quality control could be negotiated with relevant companies or mstitutrons m both developed and developmg countrtes 4. For proper trammg m apphcatton of btopesttctdes, mternattonal orgamzattons, such as FAO, UNIDO, and WHO, could be contacted 5. Use of biopesttctdes would also eliminate export restrtctlons based of restdues of certam pesticides, especially in vegetables, frmts, and other commodittes.

Developing countries should look into biopesticides for complementmg synthetic pesticides and not for replacing them. If carefully planned m certain outlets, biopesticides would bring great benefits m reducing the use of synthetic pesticides, especially those that are toxic and persistent. There should be a systematic national/regional strategy to momtor development of resistance to biopesticides, so that they could be rotated along with synthetic pesticides. With the present technology, the market share of biopesticides would be very limited, compared to synthetic chemicals. If one combines biopesticides wtth biologtcal

agents, disease, and insects, and herbicide-resistant

transgemc

crops, the overall effect of biotechnology on crop production and protection will have great impact in the next millennium. One should also apply the caution that synthettc pesticides will contmue to dominate

the market tn the near

and distant future, because of constant improvement m mventions of new pesticides, their formulattons,

and application

technology.

References 1, Shtokawa, K , Tsubot S , Kagabu, S , and Morrtya, K (1994) Imtdacloprtd a chlopraomcotmyl msecttclde, its invention and features, m Ezghth ZUPAC Znternational Congress of Pestlclde Chemutry, Options 2000, American Chemical Society, Washmgton, DC, p. 4 2. Anthony, V M , Clough, J M , Godfrey, C R A, and Godwm, J R (1994) Syntheses of fungtctdal methoxyacrylates, Ezghth IUPAC Internatzonal Con-

54

3.

4

5

6

Sugavanam and X/e gress of Pestzclde Chemistry, Optlons 2000, Amertcan Chemtcal Soctety, Washmgton, DC, p 5 Barbosa, S (1993) Potential of Bt m mtegrated pest managementfor developmg countrtes, m Blopestlclde Bt and rts Appllcatlons in Developing Countries (Salama, H S , Morris, 0 N , and Rached, E , eds), National Research Centre, Catro, Egypt, and IDRC, Canada,pp 59-7 1 Dhua, S P (1992) Regtonal network on pesttctdes for Asta and the Pacific (RENPAP) an overview, m Recent Developmentsin the Field of Pestwdes and thew Application to Pest Control (Holly, K , Copping, L S , and Brooks, G T , eds ), UNIDO, Vtenna, pp 277-284 Report on Workshop on Productton and Qualtty Control of Btopesttctde Bt Wuhan, Republtc of Chma, November, 1995 UNIDO Report, DP/ID/Ser A/ 1765,1995. Vandekar, M and Dulmage, H T , eds (1983) Guzdellnesfor Pvodzxtzon of Bacillus thurznglensls H-14, UNDP, World Bank, WHO

5 Pesticide Policy Influences on Biopesticide Technologies Noel D. Uri 1. Introduction Pesticide pohcy has a substantial impact on the development and use of btopesttctdes. Before explormg the nature and extent of this impact, it 1sworthwhile to examine btopestictde use, m order to put policy influences m proper perspective The constant evolution and adoption of new productton practices, includmg relatively pesticide-intensive farming, has led to a sustained increase m output and comctdent benefits to the American consumer m a variety of ways, mcludmg the price of food The conventional pesticides used today were new and uncertam m prevtous periods In an analogous way, biopesttcrdes and other forms of biologrcal control being developed today will be conventtonal pesttctdes m the future. The growth of biopesttcide use is an Integral part of the technologtcal revolutton m agriculture that has generated major changes m production techniques, shifts in input use, and growth m output and producttvtty (I) Predicting the growth m biopestlctde use, however, IS difficult, because of recent changes in federal laws affecting the farm sector (a major consumer of btopesttctdes) and regulating the registration and use of pesticides. Addtttonally, accurately forecasting the changes m the price of biopesticldes relative to the price of conventtonal pesticides complicate the predictron problem. These issues will be discussed below The market for brologtcally based pest controls IS small but fast growing The market value of btologtcally based products-natural enemies, pheromones, and mtcrobtal pesticides-sold m the United States during the early 1990s was estimated at $95-147 mtllton, 1.3-2 4% of the total market for pest control products (2). At least 30 commercral firms produce natural enemies. From Methods ,n Biotechnology, vol 5 B,opest!ades Use and Dehvery Edited by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

55

56

U-1

Even though the current market for brological products is growing, and large pest-control companies are begmnmg to participate, the market IS still so small that biologlcals are unlikely to replace conventional chemical pesticides m the foreseeable future, unless major research and development activities are started (3). Biologrcal pest management includes the use of pheromones, plant regulators, and microbial organisms, such as Bacillus thuringlenszs (Bt), as well as pest predators, parasites, and other beneficial organisms. The US Envnonmental Protection Agency (EPA) currently regulates biochemicals and mrcrobral organisms, and classifies them as biorational pesticides. 7.7. Microbial

Pesticides and Pheromones

Brorational pesticides have differed significantly from conventional pesticides, because they have generally managed rather than eliminated pests, have a delayed impact, and have been more selective (4). Thus, for example, microbial pesticides have not been successful as herbicides, because target weeds are replaced by other weeds not affected by the microbial pesticide Among the most successful mrcrobials has been Bt, which kills Insects by lethal infection. Growers have dramatically mcreased then use of Bt during the 199Os, especially under blomtensive and resistance-management programs, because of its environmental safety, improved performance, selectrvity, and activity on insects that are resistant to conventional pesticides. It is one of the most important msect management tools in certified organic production. Bt was used m more than 1% of the acreage of 12 fruit crops m 1995, up from 5 crops m 199 1 (Table 1). Between 12 and 23% of the apple, plum, nectarine, and blackberry acreage received Bt applications m 1995, and it was applied on over half of the raspberry acreage. Among vegetable crops, the acreage treated with Bt increased for 13 of the 20 crops surveyed between 1992 and 1994, and was used on about half or more of the cabbage, celery, and eggplant acreage Bt has been used on only a couple of field crops. Corn acreage treated was steady at 1% m 1994 and 1995, and treated cotton acreage increased from 5% m 1992 to 9% m 1994 and 1995 New Bt strains, with activity on insects not previously found to be susceptible to Bt, have been discovered in recent years. Current research is devoted to improving the delivery of Bt to pests and to increasing the residual activity and efficacy of Bt. Pheromones are used to monitor populatrons of crop pests and to disrupt mating m organic systemsand some IPM programs. Pheromones were used on 37% of fruit acreage to monitor and control pests, and on 7% of vegetable acreage to control pests (Table 2).

Table 1 Agricultural Applications of Bacillus fhuringiensis Selected Crops in Surveyed States, 1991-1995

Crop Field crops Corn Cotton (upland) Fruit Grapes Oranges Apples Peaches Prunes Pears Sweet cherries Plums Nectarmes Blueberries Raspberries Blackberries Vegetables Tomatoes, processed Lettuce Sweet corn Onion Broccoli Tomatoes, fresh Cantaloupe Snap beans Cabbage Bell peppers Cauliflower Cucumbers Strawberrres Celery Honey dew Spinach Eggplant

Planted acres (in thousands)

(Bt), Area receivmg apphcation (in % acres)

1991

1992

64,105 1 I .650

a

796 760 345 144 94 68 47 44 36 30 11 4

a

b

2 3 C, 0 a n a L1

b

0

a

11 49 18 b

323 191 164 128 111 104 98 71 70 61 54 51 46 36 26 10 4

b 6 b

b b b b b b b b h b b b b

OApplred on CO5% of the acres *Not a survey year for that commodny (Adapted wtth permlssron from refs. 40 and 42.)

57

1993 ‘2

5

b b b

8

b

b

b

10 8 45 n

b

6 18 3 ” 7 31 32 20 48 35 12 19 24 51 28 13 13

1 9

1 9 6 3 12 5 9 2 9 14 22 5 52 23

1 8

b

b

1995

2 7 13 3 0

b

b

1994

h b b b b

b b

b h b b b b b b b b

5 20 3 1 14 39 8 29 64 37 20 22 33 61 10 21 48

b

b b b b b 6 b b b b h b b b b b

58

Ur1

Table 2 Use of Selected Biological Pest Management Practices on Fruit and Vegetable Crops in Major Producing States in the 1990s % of acres Crop Fruit Grapes Oranges Apples All frurts Vegetables Sweet corn Tomatoes Lettuce All vegetables

Planted acres (m thousands)

Beneficial Insects

Pheromone traps

Resistant varietres

18 22 2 19

14 28 66 37

31 21 16 22

NA 5

17

NA NA NA NA

730

613 381 3251 640 357 259

2914

3 3

6

1 7

NA not avarlable (Adapted with permlsslon from refs. 40 and 41 )

1.2. Beneficial

Organisms

Natural enemres of crop pests or beneficrals may be Imported, conserved, or augmented. Many crop pests are not native to thts country, and the US Department of Agrrculture (USDA) issues permits for the natural enemies of these pests to be Imported from thetr country of ortgm. Natural enemy rmportatron and establishment, also called classrcal brologrcal control, has been undertaken prrmartly m umverstty, state, and federal projects. Twenty-erght states operate brocontrol programs, and most have cooperatrve efforts with USDA agencies (2) Some crop pests, such as the woolly apple aphid m the Pacific Northwest, have been largely controlled with this method. Natural enemies may also be conserved by ensuring that then needs for alternate hosts, adult food resources, overwmtermg habitats, a constant food supply, and other ecologrcal requirements are met, and by preventing damage from pestrctde apphcattons and other cropping practrces (5). Over one-half the certrfied organic vegetable growers m 1994 were provrdmg habitat for beneficrals. A small but mcreasmg number of companies are supplymg natural enemies of insects, weeds, and other pests to farmers. For greenhouse and agrrcultural crop productron, most natural enemies being sold, such as beneficial insects, predatory mites, parasitic nematodes, and insect egg parasites, are used for managing pest mites, caterpillars, citrus weevrls, and other insect and arthro-

Policy

Influences

on Biopesticides

59

pod pests, However, a number of natural enemies-musk thistle defohatmg weevils, for example-are bemg sold for managmg weeds on rangeland and unculttvated pastures (6). The Cahfornia Environmental Protection Agency has published a list of commercial suppliers of natural enemies m North America since 1979, and the number has increased steadily. In 1994, 132 companies were listed, mostly m the United States, offering over 120 different orgamsms for sale (7). 1.3. Host-Plant Resistance Corn and soybean breeding for genetic resistance to insects, disease, and other pests has been the research and development focus of major seed companies for many decades (8). US soybean acreage, for example, receives virtually no fungictdes, because of the effectiveness of the disease-resistant soybean cultivars that have been developed. The use of classical breeding programs is now being augmented wtth new plant breeding efforts using transgemc and other genetic engineering techmques. In March 1995, the EPA approved, for the first time, a limited registration of genetically engineered plant pesticides to Cuba (Base& Switzerland) and Mycogen (San Diego, CA), and, m August 1995, granted condrtronal approval for full commercial use of a transgemc pesticide to combat the European corn borer (9). Thts plant pesticide, Bt corn, is produced when the genetic mformation related to insecticidal properttes IStransferred from the Bt bacterium to the corn plant. This technology could reduce the need for conventional chemical msectrcides in corn productton. In 1995,26% of US corn-planted acreage was treated with msecticides, and corn borer is one of the top insect pests targeted for treatment. Since these new corn varieties, however, contam natural genes and genes produced from the soil bacteria Bt, many scienttstsare concerned that the new corn will hasten pest immunity to Bt That is especially a concern for the growmg number of producers who rely on the foliar-applied Bt, and has led the EPA to approve the new pestrcrdes, condrtioned on the momtormg for pest resistance and the development of a management plan in case the insects become resistant Although most classtcal breedmg programs have focused on pests resistant to chemtcals or treatments that were too expensive (10), consumer concern over pesticides m agricultural products has prompted biotechnology companies to enter the genetically engineered plant market. As agricultural biotechnology products attam commerctal success,some private Investment fundmg may shift from the smaller pharmaceutical markets toward agrtcultural crop protection (II). On the other hand, consumer acceptance of bioengmeered Bt corn, Bt cotton, and other genetically engineered crops has not yet been dem-

Uri

60

onstrated m major US markets. A 1992 survey of consumer attitudes about food biotechnology found that most consumers want mformation on labels about vartous food characteristics, including the use of biotechnology (12) Animal Plant Health Inspection Service (APHIS) has approved or acknowledged 638 field trials for insect-resistant varieties since 1987,286 field trials to test viral resistance, and 94 field trials for fungal resistance (23). 2. Biopesticide Use in the Context of Pesticide Policy Many factors affect the adoption of biopesticides as a crop productton technology. Pest cycles and annual fluctuations caused by weather and other envtronmental conditions often determine whether infestation levels reach treatment thresholds. Changes m farm biopesticide use are related to producers’ decisions on the amount and mix of crops to plant. Given this, other factors that influence bropesttcide use decisions are relative factor prices and government farm, conservation, and regulatory policies 2.1. Relative

Factor

Prices

The changing relative prices between different pesticides 1simportant, especially as biopestictdes strive to replace conventional pesticides For example, the largest pesticide market m the United States is cotton. It accounts for 35% of the msectictde market, with about $850 million m sales m 1996 (13). Bt cotton has the potential to be a challenge to conventional fohar sprays for control of the budworm/bollworm complex. The cost of Bt cotton, including Tracer (DowElanco, Indianapolis, IN), Pirate (American Cyammid, Pearl River, NY), Proclatm (Merck, Whitehouse Station, NJ), and Confirm (Rohm and Haas, Philadelphia, PA), however, averages about $34/acre. Conventional insecticide treatment averages around $10,20/acre. Given the current price differential, Bt cotton is not likely to replace conventtonal msecticides. In another example, btological control has been shown to be effective on Canada thistle, leafy spurge, the knapweeds, St. Johnswort, musk thistle, and other weeds. Biologtcal control through beneficial insects (e-g , Canada thistle stem mnnng weevtl and musk thistle rosette weevil), however, 1sprtced substantially higher than, say, atrazme. In 1996, the price of atrazme (a common herbicide used to control these weeds) was approx $3.90/acre, and btologtcal control is priced at about $70.00/acre (13). Thus, biological control will not likely to replace atrazme or one of the other triazme herbicides m the near future,

2.2. Government Commodity and Conservation Programs Federal commodity and conservation programs affect agricultural biopesticide use, m part, through the amount of acres planted. Past commodity programs were designed primarily to provide price and income protectron for

Policy Influences on Biopesticides

61

farmers. Land set-aside requirements helped restrict supply and increase commodtty prices (14). Although the Federal Agriculture Improvement and Reform Act (FAIR) of 1996 eliminated those set-aside requirements, the reauthorized Conservatton Reserve Program, designed for envtronmental objectives, pays producers to keep acreage in conserving uses, rather than in production. Fewer acres planted generally implies less biopesticides applied. Other federal agricultural conservatron programs influence productton practices on planted acreage, which in turn wtll affect biopesticide use. 2.3. Agricultural Chemical Regulations: Implications for Biopesticide Use Pesticide regulation m its modern form began with the enactment of the Federal Insecttcide, Fungicide, and Rodenticide Act (FIFRA) m 1948. Under this mandate, Congress required that all chemicals for sale m interstate commerce be registered against the manufacturers’ claims of effectiveness. The law also required manufacturers to indicate pesticide toxicity on the label. Congress amended FIFRA in 1954, 1959, and 1964, but, in practice, pesticide regulatton by 1970 meant efficacy testing and labeling for acute (short-term) toxicity. Pesticide regulation passed mto a new phase with the 1972 amendment to FIFRA, and the transfer of regulatory jurisdiction to the EPA. Under this new regulatory regime, Congress gave the EPA the responsibility of reregistermg extstmg pesticides, examining the effects of pesticides on fish and wtldlife, and evaluating acute and chronic toxicity In the 1988 amendment to FIFRA, pesticide producers were required to demonstrate, wtthm 9 yr, that all pesticides registered before November 1984 meet current standards (4). Pesticides are also regulated by various provtsions of the Federal Food, Drug and Cosmetic Act (FFDCA). Under the FFDCA, the EPA establishes the maxtmum allowable level (tolerance) of pesticide residues that can be present on foods sold m interstate commerce, and the Food and Drug Admimstratton (FDA) momtors food and feed for pesticide residues. In 1996, Congress passed the Food Quality Protection Act (FQPA), which was intended to update and resolve inconsistenctes in the two major pesticide statutes: FIFRA and FFDCA. The major components of the FQPA address the issues of settmg a single, health-based standard (i.e., a reasonable certainty of no harm) for all pesticides m all foods (although benefits can continue to be considered in certain instances when setting standards), provtdmg special protection for infants and children, regulatory relief for minor use pesticides, expediting approval for safer (reduced-risk) pesticides, requnmg periodic reevaluatton of pesticide registrations and tolerances, and reauthorizing and increasing registrant fees to fund such reevaluattons, establishing national umformtty of tolerances unless States petition for an exception, and mandating

62

U-1

the dtstrtbutton of mformatton m grocery stores on the health rusks of pesttcldes and how to avoid such rusks (24a,Z5,16). The crttical components of the FQPA, as far as btopestictdes are concerned, deal with expediting the review of mmor-use pesticides and expediting the approval of reduced-risk pesttcides. Both sections of the legrslatton should serve to accelerate the development and commercialtzatton of new btologtcal approaches to pest control EPA 1s giving hrgh prtority to rmplementmg the Minor Use Provtslons of FQPA. It has created a new program dedicated solely to coordmating minor use Issues withm the Office of Pesttcrde Programs The defimtton of a minor-use crop has been determined to be a crop produced on fewer than 300,000 acres, or a major crop (a crop grown on more than 300,000 acres) for whtch the pesticide use pattern 1s so limtted that revenues from expected sales ~111 be less than the cost of regtstermg the pesttclde, and there are msuffictent efficactous alternatives for the use, alternatives pose greater risks, the mmor use is stgmficant m managing pest resistance, or the minor use plays a stgmficant part m integrated pest management (IPM) The first part of the definition means that all but 26 of the 600-plus crops produced m the United States are minor crops. EPA will consider every crop m the United States to be a minor crop, except for almonds, apples, barley, canola, carrots, corn (field and sweet), cotton, grapes, hay (alfalfa and other), lettuce, oats, oranges, peanuts, pecans, popcorn, rice, rye, snapbeans, sorghum, soybeans, sugarcane, sugarbeets, tobacco, tomatoes, sunflowers, and wheat. Provtstons intended to help preserve the availabthty of minor use pesticides include expedrting the review of data submitted m support of mmor uses, granting time extenstons for submtttmg data on minor uses, and giving those who invest m data development for minor uses addrtronal exclusive rights to use of the data to support registration. The mmor use program at EPA, m conJunctton wtth a similar program at the USDA wtll coordinate decrstons on minor use issues m consultation with growers. A revolving grant fund is authorized at the USDA to fund the generation of data necessary to support minor use regtstratton. The Department of Health and Human Services 1s authorized to fund studtes in support of registration or reregistration of minor use pesticides that are important for public health purposes The Reduced-Risk Pesticide Inmattve 1s designed to encourage the development, regtstratton, and use of new pesticide chemicals, which would result m reduced risks to human health and the environment, compared to existing alternatives. In 1995, the average amount of time tt took to register a new conventional pesticide was 38 mo; the new reduced rusk pesticides take, on average, 14 mo (17). Since 1993,29 new chemtcal submtssrons have been received by EPA as reduced-risk pesticide candidates. Of the 29, 17 met the reducedrisk crtterta for expedited review Nme of those 17 have been registered

Policy Influences on Biopesticides

63

In November 1994, EPA established a separate division m the Office of Pestrcrde Programs, the Bropesticides and Pollution Prevention Division, to encourage the development of reduced-risk pestictdes, and to manage the regrstration and reregistration of biopesticides Biopesticides are defined by EPA to include naturally occurring and genetically engineered microorganisms, genetically engineered plants that produce then own pesticides (such as crops that produce the msectrcidal proteins from the Bt bacteria), and naturally occurrmg compounds, or compounds essentially identical to naturally occurrmg compounds, that are not toxic to the target pest (such as pheromones) EPA approved 14 new biopesticide active ingredients m fiscal year 1995 and 10 m fiscal year 1996, which represents over one-third of new active mgredients registered rn those years EPA also issued Reregistration Eligibrhty Decision documents for eight bropesticides. FQPA explicrtly recognizesthe importance ofieduced-risk pesticides and supports expedited review to help these pesticides reach the market sooner and replace older and potentially riskier chemicals. The new law defines a reducedrisk pesticide as one which “may reasonably be expected to accomplish one or more of the followmg: reduces pesticide risks to human health; reduces pestrcrde risks to nontarget organisms; reduces the potential for contammation of valued, environmental resources,or broadensadoption of IPM or makesit more effective.” Other statuteswith the potential to affect pesticide use include the Clean An Act, Clean Water Act, Safe Drinking Water Act, Coastal Zone Management Act (CZMA), and the Endangered Species Act. The Water Quality Act of 1987 (sec. 3 19) and the CZMA address nonpomt somces of pollution, such as those from farm fields These are discussed in detail in ref. 18 4. Effects of Policies on Development and Use of Pesticides In the agricultural sector, pohcy influences on pesticides take two forms. Fn-st, pohcy affects the choice of production practice, which will be more or less pesticide-intensive, depending on the policy. Second, government policy directly impacts the development of new pesticides and pest-management practices Each of these issues IS discussed m turn. 4.1. Direct and Indirect Impact of Government Policy on Use of Pesticides Government regulation attempts to mitigate the adverse effects of pesticide use Prmcipally because of concerns over the environmental problems associated with pesticide use m production agriculture, farmers, with some mvolvement on the part of government rn the form of cost sharing and educational and technical assistance, are experimentmg with a myriad of new and traditional tools, materials, and practices Short-term productivity enhancement and

64

Un

chemical use efficiency are the major goals m some of the systems; envn-onmental risk reduction and the long-run sustamabihty of farming are more promment m others To help mmimrze the trade-off between lower pesticide use and reduced net farm income, IPM, precision farmmg, and cultural and biological pest and nutrient management methods are among options encouraged by some. The first two attempt to increase efficiency of chemical use, the latter two attempt to avoid chemical use altogether. Pest scoutmg, economic thresholds, and other tools to help the farmer determine when to make pesticide applicattons, which pesticides to use, and how much to use, have been developed for decades, and expert systems and other decision-support systems to integrate these elements are emerging Prectsion farming and herbicide-resistant bioengineered crops are efficiency technologies that aim at reducing pestictde use that are Just now being developed and commerctahzed. Scoutmg and threshold use is widespread m specialty crop production (19) Half of the US fruit and nut acreage, and nearly three-quarters of the vegetable acres m the surveyed states, were scouted for insects, mostly by professional scouts. Growers reported using thresholds as the basis for making pesticide treatment decisions on virtually all of these scouted acres. Potato growers reported that 85% of their acreage was scouted and thresholds were used m making nearly three-quarters of their msecticide apphcatton decisions. A number of alternative productton practices, including crop rotation, conservation tillage, alterations m planting and harvestmg dates, trap crops, samtation procedures, irrtgatton techniques, fertihzation, physical barriers, border sprays, cold-an treatments, and habitat provision for natural enemies of crop pests, are now bemg relatively more extensively used for managing crop pests. Their diffusion IS expected to grow more widespread. These alternative production practices work by preventmg pest colonization of the crop, reducing pest populations, reducing crop injury, and enhancmg the number of natural enemies m the croppmg system (20). Crop rotation is one of the most important of these techmques that is currently m widespread use. Over half of the corn and soybeans were grown m rotation with each other during the mid- 1990s (22). Farmers rotating corn with other crops used msectictdes less frequently than did those plantmg corn 2 yr m succession (11% vs 46%). Corn IS often grown as a monocrop in areas that have high demand for livestock feed, and where climatic restrictions limit the soybean harvest period (8). Crop rotation is much less prevalent for cotton, however, which has among the highest per acre returns of the field crops grown m the United States, and less than one-third of the cotton producers use this technique.

Policy Influences on Biopesticides

65

4.2. Impact of Government Policy on Development of New Pesticides Research and development expenditures are an important factor affecting the development of new pesticides. There is a fairly extensive literature exploring the relationship between research and development expenditures on a good or service and the use of that good or service (22-24). Basically, research and development expenditures m basic mdustrtes, such as the chemical industry, are conditioned by the present value of expected net revenue. That is, research and development expenditures are made with the expectation that a profit for the firm will result. The net present value is a function of costs, such as research and development expenditures, and regulatory costs of getting a new pesticide registered, as well as the revenue generated from the sale of the pesticide. In this setting, and as apparent from the foregoing discussion, pesticide pohcy is only one, albeit an important one, of the factors affecting pesticide use, and hence the development of btopesttcides. Any government pohcy affecting relative factor prices, price responsiveness, expected returns from chemical use, conservation, and technology development and adoption will impact the development of new pesticides. The two major statutes, FIFRA and FFDCA, instruct regulators to weigh the benefits of pesticide use against unreasonable risks. This balancing process has been characterized so that the use of the term “unreasonable risk” implies that some risks will be tolerated under FIFRA; it is clearly expected that the anttctpated benefits will outweigh the potential risks when a pesticide is used according to commonly recognized, good agricultural practice (25). A study of the impact of pesticide regulation on innovation and the market structure in the United Statespesticide industry shows that pesticide regulatton in the United States has encouraged the introduction of fewer, yet less toxic pesticrdes (#j.* The 1972, 1978, and 1988 amendments to FIFRA require that new and existing pesticides meet strict health and environmental standards. Requirements for pesticide registration wtth the EPA include field testing, which can include up to 70 different types of tests that can take several years to complete and cost mtlhons of dollars. They consist of toxicologtcal studies, a two-generation reproduction and teratogenicity study, a mutagenicity study, oncogemcity studies, and chronic-feeding studies. The toxicological studies include acute (immediate), subchronrc (up to 90 d), and chrome (long-term) effects. Other tests are used to evaluate the effects of pesticides on aquatic systems and wildlife, farmworker health, and environmental fate. Recent *High toxlclty pestlcldes are those that belong to Class I acute toxlclty (indicated m the label), or are chrotucally toxic to humans, or to fish and wildlife. Lower toxlclty pesticides are all others (4)

66

U-i

estimates suggest that research and development of a new chemical pesticide (including the testing indicated above) costs between $50 and 70 million and takes 11 yr (4). As a consequence of the regulation requirements, pesticide firms refocused their research away from persistent and toxic pesticides. The number of pesticides with chronic (long-term) toxicity dropped by 86 between the 1972 and 1976 and 1987 and 1991 periods, and lower toxicity pesticides account now for more than half of pesticide sales. A 10% increase in testing costs is associated with a 2.8% increase in the proportion of less-toxic pesticides registered. In 1996, the Office of Pesticide Programs of the EPA registered 22 new pesticide active ingredients, more than half of which were considered reduced-risk pesticides. These decisions included the approval of 10 biopesticides and 12 new chemicals, which include three reduced-risk chemicals (26). The biopesticides include Bt Cotton (Monsanto), I-octen-3-01 (Armatron), Jojoba oil (IJO Products), Bt (CryMax) (Ecogen), Myrethecium verrucaria (Abbott), meat meal (Lakeshore Enterprises), red pepper (Lakeshore Enterprises), Verticillium lecanii (Abbott), NK Bt corn (Northrup King), Monsanto Bt corn (Monsanto), and Lavandin oil (S.C. Johnson). Pesticide regulation has also had undesirable consequences. Regulation discouraged new chemical registrations: The number of new pesticides registered by the EPA in 1987-l 99 1 was half that of 1972-l 976 and each 10% increase in pesticide regulatory costs caused a 2.7-% reduction in the number of new pesticides introduced (4). The higher regulatory costs contributed to an industry-wide increase in research spending, which encouraged some small firms to leave the pesticide industry. Pesticide regulation also encouraged firms to focus their research on pesticides used in larger crop markets, such as corn and soybeans, abandoning minor-crop markets, such as horticultural crops. The decline in new registrations of chemical pesticides suggests that there are market opportunities for biopesticides and genetically modified plants. These products are not only environmentally preferable, but also less costly to develop and register than chemical pesticides. Thus, it has been estimated that the average cost of developing a biopesticide ranges from $3 to 5 million vs $50 to 80 million for the development of conventional pesticides (27). Such new products, however, as noted previously, are only effective against a narrow range of pests. The 15 largest (in terms of retail sales) agricultural chemical companies are developing biopesticides, including pheromones, bioinsecticides (viruses), botanical extracts, soybean seed, corn and sorghum seed, microbial products, Bt manufacturing, microsponge formulation, and gene insertion. Cropper et al. (28) examined EPA’s Special Review Process for pesticides between 1975 and 1989, to determine whether the decision to cancel or continue the registration of pesticides could be explained by the risk and benefits associated with pesticides. Cropper et al. estimated a trade-off of $35 million

Policy

Influences

on Biopesticides

67

in producer benefits per cancer avoided among pesticide applicators. More recently, Cropper et al. (28) estimate a trade-off of $72 million per cancer avoided among pesticide applicators and $9 million per cancer avoided among consumers (in 1986 dollars). Abler (30) argues that these figures are too high, because Cropper et al. calculated the benefits at existing prices, not considering the effect of pesticide restrictions on producer prices. Abler argues that producers could even gain from pesticide restrictions if output prices increased enough. On the other hand, higher output prices caused by restrictions on pesticide use have not been empirically documented. 5. Conclusions Biopesticides developed and used in the future will emerge against the backdrop of the environmental effects associated with the use of conventional pesticides and government policies designed to control these effects. In the final analysis, farmers’ choices on pesticides will be influenced by the prevailing costs and benefits of conventional pesticides and their alternatives, including biopesticides. The outlook for pesticide use is complicated, though some directions can be perceived. There are a number of factors that will serve potentially to impact pesticide use, which in turn will affect the development of biopesticides. These include pesticide regulation, the FAIR Act, the crops planted, the management of ecologically based systems, and consumer demand for green products. 5.1. Pesticide Regulation Pesticide regulation will continue to exert a major influence on pesticide use and the development of pesticides and pest management alternatives in the United States.The number of pesticide-active ingredients for sale in the United States has decreased by 50% since 1989 because of EPA’s reregistration process (31). Moreover, regulatory changes involving the removal from the market of pesticides, which had been previously registered, because of evidence on unacceptably high health hazards from occupational exposure, may also undermine the confidence that farmers have in the safety of pesticide use (32). Finally, implementation of the FQPA of 1996 will impact pesticide registrations in general and biopesticides in particular. The FQPA is designed to expedite the review of minor-use pesticides and the approval of safer pesticides. Both sections of the legislation should serve to accelerate the development and commercialization of new biological approaches to pest control. Although the legislation does not expressly recognize the presumption of biologicals for the expedited review category, the provision will assist in promoting new research and development activities (17).

68

Uri

5.2. The FAIR Act Short-term effects of the FAIR Act on blopestrcideuse stemfrom greater flexlbllity provided to producers through elimination of baseacreageand set-asides.Elimlnation of base acreage will faclhtate rotations, which could reduce insecticide use. Without the concern of mamtaming base acreageto receive federal deficiency payments,one would expectproducers to plant crops for which returns are higher (e.g., corn rather than wheat), where producershave suchoptions. Hence,blopestlcide use will change based on how and where the mix of crops will change. For example, If producers plant more corn, a more chermcally intensive crop, rather than wheat, which generally requires lesschemicals,one would expectchemical use to increase (13) Elimination of set-asides,other acreagereduction programs, and a potentially smaller Conservation Reserve Program could result m increased planted acreage, with resulting increasedblopesticide use. Set-asides,however, have been relatively low, If not zero, for several program crops recently, sothe increasein planted acreage would not be dramatic. In net, greater biopestlclde use would be expected if more chemically intensive crops areplanted on existing acreageand greater acreageoverall ISplanted, but greater crop rotations would curb such growth (26). In the long-term, input use will hinge on the relative marginal productlvlty and cost among labor, pesticides, fertilizer, and other factors, which the FAIR Act will not alter. As nominal prices decline from 1995-1996 peaks, and real prices are anticipated to continue to decline, there will be less incentive to apply inputs whose value of the marginal product does not increase. Although federal support will fall, market demand 1sexpected to keep commodity prices relatively high by historical standards. Thus market mcentlves will replace federal mcentlves regarding apphcatlon of chemical inputs 5.3. Crops Planted The USDA projects that crop acreage of eight major crops will rise between 5 and 10% by the year 2005 from 1995 levels (33).* Both corn and wheat acreage are expected to rise by about 10 million acres each; that of cotton 1s expected to decline by about 3 mllhon. Hence, although the changing mix of crop acreage complicates a proJection of pesticide use, it seems likely that, from increased planted acreage, blopesticlde use will continue to rise, if other influences are held constant (13). 5.4. Ecologically Based Management Systems The USDA announced several years ago that switching to an ecosystembased approach for managing natural resources 1samong its major agricultural prloritles for the 1990s (34). The new approach, which 1sto be adopted gradu*Cropsarebarley,corn,uplandcotton,oats,rice,sorghum,soybeans,

and wheat

Pohcy Influences on Biopesticides

69

ally, adds the goal of mamtammg or improving the condttton of the land as the context for providing sustamable levels of goods and services. Ecosystem management 1spartly based on the emergmg research on btodiverstty, ecosystems, and environmental accountmg from new scientific disciplines, such as conservation biology, landscape ecology, and ecological economics. The impacts of species loss on crop-breeding programs, and more complete accountmg of the costs of food and fiber productton, are some of the issues that are addressed. Much of the imttal application of this approach has been for national forests and other pubhc resources, but tts potential use for crop productton systems 1s also being explored The National Research Council (NRC) (35) has published the results of a study to examme ecologtcally based pest management practices for agriculture and forestry. The NRC report concludes that pest resistance and other problems created by pesticide use has created a need for an alternative approach to pest management that can complement and partially replace current chemically based pest-management practices. Ecosystem-basedpest management is the approach that was recommended. The USDA’s Forest Service adopted an ecosystem management philosophy in June 1992, and this approach has been influencing the development of forest management plans m the Pacific Northwest and other areas. For example, ecosystem design-arrangmg landscape structures m the watershed to support species biodiversity, as well as timber production and recreation-was used in the recently developed forest management plan for a 30,000-acre watershed in Mt. Hood National Forest (31). A number of stateshave begun to examine ecosystem-based pest management m specific agricultural cropping systems. Maine’s Agricultural and Forest Experiment Statton, for example, has recently reported results from the first 4 yr of its pioneering industry-supported ecosystem proJect on sustainable potato production (36). Various states and regions also have some ecosystem research underway, mcludmg some federally funded IPM proJects, which are lookmg at btologtcal alternatives, as well as public-private efforts at the local level, such as the Chesapeake Bay watershed project, and the BIOS proJect for California almond growers. 5.5. Consumer Demand for Green Products and the Market Response The market for food products with a green label, such as certified organic and IPM, has been growing m the United States.Although organic food products only account for about 1% of total retail food sales at the present time, overall orgamc sales reached $2.8 billion in 1995, and have increased over 20% annually since 1989 (37). Orgamc foods are becoming more widely avatlable to United States consumers through the growing number of large natural

70

Uri

food stores, mainstream supermarkets, and direct outlets, such as farmers markets. A consistent set of national standards for organic productton and processmg, which 1s currently being developed by USDA, IS expected to enhance consumer confidence in the United States. Development of these standards was required by the Orgamc Foods Productton Act, which was passed by Congress as part of the Food, Agriculture, Conservation, and Trade Act of 1990. This legtslatton requn-es that all except the smallest organic growers will have to be certified by a state or private agency accredited under the national standards. The National Organic Standards Board, which was appointed by the USDA to help implement the provisions of the act, currently defines organic agriculture as a sustainable production management system that promotes and enhances biodiversity, biological cycles, and so11 biological activity. It 1s based on mmtmum use of off-farm production inputs and on practices that mamtam organic Integrity through processing and distributton to the consumer (38). In tandem with the growth m demand for food with a green label is the demand for food with less pesttctde residue. Btopesttctdes are viewed as being safer than chemical products (27). Early m then development, biopesticide companies, mcludmg Biosys, Consep, Ecogen, and Mycogen, were forced to concentrate then marketing efforts on niche markets (prtmarily vegetables and fruits) because products did not have the price/performance (efficacy) characteristics necessary to compete m the larger pesticide markets They are now estabhshmg themselves in major markets lrke cotton and corn. They found the less-competitive niche markets were a sheltered corner where they could mature. These small markets gave biopesticides a sales base for products that did not have the attributes desired m larger markets. Biopesttcide companies have since invested heavily m upgrading products, so they can move beyond then- niche market base mto major pesticide markets (39). Biopesttcide companies have made substantial improvements m recent years and have become more competitive with conventional chemical companies Most btopestlctdes have improved prtce/performance charactertsttcs as a result of improved technology, which has resulted m better efficacy and lower costs. For example, recently developed Bt products, mcludmg Bt (CryMax) (Ecogen) and Mattch (Mycogen) have more concentrated toxins, give more consistent performance, have longer residual formulations, and contam relatively more potent Bt strains. These improvements have allowed the companies to lower prices, Finally, companies have reposmoned products m markets where the products add value, and have focused on educating growers on the use of the products. For example, nematodes have been repositioned to address selected citrus

Policy

Influences

71

on Biopesticides

and ornamental markets. These sorts of reposltloning, markets, should continue m the future, as companies products are most effective.

focusing recogmze

on selected where their

References 1 Carlson, G A and Castle, E. N (1972) Economics of pest control, m Pest Control Strategres for the Future, Commlttee on Pest and Pathogen Control, National Academy of Sciences, Washmgton, DC, pp 72-l 03 2 US Congress, Office of Technology Assessment (1995) Bzologzcally Bused Technologzesfor Pest Control, OTA-ENV-636, US Government Prmting Office, Washmgton, DC 3 RIdgeway, R , Inscoe, M , and Thorpe, K (1994) Bzologzcally Bused Pest Controls Markets, Industrzes, and Products, US Department of Agriculture, Agl ~cultural Research Service, Washington, DC 4 Ollmger, M and Fernandez-CorneJo, J (1995) Regulatzon, lnnovatzon, and Market Structure zn the US Pestzczde Industry, Agricultural Economic Report 7 19, US Department of Agriculture, Econom.lc Research Service, June 5 LandIs, D A and Orr, D B (1996) BIologIcal control approaches and apphcatlons, Electronzc IPA Textbook (Radchffe, E. B and Hutchlson, W D., eds ), Unrverslty of Minnesota and the Consortium for International Crop Protection, Mmneapolls, MN 6 Poritz, N (1996) Biological control of weeds, m Bzologzcal Control of Weeds 1996 (Pontz, N , ed ), Montana State Umversity Press, Bozeman, MT, pp l-21 7 Hunter, C (1994) Supplzers of Beneficzal Organzsms zn North America, PM 9403, Cahforma Envlronmental Protectlon Agency, Department of Pestlclde Regulatlon, Sacramento, CA. 8 Edwards, C R and Ford, R. E. (1992) Integrated pest management In the corn/soybean agroecosystem, m Food, Crop Pests, and the Envzronment (Zalom, F G and Fry, W E , eds ), American Phytopathologlcal Society, St Paul, MN 9 Environmental Protection Agency (1995) EPA Issues condltlonal approval for full commercial use of field corn plant pestlclde targeting the European corn borer, EPA Press Release, Washington, DC 10 Zalom, F and Fry, W (1992) Food, Crop Pests, and the Envzronment, American Phytopathologlcal Society, St Paul, MN. 11. Nleblmg, K (1995) Agricultural biotechnology compames set their sights on multi-bllhon $$ markets, Genet Eng News 12, I,2 12 Hoban, T and Kendall, P. (1993) Consumer Attztudes about Food Technology Project Summary, Extension Service, North Carolma State Umverslty, Raleigh, NC 13. Economic Research Service (1997) Agrzcultural Resources and Envzronmental Zndzcators, US Department of Agriculture, Washmgton, DC 14 Aspelm, A L (1984) Pestzczde Industry Sales and Usage I992 and 1993 Market Estzmates, BEAD, Office of Pestlclde Programs, US EnvIronmental Protectlon Agency, Washington, DC, June

Uri

72

14a. (1996) Pesttctde and Toxzc Chemtcal News, August 21, FCN Publrshmg, Washmgton, DC. 15 US Envtronmental Protectton Agency (1996) Major issues tn the Food Qualzty Protectton Act of 1996, Preventron, Pesttctdes, and Toxic Substances, Washmgton, DC, August 16 Jaemcke, E (1997) The Myths and Realtttes of Pesttctde Reductton, Henry A Wallace Instttute for Alternattve Agriculture, Beltsville, MD 17 US Environmental Protection Agency (1997) 1996 Food Quality Protectton Act Implementatton Plan, Preventton, Pesticides, and Toxic Substances, Washmgton, DC 18 (1996) Farm Chemrcals Handbook, Meister, Wtlloughby, OH 19 Vandeman, A , Fernandez-ComeJo, J , Jans, S , and Lm, B H ( 1994) Adoptton of Integrated Pest Management tn US Agrtculture, AIB-707, US Department of Agrtculture, Resources and Technology Divrsion, Economic Research Service 20 Ferro, D N (1996) Cultural controls, m Electrome IPM Textbook (Radcliffe, E B and Hutchison, W D , eds ), Umversity of Mmnesota and Consortium for International Crop Protection, Mmneapohs, MN 21 Lm, B H and Delvo, H (1994) Pest Management Practices on 1993 Corn, Fall Potatoes, and Soybeans, Natural Resources and Environment Drvision, Economic Research Service, US Department of Agriculture, Washmgton 22 Kahn, A (1971) Economtcs ofliegulatton, Wiley, New York 23 US Congress, Office of Technology Assessment (1986) Technology, Publzc Poltcy, and the Changing Structure ofAmerican Agrtculture, OTA-F-285, US Govemment Prmtmg Office, Washmgton, DC, March. 24 Scherer, F M. (1980) Industrial Market Structure and Economtc Performance, Rand McNally, Chtcago 25 National Research Council (1993) Pesttcides tn the Diets ofInfants and Children, National Academy, Washington, DC 26 Office of Pesticide Programs (1996) Office of Pestrctde Programs Annual Report for 1996, Environmental Protectton Agency, Office of Pesticide Programs, Washington, DC, November 27 Anonymous (1996) Future development of biopesticides expedited by FQPA Pestrctde Toxic Chem News 24, 19-20 28 Cropper, M L., Evans, W N , Berardi, S J , Ducla-Soares, M M , and Portney, P R (1992) Determinants of pesticide regulatton a statistrcal analysts of EPA decision makmg, J Poltt Econ 100, 175-197. 29 Cropper, M L., Evans, W. N., Berardi, S. J., and Ducla-Soares, M. M. (1992) Pestictde regulation and the rule-making process, Northeastern J Agrtcultural

Resource Econ 21,77-82 30 Abler, D G (1992) Issues m pesttcide policy Northeastern J Agrrcult Resource Econ 21,93,94 3 1 Pease, W S , Liebman, J , Landy, D , and Albrtght, D (1996) Pesttctde Use tn Caltfornta. Strategies for Reducing Environmental Health Impacts, Calrforma Pohcy Seminar, Umversrty Environmental Health

of Califorrna,

Berkeley, Center for Occupatronal

and

Policy Influences on Biopesticides

73

32. Bender, J. (1994) Future Harvest Pesttctde-Free Farmtng, Umversity of Nebraska Press, Lmcoln 33. World Agricultural Outlook Board (1997) Agrtcultural Baselzne Projectzons to 200.5, Refecttng the 1996 Farm Act, WAOB-97-1, US Department of Agrtculture, Washmgton, DC, February. 34. Comanor, S and Gelburd, J (1994) Ecosystem approach to resource management, Agrzculturai Outlook, AO-204, Economrc Research Service, US Department of Agrtculture, January-February 35 National Research Counctl (1995) Ecologtcally Based Pest Management New Soluttons for a New Century, National Academy, Washington, DC. 36 Umverstty of Mame (1996) The Ecology, Economtcs, and Management of Potato Cropptng Systems’ A Report of the Ftrst Four Years of the Maine Potato Ecosystem Project, Bulletm 843, Mame Agricultural and Forest Expertment Statton, Orono, ME, Aprtl 37 Natural Foods Merchandiser (1996) Widening market carries orgamc sales to $2.8 billion m 1995, Nat Foods Merchandiser 17, 5-7 38 Ricker, H S (1997) Nattonal organic program: status and Issues, m Proceedtngs of the Thzrd National IPMSympostum/Workshop, US Department of Agriculture, Economic Research Service, Washmgton, DC, pp. 63-65. 39. Btosctence Securmes (1996) The Outlook for Bzopesttctdes, Btoscience Securtties, Inc , Ormda, CA 40 US Department of Agriculture (various) Agrzculturaf Chemical Usage Fruzt Crops Summary, Nattonal Agricultural Stattstics Service and the Economtc Research Servwe, Washmgton, DC 41 US Department of Agrtculture (various) Agrzcultural Chemtcal Usage* Vegetable Crops Summary, National Agrtcultural Stattsttcs Servtce and the Economic Research Service, Washmgton, DC

II BIOFUNGICIDES

Commercial

Development

of Biofungicides

Raphael Hofstein and Andrew Chapple 1. Introduction The commercial development of biofungicrdes received a significant boost m recent years, primarrly because of impressive progress in the rsolatron and charactertzatron of novel strains of microorgamsms that can fulfill the mam characteristics of a brofungrctde, which are the consistent suppression of pathogens under field conditions, and easy mass production in standard fermentation facrlrties (1,2). Progress was imtially slow because of the mherent properties of the natural isolates, most of which are obligatory parasites that require the presence of a host for propagation (3). With the meteoric development m recent years of new tools m industrial molecular biology, and, more specifically, m the area of fermentation technology, srgmficant progress has been made m the commercral development of cost-effectrve brofungrcrdes. The conceptual and methodological consideratrons leading to a commercially vtable product for biological control of fungal diseasesIS the theme of the followmg discussion. The arena of btofungicide development has been classified mto three categories: sorlborne pathogens; foliar diseases;and postharvest rots durmg storage All three have been extensively documented over the past decade (4.Q. However, only a few brofungrcrdes have been successfully promoted through regrstratron, and even fewer have managed to cross the barrier between basic and commercral development. Those that have managed to overcome the hurdles of actually becoming a commercral bropesticrde have included in then development, right from the inception of an research and development program, a crmcal analysts of market needs and potential, as well as financtal conslderatrons, such as cost of goods (9). From Edited

Methods UJ B~ofechnology, vol 5 Bmpestmdes by F R Hall and J J Menn 0 Humana Press

77

Use and Delwery Inc , Totowa, NJ

78

Hofstein and Chapple

Historically, the list of commercially viable blopestlcldes Included almost exclusively Baczllus thurzngzenszs (Bt)-based products for the control of lepldopteran pests (20). The hst of Bt products became very lmpresslve m recent years, primarily because of DNA recombinant technology, and products entermg the market have proven to be very effective against a whole array of economically important pests (21). However, unlike Bt products that rely on an inert protemaceous crystal toxin as the active ingredient, most of the blofungicides under dlscussion require an Intact, hvmg cell for function (12). Moreover, since Bt-based products are relatively inert, they can be readily incorporated mto a user-friendly formulation. Intact-cell-dependent blofunglcldes, on the other hand, require very delicate and rather sophisticated approaches m the design of a commercial formulation, one that will ensure product stablhty durmg storage (I.e., shelf life), as well as rellablhty during apphcatlon (23). As discussed by Baker (2415) and Baker and Scher (26), It IS well recognized that a substantial amount of biological control occurs m nature, and that increased or even spectacular suppresslon of plant disease can occur naturally m some agricultural situations These events, though well documented and potentially useful for the isolation of blologlcal control agents, did not result m the development of many commercially viable products. Likewise, the delicate equlllbrlum between saprophytlc and parasitic mlcroflora on the phylloplane has been dlscussed by Dubos (17), who brilliantly made reference to blocenosls as the backbone of such an equlllbrmm Again, because of the natural balance between antagonists and pests, disease may be suppressed. However, m order to Implement these features mto economically feasible control of fohar diseases, a better understanding of the dynamics of antagonist-pathogen interactions 1s required. Moreover, the antagonist has to fulfill certain criteria before it can be considered a legitimate candidate for promotlon to a commercial blofunglclde Several reviews have addressed aspects of maxlmlzmg the chances of developing a naturally occurring organism (18-20). The criteria for a successful blofunglclde can be summarized as follows. Effective suppression of the fungal pathogen before It causes economically important damage to a crop Consistent performance under authentic condltlons of crop management and the crop environment Adaptation to exlstmg integrated pest management (IPM) schemes of disease control Price competitiveness with other means of combatmg the same target pest Compatlblhty with other chemical or blologlcal treatments targeting other pest(s) Adaptation to commonly used farm agronomic methodologies Preservation of naturally occurrmg beneticlal antagonist(s) of related or nonrelated pests User and environment frlendlmess

Commercial

Development

of Biofungicides

79

The hkehhood for successful development of a commerctally vtable btofungtctde 1smaxtmtzed once all these crtterta are fulfilled. To obtain a reliable program that uses such products as prominent building blocks of a comprehensive IPM approach, it 1scrucial, right from the onset of the research program, to evaluate the adaptation of a naturally occurring orgamsm mto an economttally feasible strategy of pest control. The prmctpal drawback to developing btofungtctdes is that it is very difficult to harness the organisms to mdustrral processesof mass productron, especially fermentation m sterrle vessels, as well as incorporatton mto user-friendly formulations (21). Clear understanding of the delicate balance between antagonistic mtcroorgamsms and crop pests could eventually lead to manipulation of ecosystems to enhance crop protectron (22-24). Regarding sotlborne and foltar pathogens, the impact of naturally occurring saprophytes on disease suppression is well documented. However, tt 1s now established that it IS insufficient to rely on mnocuous mrcroflora to ensure the high degree of plant protection required m modern agriculture. The balance has to be shifted m a direction antagonistic to the fungal pathogen, or else to impose quantttattve advantages to the antagonist. The latter is the tmpetus for seeking an alternative approach by vn-tue of the development of commercial biofungtcides, and this can be most eastly attained by directed mtroductton of a selected antagonist into a given btocenosts, as prevtously proposed m excellent review articles (see refs. 25 and 26). 2. Steps in the Development Process The development of a cost-effective btofungtctde is an intricate undertakmg. The various steps involved in the process are m a quote rrgid order, which IScritical tf failure IS to be avoided. The authors describe a sequence of steps m the development strategy based on experience reflecting successes,but also failures, m several campaigns. These steps are depicted m a rather rtgorous and critrcal fashion. 2.1. Screening for Naturally Occurring Microorganisms The best source for potentially effective biofungrctdes is the site of natural interactton between a pathogen and tts antagonist. For instance, hypovirulent strains of a plant parasite were sought in sites of declining disease, or, in the case of soilborne pathogens, screening for mtcrobtal antagomst m suppressive soAs (27). In the case of hyperparasttic antagonists such as Ampelomyces quzsqualu, the hyperparastte of various powdery mildews (PMDs), the pathogen could serve as a carrier, and, therefore, colonies of the pathogen can actually become an ideal site for screenmg (28-30). Several criteria should be used as guidance from onset of the screening process, including extent of antagonism m a bioassay that best resembles the commercial envnonment (31,32), adapt-

80

Hofstern and Chapple

abtltty of the antagomsttc mtcroorgamsm to artttictal mass-productton factlttres (33), and versattltty of antagomsm to vartous spectes of the pathogen on dtfferent crops (34). 2.2. Selection of the Most Cost-Effective Fermentation Process Several methodologies have been adopted for mass productton of pesttcidal mtcroorgamsms, but the only economtcal approach currently avatlable IS one that relies on submerged fermentatton m a sealed fermentation tank. All others, including fermentation m a solid matrix, require an unacceptably long process m especially dedicated equipment. The latter has a detrimental effect on cost of goods, and therefore becomes prtce-prohtbttlve. Submerged fermentatton has proven to be the preferred technology, and one that allows for hrgh ytelds m a relattvely short period of time To attain a cost-effective process, tt IS rmperatrve to select a growth medium that rehes on mdustrtal waste products wellbalanced to supply an optimal ratio between nitrogen and carbon (35) Every development program must address such elements as low-cost nutrients and opttmizatton of the time-course. The prmcrpal author has reviewed these elements (36) describing the development of AQ 10, a btofungtctde developed for the control of PMD, based on A quuquah. 2.3. Development of Bioassays A system that can adequately simulate an authentic snuatlon m a commercoal setting 1sa crtttcal element in a proper research and development program (37). The chances of selectmg the mtcrobtal tsolate of choice to be promoted from a ubtqurtous microbial strain mto a cost-effecttve btopestictde can be maxrmtzed by a versatile assay system The btoassay has to reflect the mode of actton of the product. Although most research efforts resulted in the eluctdation of anttbrottc-producing bacteria for pathogen control that can be simulated m a Petri dish, progress made m recent years m industry, gave rise to more complex fungal agents, whose mode of action ISbased on parastttsm or competttion for space and nutrients. A preferable assay system for such agents IS the intact-plant simulator, tested m a growth chamber or greenhouse. 2.4. Development of a User-Friendly Formulation Product formulatton IS a most crtttcal aspect of the entire development program. A user-friendly formulatron has to fulfill several crrterra, mcludmg allowmg a mtcroorganism to retain and express tts pesttctdal properties; providing a stgmficant extension of shelf life of at least 6, but preferably 18 mo, at ambient conditions, so that the biofungtctde can be stored over two seasons; and allowing the active ingredient to be applied wtth extstmg appltcatton equtpment. During the course of developmg AQ 10 for commercial acceptabtltty,

Commercial Development

of Biofungicides

81

several major changes m the features of the formulation were requtred m order to fulfill the above criteria. After screening several options, a water-dispersible granular formulation proved to be most adequate for the product. It is chtefly because of this formulatton that AQ 10 reached commerctaltzatton relatively quickly Better understandmg of market needs durmg the prehmmary stage of product development may asstst in the process of tatlormg the right formulatron to each product 2.5. Establishment

of an Extensive

Field-Testing

Program

It IS cructal, and yet insufficient, to design a powerful screening broassay When adopted for the development and fine-tuning of a chemical pestictde, such bioassays often adequately predict how a new product will manifest Itself m the field; however, this ISdefimtely msuffctent m the caseof biofungtctdes. Modes of action, such as micoparasitism (e.g., Ampelomyces [38], T~rcI?odevma harzlanni 1391, and Giocladwm roseum [40l) or competttton for space and nutrients, such as in the case of, e.g., ASPIRE (Ecogen, Langhorne, PA) (41) and BIOSAVE (Ecosctence, New Brunswick, NJ) (42), for postharvest decay control, can only be determined partially in a simulator or bioassay. To demonstrate its full potential, a product must be tested m the field, and preferably in large plots, rather than the small replicated blocks commonly used for statistical analysts. Field trials are also the best stage to determme how a biofungicide can be incorporated mto an IPM program (43-46). 2.6. Preparation of a Registration Package Since biofungicides are considered envtronmentally friendly, most registration authorrttes view them as safe, and hence justify a relatively fast track of revtew (47). It 1s the authors’ experience that each case IS different, and, according to the purpose of the program and the target pathogen, the registration authorities will select the evaluation process to meet the nature of the appltcatton. For instance, biofungicides for postharvest decay control m packmghouses were relieved of the need to conduct ecological testing, since the orgamsms are expected to be applied only in a confined environment (48). Btopesticides can be exempted from tolerance requirements, and usually are relieved from any significant periods of field re-entry intervals (REI). Such elements are clear advantages of a biopesttctde, and will be obtained only when properly represented during the registration process. 2.7. Establishment of Demonstration Programs Small-plot, randomly replicated trials are part of the prerequtstte for product registration, but demonstration trials in which the product manifests itself under the authentic conditions of a commercial setting are crucial for biopesticides m

82

Hofstein and Chapple

general, and biofungicides m particular Once registered, a btofungicide can, at a relattvely early stage of its development, be subjected to such demonstration programs without having to absorb the burden of crop destruction. The benefits of such trials are threefold: first and foremost, they help the end user get acquainted with the prmciple of disease control through competition and/or parasrtism; they are also essentral for an adequate analysis of the formulation and its adaptability for a commercial setting; last, they are the only setting that allows for proper design of IPM schemes (49,50). 2.8. Design of User-Friendly Protocols Since biofungicides are a relatively new component m the arsenalof tools avallable to the farmer as part of the agronomic routine, the protocols must be very detailed, and certain issues,such as rates, application intervals, and methods of application, must be addressed cautiously. It is, conceptually, as well as techrntally, erroneous to assumethat the extrapolation from a standard protocol representmg a chemical fungicide to that of a btofungicide ISlmear. It definitely is not. A good example is the matrix of permtssible tank mrxesof AQ 10 and chemical nutrients, msecticides, and fungicides (51). Incorporatton of AQIO into an IPM program calls not only for legitimate alternation between a chemical and biological agents: In general, the two have to be mcluded m the sametank mix, otherwise, the whole program can be questioned from an economic standpoint. AQlO became commerctally recogmzed and acceptable only when a breakthrough had been made m the area of compatibility with other treatments. Compatibthty, together with fine-tuned ratesofapphcation, and optimization of Intervals, are key elements in the protocol as it 1spresentedto the end user. It must be stressedthat the primary concern of the end user is the cost-effectiveness of the product. 3. Product Development: AQ10 as a Test Case The development of A guzsqualzs mto a commercial product (AQl 0) for the control of PMD on a variety of crops is an mstructive example of the authors’ views of the learning curve related to the commercial development of biofungicides m general. It is a truism that the most difficult step for any biological control agent is that from glasshouse to the field, something very few biocontrol agents have managed. Because AQ 10 has successfully crossed this hurdle, a large part of this chapter is devoted to a description of the development trials, and, more importantly, the thmkmg behind the trials and the analysis that followed It is certainly not enough to conduct trials merely to see if it works, and at what rates, and so on. Especially for brologicals that will be combined into IPM programs, bringing the real world mto the trials and designmg assessmentand scoutmg schemes that are realistic for the end user are critical components of trial design.

Commercial Development

of Biofungicides

83

3.1. From the Glasshouse to the Field A. quisqualis has a field- and glasshouse-testing history quite at odds with the conventional academic approach. Although it started its developmental career in the laboratory and the glasshouse (52,53), the speed at which it progressed to the field was very different from that of the usual process for microbial agents. Most other agents have undergone academic laboratory and glasshouse study, and, while learned publications are prepared, passthrough a great number of generations prior to being subjected to the intended environment: the field. The latter route, namely the development of the agent with the glasshouse setting in mind, can be detrimental, because commercial application is much broader than just tailoring a product for the glasshouse industry. Such an approach may slow down commercial development, and the strain may go through even more generations before a desire to increase the range of use of the organism (or the realization that the glasshousemarket cannot sustain an on-going research program, let alone a company) causes the researchers to expand their horizons. By now the strain is very much at home in the test tube and flask, and the sort of selection pressures put on the active ingredient, the spores, in the fermenters can cause problems. The organism is first selected for an environment completely at odds with the real world-the small-scale fermenter or lab production system-and then is expected to perform in an environment where a pathogen has many advantages. In the case of PMD, these advantages include continued selection pressure by the grower for not only resistance or tolerance to any conventional pesticides, but also speed of infection and subsequent spread throughout the crop, especially in situations like the glasshouse and vineyards, where regular prophylactic sprays are the norm. AQlO bypassed some of this by being tested in the field very early. As a result, development decisions were field-driven, not laboratory- or glasshouse-driven. Not only do growers’ habits and cultures have an impact, but so does mass production in a fermentation vessel. The fermenter is analogous to a two-edged sword. Selecting for fast-growing, high-yielding strains that are tolerant to fermenter conditions may be an advantage. In the case of insecticidal nematodes or bacilli, fermenter-targeted strains may be narrowing the range of expression of the organism toward the faster-acting end of the spectrum. Taking the natural population ofA. quisqualis, there is probably going to be a normal distribution of behavior regarding the likelihood of the spore to germinate, or other characteristics. Clearly, a product for the control of PMD is best composed of spores that germinate at the least provocation, and it is presumed that the fermenter selection pressure drives the organism toward rapid response to signals for growth or germination. To put it crudely, those cells in the original inoculum that grow fastest will have a significant advantage over any slowerresponding cells, so that the progeny (i.e., the AQlO formulation itself) is

84

Hofstein and Chapple

largely made up of fast-responding members of the natural population. This may or may not act in favor of the preferred selection of the population that ought to be manifested in the authentic environment of the open field. Such considerations must be included in the design for optimal screening of the best naturally occurring strain to be selected for commercial development. The AQlO strain of A. quisqualis, which eventually became the strain of choice, was isolated from a semiarid part of Israel (54), and then transferred to other parts of the world for further field development. The advantage is that upon transfer to other locations, such as vine-growing areas of Italy, France, California, and South Africa, the spores had to adjust to less extreme, more humid environments than those of the original locale. Hence, the AQ 10 strain of A. quisqualis is probably performing reasonably well, having been taken from an environment where rapidly exploiting such conditions would be important. The PMD present in the vineyard has been under a very different set of environmental pressure in different parts of the world. Therefore, and almost as a guideline, if a replacement to the current strain of AQ 10 should be required, or if a biological control agent for downy mildew or gray mold in vines is needed (55), then the organism or strain chosen should be sought in semiarid climatic zones,where the optimal conditions for the control organism are close to those required for normal conditions of the targeted disease. Similar thinking was applied to the development of a yeast strain (Aspire) for the control of postharvest decay pathogens. The two may only be differentiated by the fact that the latter is destined for the confined environments of a packinghouse (56). It is important to note that biofungicides, such as AQ 10 and ASPIRE, which affect the pathogen via hyperparasitism and competition, respectively, are at a disadvantage whenever the environment favors the fungal pathogen, simply because the pathogen develops at an explosive, almost epidemic, rate (57). To avoid such scenarios in vines, and in order to establish a more persistent level of disease control, AQlO was applied frequently (Le., every 1O-14 d), and as part of a more comprehensive IPM approach in which sulfur, sterol-inhibitor fungicides, and AQlO offer a variety of tools for PMD control. Multiple applications apparently work by establishing a critical mass of spores in the treated area. These spores are not expected to survive long in the environment, and must interact with existing colonies of PMD quite rapidly, It has been observed in several different locations that, on the surface of leaves, AQ 10 can remain dormant in the absence of the pathogen for up to 14 d, but only when humidity is constantly high (i.e., >90%). In the normal conditions of the greenhouse or, even more so, the open field, the spores will have to germinate and interact with the host within several hours. This is the rationale behind the strong recommendation to apply the product at cooler or more humid periods of the day.

Commercial

Development

of EGofungicides

85

Early m the development of AQlO as a field control agent for PMD control in grape vmes, it was assumed that it could be used as a stand-alone treatment, with the idea of replacing chemical fungicides almost entirely Although AQ 10 gave consistently improved control of PMD when compared with untreated vines, disease control was too often unacceptable from a commercial viewpoint, especially toward the end of the growmg season.It was therefore decided that the chief strategy of use for AQ 10 would be as a replacement for conventional chemical control during certain morphological stages of vme and grape development (e.g., bloom to bunch closure, or from closure to verarson). For example, for a season of seven or eight 14- or 10-d interval sprays, the first two would be sulfur (m keeping with general PMD control practice), the next three AQlO, and the last 2-3 sprays would be an ergosterol biosynthesis inhrbitor (EBI), or similarly powerful chemical. However, field results were still extremely variable using this approach. Although, in general, AQ 10 performed less effectively m high disease pressure situations and with highly PMD-susceptible vine varieties, there were fatlures of control m situations that appeared optimal for AQlO, and other, excellent and reproducible results against high disease pressure m non-PMD-tolerant varieties. To resolve these mconsistenties and come up with user-friendly guidelmes, the results from several trials were closely analyzed. Two methods of disease assessmentwere used m the field trials. disease mctdence and disease severity. Disease Incidence reflects the percentage of leaves or grape bunches showing any PMD symptoms. Disease severity, on the other hand, scores the average diseased surface area of leaf or berry covered with PMD symptoms One of the prmcipal problems with the analysis of a season’s field data from multiple sites is the variability m assessors’estimates of disease severity. It is far easier to standardize disease incidence between cooperators or field-trial operators. However, Ecogen was fortunate, because, for two consecutive years, five of SIX trials m one of the programs were all assessedby the same highly sktlled and experienced cooperator, who on several occasions had shown a very high level of consistency of disease estimation, when results were reassessedby outside observers. It was therefore possible to assessthe relationship between disease mcidence and the underlying disease severity, and also the relationship between disease mcidence at one assessment date and dtsease severity at the next. The former is important as a measure of the accuracy of disease incidence for estimating severity; the latter as a measure of the reliability of predicting later disease severity based on current estimates of disease incidence. The data was collected from four different sites, whtch represent highly qualified grapevine growing areas in Italy. (It should be noted that a similar analysis has been made for other parts of the world, with virtually semi-

86

Hofstein and Chapple

lar results.) In each sate, AQlO was tested as part of an IPM scheme with two sulfur sprays at bud-break, followed by 2-3 AQlO sprays and followed by 23 sprays of an EBI. One other component m the blend of treatments was a selection of surfactants chosen to enhance the germmatton of AQ 1O’s spores. For SIX mdtvidual trials at the four sites, It IS clear from Fig. 1A that, wtthm sites, there IS a good correlatton between the underlymg disease severity and the disease incidence assessed. However, there was no such correlatton when considering the individual treatments (Fig. 1B). Therefore, the data was reanalyzed, taking the individual blocks from the sues, and treating these as expertmental units This approach makes sense when considering the variability wtthm sites* Although overall mean effects might differ from sate to site, the variability within any given site was always high, and blocks from one site gave levels of disease severrty that might be expected at another Analyzmg the data this way, Fig. 2A shows a clear linear relattonshtp between assessment of disease incidence and the underlying disease seventy. Similar lmeartty IS attained for the same data when broken up with respect to treatment regime, as opposed to site (Fig. 2B). However, It was the predictive power of disease incidence that was of particular significance for the future of the product development. As It appears m Fig. 3, from the relationship between the dtsease inctdence at the first assessment date and the disease severity at the second assessment date, there was a clear and reliable relationship between the two measurements. It was, therefore, decided that incidence could be used by a grower as a disease-threshold mdtcator for spraying The data has also been depicted to reflect the prolonged impact of AQl 0 on disease progresston. Figure 3 represents the correlation of severtty m second assessment as a function of percent mcrdence m the first assessment. The question reman-red. What threshold should be used? An alternative way of consrdermg the data presented IS as percent effective control (i.e., the reduction m PMD severity when compared with the untreated control plot included m the trial). Figure 4 represents the analysts of the data when the latter approach has been adopted. It reveals two important pomts. First, the regresston line of percent effective control at the second assessment date vs disease incidence, measured at first assessment date, IS far shallower for the standard chemical regtmen than for any IPM regimen including AQ 10 as a treatment. The difference m slope 1s a measure of the relative fragility of the btologrcal control agent, which should be interpreted as an mdlcatton that, at least m the case of btologrcal fungicides, there IS lrttle room for error m timing of appltcation. Once the PMD reaches a certain level of mfestatlon, the AQIO regime loses acceptable control. Second, rt 1sclear from these data (Fig. 3) that the threshold for effective control by AQ 10 IS under 30% disease mcldence. growers typically accept about 10% severity as a legitimate level of

0

10

20

40 50

% Disease Severity

30 60

70 80

. _.

I II III IV V VI

I

60 -

0

I

B

8

10

20

0

30

0

40

v 0 0 0

n

A

Regime:

50

60

70

IPM Regime I IPM Regime II IPM Regime III IPM Regime IV Standard chemical Untreated Check

Application

% Disease Severity

0.

80

Fig 1 Correlation of % diseasemcldence at the first assessmentdate with diseaseseventy at the sameassessment date Data pomts are mean values for trial sites(A) or treatment regimens (B)

,

Trral Locatron: - l Trial site ---------- Trial site -- -- -- -- m TrIalsite -- -- -- -- * Trialsite --- --- -- 0 Trial sde -0 Trial site

too

. .

20

40

60

60

* Trial site IV 0 Trial site V 100

1

0

0

0

I 20

a

.o

.

.

40

.

.

60

* . v o o o

60

100

IPM Regime I IPM Regime II IPM Regime Ill IPM Regime IV Standard chemical Untreated Check

Fig 2. Correlation of disease mcldence (x-axis, first assessment date) with disease severity (y-axis, loglO, same assessment date), usmg mdlvldual blocks from trial sites (A) or treatment regimens (B) as the data pomts

0

.

0 0 0 000

:

Commercial Development

of Biofungicides

89

Trial Location: Trial site I . Trial site II 0 Trial site Ill * Trial site IV . 0, I 0 Trial site V . .;;.,“p 0 0 Trial site VI -- -- -- -_ -- ‘m; &a’;-~,,‘o Qo . jJ *,-‘I Ao . + ’ xv ‘I I* 41 .* b, f 0 P,, o,.9x”o Q ’ l

- I ,f; , . I 0

.

,’ .* .

, 4’ . ‘a

II1 I1

i

: I I I ,

. . l--I--.-.-20

40

0

x

u

y=O.226+0.0148xX r2. 0.855; PcO.0001

60

,-..L-.L 80

_ 100

% Disease Incidence, 1st Assessment Fig. 3 Correlation of % disease mcldence at the first assessment date with the dlsease severity at the second assessment date (logI scale), using mdlvldual blocks from trial sites as data pomts (I e , data points are registered with respect to trial site) Correlation coefficient (r?) = 0 855

damage to the crop at the end of the wine-grape season. Since these observations were made, and as a critIca measure during commerclahzatlon of AQ 10 for a variety of crops with much lower damage tolerance, the product has been presented as a preventative treatment that is a very effective tool for PMD control, as long as it ISapplied when visible incidence is nearly zero (mstructions to growers are to use the product when diseaseincidence doesnot exceed 3%). % Disease seventy assessment 1(treatmentn) % Disease seventy assessment2 (treatmentn) % Disease seventy assessment 1 (untreatedcheck) - % Disease seventy assessment2 (untreatedcheck) (1)

Figure 4 depicts the percent disease suppression, but with no reference to the existing level of disease;Figure 5 shows the percent relative control, 1.e, the percent suppression of disease over the period of the two assessmentsas a function of the background disease levels at the time of the assessment(Eq. 1). For example, if a treatment had 12% disease severity at the first assessment, with a background of 20% m the untreated check in that block, and 16% severity at the second assessment,with a background disease severity of 45%, then

90

Hofs tein and Chapple 100

~+m.&O ---_8 -. t3 0 O--------Oa-----_ ‘. ---------~~~ ‘. ----.-____ .JI/l. v . --__ 0 *‘ j .A,Qy)bf am0 . . .i . . ” . ‘. ‘.. . . =ll . . .. . . -\ A * -. . t. Ii- . . . . A .. ’ ‘. I G . -\ ‘. - - - AQ regression line “9 _ Chemical regression line ’ “=y-; . * . A IPM Regime I , .. n IPM Regime II . T IPM Regime III . . _ * IPMRegime IV . o Standard chemical I

I

0

I

20 % Disease

40 incidence,

I

I

I

60

80

100

1st assessment

Fig 4 Correlation of disease mcldence at first assessment date with % effective control at the second assessment date (see text), usmg mdlvldual blocks from trial regimes as data points (1.e , data points are registered with respect to treatment) the PMD will have been suppressed by 24 5% (a posltlve value; Fig. 5) However, if the PMD 1sescaping control, percent relative control will be negative.

It appears from Fig. 5 that the higher the disease mcldence at the first assessment date, the more likely it IS that the PMD will not be controlled There are several caveats. The first disease assessmentswere all performed after at least one of the AQ 10 sprays had been completed, and the second assessmentsafter at least one or two more AQlO sprays Intervals between assessment dates ranged from 7 d when disease pressure was very high, to 21 d, when disease pressure was much lower. Hence, the mltlal disease pressure at the first assessment date was generally higher in the AQ 10 treatments than m the chemical ones, and subsequent control of disease would have been more difficult to accomplish m the AQ 1O-treated plots than m the chemically treated ones. In the cases presented, the disease was allowed to run Its course. Useful mformatlon was obtamed about control effects at high disease threshold, but these would not normally be obtained m a grower’s vineyard, because AQIO would have been replaced by a more robust method earlier m the season as an act of eradication. The question still remains: What exactly 1sthe threshold to recommend to growers?

Commercial

Development

of Blofungicides

r

-60 % Disease incidence, 1st assessment

Fig 5 Relatronshlp between % disease incidence at the first assessment date and % relative control at the second assessment date (see text), comparing the chemical standard regimen (filled triangles and all the AQIO treatments combmed (hollow circles) PMD infectlon 1s microscopic. The observable symptoms (small powdery colomes on leaves or berries) are evidence that the infection has been present for some time. Requtring a grower to use a microscope is clearly unreasonable, so the threshold for changmg from AQlO to conventional chemicals, based on observable symptoms reflected m scoring for percent incidence, has been set lower than the data presented actually mdlcates. In practice, and from other research, the manufacturer has been recommending a much lower threshold, approx 3% Incidence. This has two advantages: First, scouting need not be at too short an interval, requiring excessive time and labor, second, the grower has confidence m the control of the disease, even late in the season. The above threshold became a firm guideline. Figure 6 shows the result from trials in an area heavily infested with PMD. Yet, because of careful disease management, and with a much better understanding of the attributes and hmitatlons of AQ 10, it became the first blofunglclde to be officially included in IPM programs. It may, after all, appear a trivial discovery, but, because of Its mode of action, bringing hyperparasltlsm onto existing colonies of PMD, it took many field trials and much sophisticated data analysis to be able to reach such conclusions, and to draw practical guidelines from them.

Hofstein and Chapple

92 Data on bars is % disease incidence associated with the

Fig 6 Trial results for Bourgogne (France) site, showing % disease severity for chemical, AQ 10, and untreated regimens, for three assessment dates Numbers on bars are % disease severity

3.2. Timing and Quality of Spray Applications Spray application techniques have been known to play an important role m the performance of many pesticides (58). It appears that it is an even more important consideration in the commercial development of biopestictdes, and products of AQlO’s nature, in particular. One prerequisite IS that the apphcanon equipment used m trials must match that of the end user. Aspects of thts problem have been demonstrated m preliminary work (59), which shows that the distribution of particles of A quuqual~s m a spray tank differs markedly before and after passage through a pump (Fig. 7). Because efficiency of application of AQlO is dependent on distributing single spores evenly through a canopy, clumpmg of spores can decrease the efficiency of apphcatton stgmficantly. It was also shown that the dtstrtbutton of particles of a mineral surfactant (60), considered an enhancer of spore germinatton, also differs dramatically preand postpump. The particle distribution for the surfactant is depicted in Fig. 7. Surfactant systems for emulstfymg oils that are nonfungnoxtc are difficult to find, and, although the mineral oil used in most AQlO studies has some

01

0.15

0

Adjuvant oil 0.08

0.1

Log,, particle size pm

A. quisqualis

Ftg 7 The effect of pump sheer stress and atomtzatton on the frequency dtstrtbutton of A qusqualzs sporesand adJuvant or1(ADDQ) m the spray volume, pre- and postpump B, stirred m a beaker; T, stirred m 50 L of water m a spray tank; R, after 5 min recycling through a diaphragm pump; and S, after recycling and then spraying through a hydraultc flat fan nozzle (Reproduced wtth permtsston from ref. 59)

c

0

g. 0.05

.-g r

va

.-R N"

3 v

8

z-

94

Hofstern and Chapple 12 days after inoculation a

m

Treated at 1st symptoms and at 1st symptoms.

Fig 8. Compartson of two strategres for trmmg start of A quuqualls apphcatron applymg only at first stgns of disease, or applying 1 d after moculatton again at first signs of dtsease (Glasshouse trral In zucchnn, agamst Sphaerotheca fuhgznea ) Bars with the same letter do not doffer at the P < 0 05 level (Tukey’s HSD test) surfactant properties, these were found to be msufficrent to form a reasonable emulsion. When drops are created by the atomtzatron system, the probabrhty that a spore IS delivered with an or1 droplet must be high (near 1.O) for the enhancement properttes of the 011to take effect. It 1snoticeable that the small-scale trtals n-tthe glasshouse were done with an-pressurized systems that lack pumping or recycling. These hmttanon might have hampered the performance of AQ 10 Indeed, later on, all tnals were conducted with growers’ equipment, and, as tn the vane trials, efficacy has Improved on several crops, mcludmg roses and zucchnn, growmg m a glasshouse The trmmg of AQlO apphcattons has been addressed m more detarled test trials on cucumbers m glasshouses. One set of trials (Fig. 8) demonstrated that prophylactic treatment of cucumber resulted m effective suppression of PMD Essentially, AQlO apphcatton as early as arttficial mfection, followed by another appltcatton at first disease symptoms, gave very good results, whereas delaying application until disease symptoms were observed fatled to do so The results given m Fig. 8 mdtcate that spores of AQlO ought to be available at onset of PMD

Commercial

Development

of Biofungrcides

95

The tmphcation of these trtals for a standard recommendation to the glasshouse industry appears to be that A. quzsqualzs is effective for a short time after application, and that obtaining control very early m the disease progresston is crmcal. Hence, the grower must have a good idea of the ltkehhood of the dtsease occurrmg m the glasshouse, m order to trme the first apphcations. Again, coverage and even distribution of AQlO are crtttcal apphcatton parameters. Several risk-assessment models have been developed for the forecast of disease development. One of the most powerful examples has been described by Gubler et al. for PMD (62). This model has been translated into practical terms, and telemetry statlons have been mstalled rn several locations, to provtde warnmg of the onset of disease progression (i.e., ascospore release) These phenomena have been coupled to the prophylacttc concept of AQ 10, and tt was proved again, whenever the product was applied at relatively low levels of incidence, that tt provided commerctally acceptable disease control. Timing of appltcation and quality of applicatton are two major concerns in any program m proficient design of plant protectton, m general, and to btological pathogen control, m particular. However, in addition to inclusion withm IPM programs, assessmentof combmatton treatments of btofungictdes with environmental chemistry has also become a prtority, e.g., tank mtxes plus rates, advantages of addtttves H3, tdentificatton of alternate regtmes An exhausttve list of combinations has been screened, and the results offered to end users of AQlO (51). It became apparent that the latter mformatton contributed such value to the program that this has become a key component m every research and development program at Ecogen. 4. Current Status and Future Prospects Past reviews concernmg commercral development of btofungtcldes have highlighted the importance of cost-effective production of the active mgredient (62,63). Stgmficant progress has been made m recent years, and tt IS worth emphastzmgthe breakthroughs m massproductton of mtcrobial active mgredients through submerged fermentation, a method yteldmg large amounts of product m a cost-effective process. It has been demonstrated by the senior author, as well as other research groups, that a product-tailored formulation IS a key element m the whole program. However, what we have tried to accomplish m this chapter, which discussesthe experrence of developmg A. quuquah mto a commercial product, is to draw some basic conclusions that might be useful for others with a similar objective. It appears from this case study that the only way to learn about the attributes of a tentative btofungtctde is to throw tt mto an authentic commerctal sttuation as early in the program as possible. Such an approach, even though possessmgnumerous logtsttcal hurdles, apparently offers some significant shortcuts. This concept could be viewed as a useful gmdelme

96

Hofstein and Chapple

m any development program, but appears to be more profound m the development of blofunglcldes, since most of the small-scale slmulatlon assays commonly used m the development of chemicals fail to represent the commercial arena during the development of blologlcals. ASPIRE, a naturally occurring yeast for the postharvest control of decay, 1s a useful example of how the experience gathered during the development of AQ 10 helped accelerate the process of lmprovmg another blofunglclde The yeast suppresses decay formatlon via competition with germinating spores of prominent pathogens (e g., Penzczllzum spp) for space and nutrients. Although a standard tray assay, whereby artificially wounded and inoculated fruit ought to represent a situation in the packmghouse, has always been useful m the commerclal development of chemical funglcldes, the same has not been the case for blofungicldes In order for ASPIRE to express its full funglcldal potential, we had to rely on natural wounding and natural infestation (19). Again, when we consider the best approach to the development of blofunglcldes, taking the program to the end user 1sthe preferred approach. ASPIRE received a boost m the rate of development when it was subjected to testing m pilot- and commercial packing lines. Under these circumstances, it became apparent once agam that a blofunglclde cannot be promoted as a stand-alone program, but rather must be offered as part of a comprehensive disease control program, namely a tool m IPM Several accomplishments have been made m recent years m the area of commercial disease control Tnchodex@ (Makteshum, Israel) has been developed for gray mold (Botrytis) control, and Fusarium prollforatum was recently proposed for downy mildew control (on grapes) Other products have been commerclahzed for the suppression of soilborne pathogens It 1sstriking to review recent developments and realize that a better understanding of disease etiology, together with realization that no product can combat devastating pathogens as a stand-alone treatment, reveal an underlying opportunity for new discoveries m the area of plant pathology. In fact, at a time when many chemicals already suffer loss of performance because of pesticide resistance, blopestlcldes in general, and blofungicides m particular, are becoming recognized tools for reslstance management wtthm IPM systems. Molecular biology has produced many powerful tools for lmprovmg the understanding of pathogen-host recogmtlon and interaction. The resolution has already had a tremendous impact on expression of Bt genes m transgemc plants, offermg a whole new concept m defense against msect pests (66). Slmllarly Impressive is the progress made m an area of plant defense mechanisms and the molecular processesinvolved in defense induction (65). However, methodologles with higher resolutions are still required, to enable a better understandmg of the molecular attributes of microbial fungicldal agents known to date

Commercd

Development

of Biofungicides

97

These will, in turn, allow for a more focused effort m improving the performance of the products. AQ 10 and Trichodex are two examples of brofungrcides developed in recent years, and mtroduced already as commercial products. In reviewing the status of these examples, together with a whole host of biofungtcides for the control of soilborne pathogens, it is quite obvious that their potential to combat fungal pathogens has not been expressed. Therefore, and primarily as a safety compromise, they are promoted as a component within IPM systems,and are greatly dependent on amendment with chemical fungicides or enhancers. However, once the analytical resolution of dissecting the molecular elements that contribute to the fungicidal activity are revealed, it is presumed that the potential is better utilized. Indeed, recent developments m genetic engineering have already provided significant potential to improve some of the attributes. Future prospects are viewed as two parallel directions of product developments: intensify the expression of active gene products m better hosts, e.g., walldegrading enzymes, such as chitinases and glucanases of bacterial origin expressed m Trichoderma (66), and dissect fungicidal genes and their expression in transgemc plants. The latter approach has already proven extremely fruttful m producing herbicide-resistant or lepidopteran-pest-resistant crops, as mentioned above. It will be possible to achieve similar successeswith respect to fungal disease control, regardless of the direction selected for development, only when there is a better understanding of the molecular basis of disease suppression. As a result of increasing knowledge of the mechamsms of pathogenicity, more progress IS expected to be made m developmg new types of diseaseprotectants, as well as the genetic engineering of resistant plant varieties. The key to progress is the successm extensively interfering with the virulence of the pathogen. Agam, all that IS known to date about the mode of action of AQlO is that it exerts its suppressive effects via hyperparasitism of PMD. We have not been able to determine the molecular cues of host-pathogen recogmnon, let alone the process by which A quzsqualis governs the metabolic function of the pathogen Progress m elucidation of the mode of action will have nnmediate implications on the quality of the biofungicide product. It ~111still remam an overridmg concern of those who translate scientific achievements mto commercial products to ensure that all inherent attributes are expresseddespite hurdles of mass production, and that it is all executed m a cost-effective fashion. References 1 Chet,I. (1987) ZnnovatzveApproaches To Plant Disease Control. Wiley,Toronto,Canada. 2 Papavizas,G. C (1984) Soilborneplant pathogens new opportunities for biological control. Proceedings 1984 Brltlsh Crop Protection Conference-Pests and Diseases,

pp 371-378.

98

Hofstein and Chapple

3. Sundheim, L. and Tronsmo, A. (198X) Hyperparasites in biological control, in Biocontrol qf Plant Diseases (Mukerji, K. G. and Garg, K. L., eds.), CRC, Boca Raton, FL, pp. 53-69. 4. Cook, R. J. (1993) Making greater use of introduced microorganisms for biological control of plant pathogens. Annu. Rev. Phytopathol. 31, 53-80. 5. Cook, R. J. (1993) The role of biological control in pest management in the 21st century, in PestManagement: Biologically BasedTechnologies,vol. 18 (Lumsden, R. D. and Vaughn, J. L., eds.), Beltsville Symposium. American Chemical Society, Washington, DC, pp. l&20. 6. Fokkema, N. J., Gerlagh, M., Kohl, J., Jongebloed, P. H. J., and Kessel, G. J. T. (1994) Prospects for biological control of foliar pathogens, in Proceedings qfthe Brighton Crop Protection Conference: Pests and Diseases,BCPC Publications, Major Print Ltd., Nottingham, UK, pp. 1249-1258. 7. Janisiewics, W. J. (1988) Biological control of diseases of fruit, in Biocontrol qf Plant Diseases,vol. 2 (Mukergi, K. G. and Grag, K. L., eds.), CRC, Boca Raton, FL, pp. 228-235. 8. Wilson, C. L, Wisniewski, M. E., El-Gaouth, A., Droby, S., and Chalutz, E. (1996) Commercialization of Antagonistic yeasts for the biological control of postharvest diseases of fruit and vegetables. J. Ind. Microbial. 46,237-24 1. 9. Cook, R. J., Bruckart, W. L., Coulson, J. R., Goettel, M. S., Humber, R. A., Lumsden, R. D., et al. (1996) Safety of microorganisms intended for pest and plant disease control: a framework for scientific evaluation. Biol. Control 7,333-35 1. 10. Carlton. B. C. (1990) Economic consideration marketing and application of biocontrol agents, in New Directions in Biological Controls-Alternatives ,for SuppressingAgricultural Pestsand Diseases(Baker, R. R. and Dunn, P. E., eds.), Liss, New York, pp. 4 19-434. 11. Bravo, A. (1997) Minireview: phylogenetic relationships of Bacillus thuringiensis b-endotoxin family proteins and their functional domains. J. Bacterial. 179, 2793-280 1. 12. Lumsden, R. D. and Walter, J. F. (1995) Development of the biocontrol fungus gliocladium virens: risk assessment and approval for horticultural use, in Biological Control: BeneJitsand Risks(Hokkanen, M. T. and Lynch, J. M., eds.), Cambridge University Press, Cambridge, UK, pp. 263-269. 13. Lewis, J. A. and Papavizas, G. C. (1987) Application of Trichoderma and Gliocladium in alginate pellets for control of Rhizoctonia damping-off. Plant Pathol. 36,438-446.

14. Baker, R. (1982) Induction of suppressiveness,in SuppressiveSoilsand Plant Disease (Schneider, R. W., ed.), American Phytopathology Society, St. Paul, MN, pp. 35-50. 15. Baker, R. (1983) State of the art: plant diseases, in Proceedingsqf the National Znterdisciplinary Biological Control Conference(Battenfield, S. L., ed.), Cooperative State Reservation Service, U. S. Department of Agriculture, Washington, DC, pp. 14-22. 16. Baker, R. and Scher, F. M. (1987) Enhancing the activity of biological control agents, in Innovative Approaches to Plant DiseaseControl (Chet, I., ed.), Wiley, Toronto, Canada, pp. l-l 7.

Commercial

Development

of Biofungicides

99

17. Dubos, B. (1987) Fungal antagonism in aerial agrobiocenoses, in/nnovativeApproaches to Plant Disease Control (Chet, I., ed.), Wiley, Toronto, Canada, pp. 107-I 35. 18. Hemming, B. C. and Houghton, J. M. (1993) Influence of biotechnology on biocontrol of take-all disease of wheat, in Biotechnology in Plant Disease Control (Chet, I., ed.), Wiley-L&, New York, pp. 15-38. 19. Wilson, C. L. and Wisniewski, M. E. (1994) Large scale production and pilot testing of biological control agents for postharvest diseases, in Biological Control of Postharvest Disease, Theory and Practice (Chet, I., ed.), CRC, Boca Raton, FL, pp. 89-100. 20. Cate, R. (1990) Biological control of pests and diseases: integrating a diverse heritage, in New Directions in Biological Control: Alternatives for Suppressing Agricultural Pests and Diseases (Baker, R. R. and Dunn, P. E., eds.), Liss, New York, pp. 333-344. 21. Lumsden, R. D. and Lewis, J. M. (1989) Selection, production, formulation, and commercial use of plant disease biocontrol fungi, problems and progress, in Biotechnology of Fungi for Improving Plant Growth (Whips, J. M. and Lumsden, R. D., eds.), Cambridge University Press, Cambridge, UK, pp. 17 l-l 90. 22. El-Ghaouth, A. E., Wilson, C. L., and Wisniewski, M. E. (1995) Sugar analogs as potential fungicides for postharvest pathogens of apple and peach. Plant Dis. 79, 254-258. 23. Fokkema, N. J. (1976) Antagonism between fungal saprophytes and pathogens on aerial plant surfaces, in Microbiology ofAerial Plant Surfaces (Dickinson, E. and Preece, F., eds.), Academic, New York, pp. 487-506. 24. Katan, J. (1985) Solar disinfestation of soils, in Biology and Management of Soilborne Plant Pathogens (Parker, C. A., Rovira, A. D., Moore, K. J., Wong, P. T. W., and Kollmorgen, J. F., eds.), APS, St. Paul, MN, pp. 274-278. 25. Blakeman, J. P. and Fokkema, N. J. (1982) Potential for biological control of plant diseases on the phylloplane. Annu. Rev. Phytopathot. 20, 167-l 92. 26. Olivier, J. M. (1983) Les organismes antagonistes d’agents phytopathogenes, in Faune et Flore Auxiliaires en Agriculture. ACTA, Paris, pp. 145-164. 27. Chet, I. and Baker, R. (1981) Isolation and biocontrol potential of Trichoderma hamatum from soil naturally suppressive of Rhizoctonia solani. Phytopathology 71,286-290. 28. Falk, S. P., Gadoury, D. M., Cortesi, P., Pearson, R. C., and Seem, R. C. (1995) Parasitism of uncinula necator Cleistothecia by the mycoparasite Ampelomyces

quisqualis. Phytopathology

85, 794-800.

29. Galper, S., Sztejnberg, A., and Lisker, N. (1985) Scanning electron microscopy of the ontogeny ofAmpelomyces quisqualis pycnidia. Can. J. Microbial. 31,961-964. 30. Taber, R. A., Smith, D. H., Petit, R. F., and Johnson, J. D. (198 1) Potential for biological control of Fulvia,fulvum by Hansfordia in Texas. Phytopathology 71, 908-912. 3 1. Falk, S. P., Gadoury, D. M., Pearson, R. C., and Seem, R. C. (1995) Partial control of grape powdery mildew by the mycroparasite Ampelomyces quisqualis. Plant

Dis. 79,483-490.

100

Hofstein and Chapple

32. O’Neill, T. M., NIV, A , Elad, Y., and Shttenberg, D (1996) Btological control of Botrytzs cznerea on tomato stem wounds with Trtchoderma harztanum in Israel Eur J Plant Path01 102,6355643 33 Jackson, M and Schtsler, D. H. (1992) The composition and attrtbutes of Colletotrzchum truncatum spores are altered by the mutational envtronment Appl Envtron Mtcrobtol. 58,2260-2265. 34 Phthpp, W D and Cruger, G. (1979) Parasitismus von Ampelomyces qutsqualzs auf Echlen Mchltauptlzen an Gurken und andern Gemtisearten 2 PfZanzenkr Pflanzenschutz 86, 129-142 35. Engelkes, C. A , Nuclo, R L , and Fravel, D. R. (1997) In effect of carbon, mtrogen, and C N ratto on growth, sporulatton, and btocontrol efficacy of Talaromyces jlavus Phytopathology 87,500-508 36 Hofstem, R and Fridlender, B (1994) Development of production, dehvery, formulation and dehvery systems for biohnrgtcides, m Brtghton Crop Protectron Conference Pests and Dueases, BCPC Publications, Major Print Ltd , Nottmgham, UK, pp 1273-l 280. 37 Abu-Foul, S , Raskm, V. I , Sztejnberg, A , and Marder, J B (1996) Disruption of chlorophyll organization and function m powdery mildew dtseased cucumber leaves and its control by the hyperparasite Ampelomyces quzsqualts Phytopathology 86, 195-l 99 38 SzteJnberg, A , Galper, S., Mazar, S , and Ltsker, N (1989) Ampelomyces quzsqualzs for biologtcal and integrated control of powdery mildew m Israel J Phytopathol. 124,285-295 39 Zimand, G , Elad, Y , and Chet, I. (1996) Effect of Trrchoderma harzranum on Botrytts ctnerea pathogenictty Phytopathology 86, 1255-1260 40 Sutton, J C , LI, D., Pang, G , Yu, H., Zhang, P , and Valdebemto-Sanhueza, R M. (1997) Gltocladtum roseum. a versattle adversary of Botrytts cwerea m crops Plant Dts 81,3 16-328 41 Wtlson, C L and Wtsmewskt, M E., eds (1994) Bzologzcal Control OfPostharvest Dtsease of Frutts and Vegetables-Theory and Practtces CRC, Boca Raton, FL 42 Jamsiewicz, W J , Usall, J , and Bors, B (1992) Nutrittonal enhancement of btocontrol of blue mold on apples Phytopathology 82, 1364-l 370. 43 Elad, Y , Zimand, G , Zaqs, Y , Zuriel, S , and Chet, I (1993) Use of Trtchoderma harztanum in combmatton or condmons alternation wtth fungictdes to control cucumber gray mold (Botrytzs cznerea) under commercial greenhouse condittons Plant Pathol. 42,324332 44 Elad, Y., Shtienberg, D., and Niv, A. (1994) Trichderma hrzzanum T39 integrated with fungicides: improved btocontrol of gray mold, m Brighton Crop Protectton Conference Pests and Dweases, BCPC Publications, MaJor Print Ltd , Nottingham, UK,pp 1109-1114 45 Daoust, R A and Hofstein, R (1996) Ampelomyces quzsqualts, a new biofungicide to control powdery mildew m grapes, m Brtghton Crop Protectton Conference Pests and Diseases, BCPC Publications, MaJor Print Ltd , Nottingham, UK, pp 33-40.

Commercial

Development

of Biofungicides

101

46 Lmdow, S. E., McGourty, G., and Elkms, R (1996) InteractIons of antlblotlcs with Pseudomonasfluorescens strain a506 m the control of fire bhght and frost mJury to pear. Phytopathology 86, 841-848. 47 Federal register EPA (1989) Data Requirements for Pesticide Registration, Final Rule. 53, 15,952-l 5,999. 48 Federal register, EPA ( 1995) Candzda oleophlla Isolate I- 182 (ASPIRE) Exemptlon from the Requn-ement of a Tolerance. 60, 11,032-l 1,033. 49. Katz, M (1997) Powdery mildew. biological enhances growers’ options. Grape Grower

29,4-7

50 Cavanaugh, P (1997) New blofunglclde reduces resistance, late sulfur. Am. Vzneyard 6,3-5

5 1 Hudson, R A (1997) Compatibility of Ampelomyces quxqulzs spores with commercial chemical products for use in IPM programs Phytopathology 87, S45. 52 Sundhelm, L and Kreklmg, T (1982) Host-parasitic relationships of the hyperparasite Ampelomyces quzsqualts and Its powdery mildew host Sphaerothzca fullgznea Phytopathology 104,202-2 10 53 Jarvls, W. R. and Klmsby, K. (1977) The control of powdery mildew of greenhouse cucumber by water sprays and Ampelomyces qulsqualzs Plant Du Rep 61,728-730

54 SzteJnberg, A , Galper, G., and Lisker, N (1990) Condltlons for pycmdlal production and spore formation by AmpelomycesquzsqualzsCan J Mcroblol 36, 193-198 55 Falk, S. P , Pearson, R C , Gadoury, D. M , Seem, R C , and SzteJnberg, A (1996) Fusarlumprollferatum as a blocontrol agent against grape downy mildew Phytopathology 86, 101&1017 56 Droby, S , Chalutz, E , Wlsmewskl, M E , and Wilson, C E (1996) Host response to mtroductlon of antagonistic yeasts used for control of postharvest decay, m Mzcroblology ofAerza1 Plant Surfaces (Moms, C E , Nlcot, P , and Nguyen-The, eds ), Phytopathology Press, St Paul, MN 57 Wilson, C L , El-Ghaouth, A E., Chalutz, E., Droby, S., Stevens, C , Lu, J Y , Khan, V , and Arul, J (1994) Potential of induced resistance to control postharvest diseases of fruits and vegetables Plant Du 78, 837-844 58. Chapple, A C , Downer, R. A, and Hall, F R (1993) The effect of spray adJuvants on swath patterns and droplet spectra for a flat-fan hydraulic nozzle. Crop Protection 12,579-590

59. Chapple, A. C. and Bateman, R. P (1997) Application systems for mlcroblal pesticides. necessity not novelty, m Brztzsh Crop Protection Councd Symposium Mzcrobzal ZnsectzczdesNovelty or Necesszty?(Evans, H F , ed ), BCPC Publlcatlons, Major Print Ltd , Nottingham, UK, pp 18 l-l 90. 60 Phlllpp, W D , Beuther, E , Hermann, D , Klmkert, E , Oberwalder, C , and Schmldke, M (1990) Formulation of the powdery mildew hyperparaslte Ampelomyces qulsqualzs J Plant Dls Protectzon 97, 120-132 61 Gubler, W. D., Thomas, C. S., Weber, E., Luvisl, D., Leavitt, G , and Smith, R. (1997) Use of a weather station based disease risk assessment for control of grapevine powdery mildew m California Phytopathology 87, S36

102

Hofstein

and Chapple

62 Whtpps, J. M (1992) Status of biologtcal disease control Hortudtural Blocontr-ol Scz Technol. 2,3-24. 63 Lumsden, R D , Lewis, J A , and Fravel, D R (1995) Formulation and deltvery of btocontrol agents for use against sotlborne plant pathogens, m Bloratlonal Pest Control Agents Formulattolz and Delwery (Hall, F R. and Barry, S , eds ), Symposium Series, American Chemtcal Society, Washmgton, DC, pp 16&l 82 64 Perlak, F J , Deaton, R W , Aremstrong, T A, Fuchs, R L , Sums, S R , Greenplate, J T., and Ftschhoff, D A (1990) Insect reststant cotton plants Brotechnology 8,939-943 65 Broghe, K., Broghe, R , Benhamou, N , and Chet, I (1994) The role of cell wall degrading enzymes m fimgal dtsease reststance, m Biotechnology In Plant Dzsease Control (Chet, I , ed ), Wiley-Ltss, New York, pp 139-156 66 Chet, I , Barak, Z , and Oppenhetm, A (1994) Genetic engmeermg of mtcroorgamsms for improved biocontrol acttvrty, in Bzotechnology zn Plant Disease Control (Chet, I , ed ), Wtley-Lrss, New York, pp 21 l-235

Biological

Control of Seedling Diseases

K. Prakesh Hebbar and Robert D. Lumsden 1. Introduction Seedlmgs of economically important crop plants are attacked by various soilborne pathogenic fungi, such as Pythium, Fusanum, Rhlzoctorua, Phytopthora, and others, which cause either seed rot before germmation or seedling rot after germmation, resultmg m billions of dollars m cumulative crop losses. These diseases are often termed pre- and postemergence damping-off, or seedlmg blights Greenhouse crops grown in soilless cultures, as well as field crops, are susceptible to soilborne fungal pathogens, resulting m considerable economic losses.Currently, the most widely used control measure for suppressing soilborne diseases IS the use of environmentally hazardous fungicidal treatment of seed, seedlings, or soils. However, problems encountered, such as development of pathogen resistance to fungicides, mabillty of seed-treated fungicides to protect the roots of mature plants, rapid degradation of the chemicals, and a requirement for repeated applications, have given impetus to alternative remedies (2). One approach to address this problem is to use naturally occurrmg and environmentally safe biological control microorganisms, used alone or m conluncnon with integrated pest management (IPM) strategies (2). Several biological control agents have been commercialized or have been registered for commercial field trials (3) However, a major problem encountered m the area of biological control is the mconsistencies m performance of biocontrol products (4). According to Bowen and Rovtra (5), rapid migration to newly formed root surfaces from the point of inoculation, as well as rapid growth rates, can also be helpful in improvmg the performance of biocontrol agents Suslow and Schroth (6) have reported, m a dose response study, that the mmtmum number of viable cells needed for uniform colonization and plant growth promotion m From Methods m Biotechnology, vol 5 Blopestmdes Use and Debvery Edtted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

103

Hebbar and Lumsden

104

sugarbeets with Pseudomonasfluorescens was IO5 bacteria/seed or 10’ bacteria/g dry wt moculum. Failure m moculum trials may be caused by selection of moculants that do not have some of the above charactertstics (7). This may also be caused by factors such as lack of knowledge related to the ecology of btocontrol agents, poor sot1colomzatton and persistence, and limited dehvery technology currently available. The obJeCtiVe of this chapter is to describe the successful discovery and commercial development of two btological control agents: a bacterium, Burkholderra cepacla; and a fungus, Ghocladium wrens. 2. Examples of Biological Control Agents Examples of bacteria that are currently commercially available are Gramnegative B cepacza (Deny TM, CCT, Carlsbad, CA), and Agrobacterlum radzobacter (Nogall TM, Bro Care Technology, Australia) and Gram-positive Baczllus subtzlzs (Kodiak HBTM and EptcTM, Gustafson) and Streptomyces gnseovwcd~s (Mycostop TM,Kemira OY, Finland). The fungal btocontrol products available commercially are Glzocladzum wrens (SoilGardTM, ThermoTrilogy, Columbia, MD) and Trichoderma harzranum (T22TM and Root ShteldTM, Bioworks, Geneva, NY). Most of the above mentioned products are effective m controlled envtronments, such as greenhouses, but only a few of these products have been used extensively for biologtcal suppression of fungal diseases m major fruit (Nogall) and field crops (Kodiak) (3). This chapter wtll discuss the discovery and development of two biocontrol agents: a bacterium, B cepacla, and a fungus, G wrens. 2.7. Burkholderia 2 1.1. Ecology

cepacia

Studies of corn monoculture soils in the midwestern United States(8) and m France (9) showed that high populations of B. cepacza (syn Pseudomonas cepacza) (IO), a ubiquitous sot1bactermm, were associatedwith the rhizosphere and roots of corn, In addition, successive corn culttvation increased populations of B. cepacla (9). The rhizosphere/nonrhizosphere (R/S) population ratio was 4000, indicatmg its affimty for corn roots. When applied as seed moculants, B cepacza strains isolated from corn colomzed the rhizosphere and roots of corn extensively, with seed moculum levels as low as 10 bacteria/seed Within 2 wk, populattons proliferated to 10’ colony formmg units (CFU)/g dry wt of root (II). B. cepacza 1s also an efficient colonizer of roots and rhizosphere of radish (12), pea (231, sunflower (14), and soybean (IS). Although B cepacia has been described as a phytopathogen causing sour skm of onion bulbs (16), and also as a secondary pathogen m humans with cystic fibrosis (27), strains isolated from the rhtzosphere of corn do not cause necrosis of onion tissue (18), and are quite different from the chmcal strains (19).

Biological Control of Seedling Diseases

105

2.1.2. Mode of Action B. cepacia strains isolated from the rhizosphere of corn have a broad-spectrum antifungal activity against a range of fungal pathogens (18). Pyrrolnitrin has been reported as the major metabolite responsible for the broad-spectrum antifungal activity of B cepacia (5,12,20). On the contrary, the reference strain ATCC 254 16, an onion pathogen, does not produce pyrrolmtrm (21) Recent reports indicate that, in addition to pyrrolnitrin, B. cepacia produces antifungal compounds such as siderophores (22-24) and chlorinated phenylpyrrole antibiotics (29, as well as the hydrolytic enzyme P-1,3-glucanase (26).

2.1.3. Biological Control Numerous reports now exist on the isolation and utilization of B cepacza for biological control of vartous sorlbome fungal pathogens in different crops (22,24,18, 21,23,27-31) In growth chamber studies, Hebbar et al. (28) determmed that the majority of sot1strains of B. cepacia were unable to suppressearly corn seedling infection by Fusarium moniliforme, but those isolated from corn roots could do so. Isolates of B cepacaa from lettuce roots, when used as seed treatments,reduced radish seeddamping-off causedby Rhizoctonia solam AG4 by 50% (12). Although B. cepacia ISeffective againsta wide range of soil pathogens,its success asa seedtreatment is determmed by various factors, such astotal number of bacteria addedto coat the seeds,soil temperature,and, in somecases,the plant cultivar used. In greenhousestudies,the optimum concentration of B. cepacia necessaryto reduce seedlingdamping-off in corn causedby acombmation of three pathogens(Fusanum graminearum, Pythium ultimum, andPythium arrhenomanes), was determmed to be 1OSCFU/seed(30). In the samestudy,the optimum temperature(25°C) for B. cepacia to be effective was higher than that (18’C) for the fungal biocontrol agent G. virens. Kmg and Parke(28) determined that, although the efficacy of B. cepacza to suppress Aphanomyces eutezches root rot andPythium damping-off m peaswas not limited to a single cultivar, the differential effects of biocontrol were related to the degree of suscepttbthtyto the pathogen of eachof the four cultivars tested. Although B. cepacia is consideredas a biocontrol agent with potential for largescaleapplication, there is only one commercial product available, marketed by CCT under the trade nameDeny. Recently,promising resultswere obtained (32) m extensive field trials conductedby Agrium (Saskatoon,Canada)to evaluate, the feasibility of using B. cepacia strain Ral-3 for biological control and growth enhancementof conifer seedlings. 2.2. Gliocladium 2.2.1. Eco/ogy

virens

The fungus G. wrens Miller, Giddens, and Foster (=Trichoderma wrens, Miller, Giddens, Foster, and von Ark) was originally isolated from a sclero-

Hebbar and Lumsden

106

tmm of the plant pathogemc fungus Sclerotinza manor burled and recovered from a Beltsville, MD, soil G vzrens 1snative to all parts of the United States and 1swidely distributed throughout the world (33,341. G vzrens is a hyphomycete with no confirmed sexual stage The possible sexual stage is Hypocrea gelatznosa (33,35). It proliferates as asexual conidta that are held m masses of moist spores. It survives as vegetative segments of the mycehum, termed chlamydospores, usually embedded m organic matter. The spores are not airborne and are dispersed only as spore suspensions m water, or carried m sot1 or m organic debris Recently, molecular evidence Indicated that G vzrens is more closely related to Trzchoderma than to the type spectes of the genus Glzocladzum. This supports suggestions to refer to the fungus as T wrens, rather than G wrens (35). Because of the prevalence m the literature for the established name, G wrens, the authors prefer to use this name. However, the probable relationship to Trzchoderma 1srecognized. 2.2.2.

Mode of Action

G wrens is a common sot1 saprophyte, and, as with many other sotlborne fungi, produces several antibiotic metabolites (36-41) that are thought to enhance tts soil competittveness The metabohte most likely associated with control of Pythium and Rhtzoctonia damping-off is gltotoxm, an eptpolythiopiperazine-3,6-dione antibiotic (36,40,42, 42a). Ghotoxm has antibacterial, antifungal, antiviral, and antttumor activity It also interferes with phagocytic cells, and is unmunosuppressive (41). Smce gliotoxin has moderate mammahan (rats) toxtctty (50 mg/kg) (41), mgestion directly by an animal or human 1sof some concern. However, thorough evaluatton of formulattons indicated only traces of gltotoxin m the product (43), and the formulation is not toxic to rats (44). Consequently, the wheat-bran-based product would not be harmful tf ingested. Moreover, gltotoxm 1sproduced after mcorporation of the granular product mto the soil, remains active for a short period of ttme, and is inactivated (Lumsden, unpublished results). A defimttve role for ghotoxm mvolvement m the mechanism of antagomsm is supported by recent mutattonal analysts studies of G wrens (G-20 = GL-21) (45). In that case, at least for action of Gl-21 against P ultimum, about 60% of the biocontrol efficacy of the wild-type strain was lost when strains were mutated to no longer produce glrotoxm. The remammg bto-control effecttveness might realisttcally be attributed to competition of G wrens with P ultimum for nutrients. Conversely, other mutational analysis evtdence, m the case of R. solani, suggests that competition for nutrients might account for a greater effect on achieving biocontrol than production of antibiotic metabohtes (46)

Biological Control of Seedling Diseases

107

2.2.3. BiologIcal Control Several reports have been published on btologlcal control of a wade range of sotlborne fungal pathogens by G vzrens under both field and greenhouse condmons. Recently, the development of G vzrens for damping-off disease control m greenhouse applications was described (47). The commercial development of G vzrens as a granular formulatron, wrth the trade name SotlGard, was accompltshed by a process that included drscovery of the btocontrol fungus, product development, marketing assessment, product formulattons, process development, extensive efficacy assays, regtstratton wtth the U. S. Environmental Protection Agency (EPA), scale up, and test marketing. The biologtcal control properties of the fungus, G vixens, were aimed at controllmg damping-off diseases of seedlings caused by P ultlmum and R soianl m greenhouse production of seedlings and bedding plants (47). The biological control efficacy of G vzren~ was tested extensively and determmed to be consistent and reliable when used as prescribed for control of damping-off m greenhouse apphcattons (2,43,47-49) Also, G. vzrens reduced disease caused by R solanl on potato (SO) and ornamental crops (48,49) In field trials, appltcatton of G vzrens reduced the mctdence of southern blight caused by Sclerotzum rolfszi in carrots and tomatoes, and increased yield (51). Preliminary results have shown that seed treatment with G virens reduced damping-off in corn caused by a combinatton of P. ultimum, P. arrhenomones, and F gramznearum both in greenhouse (30) and field trials (W. Mao, personal communication). 3. Production and Application of Biocontrol Agents Certain crtterta were considered important in the early stages of development of G. vzrens and B cepaczaas brocontrol agents (4,s). For G vzrens, thts mcluded a bioassay method for selectmg the best strain of G wrens, which also considered: the use of a commercially available soilless medium, used m commercral glasshouses where the disease problems (damping-off) occur; the study of appropriate pathogens, such as P u&mum and R. solam, which are important in greenhouse operations in which the use of a btologrcal control agent would probably be most successful because of fairly uniform culture condrttons; selection of indigenous microorganisms to the United States, because nomndrgenous mlcroorgamsms might be conceived as more likely problems for the United Statesenvironment; the potential for the use of a single isolate of a btocontrol agent for control of both pathogens would be preferred over a mixture of isolates; and the utilizatron of high-value crops, important m the ornamental productton industry to defray the cost of development and registration Similarly, the above logic was also used m the development of B. cepaczaas a btocontrol agent. However, the crop was a field crop, corn, and

Hebbar and 1umsden

108

the targeted pathogens were Pythzum and Fusarium spp (8,11,28). On the basis of these factors, it would make the process for registration and commerciahzation of an agent for control of plant diseaseseasier. 3.7. Liquid Fermentation Major advances have been made m the production of biocontrol agents using hqutd fermentatton (52). The strategy used for large-scale production of Grampositive bacterial agents (Baczllus spp) has been to obtain heat- and desiccation-resistant endospores of bacteria or chlamydospores of fungi. Large-scale production of the resistant chlamydospores of fungal antagonists, such as Trzchoderma and Glzocladzum spp are now possible using llqutd fermentation (53). Although large-scaleproduction of Gram-negative bacterial agents(B. cepacla and P f2uorescen.s) is feasible, a major problem is encountered because of then sensitivity to desiccation. This constraint greatly affects the next step, which IS the delivery and application technology 3.2. Formulation

Development

Upon selection of an antagonist of choice, an appropriate formulation of the biocontrol agent for ease of preparation, application, and maximum efficacy should be chosen. Formulation is a key to product success, because it can determine success of delivery, shelf life, and stability of its effectiveness against plant pathogens. The first formulation for G wrens was based on alginate-wheat-bran granules (prill) (7,54), and was called GlioGardTM. Later in product development, certain quality control problems were encountered m the scale-up process. Difficulty was encountered because of the holding times for biomass, and for drying the alginate prill preparations m large volumes. For this reason, and because of an increase m the cost of alginate, the formulation was modified by including dextrm as a binder, reducing the algmate content, and preparing the biomass mixture by a fluid-bed granulation method. These changes did not affect the shelf life and efficacy characteristics, and thus were adopted for a modified product, SoilGard (47). Quality control 1sessential for formulatton development. A simple, but welldefined quality-control program should be m place for comparison of different formulations, and should examine, among other factors, viabihty, stability, and efficacy (43). If a quality control program is not m place during formulation research, product development can be severely delayed, which will ulttmately impact profitability. Quahty control is also extremely important during commercial productton of biocontrol agents (43). With favorable results obtained m detailed efficacy trials (49), additional trials were expanded to include several other cooperators, to ensure that results could be replicated (43). When similar results were obtained, trials were con-

Biological Control of Seedling Diseases

109

ducted with several bedding-plant grower cooperators using different protocols (43). With the exception of artrficral mfestatron with pathogens, whrch was used in the earlier trials, researchers were at liberty to use whatever plants they desired, and their own apphcatton schedules. The design of the trials in step arrangement allowed for less control m the way the product was handled with each phase of testing. By the time commercral growers were included m the trials, natural mfestatron was used to determine efficacy, and the grower was srmply given the product with a basic set of mstructrons, srmilar to those now provided on the product label. So far, other than a liquid-based or peatbased formulatron, further progress has not been achieved m formulatmg B cepacia strains (G. Growell, CCT, personal commumcation). 3.3. Application

of Biocontrol

Agents

The most commonly used methods for delivermg btologrcal control agents, especrally under controlled greenhouse conditions, are soil amendments usmg granulated formulations (SorlGard) or drenching with liquid formulations (Deny). Seed applications of dry formulatron have been successfully used for field crops such as cotton and peanuts (Bacillus spp, Kodiak). Seed apphcatron of G vzren~IS currently bemg investigated in the Btocontrol of Plant Diseases Laboratory (US Department of Agriculture, Agrrcultural Research Service). Preliminary results have shown that seed treatments with dry fungal biomass rich m chlamydospores are equal to, or better than, fungrctdes m reducing damping-off m corn (30). However, delivery of Gram-negative bacterial agents as seed treatments for field crops is still a major problem. Brocontrol agents, such as P.fluorescens or B. cepacia, coated on the seeds,have a short shelf hfe (at room temperature), and are readily killed by desiccation. Unless methods are found for their delivery as seed apphcattons, then large scale use may not be feasible However, liquid formulattons (Deny) are bemg tried usmg drip nrtgatton for a few high-value crops, such as strawberry and melons. 3.4. Registration

of Biocontrol

Agents, G. virens as Example

Regrstratron of G virens with the EPA was mtttated by W. R. Grace (now Therm0 Trilogy, Columbia, MD) (44). The EPA reviews applications and regulates mrcroorganisms used as biocontrol agents if they are genettcally engineered, nonmdtgenous to the United States, or if they will be field-tested on more than 10 acres (4.05 ha) of land or 1 acre surface of water (0.41 ha) (55). Subdivision M of the EPA Pesticide Testing Gurdelmes (56) treats mrcrobial agents for the control of plant pests in ways similar to those for chemical pesticides. Companies applying for registration must provide extensive mformation for approval for commercial use of microbial pesticide products. Product testing 1sset up in a tier systemthat recogmzesthe mherent rusksand degrees

110

Hebbar

and Lumsden

of exposure associatedwith different usesof pesticides In addition to production and taxonomrc data, long- and short-term effects on a variety of orgamsms,mcludmg plants, ammals, and other nontarget organisms, may be necessary. Three formulattons of G wrens (Gl-2 1) were approved by the EPA WRCGL-2 1, a manufacturmg-use product (fungal biomass), can be used m formulations of btocontrol products. WRC-AP- 1 (GhoGard) is an end-use formulated prtlled product containing calcmm algmate, wheat bran, and proprietary additives to prolong shelf life. The prtlls or granular material is mixed with sot1or soilless plant growmg media at least 1 d prior to planting, or incorporated mto the medium surface m plant beds prior to, or at, planting. The formulation 1s used at the rate of l-l 5 lb/yd3 (approx 1 g/L) of media when mixed, or at the rate of 0.75-l ounce/sq ft when applied to the bed surface. WRC-AP-2 (SoilGard) is an improved granulated product, with somewhat improved efficacy, and which is more economical to manufacture. GhoGard is no longer being produced, and is replaced on the market with SoilGard (47). 3.5. Compatibility of Biological Agents with Chemical Pesticides and Other /PM Strategies Recently, one of the major endeavors to improve the efficacy of btologtcal control agents, has been to use them m an IPM strategy, such as sot1solarizatton (57,58), or by using btocontrol agents resistant or tolerant to chemical pesticides (59,60) Initial results from field trials to evaluate suppression of the southern blight pathogen, S ~olfszz,m bell pepper, mdtcated that the biocontrol fungus, G wrens, was sensitive to so11solarizatton, and therefore could not be used m this combmatton. In contrast, the temperature tolerant biocontrol fungus Tularomycesflavus has been used successfully m combmatton with soil solarizatton (58). Soil solartzation, followed by the application of btocontrol agents, may improve disease suppresstveness. Seedling bioassays with corn indtcated that the btocontrol bactermm, B cepacia, could be used m combination with chemical pesticides to improve seedling emergence (9). Seedling vigor was better, when corn seedswere treated with thtram fungicide m combtnation with a peat-based formulation of B cepacia, than when either treatment was used alone. Btological agents, espectally those that are root-assoctated microorgamsms, may perform better than chemtcal or physical treatments of the soil. This ts because of then ability to colomze root tissue and have an effect on the deeper layers of the so11profile. In addmon, fungicides can protect plant seedlmgs only for a short duration, and the effect of soil solartzatton 1slimited in the upper few centimeters of the solI. Although chemical fungtctdes or soil solartzatlon has been shown to be effictent m suppressmg dampmg-off diseases on then own, there is a defimte advantage m combmmg these strategies with btocontrol agents.

Biological Control of Seedling Diseases

171

4. Current Status and Future Prospects Despite problems associated with the shelf life and delivery of blological agents, progress has been encouraging in terms of their acceptance as alternatives to chemical seed treatments when used alone or in conjunction with chemicals m an IPM strategy Currently, these questions are being asked. Can pathogens develop resistance to biocontrol agents; can their effectiveness be improved by combmmg two or more biological agents; will their effectiveness be restrlcted to certain kinds of environments; and 1sblologlcal control economically feasible? Commercialization of biocontrol agents 1sdependent on marketing assessment and determination of market availability and profit margins. Several factors should be considered, including the patentability of the formulation or strain of the blocontrol agent; the need m agricultural production systems for safe, reliable, nonchemical treatments for controlling plant diseases; the requirement for simple, mexpenslve fermentation systemsto produce biomass of blocontrol agents m large commercial scale fermenters; and the ability of biological pest control agents to be generally less damaging to the envlronment, and cheaper to develop, register, and market than chemical control compounds. Considering all of these factors, biological control products are begmmng to be recognized as commercially feasible for agricultural markets, and for plant protection strategies in general. The commerclahzation of the two blologlcal products mentioned in this chapter (Deny and SollGard) are good examples of how successful blologlcal control projects can matenahze, if a logical step by step approach, from the mltlal discovery, and ecological study, to the final development, 1sused. Also, close cooperative work between pnvate companies and public research institutions facilitates development of new and mnovatlve btocontrol products. 5. Note Mention of a trademark or proprietary product does not constitute a guarantee or warranty of the product by the USDA, and does not imply approval to the exclusion of other products that may also be suitable. References 1. Cook, R J. and Baker, K. F (1983) Nature and Practzce of Bzologzcal Control of Plant Pathogens APS, St.Paul, MN. 2 Lumsden, R D and Locke, J C ( 1989) Blologlcal control of damping-off caused by Pythum ultlmum and Rhuoctonza solam with Glzocladzum wrens m soilless mix Phytopathology 79,361-366 3. Lumsden, R. D., Lewis, J. A., and Fravel, D. R. (1995) Formulation and delivery of blocontrol agents for use against soilborne plant pathogens, m Bloratlonal Pest

712

4

5 6 7 8

9

10

11

12.

13 14. 15

16 17

Hebbar and Lumsden

Control Agents (Hall, F R. and Barry, J. W , eds ), American Chemtcal Society, Washmgton, DC, pp. 16&l 82 Lumsden, R D and Lewis, J A. (1989) Selection, productton, formulation and commerctal use of plant disease biocontrol fungi, problems and progress, m Btotechnology ofFungifir Improvtng Plant Growth (Whtpps, J. M. and Lumsden, R. D , eds ), Cambridge University Press, Cambridge, UK, pp 17 I-l 90 Bowen, G D and Rovira, A D (1976) Mtcrobtal colonization of plant roots Annu Rev Phytopathol 14, 121-144 Suslow, T V and Schroth, M N (1982) Rhrzoctoma of sugar beet effects of seed apphcatlons and root colomzatton on yield Phytopathology 72, 199-206 Kloepper, J W., Ltfshttz, R., and Zablotowtcz, R. M (1989) Free-living bacterial mocula for enhancing crop producttvtty Trends Btotechnol 7,3944 Hebbar, K P., Davey, A G , and Dart, P J ( 1992) Rhtzobacterra of corn antagonistic to Fusartum montforme, a sotlbome fungal pathogen tsolatton and tdenttfication Sod Btol Btochem 24, 978-987. Hebbar, K P , Martel, M. H., and Heulm, T (1994) Burkholderta cepacta, a plant growth promotmg rhrzobacterial assoctate of corn, in Improvtng Plant Producttvzty wzth Rhzzosphere Bacterza (Ryder, M H., Stephens, P M , and Bowen, G D , eds.), Proceedzngs, Thud International Workshop on Plant Growth Promoting Rhtzobacterta, Adelaide, Austraha, pp 201-203 Yaabucht, E , Kosako, Y., Oyalzu, H , Yano, I , Hotta, H , Hastmoto, Y , Ezakt, T , and Arakawa, M. (1992) Proposal of Burkholdena gen Nov and transfer of seven species of the genus Pseudomonas homology Group II to a new genus, wtth type species Burkholderza cepacra (Pallerom and Holmes, 1981) comb Nov Mtcrobtol Immunol 34, 125 1-l 275 Hebbar, K P , Davey, A. G., Merrm, J , McLaughlin, T. J., and Dart, P J (1992) Pseudomonas cepacta, a potenttal suppressor of corn sotlborne diseases Seed moculatton and corn root colomzatton Sod B1o1 Biochem 24, 999-l 007 Homma, Y , Sata, Z , Htrayama, F , Konna, K , Shnahama, H., and Suzuki, T. (1989) Productton of anttbtotrcs by Pseudomonas cepacra as an agent for biological control of sotlbome plant pathogens. So11Btol Btochem 21, 723-728 Parke, J. L. (1990) Population dynamics of Pseudomonas cepacta m the pea spermosphere m relation to blocontrol of Pythzum Phytopathology 80, 1307-l 3 11 Hebbar, K P , Berge, O., Heulm, T , and Smgh, S P (199 1) Bacterial antagonists of sunflower (Helzanthus annuus L) Fungal pathogens Plant Sod 133, 13 l-140 Hebbar, K P , Hackett, J. D , Fravel, D R , and Lumsden, R D (1995) Assoctatton of Burkholderta cepacta and Pseudomonas with corn and soybean roots and their role m suppressmg pre-emergence damping-off of soybean Abs Phytopath01 85, 1136 Burkholder, W H (1950) Sour skin, a bacterial rot of onion bulbs Phytopathology40,115-117 Lessee, T. G , Hendrtckson, W., Manning, B D , and Devereux, R (1996) Genomtc complextty and plasttcrty of Burkholderta cepacta FEMS Mcrobtol Lett 144, 117-128

Biological

Control

of Seedling

Diseases

113

18 Hebbar, K. P , Atkinson, D., Tucker, W , and Dart, P J (1992) Suppression of Fusarcum monrlrforme by corn root-associated Pseudomonas cepacla SolI Blol Bzochem 24, 1009-l 020 19. G~lhs, M., Van, V T., Bardm, R., Goor, M., Hebbar, P., Wrllems, A , et al. (1995) Polyphastc taxonomy m the genus Burkholderza leadmg to an emended descriptron of the genus and proposition of Burkholderza vletnamlensls sp Nov. for N2-fixing isolates from rice m Vietnam Int J Systematzc Bacterial 45,274-289 20 Jamsiewrcz, W J and Roitman, J (1988) Biological control of blue mold and grey mold on apple and pear with Pseudomonas cepacla Phytopathology 78, 1697-1700 21. Homma, Y., Chtkuo, Y , and Ogoshl, A (1990) Mode of suppression of sugar beet damping-off caused by Rhlzoctonla solam and Aphanomyces cochllodes by seed bactertzatton with Pseudomonas cepacla, in Plant Growth Promoting Rhczobacterza-Progress and Prospects (Keel, C., Koller, B , and Defago, G , eds ), Proceedzngs, Second Internattonal Workshop on Plant Growth Promoting Rhtzobacteria, Interlaken, Switzerland, pp. 115-l 18. 22. Barelmann I , Meyer, J M , Tarez, K., and Budztktewtez, H (1996) Cepaciachelm, a new catecholate siderophore from Burkhoiderla (Pseudomonas) cepacla. Z Naturforschung

51,627-630

23 Smirnov, V V , Kiprianova, E. A , Gargulya, A D , Dodatko, T A , and Ptlyaschenko, I I (1990) Anttbtotic activity and siderophores of Pseudomonas cepacla Appl Blochem and Mxroblol

26, 5843.

24 Meyer, J. M , Hohnadel, D., and Halle, F (1989) Cepabactm from Pseudomonas cepacla, a new type of siderophore J Gen Mlcroblol 135, 1479-1487. 25 Rottman, J N , Mahoney, N E., Janistewtcz, W J., and Benson, M (1990) A new chlormated phenyl pyrrole antrbiotic produced by antifungal bacterium Pseudomonas cepacla J Agrlc

Food Chem 38,538-541.

26 Fravel, D R , Marois, J. J , Lumsden, R D., and Conmck, W J , Jr (1985) Encapsulation of potential btocontrol agents m an algmate-clay matrix Phytopathology 75,774-777 27. Cartwrtght, K. D and Benson, D M (1995) Optimization of biological control of Rhizoctoma stem rot of Pomsetta by Paecllomyces ldaclnus and Pseudomans cepacla. Plant Dzs 79, 30 l-308 28 King, E B and Parke, J. L (1993) Btocontrol of Aphanomyces loot rot and Pythzum damping-off by Pseudomonas cepacza. Plant Dls 77, 1185-l 188 29 Lumsden, R. D , Garcia, E R., Lewis, J A., and Frtas, T G. A (1987) Suppression of damping-off caused by Pythzum spp m soil from the indigenous Mexican Chmampa agricultural system. Soul Bzol. Blochem 19, 50 l-508 30. McLaughlin, T. J , Quinn, J P., Betterman, A., and Bookland, R. (1992) Pseudomonascepacla suppression of sunflower wilt fungus and role of antifungal compounds m controlling dtsease. Appl Envzron. Mlcroblol 58, 1760-l 763 3 1, Renato-de-Frettas, J and Germida, J. J. (1991) Pseudomonas cepacla and Pseudomonasputlda as winter wheat inoculants for btocontrol of Rhzzoctonla solam Can J Mlcroblol 37,780-784.

114

Hebbar and Lumsden

32 Reddy, M S., Funk, L M , He, D N , and Pedersen, E A (1996) Status on commerclal development of Burkholderla cepacza for bIologIca control of fungal pathogens and growth enhancement of comfer seedlings for a global market, m Advances zn Bzologzcal Control of Plant Diseases (Whehhua, T , Cook, R J , and Rovira, A., eds ), Proceedings, International Workshop on BIological Control of Plant Diseases, BelJmg, May 22-27 33 Domsch, K H , Cams, W , and Anderson, T (1980) Cornpendzum of Sozl Fungi, vol 1 Academic, London 34. Farr, D F , Bills, G F , Charnuns, G P , and Rossman, A Y. (1989) Fungi on Plants and Plant Products zn the Unzted States American Phytopathologlcal Society, St Paul, MN. 35 Samuels, G J and Rehner, S A. (1993) Toward a concerpt of genus and species of Trzchoderma, m Pest Management Bzologzcally Based Technologzes (Lumsden, R D and Vaughn, J. L., eds.), American Chemical Society, Washmgton, DC, pp 186-l 88 36 Aluko, M 0 and Hering, T. F (1970) The mechamsms associated with the antagomstlc relationship between Cortzczum solanz and Glzocladzum wrens Trans Br Mycol Sot 55, 173-179 37 Howell, C R and Stlpanovlc, R D (1983) Ghovirm, a new antibtotlc for Glzocladzum wrens, and Its role m the bIologIca control of Pythzum ultrmum Can J Mzcrobzol 29, 321-324. 38 Jones, R W and Hancock, J G (1987) ConversIon of vmdm to vlrldlol by vlrldmproducing fungi Can, J Mzcrobzol 33,963-966 39 Lumsden, R D , Locke, J C., Adkms, S T , Walter, J F , and Rldout, C J (1992) Isolation and locahzatron of the antlbiotlc ghotoxm produced by Gliocladzum wrens from alginate prrll m so11and soilless media Phytopathology 82,23&235 40 Lumsden, R D , Rldout, C J , Vendemla, M E , Hamson, D J , Waters, R M , and Walter, J F (1992) Characterlzatlon of major secondary metabohtes produced m sollless mix by a formulated strain of the blocontrol fungus Glzocludzum wrens Can J Mzcrobzol 38, 1274-1280 41 Taylor, A (1986) Some aspects of the chemistry and biology of the genus Hypocrea and Its anamorphs, Trzchoderma and Glzocladzum Proc Nova Scotia Inst Scz 36,27-58 42 Roberts, D P and Lumsden, R D (1990) Effect of extracellular metabolltes from Glzocladzum wrens on germmatlon of sporangla and mycellal growth of Pythzum ultzmum Phytopathology 80,46 l-465 42a. Warmg, P , Elchner, R D , and Mulbacher, A (1988) The chemistry and biology of the unmunomodulatmg agent ghotoxm and related eprpolythlodloxoplperlzmes Med Res Rev 8,499-524 43 Mintz, A and Walter, J F (1993) A private Industry approach development of GhoGardTM for disease control m horticulture, m Pest Management Bzologzcally Bused Technologres (Lumsden, R D and Vaughn, J L , eds ), American Chemlcal Society, Washmgton, DC, pp 398-403. 44 Lumsden, R D , Locke, J C , and Walter, J F. (1991) Approval of Glzocludzum wrens by the U S Envlronmental Protectlon Agency for blologlcal control of Pythlum and Rhlzoctoma dampmg-off. Petrza 1, 138

Biolog~cai Control of Seedltng Diseases

115

45 Wtlhtte, S E., Lumsden, R D , and Straney, D. C (1994) Mutational analysts of gliotoxm productton by the biocontrol fungus Gltocladtum wrens m relation to suppression of Pythmm damping-off Phytopathology 84,816-82 1. 46 Howell, C R and Stipanovic, R. D. (1995) Mechamsms m the biocontrol of Rhtzoctonta solani-mduced cotton seedlmg disease by Gltocladtum wrens antibiosts Phytopathology 85469-472 47 Lumsden, R D., Walter, J. F., and Baker, C. P (1996) Development of G/locladtum vzrens for dampmg-off disease control Can J Plant Path01 18, 463-468. 48. Lumsden, R D. and Vaughn, J. L. (1993) Pest management’ biologically based technologies, m Proceehngs ofBeltsvtlle Symposzum XVIII, Beltsville, MD, May 2-6, American Chemtcal Soctety, Washington, DC, p. 435 49. Lumsden, R. D., Locke, J. C , Lewis, J A., Johnston, S. A , Peterson, J. L , and Rtstamo, J. B (1990) Evaluation of Gfzocladtum wrens for biocontrol of Pythmm and Rhtzoctoma dampmg-off of bedding plants at fout greenhouse locations Brol. Cult Control Tests 590 50 Beagle-Ristamo, J. E and Papavizas, G. C (1985) Biological control of rhtzoctoma stem canker and black scruf of potato Phytopathology 75, 560-564 51 Rtstamo, J B., Lewis, J. A., and Lumsden, R D. (1994) Influence of isolate of Gltocladtum vtrens on sclerotra of Sclerotwm rofin, soil mtcrobiota, and the incidence of southern blight Phytopathology 81, 1117-I 124 52 Jackson, M and Schisler, D H (1992) The composttton and attributes of Colletottwhum truncatum spores are altered by the mutational environment Appl Envtron Mtcrobtol 58,226@-2265. 53 Eyal, J., Baker, C F , Reeder, J D , Devane, W E , and Lumsden, R D (1997) Large scale productron of chlamydospores of Gltocladtum wrens strain Gl-2 1 m submerged culture. J. Zndust Mtcrobtol Brotechnol 19, 163-l 68 54 Lewis, J A and Papavizas, G. C. (1987) Application of Trtchoderma and Gllocladrum in alginate pellets for control of Rhizoctonia dampmg-off Plant Path01 36,438-446 55. Betz, F , Rispm, A , and Schneider, W (1987) Biotechnology products related to agrtculture Overview of regulatory decisions at the U S. Envnonmental Protection Agency. ACS Symposium series 334, American Chemical Society, Washington, DC, pp 3 16-327 56 (1989) Data requirements for pesticide registration; final rule. Federal Register 53, 15,952-15,999 57. Ristamo, J. B., Perry, K B., and Lumsden, R. D. (1996) So11 solarization and Gltocladtum wrens reduce the mcrdence of southern blight (sclerotrum rolfirr) m bell pepper m the field. Btocont Sci and Technol 6,583-593 58 TJamos, E. C. and Fravel, D. R. (1995) Detnmental effects of sublethal heatmg and Talaromyces flavus on microsclerotia of Verttctlltum dahltae Btol Control 85,388-392. 59 Papavtzas, G C , Lewis, J A , and Abd-El Moity, T. H (1982) Evaluation of new biotypes of Trtchoderma harztanum for tolerance to benomyl and enhanced biocontrol capabilities. Phytopathology 72, 126-132

116

Uebbar and Lumsden

60 Postma, J and Luttrkholt, A. J. G (1993) Selectton of benomyl-resrstant Fusarcum tsolates for ecologrcal studies on biological control of Fusarzum welt of carnation Neth J Plant Path01 99, 175-188 61 Burkhead, K D , Schtsler, D A , and Slmmger, P J (1994) Pyrrolmtrm productron by biologtcal control agent Pseudomonas cepacla B37w m culture and colomzed wounds of potatoes Appl Envwon hkcroblol 60,203 I-2039 62. Frrdlender, M Inbar, J., and Chet, I (1993) Btologrcal control of soilborne plant pathogens by beta-l ,3-glucanase-producing Pseudomonas cepacra. So11Bzol 63 Mao, W., Lewis, J A, Hebbar, K. P , and Lumsden, R D. (1997) Seed treatment with a fungal or a bacterral antagomst for reducmg corn dampmg-off caused by species of Pythzum and Fusarwm Plant Dis 81,450-454.Blochem 25, 1211-1221.

Joint Action of Microbials Claude Alabouvette

for Disease Control

and Philippe Lemanceau

1. Introduction

During the past 20 yr, more attention than ever has been given to the development of btological methods to control plant diseases.Indeed, the concern for food of high quality, wtthout residues of pesticides, and for a sustainable agrtculture that will preserve the fertthty of soil, and prevent the pollutton of the environment, has stimulated research dealmg with btological control At the same ttme, progress made m molecular technology has provided tools to study the modes of action of btocontrol agents. It 1snow possible to understand the mechanisms by which an antagonist can hmtt either the density or the activity of the target pathogen, and can induce resistance of the host plant. This knowledge should help to identify the environmental condmons favorable for application of biocontrol, and improve efficiency and consistency of biocontrol methods. Until now, practical apphcatton of btological control has been limited to a very few commerctal products effective against a limited number of pathogens in a few crops. One strategy to make biological control successful would be to associate in a single product several btological control agents having complementary or even synergistic modes of action against the same pathogen, or having antagonistic effects on several pathogens affecting the same crop. Usually, biological control agents have been selected for their efficacy toward a given pathogen, and therefore have a limited spectrum of targets. This target specificity is a disadvantage for practtcal use of these biocontrol agents, smce they have to be compattble with the pesttctdes required to control other pests and diseases affecting the same crop. An associatton of mtcroorganisms able to control several diseaseswill represent a great advantage for practical application of biological control Association of several microorganisms in a single product would also improve the consistency of the control. Indeed, btologtcal From Methods m Biotechnology, vol 5 B/opeshodes Use and Debvery Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

717

118

Alabouvette and Lemanceau

control IS often considered as less conststent than chemtcal control. Even tf It IS not always true, one must admit that brologrcal control 1smore dependent on envtronmental factors than chemical control. The idea to utthze combmatrons of antagomsttc mtcroorgamsms came from the study of soulsthat are naturally suppressive to diseasesinduced by sotlborne plant pathogens. Indeed, the control provided by the complex mteractrons responsible for sot1suppresslveness 1salways more consistent and stable with time than the control provided by a gtven strain of antagomstrc mtcroorgamsm, Isolated from the suppresstve so11 and mvolved m the mechanisms of suppressron. In this chapter dealing with Joint action of mtcrobtals for disease control, the first part will be dedrcated to a summary of knowledge resulting from the study of soils suppressive to fusarmm w&s, m order to show the great complextty of mechanisms responsible for consrstent control of a disease The chref modes of action by which antagomsttc mtcroorgamsms control diseases~111be presented m the second part Then, examples of the beneficial effects of assoclatton of several mtcroorgamsms having different modes of action will be presented, before dtscussmg prospects for then application as blologtcal control products, 2. An Example of Joint Action of Microbials: Soils Naturally Suppressive to Fusarium Wilts Suppressive soils are soils m which diseaseincidence or diseaseseventy remams low despite the presenceof the pathogenand environmental conditions favorable to diseaseexpressionon a susceptiblevartety of hostplant. Soils suppressiveto someof the most tmportantdiseasescausedby sotlbomeplant pathogenshave beendescribed, mdrcatmgthat so11suppresstvenessISnot a rare phenomenon(Z-3). Among the bestknown examples are soils suppressive to msarium wilts. The study of these ~011s demonstrated clearly that suppresstvenessIS based on mteractronsamong several mtcroorgamsmshaving different modesof action and acting togetherto conststently control the disease.The followmg summary of studiesdealing with soils suppressive to fusarrum weltswill not only illustrate the complex@ of mechanismsmvolved, but also the dtfficulty m reproducing sucha phenomenonby mtroducing selectedstrains of antagomsttcrmcroorgamsm m a conducive sot1 The first reports of fusartum-wilt-suppressive soils were made by Stover (4) m Central America, where banana planted m soils having a high content m Montmortllomte-type clays were less affected by the Panama disease than banana planted m solIs from another type Later, Smith andSnyder(5,6), studymg a suppressive sot1from California, made the observatton that thts so11was rich m Fusarium oxysporum, leading Toussoun (7) to state that soils suppressive to fusarmm wilts induced by pathogenic F oxysporum harbor high levels of saprophytrc fusarra. But It was still unclear whether so11suppressrveness was linked to sot1ablotrc characterrsttcs or to the sot1mtcroflora. Louvet et al.

Joint Action of Microbia/s

119

(8) were the first to establish the microbial nature of soil suppressiveness to fusarium wilts They demonstrated that suppressiveness disappears after heat treatments that destroy most of the microorgamsms, and is restored by mtroducmg a small proportion of the suppressive soil mto the heat-disinfested suppressive soil. It was also possible to make a conducive soil suppressive by mixmg 10 p 100 of suppressive soil in it. This global transfer of suppressiveness is achieved by the introduction of a sample of suppressive microflora mto the conducive soil. To determine the role of each population constitutmg the mixture, Rouxel et al. (9) isolated different types of microorgamsms from the suppressive soil and reintroduced them mto the heat-treated soil. Results showed that nonpathogemc F oxysporum and Fusarwn solam wet e involved m the mechanisms of soil suppressiveness, but the mechanisms by which the soils suppress disease remam obscure. Constdermg that addition of glucose to the suppressive soil made it conducive and stimulated the germination of chlamydospores of F oxysporum, Alabouvette et al (10) suggested that competition for carbon (C) could be one of the mechanisms by which soils suppress fusartum wilts. Alabouvette et al. (22) also demonstrated that the microbial biomass was greater and more responsive to glucose amendment m a suppressive than in a conducive soil, and concluded that both the general suppression caused by the activity of the total biomass of the soil and the specific suppression caused by the activity of the nonpathogemc Fusana were responsible for the suppressiveness of the soils from the Chlteaurenard area (12) At the same time, Kloepper et al. (13), studying the role of the populations of fluorescent pseudomonads m the rhizosphere of plant, suggested that they could contribute to control diseases.Following the same track, Scher and Baker (14,15) established that either a fluorescent strain of Pseudomonas sp, isolated from a suppressive soil, or its siderophore, can causea conducive soil to become suppressive to fusarmm wilts. Based on the fact that addition of a strong u-on (Fe)-chelator (EDDHA) also made a conducive soil suppressive, it was concluded that competitton for Fe was the chief mode of action of the siderophoreproducmg pseudomonads. Finally, Elad et al. (16,17), having established that the growth of germ tubes arising from chlamydospores was reduced m the presence of siderophore-producing pseudomonads, concluded that competition for Fe was the mechanism responsible for soil suppressiveness to fusarmm wilts. Almost at the same time, Schneider (18) isolated nonpathogemc strains of F. oxysporum from suppressive soils m California, and demonstrated that their addition to a conducive soil infested with F oxysporum f.sp. aplz contributed to hmttmg the severity of fusarium wilt of celery. Later, Pauhtz et al. (19) estabhshed that a Metz sandy loam from the Salmas Valley, suppressive to fusarium wilt of several crops, supported large populations of nonpathogemc F oxysporum, and suggested that they could contribute to soil suppressiveness.

120

Alabouvette and Lemanceau

Together, these data pointed clearly to the role of nonpathogemc fusarta and fluorescent pseudomonads m the suppresston of fusarmm writs m suppressive soils, and tt was necessaryto consider the role of both types of microorganisms. Park et al. (20) and Mandeel and Baker (21) considered the Joint action of nonpathogentcfusarta and fluorescent pseudomonadstn the mechanismsof suppression of t%sanumwelts,but the clearestdemonstrationof the postttve mteracttonbetween theseantagonistswas provided by Lemanceauet al. (22,23; see Subheading 4.). As demonstrated for soil suppressive to htsarium wilts, examples of sot1suppressive to other diseasesalso showed that several mtcroorganisms acting together or successively,and having different modes of action, are responsible for disease suppression.Therefore, microbtal assoctattonsmay be proposed to mimic the complex mteracttonsexisting m suppresstvesoils and achteve btological control 3. Modes of Action of Biological Control Agents To improve efficacy and consistency of btologtcal control, mtcrobtal assoclattons can be proposed However, the btocontrol agents have to be chosen m conJunctton with then modes of actton, whtch should not exclude each other, but, on the contrary, show a complementary or even a synergistic effect. Diseasecontrol may result from a dtrect antagontsmdirected against the pathogen,especiallydunng its saprophyttcgrowth phase,or from an indirect action through induced reststanceof the hostplant. The chief modes of action by whtch antagomsttc mtcroorgarusmscould control diseaseswill be reviewed quickly, keeping m mind that a smgle stramof btocontrol agent may expressseveral modes of actions. 3.1. Microbial

Antagonism

Mtcrobial antagonism implies direct interactions between two mtcroorgantsms that share the same ecologtcal niche. Three mam types of direct mteracttons may be characterrzed: parasitism, competition for nutrtents, and anttbiosts. These mteractions are not exclustve of each other; on the contrary, a gtven strain may possessseveral modes of actions, and tt IS often difficult to dtstmgutsh the relative tmportance of each of them m the efficiency of the observed antagomsm. Microbial antagonism occurs mostly during the saprophyttc phase of plant pathogens, and contributes to reducing the moculum density and/or the saprophyttc growth of the pathogen in the so11and at the root surface, resultmg in a decrease of the probabthty for the pathogen to achieve successful mfecttons of the host plant. 3.1.1. Parasitism Parasitism of plant pathogen by other mtcroorgantsms, including vnuses, IS a well-distributed phenomenon, but its significance in relation to btological control of plant diseases 1sstill controverstal.

Joint Action of Microbials

121

Many fungi contam vu-uses or vu-us-hke particles, and m a few cases this parasitism has been associated with reduced virulence of the pathogen. The best example ISthat of Cryphonectria parasitica, hypovirulent strains of which are being used as blologlcal control agents m several countries (24). Mycoparasltes, such as Conzothyrzum minitans and Sporldesmium sclerotlvorum, have been tested as blocontrol agents, and some of them are efficient in controlling diseases caused by Sclerotuzla spp and other sclerotiaforming fungi (25,26). The parasitic activity of strains of Trzchoderma sp has been extensively studled, and plays a major role in the antagonism expressed against Rhzzoctonza solani (27). But discrlmmation between parasitism and other modes of action 1sdifficult to establish, since cell-wall-degrading enzymes, such as chltinases and glucanases, are mvolved m the process of parasitism. Most strams of Trlchoderma spp possessseveral modes of action contributing to their blocontrol activity (28). Whether parasitism is one mode of action that can be deliberately associated with other modes of action to improve efficacy of blologlcal control has not yet been investigated. 3.1.2. Competition for Nutrients Because so11IS an ohgotrophic milieu, and because so11microorganisms and plant pathogens are heterotrophic, competition for C and energy IS strongly expressed m SOIL 3 1.2.1

COMPETITION

FOR CARBON

As already stated (see above), competltlon for C IS one of the mechanisms responsible for so11suppresslveness to fusarmm wilts. Competltlon for C IS expressed m every soil, and ISconsidered responsible for the well-known phenomenon of funglstasls, which describes the mhlbltlon of fungal spore germlnation m soil (29,30). Energy deprivation in sol1 is also partly responsible for “general suppression of a pathogen that is directly related to the total amount of mlcrobiologlcal activity at a time critical to the pathogen” (2). This general suppression results from the combined activity of several microbial populations, and, even if the mechanisms are not clearly understood, application of mlcroblal assoclatlons will increase the intensity of competltlon for nutrients. Any specific antagonism IS expressed on this background of general suppression, and any blocontrol agent apphed to sol1 will be submitted to soil fungistasis. Some species or strains of antagonists are more competltlve than others, and should be selected for biological control. For example, Couteaudler and Alabouvette (31) have shown that a great diversity exists among strains of nonpathogemc F. oxysporum in relation to their ability to utilize C efficiently. A slgmficant correlation was established between the ability of several strains

122

Alabouvette and Lemanceau

of nonpathogemc F oxysporum to mhrbtt the germmatron of the pathogen m the rhlzosphere, reduce drseaseIncidence of fusartum writ of flax, and compete efficiently for C, with the pathogenic F oxysporum m sot1 (32). Competitton for C has also been involved in the determtmsm of the antagonism expressed by different strains of Trcchoderma sp against several plant pathogens, especially F. oxysporum (33). 3.1 2.2. COMPETITION FOR MINOR ELEMENTS

Competttion for minor elements also frequently occurs m so11 As already stated (see above), competmon for Fe 1sone of the modes of action by whtch fluorescent pseudomonads hmtt the growth of pathogenic fungi and reduce disease incidence or severity. Under conditions of Fe stress,these bacteria synthesize stderophores, called pseudobactms or pyoverdms, which show a higher affinity for Fe3+than fungal srderophores. Numerous studies have associated the bacterial antagonism to pseudobactm synthesis, and several review papers are available (34-3 7). Other micronutrtents (Cu, Mn, Zn) also play a role in controllmg diseases induced by soilborne pathogens. Then- mode of action is not clearly established, but they contribute to some extent to sot1suppresstveness (38). The most important point to stress 1sthat competrtton for a given nutrient IS not exclusive from competmon for another nutrient, and, therefore, assoctatron of two antagomsttc mtcroorgamsms competing with the pathogen for two dtfferent nutrients may result m an increased efficacy of biocontrol. Moreover, competttton for nutrients 1snot exclusive from other modes of action, and may play an important role m the effictency of a btocontrol agent, even tf another mode of action has been investigated 3.1.3. Antibiosis Antrbtosts is the antagonism resulting from the production by one mtcroorgamsm of secondary metabolites toxic for another mrcroorganism. Antrbrosis 1sa very common phenomenon responsible for the btocontrol activity of many organisms, such as Pseudomonas spp, Bacillus spp, or Trlchoderma spp developed as btocontrol agents. A variety of different anttbtottcs, bactertocms, enzymes, and volatile compounds have been described, and are Involved m the suppressron of different pathogens. Several review articles are available (39-42). A given strain of brocontrol agent may produce several types of antifungal compounds, effective against certain species of fungal pathogens. For example, the production by fluorescent Pseudomonas sp of phenazmes and 2-4-dtacetylphloroglucmol was shown to be the prrmary mode of antagonism agamst Gaeumannomyces gramznzs var. triticz (43-45), but 2-4-dtacetylphloroglucmol and cyanide were

Joint Action

of Microbials

123

mvolved in the antagonism expressed against Chalura elegans (44,46,47). On the contrary, these secondary metabolites have not yet been implicated m the inhibition of the growth or activity of F oxysporum (37). Since a given strain often produces several of these metabolites, the best procedure to demonstrate the involvement of a given molecule in the antagonistic activity of the biocontrol agent is to produce mutants affected m their ability to synthesize the molecule, and to demonstrate that the deficient mutant is no more able to control the disease (41). But it is important to emphasize that a single antifungal metabohte generally does not account for all the antagomstic activity of a biocontrol agent (40). Therefore, it may be very useful to associate several strains of btocontrol agents producing different types of antifungal metabolites, to improve the efficacy or enlarge the acttvity spectrum of biological control. 3.2. Induced Systemic Resistance More and more studies are devoted to the resistance induced m the host plant by apphcatton of btocontrol agents. Induced systemic resistance classically occurs when an inducing agent is applied prior to challenge maculation with a pathogen, resulting in reduced diseasem compartson to the nomnoculated control. Kuc et al. (48) reported systemic protection of cucumber against Colletotricum orbzculare when the cotyledons or the first leaves of the plant were premoculated with the same pathogen. It has also been well established that the premoculatton of an host plant with an incompatibleforma speczalzs or race of F oxysporum will result in reduced disease severity when the plant IS moculated with the compatible pathogen (49). Therefore, it was suggested that the nonpathogemc fusaria used to control fusarium welts may be effective through induced resistance (22). Using one experimental design wtth a split-root system, which allowed application of the biocontrol agent on one side and the pathogen on the other side, Ohvam et al. (50) demonstrated that mduced systemic resistance contributes to the biocontrol efficacy of a nonpathogemc strain of F oxysporum The fluorescent pseudomonads, selected for their plant-growth-promotmg capacity or for their biocontrol activity, have been shown to induce systemic resistance in the plant (51). The first evidence was given by Van Peer et al. (52), who demonstrated that root colomzation of carnation by a stram of fluorescent Pseudomonas sp resulted in an accelerated and increased accumulation of phytoalexms m the stem of carnation after moculatton with F oxysporum f.sp. dianthl. Many other biocontrol agents are able to induce resistance m the host plant, and several recent review papers are available (48,51,53). Induced resistance is not exclusive from other modes of action, and may exert a complementary

124

Alabouvette and Lemanceau

effect to microbial antagonism. Indeed, direct antagonism usually limits the saprophytic growth of the pathogen, resulting m a decreased number of infection sites, and induced resistance ltmrts the growth of the pathogen during its parasitic phase inside the plant. Because the control resulting from an associatton of several modes of action is generally more effective and consistent than the control provided by a single mode of action, it would be interesting to associate several modes of action m combmmg several microorgamsms for biologrcal control. 4. Associations of Microorganisms for Biological Control and Growth Promotion 4.7. Microbial Associations for Biological Control One of the best-documented examples of a microbial associatton used to improve efficacy and consistency of btological control is provided by the association of strains of nonpathogemc F. oxysporum with strains of fluorescent Pseudomonas spp, tsolated from soils suppressive to fusartum wilts. From a theoretical point of vrew, competition for C between pathogemc and nonpathogemc F oxysporum, the existence of which has been demonstrated in the suppressive soils from Chiteaurenard, was not contradictory with the existence of competmon for Fe, as demonstrated m the suppressive soils from the Salinas Valley. Therefore, considering the two hypotheses, Lemanceau et al (54) established that both competition for C and competmon for Fe drd exist m the suppressive soils from Chateaurenard, even if the populatrons of fluorescent Pseudomonas spp isolated from the suppressive soil were not more competitive for Fe than the populations isolated from a conductve sot1 (54). Addition of C with EDDHA in a conducive soil resulted m an intermediate level of receptivity between the htgh conduciveness observed after addition of C and strong suppressiveness after addition of EDDHA (55). These observations prompted the hypothesis of a complementary effect of nonpathogemc F oxysporum with fluorescent Pseudomonas spp. Indeed, followmg a specific screening procedure, Lemanceau and Alabouvette (56) isolated from the suppressive soil strains of fluorescent Pseudomonas spp able to improve the efficacy of biological control achieved by the application of a strain of nonpathogemc F oxysporum. Park et al. (2U), followmg another approach, also showed that interactions between Pseudamonasputzda and strains of nonpathogemc F. oxysporum could achieve biocontrol of fusarmm wilts. The mechanisms of this beneficial mteraction remained obscure until Lemanceau et al (22,23), using a siderophore-deficient mutant ofP putzda strain WCS358, demonstrated that competition for Fe resulting from the activity of the bacterial strain enhanced the efficacy of competition for C between strains of F. oxysporum. Indeed, the growth yield of a stram of F oxysporum growmg on a single source

Joint Action

of Microbials

725

of C was greatly reduced m the presence of the bacterial siderophore pseudobactm 358. Moreover, it was shown that the nonpathogemc strain Fo47 was less sensitive to pseudobactm-mediated Fe competition than the pathogenic F. ox~~~orurn f.sp. dzanthi strain WCS816. These data together demonstrated that competition for Fe, resultmg from stderophore production by Pseudomonas spp renders more severe competrtton for C, resulting from the activtty of both the total biomass and the nonpathogenic F oxysporum. These mechanisms, which exist in naturally suppressive souls,may be used to achieve biological control of fusarium wilts by introduction of selectedstrams of nonpathogemc F oxysporum associated with fluorescent Pseudomonas spp mto conducive substrates. Several experiments carried out under commercial-type conditions have demonstrated the validity of such an approach. The control provided by the association of the nonpathogemc F oxysporum strain Fo47 with the P fluorescens strain C7 was always better and more consistent than the control achieved by either one or the other biocontrol agent (56,57). Fungi other than F. oxysporum can be associated to fluorescent Pseudomonas spp to achieve biological control of fusarium wilts. Coinoculation m pot bioassays ofAcremomum rutdum and Verticillium lecanil with different strams of Pseudomonas spp, significantly suppressed disease,compared with the control treatment, if the microorganisms were applied m moculum densities that were ineffective m suppressmg disease as separate inocula (58). Nonpathogemc strains of F. oxysporum have also been associated with other bacteria, such as Bacillus sp, but, rn the case of fusarium wilt of chickpea, the association of a strain of Baczllus sp did not improve the control achieved by a strain of nonpathogenic F. oxysporum (59). Other microbial associations, such as Trichoderma spp with Pseudomonas spp have also been studied to improve efficacy of biological control, but the results were often contradictory and ~111be discussed below (see Subheading 5.). On the contrary, several studies have demonstrated that association of bacteria with mycorrhizal fungi have a beneficial effect on plant growth. 4.2. Microbial Associations for Plant Growth Promotion The roots of most terrestrtal plants are inhabited by symbiotic fungi forming specialized structures known as mycorrhizas. Based on the mteractions established between the fungus and the plant root, two mam types of mycorrhtzas are distmguished: the ectomycorrhizas and the vesicular-arbuscular mycorrhizas.In both cases,the association between the fungus and the root occurs m the soil, and therefore can be influenced by other soilborne microorganisms Recently, some beneficial associations have been described. A recent review by Garbaye (60) shows how the symbiotic establishment of mycorrhizal fungi on plant roots is affected in various ways by the other micro-

126

Alabouvette

and Lemanceau

organtsms of the rhtzosphere Some bacteria, mostly fluorescent pseudomonads, consistently promote mycorrhtzal development. These bacteria have been called mycorrhizatton helper bacteria (MHBs) Several modes of action have been proposed to explain these postttve mteractions between bacteria and a mycorrhtzal fungus: The bactermm prohferatmg tn the rhizosphere before any mvolvement of the symbtottc fungus improves the recepttvity of the root to the mycorrhrzal formation; the bacterium interferes with the plant fungus recogmtion mechamsms, which are the first steps of the mteractton process leadmg to the symbtosts, the bacterium helps the growth of the fungus in its saprophytic stage in the rhizosphere, or on the root surface; the metabolic activtty of the bacterium multtplymg m the rhtzosphere modifies the physicochemical properties of the soil m a way to facilitate mycorrhizal mfectton, the bacterium triggers and accelerates the germination of the spores or other dormant propagules specialized m the conservation of the fungus m the soil. Whatever the mechanism, the use of helper bacteria as an adjuvant of fungal moculum is considered m order to improve mycorrhtzatton of trees There are also several papers reporting the synergtsttc effect of the association of VA-mycorrhiza with nitrogen (N)-fixing bacteria For example, Azcon et al. (61) described selecttve interacttons between different species of mycorrhizal fungi and strains of Rhizoblum meliloti apphed to Medzcago satwa. Depending on the combmatton of strains tested, there was a significant increase of the concentration and/or content of N m the shoots. This increase may be a consequence of a phosphorus-mediated sttmulation of N-fixation by VA-mycorrhiza, but m other cases the increase m mtrogen content seems to reflect a VA-mycorrhtzal-mediated enhancement of N uptake from the soil. In another example, Paula et al. (62) reported synergistic effects of VA-mycorrhtzal fungi and drazotrophtc bacteria on nutrition and growth of sweet potato Tuber production and N and phosphorus accumulatton were increased when diazotrophtc bacteria were applied together with VA-mycorrhtza-fungal spores. Thts beneficial effect seems to be caused by an enhanced mycorrhization of the plant in the presence of the bacteria. These few examples show that assoctatton of several mtcroorgamsms IS not only useful for biocontrol of plant drseases, but tt may also contribute to enhance plant growth.

5. Use of Microbial Associations for Biological Control: Prospects and Constraints Although experiments conducted under commerctal-like condtttons have shown great mterest m using microbtal assoctations to improve efficacy and consistency of btological control, practical appltcations of these mixtures need

Joint Action

of Microbials

127

further mvestlgatlon. It 1s first necessary to evaluate the compatlbilrty of the microorganisms under various envlronmental condltlons, and, second to develop production and formulation processes leading to a commercial product easy to handle and apply. 5.1. Compatibility

Between Microorganisms

To show an additive effect, the mlcroorgamsms to be associated must be fully compatible. They must establish together m the rhlzosphere of the host plant, without excluding each other by competltlon or antiblosls. This compatlblllty may be influenced by environmental condltlons, as shown by the contrasting data resulting from the assoclatton between strains of Trzchoderma spp with fluorescent pseudomonads Dandurand and Knusden (63) failed to demonstrate any beneficial effect of the associatlon of a strain of Pseudomonasfluorescens with a strain of Trlchoderma harzzanum m controllmg Aphanomyces root rot of peas, although the presence of the bacteria stimulated the hyphal growth of the fungus orlgmatmg from the coated pea seeds. Hubbard et al (64) reported that a strain of Trzchoderma hamatum applied as comdla to pea seedswas effective in controllmg seed rots caused by Pythzum spp in some soils, but not m other soils. This failure was a result of the antagonism exerted by fluorescent pseudomonads that colomze seed coats and lyse germlmgs of T hamatum on treated seeds This antagonism between fluorescent pseudomonads and T. hamatum was controlled by Fe avallabihty m the soil, and addition of ferrous oxalate to soil permitted T hamatum to protect pea seeds. In vitro experiments confirmed that extracellular compounds produced by fluorescent pseudomonads, under condltlons of Fe deprivation, were responsible for mhlbltlon of T hamatum, and suggested that pseudomonads mhlblt T hamatum through the production of slderophores. Studying the interactions between fluorescent pseudomonads and VA mycorrhlzal fungi, Pauhtz and Lmdermann (65) failed to show any beneficial effect of the VA mycorrhizal fungi on the population density and activity of the fluorescent pseudomonads. The mteractlons depend both on the fungal species and the bacterial strain. The population density of the bacteria was lower m the rhizosphere of cucumber roots colonized by Globus intraradices than m nonmycorrhizal plants. But this difference was not detected when the mycorrhizal fungus was Globus etunacatum. If some strains of fluorescent pseudomonads delayed the germination of G etunicatum spores, none of the bacterial strains affected the colomzation of cucumber roots by the mycorrhlzal fungus These results show how complex and specific are the mteractlons between several strains of microorganisms benelklal for plant growth. It is therefore impossible to draw any general conclusion.

128

Alabouvette and Lemanceau

Havmg selected mdependently, two efficient microorgamsms do not guarantee that their association will result m an increased beneficial effect. The best approach to create effective combinattons of beneficial microorganisms IS to isolate the candidate microorgamsms from the same ecological niche, and to develop specific screening procedures. For example, the nonpathogemc F. oxysporum strain Fo47 and the P jluorescens strain C7 have been isolated from the same suppressive soil, and then selection has been conducted in a biotest m which the two mtcroorgamsms have been confronted alone or together to the host plant and its pathogen (56). Only 8% of the bacterial strains associated with Fo47 increased the biocontrol capacity of the nonpathogemc F oxysporum. Then their compatibihty has been studied under various environmental conditions, m soil and m rockwool, m the rhizosphere of different plant species (66). Moreover, it has been established that the nonpathogemc F oxysporum strain Fo47 was much lesssusceptibleto the pseudobactm produced by a strain of P.Juorescens than a strain of F. oxysporunzf.sp. dzanthi (23). More attention should be given to the natural associations of microorganisms m order to select antagonists that could be used m combmations.

5.2. Development of Microbial Products Development of microbial products based on association of microorgamsms has not yet been achieved Obviously, the microorganisms will have to be produced separately by liquid or solid-state fermentation, but the question is whether they can be mixed m a single formulation, or have to be formulated and stored separately, to be mixed at the moment of application. Fermentation and formulation processes represent important steps in the development of a biocontrol product. Indeed, not only should a high biomass be produced at the lowest cost, but the properties of this biomass, i.e., tts capacity to control the disease, must be conserved during the processing and the storage Several review papers are available that discuss the principal requirements for producmg and formulatmg an active biomass (67,68). Obvtously conditions of production and formulation are different for fungi and bacteria, but, even for the same species, the results may differ from one strain to another. For example, Hebbar et al. (69) showed that the proportton of comdia vs chlamydospores of F oxysporum erythoxzll, used to control coca, varies according to the composition of the growth medium, and, for the same medium, varies from one strain of F oxysporum to another Formulatton of mtcroorgatnsms as seed coating requires spectfic conditions to preserve the viability of the microorgamsms, and to enable then growth in accordance with seed germmatton. The water activity 1san important factor to control, and it would be difficult to determine conditions favorable to the survival of a mixture of bacteria and fungi at the seed surface.

Joint Action

of Microbials

129

At the present time, cultures grown in hydroponics on artificial substratum (rockwool) represent an unique opportunity to apply microbial associations. In hydroponics, the microorganisms can be applied through the drip irrigation system, as a mixture, or successively, and repeated applications can easily be done. Microbial associations can also easily be applied to potting mixtures. It 1spossible to mtx each microorgamsm with one of the constituents of the mixture, and to use this organic matter as the support of the microorgamsms Peat and bran-peat mixtures are classically used to grow microorganisms, such as Trzchoderma and Glzocladzum, and peat is also used as a vector to apply Bradyrhizobium to seeds (70). This strategy is presently used m this laboratory to prepare a substratum, enriched with growth-promotmg microorganisms, to improve the successof acclimatation of vitro plants. The last step before commercial use of such microbial mixtures will be registration, and one must admit that it would probably be more difficult to prove the mocuity of a mixture than that of a single strain of microorganism. Today, there is no example of commerctal application of a biocontrol product assoctatmg several strains of antagonistic microorganisms.

5.3. Association of Several Modes of Action in a Single Strain Recently, another approach has been proposed to make biological control more successful. Rather than associatmg several strains of biological control agents m a single product, it has been proposed to associate several modes of action in a single strain of biocontrol microorganism. Accordmg to Roberts (71), three strategies can be followed to enhance the btocontrol performance of a bacterial strain through genetic engineering: adding biocontrol traits to bacterial biocontrol agents, modifying the regulation of expression of traits important to biocontrol, and enhancing the stability of the biocontrol activity. Several papers report attempts to associatein a given strain 0fP jluorescens or P. putzda the capacity to produce several of the secondary metabolites havmg antifungal properties. The mtroduction of DNA encoding the synthesis of an antagonistic metabolite m a nonproducmg strain usually resulted in an increased disease suppression by the transformed strain, in comparison with the parental strain. For example, insertion of the locus responsible for the synthesis of 2,4-dtacetylphloroglucmol mto nonproducer strains resulted m synthesis of phloroglucinol and increased mhibition of G. gramznzs var trztzcz and R. solani m vitro (43). The strain P3 of P j’haorescens had only a slight activity in controlling black root rot of tobacco, but the recombinant with cyanide-encoding genes from the strain CHAO showed an increased btocontrol activity (46).

130

AlabouvetteandLemanceau

Having demonstrated the synergistic effect of the assoctatton of two chttmolytic enzymes from T. harzzanum with cells of Enterobacter cloacae on the mhtbttion of spore germmatton of several pathogenic fungi, Lorito et al. (72) suggested transformmg the bacteria with the fungal genes encoding cellwall-degrading enzymes to improve the efficacy of those btocontrol bacteria. Today, the large possibilities offered by genetic engineering stimulate this type of research dealmg with improved strains for btological control. However, before practical application of these strains, many problems have to be solved and the successof this strategy will also depend on the acceptance by the public of the use of transgemc mtcroorganisms. 6. Conclusion Biological control of plant diseases,especially of plant diseasesinduced by soilborne pathogens, is presently restricted to a ltmtted number of commerctal preparations based on a single antagonistic microorganism Most, if not all, of the biocontrol agents control plant pathogens through several modes of action, havmg a complementary beneficial effect. Since a single antagonist does not possessall the possible modes of action, tt would be very Interesting to assoctate several antagonists, to improve efficacy and consistency of btologtcal control. Indeed, m compartson to biological control based on the application of a smgle microorgamsm, natural control of diseases,such as control provided by suppressive soils, appears complex, always involving several microorganisms, and several mechanisms. It seems possible to conclude that the more complex are the mteracttons, the better is the control achieved by the microorgamsms. However, this brief review of the literature shows that there are only a few examples of potential uses of microbial associations for biological control, or for promoting plant growth. Indeed, the selection of compattble antagonists requires a good knowledge of the modes of action of the antagomsts,and also of their ecological behavtor m different soils and m the rhizosphere of several plant species. Much more research 1sneeded before a mixture of beneficial microorganisms will be put on the market. However, as stressed above, there are a few specific niches, such as potting mixes and hydropomcs, m which apphcatton of mtcrobtal associations should be possible m the near future References 1 Baker, K F. and Cook, R J , eds (1974) Biological Control of Plant Pathogens American PhytopathologySociety,St Paul, MN 2 Cook, R J and Baker, K. F., eds (1983) Nature and Practice ofBlologwa1 Control of Plant Pathogens American PhytopathologySociety,St Paul, MN 3 Schneider,R W., ed (1982) Suppresswe Sods and Plant Disease American Phytopathology Society,St Paul, MN

Joint

Action

of Microbials

131

4 Stover, R H (1962) Fusarial wilt (Panama disease) of bananas and other Musa species CMI Phytopathol Pap 4, 117 p. 5 Smith, S N and Snyder, W C (1971) Relattonshtp of moculum density and soil types to severtty of fusarmm wilt of sweet potato Phytopathology 61,1049-l 05 1 6. Smith, S N and Snyder, W. C. (1972) Germination of Fusarzum oxysporum chlamydospores m soils favorable and unfavorable to wilt estabhshment Phytopathology 62,273-277 7 Toussoun, T A. (1975) Fusarmm-suppressive soils, m Bzology and Control of So&Borne Plant Pathogens (Brnehl, G. W , ed.), American Phytopathology Society, St Paul, MN, pp 145-15 1 8 Louvet, J , Rouxel, F., and Alabouvette, C. (1976) Recherches sur la resistance des sols aux maladies I-Mise en evidence de la nature microbtologtque de la resistance d’un sol au developpement de la fusartose vasculaire du melon. Ann. Phytopathol 8,425-436 9 Rouxel, F , Alabouvette, C , and Louvet, J. (1979) Recherches sur la r&stance des sols aux maladies IV-Muse en Cvtdence du role des Fusarzum autochtones dans la reststance d’un sol a la Fusariose vasculane du Melon. Ann Phytopathol 11, 199-207. 10. Alabouvette, C., Couteaudler, Y , and Louvet, J. (1985) Recherches sur la reststance des sols aux maladies XI-Etude comparative du comportement des Fusarzum spp dans un sol resistant et un sol sensible aux fusartoses vasculaires enrrchis en glucose. Agronomte 5,63-68. 11 Alabouvette, C , Couteaudier, Y., and Louvet, J (1985) Recherches sur la resistance des sols aux maladies XII-Acttvtte resptratotre dans un sol resistant et un sol sensible aux fusarloses vasculaires enrichis en glucose Agronomze 5,69-72 12. Alabouvette, C., Couteaudter, Y , and Louvet, J. (1985) Soils suppresstve to Fusarmm wilt mechanisms and management of suppressiveness, m Ecology and Management of Soil Borne Plant Pathogens (Parker, C A , Rovtra, A D., Moore, K. J., Wong, P. T. W , and Kollmorgen, J. F., eds ), American Phytopathology Society, St. Paul, MN, pp. 101-106. 13 Kloepper, J W , Leong, J , Temtze, M , and Schroth, M N. (1980) Pseudomonas stderophores a mechamsm explaining disease-suppressive soils Cur-r Mtcrobtol 4,3 17-320 14 Scher, F. M and Baker, R. (1980) Mechanism of biological control m a Fusarmmsuppressive soil Phytopathology 70,4 1224 17. 15. Scher, F. M and Baker, R (1982) Effect of Pseudomonasputtda and a synthetic iron chelator on mductton of so11suppresstveness to Fusarmm wilt pathogens Phytopathology 72, 1567-l 573. 16. Elad, Y. and Baker, R. (1985) Influence oftrace amounts of cattons and stderophoreproducing pseudomonads on chlamydospore germination of Fusartum oxysporum. Phytopathology 75, 1047-1052. 17. Elad, Y and Baker, R (1985) The role of competition for iron and carbon m suppression of chlamydospore germination of Fusartum spp by Pseudomonas spp. Phytopathology 75, 1053-l 059

132

Alabouvette and Lemanceau

18 Schneider, R. W (1984) Effects of nonpathogemc strams of Fusarrum oxysporum on celery root mfectlon by F oxysporum f. sp apll and a novel use of the lmeweaver-burk double reciprocal plot technique Phytopathology 74, 646-653 19. Pauhtz, T. C., Park, C S , and Baker, R (1987) BiologIcal control of Fusarmm wilt of cucumber with nonpathogemc isolates of Fusarlum oxysporum Can J Mzcrobzol 33,349-353 20 Park, C S , Pauhtz, T C , and Baker, R (1988) Blocontrol of fusarmm wilt of cucumber resulting from mteractlon between Pseudomonasputlda and nonpathogemc Isolates of Fusarlum oxysporum. Phytopathology 78, 190-l 94 21 Mandeel, Q. and Baker, R (1991) Mechamsms involved m blologlcal control of fusarmm wilt of cucumber with strains of nonpathogemc Fusarlum oxysporum Phytopathology 81,462-469 22, Lemanceau, P , Bakker, P. A. H M , De Kogel, W J , Alabouvette, C., and Schlppers, B (1992) Effect of Pseudobactm 358 production by Pseudomonas putzda WCS358 on suppression of Fusarmm wilt of carnations by nonpathogemc Fusarlum oxysporum Fo47 Appl Enwon Mlcroblol 58,2978-2982. 23 Lemanceau, P , Bakker, P A. H M , De Kogel, W J , Alabouvette, C , and Schlppers, B (1993) Antagonistic effect on nonpathogemc Fusarrum oxysporum strain Fo47 and pseudobactm 358 upon pathogenic Fusarlum oxysporum f sp dzanthz Appl. Environ Mlcrobrol 59,74-82 24 Van Alfen, N K., Jaynes, R A , Anagnostakis, S L , and Day, P R (I 975) Chestnut blight. biological control by transmissible hypovirulence m Endothla parasltlca Science 189,890,89 1 25. Adams, P B and Fravel, D R (1993) Dynamics of Spondesmwm, a naturally occurring fungal mycoparaslte, m Pest Management Bzologlcally Based Technologzes (Lumsden, R D. and Vaughn, J. L., eds ), American Chemxal Society, Washmgton, DC, pp. 189-195 26 Whlpps, J M and Lewis, D H (1980) Methodology of a chitm assay Trans Br Mycol sot 74,41&417 27 Elad, Y., Chet, I., Boyle, P , and Hems, Y. (1983) Parasitism of Trzchoderma spp on Rhlzoctoma solanr and Sclerotrum rolfsu. Scanning electron mlcroscopy and fluorescence mxroscopy Phytopathology 73,85-88 28 Lonto, M., Harman, G E., Hayes, C K., Broadway, R M , Tronsmo, A , Woo, S. L., and DI Pletro, A (1993) Chltmolytlc enzymes produced by Trrchoderma harzlanum antifungal activity of purified endochltmase and chltoblosldase Phytopathology 83, 302-307 29 Lockwood, J L (1977) Funglstasts m solIs Blol Rev 52, l-43. 30 Lockwood, J L (1988) Evolution of concepts associated with sodborne plant pathogens. Annu Rev Phytopathol 26,93-121 3 1 Couteaudler, Y and Alabouvette, C (1990) Quantitative comparison of Fusarlum oxysporum competltlveness m relation with carbon utlhzatlon FEMS Mlcroblol Ecol 74,261-268 32. Alabouvette, C. and Couteaudler, Y (1992) BiologIcal control of fusarmm wilts with nonpathogemc Fusana, m Blologzcal Control of Plant Diseases (TJamos, E C., Cook, R J , and Papavlzas, G C , eds ), Plenum, New York, pp. 415426

Joint Action

of Microbials

133

33 Sivan, A and Chet, I (1989) The posstble role of competmon between Trtchoderma harztanum and Fusarmm oxysporum on rhizosphere colonization P&opathology 79, 198-203. 34 Leong, J. (1986) Siderophores* then btochemistry and possible role m the btocontrol of plant pathogens Annu Rev Phytopathol 24, 187-209. 35 Schtppers, B., Bakker, A W , and Bakker, P A H M (1987) Interactions of deletertous and beneficial rhtzosphere mtcroorgamsms and the effect of cropping practices Ann Rev Phytopathol 25,339-358 36 Bakker, P A H M , Van Peer, R , and Schippers, B (199 1) Suppression of sonborne plant pathogens by fluorescent pseudomonads: mechanisms and prospects, m Development tn Agrtculturally Managed-Forest Ecology (Beemster, A B R , Bollen, G. J , Gerlach, M., Ruissen, M. A., Schippers, B , and Tempel, A , eds.), Elsevter, Amsterdam, pp 2 17-230. 37. Lemanceau, P. and Alabouvette, C. (1993) Suppression of msanum wilts by fluorescent pseudomonads. mechanisms and applicattons. Btocontrol Scz Technol 3,2 19-234 38. Hoeper, H. (1996) Importance of physical and chemical soil properties m the suppressiveness of soils to plant diseases Eur J Sot1 Bzol 32,41-58 39. Fravel, D. R. (1988) Role of anttbiosis in the btocontrol of plant diseases Ann Rev Phytopathol 26,75-9 1 40 Loper, J E and Lmdow, S E (1993) Roles of competmon and anttbtosts m suppression of plant diseases by bactertal brologtcal control agents, m Pest Management Biologtcally Based Technologtes (Lumsden, R. D and Vaughn, J L , eds ), American Chemical Socrety, Washington, DC, pp 144-l 55 41. Weller, D. M and Thomashow, L S. (1993) Microbial metabolttes with blological activity against plant pathogens, in Pest Management Btologically Based Technologzes (Lumsden, R D. and Vaughn, J L., eds ), American Chemical Society, Washington, DC, pp 173-l 80. 42. Alabouvette, C., Hoeper, H., Lemanceau, P., and Steinberg, C (1996) Soil suppressiveness to diseases induced by soil-borne plant-pathogens, m Sorl Bzochemistry, vol 9 (Stotzky, G. and Bollag, J. M , eds.), Marcel Dekker, New York, pp 371-413 43 Vincent, M N , Harrison, L. A., Brackm, J. M., Kovacevlch, P A , MukerJt, P , Weller, D M., and Pierson, E A (1991) Genetic analysts of the antifungal activity of a soil-borne Pseudomonas aereofactens strain Appl Environ Mtcrobtol 57,2928-2934 44. Keel, C., Schmder, U , Maurhofer, M., Votsard, C , Lavtlle, J , Burger, U , et al

(1992) Suppresston of root diseases by Pseudomonasfluorescens CHAO. Importance of the bacterial secondary metabohte 2,4-dtacetylphloroglucinol Mol PlantMtcrobe Interact. 5,413. 45 Harrtson, L , Teplow, D B , Rinaldi, M., and Strobel, G. (1991) Pseumycms, a famtly of novel pectides from Pseudomonas syrzngae possessing broad-spectrum antifungal activity J Gen Microbtol 137, 2857-2865 46. Voisard, C , Keel, C , Haas, D , and Defago, G. (1989) Cyanide productton by Pseudomonasfluorescens helps suppress black-root rot of tobacco under gnotoblottc condmons EMBO J 8, 35 l-358

134

Alabouvette and Lemanceau

47 Keel, C , Wnthner, P H , Oberhansh, T H , Volsard, C , Burger, U , Haas, D , and Defago, G (1990) Pseudomonads as antagonists of plant pathogens m the rhtzosphere role of the anttbtottc 2,4-dtacetylphloroglucmol m the suppression of black-root rot of tobacco Symbrosts 9,327-34 1 48 Kuc, J (1987) Plant tmmumzatton and its applicability for disease control, m Innovatzve Approaches to Plant Dtsease Control (Chet, I , ed ), Wiley, New York, pp 255-274 49 Btles, C L and Martyn, R D (1989) Local and systemtc reststance induced n-r watermelons by formae speciales of Fusarium oxysporum. Phytopathology 79, 856860 50 Olivam, C., Steinberg, C , and Alabouvette, C (1995) Evtdence of induced reststance m tomato inoculated by nonpathogemc strains of Fusarrum oxysporum, m Envtronmental Btottc Factors tn integrated Plant Dtsease Control (Manka, M , ed ), Polish Phytopathologrcal Society, Poznan, pp 427-430 51 Kloepper, J W , Zehnder, G. W., Tuzun, S , Murphy, J F , We], G , Yao, C., and Raupach, G (1996) Toward agrtcultural tmplementatton of PGPR-mediated induced systemic resistance against crop pests, m Advances rn Btologtcal Control ofPlant Dtseases (Tang, W., Cook, R. J., and Rovira, A , eds ), Proceedings of the International Workshop on Biologtcal Control of Plant Diseases, China Agricultural University Press, BeiJmg, May 22-27, pp. 165-174. 52 Van Peer, R , Ntemann, G J., and Schippers, B (1991) Induced resistance and phytoalexme accumulation m btologtcal control of fusarmm wilt of carnation by Pseudomonas sp strain WCS417r Phytopathology 81,728-734 53. Van Loon, L C. (1996) Drsease-suppressive actions of Pseudomonas bacteria induced resistance, in Proceehngs of a Workshop on Btologtcal and Integrated Control ofRoot Dtseases tn Sotlless Cultures (Alabouvette, C , ed ). IOBC/WPRS Bulletm, DlJon, September 18-2 1, 1995, pp 53-6 1 54 Lemanceau, P., Alabouvette, C , and Couteaudter, Y (1988) Recherches sur la resistance des sols aux maladies XIV-Modtficatton du mveau de receptivtte d’un sol reststant et d’un sol sensible aux fusartoses vasculanes en rtponse a des apports de fer ou de glucose Agronomte 8, 155-l 62 55. Lemanceau, P. (1989) Role of competition for carbon and iron m mechamsms of soil suppresstveness to fusarmm wilts, m Vascular Wilt Diseases ofPlants-Baste Studies and Control (TJamos, E C and Beckman, C H , eds ), NATO ASI Series, Springer-Verlag, Berlin, pp. 386-396 56 Lemanceau, P and Alabouvette, C (1991) Btologtcal control of fusarmm dtseases by fluorescent Pseudomonas and non-pathogemc Fusartum Crop Protectton 10,279-286 57 Alabouvette, C , Lemanceau, P , and Steinberg, C (1993) Recent advances m btologtcal control of fusarmm wilts Pestzctde Set 37,365-373 58 Leeman, M , Den Ouden, F M , Van Pelt, J. A., Comehssen, C , Bakker, P A H M , and Schtppers, B (1995) Suppresston of fusarmm welt of radtsh by co-moculatton of fluorescent Pseudomonas spp and of root colonizing fungi Eur J Plant Path01 102,2 l-3 1

Joint Action

of Microbials

135

59 Hervas, A , Landa, B , and Jlmenez-Dtaz, R. (1997) Influence of chtckpea geno-

type and Bacillus sp on protection from Fusarmm wilt by seed treatment wtth nonpathogemc Fusarmm oxysporum. Eur J Plant Pathol , 103,63 l-642 60 Garbaye, J (1994) Transley Revtew No. 7&--Helper bacteria a new dimension to the mycorrhizal symblosts. New Phytol 128, 197-2 10 61 Azcon, R , Rubto, R , and Barea, J M. (1991) Selecttve mteracttons between different species of mycorrhtzal fungi and Rhlzobzum melzlotl strains, and their effects on growth, N2-fixation (15N) and nutrrtron of Medlcago satlva L New Phytol 117,399404. 62 Paula, M A , Urqutaga, S , Srquerra, J 0 , and Doeberemer, J (1992) Synergistic

effects of vesrcular-arbuscular mycorrhrzal fungi and dtazotrophtc bacteria on nutrmon and growth of sweet potato (Ipomoea batatas) B~ol Ferttl SOJ~S 14,61-66 63 Dandurand, L M and Knudsen, G. R (1993) Influence of Pseudomonas ji’uorescens on hyphal growth and brocontrol activity of Trlchoderma harzzanum m the spermosphere and rhrzosphere of pea. Phytopathology 83,265-270 64 Hubbard, J P., Harman, G E , and Hadar, Y (1983) Effect of soilborne Pseudomonas spp on the btologtcal control agent, Trlchoderma hamatum, on pea seeds Phytopathology 73, 655-659 6.5 Pauhtz, T C and Lmderman, 66

67

68

69

R G (1989) lnteractlons

between fluorescent

pseudomonads and VA mycorrhizal fungt New Phytol 113,3745 Eparvrer, A., Lemanceau, P., and Alabouvette, C (1991) Populatton dynamtcs of nonpathogemc Fusarlum and fluorescent Pseudomonas strains m rockwool, a substratum for sotlless culture FEMS Mzcoblol Ecol 86, 177-l 84 Lumsden, R D and Lewis, J A (1989) Selection, productton, formulation and commercial use of plant disease btocontrol fungt. problems and progress, m Bzotechnology of Fungi for Improvmg Plant Growth (Whipps, J M and Lumsden, R D , eds ), Umverstty Press, Cambridge, UK, pp 17 l-2 17 Harman, G E. and Taylor, A G (1990) Development of an effectrve btologrcal seed treatment system, in Blologlcal Control of Sod-Borne Plant Pathogens (Hornby, D , ed.), C A B Internattonal, Wallmgford, pp 415426 Hebbar, K P , Lewts, J. A., Poch, S. M , and Lumsden, R D. (1996) Agrtcultural byproducts as substrates for growth, comdratton and chlamydospore formation by a potential mycoherbrctde, Fusarmm oxysporum strain EN4 Bzocontrol Scz

Technol. f&263-275 70 Catroux, G , Revellm, C , and Hartmann, A. ( 1996) Possible strategtes to improve the efficacy of mtcrobtal moculants and inoculatron methods, m Blologzcal and Integrated Control of Root Diseases zn Sodless Cultures (Alabouvette, C., ed ),

Working Group “Btological Control of Fungal and Bactertal Plant Pathogens”, IOBC/WPRS Bulletm 19, DlJon, France, September 18-2 1, 1995, pp 159-l 63 71 Roberts, D P. (1993) Genetically modified bacteria for btocontrol of sotlborne plant pathogens, in Pest Management Blologlcally Based Technologies (Lumsden, R D and Vaughn, J. L , eds.), Amencan Chemical Society, Washmgton, DC, pp 33&346. 72 Lortto, M , DI Ptetro, A , Hayes, C K., Woo, S. L , and Harman, G E (1993) Anttfungal, synergisttc mteractton between chttmolyttc enzymes from Trzchoderma harzlanum and Enterobacter

cloacae Phytopathology

83,72 1-728.

III BIOHERBICIDES

9 Neem and Related Natural Products Murray 6. lsman 1. Introduction Although botamcal msectlcldes once held a positlon of Importance m the grower’s arsenal of plant protectlon products, they were almost completely dlsplaced m most mdustrlahzed countries by synthetic insectrcides m the 1950s and 1960s. However, Increasing documentation of the negative environmental and health Impacts of synthetic neurotoxlc msectlcides and increasingly stringent government regulation of pesticides has resulted in renewed interest in the development and use of botamcal pest management products In spite of this interest, there remam only a handful of botamcal msectlcldes m use m North America and Europe, with few new products on the threshold of commerciahzatlon (Z-3). To this point, the market for botanical msectlcldes has been dominated by two plant preparations whose commercial productlon goes back over 150 years. pyrethrum and rotenone While synthetic pyrethrolds (chemicals loosely modelled after the natural msectlcldal constituents m pyrethrum) are among the most potent and widely used conventional msectlcldes, natural pyrethrum (from Chrysanthemum cinerariaefolium; Asteraceae) has maintained a small but consistent market share among so-called “alternative” pest control products Rotenone (from Derris elliptzca and Lonchocarpus spp; Legummosae) 1s still used to a small extent for insect control, but 1snow primarily used as a commercial plsclclde (fish poison), reflecting its original use over 300 years ago. Neither of these products enjoys wide use m conventional crop production, but they have been embraced by organic food producers. Other botanical msectlcldes have seen use in mdustrlahzed countries, but for various reasons have slipped from the marketplace or are used on a very hmlted scale. These include nicotine (from Nzcotzana tabacum, Solanaceae), From Edited

Methods by

m Biotechnology,

F R Hall

vol 5 Btopestmdes

and J J Menn

0 Humana

139

Press

Use and DeBvery Inc , Totowa,

NJ

140

lstnan

quassia (from QUUSSEU amara; Stmaroubaceae), ryama (from Ryan~~speclosa; Flacourtiaceae), and sabadilla (from Schoenocaulon officianale; Lihaceae). Why have so few botanical insecticides been developed? One overwhelmmg reason is that the discovery process, based on screening samples through bioassays with pests,has focused on materials that are acutely toxic to insects. However, acute toxicity 1snot the usual modus operandi m the real world of plant defensive chemistry, m which selection seems to have favored a more moderate approach: herbivore deterrence or dtscouragement. If an mvesttgator uses Insect mortality as the bioassay end point, as one would m the case of synthetic msecttctdes, it is not surprismg that so few plant preparattons have been discovered that have efficacy (in the laboratory) comparable to conventional msecttctdes. However, the demonstrated field efficacy and subsequent commercial development of botanical msecticides derived from the Indian neem tree (Azadzrachta zndlca; Meltaceae) have changed our basic assumptions about how a natural product must affect insects to be useful for plant protection on a commerctal scale. Neem functions primarily as an insect growth regulator, but also as a behavior-modifying substance, deterring feeding and/or oviposition m certain pest species (4). Of equal importance, neem has mmtmal toxtctty to vertebrates, is soft on natural enemies (5) and pollmators (61, and degrades rapidly m the environment. Neem serves as a paradigm for the development of other botanical msectictdes having nonneurotoxic modes of action (I), Environmental nonpersistence is an important trait that neem shares with other botanical insecticides. Although neem and perhaps other botanical preparations will prove to have superb efficacy in certain pest management contexts, for the most part it is unrealistic to expect botamcals to displace conventional msecttcides m agriculture and forestry, except where protection of the envtronment is paramount. The extent to which mainstream agriculture is prepared to accept botanical insecticides as legitimate products for plant protection may well depend on the successof neem m the marketplace m the next decade. 2. Materials 2.1. Choice of Plants There is a long list of properties that would be desirable for an ideal msectttide (for example, see refs. 2 and 7); to some extent, the mmimal presence of existmg botanical msecticides m the marketplace reflects the fact that they fall to meet some of the criteria necessary for commercial success.Neem has many desirable attributes, including efficacy at low concentrations, broad spectrum of action, mmimal nontarget toxicity, and no environmental persistence-but even neem has limitations, and is not a panacea for pest management.

Neem

and Related

Natural

Products

747

In reviewing the barriers to the commercialization of new botanical msecticldes (3), the following are listed as major barriers* abundance of the natural resource, standardization and quality control, and registration. In the final analysts, two constderattons are foremost: efficacy against one or more pests, and ongoing availability of the natural resource. Although many plant preparations can be used to mitigate pests, only a select group of these are sufficiently and reliably efftcacious to the point that people will actually purchase and use them repeatedly. However, the products that can deliver suitable performance must also be available to the manufacturer in quanttties sufficient to justify the costs of product development and production. To some extent, users will accept a Iowet absolute level of efficacy, but only if the product is safe to the user and the environment, mexpensive, and easily obtained and applied. Neem seeds, as the starting material for botanical msecticides, benefitted greatly not only from the widespread availabilrty of neem trees in India, but also from the fact that seedswere previously harvested and traded for the manufacture of soap. The foliage of Ginkgo biloba produces both medtcmal and msecticldal compounds (8,9); the latter (dtterpene lactones) can be obtained as a byproduct followmg extracting and removal of the more lucrative flavone glycostdes and proanthocyanidms, the compounds of pharmaceutical interest. If the starting plant material IS used exclusively for productton of insecticides, then it must be naturally plenttful, or, preferably, readily cultivated. In these cases, the cost of producmg the plant material can be a significant factor for development. 2.2. Tissue Harvested If the plant is grown strtctly for the production of natural msecttctdes, the best tissue to harvest must be established. This is especially the case for woody perennial plants, but the reader need only be reminded that pyrethrum is an extract of chrysanthemum flowerheads, and rotenone is obtained specifically from the roots or rhizomes of derrts. If the plant has other uses, then the harvest of biomass for msectictde productton has to be both sustainable and compatible with the other uses. The harvested part of the plant should be relatively abundant, at least seasonally, and easily harvested to mirnmize labor costs. Ideally, one wants to select the plant tissue that provides the optimal concentration of active mgredtent(s), and requires the mmtmal extraction, cleanup, and refinement. Seeds tend to be a good starting material, because they are often well protected m nature by secondary compounds. However, seed productton is often limited in species that are not prolific On the other hand, seeds can be a waste product of, for example, the fruit juice industry. Botanical insecticides can be

142

lsman

prepared from the seedsof soursop (Annona muncata, Annonaceae) and grapefruit (Cztrus paradzsz; Rutaceae), and m certam regrons the seeds of these species can be sourced m large quantmes at mmtmal (or no) cost. For species that have no other use, tt 1s advisable to screen vartous plant ttssues to determine the ttssue having the best combmatton of btoactivity and biomass avatlabrltty. McLaughlin and coworkers have done such an assessment on various plant parts of the pawpaw tree, Aszmzna trzloba (Annonaceae) (20) Although the seeds and unripe fruit produce the most acttve extracts, they are also the least available forms of btomass. The tissues found to gave the best balance of yield and btoacttvtty are the stem bark and wood, but because separation of the bark from the wood could be costly, the most appropriate ttssue for harvest on a sustainable basrs was deemed to be stems Further broassays establrshed that the smallest-size class of stems, namely, twrgs of C6.5 mm m diameter, were the most btoacttve. 2.3. Collection Sites Just as the quahty of agricultural commodmes varies between seasons and between locattons, so do the msecttctdal constituents of the plants that produce them. Although chemical vartabthty is a natural phenomenon, It 1s one that must be managed if a botanical msecttctde IS to be efficiently produced Chemtcal composttton of plants can vary at all levels among spectes, among and within populations, and between tissues Vartatton m the msecttctdal constttuents of Hawanan Zanthoxyhun (rutaceae) sp (II), and of the shrub Aglaaa odor&a (mehaceae) (12), are but two well-documented examples In the case of neem, numerous studies have been undertaken armed at determmmg the factors regulating variabthty of azadtrachtin content m the seeds, but none to date have been conclustve. For all intents, tt remains necessary to assessthe quality of neem seed, erther by chemical analysts (llqutd chromatography) or vta bioassay prior to commerctal-scale extractton. There ~111always be good and bad batches of starting maternal, but achieving a mmtmum acceptable level or standard can be accomphshed by blending lots of dtffermg qualtty, as 1sdone wtth other commodmes, such as coffee beans or tea Plant secondary compounds can be obtained through other means, though perhaps with decreased variabtlity. For example, pyrethrms can be produced by chrysanthemum tissue culture (13), as can azadtrachtms from Azadzrachta cell-suspenston cultures (14). Plant-ttssue culture offers several advantages for the productton of secondary metabolttes (opttmrzatton of productton, no seasonaltty, fewer co-extractives), but tt remains to be seen whether thts method can be cost effective, compared to harvestmg tissue from growing plants

Neem

and Related

Natural

Products

143

3. Methods 3.7. Screening for Bioactivity The wide range of insect species used for screening btoassays, and, stmtlarly, the wide range of bioassay types themselves, usually ensure that comparisons of results between laboratortes are tenuous at best. Assuming that the goal of the research is the discovery and development of an msecticide for the management of phytophagous pests of agriculture and forestry, tt makes sense to use a plant-feeding pest spectes as the primary screening organism. However, if screening 1s limited to one species, tt is easy to miss potentially useful bioacttvity against other types of pests, so a more thorough approach is to use a battery of bioassay species, Examples of this approach are the screening programs at the Rothamsted Experimental Station in the United Kingdom (15), the National Chemical Laboratory in India (16), and the Research and Development Corporation in Japan (17). In industry, tt IS not uncommon for as many as 10 pest species to be tested with candidate compounds. Although desirable, few laboratories can afford to mamtam contmuous cultures of more than three insect species, because of the human resources and direct costs required. Some mvestigators who have screened extracts from many plant species have relied on a single organism, e.g., brine shrimp (18) or mosquito larvae (17,19), largely because of the convemence m using these species. Though these bioassays are sensitive and reproducible mdicators of cytotoxtctty, they are not necessarily good predictors of btoactivity against agricultural pests (17). In the author’s research program, the primary screening species is the tobacco cutworm (Spodoptera Iztura), an noctuid pest of tobacco and vegetable crops m tropical and subtropical Asia. Active plant extracts are subsequently evaluated against a range of insects, rncludmg the migratory grasshopper (Melanoplus sanguinipes), the green peach aphid (Myzuspersrcae), the yellow mealworm (Tenebrzo molitor), and the large milkweed bug (Oncopeltus fasciatus). The primary bioassay, utilizing the cutworm, measures larval growth and survival of neonate larvae reared at 26°C for 10 d on arttfictal media, to which plant extracts (or fractions thereof, or pure compounds) are admixed Crude extracts are screened at 1000 ppm fresh wt (= 0 l%), pure compounds at 50 ppm. For compounds active at these concentrations, ECsO values (effective concentration reducing larval growth by 50% compared to controls) are established based on dose-response relattonships obtained usmg four or five lower concentrations An advantage of the larval growth bioassay is that tt can detect effects on the insect that have either behavioral or phystological bases,and effects resultmg from a wide range of modes of action. Many investigators have used antifeedant bioassays (either choice or no-choice feeding tests) for screenmg plant extracts. These strictly behavioral

744

lsman

bioassaysare no longer conducted on a routine basis for three reasons: A strong antifeedant wrll result m suppressedgrowth m the aforementioned diet bioassay, so a separate bioassay measuring feeding behavior alone IS not needed, differences between insect speciesare far greater for behavioral effects (feeding deterrence) than for physiological effects (toxrcity), and feeding behavior can be a function of bioassay duration, because insects can quickly habituate to feeding deterrents, as has been demonstrated in the case of azaduachtm, the prmcipal antifeedant from neem (20). The utility of antifeedantsper se as crop protectants is therefore questionable, given the plasticity of insect feeding behavior This variation in behavior response IS well exemplified by azadnachtm, the most potent insect antifeedant yet discovered Even closely related species of noctuid caterpillars differ srgmficantly m then behavioral responses to this substance, but their physiological responses are far more consistent (21). And, although azadirachtm is a profound antifeedant for the desert locust, it has no antifeedant effect agamst the migratory grasshopper (22) or the strawberry aphid (23), although both of the latter species are susceptible to the phystological actions of the compound Acute toxicity can be determmed through different modes of admmistranon, i.e., via direct topical application, or via exposure of insects to residues of test materials applied to glass plates or vials (24). A less precise but perhaps more realistic approach is to spray-test materials (m dilute alcoholic solutions) onto plants, either naturally or artificially infested with insects. Placmg insects onto freshly treated plants can indicate contact action of residues of the test material, rather than assessing the impact of the material hitting the pests directly. In these types of bioassays, mortality is normally assessedafter 24 or 48 h An important shortcommg of such experiments is that not all useful crop protectants result m pest mortality within 48 h; significant chronm effects can therefore be missed if acute mortality is the only bioassay endpoint considered. For example, neem msecticides often take 4-7 d to kill lepidopteran larvae (basically, at the time of the next molt), but these insects, though ahve m the mterim, often cease feeding almost immediately, so no further damage to the crop is inflicted. If we are to seriously consider the dtscovery and development of botanical msecticides, we must be prepared to consider many modes of action that do not result in acute mortality. 3.2. Extraction

Active prmciples from most existmg botanical msecticides are of moderate polarity and can be readtly extracted usmg alcohols of various origm. Important considerations m the choice of solvents include cost,safety (low flashpoint), and the potential for recycling Effluent disposal and flammability may preclude the use of organic solvents m some cases,but, for example, petroleum disttl-

Neem

and Related

Natural

Products

145

lates are used for the extraction of pyrethrum. There have been very few attempts to use exotic solvents and/or extraction technologies to obtain botamcal insecticides, though a supercritical fluid-extraction process (utilizmg liquid C02) for yielding concentrated pyrethrins from Chrysanthemum flowers has been patented (25). In the case of neem, the seedscan contam up to 40% by weight of oil, and tt is preferable to remove the oil, either through cold pressing or hexane extraction, prior to initial extraction for the azadirachtms. Note that neem oil itself, refined or clarified, can be used as a separate crop protectant targeted at certain soft-bodied pests. For some materials, it may be possible to use a crude extract, followmg removal of the extracting solvent, if the active constituents are present m sufficient concentration However, m most cases, some additional cleanup or refinement is necessary. This can normally be accomplished by some form of hquid-liquid partition, including the use of countercurrent extraction, on a commercial scale. The goal, though, should be to mmimize the number of steps needed to obtain a technical grade extract with an acceptable concentration of active ingredients. Extraction of insecticidal acetogenins from the bark of A. triloba provides an excellent example (28). Bioactivity (measured as brine shrimp toxicity) of the crude ethanohc extract is increased 4.5x by solvent partttion, but an additional partition of the organic phase leads to a further 42-fold increase m acttvity-almost to the same level as that of the mam constituent m purity. Both cost and yield of each step must be factored into the equation before the manufacturer can decide what level of refinement is justified. 3.3. Standardization For a botanical to be approved for use (i.e., registered) m industrialized countries, the putative active ingredient(s) must be specified and Its concentration guaranteed on the product label. It is therefore necessaryto standardize the technical grade material (refined plant extract). There are various chromatographic means of quantifying mdivtdual constituents of complex mixtures such as plant extracts, but high-performance liquid chromatography appears to be the most widely applicable for most types of insecticidal compounds found m plants. The active prmctples m plants almost always occur as suites of closely related structures, often comparable to one another in bioactivity when isolated. Thus, they are collecttvely considered to be the active ingredient, viz., pyrethrins in pyrethrum and azadirachtms m neem. Quantification of active ingredients is not only important for regulatory purposes, but also for trade; m the case of neem, the (total) azadirachtin content determines the price of the refined seed extract.

146

lsman

Neem kernels contam at least a dozen analogs of azadn-achtm.The chemlstry and biological activity of these are extensively revlewed by Kraus and Rembold (4). There are some significant differences m the behavioral (antlfeedant) and physlologlcal (growth-disrupting) effects of these compounds, but mvestlgatlons of structure-activity relations have revealed that the entlre carbon skeleton 1sessential for insect-growth-regulatmg activity. This reahzatlon, combined with the structural complexity of the azadirachtm molecule (containing 13 chu-al centers and 4 oxygenated rings), diminishes the prospect of syntheslzmg a simpler compound retaining the outstanding bloactivlty of the natural product. From a practical standpoint, the issue of structure-activity relations of the naturally occurring azadirachtms is largely a moot point, because, of the dozen or so compounds m neem kernels, two account for about 99% of the total. These are azadlrachtm proper (sometimes referred to as “aza A”), and 3-tlgloylazadlrachtol (frequently referred to as “aza B”). In analyzing 20 partially- to highly-refined neem kernel extracts via HPLC for hmonoid constltuents, we found that these two compounds occur m ratios of 2-6 to 1 (average 2 5.1), with azadlrachtin dominating (Isman et al., unpubhshed data). As an Insect growth regulator, 3-tigloylazadlrachtol ls substantially more active than azadlrachtm agamst some pest species (e.g., S. litura and Epzlachna vanvestu), but less active against others (e.g., Schzstocerca gregarza and Helzothis vzrescens). As an antlfeedant against noctuld larvae, 3-tlgloylazadlrachtol appears to be somewhat less active than azadlrachtm Given the above observations, considering the azadlrachtms collectively in quantitatmg the active ingredients m neem preparations seems to be a reasonable approach. 3.4. Formulation Some of the existing botamcals have been sold primarily m the form of dusts or powders, but these tend to be relatively inefficient with respect to delivery of toxlcant to the target pest and residual action on plant foliage More desirable, and more widely used, are emulsifiable concentrate (EC) formulations of botanicals. Most botamcals lend themselves well to the preparation of EC formulations. Because of then- moderate polarity, the technical grade extracts often dissolve readily m conventional alcohol-based carriers. There are also numerous conventional food-grade emulsifiers (e.g., ethoxylated glycerides or esters) that will produce stable aqueous emulsions of the dissolved extract. Compared to many synthetic msectlcides, the active principles m many botanical msectlcldes, mcludmg neem, tend to be very susceptible to photodegradation, and labile to oxldatlon m storage. To counter these deleterious actions, UV-absorbing adjuvants (i.e., sunscreens) and food-grade antloxldants

Neem and Related Natural Products

147

can be added to the formulation, although then ability to protect the active ingredients should be determined empirically. Some actives may be sufficiently stable to allow the preparation of ready-to-use formulations, with or without a propellant, for use m the home and garden.

3.5. Applications The predominant botamcals m current use, namely pyrethrum and rotenone, enjoy widespread use, at least in part, because they are broad-spectrum msectictdes. Newer botanicals, including several under development, have more subtle and varied modes of action, functionmg as moltmg disruptants, proteinsynthesis mhtbitors, or inhibitors of other specific enzyme systems. In some cases, it 1slikely that several modes of action are possible for a single compound. In the field, neem acts as a crop protectant largely through tts action as an insect growth regulator, but suppression of feeding through the insect’s central nervous system and behavioral effects (deterrence of feedmg and oviposition), and reduced mobility or vigor, cannot be discounted in some applications. Neem is active against a wide array of pest spectes, including members of most of the economtcally important insect orders, but, like some other botamcals, has poor contact action and IS efficacious only when ingested by the target pest Another consequence of the anti-hormonal actton of azadirachtm can be reduced fecundity following ingestion of sublethal doses by either larvae (Leptdoptera) or adult insects (Coleoptera). Whtle the Impact of this effect may not be nnmediately apparent (I.e., within a growing season), It can contribute to long term populatton reduction of pests. Such an effect could be particularly important for multtvoltme species where latter generations are the most potentially damaging to the crop. Against certain pests such as aphids, efficacy IS influenced to a large degree by the host plant, presumably reflecting the relative systemic movement of azadnachtm m different crop species (26). Though the systemic action of neem has been demonstrated in some important crop species and even m certain tree species, it 1sdangerous to assume this applies to all plants. Empirical studies with specific crops are clearly warranted. Other botamcals under development are more selective in their efficacy; if commercrahzed they will have to be aimed at mche markets where they can compete with conventional products. For example, the thiophene a-terthienyl is extremely effective against mosquito larvae, but variable against plant-feedmg pests (27), gmkgohdes, msectictdal diterpenes from Ginkgo bzloba foliage are effective against planthoppers, but only weakly active against lepidopteran pests (9); and hmonm, an abundant triterpene from grapefruit seed is a potent antifeedant for the Colorado potato beetle Leptinotarsa decemlineata, but relatively ineffective against leptdopteran pests (28).

lsman

148 3.6. Compatibility

with Integrated

Pest Management

Most botanical msecticides are relatively soft on nontarget arthropods, probably because they have to be Ingested to be effective (1.e , they lack contact toxtctty) and owing to their limited persistence on foliage. In particular, neem has been shown to have mmtmal impact on natural enemies of aphids (5) and predatory mites commerctally reared as biocontrol agents (29). Equally important, neem does not disrupt foraging by honeybees and other pollmators, nor does tt appear to pose a risk to bees (6,30). These properties suggest that neem insecticides will be quite compatible with integrated pest management in many crop ecosystems.Less is known regarding the effects of other botamcals under development on natural enemies and pollinators. If these products can be demonstrated to have reduced impact on nontargets, their attractiveness to growers will be enhanced. 4. Results 4.7. Recent Products Renewed interest m botanical msecticides is ostensibly a consequence of the recent mtroduction of neem into the market. The ortgrnal neem product, Margosan-OTM was introduced into the United Statesby W R. Grace (Columbra, MD) m 1990. Developed by Robert Larson of Vtkwood Botamcals (Sheboygan, WI) with the assrstanceof the U.S. Department of Agrtculture, tt consisted of an ethanol extract of ground neem seeds mixed with an emulsifier to a final concentratton of 0.3% azadtrachtm and approx 20% neem oil. This product contmues to be sold by Ringer (Mmneapolts, MN) under several trade names, including Safer Bto-Neem TM Margosan-0 was followed mto the marketplace by a range of products based on oil-free neem seed extracts, developed by AgrtDyne Technologies (Salt Lake City, UT) These products, contammg 3% azadnachtm as the active ingredient, mcluded AzatmTMfor nonfood crops and AhgnTM for food crops. Since 1992, several companies in India, as well as firms m Germany and Australia, have independently developed processes to obtain oil-free concentrated neem seed extracts m a dry form, typically contammg between 10 and 30% azadirachtm(s). From these technical grade extracts, EC formulations containing between 2 and 5% azadrrachtin can be readily prepared. ThermoTrrlogy (Columbia, MD) (having acquired both the Biopestictdes Division of W. R. Grace and AgriDyne Technologres) is currently marketing an oil-free neem msecticide contammg 4.5% azadtrachtm(s) under the name Neemtx 4.5TM Among other companies, Trifolio-M GmbH in Lahnau, Germany has developed products under the tradename of NeemAzalTM that are currently being sold m India, and for which regulatory approval in Germany is imminent. For-

Neem and Related Natural Products

149

tune Btotech (Secunderabad, India) is pursuing registration of neem msectitides in the United States. Closely related to the Indian neem tree IS the chinaberry tree, Melza azedaruch (meliaceae). The seeds of this tree contam msectictdal constttuents from which a pest-control product can be prepared, but the seeds also contam meliatoxms, substancesthat are toxic to vertebrates (31). However, the bark of this tree and that of Melza toosendan (considered a race of A4. azedarach by some authors, a distinct species by others) contam hmonoid triterpenes that can be used to produce a botanical msecttcide (32). Such a product 1scurrently manufactured m the Peoples’ Repubhc of China, for use against frutt, nut, and vegetable pests. The active ingredient listed on the label is 0.5% toosendanm, but the commercial product is known to contain 0.1% of a synthetic pyrethroid or other conventional msecticide. Trials m North America using the Chinese product, and formulated bark extract alone, indicate that the latter has only limited efficacy as a stand-alone product, suggesting that the limonoids function primarily as a synergist to the small amount of synthetic insecticide m the commercial formulatton. 4.2. Botanicals Under Development Several other botanical materials have been the subject of considerable SCIentrfic investigation and could proceed to commerctal production if suitable parties m the private sector are prepared to shepherd them through the regulatory process. A number of these botamcals are reviewed elsewhere (2). Insecticides prepared from the seeds of soursop fruit (A muricata) and the twigs of the pawpaw tree (A tnloba) have been the subjects of U.S. patents (33,3#). In both cases, the active ingredients are annonaceous acetogenins. Field trrals demonstrate that plant extracts containing these compounds are effective against a wide range of insect and mite pests, and they act synergistttally when combined with pyrethrum or neem (35). In purity, the acetogenins show stgmficant mammalian toxicity, although some are particularly effective as antitumor agents. It can be argued that, as crude, complex mixtures in relatively low concentrations, annonaceous acetogenms could be effective for pest management without posmg appreciable risks to humans and wtldhfe, a situation not unlike that for pyrethrum and rotenone. Crude foliar extracts of iVicotiaaa gossez and related species, prepared by rinsing foliage with dichloromethane, are very effective for the control of soft-bodied arthropod pests (36). The active compounds m the extracts are sucrose esters having acyl substituents (Ct-ClO) on both the fructose and glucose moietres. However, commercial synthesis of these compounds is not economically feasible, and their recovery from Nicotmna foliage may not be, either.

150

lsman

A standardized extract from seeds of Meha volkenm, a species native to East Africa and related to the neem tree, shows good potential for management of a number of pest insects, especially the yellow fever mosquito, Aedes aegyptz, and the desert locust, S. gregarza (37) Prehmmary results of laboratory and field trials m the United States suggest that this material could be efficacious againstagricultural pestsm temperatecountrtes,aswell (H. Fescemyer, personal communication) Other botanical preparations that have been touted for pest management include root extracts of marigolds (Tagetes spp), rich m throphenes, as a mosquito larvrcide (27), fohar extracts of G biloba for control of the brown planthopper on rice (9), limonin from grapefruit seeds for control of the Colorado potato beetle (38), and seed extracts of lupms (Lupznus spp), rich in qumohzidme alkaloids, which function as feeding deterrents and msecticides to a wade range of pests (39) 4.3. Future Trends With consumer and polmcal pressure for reductions m pesticide usage m agriculture and forestry, and increased awareness of the nontarget impacts of garden pesticides among homeowners, the prospectus for botanical msecticides 1sthe most favorable it has been for 50 yr. As we reach the mtllenmum, with proper marketing and continued refinement, we should see neem msecticides approach the current use levels of pyrethrum. Neem 1scurrently being evaluated as an alternative to synthetic pyrethroids for protection of cotton in China and Australia Cotton represents the largest single market sector for insecticides, so even modest successesm these trials could lead to an explosive Increase m the global demand for neem products. Increasing use of any new msecticide raises the specter of pest resistance In this regard, botanical materials consistmg of mixtures of active prmciples may have an advantage over conventional synthetic insecticides. Artificial selection experiments with diamondback moth larvae (Plutella xylostella) (40) and green peach aphid nymphs (A4 permae) (41) suggest that pest species cannot readily evolve resistance to neem-based msecticides, even though, m the same experiment, selection with pure azadirachtm led to the development of nmefold resistance to this natural product m the aphid. Other botamcal products, mcludmg some of those mentioned m this chapter, should reach the market for specialty uses; few of the products under development ~111have the widespread apphcabtlrty m agriculture and forestry that neem appears poised to attam. Also, botamcals should Increase their share m the home and garden (domestic) msecticide market, perhaps reaching 50% by volume in 1&15 yr. Up to the present, regulatory approval, designed around synthetic pestictdes, has constituted a barrier to the introduction of new botamcals, primarily because

Neem

and Related

Natural

Products

151

botamcals usually consists of complex mixtures of active ingredients, and can have more than one mode of action in pests. However, there IS evidence that regulatory authorities, having gamed experience with botanicals through evaluatton of neem insecticides, are likely to look more favorably on these alternative products, with procedures aimed at moving new products into the marketplace with fewer impediments.

References 1 Isman, M B (1994) Botanical msecticides Pestzczde Outlook 5, 26-3 1 2 Isman, M B (1995) Leads and prospects for the development of new botanical msecticides Rev Pestle Toxzcol 3, l-20 3 Isman, M B (1997) Neem and other botamcal msecticides barriers to commercialization Phytoparasltzca 25, 339-344. 4 Schmutterer, H , ed (1995) The Neem Tree VCH, Wemheim, Germany 5 Lowery, D. T and Isman, M. B (1995) Toxictty of neem to natmal enemies of aphtds Phytoparasltzca 23,297-306 6. Naumann, K and Isman, M B. (1996) Toxicity of neem (Azadwachta wzdrca A Juss) seed extracts to larval honeybees and estimation of dangeis from field appllcations Am Bee J 136,5 18-520 7 van Beek, T A and de Groot, A (1986) Terpenoid antifeedants Part I An overview of terpenoid antifeedants of natural origm Recued des Travaux ChzmlqueF des Pays-Bas 105,5 13-527 8 O’Reilly, J (1993) Gznkgo bzloba-cultivation, extraction and therapeutic use of the extract, m Phytochemlstry and Agrzculture (van Beek, T. A and Bretelei, H , eds ), Clarendon, Oxford, pp 253-270 9 Ahn, Y J , Kwon, M , Park, H. M , and Han, C G (1997) Potent msectictdal activity of Ginkgo bzloba derived trilactone terpenes against Ndaparvata lugens, m Phytochemzcalsfor Pest Control (Hedin, P A , Hollmgworth, R M , Masler, E P , Miyamoto, J., and Thompson, D G , eds.), American Chemical Society, Washmgton, DC, pp 90-105. 10 Ratnayake, S , Rupprecht, J K , Potter, W. M , and McLaughlm, J L (1991) Evaluation of the pawpaw tree, Aszmznu trlloba (Annonaceae), as a commercial source of the pesticidal annonaceous acetogenms, m New Crops (Jamck, J and Simon, J E , eds.), Wiley, New York, pp. 644-648 11 Marr, K L and Tang, C. S. (1992) Volatile insecttcidal compounds and chemical variabihty of Hawaiian Zanthoxylum (Rutaceae) species Bzochem Syst Ecol 20, 209-217 12 Satasook, C , Isman, M B , Ishibashi, F , Medbury, S , Wniyachitra, P., and Towers, G H N (1994) Insecticidal bioactivity of crude extracts of Aglaza species (Mehaceae) Bzochem Syst Ecol 22, 121-127 13. RaJasekaran, T , Perena, J , Ravishankar, G A., and Venkataraman, L V (1996) Repellency of callus derived pyrethrins to mosquito Culex qurnquefasclatus Say and red flour beetle Trlbollum castaneum Herbst Int Pest Control 38, 155-159

152

lsman

14 Morgan, E. D., van der Esch, S A., Jarvts, A P , Macctom, O., Gtagnacovo, G , and Vttale, F (1996) Productton of natural msecttctdes from Azadzrachta spectes by tissue culture. Abstract, International Neem Conference, Lawes, Australia. 15. Khambay, B P. S. and O’Connor, N (1993) Progress m developmg msecttcides from natural compounds, in Phytochemzstry and Agrzculture (van Beek, T A and Breteler, H , eds ), Clarendon, Oxford, pp 40-61 16 Sharma, R N (1984) Development of pest control agents from plants a comprehenstve workmg strategy, m Natural Pesticides from the Neem Tree and Other Tropical Plants (Schmutterer, H and Ascher, K R S , eds ), GTZ, Eschbom, pp 55 l-563 17 Escoubas, P., LaJide, L , and Mizutam, J (1994) Insecttctdal and anttfeedant activtttes of plant compounds. potenttal leads for novel pesttctdes, m Natural and Engineered Pest Management Agents (Hedm, P. A., Menn, J. J., and Hollmgworth, R M , eds ), American Chemical Society, Washmgton, DC, pp 162-l 7 1 18 Alkofahl, A., Rupprecht, J. K., Anderson, J E., McLaughlin, J L , MikolaJczak, K L , and Scott, B. A (1989) Search for new pesttctdes from higher plants, m Insectwdes of Plant Orrgrn (Amason, J T , Phllogene, B J R , and Morand, P , eds.), American Chemical Society, Washmgton, DC, pp 2543 19. Cepleanu, F , Hamburger, M 0 , Sordat, B , Msontht, J D , Gupta, M P , Saadou, M , and Hostettman, K (1994) Screening of troptcal medtcmal plants for mollusctctdal, larvtctdal, fungtcidal and cytotoxic activities and brme shrtmp toxtctty Int J Pharmacog 32,294-307. 20. Bomford, M. K and Isman, M. B. (1996) Desensmzatton of fifth mstar Spodoptera lztura (Lepidoptera. Noctutdae) to azadtrachtm and neem Entomol Exp Appl 81,307-313 21 Isman, M. B. (1993) Growth mhtbttory and anttfeedant effects of azadtrachtm on SIX noctutds of regtonal economic tmportance Pestlclde Scl 38,57--63 22. Champagne, D. E., Isman, M B , and Towers, G. H. N (1989) Insectictdal acttvtty of phytochemicals and extracts of the Mehaceae, m Insectzczdes of Plant Ongzn (Arnason, J. T., Phtlogene, B. J. R., and Morand, P., eds.), Amertcan Chemical Society, Washington, DC, pp 95-109 23 Lowery, D. T and Isman, M B. (1993) Antifeedant acttvity of extracts from neem, Azadwachta mdlca, to strawberry aphtd, Chaetoslphon fragaefolu J Chem Ecol 19, 1761-1773. 24. Isman, M B , Proksch, P , and Yan, J -W. (1987) Insecttctdal chromenes from the Asteraceae: structure-activity relattons. Entomol Exp Appl 43,87-93 25. Sims, M. (1981) Ltqutd carbon dioxide extraction of pyrethrins US Patent No 4,281,171 26 Lowery, D. T and Isman, M B (1994) Insect growth regulatmg effects of neem extract and azadtrachtm on aphids Entomol Exp Appl 72, 77-84 27. Amason, J. T., Phtlogene, B. J. R , Morand, P., Imne, K., Iyengar, S., Duval, F., et al (1989) Naturally occurrmg and synthetic thiophenes as photoactivated msecttctdes, m Znsectzcides of Plant Orzgzn (Amason, J T , Phtlogene, B J R , and Morand, P , eds ), American Chemical Soctety, Washmgton, DC, pp, 164-l 72

Neem

and Related

Natural

Products

153

28. Mendel, M J., Alford, A. R., and Bentley, M. D. (1991) A comparison of the effects of hmonm on Colorado potato beetle, Leptrnotarsa decemltneata, and fall armyworm, Spodoptera frugtperda, larval feeding. Entomol Exp. Appl 58, 191-194. 29. Spollen, K. M. and Isman, M. B. (1996) Acute and sublethal effects of a neem msecticide on the commercial biocontrol agents Phytosetulus perstmtlts and Amblysetus cucumerts (Acari Phytoseiidae), and Aphzdoletes aphtdtmyza (Rondam) (Diptera. Cecidomyiidae) J. Econ. Entomol 89, 1379-1386 30 Naumann, K , Currie, R. W , and Isman, M. B. (1994) Evaluation of the repellent effects of a neem msecticide on foraging honey bees and other pollinators Can Entomol 126,225-230. 3 1 Ascher, K R S , Schmutterer, H., Zebitz, C. P. W., and Naqvi, S. N H (1995) The Persian lilac or chmaberry tree: Melta azedarach L , m The Neem Tree (Schmutterer, H , ed ), VCH, Wemheim, pp. 605-642 32 Chiu, S -F. (1995) Melra toosendan Sieb. & Zucc., m The Neem Tree (Schmutterer, H , ed), VCH, Weinheim, pp. 642-646. 33. Moeschler, H. F., Pfuger, W., and Wendlisch, D. (1987) Pure annonm and a process for the preparation thereof. US Patent No. 4,689,323 34 MikolaJczak, K L , McLaughlin, J. L., and Rupprecht, J. K. (1988) Control of pests with annonaceous acetogenins. US Patent No. 4,721,727. 35. McLaughlin, J. L., Zeng, L , Oberlies, N H , Alfonso, D , Johnson, H A , and Cummings, B. A. (1997) Annonaceous acetogenms as new natural pesticides. recent progress, m Phytochemicals For Pest Control (Hedm, P. A., Hollingworth, R. M , Masler, E. P , Miyamoto, J , and Thompson, D G , eds ), American Chemical Society, Washmgton, DC, pp. 117-133. 36. Pittarelh, G W , Buta, J. G., Neal, J W , Jr., Lusby, W. R., and Waters, R M (1993) Biological pesticide derived from Nicottana plants US Patent No. 5,260,28 1 37 Rembold, H and Mwangi, R W. (1995) Melta volkensu Gurke, m The Neem Tree (Schmutterer, H., ed ), VCH, Wemheim, Germany, pp. 647-652. 38. Murray, K. D., Alford, A. R., Groden, E., Drummond, F. A., Starch, R H., Bentley, M. D., and Sugathapala, P. M. (1993) Interactive effects of an antifeedant used with Bacrllus thurmgtensu var. san dtego delta endotoxm on Colorado potato beetle (Coleoptera: Chrysomehdae). J Econ Entomol 86, 1793-l 801 39. Wink, M. (1993) Production and application of phytochemicals from an agricultural perspective, m Phytochemtcals and Agrtculture (van Beek, T A. and Breteler, H , eds ), Clarendon, Oxford, pp 171-213. 40 Vollinger, M. (1995) Studies of the probability of development of resistance of Plutella xylostella to neem products, m The Neem Tree (Schmutterer, H., ed ), VCH, Wemhelm, Germany, pp. 477483 41. Feng, R. and Isman, M. B. (1995) Selection for resistance to azadnachtm m the green peach aphid, Myzus perstcae Expertentta 51,83 l-833

10 Commercial

Experience

with Neem Products

James F. Walter 1. Introduction The agricultural mdustry of the 1990s IS challenged to find new methods and materials for controllmg pests and diseases. New legtslatton, mcludmg the 1996 Food Quality Protection Act, Worker Protection Standard, and Pesttctde Reregtstratton, are limiting the avatlabtltty of traditional chemical pesticides. Governmental policies committed to the institution of integrated pest management (IPM) programs, the mcreasmg resistance developed by msects and pathogens to chemical pesticides, and the public concern about chemicals m general has u-utiated a re-evaluation of pesticide use. increasingly, farmers m developed and developing nations are looking toward the use of natural materials as pest-control agents (I). Neem-based msectlcldes containing azadtrachtm address these concerns. The insect-growth regulator (IGR), azadtrachtm, affects over 300 species of Insects, including such important pests as armyworms, leafmmers, aphids, whiteflies, psyllids, and numerous other insect pests (2). In addition to controllmg these pests, many azadirachtm-based msecttcides have negligible effect on natural beneficial insects, and low environmental impact (3). These properties make azadlrachtm a sensible material to use m most pest-management programs. However, significant manufacturing, regulatory, and application problems had to be solved before azadxachtin could be brought to the market. Since then mtroduction mto the agricultural market m the United States m 1993, azadirachtmbased pesticides are fast becoming an important tool in crop protection, although the total amount of azadirachtm sold 1smuch less than 1% of all Insecticides sold.

From Methods in Blofechnology, vol 5 B/opestmdes Use and Delwery Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

155

156

Walter

2. Where Does Azadirachtin Come From? Extracts and parts from the neem tree have been used for centuries to control numerous insect pests and diseases,and as a therapeutic substance The neem tree, or Azadzrachta zndzca (A. Juss), 1sa member of the mahogany family (4). The trees are a hardy, broad-leaved evergreen that can reach heights of 100 ft It grows prlmarlly m tropical regions where the rainfall 1s<32 m./yr, and IS native to such areas as India, Pakistan, Indonesia, and Thailand. Neem has spread throughout the world, and various amounts of neem can be found m such diverse areas as Australia, Sudan, Senegal, Saud1Arabia, Haiti, Mexico, Nicaragua, and Hawaii (5). The dlstrlbutlon of neem 1slimited, because it 1sa tropical tree and cannot withstand cold winters. Azadlrachtin is found m varymg degrees in all parts of the plant. However, the highest concentration of azadn-achtm 1susually m the seeds. Currently, the maJorlty of neem-based msectlcldes are produced m India, with smaller amounts being produced m Sri Lanka, Australia, Nicaragua, and Thalland In India, seeds from the neem tree fall m early spring, after the monsoon rams The seedsare collected by local farmers, cleaned, dried, and brought to local trading centers, where they are auctioned to the highest bidder. From the trading centers, the seeds are moved to the extraction plant, where manufacturmg-grade or end-use products are produced. The vast maJorlty of neem seed collected in India goes mto the extraction of neem oil, which is used to produce soaps and toothpastes (6). Less than 5% of the neem seed collected in India goes into the production of neem insecticides. There are many different grades of neem-based extracts produced m India, ranging from crude products that primarrly contam oil from the neem tree and only a small fraction of azadlrachtm (<0.15%), to highly refined extracts of relatively high purity (>20%). In general, the cruder materials are uncharacterlzed and thus have little, if any, quality control. The more highly refined products are, m general, more highly characterized and thus more consistent. There are obvious differences between the characterlstlcs of neem-based pestlcldes, but even mmor varlatlons m formulation can effect the overall product quality (7) A great variety of manufacturing processes are used to produce neem msectlclde, and the varlablhty m product quality has led to a series of inconsistent reports on the activity, efficacy, and toxicity of azadlrachtm. For instance, Schafer and Jacobson (8) report that the acute oral toxlclty (LD,,) for expressed or extracted neem oil 1s>lOOO mL/kg for red-winged blackbirds. But Sharma et al. (9) report that the acute oral LD,, for neem seed 0111s 39.9 mL/kg for chickens. Larson (6) reports the LDsO for the commercial product Margosan-0 (Vlkwood Botamcals, Sheboygan, WI) 1s>I 6 mL/kg for mallard ducks. &mllarly, Chopra (10) reports that neem-seed 011produced occasional diarrhea,

Commercial

Experience

with

Neem

Products

157

nausea, and general discomfort when given orally to human adults. A 4-mo-old infant died after being administered 12 mL of neem oil for a cough (II). Yet neem oil is commonly used as an oral medication in India, and oral doses of neem as high as 6 mL/kg caused no mortality in female albino rabbits (22), Margosan-0 produced no mortality for rats when fed at a rate of 5 g/kg (10) These reports mdicate that the toxicity and chemical composition of neem extracts can vary significantly. This makes the comparison of different extracts difficult. Although azadirachtin has been identified as a key mgredient, neem extracts can contain dozens of other extractable materials that can influence the toxicity and efficacy of the extract. In particular, aflatoxms have been identified as a potentially toxic contaminant of neem extracts. Other limenotds include such compounds as salannm, mmbandiol, nimbm, and deacetyl mmbinbandiol, all of which have insecticidal activity. This has created confusion in the literature and led to unsubstantiated claims regarding azaduachtin. In order for a product to be registered and suitable for use m developed countries, a consistent,reliable, efficacious product had to be developed. The development of neem msecticides was expensive and drawn out, and took many years of development. 3. The History of the Commercialization of Azadirachtin-Based Pesticides in the United States The development and commercialization of refined azadirachtm-based msecticides has been spearheadedby two companies m the United States.This is somewhat ironic, since there is very little neem grown in the United States,and the regulatory environment in the United States is very strict. However, these two factors probably stimulated the development of efficient, consistent, azaduachtm extraction techniques. The history of azadnachtm-basedmsecticides m the United Statesis very convoluted, with several parties changing hands and products changing names, and constant improvements m technology. The first commercial use of an azadnachtm-basedpesticide for nonfood use was approved by the U.S. Environmental Protection Agency (EPA) in 1985. Vikwood Botamcals, owned by Robert Larsen, introduced Margosan-0 for use on trees and shrubs to control leafmmers and gypsy moths. This product was developed m part with the assistance of the United States Department of Agriculture (USDA) m Beltsville, MD, and was tested throughout the world Margosan-0 contamed an ethanohc extract of neem seedswith 0.3% azadirachtin (6). However, because of manufacturing and formulation problems, Margosan-0 production was limited to sample quantities, and, although it became an academic international standard, it had little commercial impact. In 1988, W R. Grace (New York, NY) purchased the patent, registration, and technology for Margosan-0 from Larson. Over the next several years, Grace improved the manufacturing and formulation technology for azadnachtin, which resulted in a much more consistent product.

158

Walter

In 1990, Grace changed the formulatton of Margosan-0, reducmg the active content to 0.25%, expanded the registration to include several important insect pests, mcludmg whiteflies, aphids, and armyworms, and expanded its use to include the greenhouse and mteriorscape environments Working through its partially owned subsidiary, Grace/Sierra Horticultural Products, Grace mtroduced Margosan-0 m the greenhouse/nursery mdustry (14). Margosan-0 had a caution label and no specific handling requirements Sales of Margosan-0 were primarily targeted to the control of whiteflies on pomsetttas and other ornamental crops, and the product successfully established the commercial viability of azadtrachtm-based insecticides. In 1992, Agrtdyne Technologies (Salt Lake City, UT) received registration for, and introduced, Azatm to the greenhouse market and Turfplex to the lawn care Industry Both Azatm and Turfplex contamed 3% azadn-achtm m a naphalene solvent, and carried a warning label. Azatm made inroads mto the greenhouse market because of its higher concentration and price position, but sales of Turfplex were weak and the product was dropped after a few seasons. Grace introduced the product Bioneem (0 09% azad), through Ringer, for the consumer pesticides market. Because of their nontoxic mode of action on insects and then inherent low toxictty, the EPA has created rational guidelmes for the registration of neembased msecttcides In 1993, the EPA granted an exemption of tolerance for using azadn-achtm on all food crops at ~20 g azadirachtm/acre. Grace received approval for, and Introduced, Neemix (0 25% azadirachtin) for use on food crops. Inmal sales efforts were targeted on vegetables m Florida. Simultaneously, Grace, m collaboration with an Indian partner, started up the world’s largest azadirachtm plant m Tumkur, India (15, Neemtx found good acceptance m the citrus and vegetable markets for control of such pests as armyworms, leafmmers, and aphids. In 1994, Grace received registration for a 4.5% azadirachtm formulation, and introduced Neemix 4.5 to the agricultural market The 4.5% formulatton, being 18 times more concentrated than the 0.25%, reduced the difficulty of handling large volumes of material and simplified package dtsposal. Also m 1994, Grace sold Grace/Sierra Horticultural Products to the Scotts Company (Marysville, OH). Grace retained ownership of the biopesticide business, but Margosan-0 was renamed Neemazad. In March 1996, AgrtDyne Technologies was sold to Biosys of Columbia, MD. Biosys is a nematode and pheromone producer, and this acquismon was seen as a strategic fit In May 1996, while divesting itself of noncore businesses, Grace sold its biopesticide busmess to Therm0 Ecotek, which formed a new company, ThermoTrdogy (Columbia, MD), to market and develop these biopesticides (16). In September 1996, Biosys filed for bankruptcy (17) In

Commercial

Experience

with Aleem Products

159

January 1997, ThermoTrilogy bought the assetsof Biosys out of bankruptcy, and thus became the sole producer and supplier of azadrrachtm m the United States.With the acquisition of the Biosys assets,ThermoTrilogy has become a much more dtversrfied bropesticide company. It is drffrcult to predict rf azadirachtm will remam a focus of the company. 4. Development of Azadirachtin-Based Insecticides Outside of the United States Outside of the United States, India is the largest user of azadn-achtm-based msecticrdes. Because of familiarity and a captive supply, companies in India have produced and distributed for many years neem-based pesticides that claimed to contain azadirachtin. Various manufacturers, including West Coast Herbochem (Neemark) and Godrej Soaps (Neemacure), produced neem-oilbased pesticides that were unregistered and did not specify an azadirachtm content, because no formal registration system had been established. In 1993, the government of India formalized the registration process for neem-based msecticides. They created a registration system, unique to India, based on the azadirachtm content of the product. Crude neem-oil-based formulations are required to contam 0.0 15% azadirachtm More refined extracts were allowed to contam 0.15,0.3, 1,2, or 5% azadirachtin. Acute toxicity and chemical testmg is required for all formulations. Many small companies entered the market with oil-based azadirachtm formulations. The overall quality and effectiveness of these products has been rather mconsistent, because of variabilmes m oil quality. Several large Indian companies have entered the azadirachtm pesticide market with higher-concentration, and higher-quality, products, mcludmg Spit (Neem Gold), E.I.D. Parries (Neemazal) and MargoBrocontrols (Econeem). Most of the azadirachtm-based pesticides produced m India are used on three mam crops: tea, cotton, and vegetables (28). Other than m the United States and India, the use of azadu-achtm-based msecticides is sparse, but the approval of new registrations in many countries may change this situation. Registration for neem-based pesticides manufactured by ThermoTrilogy or Biosys have been approved m Saudi Arabia, Taiwan, Israel, Spain, Chile, Mexico, Nicaragua, Costa Rtca, and Ecuador A German company, Trrfoho, has received registration of an azadirachtm-based msecticide (Neemazal) in Switzerland, and has applied for registration m Germany. Despite the great amount of fundamental research conducted on azadirachtm m Europe, only recently has azadirachtm been registered m a few countries there. These new registrations should allow for the expanded use of azadn-achtm (19). Along with the registered uses for azadirachtm, there are several countries where azadirachtin is used without formal registrations Australia and Indone-

760

Walter

sta have several plantings of neem, and mdrviduals have been sellmg unregistered azadu-achtm-based msecticides for a number of years. Although most of the products are used as mosqmto repellents or cures for head lice, some have been used on food crops. Development projects in Kenya, Senegal, Thailand, Nicaragua, Phihppmes, Ham, and other countries, supported by European or American governments, have started up rudimentary neem extraction plants or tramed local farmers to produce neem-based msecticides (18). The exact amount of azadnachtm used at this level is unmonitored, and, thus, it IS difficult to estimate the true extent to which neem is used. 5. How Does Azadirachtin Work? The primary active ingredient m most neem pesticides ISa compound called azadnachtm. Although neem extracts can, and usually do, contain other compounds that can control insects or influence the activity of azadnachtm, neem pesticides generally only specify their azadtrachtm content. Azadtrachtm is a hmmotd or, more spectfically, a tretranor triterpenoid with great msecttctdal activity. Azadnachtm 1schemically very complicated and has not been chemically synthesized. Azadirachtm has numerous effects on msects; however, Its major modes of actton are that of a powerful IGR, a feedmg deterrent, and an oviposttion deterrent. These three modes of action give azadu-achtm unique properties that make it very useful m today’s agricultural mdustry. Most farmers are not familiar with these modes of action, and need to understand them, so that they will know what to expect when they use the product. The most pronounced mode of action of azadirachtm is as an IGR. IGRs effect the hormonal system of insects, preventmg them from developmg normally mto mature insects. However, thts IGR property will not cause the immediate death of the insect pests Azadnachtin is structurally similar to the natural insect hormone ecdysone. Ecdysone regulates the development of insects, and any disruptton m its balance will cause improper development, Azadnachtm interferes with the production and reception of this msect hormone durmg an insect’s growth and molting. Thus, m thts manner, azadirachtm blocks the molting cycle, causing the msect to die (20). Because of its IGR effect, azadirachtin does not immediately kill msects and does not ktll adult insects. Immature insects die durmg their development, thereby reducmg the overall populatton over a period of time. The length of time depends mostly on the species of insect, age of insect, and the size of the population. Mortality can be seen m as little as l-2 d, to as long as a few weeks. Azadirachtm has its greatest effect on the early mstars. However, azadnachtin has effects on the emergence of pupae of some insects. It has been observed that the pupae of the leafmmer Liromyria trzfolia (21) and the frutt fly Ceratztls capztata (22), treated with azadnachtm, die before they emerge as

Commercial Experience with Neem Products

167

Table 1 Control of Beetarmy Worm (BAW), Cabbage Looper (LOOPER), and Diamond Back Moth (DBM) on Broccoflower in Oxnard, CA Treatment

Rate (per acre)

BAW

Average number Loopers

Untreated Xentari Neernlx Asana

1 lb 0.5 gal 9 6 oz

0 67 A 0.47 AB 0.83 A 023 B

09A 0.73 A 0.90 A 03B

DBM 0.47 0 23 0 37 0 07

A B A c

Plant damage P-3 45A 31A 1.2 B 08B

Three treatments made weekly at 100 gal/acre and evaluated after the third treatment Treatments followed by the same letter do not slgmficantly differ (P = 0 05 Duncan’sMRT)

adults. This variability in the expression of the activity of azadrrachtm can confuse farmers. Many IGRs have the drawback that they do not immediately kill the pest insect, thus leaving the insect to further damage the plant until it succumbs to the IGR. However, m the case of azadirachtm, the additional modes of actlon help protect the plants from damage while the IGR works on the insect Many insects exposed to azadlrachtm will stop feeding shortly after exposure. This, m effect, ends the damage to the plant, even though the insect larvae are still present This effect 1sexperienced m the field, where insect counts may not be significantly reduced, but plant damage is not occurring. In field trials conducted in California, the author has noted that broccoflower treated with Neemlx had amounts of worms present equivalent to the untreated control. But the damage caused to the plant was significantly reduced in the Neemlx-treated plots, and similar to that observed in plots treated with conventional msecticldes (Table 1). This mode of action complicates scouting, because insect counts are not totally representative of potential damage. 6. Formulation Effects Currently, m the United States, at least four different formulations contaming azadirachtin are registered for commercial use. They vary in active ingredient content, manufacturing process, as well as formulation components. ThermoTrilogy produces three azadirachtin forrnulatlons, including BloNeemT”, containing 0.09% azadirachtm m an alcohol base and supplied to the homeowner market. NeemixBTM and NeemazadBTM,contammg 0.25% azadlrachtm m an alcoholic base with 5% neem 011,are supplied to the agricultural and greenhouse markets, respectively. Neemlx and Neemazad were previously called Margosan-0. Neemix 4.5 and Neemazad 4.5, contaming 4.5% azadn-achtm m an acetate base, are marketed m the agricultural and greenhouse markets, respectively. Biosys produces Azatm and Align, which contam 3% azadlrachtm

Walter

162 Table 2 Final Population Density of A. pisum Exposed to Broad Beans Treated with Several Neem Insecticides at Equivalent Rate of 100 mg of Azadirachtin/L No. aphids Control

Azatm

Neemix

RH-9999

1392.25 A

654 B

232 C

1378 5 A

Table 3 Toxicity of Neem Insecticides to Immature A. pisum Exposed as First lnstars to Neem Insecticides at 100 mg of AzadirachtML Control

Azatm

Neemlx

RH-9999

2oc

68 0 B

90.0 A

80C

Means followed by same letter are not stgnlficantly different Based on five rephcates m a naphthalene base. Azatm is supplied to the greenhouse trade, and Align 1s supplied to the agricultural trade. Despite the fact that all these products contain azadlrachtin, which has little mammalian toxicity, the different mert Ingredients create products that have different levels of toxicity and may not perform similarly. Wan et al. (23) has recently shown that the naphthalene carrier m Azatm makes It 10 times more toxic to Juvenile salmon than is Neemix (Margosan-0). Formulation differences also impact the ability of these materials to control insects. Stark and Walter (24) demonstrated that three different azadlrachtm contaming formulations, when applied at the same rate of azadlrachtm, showed very different ablhtles to control the pea aphid, Acyrthoszphonpzssum (Harris). In these tests, they examined the effects of Neemlx (Margosan-0), Azatm, and an experimental formulation RH-9999 (Rohm and Hass, Phlladelphla, PA) on the pea aphid on broad beans. Trials conducted on mixed-age populations, as well as first mstar nymphs, showed that, when applied at equal rates of azadlrachtm, Neemlx was statistically more effective than the other materials, and RH-9999 had little effect (Tables 2 and 3). Slmllar results were reported by Eckberg et al (25), who noted that Neemlx (Margosan-0), used at a low rate, killed forest tent caterpillars, Mulacosoma disstria, more quickly than Azatzn at a much higher rate (Table 4). Further analysis conducted by Stark and Walter (26) suggests that the presence of llmmlods other than azadlrachtin, present m Neemlx, including

Commercial

Experience

with Neem

Products

163

Table 4 Effect of Azatin EC and Neemix (Margosan-0) on Mortality of Forest Tent Caterpillars % Mortahtv Treatment Untreated Azatin Neemlx

Rate 50 PPM 125PPM

7 DAT

11 DAT

15 DAT

19 DAT

24 DAT

OOA OOA 160A

OOB 10.5 B 61 3A

40B 30.0 B 68.0 A

80B 70 0 A 82.0 A

380B 86 0 A 90 0 A

Fourth mstar larvae were placed m a Petri dish with ash leaves treated with the products for 6 d Then the treated leaves were replaced with untreated leaves Numbers within a column followed by the same letter are not slgmficantly different (P = 0.05) by SNK

Table 5 Effect of Removing Neem Oil Components from Neemix and Adding to Aratin and RH-9999 No aphids Control 1437.75 A

Azatm

Azatm + 011

RH-999

RH-9999 + 011

625.25 C

281.75 D

1416.0 A

810.0 B

Means followed by same letter are not slgmficantly different Population density of,4 p~sumexposed to several neem msectlcldes at 100 mg of azadlrachtmil

mmbandlol, deacetylsalanmn, deacetylnimbin, mmbm, 6-acetylnimbandtol, and salanmn, and the or1component, are responsible for the enhanced acttvity of Neemtx. These hmmotds have little or no msecttcldal of then own at the levels present, but appear to strmulate the activity of azadnachtin Removal of these components from Neemrx reduces its activity, but then addition to the other azadnachtm formulattons Increased the activity of azatm and RH-9999 (Table 5). This 1sfurther amplified by the fact that recommended apphcation rates for azatm (S-21 oz/lOO gal, equivalent to 8-18 g azadtrachtm/lOO gal) are roughly 3 ttmes higher than those recommended for Neemtx (2.5-5 pt/ 100 gal equivalent to 2.8-5.6 g azadirachtin/lOO gal). Thus, azadtrachtm-based msecttctdes cannot be compared solely on this azadirachtin content. Other factors Influence the acttvlty of azadirachtin. 7. Adjuvant

Effects

An adjuvant is used to aid the operatton or improve the effectiveness of a pesticide. The term mcludes such matertals as wetting agents, spreaders, emulsifiers, dispersing agents, and penetrants. These materials are commonly used

Walter

164 Table 6 Effect of Adjuvants on Control of Greenhouse Whitefly on Tomato Treatment Untreated Neemlx Neemlx + Bond Neemlx + Plyac Neemlx + Intact Neemlx + Kmetlc Neemlx + Sllwet Neemlx + Soydex Neemlx + Dynamic Neemix + Joint venture

by Neemix

Adluvant rate

% Mortahty 19 35 A

002% 002% 025% 0.02% 0.03% 025% 025%

05%

3288 B 4078 B 37 1B 5007c 5321C 7452 CD

92.11 D 83 02 D

91 27D

Mortahty of first mstarnymphsexposedto Neemtxat a rateof 0 5 gal/100gals (5 ppmazadlrachtm) Meansm the samecolumnfollowedby the sameletterare statlstlcallyequal, P=OO5

to enhance the performance of conventional pestlcldes. Although azadirachtm works differently from conventional pesticides, because of its broad spectrum of activity, many applications of azadirachtin can benefit from the addition of adJuvants. As with most pesticides, getting the material to the pest 1s of utmost importance Although azadirachtm has some systemic propertles and will translocate across the leaf cuticle (26), it does not effectively move from leaf to leaf The use of spreading agents has been shown to improve coverage and enhance the activity of azadlrachtin m the laboratory and m the field. Laboratory trials conducted on greenhouse whitefly (Trraleurodes vaporariorum) indicate that not all adjuvants mfluence azadlrachtin similarly. In these trials, tomato plants were introduced into a greenhouse whitefly colony and the adults were allowed to lay eggs. The plants were then removed and sprayed with solutions of Neemlx (0.5 gal/l00 gal) containing various adJuvants. Seven d later, the mortality of the immature nymphs was evaluated. The results presented m Table 6 indicate that most of the adJuvants Improved the performance of Neemlx, but the adJuvants contammg vegetable or mineral 011 (Joint Venture and Soydex) gave the most significant improvement. Field tnals conducted m Florida show that the addition of Joint Venture to Neemlx 4.5 improved whitefly control (Table 7). Other effects have been noted with other adjuvants. For instance, the addltlon of Cell-u-wet dramatically improved the efficacy of Neemix to control pepper weevil damage, but did not Improve the control of armyworm m field

Commercial

Experience

with Neem

165

Products

Table 7 Control of Silver Leaf White Fly on Tomatoes with Neemix and Joint Venture No dead nymphs/leaf Treatment Neemlx Neemlx + Joint venture

Thiodan + Ambush + Soydex

Rate/acre

ODAT

4DAT

0.5 gal

0 05 0 05 0.12

0.14 0.37 0 10

0 5 gal + 0.25 gal

1 qt + 8 02 +1 qt

lIDAT 0 02 0 50 0 06

18DAT 0 25 1 10 0 32

Treatments applied on d 0,8, and 15 at 100 gal/acre. Evaluations made on d 0,4, 11, and 18.

Table 8 Effect of Additives on Control of Rosy Apple and Green Peach Aphids at Two Locations (Sebastopol, CA and Graton, CA) Treatment

Untreated M-Pede Stylet 011 Neemix Neemlx + Stylet oil Neemlx + M-Pede

Sebastopol

Graton

Rate/acre

Damaged shoots

Infested shoots

2 gal 2 gal 0 5 gal 0.5 gal + 2 gal 0.5 gal + 2 gal

1005 7 65 8 65 65 3 30

6.25 A 3.17 AB 2.75 B 358AB 0.5 B 3 17AB

Apphcatlons apphed 3 times, 10-14 d apart, at 100 gal/acre The Sebastopol trial was evalu-

ated for damaged shoots and the Graton trial for infested shoots trials m Florida.

This result 1s sigmficant,

because Neemrx

does not ktll the

adult pepper weevil, but in some way deters their ovipositton on the fruit. The addmon of spreading agents that slow the rate of drying, such as Cell-u-wet, enhances this activity. In field trials in California, the addition of materials like Stylet 011 and M-Pede

improved

the control

of aphids by Neemlx.

In these

trials, apple trees infested wtth aphids were sprayed 3 times over a 1mo period, and then evaluated for aphid damage. Four replications were included per treatment. In one trial, the addition of Stylet oil at 2% dramatically improved the performance of Neemtx, and, m another case, M-Pede provided a stmilar response (Table 8). Both M-pede and Stylet 011have some insecticidal activtty on their own, but the combinatton with Neemix IS much more effective. All of these trials indicate that, for some pests, the addmon of adjuvants that increase coverage and delay drying of the application, enhance the activtty of Neemix.

Mixtures of azadn-achtmwith other pesticides often show an apparent synergy.

Walter

166

Table 9 Neemix Has Been Found to Have Little or No impact on the Following Beneficial Organisms When Used According to Label Directions General beneficial organisms

General insect predators and parasites Delphastus pusdlus Scymnus sp

Honeybee worker adults (Apes melllfera) Spiders (Lycosa pseudoannolata, Chwoanthlum

mzldzz)

Nematodes (Steznernema, Heterobdztus) Lady beetles (Cocclnellldae, Hlppodamla

Phytosendus persrmllus Lyslphlebus tesacelpes

convergens)

Earthworms (Elsema foetlda) Nonorlbated mites Sprmgtalls (Collembola) Ground beetles (Carabtdae) Rove beetles (Staphylrnldae)

Aphelznus asychzs Eretmocerus callfornlus Encarsla formosa Pysttalla lnclsls Encarsla translena

Clearly, the influence of extractton processes,formulation solvents, and even formulation adjuvants can make the comparison of neem-based msecttctdes difficult, tf not imposstble. Although researchers and manufacturers have tried to use azadtrachtm as a marker mdtcattve of a product pesticide acttvtty, practical experience indicates that this has been unsuccessful. Farmers and other applicators must rely on the manufacturer’s recommendatton and data developed by local extension agents when selecting application rates and adjuvants for controllmg specific pests. The lack of a true standard of acttvtty for neem extracts is a clear difficulty m the commerctahzatton of neem Substandard or degraded products have too often left farmers with the impression that neem products are not effective. However, with high-quality products and the proper use of adjuvants, these impressions can be overcome In the United States market, because of the presence of only two neem suppliers, confuston is somewhat limited But m the world market, the differences m neem pesticides are not well understood and are a major obstacle to the acceptance of neem msecttcides 9. Use of Neemix

with Beneficial

Insects

Neem pesticides, and, m particular, Neemtx, are an excellent choice of materials to use when the preservation of beneficial msects 1s important In both laboratory and field tests,Neemtx was found to have little or no effect on benefictals, such as spiders, lady beetles, parasitic wasps, and predatory mites (27,28). Table 9 presents a list of beneficial

insects compatible

with Neemlx.

Azadnachtin has also been demonstrated not to reduce honeybee pollmatton m the field, nor to effect worker bees through direct contact sprays In fact, recent

Commercial Experience with Neem Products

167

work suggests that the consumption of Neemix can actually be beneficial to bees through the control of Chaulk brood and Varoa mites (29). In laboratory testsperformed on aphid parasites,it was found that, even after four applications of azadirachtin, the wasps inside the parasitized aphids exposed to the azadirachtm emerged m equal numbers to the untreated aphids (30). In other studies, Neemix did not significantly affect lady beetles or parasitism by P jlavzpes. The compatibility of neem with beneficial insects IS especially important because of today’s expanding IPM programs. Neemtx’s low impact on beneficials allows it to be used m conjunction with these predators and parasitoids with little worry about the effect rt ~111have on them. This strategy of combmmg Neemix with beneficial insects has been used successfully m the field to control insect pests. Pepper growers in Florrda have learned thts by using Neemix early in the seasonto control armyworms and aphids and protect beneficial insects. The loss of beneficial msectscaused by the use of mcompatible msecticides causes secondary pests to multiply unchecked, and typically result in late seasoncontrol problems. The use of Neemrx preserves the natural balance of beneficial insects, and thus ehmmates the need for additional pesticrde apphcattons, whrch ultrmately saves the grower money (31). It has been speculated that the msensmvity of beneficial insects to Neem extracts, m general, 1sbecause neem products must be ingested to be effective Thus, insects that feed on plant tissues will be effected by the extracts, but those that feed on nectar or other insectsrarely contact lethal concentrates (32). This is clearly not the case, because bees can be fed concentrations of Neemix m a sugar solution that would kill whiteflies or other insect pests (29). Beneficial insects m some way have a defense mechanism agamst azadnachtm, probably because of a different evolutionary history of carnivores and herbivores. 10. Summary Researchers have for years extolled the potential of using azadirachtm-based pestictdes on ornamental and food crops. After several years of regulatory review, azadirachtm was approved for use on food crops m the United States, m 1993, and was mtroduced for sale in Florida, and later California and Texas, for use on vegetables under the trade name of Neemix. In 1995, sales expanded to the Northwest and Northeast, and a new, more concentrated formulation, Neemix 4.5 (4.5% azadnachtm), was introduced. Because of its inherent safety, Neemix has been granted an exemption from tolerance for use at ~20 g a.1.per acre, and has the shortest re-entry time allowed by the EPA. Similarly, azadirachtin-based msecticides have been registered in several other countries, mcludmg Mexico, Saudi Arabia, Taiwan, Chile, and India. These products would never have been developed without the contributions of several mtemational researchers, mcludmg H. Schmutterer, W. Kraus, M. Jacobson, H. Rembold, R. C. Saxena, E. D. M Saxena, and K. R. S. Asher.

168

Walter

As a pesticide, azadnachtm kills Insects slowly and does not krll adult insects. Thus, neem-based msectrctdesare often used m combmatron with adulticrdes or beneficial insects, which expand the efficacy of the product. Education of farmers, and the development of practical strategies are essential to the effective use of azadirachtm-based pesttctdes.Manufacturers and extension agents must work m cooperation for azadirachtm-based pesttcrdesto succeed. Commercralization of azadirachtm 1splagued by the problem that the brological actrvrty of neem-based pesticides cannot be Judged solely by their azadnachtin content. Differences in extraction process, formulatton solvents, and adjuvants can dramatrcally influence the toxrcrty and pestrcidal activity of neem-based pestrcrdes. Detailed field experience with specific azadirachtmbased msecticides IS critical for both manufacturers and extension agents, to effectively recommend applicatron rates sufficient for adequate pest control. Despite these limitations, azadirachtin based msectrcides have established niche uses for controlling pests on peppers, melons, lettuce, tomatoes, pears, cm-us, celery, and other crops. The future demands for safe, environmentally sound pesticides will undoubtedly offer additional uses for azadirachtin. References 1 Mau, R F. L , Gusukuma-Minuto, L R , and Shimabuku, R. S (1994) Laboratory evaluations of biomsecticides against DBM larvae Arthropod Manage Tests 20, 327 2 Schmutterer, H and Smgh, R. P (1995) Uses of Neem, m The Neem Tree (Schmutterer, H , ed ), Verlagsgesellschaft, Wrenham, GDR 3 Schmutterer, H (1995) List of Insect pests susceptible to neem products, in The Neem Tree (Schmutterer, H , ed ), Verlagsgesellschaft, Wrenham, GDR, pp l-29 4 Benge, M D (1989) The tree and its characteristics, cultivation and propagation of the neem tree, m Focus on Phyotochemrcal Pestlcldes (Jacobson, M , ed.), CRC, Boca Raton, FL, pp. l-l 6 5 Schmutterer, H (1995) Introductory remarks, m The Neem Tree (Schmutterer, H , ed.), Verlagsgesellschaft, Wienheim, GDR, pp. IX-XII 6 Larson, R 0. (1989) Commercialization of neem, m Focus on Phytochemlcal Pestzcrdes (Jacobson, M , ed.), CRC, Boca Raton, FL, pp. 155-160 7 Isman, M. B , Koul, 0, Lowery, J. J., Arnason, D , Gagnon, J G., Stewart, J , and Salbum, G S (1990) Development of a neem-based msecticrde m Canada m neem’s potential m pest management programs, Proceedmgs of the USDA Neem Workshop (Locke, J C , ed.), USA ARS, pp. 32-39 8 Schafer, E W and Jacobson, M (1993) Repellency and toxrcity of 55 Insect repellents to red winged blackbirds (Angelausphoenzceus) J Envwon SCI Health l&493-497 9 NCR (1992) Neem, m A Tree for Solving Global Problems (Ruskm, F R , ed ), National Academy, Washmgton, DC, pp l-l 37

Commercial

Experience

with Neem

Products

169

10. Chopra, R N., Badhwar, R L , and Ghosh, S. (1968) in Pozsonous Plants oflndza, vol 1. (Pravad, J , ed ), Indian Council of Agricultural Research, New Delhi, pp. 248-270.

11, Sinmah, D , Baskaran, B. G , Looi, L. M., and Leong, K. L (1983) Reye-like syndrome due to Margasa oil poisonmg: report of a case with post mortem lindings Am J. Gastroenterol. 77, 158-164. 12. Jacobson, M. (1989) Pharmacology and toxrcology of neem, in Focus on Phytochemtcal Pestrctdes (Jacobson, M , ed ), CRC, Boca Raton, FL, pp 133-l 83 13 Smniah, D., Baskaran, G., Looi, L. M., and Leong, K. L. (1983) Fungal flora of neem seeds and neem oils toxicity, Malaysia, Appl Btol 12, l-l 2 14 Walter, J. F. and Knauss, J. F. (1990) Developing a neem-based pest management product, m Proc. Neem ‘s Potenttal tn Pest Management Programs (Locke, J. C., ed.), USDA ARS, USA, pp 29-3 1. 15. Anon. (1993) “New York Times News Service,” June 6. 16. Anon (1996) “Company News on Call,” May 14. 17 Anon. (1996) “Busmess Wire,” September 30 18. Walter, J. F (1996) Proceedings International Neem Conference, Austraha (Smgh, R P , ed.), m press. 19. Walter, J. F (1996) International Conference on Standardization of Neem Pesticides, Stuttgart, Germany 20. Mordue (Luntz), A. J. and Blackwell, A. (1993) Review of the activity of azadnachtm. J. Insect Phystol 39,903-924. 21. Smith, R. and Chaney, W. (1995) Update on leafmmer pest control U C Crop Reporter February 3, l-3. 22 Stark, J. D , Vargas, R. J., and Wong, T. Y (1990) Effects of Neem Extracts on Trephiretha Fr. #Fhes and then Parasitoids m Hawaii, Nemo Potential in Pest Management Program USDA ARA, pp 36-42 (Locke, J C , ed ) 23. Wan, M. T., Watts, R G., Isman, M. B , and Strub, R. (1996) Evaluation of the acute toxicity to juvenile Pacific Northwest salmon of azadnchtm, neem extract, and neem-based products. Bull Environ Contam Toxtcol 56,432-439 24. Stark, J D. and Walter, J. F. (1995) Neem oil and neem 011 components affect the efficacy of commercial neem insecticides J Agrtcult Food Chem. 43,507-512. 25. Eckberg, T. B., Cranshaw, W., and Sclar, D. C. (1994) Evaluation of neem msec-

ticides and persistence for control of forest tent caterpillar, Ft Collins, CO Arthropod Manage Tests: 1995,20, 327. 26. Larew, H. G., Knodel, J. J., and Marion, D. F. (1987) Use of fohar-applied neem (azadirachta zndica A. Juss) seed extract for the control of the birch leafmmer, Fenusa pustlla (Lepeletier). J Envtron Hort 5, 17-19. 27 Stark, J. D., Vargas, R. I., and Wong, T. Y. (1990) Proc. Neem ‘s Potenttal tn Pest Management Programs (Locke, J C , ed ), USDA ARS, USA, pp 106-l 13. 28 Hoelmer, K. A., Osborne, L. S , and Yokomu, R. K (1990) Effects of neem extracts on beneficial insects m greenhouse culture, Proc Neem s Potenttal zn Pest Management Programs (Locke, J. C., ed ), USDA ARS, USA, pp 100-106

170

Walter

29 Lru, T P (1995) Possible control of chalkbrood and noseme drsease of the honey bee wrth neem. Am. Bee J 134, 195-198 30 Schauer, M (1985) Die Wrrkung von Ntemmhaltsstoffen auf Blattlause und dte Rubenblattwanze. Doctor thesis, Umversrty of Gressen, Germany 3 1 Walter, J F (1996) Use of botanical pestrcides for the control of pepper pests, Internatronai Pepper Conference Proceedmgs (Maynard, D , ed ), pp 17-19 32 Wrlhams, L. A D and Manstgh, A (1996) Insecttctdal and acamedal actron of compounds from Azahrachta zndzca (A Jus) and thetr use m troprcal pest management Integrated Pest Manage Rev 27, 133-145

11 Fermentation-Derived

Insect Control Agents

The Spinosyns Thomas C. Sparks, Gary D. Thompson, Herbert A. Kirst, Mark B. Hertlein, Jon S. Mynderse, Jan R. Turner, and Thomas V. Worden 1. Introduction Many insect pests present an ongoing battle between the grower’s ability to control the pest and the pest’s ability to resist the available control methods Several well-known examples include the Colorado potato beetle on potatoes, the diamondback moth on vegetable crops, and the Helzothzs complex on cotton. The discovery of new, novel insect control agents for use against these and other Insect pests has served as a focal pomt for insecticide research for more than four decades There are a number of approaches that can be taken m the discovery of Insect control agents, and these have been discussed from a variety of vlewpoints (Z-4). A key component m all of these various vlslons for the discovery of new potential Insect control agents IS natural products A variety of natural products have been or are used as insect control agents (5), mcludmg pyrethrum, abamectm, milbemycin, azadirachtm, mcotme, rotenone, and ryania (6-10). LikewIse, natural products have served as leads for a variety of insect control agents, mcludmg the pyrethrolds, Juvenoids, chlorfenapyr, and, arguably, the phenyl carbamates (9,11-23). Thus, it IS reasonable to expect that natural sources will continue to provide other new insect control agents. It is the purpose of this chapter to outline the discovery, chemistry, and biology of a new class of novel, fermentation-derived Insect control agents the spmosyns. From Methods in Biotechnology, vol 5 Bopeshodes Use and Delwery Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

171

172

Sparks et al.

2. Discovery of the Spinosyns A key aspect of searching natural sources for new products 1s to mnnmaze the ltkelthood of redtscovermg known, unmterestmg compounds. This IS critical because of the time and difficulty of lsolatmg and identtfymg the compound or compounds that are associated with a given blological activity. Just as deconvolutron can be a key pinch point m the utilization of combinatortal chemistry, the isolation and tdenttficatlon of components m a naturally occurring source represents a deconvolutlon of nature’s combinatortal chemistry. Thus, it is prudent to take measures that increase the probablllty of the isolated component being a new, novel compound when finally identified. Two aspects that directly influence fmdmg new, novel compounds are screening new or unusual sources of natural products, such as marine algae or invertebrates, uncommon fungal or bacterial sources, and so on, and employmg a screening tool that imparts some measure of selectivity or novelty. By using one, or, better yet, both aspects, the chances that the isolated compounds will be new and interesting are greatly improved. During the 1980s Lilly Research Laboratories (LRL, Indtanapolts, IN) operated a program directed at finding new natural products that possessed utility m the pharmaceutical and agricultural mdustrtes. Soil samples from all over the world were collected, fermented, extracted, and screened m a variety of assaysystems.Where possible, the sot1samples were collected from unusual habitats, to improve the chances of findmg new mlcroorgamsms. Among the assaysemployed m the LRL screening program was a mosquito larvlclde assay that was used to detect msecticrdal activity (24). A multtspecles, on-plant assay was used as a follow-up to any actrves detected m the mosquito larvtcide assay. This multispectes assay provided a measure of spectrum for any mosquitoactive broth extracts. During the course of this fermentation screening program, extracts from the fermentation broth of a sol1 sample (designated A83543) collected m 1982 on a Caribbean island were found to be active on mosqurto larvae (14). More importantly, these extracts were active on southern armyworm (Spodoptera endanza) when tested m the multrspecles follow-up assay. Subsequent testing suggested that the insect activity was caused by a low-abundance, htgh-acttvtty substance that appeared to have activity at the level of some commerctal msectrcrdes. The mtcroorgamsm Involved was identified as an actmomycete belonging to a less common genus, Saccharopolyspora. The mtcroorgamsm, Saccharopolyspora spmosa, was identified as a new species (15), and rt produced a family of new, unique macrohdes (molecules contammg a macrocyclic lactone) (14), ortgmally referred to as A83543 factors, but now called spmosyns (16)

Spinosyns

173 2’,3’,4’-tw0-IMethyl Rhamnose R2 ’

Tetracycllc Rmg

Fig

1.

Generalstructureof the spmosyns

3. Chemistry of Spinosyns A variety of techniques were used to establish the structure and stereochemistry of spmosyn A, mcludmg mass spectrometry, extensive NMR spectroscopy, X-ray crystallographic analysis, and hydrolysis of the forosamine sugar, to establish absolute configuration (1417). The spmosyns are composed of a 12-membered macrocyclic rmg as part of an unusual tetracychc ring system, to which two different sugars are attached (Fig. 1); an ammo sugar, forosamme, and a neutral sugar, 2’,3’,4’-tri-O-methylrhamnose. These attributes set the spmosyns apart from other macrocyclic compounds, such as erythromycin A (14-membered monocyclic macrocyclic ring), tylosm, and spiramycin (all 16-membered monocyclic macrocyclic rmgs), the avermectm-mllbemycm family, and lkarugamycin (a tetracyclic macrolactam lacking any sugars) (I4,16, Fig. 1). Spmosyn A (A83543A) was the first spmosyn isolated and identified from the fermentation broth of S. spznosa. Subsequent exammation revealed that the original parent strain of 5’. spznosa (wild-type, WT) produced a number of spinosyns (A-J) (Table 1). Compared to spmosyn A, these other spmosyns (B-J) are characterized by differences m the substitution patterns on the amino group of the forosamme, at selected sites on the tetracychc ring system, and on 2’,3’,4’-tri-O-methylrhamnose (Fig. 2). The WT stram of S spinosa produces a mixture of spmosyns (Table l), of which the primary components are spmosyn A and spinosyn D. An extract of the fermentation broth that contains this naturally occurrmg mixture of spinosyns A and D IS called spmosad (Tracer@), the first product in Dow Agrosciences Naturalyte@ lme of insect control products. Spmosad received U. S. Environmental Protectlon Agency (EPA) registration for use in cotton insect control m February 1997. Spmosyn A is the most active of the naturally occurring spmosyns (listed In Table 1) agamst larvae of the tobacco budworm, Helzothis vwescens, followed closely by spmosyn D. Thus, the most insecticldally active spmosyns are also those that the mlcroorgamsm naturally produces m largest quantrty

Rl

Sources,

R21

R16

Activity

R6

R2’

of Spinosyns

Me Et Me H OMe Me Et Me H OMe H Et Me H OMe Me Et Me Me OMe Me Me Me H OMe Me Et H H OMe Me Et Me H OMe Me Et Me H OH Me Et Me H OMe nonfunctional 2’-O-methyltransferase Me Et Me H OH Me Et Me ine OH Me Et Me H OH Me Me Me H OH Me Et Me H OH nonfuncttonal 3’Gmethyltransferase Me Et Me H OMe Me Et Me Me OMe Me Et Me H OMe Me Et Me Me OMe

R2

and Biological

Spmosyns from wild-type A Me B H C H D Me E Me F Me G” Me H Me J Me Spmosyns from H mutant H Me Me Q R H S Me T Me Spmosyns from J mutant. J Me L Me M H N H

Table 1 Structures,

OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe

OMe OMe OMe OMe OH OH OH OH OH

R4’

~80 26 22 6 40

5.7 05 14 5 53 >64

0.3 04 08 08 4.6 45 71 57 >80

TBWa Neonate drench LC$

to Spinosyn

OMe OMe OMe OMe OMe OMe OMe OMe OH

R3’

Compared

>25 -25 >25

>25 31 -

1.1 60 18 31 66 18 >25 >25

TBW Cotton leaf-drp LCSClb

-63 67 >50

114 -

95 14

0.9 04 84 29 >50 16 95 -63

>lOO >lOO 61

12 >lOO

-

6.9 51 15

ALH Vial contact LC50b

Standards

TSSM Acute LC50b

A and Selected

“TBW, tobacco budworm, TSSM, two-spotted spider mite, ALH, aster leafhopper ‘wm ‘The ammo sugar IS ossamme Instead of forosamme dForosamrne is missmg ‘2’,3’,4’-tn-O-methylrhamnosyl moiety IS mlssmg

Spinosyns from K mutant: nonfuncttonal4’-0-methyltransferase K Me Me Et Me H OMe OMe 0 Me Me Et Me 1Me OMe OMe Y Me Me Me Me H OMe OMe Spinosyns from H and J mutants. nonfunctional 2’ or 3’-0-methyltransferase Me H OH OMe U Me Me Et Me Me OH OMe V Me Me Et Me H OMe OH P Me Me Et Me Me OMe OH W Me Me Et Spinosyns lacking one or more sugars Me H OMe OMe Psa-17 d d Et Me H e e Psa-9 Me Me Et d d Et Me H e Aglycone Psa-17 D d d Et Me Me OMe OLe Psa-9 D Me Me Et Me 1Me e e Aglycone D d d Et Me Me e e Cypermethrm Propargite Ethofenprox OMe e e OMe e e

>64 >64 >64 >64 >64 >64 061 -

-

100 -

19

13 -

03

82 28 -

x50 -

32 69 -

10 04

-

>25 OH 3.5 14 13 OH OH 20 in combmatton wtth smefungm OH 22 OH 17 OH >64 OH >64

Sparks et al

776

Spmosyn A

Avermectm

B 1a

Erythromycm

.4

Spxamycm I

Ikarugamycm

Fig 2 Structures of spmosyn A and other macrocychc compounds

In addition to the spmosyns produced by the WT stram, other spmosyns were identified from several mutant strains. Because the parent strain produced the spmosyns only m very minute quantities, LRL began a program of strain improvement to increase the yield. One offshoot of the strain selection program was the tdentifkation of several mutant strams, possessmgnonfunctional 2’- and/or 3’- and/or 4’-O-methyltransferases. Because these mutants were unable to methylate particular hydroxyl groups on the 2’,3’,4’-tri0methylrhamnose, a variety of spmosyns were produced, most of which were not pro-

Spinosyns

177

duced by the WT stram (18). The spinosyns isolated from these mutant strains include spmosyns H, Q, R, S, and T; spmosyns J, L, M, and N; and spinosyns K, 0, and Y from mutants with nonfunctronal2’-, 3’-, or 4’0methyltransferaseq respectrvely (18,19). Further varratrons m the methylatron of the rhamnose sugar were observed with the addition of smefungm (ZO), whrch was found to specifically block the 4’0methyltransferase during the fermentatron process of the WT strain. When coupled (sinefungin) with the H and J mutant strams, several other new spmosyns, P, U, V, and W, were produced (18). Again, the variations in all of these spmosyns center around methylation of the forosamme ammo group, presence or absence of O-methyl groups on the rhamnose sugar, and presence or absence of methyl group(s) in the C6, C 16, and C2 1 positions of the tetracychc rmg system. To date, more than 20 spmosyns have been identified from the WT and mutant strains (Table 1). Physical characterrsttcs of the spmosyns have been published elsewhere (14,22,22). Because of their chemically complex nature, the spmosyns are efficrently obtained only through the process of fermentation Although not a useful process for commerctal productron, the first total synthesis of spinosyn A as its unnatural levorotatory enantiomer was accomplished by Evans and Black (23,24; the term “spinosyns” 1snow the preferred name for this chemical class) The naturally occurring spmosyn A is dextrorotatory and 1sbrologrcally active against insects such as H. vlrescens, but the unnatural levorotatory enantromer IS brologically inactive against H vzrescenslarvae. 4. Insect Spectrum of Spinosyns Spinosyn A is active on a variety of insect species, but especrally on leprdopterous pests such as the tobacco budworm (H. vzrescens), the cotton bollworm (Helicoverpa zea), American bollworm (Helzothu armigera), armyworms (Spodoptera exigua, Spodoptera littoralis, Spodoptera frugzperda), loopers (Trzchoplusza nz), diamondback moth (Plutella xylostella), and rice stemborer (CItllo suppressah) (Tables 1 and 2; refs. 16,21,22,25). Good activity IS also observed against a variety of dipteran pests, thrips, fleas, and hymenopteran pests; activity IS variable against coleopterans (21). At a screenmg rate of 400 ppm, spinosyn A 1s active against leprdopteran pests, spider mites, planthoppers, and cockroaches, but no actrvrty is observed on aphids or nematodes (Table 2; ref. 22). Although the relatively broad activity spectrum of the spmosyns is interesting, the truly exciting aspect of this novel chemistry is the level of activity observed against leptdopteran pests such as H. virescens. Bioassays using a standard topical bioassay show spinosyn A to be far more active than a variety of organophosphorus, carbamate, cyclodiene, or other insect control agents commonly used m crops such as cotton (Table 3). In these and other assays,

178 Table 2 Screening

Sptnosyn A B C D E F H J K L M N 0 : R S

Sparks et al. Activity

TBW + + + + + + + + + + + + + + + -

of Spinosyns

BAW + + + + + + + + + + + + + + +

T

+ +

U

+

+

v

+

nt

W Y

+ +

+

Against

CPH + + + + + + + + + + + + + nt +

Selected

-

GECR -t -t + + + + -

+

-

-

nt nt

-

-

-

nt nt

-

-

-

nt

nt

+

-

nt nt

nt

nt -

nt

+

-

nt

nt

nt

TSSM + + + +

Pest Species

+ +

CA -

RKN nt nt

nt

+ = >80% mortality at 400 ppm, - = ~80% mortality at 400 ppm; nt = not tested TBW = tobacco budworm, H wrescenr, BAW = Beet armyworm, Spodopteta exlgua, CPH = Corn plant hopper, Peregrrnus mardls, TSSM = Two-spotted spider mne, T wtlcae, GECR = German cockroach, Blatella germamca, CA = Cotton aphid, Aphls gossyplz, RKN = Root knot nematode, Meloldogyne mcogmta

sptnosynA ISasactive asmany of the pyrethroid msecttctdes(Table 3). When tested n-rassaysthat are prnnartly contactm nature (such as toptcal, glassvial, and so on), spmosynA ISabout as active or slightly less active than pyrethroids, such as cypermethrin; however, tn assaysthat incorporate an oral component (leaf or diet assays), spmosyn A can be more active than cypermethrin (refs. 28,26, Table 1). Thus, depending on the assaysystem,sptnosyn A 1scomparable to, and m some cases supenor m acttvtty to, pyrethrotdssuchas cypermethrm(refs. 18,26; Tables 1 and 2). 4.7. Spinosyn Structure-Activity Relationships Very simple changes in structure can profoundly alter the acttvtty of the spmosyns toward larvae of H vzrescens.The presence or absence of N-methyl

179

Spinosyns Table 3 Acute Insect (H. wirescens, Topical, 48-72 h, yglg) and Mammalian (Rat Oral, Acute, mg/kg) Toxicity of Selected Insect Control Agents Compound Spinosyns Spmosyn A DDT and Pyrethrolds DDT Permethrm Fenvalerate Cypermethrm LCyhalothrm Blfenthrm Cyclodlenes Endosulfan Organophosphates Me Parathion Azmphosmethyl Acephate Profenofos Carbamates Methomyl Avermectms Abamectm Emamectm

H vwescens L&o

Rat oral L&o

VSRa

Refs

1 28-2 25

3783-5000

1681-3906

18

52-152 1 33-2 79 0 870-l -89 0 241-1.61 0 929 1 32

87 >4000 451 247 56 55

0 9-2 8 >1434-3008 239-l 139 153-1025 60 42

42,43,45 39,49,45 39,45 39,4.5 39,45,47 39,47,48

73 3

18

03

42,45

11&650 29 3-33 3 74 3 11.0

9 5 866 400

0148 02 11 7 36

39,45 4294549 45,so 39,45

4 33,26 7

17

0 6-3.9

39,45,49

1 16 0.10

10.6-l 1.3 70

9 l-97 700

41,44 46,30

‘VSR, vertebrate selectwty ratlo for Rat oral/Hv toxlclty Not necessarily representatwe for other insect pests

groups on the ammo group of forosamme (spinosynsB and C) or a methyl group at C6 (spmosyn D) does l&e to alter the biological activity relative to spmosyn A. However, lossof a methyl group at C2 1 or Cl 6 of the tetracyclic ring (spinosynsE and F, respectively) reduces activity (Table 1). Most dramatic, however, are the activity changesassociatedwith the loss of O-methyl groups on the rhamnose nng Loss of the methyl group m the 2’ or 4’ position (spinosynsH and K, respectively) reducesactivity by about an order of magnitude; loss of the methyl group at the 3’position (spmosyn J) decreasesactivity by more than two orders of magnitude (>2OOx),compared to spinosyn A (Table 1). Most of the other spinosynsrepresent combmatlonsof spinosynsH, J, or K associatedwith the altered methylation patterns found m spinosynsB, D, E, and F. Thus, the natural spmosynsexhibit simultaneous variations in N-, C-, and 0- methylations.These spmosynsare all much lessactive than spinosyn A, with one notable exception: spinosynQ (spmosynH with a methyl

180

Sparks et al.

group at C6) (Table 1). There is also a general trend m which the C6-methyl analogs of the 2’-, 3’-, or 4’0demethylrhamnose derivatives (spinosyns Q, L, 0, respectively) are more active than their respective parents (spinosyns H, J, K) (Table 1) Loss of either sugar (forosamme or the trt-O-methylrhamnose) to yield the C 17 and C9-pseudoaglycones, respectively, or both sugars (aglycone), results m a loss of activity (ref. 27, Table 1). A stmtlar trend 1sobserved for the two pseudoaglycones and the aglycone of spmosyn D (methyl at C6) Thus, none of the more than 20 naturally occurring spinosyns, their aglycones, or pseudoaglycones, listed m Table 1, 1smore active against neonate larvae of H v~~~cens than spmosyn A. Cotton leaf-dip activity of the spmosyns toward larvae of H vzrescens revealed a pattern that was very simtlar to that of the neonate drench assays In the cotton leaf-dip assay, spmosyns A and D clearly show then superior activity compared to the other spmosynsexamined (Table 1) In both of the H vzre,scey1s assays,spmosyn A was as active as the commercial pyrethroid cypermethrm, a trend observed in a variety of other assaysystems(18) Although they apparently lack the residual activity crtttcal to succeed as an acartctde, the spmosyns do exhibit acute toxtctty to two-spotted spider mites, Tetranychus urtzcae (acute LCsO= 0.9 ppm, Table 1) Structure-activtty relattonshtps for T urticae present a significant departure from that of H vzrescens, m that the best btologtcal activity against T. urtzcae is exhibited by spmosyns K and 0, but spmosyns H and Q are relatively weak (Table 1). Regression analysts of spmosyn activity toward neonate H vzrescenslarvae vs acute toxictty to T. urtzcae revealed only a relatively weak correlatton (r’ = 0.5 18,s = 0.655, F = 0.0037). Thus, there appears to be little relationship between spmosyn acttvtty against mites and activity against H. vzrescens Although the spmosyns possesssome acute acartcidal activity, residual activity toward mites is weak compared to commerctal standards, rendermg the spmosyns unsuttable as acaricides within the context of our current knowledge of this chemistry (22). In addition to their demonstrated activity on lepidopterans, mites, and dtpterans (21,28), the spmosyns are also active on some homopteran species, such as the aster leafhopper, Macrosteles severznz, which 1sused as a model for Asian plant- and leafhopper pest species Although the current data set 1smcomplete, an exammatton of the data m Table 1 shows that several spmosyns (A, B, K, 0, V) possessnearly equivalent contact activity against adults of M severznr, and spmosyns K and 0 are again among the most active of the series. Although the contact activity of the spmosyns toward M severznz 1sinteresting, they are much less active than extstmg commerctal products, such as ethofenprox (Table 1). Similar conclustons would also apply to other hopper pest species (22) 5. Mode of Action A number of modes of actton for insect control agents are known, mcludmg acetylcholmesterase mhibitors (organophosphorus and carbamate msecticides),

Spinosyns

181

sodmm-channel blockers (DDT, pyrethroids), channel blockers for y-ammobutyrtc acid (GABA)-gated chloride channels (cyclodlenes, fiproles), mcotmlcreceptor agonists (tmtdacloprrd), octopamine receptor agonists (formamtdmes), and inhibitors of mttochondrtal respiration (chlorfenapyr, rotenone) (29,30). Currently available mformatron based on extensive mode-of-action studies clearly indicates that the mode of action for spmosad is drstmct from all of the forementioned groups, or any other insect control agent whose mechanism IS known. Electrophysrological studies have shown that spmosyn A acts on the insect central nervous system to increase spontaneous acttvtty, leading to mvoluntary muscle contractrons and tremors (31). This increase in excttatton appears to result from the persistent activation of mcotmtc acetylcholme receptors and prolongatton of acetylcholine responses, in a manner that IS drstmct from other mcotmtc active molecules, such as tmidacloprrd and mcotme (31). In addmon, the spmosyns can also alter the function of GABA-gated chloride channels, again m a manner distinct from all known insect control agents (31), which may or may not also contribute to the btologtcal activity of this novel class of insect-control agents. Thus, based on our present understanding, the mode of action of spinosad appears to be unique. 6. Environmental and Toxicological Profile The pyrethrotd levels of msecttctdal activity observed for spmosyn A and spinosad contrast sharply with Its relative safety to many beneficial insects. Spmosad was an order of magnitude less toxic to honeybees and a whitefly parasitold (24 h LCsO = 11.5 and 29.1 ppm, respectively) than cypermethrin (24 h LC50 = 1.2 and 1.9 ppm, respectively) (25). For the hemipteran, coleopteran, and neuropteran benetictals studied (e.g., minute pirate bug, convergent lady beetles, common green lacewing), spinosad was nontoxtc at the highest dose tested (24 h LC5,, = >200 ppm), in contrast to cypermethrin, which was htghly toxic (24 h LCsO= ~0.8 ppm) (25,32,33). Thus, spinosad provides a high degree of selecttvity toward beneficial insects, making it an attractive tool for insect-pest management programs that seekto preserve beneficial populations as a means to improve overall pest-insect control, and reduce the risk of secondary pest outbreaks. A compound can possessexcellent levels of activity against target msectpest species and yet have limited utility, tf tt IS highly deleterious to beneficial insects, mammals, or other nontarget species. The fact that a compound IS a natural product does not necessarily guarantee good envuonmental and/or pest management compatibility. However, m the case of spinosyn A and spmosad, excellent levels of activity against target-pest species are indeed coupled with an excellent envn-onmental/toxrcologrcal profile. In testmg conducted for EPA registration, spinosad was not shown to leach or persist m the environment. Toxicity to mammalian, avran, and aquatic species is relatively low when com-

Sparks et al.

182

pared to many currently used insect control agents (Table 4) One approach to quanttfymg the relative selectivity of a compound for target (insect) vs nontarget (i.e., mammalian) species is to calculate a therapeutic Index. Although such mdices certamly possess hmitations (34), such as dependence on only one pest/nontarget species and one apphcation method, a therapeutic index, such asvertebrate selectivity ratio (VSR = acute rat oral LD,, mg/kg/msect LD50 pg/g; 34), can illustrate the relative selectivity of an insect control agent for the target-insect species vs nontarget mammals in a given cropping system. Certainly, many of the older msecticides, such as methyl parathion and EPN have comparatively high toxicity to mammals, and only comparatively moderate activity on the target-insect pest, resultmg m low VSRs (VSR =
and Resistance

Management

A consideration m the development of any new msecticidal chemistry IS the potential for pest resistance development. In many crop systems, such as cotton and vegetables, msecticide resistance has been an area of concern for more than three decades. The development of pyrethroid resistance m H vzrescens m many parts of the United Statescotton-growing region is only the most recent instance m a long hst of such msect control agents to be affected (35,36), and the problem appears to be expanding to include other currently available msectcontrol agents (36,37) Thus, resistance to any insect control agent is a real possibihty. In the case of spmosad, presently available mformation suggests that, currently, no cross-resistance IS observed m larvae of H vzreScenSresistant to any other insect control agents, nor m field or resistant strains of other lepidopterans (18,381. When spmosad was evaluated on a number of field strains of H. vzre,scenSfrom the southern United States, the diagnostic dose of spmosad used provided good control for all of the strains examined, and clearly provided much better control than other insect control agents, such as cypermethrm or profenofos (28). Other studies have shown little or no cross-resistance in laboratory strains previously selectedfor high levels of resistanceto msecticides

3783-5000 >2000 Not irritant >2000 >2000 30 96 96 33,51

Compared

Rat oral (mg/kg) Rat dermal (mg/kg) Rabbit skm irritation Mallard duck, acute oral (mg/kg) Quarl, acute oral (mg/kg) Rambow trout, acute 96 h (mg/L) Carp, acute 96 h (mg/L) Daphma magna, 48 h (mg/L) Refs

A (Technical) Spinosyn A

of Spinosyn

Test

Table 4 Toxicity Profile

106-113 84 2000 0.0032 0 042 0 00034 44,52,53

Abamectm

with Selected

247 >2000 Moderate irritant >10,000 0.025 0 0016 0 0013 45,48,54,55

Cypermethrin

LW .

Insect Control

(Technical)

97 >2000 Not rrritant >2150 11.3 0.25 0.43 0 19 56

Fiprornl

Agents

450 HO00 31 >32 57.58

Imidacloprid

184

Sparks et al.

(18), or known to possessspectfic target site- or metabolism-based mechamsms.

However, some variation m susceptibihty is expected for any new insect-control agent, as demonstrated by the range of LDso values observed for several pyrethroids 2 yr prior to their commercial mtroduction (39). Likewise, some variation has also been observed for spmosad, with some field strains having higher and some lower LDsOs,compared to a laboratory reference colony (38). The unique mode of action of spinosadand the spmosynscertainly renders crossresistancecausedby altered target site (I.e., knockdown resistance,altered acetylcholmesterase)asvery unlikely. Likewise, laboratory colomes of the diamondback moth possessmg enhanced detoxification systems (glutathione transferases, monooxygenases; Chih-Nmg Sun, personal communication) showed little crossreststanceto spmosyn A (IS). These observations are consistent with recent studies indicating that pest lepldopterans, such as H vzrescens, do not readily metabolize spinosyn A (40). Thus, the potential for the rapid development of pestInsect resistanceto spinosadmay be reasonably low, making spinosadan attracttve and a potentially useful tool in insecticide-resistancemanagement programs However, as the history of insecticide resistancehas aptly demonstrated, resistanceto any insectcontrol agent can occur If sufficient selection pressure1sapplied through overuse or misuse. In an effort to ensure the long-term utilrty of spmosad, DowAgrosciences is mvesttgatmg and, where appropriate, providing use recommendations designed to munmize the chanceof resistancedevelopment.

8. Summary The discovery and subsequent development and registration of spmosad (Tracer@), demonstrates that natural products contmue to provide a fertile source of new, novel insect control agents. Although more than 20 spmosyns have been isolated and identified, thus far spinosyns A and D (the primary components of spmosad) remain the most active against lepidopterous larvae such as H. virescens. The different spmosyns arise from varrations in substitution patterns on the two sugars (forosamme and 2’,3’,4’-trr-O-methylrhamnose) and the tetracychc ring system. Many of the spinosyns were isolated from mutant strains lacking a specific O-methyltransferase for one of the hydroxyl group positions on rhamnose. Among the spmosyns, small changes m their structure can result in large changes m biological activity, especially modrfications to the rhamnose moiety and at C 16and C2 1of the tetracychc ring. Although spinosyn A IS the most active of the spmosyns toward lepldopterous larvae, spinosyns K and 0 are among the most active for T urtzcae and M sevennz. However, at this time, thesespmosyns lack the necessaryacttvny and/or residual properties to be considered as potenttal products for mite and leafhopper pests Available mformation indicates that the mode of action of spmosyn A is unique. Compared to other insect control agents, the spmosyns also possess

Spinosyns

185

very favorable environmental and toxtctty profiles, and are also comparattvely safe to beneficials Vertebrate selectivity ratios (a type of therapeuttc index contrasting vertebrate and insect toxtcrty levels) for spinosyn A on H vzrescens larvae are among the most favorably observed to date for insect control agents. Thus, these novel compounds represent a new genre of unique, naturally derived insect control agents that possess pyrethrotd levels of acttvity, an excellent toxtcologtcal and environmental profile, and a lack of cross-resistance to the currently available insect control agents (based on available data) (28).

Acknowledgments We gratefully acknowledge the assistance of our many colleagues at Dow Agrosctences (DAS) and Lilly Research Laboratories (LRL), includmg Larry L. Larson (DAS), James Gtfford (DAS), Joe Schoonover (DAS), John Babcock (DAS), James Dripps (DAS), John R. Skomp (DAS), Vmce Salgado (DAS), Larry Creemer (LRL), Patrick J. Baker (LRL), M. Chris Broughton (LRL), Mary L. Huber (LRL), James W. Martin (LRL), Walter M. Nakatsukasa (LRL), Karl Michel (LRL), Raymond Yao (LRL), and Jonathan W. Paschal (LRL).

References 1 Menn, J J. (1983) Present msecticides and approaches to discovery of environmentally acceptable chemicals for pest management, in Natural Products&r Innovatwe Pest Management (Whitehead, D. L , and Bowers, W. S , eds ), Pergamon, New York, pp 53 1 2 Geissbuhler, H., d’Hondt, C , Kunz, E., Nyfeler, R , and Ptister, K. (1987) Reflections on the future of chemrcal plant protection research, m Pestlclde Sczenceand Blotechnology (Greenhalgh, R. and Roberts, T R., eds ), Blackwell, Boston, pp 3-14 Hodgson, E and Kuhr, R J (1990) Introduction, m Safer Insecticzdes* Development and Use (Hodgson, E and Kuhr, R. J., eds.), Marcel Dekker, New York, pp l-l 8 Hedm, P A., Menn, J J , and Hollmgworth, R. M , eds. (1994) Natural and Englneered Pest Management Agents American Chemical Society, Washmgton, DC Godfrey, C. R A., ed (1995) Agrochemzcals from Natural Products. Marcel Dekker, New York Casida, J E., ed. (1973) Pyrethrum The Natural Insecticide. Academic, New York Hansen, D. J., Cuomo, J., Khan, M., Gallagher, R. T , and Ellenberger, W P (1994) Advances m neem and azadnachtin chemistry and bioactivity, m Natural and Engmeered Pest Management Agents (Hedm, P A., Menn, J. J , and Holhngworth, R. M., eds.), American Chemical Society, Washmgton, DC, pp. 103-129 8 Mrozik, H (1994) Advances in research and development of avermectms, m Natural and Engmeered Pest Management Agents (Hedin, P A., Menn, J J., and Holhngworth, R M., eds.), American Chemical Society, Washington, DC, pp 54-73 9 Addor, R. W. (1995) Insecttcides, m Agrochemlcals from Natural Products (Godfrey, C R. A , ed ), Marcel Dekker, New York, pp l-62 10 Kornis, G. I. (1995) Avermectms and Milbemycins, in Agrochemzcals from Natural Products (Godfrey, C R A., ed.), Marcel Dekker, New York, pp 215-255

186

Sparks et al.

11 Kuhr, R J and Dorough, H. W (1976) Carbamate insecttctdes Chemutry, Btochemtstry and Tox~ology CRC, Cleveland, OH 12 Henrrck, C A (1995) Pyrethrords, m Agrochemrcals from Natural Products (Godfrey, C R A., ed.), Marcel Dekker, New York, pp. 63-145 13 Henrrck, C A (1995) Juvenotds, m Agrochemzcals from Natural Products (Godfrey, C. R A , ed ), Marcel Dekker, New York, pp, 147-2 13 14 Knst, H A , Mrchel, K H , Mynderse, J. S., Chro, E H., Yao, R. C , Nakatsukasa, W M , et al (1992) Discovery, isolation and structure eluctdatron of a family of structurally unique, fermentanon derived tetracychc macrohdes, m Syntheses and Chemzstry of Agrochemtcals Iii (Baker, D R , Fenyes, J G , and Steffens, J J , eds ), American Chemical Society, Washmgton, DC, pp 2 14-225 15 Mertz, F P and Yao, R C (1990) Saccharopolyspora spznosa sp nov. isolated from so11 collected m a sugar ml11 rum still Int J Cyst Bactertol 40, 34-39 16 Sparks, T C , Thompson, G D , Larson, L. L , Kn-st, H A, Jantz, 0 K , and Worden, T V (1995) Brologrcal characterrsttcs of the spmosyns a new class of naturally derrved insect control agents, m Proceedmgs of the 1995Beltwtde Cotton Production Conference,Nattonal Cotton Councrl, Memphis, TN, pp 903-907 17 Krrst, H A , Mtchel, K H , Mynderse, J. S , Creemer,L C , Chro, E. H , Yao, R C., et al (199 1) A83543 A-D, unique fermentation-derrved tetracychc macrohdes Tetrahedron Lett 32,4839-4842 18 Sparks, T C , Ktrst, H A , Mynderse, J. S., Thompson, G D , Turner, J R , Jantz, 0 K , et al. (1996) Chemistry and btology of the spmosyns components of spmosad(Tracer@), the first entry mto DowElanco’s Naturalyte class of Insect control products, m Proceedtngs of the I996 Beltwtde Cotton Productton Conference, Natronal Cotton Counctl, Memphts, TN, pp 692-696. 19 Mynderse, J S , Martm, J W , Turner, J R., Creemer, L C , Knst, H A , Broughton, M C , and Huber, M L B (1993) US Patent 5202242 20 Chen, S. T , Hensens,0 D , and Schulman,M D (1989) Biosynthesis,m Zvermectm and Abamectm (Campbell,W C , ed ), Sprmger-Verlag,New York, pp 55-72 21 Thompson, G D , Busacca, J D , Jantz, 0. K , Borth, P. W., Noltmg, S P , Wmkle, J R , et al (1995) Freld performance m cotton of spmosad.a new naturally dertved insect control system, m Proceedings of the 199.5Beltwtde Cotton Production Conference, National Cotton Councrl, Memphts, TN, pp 907-910 22 DeAmtcrs,C V , Dnpps,J. E., Hatton,C J., andKarr, L L (1997)Physicalandbtologlcal propertiesof the spmosynsnovel macrohdepestcontrol agentsfrom fermentatron,m PhytochemtcalsforPestControl(Hedm,P A ,Hollmgworth,R ,Masler,E P, Mlyamoto, J , andThompson,D , eds.),Amerxan ChemrcalSociety,Washmgton,DC, pp. 144-154 23 Evans, D A and Black, W C. (1992) Asymmetric synthesesof macrohde (+)A83543A (leprctdm) aglycon. J. Am Chem Sot 114,2260-2262 24 Evans, D A and Black, W C (1993) Total synthesis of (+)-A83543A [(+)leptcrdm A] J Am Chem Sot 115,4497-45 13 25 Thompson, G D , Busacca, J D , Jantz, 0. K , Larson, L L , and Sparks, T C (1995) Spmosyns an overview of new natural managementsystems,In Proceedmgsof the 199.5Beltwtde Cotton Productron Conference, National Cotton Councd, Memphis, TN, pp 1039-1041

Spinosyns

187

26 Larson, L L (1995) Laboratory toxicity of spmosad to late second mstar tobacco budworm compared to commercial standards, 1994. Arthopod Manage Tests 20,356 27 Klrst, H A. (1998) Fermentation-derived compounds as a source of new products Pure Appl Chem , m press 28 Edwards, J M , Karr, L. L , Schneider, M B , and Paterson, E (1995) Potential of spmosad, a Naturalyte insect control product, as a control agent for Dlptera. Entomological Society ofAmenca National Meetmg, December 17-20, 1995, Las Vegas, NV 29 Eldefrawl, M E. and Eldefrawl, A. T. (1990) Nervous system-based Insecticides, m Safer Znsecticzdes Development and Use (Hodgson, E. and Kuhr, R J , eds ), Marcel Dekker, New York, pp 155-207 30 Sparks, T C (1996) Toxicology of insecticides and mltlcldes, m Cotton Znsects and Mites Characterlzatzon and Management (King, E. G., Phillips, J. R., and Coleman, R J , eds.), Cotton Foundation, Memphis, TN, pp 283-322 31 Salgado, V L , Watson, G. B , and Sheets, J J (1997) Studies on the mode of action of spmosad, the active ingredient m Tracer@ Insect Control, in Proceedings of the 1997 Beltwlde Cotton Production Conference, National Cotton Council, Memphis, TN, pp. 1082-I 086 32 Schoonover, J R and Larson, L L (1995) Laboratory activity of spmosad on non-target beneficial arthropods, 1994 Arthropod Manage Tests 20,357 33 Borth, P W , McCall, P J , Blshoff, R F , and Thompson, G D (1996) The environmental and mammalian safety profile of Naturalyte Insect Control, m Proceedmgs of the 1996 Beltwide Cotton Productzon Conference, Natlonal Cotton Council, Memphis, TN, pp 69&692 34 Hollmgworth, R M (1976) The biochemical and physlologlcal basis of selective toxlclty, m Insectrczde Bzochemtstry and Physiology (WIlkinson, C F , ed ), Plenum, New York, pp 43 l-506 35. Sparks, T C (1980) Development of msectlclde resistance m HellothIs zea and Hellothzs vlrescens m North America Bull Entomol Sot Am 27, 186-192 36 Sparks, T C , Graves, J B , and Leonard, B R (1993) Insectlclde resistance and the tobacco budworm. past, present and future, m Reviews In Pestlclde Toxrcology, vol 2 (Roe, R M and Kuhr, R J, eds.), Toxicology Commumcatlons, Raleigh, NC, pp 149-183. 37 Martin, S H , Graves, J B , Leonard, B R , Burns, E , Mlcmskl, S , Ottea, J A , and Church, G. (1994) Evaluation of msectlclde resistance and the effect of selected synergists tn tobacco budworm, in Proceeding of the 1994 Beltwlde Cotton Productzon Conference, Natlonal Cotton Council, Memphis, TN, pp. 8 l&823 3x Leonard, B R , Graves, J B., Burrls, E , Mlcmskl, S , and Mascarenhas, V (1996) Evaluation of selected commercial and experimental msectlcides agamst lepldopteran cotton pests m Louisiana, m Proceedmgs of the 1996 Beltwzde Cotton Production Conference, National Cotton Council, Memphis, TN, pp 825-830 39 Leonard, B. R , Graves, J. B., Sparks, T. C , and Pavloff, A M (1988) Varlatlon m field populations of tobacco budworm and bollworm (Lepldoptera. Noctuldae) for resistance to selected msectlcldes. J Econ Entomol 81, 152 1-l 528 40 Sparks, T. C., Sheets, J. J , Skomp, J. R., Worden, T V., Larson, L L., Bellows, D., Thlbault, S., and Wally, L (1997) Penetration and metabolism of spinosyn A

188

41

42 43 44 45 46 47 48 49

50

5I 52

53 54

55 56

57.

58

Sparks et al

into leptdopterous larvae, III Proceedings of the 1997 Beltwzde Cotton Production Conference Natlonal Cotton Council, Memphis, TN, pp 1259-1264 Bull, D. L. (1986) Toxicity and pharmacodynamlcs of avermectm m the tobacco budworm, corn earworm and fall armyworm (Nocturdae Lepldoptera). J Agrzc Food Chem 34,74-78. Graves, J B , Clower, D F , Bagent, J L , and Bradley, J R (1964) Bollworms increasing m resistance to msectlcides. LA Agrlc 7,3,16 Graves, J. B , Clower, D. F , and Bradley, J R. (1967) Resistance of the tobacco budworm to several insecticides m Louisiana J Econ Entomol 58, 583,584. Lankas, G. R and Gordon, L R (1989) Toxicology, m Ivermectln and Abamectln (Campbell, W C., ed ), Springer-Verlag, New York, pp 89-112 Ware, G W (1983) Pestlcldes Theory and Applzcation W H Freeman, San Francisco Merck (1995) Emamectln Benzoate Inseclzczde, Technzcal Data Sheet. Merck, Three Bridges, NJ Melster, R. T., ed (1996) Farm Chemzcals Handbook ‘96 Meister, Willoughby, OH Naumann, K (1990) Synthetic Pyrethrozd Insectlczdes Structures and Properties Sprmger-Verlag, New York Palazzo, R. J (1976) Comparison of the responses of adults and larvae of five lepldopteran species to seven msectlcldes. M.S Thesis, LouIslana State Umverslty, Baton Rouge Rose, R L and Sparks, T C (1984) Acephate toxlclty, metabohsm, and antlchohn-esterase activity m Helrothzs vzrescens (F ) and Anthonomus grandis grandts (Boheman) Pestle Blochem Physlol 22,69-77 DowElanco (1994) Spmosad Technical Gmde DowElanco, IndIanapolls, IN Wlslockl, P. G., Grosso, L S , and Dybas, R A. (1989) EnvIronmental aspects of abamectm use in crop protectlon, in Ivermectin and Abamectin (Campbell, W. C,, ed ), Springer-Verlag, New York, pp 182-200 Lasota, J A and Dybas, R. A. (1991) Avermectins, a novel class of compounds implications for use m arthrodop pest control. Ann Rev Entomol 36, 9 1-I 37 Hill, I R (1985) Effects on non-target organisms m terrestrial and aquatIc envlronments, m The Pyrethrold Insectlcldes (Leahey, J P , ed ), Taylor and Francis, Philadelphia, pp. 15 l-262 Lltchfield, M. H. (1985) Toxlclty to mammals, m The Pyrelhrold Insectlczdes (Leahey, J P , ed ), Taylor and Francls, Phlladelphla, pp. 99-150 Hamon, N., Shaw, R., and Yang, H (1996) Worldwide development of fiproml msectlclde, m Proceedings of the I996 Beltwzde Cotton Productlon Conference, Natlonal Cotton Council, Memphis, TN, pp. 759-765. Elbert, A , Overbeck, H , Iwaya, K., and Tsuboi, S (1990) Imrdaclopnd, a novel systemic mtromethylene analogue msectlcide for crop protectlon. Brighton Crop Protect Conf Pests Dls 1,21-28 Mullms, J. W (1993) Imldaclopnd. A new mtroguamdme msectlclde, m Pest Control wzth Enhanced Environmental Safe& (Duke, S 0 , Menn, J. J , and Plmuner, J R., eds ), American Chemical Society, Washington, DC, pp. 183-l 98.

12 Bacillus

thuringiensis

Na fural and Recombinant Bioinsecticide Products James A. Baum, Timothy B. Johnson,

and Bruce C. Carlton

1. Introduction Worldwide sales of Bacdlus thurznglensis (Bt) dwarf those of any other biopesticide product. Annual sales in the early 1990s were estimated at $100 million, accountmg for l-2% of the global insecticide market (1,2). The largest market for Bt-based bioinsecticides, estimated by van Frankenhuyzen (3) to be -60% of the total Bt market m 1990, is m the protection of vegetable and horticultural crops from lepidopteran pests. The remainder of the Bt market Includes applications for the control of forest pests (3), particularly m North Amertca, dtpteran pests that act as vectors of human diseases (2), lepidopteran pests on cotton, and coleopteran pests on solanaceous crops, Lambert and Peferoen (2) and van Frankenhuyzen (3) provide fine historical overviews of the development of Bt as a commercial bioinsecticide. Over the past 15 yr, much progress has been made in understanding the molecular and genetic basis of Bt insecticidal activity. The recent review by Cannon (4) covers many aspects of Bt molecular biology. In this chapter, we will highlight advances in the development of improved btoinsectrcide products based on recombinant or genetically modified strams of Bt. 2. The Bacterium and Its Crystal Proteins Bt is a Gram-positive spore-forming bacterium that produces parasporal inclusions (or crystals) during stationary and/or sporulatron phase, the mclusions bemg composed of crystal (Cry) proteins that are toxic to a wide variety of insect species. The presence of parasporal rnclusions distmguishes Bt from From Edlted

Methods by

tn B!ofechnology,

F R Hall

vol 5 &opesttcdss

and J J Menn

0 Humana

189

Press

Use and De/wry Inc , Totowa.

NJ

190

Baum,

Johnson,

and Car/ton

the common so11bacterium Baczllus cereus Although Bt can be isolated from the sot1 (5) and from fohar surfaces (6), it IS most abundant m gram dust, the debris recovered from gram ~110sand other gram storage faclhtles. Bt IS classtfied primarily on the basis of flagellar antigen serotypmg (7) This classlficatlon system, comprising 45 distinct serotypes representmg 55 serovars of Bt (8), correlates well with morphologlcal and blochemlcal characteristics of the bacterium, but IS a poor predictor of msectlcldal actlvlty Dlstmct serovars are classified as subspecies of Bt, for Instance, subspecies kurstakt, a~zawu~, mornsonz, and wuelensu. Although certain CT genes are commonly found m certain serovarsor subspecies(e.g.,cvyl Cu m subspeciesu~zawuz and entomoczdus), the correlation IS, m general, very poor. For instance, strains of the subspecies mowzsonzhave ylelded crystal proteins with lepldopteran, coleopteran, or dlpteran toxlclty Fmally, the use of subspecies designations as an Indicator of msectlcldal actlvlty IS even less reliable when discussing recombinant or genetically modified Bt strams, since virtually any combmatlon of crystal protem genes may be constructed usmg molecular genetic techniques Corporate, mstttutlonal, and government strain collections of Bt contain thousands of strain Isolates from around the world. The rapid growth m new cry genes reported m the sclentlfic and patent literature over the past few years, mostly becauseof the genediscovery programs of companiesmvolved m Bt blomsectlclde and transgemc plant development, has prompted the adoption of a new nomenclature system that categorizesthe encoded Cry proteins on the basis of ammo-acldsequenceidentity (9), rather than on msectlcldal activity (IO) The Cry proteins of Bt, also referred to as &endotoxms, comprise a diverse group of msectlcldal agents. As of this writing, there are -70 different classes/subclasses of Cry proteins, representing at least four distinct protein families that have apparently co-evolved toxicity toward insects (II). Cry proteins with toxtclty toward leprdopteran, dlpteran, and coleopteran insect larvae have been well documented. Proteins with toxicity toward nematodes,protozoans, flatworms, and mites have also been reported (22). Strategies for identifying new Cry proteins and their genes rely heavily on bioassay screening and on molecular methods employing the polymerase chain reaction (PCR) or colony blot hybridlzatlon and gene-specific ohgonucleotldes as PCR primers or as hybridization probes (13-15) Insect colonies resistant to certain classes of Cry proteins (26) can be particularly valuable m ldentlfymg toxins with different modes of action Recent reviews are available that discuss cry gene expression m Bt (27,18), cry gene dlverslty and evolution (12,19), and Cry protein structure and function (20). 3. Development of a Successful Bioinsecticide The emergence of Bt as a successful blomsecticlde necessitated technical advances m a number of dlsclplmes outside of molecular biology. Among those

Bt. Bioinsecticide

Products

191

key developments was the lsolatton of strain HDl (21), a kurstakz strain with potent toxicity toward a number of important lepidopteran pests, which for many years has been a standard mdustrtal production strain. Other important developments included the adoption of an international standard for potency, improvements in fermentation yield, the introduction of standardized, and, subsequently, more concentrated formulattons, and developments m field apphcation technology (3). Finally, the tdentificatton of Bt strains with toxtctty toward dipteran (22) and coleopteran (23) pests expanded the use of Bt-based btoinsecttctdes to other markets. Most Bt-based btoinsecttctde products are produced using naturally occurrmg strains of Bt, and utrhze only a small fraction of the known Cry proteins. In the United States and Canada, derivatives of Bt strain HDl subsp kurstakz have become the major pesticide used to control the gypsy moth, Lymantrla dlspar (24), and the spruce budworm, Choristoneura fumzferana (3), respectively. Examples of strain HD 1-based products for forestry use are marketed under the registered product names Foray@ 48B and Dtpel@ 6AF (Table 1) Other forest pests controlled by Bt include the nun moth (Lymantrza monacha), the Asian gypsy moth (L dispar), the pme processtonary moth (Thaumetopoea pztyocampa), and the European pine shoot moth (Rhyaczonzabuohana) (25). Bt products based on Bt subsp zsraelensis(“Bti”) have proven to be effective m the control of mosquitoes and black flies worldwrde (26). Examples of registered Bti-based products include Vectobac@(Abbott), Bactlmos@(Solvay/ Duphar), Teknar@ (Therm0 Trtology), and Skeetal@(Abbott). In agrtculture, Bt products have been used successfully m the vegetable, cotton, and specialty crop (frurts, nuts) markets, almost exclustvely for the control of fohar-feeding leptdopteran pests. Although this use 1slimited, compared to that of conventional msectictdes, renewed interest in integrated pest management to slow insect-resistancedevelopment, public concern about conventtonal pesticide use, and the rising costs of developing new synthettc msecttcides all suggest that this btologtcal control agent will become increasingly important in the years to come. Table 1 provides a listing of some of the better-known Bt-based biomsectictde products, as well as some recently registered products. 4. Opportunities for Improving Bt-Based Bioinsecticides Htstorically, several factors have limrted the use of Bt m plant protectton, particularly m agriculture Bt strains have a narrow spectrum of msecttctdal activity when compared to conventional msecttctdes, typically exhtbttmg slgnificant toxtctty toward only one order of insect species. Even wtthm an order of insects (e.g., Lepidoptera), dramatic differences in sensitivity are exhibited among species. For instance, the beet and fall armyworms (Spodoptera spp) are notortously difficult to control with Bt-based biomsectlcides based on strain

Baum, Johnson, and Carlton

192 Table 1 Registered

B&Based

Bioinsecticide

Products

Strain

Product Able Agree BlobIt Bactospeme Condor Costar CRYMAX Cutlass Design Dlpel Foil Foray Florbac Futura Javelin Lepmox MATTCH MTRAK MVP Novodor Raven Steward Thurlcide Trident Vault Xentari

for Agricultural

Use

Insect

background

CompanyC

order

kurstakl azzawai HD 1 kurstakz kurstakc kurstakl kurstakl kurstakz kurstakz alzawal HD 1 kurstakz kurstakz HD 1 kurstakz azzawal kurstakl HDl kurstah kurstaki Pseudomonas Pseudomonas Pseudomonas tenebrlomb kurstakl HD 1 kurstakr HD 1 kurstah tenebnoms HD 1 kurstakz aizawal

Therm0 Trlology Therm0 Trlology Abbott Abbott Ecogen Therm0 Trlology Ecogen Ecogen Therm0 Trlology Abbott Ecogen Abbott Abbott Abbott Therm0 Trlology Ecogen Mycogen Mycogen Mycogen Abbott Ecogen Therm0 Triology Therm0 Trlology Therm0 Trlology Therm0 Triology Abbott

L L L L L L L L L L L/C L L L L L L C L C L/C L L C L L

Comments Transcorqugant Bt TransconJugant Bt -

Recombinant Bt Transcoqugant

Bt

Transcotqugant Bt Transcoqugant Bt Recombinant Bt EC” EC EC Recombinant Bt -

“Encapsulated crystal proteins. %enebnoms = subspecies mormonc ‘Abbot Laboratories, Chlcago, IL, Ecogen, Inc , Langhome, PA, Mycogen Corp , San Dlego, CA, Therm0 Trilogy Corp , Columbia, MD L, lepldopteran-toxw, C, coleopteran-toxvz

HDl,

but the tobacco budworm (Helzothu vzrescens) and the diamondback xylostella) are not. In agriculture, these sensitlvlty dtfferences have a sq+?cant impact because multispecies pest complexes are typically the rule rather than the exception. In contrast, the speclfictty and safety of Bt 1s an advantage m the forestry and vector control markets, m which major target pests are fewer m number, and in which Bt-based btoinsectictdes are sprayed on relatively complex ecosystems, where nontarget orgamsms abound.

moth (Plutella

Bt Bioinsecticide Products

193

The efficacy of Bt is also limited by the nature of its mode of action. The Cry protems must be Ingested m order to effect mortality. The longer the Cry protem is presented to feeding larvae, the greater the chances for insect control. Thus, the efficacy of Bt-based bioinsecticides is affected by the timmg of spray application, spray coverage, larval feeding behavior, the ram-fastness of the formulation, and by the inactivation of both the spore and crystal by sunlight Improvements m Bt formulation can address many of these issues, including sunlight inactivatron (27). Encapsulation of crystal proteins within the host cell has been advanced as a means to improve persistence (28,29), yet the major environmental factor impactmg field persistence of Bt Cry proteins is almost certainly UV light (30,31), and encapsulationper se offers no protection against UV mactivation. Indeed, reports of a twofold increase m foliar persistence of encapsulated Cry protems (28) may not be as dramatic as improvements in msecticrdal potency and Cry protein yield resulting from genetic manipulation of the Cry protein genes m Bt (see Subheading 7.)

5. Genetic Manipulation of Bt The genetic manipulation of cry genes m Bt offers a promising means of improving the efficacy and cost-effectiveness of Bt-based bioinsecticide products. Certain combmations of Cry proteins have been reported to exhibit synergistic toxicity toward lepidopteran (32,33) and drpteran pests (34-36). In addition, the presence of spores can also synergize the activity of Cry protems against certain lepidopteran pests (37-40), and may forestall the development of msect resistance to Cry proteins (38). The contrrbution of the spore to Insect mortality, and its posstble utrlrty m resistance management, has been largely ignored by those advocating the expression of cry genes in alternative hosts. Other factors may contribute to the entomopathogemc character of Bt, mcluding the vegetative insecticrdal proteins (VIP) (#I), a-endotoxm (42), and a variety of secondary metabohtes (43), including Zwittermycm (44,45). These too may be amenable to genetic manipulation. The cry genes are almost exclusively localized on large plasmids (46,47), frequently on multiple plasmids, some of which can be transferred from one Bt strain to another by a conjugation-like process (4849). Thus, the natural processes of plasmid curing and plasmid transfer have been used to construct transconlugant strains with improved msecticidal properties (50). The curing and transfer of native or resident cry plasmids has hmrtattons, however, from the standpoint of product improvement. Most cry genes are not readily transferred by conmgal transfer. Furthermore, cry genes tend to be linked on the same large plasmrd (e.g., the -110 MDa plasmid of strain HDI, the -75 MDa plasmid of Bti), so that genes encoding superior toxins (for a partrcular target pest) cannot be readily separated from genes encodmg inferior toxins.

194

Baum, Johnson, and Carlton

The use of recombmant DNA methodologies can circumvent these problems associated with Bt strain improvement. In addition, the ability to transfer cloned genes into Bt means that modified Cry proteins engineered for improved productron or toxicity can now be readily used as active ingredients. Coqugation (48,49) and transduction 1.51)have been used to transfer recombinant plasmids from a donor Bt to a recipient Bt; however, the preferred method of gene transfer employs the use of electroporatron, for which numerous protocols are available (see ref. 50 for review). A variety of Eschenchm c&z-Bt shuttle vectors have been constructed to facrhtate the mtroduction of cloned cry genes in Bt Some of these employ plasmid rephcons derived from other Gram-positive bacteria (e.g., pBC 16, pC 194); others employ rephcons isolated from native Bt plasmtds (52-54) In addition to these shuttle vectors, mtegrattonal vectors have been used to insert cloned cry genes into resident plasmtds (55,56), or mto the chromosome (57), by homologous recombmatron Figure 1 illustrates the use of a temperature-senstttve mtegratronal vector for thrs purpose. In several Instances, the transfer of a cloned cry gene mto a Bt host strain has resulted in an improved spectrum of msectlcidal toxicity (52,54,55,57) Heterologous promoters may be used to improve the expression of certain cry genes, mcludmg the promoters for the B. subtrlzs a-amylase gene (54), cry3Aa (18), and cry3Bb (17). Unlike most cry genes, the cry3A’a (and presumably cry3Bb) gene IS sporulation-independent and is induced or derepressed during stationary phase, presumably by transition-phase regulators (for reviews, see refs. 2 7 and 28) The use of these sporulatron-independent promoters may be useful in improvmg the production of sporulation-dependent Cry proteins Homologous recombmatron may be used, not only to integrate cry genes mto a resident plasmid, or mto the chromosome, but also to disrupt genes of interest Integrational vectors based on temperature-sensitive plasmid rephcons, such as pE194ts (58), are well suited for this purpose Examples of successful gene disruption experiments mclude disruptions of cytlA (cytA; 59), cryilA (cry1 VD, 60) hknA (61), spoOA (62), and apr (63), an alkaline protease gene m Bt. Dtsruption of spoOF (64) and a mutation m an uncharacterized spo0 gene (62) have each been shown to increase the productron of Cry3Aa encoded on a native plasmid by -2.5-fold. Dtsruption of apr by homologous recombination has been shown to enhance the production of Cry1 proteins m some instances (63). The use of recombination m Bt strain development was advanced further by the deployment of a site-specific recombination (SSR) system, to selectively delete ancillary or foreign DNA elements (e g , antibiotic resistance genes) from recombinant cry plasmtds after their mtroduction mto a Bt host (65,66) This SSR system IS composed of the TnpI recombmase of Tn.5401 (67,68) and its cognate site-specific recombmatton site, or internal resolution site (IRS)

Bt: Bioinsecticide

Products Of/-f.5

f

195 cat 1

restdent plasmid

Integrated

gene

Fig 1. Schematic diagram depicting the use of homologous recombmatlon to target cloned genes to specific genomic sites m Bt, employmg a pE 194ts mtegratlonal vector A cloned cry gene (solid box) IS subcloned mto a target DNA fragment (shaded) having sequence slmllarlty with a Bt plasmid or chromosome Preferably, at least 1 kb of target DNA 1spresent on either side of the cloned cry gene. Recombmatlon m Bt 1s allowed to occur by cultivating the recombinant strain at 30°C under selection for chloramphemcol resistance Subsequent plating under chloramphemcol selection at 41°C, the restrlctlve temperature for pEl94ts replication, allows for the isolation of colonies m which recombmation has occurred. Indlvldual colonies are then cultivated m 200 mL of 1X bram-heart mfuslon (Dlfco Laboratones, Detroit, MI), 0.5% glycerol (BHIG) at 3O”C, with subsequent passages m 2 x 200 mL of BHIG at 4 1‘C, to allow for resolution of the co-integrate intermedlate and loss of the temperature-sensitive mtegratlonal vector, giving rise to two possible outcomes that may be dlstmgulshed, for example, by PCR amphfication using flanking primers (arrowheads) and subsequent restriction enzyme analysis Structural maps of two E. co/z-Bt shuttle vectors, contammg duplicate copies of the IRS, are depicted m Fig. 2. The SSR plasmlds pEG939 and pZG940 are distinguished only by the Bt plasmld replication orlgm used to ensure stable maintenance m Bt. In this gene-transfer system, depicted m Fig. 3, a cry gene 1s inserted rnto the SSR vector, and introduced into a suitable Bt host by selectmg for tetracycline resistance The resulting recombinant strain is then trans-

196

Baum,

Johnson,

and Carlton

Sfil - Eagl . Clal Sstl Xhol BamHl Bhll Smal Asp71 8 Pstl Yh3l i

-___-.

pTZ19u

\

Xbal Sfil Salt (NsPISP) ’

B Sfil Eagl Sphl Hpal Clal Sstl Xhol BamHl Blnl Smal Asp71 8 Pstl Xbal

9u

Fig. 2. Schematic diagrams of the Bt-E. coli shuttle vectors pEG939 (A) and pEG940 (B). The pTZ19u fragment contains a replication origin functional in E. coli and the b-lactamase gene-encoding ampicillin resistance. Designations: oui43 and ori = Bt plasmid replicons (82), tet = tetracycline-resistance gene from the B. ce7eu.s plasmid pBC16, IRS = internal resolution site region from Tn5401 containing the TnpI recombination site. The restriction endonuclease sites Asp718, BamHI, BlnI, PstI, SmaI, SphI, and X/z01 (bold type) occur only once in the plasmids.

Bt: Bioinsecticide Products

197

IRS ori-Ec

cat

amp

tet

IRS

ori- ts 30 c tnpl

~

b

cat 37 c ori-ts

tnpl ori

pEG348 A

recipient cell Fig. 3. Use of the SSR vector system to introduce cloned cry genes into Bt. A crylC gene was inserted into the SSR vector pEG940 to yield plasmid pEG348. Plasmid pEG348 was introduced and maintained in Bt by selecting for tetracycline resistance at 30°C. The temperature-sensitive plasmid pEG922, encoding the Z’npI recombinase, was introduced into the recombinant strain by electroporation, selecting for chloramphenicol resistance at 30°C. The introduction of pEG922 resulted in expression of the TnpI recombinase, and a subsequent site-specific recombination event between the duplicate copies of the IRS. Plating of the recombinant strain at 37-41”C allowed for the isolation of colonies that have lost pEG922, but have retained the crylC plasmid pEG348A. Designations: amp, ampicillin resistance gene; ori-Ec, E. coli replication origin; cat, chloramphenicol acetyltransferase gene conferring chloramphenicol resistance. Other designations are described in Fig. 2.

formed with the temperature-sensitive Tn.5401 vector pEG922 (67), this time selecting for chloramphenicol resistance. The TnpI recombinase protein encoded by Tn5402 catalyzes the recombination event between the duplicate copies of the IRS on the cry-encoding SSR plasmid, resulting in the deletion of the E. coli replicon, ampicillin

resistance gene, and tetracycline

resistance gene,

198

Baum, Johnson, and Carlton

Table 2 Recombinant Bt Bioinsecticide Products Product

Strain

Raven

EG7673

CRYMAX

EG7841-I

Lepinox

EG7826 construct 11724

cry genes (no ) crylAc (2) cry3A cry3Bb” ciylAc (3) cry2A crylca crylAa crylAc (2) cry2A

US EPA reglstratlon January 1995

February 1996

December 1996

cryIF-IAc” “Encoded by recombmant plasmld

and the generation of a cry plasmld composed of a CT gene, a Bt plasmld rephcation origin, and a vestigial copy of the Tn5401 IRS region. The Tn.5401 vector pEG922 1ssubsequently cured by cultivation of the recombinant stram at 37’C, the restrlctlve temperature for pEG922 rephcatlon (67). The resulting recomblnant Bt strain contams only the modified cyyplasmld, and 1sfree of foreign DNA elements. This gene transfer system has been employed m the development of several new Bt-based blomsectlclde products (see Table 2). A stmllar SSR plas-

mid based on the Bt transposon T&430 has recently been described (69) 6. Bioinsecticide Products Based on Genetically Modified Bt Strains A number of bloinsecticlde products are based on transconjugant strains of Bt, including Agree @,Condor@, Cutlass@,Design@, and Foil@ (Table 1) For the constructIon of Condor and Cutlass, a self-transmissible cry1A plasmld from an aizawaz strain was transferred via conjugation to a kurstakz recipient strain. In the case of Agree/Design, a cry1 plasmld from a kurstaki strain was transferred to an alzawaz recipient strain. The active ingredient m Foil OF,

strain EG2424, produces both Cry1 AC and Cry3A protein, and exhibits toxlclty toward both lepidopteran and coleopteran pests This expanded insecticidal host range was accomplished by transferring a -88 MDa Cry3A-encoding plasmid from EG2 158 subsp mormoni to an HD263 subsp kurstakz-derived reclpl-

ent strain (50). The yield of Cry3A protein in large-scale production has been increased through the use of genetically modified strains. Bt strains used to produce Novodor@ FC (NB 176) and Foil BFC (an EG2424 variant) exhibit an oligo-

Bt. Bioinsecticide

Products

199

sporogenous phenotype and produce unusually large rhomboid crystals composed primarily of Cry3A protein Strain NB 176 was obtained by gamma Irradlatlon of Bt tenebrionls strain NB 125 (70); the Foil BFC strain was isolated as a spontaneous variant (Ecogen, unpublished data) The Cry3A overproduction phenotype of these strains appears to be caused in part by the sporulatlon-mdependent nature of cry3A expression, the prolonged synthesis of Cry3A protem during the terminal stationary phase of asporogenous cells, and the absence of sporulatlon-dependent proteases m asporogenous cells (17,18, 61,62,6#). In the case of Novodor stram NB176, a duphcatlon of the cry3A gene on its native plasmid probably contrlbutes to Cry3A overproductlon (56). 7. Bioinsecticide

Products Based on Recombinant Bt Strains Field testmg of recombmant Bt strains began m 1990, with small plot trials conducted by Sandoz Crop Protection. Since that time, a number of companies, including Abbott (711, AgrEvo USA (72), CIBA (73), Ecogen (74), and Sandoz Agro (75) have pursued the development of recombmant strams for commercial use Currently, there are three blomsectlclde products based on recombinant Bt strains that are registered with the US Environmental Protection Agency (EPA) (Table 2). In general, the reglstratlon of these products was obtained within 1 yr of submlsslon of the registration packet to the EPA. It may be concluded that there IS no serious impediment to the registration of recomblnant Bt-based biomsectlcldes of this nature m the United States.A brief descriptlon of the recombinant strains follows. The T&401-derived SSR system (described above) was used m the construction of strain EG7673, a recombmant Cry3-overproducmg strain that was approved as the active mgredtent m Raven TM Biomsectlcide by the EPA m January of 1995. Strain EG7673, contammg the recombinant cry3Bb plasmld pEG930.96, produces 3-4 times more Cry3 protein than the progemtor strain EG2424, the active mgredlent m Foil OF bloinsecticlde (66). This increase m crystal protein yield, presumably caused by the high copy-number of the cry3Bb plasmid (I 7), allows for more cost-effective use of this product for the control of Colorado potato beetle (Leptinotarsa decemlineata) larvae. The construction of CRYMAX strain EG784 I- 1 (Table 2) involved a series of genetic manipulations that included plasmid curing, conJugatlon, transformation, and site-specific recombmatlon. EG7841-1 IS a derivative of strain EG3 125, a naturally occurring kurstakz strain Isolated from a North American gram dust sample. EG3 125 contains two crylh genes and a cry2Aa gene on a - 110 MDa plasmid and a cry/& gene on a -46 MDa plasrmd. A cured denvatlve of EG3 125 missing the 46 MDa plasmld, designated EG60 12, was used as a recipient m a series of conJugation experiments to Introduce cry-encodmg plasmlds from other strains of Bt. One transcoqugant, designated EG4923,

200

Baum, Johnson, and Carlton

was Identified as having improved msecticldal actrvrty compared to EG3 125 m quantitative broassaysagainst a variety of leprdopteran pests.The introduced plasmrd m EG4923, a 56 MDa plasmid from Bt strain HD74 (761, encodes a crylAc gene, Thus, EG4923 contains an unusual cry gene cornpositron: three cryIAc genes and one cry2Aa gene. The multtple copres of crylAc allow for hrgher levels of Cry1 AC production than can be achieved with strains harboring a single cryZAc plasmtd (e.g., HD73). A spontaneous colony morphology variant of EG4923 recovered from a starch agar plate was found to produce 30-40% more Cry1 AC protein than strain EG4923. This uncharacterrzed EG4923 variant was subsequently used as a host strain for the mtroductron of a cloned crylC gene on plasmid pEG940 (Fig. 2B) by the method described above (Fig. 3). The resulting recombmant strain, EG7841-1, produced an additional -30% more Cry1 protein than the progenitor strain n-r small-scale fermentation, using a standard productton medium. Since the CrylC protoxm migrates slower than the Cry1 AC protoxin on SDS-polyacrylamide gels, the proportion of Cry 1C protoxin could be estrmated to be 30-40% of the total Cry 1 protem. Transcrrptron of the cry1 C gene from its native promoter on plasmrd pEG940 did not adversely affect cell growth or sporulatron Accordmgly, rt IS not necessary to use heterologous or sporulation-independent promoters to ensure increases in Cry1 production or effictent sporulation (77). A WDG (water-dispersible granule) formulation prepared by extrusron of the cell paste, as opposed to spray drying, provides CRYMAX WDG with desirable handling properttes and good coverage of folrar surfaces. Extensive field trtals with CRYMAX have demonstrated excellent efficacy when compared to other btoinsectrcrde products at recommended usage rates. CRYMAX WDG at 0.5 lb/acre was equivalent to Xentarr WDG at 1 lb/acre, and superior to MATTCH at 1 qt/acre for the control of the cabbage looper, Trzchopluszanz (Fig. 4). Furthermore, CRYMAX WDG at 0.75-l lb/acre provided better control than either Xentarr at 1 lb/acre or MATTCH at 2 qt/acre. Against diamondback moth populatrons showing resistance to Cry1 A- and Cry1 F-type proteins, CRYMAX WDG provided superior crop protection when compared to the azzawazproducts Xentart and Florbac, and comparable control when compared to the chemical msectrcides Agrrmek@ and Regent@(Fig. 5). The Cry 1C protem in CRYMAX contributes to rts toxicity toward armyworms: Against the yellowstrtped armyworm, Spodoptera ornithogalb, CRYMAX at 0.5 lb/acre provtded supertor crop protectton (Fig. 6) when compared to either Xentari (1 lb/ acre) or MATTCH (2 qt/acre), two products that also contain Cry1 C protem. Lepmox strain EG7826 construct 11724 (Table 2), a derrvatrve of Condor strain EG2348, was constructed using a Tn.5401-based SSR plasmrd contammg a chrmerrc cryZF’vyZAc (crylF-IAc) gene. The encoded fusion protein

Bt. Bioinsecticide

Products (field

Ambush

teals from Jan 1994 - Aug

1995)

2EC 0 I lb a.1 /a

MATTCH

2 0 qtla

MA’II

CH I 0 qtla

Xentari

WG I 0 lb/a

CRYMAX

I 0 lb/a

CRYMAX

0 75 lb/a

CRYMAX

0 5 lb/a

0

20

40

60

60

100

Mean Percent Control (range)

Fig 4. Field efficacy of CRYMAX WDG for control of the cabbagelooper. Sumof small plot trials employmg a randomizedcomplete-blockdesign (RCBD)

mary

contains a portion of the carboxyl-terminal half of Cry1 Ac Introduction of cryIF’-ZAc into the Condor host background resulted m a 25% mcrease m Cry 1 production when compared to a Condor recombinant stram containing the native cryZF gene. This yield increase IS presumed to be a result of more efficient crystal formation with the resident CrylA proteins in strain EG2348, an attribute presumably contributed by the Cry 1AC portion of the fusion protein. Expression of the cry1 F-IAc gene did not adversely affect cell growth or sporulatlon. In field trials on sweet corn (Fig. 7) and bentgrass (Fig. S), Lepmox WDG provided superior control of the fall armyworm, Spodoptera frugiperda, when compared to Condor OF, and equivalent control when compared to the chemical standards Lannate@and Scimitar@,respectively. Although different formulations were tested, the difference m field efficacy between Condor and Lepmox is representative of other comparisons using equivalent formulations, and reflects the improved potency toward S. fmgiperda contributed by the Cry1 F toxin. 8. Conclusions These developments underscore the validity of genetic manipulation as a means to improve efficacy/cost-effectiveness, and to expand the markets for Bt-based biomsecticrdes. A variety of methods may be used to introduce and stably maintain cloned cry genes, and in vlvo recombmatlon techmques may be used to delete or otherwise modify resident genes in Bt. These manipulations can result in strains with improved crystal protein production and msectl-

202

Baum, Johnson, and Car/ton 25 45, z 4g 35% 3$ 251 2rB Y

1.5

4

0.5..

5

4 05

-

175

1 94

;:33

4PI 2

123

l-

o-

~$~~~~q i

0 2”6

6 3Pl

G

s

1li

5I

<

d

F3E

Fig 5. Field efficacy of CRYMAX WDG for control of the dlamondback moth Results from two RCBD small-plot trials. Agrlmek IS abamectm from Merck Agvet Regent IS fiproml from Rhone Poulenc

126

CKYMAX 0 5 LB/a

Xentarl I 0 lb/a

Fig 6 Field efficacy of CRYMAX Results from a RCBD small plot trial

8

MATTCH 2 0 qlh

LANNATE 0 45 lb a I /a

Untreated Check

for control of the yellowstrlped

armyworm

tidal potency. Furthermore, the ability to transfer cloned cry genes into Bt makes it possible to use engmeered crystal proteins that exhlblt improved charactenstlcs. In the future, It is easy to imagme that crystal protein vanants, englneered for improved toxicity (78), yield, or stab&y by in vitro mutagenesis,

Bt. Bioinsecticide

Products

203

: j

IO9u-7 6 8$ 7$, 6. ii 523 4-

905

a

B

7i

3-

Jg

2-

2 % e

l0,

f <

25 b

245b

21

b

I IOlbh Lepmox

2Olbh

751bJa

WDG

215

!

r

IO 0 lb/a

Lepmox

1% G

b

2Oplla LANNATE

Untreated Check

Frg 7 Freld efficacy of Lepmox for control of the fall armyworm on sweet corn Results from a RCBD small plot trral (S Ah,

Umreatcd

Check

Untreated

Check

Scllnttar

Kmgston,

RI - Summer

1996)

5 0 n 02/a

0

I

1

1

5

IO

15

Total larvaepersquarefoot at 4 daysafter treatment

Fig 8 Field efficacy of Lepmox WDG for control of the fall armyworm on bentgrass Results from a RCBD small plot trial. Scrmrtar IS h-cyhalothrm from Zeneca may be used to steadily improve the performance of Bt-based btomsecttctdes. In addrtion, the construction of hybrid toxins with improved msecttcrdal acttvrty (79,80) ~111 provrde novel active Ingredients suitable for commerctal exploitatron. The mtroductton m 1996 of insect-resrstant cotton and corn cultrvars expressing cry genes raises questrons about the future of spray-on Bt products m some markets Despite the already demonstrated effecttveness of in planta

204

Baum, Johnson, and Car/ton

expression of cry genes for plant protection, there are concerns about the sustainability of this technology m the face of developing insect resistance. Unlike spray-on products, mcludmg traditional msectlcides, transgemc plants expressing a cry gene are typically engineered to produce toxin constitutlvely at the site of larval feeding, thus applying contmuous selection pressure against the insect pest. A variety of strategies have been proposed to slow resistance development, including the provlslon of temporal or spatial refugla for susceptible insects, the use multiple insecticidal protems with different modes of action (e.g., proteins with different receptors m the insect midgut), and the development of transgemc plants expressing ultra-high doses of toxin protein capable of klllmg heterozygote carriers of resistance alleles (81). Although the relative merits or pragmatism of these strategies can still be debated, It IS worth noting, albeit with hindsight, that two of these three strategies are used m the application of Bt-based blomsectlcldes. The limited field persistence of Bt-based bloinsectlcldes and the mcomplete coverage of fohar surfaces ensures the presence of insect refugla. In addttlon, Bt has evolved a pyramid strategy and readily produces multiple crystal proteins, sometimes with synergistic effects, and can be manipulated far more easily than plants to overproduce certain crystal proteins. These factors, together with the low development costs associated with the construction of recombinant or genetically modified strains, makes Bt-based biomsectlcldes ideal for Integrated pest management programs in those markets where field efficacy has been demonstrated. The premium charged for transgemc plant seed, required to recoup high development costs, and the season-long (albeit variable) expression of Cry protein zn pluntu, may make insect-resistant plants less practical as a tool for resistance management Despite the complex mode of action of Bt that hlstorlcally has limited its ability to compete with chemical msectlcldesm the marketplace, genetic mampulatlon of Bt and improvements in forrnulatlon have made possible the commerclahzation of blomsecticlde products with field efficacy rlvahng that of current insecticide standards. Bt-based blomsecticldes are the insecticides of choice in the control of North American forest pests, and in the control of insect vectors, because of their efficacy and environmental compatlblllty. Future improvements in the efficacy of Bt products for agricultural use should make them attractive alternatives to existing msectlclde products, and expand theurole m integrated pest management strategies. References 1 Lambert,B andPeferoen,M (1992) InsectlcldalpromiseofBaczlZus thunngzenszs Factsand mysteriesabout a successfulblopestlclde Bzosczence 42, 112-l 22 2 Bernhard, K andUtz, R (1993) Productionof Baczllus tlzurzngzenszs msectlcldes for experimentaland commercialuses,m Bacillus thurmglensis,An Envlronmen-

Bt- Bioinsecticide Products

3

4. 5 6. 7 8

9

10 11. 12. 13 14.

15

16 17 18. 19.

205

tal Bzopestzczde Theory and Practzce (Entwlstle, P. F , Cory, J S , Bailey, M J , and Hlggs, S , eds ), Wiley, Chtchester, UK, pp 255-267 van Frankenhuyzen, K (1993) The challenge of Baczllus thurzngzenszs, m Bacillus thuringlensls, An Envzronmental Bzopesticzde* Theory and Practzce (Entwlstle, P. F., Cory, J S , Bailey, M. J , and Hlggs, S., eds ), Wiley, Chlchester, UK, pp l-35 Cannon, R. J. C (1996) Baczllus thuringzensis use m agriculture, a molecular perspective B1o1 Rev 71,561-636. Travers, R S , Martin, P. A., and Relchelderfer, C F (1987) Selective process for efficient lsolatlon of sol1 Baczllus sp Appl Envzron Microbzol 53, 1263-1266 Smith, R A and Couche, G A. (1991) The phylloplane as a source of Bacrllus thurzngzenszs variants Appl Environ Mzcrobiol 57, 3 1l-3 15. de BarJac, H. and Bonnefol, A. (1968) A classification of strains of Baczllus thurzngzenszs with a key to their differentiation. J Invert Pathol. 11, 33.5-347 Lecadet, M. M , Frachon, E , Dumanolr, V. C., and de BarJac, H (1994) An updated version of the Baczllus thurzngzensis strams classification according to H-serotypes, m Abstracts of the VIth Irzternatzonal Colloquzum on Invertebrate Pathology and Mzcrobzal Control, Society for Invertebrate Pathology, Montpelller, France, p 345. Cnckmore, N , Zelgler, D R , Feltelson, J., Schnepf, H. E , Lambert, B., Lereclus, D , Gawron-Burke, C , and Dean, D. H. (1995) Revlslon of the nomenclature for the Baczllus thurzngzenszs pestlcldal cry genes, in Program and Abstracts of the 28th Annual Meetzng of the Soczetyfor Invertebrate Pathology, Society for Invertebrate Pathology, Bethesda, MD, p, 14. Hofte, H and Whiteley, H. R. (1989) Insecticidal crystal proteins of Baczllus thuringzenszs. Mzcrobzol Rev 53,242-255 WWW site http //www,biols,susx.ac.uklHomelNeil_Cnckmore/Btl~ndex.ht~nl Feltelson, J S (1993) The Baczllus thurzngzenszs family tree, m Advanced Engzneered Pestzczdes (Kim, L., ed.), Marcel Dekker, New York, pp 63-7 1. Gaertner, F H , Sick, A J , Thompson, M., Schnepf, H. E , Schwab, G. E., and Narva, K E (1995) Probes for the identification of Baczllus thurzngzenszs endotoxin genes. US patent 5430137 Ceron, J , Ort~z, A., Qumtero, R., Guereca, L., and Bravo, A. (1995) Specific PCR primers dlrected to Identify cvi and crylll genes within a Bacillus thurzngzenszs strain collection Appl Envzron Akrobzol 61, 3826-3831 Kuo, W -S. and Chak, K -F. (1996) Identification of novel cry-type genes from Baczllus thuringzenszs strains on the basis of restriction fragment length polymorphism of the PCR-amphfied DNA. Appl. Envzron Mzcrobzol 62, 1369-1377. Tabashmk, B. (1994) Evolution of resistance to Baczllus thurzngzenszs.Annu Rev Entomol 39,47-79. Baum, J A and Malvar, T. (1995) Regulation of insectlcldal crystal protein production m Baczllus thurzngzenszs. Mol Mzcrobzol 18, 1-12 Agaisse, H and Lereclus, D. (1995) How does Baczllus thurzngrenszs produce so much msectlcldal crystal protein’? J Bacterzol. 177, 6027-6032 Schnepf, H. (1995) Baczllus thurzngzenszstoxms. regulation, actlvitles and structural diversity Curr Opzn Bzotech 6, 305-3 12

206

Baum,

Johnson,

and Car/ton

20 Thompson, M A , Schnepf, H E , and Feltelson, J S. (1995) Structure, function, and engmeermg of Bacillus thurlngrensls toxins, m Genetic Engzneenng, vol 17 (Setlow, J K , ed ), Plenum, New York, pp 99-l 17 2 1 Dulmage, H. T (1970) Insectlcldal activity of HD-1, a new Isolate of Baczllus thurzngzenszs var alestl J Invert Path01 15, 232-239 22 Goldberg, L J and Margalit, J (1977) A bacterial spore demonstratmg rapid larvlctdal actlvlty against Anopheles sergentlz, Uranotaenza ungulculata, Culex unlvltattus, Aedes aegyptl, and Culexpiplens Mosquito News 37,355-358 23. Kneg, A , Huger, A M , Langenbruch, G. A , and Schnetter, W (1983) Baczllus thurzngzenszs var. tenebrlonls Em neuer, gegenuber Larven von Coleopteren wlrksamer Pathotyp Z Ang Entomol 96,500-508 24 Twardus, D (1989) USDA forest service gypsy moth aerial suppresslon/eradlcatlon proJects-1989 Gypsy Moth News 20,2. 25 Bowen, persona1 communication 26 Becker, N and Margaht, J. (I 993) Use of Baczllus thurzngzensrs lsraelensls against mosquitoes and blackflIes, m Bacrllus thunnglenszs, an Environmental Blopestlczde Theory and Practzce (Entwlstle, P F , Cory, J S , Bailey, M. J , and Hlggs, S , eds ), Wiley, New York, pp 255-267 27. Morris, 0 N (1983) Protectlon of Baczllus thurzngzenszs from mactlvatron by sunlight Can Entomol 115, 1215-1227 28 Gelerntner, W (1990) Targetmg msecticlde-resistant markets New developments m mlcroblal-based products, m Managing Resistance to Agrochemzcals From Fundamental Research to Practrcal Strategzes (Green, M B , LeBaron, H M , and Moberg, W K , eds ), American Chemical Society, Washington, DC, pp 105-l 17 29 Gaertner, F (1990) Cellular dehvery systems for msectlcldal proteins. living and non-hvmg microorganisms, m Controlled Delzvery of Crop Protectton Agents (Wllkms, R M , ed ), Taylor and Francis, London, pp 245-257 30 Pozsgay, M , Fast, P. G., Kaplan, H , and Carey, P (1987) The effect of sunlight on the protein crystals from Bacrllus thurmngzenszs subsp. kurstah HD- 1 and NRD- 12, a Raman spectroscopy study J Invert Path01 50,246-253 31 Pusztal, M , Fast, P , Grmgorten, L , Kaplan, H., Lessard, T , and Carey, P R (1991) The mechanism of sunlight-mediated mactlvatlon ofBacll1u.s thurlnglenszs crystals Bzochem J 273,43-47 32 Lee, M K , Curtlss, A , Alcantara, E , and Dean, D H (1996) Synerglstlc effect of the Bacillus thurlnglensts toxins CryIAa and CryIAc on the gypsy moth, Lymantrla

dlspar

Appl Environ. Mlcroblol

62, 583-586

33 Bradfisch, G A , Thompson, M , and Schwab, G E. (1996) Pestlcidal composltlons US patent 5508264 34 Chang, C , Yong-Man, Y , Shu-Mei, D , Law, S. K , and G111,S. S. (1993) High level cryIVD and cytA gene expresslon m Bacillus thurmgzenszs does not require the 20-kllodalton protem, and the coexpressed gene products are synerglstlc In their toxlclty to mosquitoes Appl Envzron Mzcrobtol 59, 8 15-82 1 35 Cnckmore, N , Bone, E J , Wllhams, J A., and Ellar, D J (1995) Contrtbutlon of the individual components of the F-endotoxm crystal to the mosqultoctdal activity of Bacillus thunnglensls subsp uraelensis. FEMS Mlcroblol Lett 131,249-254

Bt- Bioinsecfictde

Products

207

36. Wu, D , Johnson, J J , and Fedencl, B A. (1994) Synergism of mosqultocldal toxlclty between CytA and CryIVD proteins using mclus~ons produced from cloned genes of Bacillus thurmglensu. Mol Mzcroblol 13, 965-912 37 Moar, W J., Trumble, J T , and Fedencl, B A (1989) Comparative toxlclty of spores and crystals from the NRD- 12 and HD- 1 strams of Baczllus thurznglenszs subsp kurstakr to neonate beet armyworm (Lepidoptera Noctuldae). J Econ. Entomol 82, 1593-l 603 38 Moar, W. J , Pus&al-Carey, M., Van Faassen, H., Bosch, D., Frutos, R , Rang, C , Luo, K , and Adang. M. J (1995) Development of Bacillus thurzngrensls CryIC resistance by Spodoptera exzgua (Hubner) (Lepldoptera Noctuidae). Appl Environ Mtcrobiol 61,2086-2092 39. Dubols, N and Dean, D H (1995) Synergism between CryIA msectlcldal crystal proteins and spores of Banllus thurmgzensis, other bacterial spores, and vegetative cells agamst Lymantrla dlspar (Lepldoptera Lymantnidae) larvae Environ Entomol 24,174 1-l 747 40 Tang, J D , Shelton, A. M , Van Rie, J , De Roeck, S , Moar, W. J., Roush, R T , and Peferoen, M. (1996) TOXIC@ of Bacillus thurzngzenszsspore and crystal protem to reststant dlamondback moth (Plutella xylostella). Appl. Envzron Mlcroblol 62,564--569 41 Warren, G W , Kozlel, M. G , Mullms, M A., Nye, G J , Carr, B., Desal, N M., et al (1996) Novel pestlcldal protems and strams. Patent WO 96/10083. World Intellectual Property Orgamzatlon 42 Beegle, C. C and Yamamoto, T. (1992) History ofBaczllus thurznglensls Berlmer Research and Development Can Ent 124,587-6 16 43 LIU, C-L , Manker, D. C , Macmullan, A. M., Lufburrow, P A., and Starnes, R L. (1995) Novel pestlcidal composltion and Bacillus thunngzensu strain Patent WO 95/25 18 1. World Intellectual Property Orgamzatlon 44 Stabb, E V , Jacobson, L M., and Handelsman, J. (1994) Zwlttermlcm A-producmg strains of Bacillus cereus from diverse soils. Appl Envzron Mzcroblol 60,440444 12. 45 Manker, D C , Lldster, W. D., Starnes, R L., and Macintosh, S. C. (1994) Potentlator of Baczllus pestlcldal actlvlty. Patent WO 94/09630 World Intellectual Property Orgamzation. 46 Kronstad, J. W , Schnepf, H. E , and Whlteley, H R. (1983) Dlverslty of locations for Bacillus thurzngzenszs crystal protein genes. J. Bactenol. 154,4 19-428 47 Carlton, B C and Gonzalez, J M , Jr (1985) Plasmlds and delta-endotoxm production m different subspecles of Bacillus thunngzensu, m Molecular Bzology of Micro&al Dcfferentlatcon (Hoch, J A., and Setlow, P , eds ), American Society for Mlcrobrology, Washington, DC, pp. 246-252 48 Gonz6lez, J M , Jr and Carlton, B C (1982) Plasmid transfer m Baczllus thurmglenszs, m Genetic Exchange A Celebration and a New Generation (Strelps, U N , Goodgal, S H., Gmld, W R , and Wilson, G A, eds ), Marcel-Dekker, New York, pp 85-95 49 GonzBlez, J M., Jr, Brown, B J., and Carlton, B. C (1982) Transfer of Baczllus thurzngzensu plasmlds codmg for &endotoxm among strams of B thurznglensu and B cereus. Proc Nat1 Acad Scl USA 79,6951-6955. 50. Gawron-Burke, C. and Baum, J (1991) Genetic manipulation of Baczllus thuringzensis insecticidal crystal protein genes in bacteria, m Genetic Engzneerzng (Setlow, J. K , ed.), Plenum, New York, pp 237-263

208

Baum,

Johnson,

and Carlton

5 1 Lecadet, M -M , Chaufaux, J , Rlbler, J , and Lereclus, D (1992) Construction of novel Bacillus thurmgienszs strams with different msecticldal speclficltles by transduction and transformation. Appl Environ Mzcrobzol 58,840-849 52 Baum, J A., Coyle, D. M., Jany, C S., Gllbert, M P., and Gawron-Burke, C (1990) Novel cloning vectors for Badus thurmgzems Appl Environ Microbrol 56,3420-3428 53 Arantes, 0. and Lereclus, D. (199 1) Construction of clonmg vectors for Bacillus thurzngzensu.

Gene 108, 115-I 19.

54 Chak, K -F , Tseng, M -Y , and Yamamoto, T (1994) ExpressIon of the crystal protem gene under the control of the a-amylase promoter m Bacdlus thurmgzenm strains Appl Enwon Mxrobzol 60,2304-23 10 55 Lereclus, D , Vallade, M , Chaufaux, J., Arantes, O., and Rambaud, S. (I 992) Expansion of the msectlcldal host range of Bacdlus thurlnglenszs by zn vzvo genetlc recombmation. BzoRechnology lo,41 8-42 1 56. Adams, L. F , Mathewes, S , O’Hara, P , Petersen, A., and Gurtler, H (1994) Elucldatlon of the mechamsm of CryIIIA overproduction in a mutagemzed stram of Bacdlus thurrngzenszs var tenebrlonls Mel Muzrobiol 14, 38 1-389 57. Kalman, S , Kiehne, K. L , Cooper, N , Reynoso, M S , and Yamamoto, T (1995) Enhanced productlon of msectlcldal proteins m Bacillus thurmgzenszs strams carrying an additIona crystal protem gene in their chromosomes Appl Envrron Mlcroblol

61,3063-3068

58 Villafane, R., Bechhofer, D H , Narayanan, C S , and Dubnau, D (1987) Rephcatlon control genes of plasmld pE 194 J Bactenol. 169,4822-4829. 59 Delecluse, A , Charles, J -F., Kher, A., and Rapoport, G (1991) Deletion by m vzvo recombmatlon shows that the 2%kilodalton cytolytlc polypeptlde from Bacillus thurzngzenszs subsp. rsraelensrs is not essential for mosqultocldal actlvIty J Bacterlol 173,3374-3381 60 Poncet, S., Anello, G., Delecluse, A., Kher, A., and Rapoport, G (1993) Role of the CryIVD polypeptlde m the overall toxicity of Bad/us thurmglensts subsp lsraelensls

Appl Envwon

Mlcrobrol.

59, 3928-3930

61 Malvar, T , Gawron-Burke, C , and Baum, J. A (1994) Overexpresslon of Bad2~s thurmglenszs HknaA, a hlstldine protein kmase homolog, bypasses early Spomutations that result m CryIIIA overproduction. J Bacterlol 176, 4742-4749 62 Lereclus, D , Agalsse, H , Gommet, M , and Chaufaux, J. (1995) Overproduction of encapsulated msectlcldal crystal protems In a Bacdlus thunnglenszs spoOA mutant. Blo/Technology 13,67-7 1. 63 Tan, Y. and Donovan, W (1995) Clomng and charactenzatlon of the alkalme protease gene of Bacdlus thunngzensu, m Abstracts of the 95th Annual Meeting of the Amerczan SocletyforMcrobzology, Amencan Society for MIcrobIology, Washington, DC, p 406 64 Malvar, T. and Baum, J A. (1994) Tn54OI dlsruptlon of the spoOF gene, Ident]fied by direct chromosomal sequencmg, results m CryIIIA overproductlon m Bacdlus thurmgzensu.

J Bacterlol

176,4’750-4753

65 Baum, J A. (1995) Bacdlus thurmglensls transposon Tn5401. US patent 5441884 66. Baum, J A., Kakefuda, M , and Gawron-Burke, C (1996) Engmeermg Bacrllus thurrnglensls blomsectuzldes with an mdlgenous site-specific recombmatlon system. Appl Environ Mzcroblol 62, 4367-4373.

Bt. Bioinsecticide

Products

209

67 Baum, J A (1994) Tn.5401, a new class II transposable element from Baczllus thurzngzenszs. J Bacterzol

176, 2835-2845.

68 Baum, J. A. (1995) Tnpl recombmase. rdentrficatton of sites within Tn5401 required for TnpI bmdmg and sue-specrtic recombmatton J Bacterzol 177,403&4042 69 Sanchrs, V., Agarsse, H , Chaufaux, J., and Lereclus, D (1997) A recombmasemedtated system for elimmatton of antibiotic reststance gene markers ftom genetically engmeered Baczllus thurzngzenszs strams. Appl Envzron Mzcrobzol 63,779-784 70. Gurtler, H. and Petersen, A. (I99 1) Mutants or variants of Baczllus thurrngzenszs producing high yields of delta endotoxin Patent WO 9110748 1. World Intellectual Property Organization 7 1 Public report 1535 IO (1995) California Environmental Protection Agency 72 Public report 160833 (1996) California Envtronmental Protection Agency. 73 Federal Register (1994) Vol. 59, No. 143, pa 38,174 74 Johnson, T , Hannan, R , Gouger, R., Colbert, F , Jany, C , Jelen, A, and Baum, J (1995) Development of CRYMAX@ WDG-a recombmant BT product for control of vegetable insect pests. Program and Abstracts, 28th Annual Meeting, Society for Invertebrate Pathology, p. 32. 75 Cerf, D , Kalman, S., Cooper, N , Shahabl-Reynoso, M., and Yamamoto, T (1995) Small scale field trials of recombinant Baczllus thuringzenszs m Caltforma Program and Abstracts, 28th Annual Meeting, Society for Invertebrate Pathology, p 12 76. Dulmage, H T , Beegle, C C., de BarJac, H , Retch, D., Donaldson, G , and Krywienczyk, J (1982) Baczllus thuringzenszs cultures available from the U S Department of Agnculture, in US D A -A R.S Agrzcultural Reviews and Manuals, ARM-S-30/0ct 1982 LJS Department of Agrrculture, Agricultural Research Service, New Orleans 77 Sanchis, V., Agaisse, H , Chaufaux, J., and Lereclus, D (1996) Constructton of new msecttcrdal Baczllus thurzngzenszs recombinant strains using the sporulatton non-dependent expression system of cryMA and a site specific recombmation vector J Bzotechnol 48,81-96 78 RaJamohan, F , Alzate, O., Cotrill, J A , Curttss, A., and Dean, D H (1996) Protein engmeermg of Baczllus thurzngzensz’s &endotoxm: mutations at domam II of CryIAb enhance receptor affinity and toxrcity toward gypsy moth larvae. Proc Nat1 Acad Scz USA 93, 14,338-14,343

79 Bosch, D , Schtpper, B , Van der KlelJ, H., de Maagd, R A , and Sttekema, W J. (1994) Recombinant Baczllus thurzngzenszs crystal proteins with new properties possibtltttes for resistance management. Bzo/Z’echnology 12, 9 15-9 18. 80 de Maagd, R A , Kwa, M. S G., Van der Kleij, H , Yamamoto, T , Schipper, B , Vlak, J M , Sttekema, W. J , and Bosch, D (1996) Domam III subsmutton m Baczllus thurzngzensis delta-endotoxm CryIA(b) results in superior toxicity for Spodoptera exzgua and altered membrane protem recogmtion Appl. Enuzron Mzcrobzol 62, 1537-l 543 8 1. McGaughey, W H and Whalon, M E (1992) Managing resistance to Baczllus thurzngzenszs toxins Science 258, 1451-1455. 82 Baum, J. A and Gilbert, M P (1991) Characterization and comparative sequence analysts of replication ortgms from three large Baczllus thurzngzenszs plasmids J Bacterzol.

173,5280--5289

13 Transgenic Plants Expressing from Bacillus thuringiensis

Toxins

Johnie N. Jenkins 1. Introduction The science of plant breeding has been a major force m the development of new and improved cultlvars and hybrids of major agricultural crops. All people have benefited from these developments. In conventional plant breeding, one must usually find the trait of interest wlthm the same or closely related plant species. Despite this limitation, plant breeders have made remarkable progress in improvmg cultlvars of crop plants. The discovery of deoxyrlbose nucleic acid (DNA) as the basic molecule of the gene, and the deciphering of the genetic code, about the middle of the 20th century, provided hope that plant breeders would be able to utlhze genetlc variation from many species in the Improvement of crop plants. This dream required further developments m molecular biology. Recent advances in this field have opened the door for major new and mnovatlve techmques m plant breeding. Totlpotency implies that every plant cell has within It the genetics for the entire life cycle of the intact plant. This IS a basic concept of plant cell genetics and a feature not applicable to most animal cells (1). Because of this feature, any plant cell, in theory, can be induced to produce a new plant. This combmation of tissue culture and molecular biology has opened the way for genetic engineering, and 1sthe basis upon which scientists today can bridge the gap across species Today the plant breeder has access to genes from almost any species of living organism. This means that a desirable gene from bacteria can now be cloned and made to express its function m a plant. Thus, the species-crossing barrier, which has limited plant breeders for years, is no longer a barrier. From Methods m Biotechnology, vol 5 Btopesbndes Use and Del/very Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

211

Jenkms

212

7.1. The Bacterium

Bacillus thuringiensis Bacillus thurzngzensis (Bt) 1sa Gram-positive sot1 bacterium characterrzed by the abiltty to produce crystallme mclustons during sporulation (2) These inclusions constst of proteins that have insecticidal activity agamst several orders of insects. The crystal protems are protoxms that require proteolyttcal conversion into smaller toxic peptides in the midgut of insects The insecticidal protein crystals are also called 6-endotoxms It is thought that these toxms cause osmottc imbalance by creating pores m the cell membrane of the midgut epithelmm of suscepttble insects. An extensive classificatton system (2) divided the crystal protems into four major classes*Lepidoptera-spectfic, Lepidoptera and Dtptera-spectfic; Coleoptera-spectfic, and Dtptera-spectfic genes. There have been 96 genes that code for 6-endotoxms descrtbed thus far (3) The first msecttctdal protein-encoding gene from Bt was cloned, sequenced, and expressed m Escherzchzacolz m 198 1. This provided the prospects for usmg these and other msecttctdal proteins m transgenic plants (4). 7.2. Transgenic

Technology

1.2 1. Background

The commercialtzation of transgemc crops (particularly corn, Zea maize L.) with insect reststance, herbicide tolerance, and viral disease reststance by the year 2000 was predicted in 1989 by Gasser and Fraley (5) As it turns out, Insect-resistant corn and cotton, Gossypium hirsutum L , reached the commercial market m the mid-1990s. It was also suggested that, m addition to use of crops for commodity purposes, they would also be engmeered to grow specialty chemical products. In 1989,22 major herbaceous dicots, 3 woody dtcots, and 5 monocots had been transformed (5). Fraley (6) suggested m 1992 that sustamable agriculture could develop only by Investment in, and development of, new agrtcultural technologtes. He then showed some of the posstbilmes via genettcally engmeered plants, and expanded the list of species to nearly 50 that have been genettcally manipulated. This Included most of the major dicotyledonous crops and a rapidly mcreasmg number of monocotyledonous crops, mcludmg race and corn. Worldwide cost associated wtth management practices and chemical control of insects approaches $10 b&on annually, yet global losses caused by insects stall account for 20-30% of total productton. The potenttal exists to vutually transform any crop with a crop tailored gene (3,7). 1.2.2

Specifics

The first reported use of the Sendotoxm gene expressed m plants for insect control occurred m 1987 (8,9). Tobacco plants, Nzcotzana tabacum L , were

Transgenic Plants Expressmg Bt Toxins

213

developed that produced enough of the endotoxm to ktll first-mstar tobacco hornworm, Manduca sexta L., larvae placed on leaves of transformed plants The combmation of tissue culture and molecular biology holds great potential for genetic improvements of crop plants. The two most wrdely used methods of transforming plants are Agrobacterium-mediated transfer of DNA and bombardment of cells with DNA-coated particles. Agrobacterium tumefuciens Infects many dicotyledonous plants and causesthe growth of tumors or galls m the infected plants. Virulent strains of the bacterium contam large tumor inducing (TI) plasmids responsible for the DNA transfer and subsequent disease symptoms. These Ti plasmrds contam two sets of sequencesnecessary for gene transfer to the plant One of these sequences IS the transferred DNA [T-DNA) region, which IS transferred to the plant. The other sequence, the vnulence genes (vu), is not transferred during infection. The T-DNA regions are flanked by border sequences that determme the definition of the regton transferred to the infected plant. A special type of the bactermm, which has the Ti gene disarmed, is used m transformatron research. Disarmmg the Tr gene allows the bacterium to infect the plant, but does not allow the formatron of tumors. The desired gene is inserted between the flanked border sequences of the Tt plasmrd, where the tumor gene is normally located, and the bacterium then inserts the desired gene into the plant chromosome m a manner slmtlar to normal infection of the plant trssue in nature. Thus, the infection process of the A tumefaciens bacteria is used to insert the gene of interest, rather than the Ti gene, and becomes one of the vehicles by which foreign genes can be inserted into plants. Partrcle bombardment and transformation were first developed by Sanford and coworkers (10); licensing rights are currently held by DuPont (I). The second development was by scientists at Agracetus (1,11). A recent review of gene transfer by particle bombardment describes how this method is being used with plants and animals, and reports important applications of the process to produce transgenic maize and soybean, Glycine max, and for the introductron of DNA into plastids and mitrochrondria (12). This process has also been used to transform cotton (13). A recent review of brotechnology of cotton describes many of these processes (I). The method by which rice, Oryza sativa L. (Id)), and corn (15) were first transformed and whole plants regenerated was by electroporatton of protoplasts in a DNA solution. This mvolves application of high voltage across the protoplast-DNA suspension, whrch apparently causes small holes to form temporarily m the protoplast membranes, through which DNA can pass. Wilhams et al. (I@, at Ciba-Geigy AG, devised and demonstrated a system in tobacco for temporally controllmg the expression of genes by havmg the gene driven by a chemrcally responsive promoter. Thus system may be useful

214

Jenkins

m delaymg the development of resistant insect populations to the plantexpressed &endotoxm. 2. Transgenic Cotton 2.7. History Somattc embryogenesis and regeneration in cotton was first reported by Trolinder and Goodin (2 7), and genetically transformed cotton plants were first reported by Umbeck et al. (28), with plants regenerated that expressed the marker enzyme, neomycin phosphotransferase. Bt &endotoxin was first expressed in transgenic N tubacum plants in 1987, and provided resistance to tobacco hornworm (8,9). The first field test of transgemc upland cotton, G hirsutum L., containing a gene that codes for the &endotoxm from Bt kurstakz, was conducted m 1989 (29). Thus and other early research on transgemc cotton plants, contaming a gene that codes for the 6-endotoxin protein, showed that the wild-type version was not very effective m insect control (19,20) Expression of the toxic protein m plant tissue was Increased loo-fold, above the wild-type gene, through sitespecific modification m the coding sequences (DNA), use of improved promoter m Agrubacterium Ti plasmrd transformation vectors, and by usmg a truncated version of the gene (20,21). In plants, their modified crylA(b) and crylA(c) genes expressed toxic protein at 0.05 and 0.10% of total soluble protein in leaves, respectively In laboratory tests (20), these modified genes were effective against cabbage looper, Trichoplusia nzHubner, and beet armyworm, Spodoptera exzgua Hubner. Subsequent field tests confirmed that plants expressing these modified genes were capable of providing effective control of tobacco budworm, Hellothzs virescens (F.), pink bollworm, Pectznophera gossypzella Saunders, and of moderate population levels of the bollworm, Hellcoverpa zea Boddie (22-32) In the breeding of cultivars that contain transgenes,the choice of a transformed lme to use asa parent is not straightforward. The processesof transformation and regeneration are such that one cannot direct where the transgene is inserted mto the plant genome. Each msertton may be at adifferent chromosome site or genetic locus. Each transformation event or msertion must be evaluated for expression of the gene of interest and for associatedagronomic effects. These are influenced by somaclonal variation, position effects, and unknown causesassociated with the transformation process (19,2.5,28-31). Because plants of the Coker cotton cultivars are more easily regenerated than other cultivars, they are generally selected for transformation and regeneration (33). The development of cotton cultrvars with transgenes usually mvolves two distinct phases. The first phase mvolves selection of transformatton events (transformed plants) that express the &endotoxm protein at the desired level,

Transgenic

Plants

Expressing

Bt Toxins

215

wnhout any major negative effects on agronomic and fiber properties. The second phase mvolves hybrldtzatron of the selected transformed plant with elite germ plasm, followed by selection and evaluation of progeny to determme the expression of the &endotoxm and the agronomic and fiber properties of the selected lures. Ideally, both laboratory and field evaluattons would be conducted for each of these two phases of breeding (34). 2.2. State of the Art in Cotton The first field test of transgenic cotton (19), although not effective in controllmg Insects, was effective m estabhshmg protocols for field evaluatton of cotton plants with &endotoxm genes from Bt. The use of a btologtcal sink, composed of 24 border rows around the field, was shown to be effecttve m controllmg the spread of pollen from the transgenic plants to plants outside the border rows. A conststent and significant reduction in pollen dlssemmatron was measured as distance from the test plot increased. Outcrossmg decreased from 5 to Cl% by 7 m away from the test plot. A low level (~1%) of pollen drspersal continued sporadically m the remaining border rows to a distance of 25 m from the test plot (35). This established the protocols used m subsequent field tests with transgetuc cotton plants. Weight and survival of tobacco budworm on SIX msertron events from Monsanto (St. Louts, MO) (36) showed that each was effective m srgmficantly reducing weight of survtvors on cotyledons, seedling stems, first true leaves, terminal leaves, old leaves, squares, and petals of transgemc plants, compared to Coker 3 12 nontransformed plants There was also a stgmficant difference m level of &endotoxin among the SIX events, ranging from 0.0042 to 0.06% of extracted protein m first true leaves of field-grown cotton plants. These expertments (36) m 1990, using transgemc Monsanto events, contrasted wrth those (19) m 1989 with the transgemc events from Agracetus. The modified and enhanced expression of the F-endotoxin gene m the lines from Monsanto exhibited great potential for control of tobacco budworm larvae under field conditrons. Third mstar larvae of tobacco budworm spend significantly less time feeding and more time resting on transgenic plants expressmg the Cry 1A(c) or CrylA(b) protein than on nontransgentc plants m the greenhouse (30). Larvae on transgenic plants also tended to spin a silken thread and spur-down to a lower area of the plant. There was a significant reduction m damage by tobacco budworm and bollworm larvae to transgenic plants. Transgemc plants expressmg the &endotoxm genes have excellent potential for suppression of these insects m cotton productron. Three experimental lines from Delta and Pine Land Seed (Scott, MS) effectively controlled very high levels of tobacco budworm m field plots and pro-

216

Jenkins

duced yields equal to or htgher than the backcross recurrent parental lmes (37). These authors suggestedthat a new era had dawned m cotton productton with the advent of commercral release of culttvars dertved from these expertmental lines. Only one ttme during the season was the selected 2% square damage threshold exceeded m the transgemc plots in research conducted m South Carolma (38) Thus, control of tobacco budworm and bollworm was quite good. The transgemc event m this expertment was Monsanto event 1076, and dtd not yield as well m these experiments as the nontransformed Coker 3 12 Transformatton events M8 1and M53 1 from Monsanto were evaluated under field condmons m North Carolma, with large natural populattons of bollworm expected to migrate from maturing fields of corn mto the cotton plots late m the season.Overall control of tobacco budworm, bollworm, and European corn borer, Ostrinza nubzlalzs (Hubner), was excellent; however, for the first ttme, large populattons of bollworm were reported to cause stgmficant damage to transgemc cotton, with 14% boll damage at one location in 1993 (39). The transgemc event at this locatton was event M53 1, later commerctalized m BollgardTM culttvars. This report was the first to raise the posstbillty that transgemc Bollgard cotton may not be able to offer acceptable control of htgh levels of bollworm populattons without the use of one or more apphcattons of insecttctde. This should have alerted the seed compames commerctahzmg these cultivars to the possibility that these cultivars may benefit from the use of some msecttctde under sttuattons of very htgh levels of bollworm 2.3. Registration and Commercialization of Transgenic Cotton Cultivars In October of 1995, EPA granted registration for transgemc cotton expressmg the &endotoxm gene from Bt subsp kurstaki Cry 1A(c). The generic name of the active ingredient IS Bt kurstakz 6-endotoxm as produced by the crylA(c) gene, and its controllmg sequences as expressed m cotton. The common name 1sBtk subsp CryIA(c) &endotoxm m cotton Target pests are listed as cotton bollworm, tobacco budworm, and pmk bollworm. Early m the development of transgemc plants, and espectally cotton plants, many researchers m academia and industry were acttvely researchmg strategies for reststance management. High-dose expresston, coupled wtth refugta, to mamtam an adequate supply of susceptible moths to mate wtth any resistant moths, have been two of the maJor components of the predommant strategy (40,41) for tobacco budworm m the United States.Refugta are also part of the plan (42) for the use of these genes m Australia. Other approaches to a resistance management strategy mcluded the use of mixtures of plants with and without the delta endotoxin gene (J. N. Jenkins et al., unpubltshed data). Field plots in Mtssisstppt seeded with’a mixture of

Transgenic

Plants

Expressing

Bt Toxins

217

90% transgenic plants plus 10% nontransgenic plants were compared to 100% nontransgenic plants. The plots that were conventional cotton required 5.5 more applications of insecticide than the 90% transgemc plots. These transgenic plots (43) averaged $78 less for msecticides and made 198 lb more lint per acre than the conventional plots. In a similar study in South Carolma (44), but with various mixes, the data showed that more than 10% nontransgemc plants may not provide acceptable control of bollworm and tobacco budworm. Other field trials in the United States showed good efficacy against tobacco budworm or bollworm (34,45,46). Transgemc cotton cultivars express the cryZA(c) gene from Monsanto and are being marketed as Bollgard. Bt fohar spray products are used in cotton for control of tobacco budworm and bollworm. Registration of transgemc cotton cultrvars for control of selected lepidoptera was handled differently from potato or corn. EPA required a fully developed resistance-management plan, which Included refugia, be implemented as a condition of registration of transgemc cotton. Growers could choose between two refuge options. One plan required that four acres of conventional cotton (refuge) be planted for every 100 acres of transgenic cotton. With this option, no msecticides for control of leptdopteran insects could be used in this refuge. The second plan required that 25 acres of conventronal cotton (refuge) be planted for every 100 acres of transgemc cotton. Any msecticide except a Bt product could be used for control of lepidopteran insects in this refuge. Specific momtormg for shifts m levels or response to the &endotoxin was also required. This resistance-management strategy was primarily aimed at tobacco budworm, for which the transgemc cotton produces a high level of toxin. These cottons are also targeted to control bollworm and pmk bollworm. The toxin dose is not considered high for these two msects,especially for the bollworm. A resistance-management strategy was put mto place before any Bollgard cotton was sold to producers. This strategy was multifold and included refugia where tobacco budworm moths or larvae are not exposed to the &endotoxm or good integrated pest management (IPM) practices, and will include one or more additional resistance genes m the plant when the strategy is fully implemented. In 1996, about 1.8 mullion acres of transgemc cotton cultivars NuCOTN 33B and NuCOTN 35B were planted in the United States.Acreage increased m 1997. These cultivars carry the Bollgard gene by Monsanto. In general, the first large scale planting was very successful in controllmg target pests. No problems were reported with control of tobacco budworm; however, some fields with high population levels of bollworm required some msecticide applicatrons for bollworm control. Bollworm damage occurred m some fields when cotton was at peak bloom and bollworm populations were high. Perhaps the bollworm found a flaw m the plant’s genetic armor and exploited this situ-

218

Jenkms

ation for insect survival. As should have been expected, when a culttvar that had never been m any state official variety tests was planted on 1.8 mtlhon acres, there was a great probability rt would be planted on some soils and farms where tt was not well adapted. This occurred in 1996. The first field trials of cotton transformed with the &endotoxin gene from Bt occurred m Australia in the 1992-1993 season.The plants expressmg either the full-length or truncated versions of three transformation events showed efficacy against field populatrons of Helzcoverpa armzgeru. There were mdications of a declme m level of &endotoxin expression in some lines as the plants began to senesce(47,48). Biosafety results of field trials m Australta, and the possible consequences of expanding planting to more hectares, is discussed by Fttt and Jones (48). In the 1996-1997 growing season, about 80,000 ha were planted to transgetnc cotton m Australia. These were primarily planted to culttvars developed by Commonwealth Screntrfic and Industrral Research Orgamsatron (CSIRO) and marketed by Cottonseed Drstrlbutors, with a limited amount planted to cultrvars developed by Delta and Pme Land Seed. The effectiveness of transgemc plants for control of H armzgera and Helzcoverpa punctzgera under large-scale plantings 1s not fully known; however, midseason reports indicated that, m some field situations, msecttctdes were being used with the transgenic culttvars m order to provide more effective control of these insects. In Europe, Centre de Cooperatton Interventtonale en Recherche Agronomtque pour le Developpment (CIRAD)/Instttute National for Research m Agriculture (INRA) 1sdeveloping transgenic cotton plants that express the &endotoxm genes from Bt and also express protemase mhtbitors as a means of control of Helzothis arrnigera and Spodoptera lzttoralzs. It is hoped that plants with these two types of toxms will delay or prevent the development of resistant strain of insects (49). Genetics of resistance to the &endotoxm from Bt in tobacco budworm mdrcates that at least four genes are mvolved in resistance m a strain selected for resistance in the laboratory over >30 generations. This strain, YHD2, was developed by selection on diet containmg CrylA(c) for >30 generations, and 1s approx 1O,OOO-foldresistant (50). One gene m linkage group nine (51) contrtbuted over 80% of the resistance m this strain. 2.4. Transgenic

Cotton: The Future

There is a need for better control of bollworm on cotton than 1snow provided by the available 6-endotoxin gene constructs. Either different &endotoxm genes or entirely different genes are needed. Control of fall armyworm and beet armyworm 1s not sufficient with present F-endotoxin genes, Although control of tobacco budworm and pink bollworm are quite good wrth the present

Transgenic

Plants

Expressing

Bt Toxins

219

&endotoxin genes, additional &endotoxin genes or genes of a drfferent nature are needed for long-term resistance management The companies developing transgenic cotton technology operate m a global market. There IS a great likelihood that transgemc cotton culttvars ~111be utthzed m most major cotton-producing countries around the world. In all countries that produce cotton, there ISa need for control of leprdopteran msects,and thts technology provrdes very effective control. The prospect of global use of transgenic Bt cotton places greater emphaseson the need for resistance-managementstrategies. 3. Transgenic Corn 3.1. Hisfory and State of the Art Using microprojecttle bombardment of immature embryos, scienttsts at CIBA Brotechnology successfully placed a synthetic gene encodmg a truncated version of the CrylA(b) protein derived from Bt mto corn plants (52). Chmese screntrsts have also used mrcroprojecttle bombardment of matze cell suspensions, immature embryos, and embryogemc call1 to transform a gene from Bt into maize, and have regenerated plants, some of which expressed the toxin when evaluated agamst the corn borer, Ostrima jiirnacalls (53) Ovaries of maize inbred lures have also been transformed with the insectresistance cvylA gene from Bt (54’. This was accomplished by iqectmg ovarres on the ear 10-20 h after pollmatton with a lO--loo-pm drameter glass needle containing l-3 pL DNA solution. From 40 ovaries mjected on each of 2 16 ears, four plants carried reststanceto European corn borer m the next generation. From 12 independently transformed hnes of transgentc matze expressmg the CrylA(b) msectlctdal protein from Btk, scientists at Monsanto were able to show that eight had significantly less damage to ears, and three lines exhibited a 75% reduction in feeding by H. zea larvae, which were also stunted in growth. Concentratron of the CtylA(b) protein was 0.0-l .28 pg/g fresh wt of silks (55). Williams et al. (56) evaluated field plots of transgemc corn plants expressmg &endotoxin msectictdal proteins for southwestern corn borer, Dzatraea grandzosella Dyar, and fall armyworm, Spodoptera frugiperdu (J. E. Smnh). These transgenic hybrids offer a high level of resistance to fall armyworm, and near tmmumty to southwestern corn borer. These transgemc corn plants have the htghest levels of plant resistance documented to these two insects. Transgemc corn lines have also been evaluated in the field for resistance to the European corn borer where both first-generation leaf-feeding and second-generation stalk tunnelmg have been observed (.57,58). 3.2. Registration and Commercialization of Transgenic Corn A truncated version of a cryZA(b) gene expressed m ehte hybrrds of maize provided excellent protectton against European corn borer (0. nubzlalis) when

220

Jenkms

challenged with over 2000 larvae/plant m the field (57). These transformed maize plants were different from most other plants expressmg 6-endotoxm genes, because these maize plants had a maize-optimized gene plus a ttssuespecific promoter, which targeted the cry1 A(b) productton to plant ttssue relevant for control of European corn borer. These plants were first grown on a commercial scale for seed productton m 1996. In 1996, transgenic corn hybrids expressmg the cryld(b) gene were sold by Mycogen (NatureGard) and CIBA (Maximizer) to growers for control of European corn borer. Transgenic corn hybrids were available to growers on a large scale m 1997. These contain the crylri(b) gene and are available from several compames. The U.S. Environmental Protection Agency (EPA) has granted three registrations for transgenic corn, Mycogen event 176, Northrup King event BT 1I (YieldgardTM, and Monsanto event MON8 10 (Y ieldgard). Expression of the Mycogen event 176 is low m the silks and ear. Expression of the event NKBTl 1 and MON810 are such that all plant parts express the toxic protein as the CaMV35S promoter is used with the gene. EPA registration requirements for these corn hybrids mclude parts of a resistance-managementplan. Mycogen is required to monitor for changes in level of response of corn borers to &endotoxm, and to continue to collect data and report to EPA. In addition, a resistance-management plan that includes a refuge must be developed by the year 2000. EPA required these same elements as conditions of registration of the event NKBTl 1 from Northrup Kmg. There were, however, two additional requirements: The corn hybrids with this event could not be sold m the South, where cotton is grown, and data were required to be obtained, to better understand the relationship of resistance development in H zeu. These two additional requirements are related to the fact that H. zea feeds on both corn and cotton. The gene in the NKBTI 1 has the CaMV35S promoter and expressesin the silks and the ear, as well other plant parts. The event MON8 10 was m hybrids sold by Pioneer, Cargtll, and Golden Harvest m 1997. EPA required all the above elements for registration of MON8 10 (Yieldgard). In addttion, Monsanto was required to collect data to validate a resistance management model Monsanto also voluntarily placed one additional requirement on the use of its Yieldgard product A refuge must be maintained on each farm that grows the product. This applies to all companies that license the product and the growers to whom they sell hybrids. Thus, EPA placed a requtrement on all three transgemc corn genes that some elements of a resistance-management plan be developed and a plan with a refugia be in place by the year 2000. 3.3. The Future of Transgenic

Corn

Expectations are that acreage of transgenic corn will expand. Current research is underway to find genes other than &endotoxm genes to use for insect con-

Transgenic

Plants

Expressing

Bt Toxins

221

trol in corn. Rootworms and cutworms are major pests of corn that are somewhat recalcttrant to the presently used family of Bt 6-endotoxms. The Vtp proteins produced during the vegetattve phase of growth of certain Bacillus spp strains may prove to be useful for control of these pests. 4, Transgenic Potato and Eggplant A modified version of a 6-endotoxin gene, cry.%4from Bt var tenebrzonzs was used to transform potato plants, Solarium tuberosum L. (59). This conferred resistance to Colorado potato beetle, Leptinotarsa decemlineata Say, under high levels of natural field infestatron. This transformation was accomplished in Russet Burbank potato cultivar without any loss in agronomtc or quality charactertsttcs. Transgenic potatoes reststant to Colorado potato beetle were first grown commercially in the Umted States m 1995, and performed very well in the field. Transgenic potato plants containmg a redesigned cvy3A gene under the mfluence of the CaMV 35Ymannopme synthetase promoter show high levels of resistance to Colorado potato beetle. The level of insect control was highly correlated with the level of Ei-endotoxm RNA and protein (60). Plants transformed with the crylA(c) gene from Bt, strain HD-73, expressed a moderate level of leaf-feeding reststanceto the tobacco hornworm m laboratory tests (61). The potato tuber moth, Phthorminaea operculella Zeller, is a major pest of potato in the tropics and subtropics. Commercial lmes of potato transformed wtth a codon-modified cry5 gene from Bt have shown high levels of reststance, and other transgenic lmes expressing the wild-type cry1 gene have shown reduced leaf feeding by the potato tuber moth (62). The transgenic potato, NewleaF”, was the first transgemc crop plant wtth a gene from Bt to be registered by EPA. It expressesthe cry3A gene, with Colorado potato beetle as the target pest. Toxic Bt protein IS not widely used as a spray product on potatoes, because rt is only effective against early mstars of the insect. In contrast, the transgemc potato plant is considerably more effective for control of this pest. As a condition of registration, EPA requtred the developing company to continue to do research on resistance management, to monitor where the NewLeaf potatoes were planted, and to monitor for shafts m levels of susceptibility to the toxin. Data from these are to be reported to EPA. In additton to this, Monsanto voluntarily placed a requirement on growers that no more than 80% of then crop could be planted with the transgemc potato. The Colorado potato beetle feeds on several plants of the genus Solanum. Eggplant, Solarium melongena L., has been transformed wtth a synthetic cry3A gene from Bt via A. lumefaciens-mediated transformation. From 300 plants, 175 were confirmed to be transformed, and some exhibited htgh levels of resistance m the field when arttfictally infested with egg massesof Colorado potato

222

Jenkins

beetle. Progeny from resistant plants with a single msertion segregated m typical Mendehan fashion (63). 5. Transgenic Rice A modified cryiA(b) gene from Bt has been inserted into a rice, 0 satzva L, Japomca cultivar, and confers resistance to two maJor rice insects, rice leaffolder, Cnaphalocrosis medlnalils, and striped stem borer, Chllo suppressalzs (64) This shows that genes from Bt should be useful m developing insect-resistant culttvars for specific insects m rice. The gene m race is extensively modified, and is a truncated version of the gene, based on codon usage m known rice genes The gene was mtroduced mto embryogernc rice protoplasts by cotransformation with the hygromycm-resistant selectablemarker gene 6. Transgenic Tomato Insect-resistant transgemc tomato, Lycoperszcon spp, developed m 1987, IS tolerant to tobacco hornworm, tobacco budworm, and bollworm (65). This same research suggested the feasibility of genetically engineering msect-tolerant transgemc crops by expressing the insect control proteins from Bt, and opened the way for research in cotton and corn. A truncated version of the gene from the HD-I strain of the bacteria was more effective than the full-length version of the gene m expressing the desired resistance trait m tomato plants Field tests of genetically engmeered tomatoes expressing a gene from Bt were conducted m 1987 m Illmois Plants were allowed to produce flowers and seed, and to decompose mto the soil. Similar tests were carried out with tomatoes m Florida and California m 1988 (66). Tomato plants were transformed to expressthe &endotoxm gene from Bt subsp tenebrionzs by exposing leaf disks to A. tumefaczens.These transformed plants expressedan msecticidal protein of 74 kDa that was active againstColorado potato beetle (67) Scientists m China have reported transformation of tomato with a CMV-cp genefor resistanceto a vnus and with a Bt gene for resistanceto insects Although their chemical data showed transformation, they did not report any F 1 or F2 data on resistanceof the plants to vnus or insects(68). 7. Transgenic Soybeans Soybean, Glyclne max L., was transformed via A tumefaciens, and plants were regenerated that expressed the transgenes (69). These plants expressed either the GUS gene or glyphosate tolerance inherited m a 3: 1 Mendehan fashion. These results showed that stable transformation was possible in soybean Somatic embryos of the soybean cultivar Jack were transformed using microproJectile bombardment with a synthetic Bt crylA(c) gene with the

Transgenic

Plants

Expressing

Bt Toxins

223

35s promoter gene linked to the HPH gene (70). These plants exhtbtted varymg levels of resistance to corn earworm, soybean looper (Pseudoplusza includens), tobacco budworm, and velvetbean caterpillar (Antzcarszagemmatalis Hubner). The corn earworm and soybean looper are more tolerant of the &endotoxm, thus, the transgenic plants were less resistant to these two pests than to the other two, which are more susceptible to the &endotoxm 8. Transgenic Trees The prospects for genettc engineering of insect resistance in forest trees was reviewed by Strauss et al. m 1991 (71). They suggested that, m addition to the Bt genes, other potentral strategies could mclude proteinase-inhibitor genes, chttmase, lectms, and baculovtrus genes. The &endotoxm of Bt m the form of CaMV 35S-Bt was stably transformed by electric dtscharge partrcle acceleratton into Populus alba x Populus grandldentata Crandon and Populus nlgra Betulifoha x Populus trzchocarpa hybrids. Transformed plants were highly resistant to feedmg by the forest tent caterpillar, Malacosonza disstria Hubner, and the gypsy moth, Lymantrza dlspar L. (72). Hybrid Populus plants (clone NC 5339), genetically engineered wtth a crylA(a) 6endotoxm gene, showed field resistance to forest tent caterpillar and gypsy moth, m the form of reduced feeding and weight gain; however, mortahty of late third-mstar larvae of gypsy moth drd not differ when fed on transgemc and control foliage (73). The gypsy moth IS a major pest of many forest trees around the world. Poplar, P nzgra L., trees have been genetically engineered to resist thts pest III China by transforming plants with A tumefaczensstrains carrymg a truncated gene from Bt driven by a CaMV35S promoter. Three transgenic clones were selected for resistance to gypsy moth and Apochemia czneraius, reduced morphologtcal changes, and promtsmg stlvrculture traits. These are under largescale field evaluation in SIXprovinces m China (74). Plants of poplar P alba x P grandidentata cv Crandon have been transformed to contain a truncated gene from Bt, plus the marze gene AC. Transgemc plants expressing AC and callus contammg the Bt gene were recovered (75). Transgemc plants contammg a modified Bt gene have been produced by transformatton and regeneration of excusedleaves of poplar hybrtd 741 (76). Populus deltoides plants were transformed using A. tumefaclens LBA 4404 strains contammg a gene from Bt, and two of three plants regenerated successfully integrated the Bt gene (77). The transfer and expression of the Bt toxin gene via A. rhz’zogenes-medrated transfer has been documented usmg Southern, Northern, and Western blots of needle tissue from transgemc plants of European larch trees, Larix decldua (78,79).

224

Jenkins

9. EPA Registration and Resistance Management Several msecticldal Bt products have been used for over 30 yr as sprays for insect control. At one ttme, It was thought that use of these spray products would probably not select for resistant populations of insects. However, with their continued use on cabbage, resistant strains of dlamondback moth, Plutella xylostella L , have developed ($0). Strains of Indian meal moth, Plodra znterpunctella Hubner, resistant to the &endotoxms from Bt, have also been reported (Sl), Most transgemc plants developed to express the Gendotoxms have a constltutlve promoter. This causes the plant to express the toxic protein contmuously, and in most or all the parts of the plant. This has the potential to place a different level of selection on the pest population than pesticides apphed as sprays. This has raised the issue of development of reslstant populations of pest species. EPA has proposed the posltion that transgemc plants may produce a plant pesticide, and, if so, EPA has the authority to regulate and register these transgemc plants under the Federal Insectlclde, Fungicide, and Rodentlclde Act (FIFRA) (EPA proposed pohcy and rule announced November 23, 1994) (82). A concern that this 1snot the correct approach has been expressed by 11 scientific socletles with approx 80,000 members (83). Theprimaryconcernwith theproposedrule ISthecreationof anewcategoryof pestlclde,called‘plantpestlclde’,solelyfor thepurposeof regulationunderexlstmgstatutes EPA proposes to designate as ‘plant pestmdes all substancesresponsible for pest remtance In plants, as well as the genes needed for production of these substances Under Its proposed pohcy, however, EPA singles out for possible reglstratlon as ‘plant pestmdes only those traits Introduced Into plants usmg rDNA techniques (83)

EPA has placed different requirements on the three transgemc, insect-reslstant field crops (cotton, corn, and potato) approved for commercial use. This variation m requirements across crops has a basis m the host range of target Insects, acreage projected to be planted to the transgemc cultlvars, and the public risk perceived to be associated with development of resistance. There are some similarities m all requirements, as well as specific dlfferences required to be implemented m each crop. A major strategy for long-term use of conventional plant resistance has always Included multiple genes and new types of genes to breed mto the crop when Insect populations become resistant to the currently used genes. The companies involved with transgemc resistance are also actively working to develop addItiona toxic protein genes with different modes of action. These transgemc cultivars and hybrids should not be thought of as standalone products. They should be used as the foundation on which to build good IPM and crop management practices, Resistance management IS m Its infancy,

Transgenic

Plants

Expressing

Bt Toxins

225

and sctenttsts do not know how to best manage or slow the development of resistant strams of insects. There is a grand opportunity to learn much with the commercialization of three maJor transgemc crops. In the near future, msectresistance genes should be m crops that are not related to the &endotoxms. The use of two or more genes with different modes of action should be a maJor tool for sustaimng the vtabrhty of transgemc insect-resistant cultivars and hybrids. The use of IPM and crop management strategies will Improve as experience is gained with the cultivation of transgenic culttvars and hybrids as replacements for organic insecticides in pest control. The future looks bright, yet there is much to learn m this uncharted adventure to use biotechnology as a useful tool m plant breeding and in how to control pests on crop plants. 10. Future of Bt Transgenic Crops Movmg mto the future with plant biotechnology, genetically engineered plants that resist msects will use novel insecticidal principles to target tmportant insect pests that have escaped Bt technology (84). The arena of genetically engineered plants to control pests is very attractive. Costs associated with pestmanagement practices and chemical control of Insects approaches $10 billion annually (84). Even with this expenditure, these same authors estimate that 20-30% of global production 1s lost to insects. Advances m transformation, tissue culture, and expression of foretgn genes m plants have so improved that the potenttal exists to vtrtually transform any crop with a crop-tailored gene. The msecticidal genes from Bt led the way, and now additional genes from this bacterium, as well as novel genes from other genera, will be used to provide improved cultivars of plants that resist insects. In addition to the &endotoxms, a second class of protems effective against certain insects, some of which are not greatly affected by the cry genes, are the Vip insecticidal proteins produced by Bt during the vegetative growth of the bacteria. These proteins are dtstmct from the &endotoxm proteins and afford acute btoacttvtty in the range of ng/mL of diet for susceptible insects (8586). These genes are available for use in genetically engineered plants, Instead of, or in conjunction with, Gendotoxin genes, The clarified culture supernatant fluids collected durmg vegetative growth of other Bacillus species are also a rich source of insecticidal-activity Vip protems (4,85). These Vip proteins are a class of msecticidal protems dtstmctly different from the delta endotoxins. V1p111A from Bt show acute bioactivity m the nanogram range against a wide spectrum of lepidopteran insects, particularly black cutworm, Agrotis zpszlon(Hufnagel), fall armyworm, Spodoptera fruglperda (J. E. Smith), and beet armyworm, Spodoptera exzgua (Hubner) (85). This makes these proteins promismg candidates for use in conjunctton with the &endotoxins.

Jenkms

226

References 1, Stewart, J McD. (199 1) Brotechnology of cotton ICAC Revrew article on cotton production research no 3. C A.B International, UK 2. Hofte, H and Whtteley, H R. (1989) Insectrctdal crystal proteins of BaczlEus thurzngzenszs Mzcrobzol

Rev 242-25.5

3. Estruch, J J , Carozzl, N. B , Desai, N , Duck, N B , Warren, G. W , and Koztel, M G (1997) Transgemc plants an emergmg approach to pest control Nature Bzotechnol

15, 137-14 1

4 Schnepf, H E and Whtteley, H R (1981) Clonmg and expression of the Baczlfus thurzngzenszs crystal protem gene in Escherzcza colz Proc Nat1 Acad Scz USA 78,2893-2897.

5 Gasser, C and Fraley, R T (1989) Genettcally engineered plants for crop improvement Sczence 244, 1293-I 299. 6 Fraley, R (1992) Sustammg the food supply Bzo/Technology 10,40-43 7 Oerke, E C (1994) Esttmated crop losses due to pathogens, ammal pests and weeds, m Crop Productzon and Crop Protectzon Estzmated Losses In MaJor Food and Cash Crops (Oerke, E C , Dehne, H. W., Schonback, F , and Weber, A, eds ), Elsevier, Amsterdam, pp 72-78 8 Barton, K A , Whlteley, H R , and Nmg-Sun, Y (1987) Baczffus thurzngzenszs delta endotoxm expressed m transgemc Nzcotiana tabacum provides reststance to leptdopteran Insects Plant Physzol 85, 1103-I 109 9 Vaeck M., Reynaerts, A , Hofte, H , Jansens, S , De Beukeleer, M , et al (1987) Transgenic plants protected from Insect attack Nazure 328,33-37. 10 Klem, T M , Wolf, E. D , Wu, R , and Sanford, J C (1987) High velocity microproJecttles for delrvermg nuclerc acids into ltvmg cells Nature 327, 70-73 11 Chrtstou, P D , McCabe, E , Martmell, B J , and Swam, B J. (1990) Soybean genetrc engmeermg-commercial productron of transgemc plants Trends Bzotechnot 8, 145-151 12 Klein, T M , Arentzen, R , Lewis, P A., and Fttzpatrtck-McEllrgott. S. (1992) Transformation of microbes, plants, and animals by particle bombardment. Bzo/ Technology 10,286291.

13 Fmer, J J and McMullen, M D (1990) Transformatton of cotton (Gossypltlm hzrsutum L ) vta parttcle bombardment. Plant Cell Rep 8, 586-589 14 Tortyama, K , Arimoto, Y., Uchrmtya, H., and Hmata, K (1988) Transgemc rice plants after direct gene transfer mto protoplasts Bzo/Technology 6, 1072-1074 15 Rhodes, C. A , Pierce, D. A , Mettler, I J , Mascarenhas, D , and Detmer, J J (1988) Genetically transformed maize plants from protoplasts Sczence 240,204-207 16 Williams, S , Frtedrich, L., Dincher, S., Carozzr, N , Kessman, H , Ward, E , and Ryals, J (1992) Chemical regulation of Baczllus thurzngzenszs delta endotoxm expresston m transgenic plants. Bzo/Technology 10,541-543 17 Trolmder, N L. and Gooden, J R (1987) Somatic embryogenesis and plant regeneratton m cotton (Gossypzum hzrsutum L). Plant Cell Rep 6, 23 l-234 18. Umbeck, P , Johnson, G , Barton, K , and Swam, W ( 1987) Genettcally transformed cotton (Gossypzum hzrsutum L.) plants Bzo/Technology 5,263-266

Transgenic

Plants

Expressing

Bt Toxins

227

19 Jenkins, J N , Parrott, W. L , McCarty, J. C., Jr., Barton, K. A., and Umbeck, P. F (199 1) Field test of transgemc cotton contammg a Bacdlus thurzngzenszs gene. MLYS Agrx and For Exp. Sta Tech Bull 174, 14. 20 Perlak, Fredrlck J., Deaton, W R , Armstrong, T. A., Fuchs, R. L., Sims, S. R , Greenplate, J T., and Fischoff, D A (1990) Insect reslstant cotton plants. Blo/Technology 8,939-942 21 Perlak, F. J., Fuchs, R. L , Dean, D. A , McPherson, S. L , and Flschoff, D. A (1991) Modlficatlon of the coding sequence enhances plant expression of insect control protein genes Proc Nat1 Acad. Scz USA 88,3324--3328. 22 Benedict, J H , Altman, D. W., Sachs, E. S , Deaton, W. R , and Ring, D R. (1991) Field performance of cotton genetically modified to express msectlcldal protein from Bacdlus thurrnglenszs, in Proceedings Beltwzde Cotton Production Research Conference, San Antonio, TX(Brown, J , ed ), National Cotton Council of America, Memphis, TN, p 577. 23 Deaton, W. R (1991) Field performance of cotton genetlcally modified to express insecticidal protein from Baczllus thurrnglensu, mtroductlon, m Proceedzngs Beltwlde Cotton Production Research Conference, San Antonio, TX (Brown, J , ed.), Natlonal Cotton Council of America, Memphis, TN, p. 576 24 Gannaway, J R , Rummel, D R., and Owen, D F (1991) Field performance of cotton genetically modified to express insectlcldal protein from Baaflus thurznglensu, in Proceedings Beltwide Cotton Productron Research Conference, San Antonlo, TX(Brown, J , ed.), Natlonal Cotton Council of America, Memphis, TN, p 578 25 Jenkins, J N., Parrott, W L., and McCarty, J C., Jr. (1991) Field performance of transgemc cotton containing the Bt gene, m Proceedmgs Beltwzde Cotton Productzon Research Conference, San Antonro, TX(Brown, J , ed ), National Cotton Councrl of America, Memphis, TN, p 576. 26 Micmskl S and Caldwell, D W. (1991) Field performance of cotton genetically modified to express msectlcldal protein from Bacdlus thurlnglensls, m Proceedlngs Beltwlde Cotton Production Research Conference, San Antonio, TX (Brown, J , ed.), National Cotton Council of America, Memphis, TN, p. 578. 27 Micmskl, S., Caldwell, W. D., Fltzpatnck, B. J., and Griffin, R. C. (1992) First Loulslana field trial of insect-resistant transgenic cotton LA Agriculture 35, 8-l 0 28. Jenkms, J. N (1993) Use of Bacillus thuringzensis genes in transgemc cotton to control Lepldopterous insects, in Pest Control wzth Enhanced Environmental Safety (Duke, S 0 , Menn, J J , and Plimmer, J R , eds.), American Chemical Society Symposmm Series 524, Washington, DC, pp. 267-280. 29 Jenkins, J N (1995) Host plant resrstance to insects m cotton, m Proceedzngs of World Cotton Research Conference I Challengrng the Future (Constable, G A and Forrester, N W., eds ), Brisbane, AU, Feb 14-17, 1994, pp 359-372. 30. Benedict, J H , Sachs, E S , Altman, D. W , Rmg, D. R., Stone, T B., and Sims, S. R (1993) Impact of delta endotoxm producing transgemc cotton on msectplant mteractlons with Helzothts vzrescens and Hellcoverpa zea (Lepldoptera Noctuidae). Environ Entomol. 22, l-9 31. Wilson, F. D and Flmt, H M (1991) Field performance of cotton genetically modified to express msectlcldal protein from Bacillus thunnglenszs, m Proceed-

228

32

33 34

35

36.

37

38.

39

40

41 42

43

Jenkins zngs Beltwlde Cotton Production Research Conference, San Antonzo, TX(Brown, J., ed.), National Cotton Council of America, Memphis, TN, p 579. Williamson, D R. and Deaton, W R. (1991) Field performance of cotton genetltally modified to express msectlcldal protem from Bacillus thunngzensu, m Proceedmgs Beltwtde Cotton Productton Research Conference, San Antonlo, TX (Brown, J , ed ), National Cotton Council of America, Memphis, TN, p 577 Trolmder, N L. and Xhixlan, C (1989) Genotype specificity of the somatic embryogenesls response m cotton Plant Cell Rep 6,23 l-234 Jenkins, J N , McCarty, J. C , Jr, Buehler, R E., Klser, J., Willlams, C , and Wofford, T (1997) Resistance of cotton with delta endotoxm genes from Bacdlus thurrngrenszs berlmer kurstakl on selected lepldoptera Insects. Agronomy J 89,768-780 Umbeck, P. G , Barton, K. A , Nordhelm, E V , McCarty, J C , Jr, Parrott, W L , and Jenkins, J N (1991) Degree of pollen dispersal by Insects from a field test of genetically engmeered cotton J Econ Entomol 84, 1943-1950 Jenkins, J. N , Parrott, W. L , McCarty, J C , Jr, Callahan, F E , Berbench, S A , and Deaton, W R (1993) Growth and survival of Hellothzs vlrescens (Lepldoptera* Noctuidae) on transgemc cotton containing a truncated form of the delta endotoxin gene from Bacillus thunnglensrs. J Econ Entomol 86, 18 l-l 85 Jenkms, J N , McCarty, J. C., Jr, and Wofford, T (1995) Bt cotton, a new era m cotton production, m Proceedings of Beltwrde Cotton Conferences, Blotechnology Workshop (Herber, D J. and Richter, D A , eds ), National Cotton Council of America, Memphis, TN, pp. 17 1-173 Durant, J. A (1994) Evaluation of treatment thresholds for control of bollworms and tobacco budworms in transgemc Bt cotton m South Carolina, m Proceedzngs ofBeEtwlde Cotton Conferences (Herber, D. J. and Richter, D. A., eds ), National Cotton Council of America, Memphis, TN, pp. 1073-1075 Mahaffey, J S , Bacheler, J. B , Bradley, J. R., and Van Duyn, J. W. (1994) Performance of Monsanto’s transgemc B T. cotton against high populations of lepldopterous pests m North Carolina, in Beltwlde Cotton Production Conference, San Antonro, TX(Herber, D. J and Richter, D. A , eds ), National Cotton Council of America, Memphis, TN, pp. 106 l-l 063. Fischoff, D. A. (1992) Management of Lepldopteran pests with insect reslstant cotton recommended approaches, in Proceedings Beltwlde Cotton Production Research Conference, San Antonlo, TX(Brown, J., ed.), Natlonal Cotton Council of America, Memphis, TN, pp 75 l-753. Gould, F. (1991) Arthropod behavior and the efficacy of plant protectants. Ann Rev Entomol 36,305-330. Forrester, N W , Hokkanen, H M T , and Deacon, J , eds. (1994) Proceedings of an OECD workshop on ecologlcal lmplicatlon of transgemc crops contammg Bt toxin genes, Jan. (1994) New Zealand Biocontrol Scl Technol 4,549-553 Davis, M K , Layton, M B., Vamer, J D., and Little, G (1995) Field evaluation of Bt transgemc cotton in the Mississippi Delta, in Proceedmgs of Beltwlde Cotton Conferences (Herber, D. J and Richter, D. A , eds.), National Cotton Council of America, Memphis, TN, pp 77 l-775

Transgemc Plants Expressing Bt Toxins

229

44. DuRant, J A. (1995) Efficacy of selected seed mtxes of transgenic Bt and nontransgenic cotton against bollworms and tobacco budworms m South Carolina, in Proceedings of Beltwtde Cotton Conferences (Herber, D. J. and Richter, D A., eds.), Nattonal Cotton Council of America, Memphis, TN, ppq 769-77 1 45. Rummel, D R , Arnold, M. D., Gannaway, J. R., Owen, D F., Carroll, S C., and Deaton, W. R (1994) Evaluatton of Bt cottons resistant to mJury from bollworm* imphcations for pest management m the Texas southern htgh plams Southwest Entomol 19, 199-207. 46. Luttrell, R G. and Herzog, G A. (1994) Potential effects of transgemc cotton expressing BT on cotton IMP programs, m Proceedings of Beltwtde Cotton Conferences (Herber, D I. and Richter, D. A., eds.), Nattonal Cotton Councti of Amertca, Memphts, TN, pp. 806-809. 47. Fttt, G P , Mares, C. L , and Llewellyn, D. J (1994) Fteld evaluation and potential ecologtcal impact of transgemc cotton (Gossypzum hrrsutum) in Australia, m Proceedings of an OECD Workshop on Ecologtcal Impltcattons of Transgenic Crops Contatntng Bt Toxtn Genes. New Zealand. 48. Fttt, G. P. and Jones, D. D. (1994) Field evaluation of transgenic cotton in Australia environmental constderations and consequences of expanding trial size. The biosafety results of field tests of genetically modified plants and microorgamsms Proceedtngs of the 3rd lnternattonal Symposium, Monterey, CA. 49 Pannetier, C., Guiderdom, E., and Hau, B. (1995) Genetic engmeermg and improvement of rice and cotton Agrtcultural Dev. 6, 16-27. 50 Gould, F., Anderson, A., Reynolds, A., Bumgarner, L., and Moar, W. (1995) Selectton and genettc analysis of a Heliothis vu-escens (Lepidoptera: Noctmdae) stram with high levels of resistance to Bacillus thurmgzenszs toxins. J Econ. Entomol 88, 1545-1559. 5 1 Heckel, D. G , Gahan, L C , Gould, F., and Anderson, A (1997) Identtfication of a linkage group with a major effect on resistance to Bacillus thurtngtensts Cry 1A(c) endotoxin in the tobacco budworm. J Econ Entomol 90,76-86. 52. Hill, M., Launis, K., Bowman, C., McPherson, K , Dawson, J., Watkms, J , et al. (1995) Btohstic mtroduction of a synthetic Bt gene mto ehte maize. Eucarpia Genetic Manipulation in Plant Breeding section meeting, Cork, Irish Republic, Euphytica 85, 119-l 23, 53 Wang, G Y , Du, T B., Zhang, H., Xie, Y. J., Dal, J. R , Mi, J. J., et al. (1995) Transfer of BT toxin protem gene mto maize by htgh velocity microproJectile bombardments and regeneration of transgenic plants. Sczence zn China Series B. Chemt., Lfi Set Earth Set. 38(7), 8 17-824. 54 Dmg, Q. X , Xie, Y. J., Dai, J. R., Mi, J. J , Li,T. Y , Tian, Y. C., et al. (1994) Introducmg Bt gene into maize with ovary InJection. Sczence tn Chrna Series B Chemt , Life Set. Earth Sci, 37(5), 563-572 55 Sims, S. R , Pershing, J C , and Reich, B J. (1996) Field evaluatton of transgemc corn contammg a Baczllus thurzngzensts Berliner msectictdal protein gene agamst Heltcoverpa zea (Leptdoptera Nocturdae). J Econ. Entomol. 31,340-346. 56. Willhams, W. P., Sagers, J B., Hanten, J. A., Davis, F. M , and Buckley, P M. (1997) Evaluation of transgenic corn for resistance to fall armyworm and southwestern corn borer. Crop Scr 37,957-962

230

Jenkms

57 Kozlel, M. G , Beland, G. L , Bowman, C , Carozzl, N B , Crenshaw, R , Crossland, L , et al (1993) Field performance and elite transgemc maize plants expressing an msectlcldal protem derived from Baczllus thurzngzenszs Blo/Technology 11, 19&200 58 Armstrong, C L , Parker, G B , Pershmg, J C , Brown, S M , Sanders, P R , Duncan, D R , et al (1995) Field evaluation of European corn borer control m progeny of 173 transgemc corn events expressing an insecticidal protein from Bacillus thurlnglensls Crop SCI 35, 550-557 59 Perlak, F J , Stone, T. B , Muskopf, Y M , Peterson, L J , Parker, G B , McPherson, S A , et al. (1993) Genetically improved potatoes. protection from damage by Colorado potato beetles Plant Mol Blol 22, 3 13-32 1 60 Adang, M J , Brody, M S , Cardmeau, G., Egan, N., Rousch, R T , Shewmaker, C K , et al (1993) The reconstructlon and expression of a Baczllus thuruzg~ensrs cry3 A gene m protoplasts and potato plants Plant Mol Biol 21, 1 13 l-l 145 6 1 Cheng, J , Bolyard, M G , Saxena, R. C , and Stncklen, M. B (1992) ProductIon of insect resistant potato by genetic transformation with a delta endotoxm gene from Bacillus thurwgzensu var kustakz Plant Scz Llmerlck 81, 83-91 62 Douches, D S , Llswldowatl, A., Hadl-Permadl, W P , Hudy, P , Westedt, A , and Grafus, E (1996) Progress m development and evaluation of BT transgemc potatoes with resistance to potato tuber moth (Phthormlnaea operculella Abstracts of The 80th Annual Meeting of the Potato Association of America, Annual Meetmg, Idaho Falls, ID, August 1 l-l 5, 1996 Am Potato J 73,8 63 Jelenkovq G , Bllhngs, S , Chen, Q , Lashomb, J , and Ghldu, G (1996) Englneermg transgemc eggplant (So/unum melogena L ) resistant to Colorado potato beetle (Leptuzotarsa decemllneta Say) HortSczence 31, 572. 64 Fujlmoto, H , Itoh, K , Yamamoto, M., Kyozuka, J , and Shlmamoto, K (1996) Insect resistant rice generated by mtroductton of a modified delta endotoxm gene of Bacrllus thunngzensu Blo/Technology 11, 115 1-l 155. 65 Flschoff, D A , Bowdish, K. S , Perlak, F. J , Marrone, P. G , McCormick, S M , Nledermeyer, J G , et al (1987) Insect tolerant transgemc tomato plants Bzo/Technology 5,808-8 13 66 Muench, S R (1990) Field release of genetically engineered plants m 1987 and 1988 Risk assessment in agricultural biology Proceedings of the International Conference, pp 97-10 1 67 Rhlm, S L , Cho, H J , Kim, B D., Schnetter, W , and Gelder, K (1995) Development of insect resistance m tomato plant expressing the delta endotoxm gene of Bacillus thurlngzenszs subsp tenebrlonls Mol Breeding 1, 229-236 68 Llang, X.-Y , JIU, M J , Zhu, Y. X , and Chen, X L (1994) Construction of plant expression vector with double resistance to vnus and insect and ldentlficatlon of transformation m tomato Acta Botanzcu Sinzca 36, 849-854 69 Hmchee, M. A. W., Conner-Ward, V. D., Newll, C. A , McDonnell, R. E , Sato, S J , Gasser, C S , et al (1988) Productlon of transgemc soybean plants using agrobacterlum mediated DNA transfer Blo/Technology 6, 9 15-922 70 Stewart, C N , Jr., Adang, M. J., All, J. N., Boerma, H. R , Cardmeau, G., Tucker, D., and Parrott, W A (1996) Genetic transformation, recovery, and charactenza-

Transgenic

71

72

73

74.

75

76

77 78

79.

Plants

Expressing

Bt Toxins

231

non of fertile soybean transgemc for a synthettc Baczllus thurzngrenszs cry1 Ac gene Plant Physzol 112, 12 l-l 29. Strauss, S H , Howe, G T , Goldfarb, B., Neale, D. B , and Kmlaw, C S (1991) Prospects for genetic engineermg of msect resistance in forest trees Forest btotechnology Forest Ecol Manage 43, 18 l-209. McCown, B. H., McCabe, D. E., Russell, D. R., Robmson, D J , Barton, K A, and Raffa, K F (199 1) Stable transformatton of Populus and mcorporatton of pest resistance by electrtcal discharge particle acceleratton. Plant Cell Rep 9, 590-594. Klemer, K W., Elhs, D. D., McCown, B H , and Raffa, K. F (1995) Field evaluatton of transgemc poplar expressmg a Baczllus thurzngzensls crylA(a) d-endotoxin gene against forest tent caterpillar (Leptdoptera, Lastocamptdae) and gypsy moth (Leptdoptera, Lymantrudae) followmg winter dormancy Environ Entomol 24, 1358-1364 Wang, G., Casttghone, S , Chen, Y., Lt, L., Han, Y , Ttan, Y , et al (1996) Poplar (Populus nzgra L ) plants transformed wtth a Baczllus thurmgzensls toxin gene msectictdal acttvtty and genomtc analysts Transgenic Res 5,289-301. Howe, Cl T , Goldfarb, B , and Strauss, S. H (1994) Agrobactertum mediated transformation of hybrrd poplar suspension cultures and regeneration of transformed plants Plant Cell Tzssue Organ Culture 36, 59-7 1 Zheng, J B., Zhang, Y M , Yang, W. Z., Pet, D T., Tran, Y C., and Mang, K Q. (1995) Plant regeneration from excrsed leaves of poplar hybrtd 74 1, and transformatton wtth Insect reststant B t. toxm gene J Hebez Agrzcultural Unw. l&20-25 Chen-Ymg, Han-YtFan, Ttan, Y.-C., Li, L , and Nte, S J (1995) Study on plant regeneration from Populus deltotdes explants transformed with Bt toxm gene Sclentla-Szlvae-Smxae 31,2,97-103 Karnosky, D. F., Shin, D I , Huang, Y H., Podtla, G K , Huang, Y H , Schmidt, W. C., and McDonald, K. J (eds ) 1995 Transfer and expresston of forergn genes m Lartx Opportumtres for genetic Improvement Ecology and management of Larix forests, a look ahead Proceedmgs of an Internattonal sympostum, Whttefish, Montana, USA. October 5-9, No INT-GRT-3 19, pp. 405-407. Shin, D I., Podtla, G K., Huang, Y., Karnosky, D F., and Huang, Y H (1994) Transgetuc larch expressing genes for herbrcide and insect resistance. Can J Forest Res 24,2059-2067

80 Tabashmk, B E., Cushmg, N L , Fmson, N , and Johnson, M. W. (1990) Fteld development of reststance to Bacillus thurzngzenszs m Dtamondback moth (Leptdoptera Plutelhdae) J Econ Entomol 83, 1671-1676. 8 1 McGaughey, W. H and Beeman, R. W (1988) Resistance to BaczZlus thurmgzenszs m colonies of Indtanmeal moth and almond moth (Leptdoptera Pyrahdae). J Econ Entomol 81,28-33 82 US Environmental Protectton Agency (1994) Plant-Pesticides SubJect to the Federal Insecttctde, Fungicide, and Rodenttctde Act and the Federal Food, Drug, and Cosmetic Act Federal Regzster 59,60,496-60,5 18 83 Cook, R. J and Qualset, C. 0 (eds ) (1996) Appropriate overstght for plants with mhertted traus for reststance to pests A report from 11 professtonal sctentific

232

Jenkins

societies. Coordmatmg society. Institute of Food Technologists, 22 1 North LaSalle St , Suite 300, Chicago, IL 60601-1291 84. Estruch, J. J., Carozzi, N B., Desai, N , Duck, N B., Warren, G. W., and Koziel, M G (1997) Transgemc plants, an emerging approach to pest control Nature Blotechnol 15, 137-141. 85. Estruch, J J , Warren, G W., Mullms, M A., Nye, G. J , Craig, J A , and Koziel, M G. (1996) V1p3A a novel Bacdlus thurznglenszs vegetative msecticidal protem with a wide spectrum of activities against lepidopteran insects Proc Nat1 Acad Scl USA 93,5389-5394

86 Warren, G W , Koztel, M. G , Mullms, M A , Nye, G J , Carr, B., Desai, N , et al. (1996) Novel pesticidial proteins and strains. World Intellectual Property Organization 96, 10,083.

14 Production, Delivery, and Use of Mycoinsecticides for Control of Insect Pests on Field Crops Stephen P. Wraight and Raymond I. Carruthers 1. Introduction In 1985, Pierre Ferron reviewed the status of a “hundred-year-old hypothesis” that the entomopathogemc fungi would one day become integral components of many insect-pest management systems. He concluded his review by stating the need to “define precisely the ecosystems in which these natural enemies effectively play a positive role, and to determine the treatment strategies necessary to the expression of their potentialities, m order to have available reliable methods suitable with other phytosamtary techmques, accordmg to the integrated protection concept” (I), The hundred-year-old hypothesis IS now more than a decade older. Our understanding of the modes of action and environmental requirements of these agents has increased, and many new strategies have been developed, and old ones elaborated, to better exploit the potential of fungal pathogens. Effective use of fungi for control of a variety of pests has been implemented on many local and regional scales. Commercial use of A4etarhizzum anisopliae in Brazil and Beauveria bassiana in China and Eastern Europe has been well documented (2-Q. Nevertheless, the widespread integration of these agents mto commercial pest control systems has remained, it would seem, ever on the horizon. Reliable mycoinsecticide products remain largely unavailable to private growers, especially in high-technology field-crop production systems. Recent years, however, have seen a dramatic increase in commercialization efforts worldwide. This increase has been stimulated not only by the persistent problem of pesticide resistance and the growing economic and environmental costs of synthetic chemical insecticide use, but by a number of specific technological advances that have the potential to greatly expand the commercial From Methods /n &otechno/ogy, vol 5 Bropestrodes Use and De//very Echted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

233

234

Wrarght

and Carruthers

feastbthty of mycomsecttctde use. This chapter ~111discuss a number of these advances in the areas of fungus productton, formulatton, and application. Wrthm the scheme of btologtcal control, use of fungal pathogens falls under all three of the broad strategies defined m DeBach (7): rmportatton, augmentation, and conservatron. Use of fungal pathogens m each of these strategies has been the subject of many recent reviews (2,4-6,8-12). This paper will focus on the mundatrve augmentation strategy and development of mycomsecttcrde products for mtcrobtal control applications against insect pests of field crops 2. Recent Advances in Development of Entomopathogenic Fungi 2.1. Development of Mass-Production Technologies This topic, though seemingly out of place in a volume on delivery and use of brorattonal control agents, IS a crmcal one m the dtscusston of the current state of the art of mtcrobtal control with fungal pathogens. The long life of the hundred-year hypothesis 1s ulttmately attrtbutable to dtfficulttes wtth field efficacy, which IS highly dependent on an affordable applrcatton rate Many years of field studies of hyphomycete fungi indicate that high rates (on the order of 1Ot3-1Ot4 spores/ha) are often required to provide acceptable levels of control (5,12). This requirement derives from a complex of factors, but one of the most Important relates to the low regression coefficients (slopes) inherently assoctated with fungal dose-host mortahty responses (13) These coeffictents typrtally vary between 0.5 and 1S. An LCSOof approx 100 spores/mm* has been reported for B bassiana against at least two key agrtcultural pests (14,15). Given a typical regression coefficrent of 1.0, the associated LC& IS nearly 4400 spores/mm*. In terms of a umform dtstrtbutton of spores applied to a planar surface, this represents a dose of 4.4 x lOI3 spores/ha. From this perspective (even without factormg m application inefficiencies or three-drmenstonal crop canopy surface areas), It is obvious that many problems of efficacy are crttrcally and inextricably linked to productron economtcs. Thts sttuatton will remam without remedy for at least some time mto the future, when tt will become possible to genetically engineer strains with greatly enhanced infecttvrty and vtrulence (16) Regarding spore-based mycoinsecticide products, efficient productton technologies exist only for select strains of a few pathogen species. Greatest production efficiencies have been achieved wtth B ~~SSZQY~U, at least m part because of the small size of the conidia relative to those of other entomopathogenic fungi. Even with this species, however, the commercial-scale production capacity necessary to support multtple applications to field crops at the high rate of 1Or3spores/ha, at costscompetrtive with synthetic chemical msecticrdes, has long represented a major production barrter. For many years, the standard technology for mass productton of M amsopliae and B bassiana

Mycoinsecticides

and Field-Crop

Pests

235

throughout South and Central Amertca, Europe, and Asta has employed a substrate of cooked race or other grams on trays or in autoclavable plastic bags or glass Jars. Average yields of approx l-5 x IO9 spores/g (dry wt) of substrate are achrevable with selected fungal strains (3,17-20), but may be substanttally less (21). This technology IS generally adequate to support application rates of l-5 x lOI* spores/ha (22,22). Labor costs and decreasing efficiency wtth increases in scale are srgmficant constramts to continued use of these low-technology production systems.Seekmg an alternative, researchers in the Soviet Union pioneered mass production m submerged culture, using conventional fermentation equipment. Operatronal scale productton systems were developed for B basszana (23), however, no significant gains m efficiency were realized. Ltqurd fermentation supported applicattons of 4 x lOI* spores/ha. Use of the fungus at higher rates was not economtcally feasible, according to Lipa (24). Many spectes of entomopathogenic fungi can be produced m submerged culture. Under these condtttons, as in the host hemolymph, thin-walled hyphae or hyphal bodies are normally produced; hyphomycetes typically produce ovalshaped, single-celled hyphal bodies termed “blastospores.” Development of liquid fermentatton technologies for entomopathogenic fungi remams one of the most active areas of mycoinsectictde research, especially because comdra of many pathogens cannot be eftictently produced on solid substrates Much progress has been made m recent years on the productron and stabthzatton of hyphal body and mycelmm formulattons of various entomophthoralean and hyphomycete pathogens (25-30) However, the commerctal-scale productton/stabrhzation systemsdeveloped thus far are not compettttve with extstmg technologres for comdia (see followmg discussions) Durrng the past decade, major advances have been made in the solid substrate culture of hyphomycete species.Ptoneenng technologies developed tn China and Eastern Europe for mechanized production of conrdia of B. bassiana were extensively reviewed and descrtbedby Feng et al. (6). In Chma, diphastc surface-culture systemsare clanned to produce nearly pure conidial powders of B. basszana suffictent to treatmany hundredsof thousandsof hectaresannually. Cost of an application of 1.53 x 1Ot3conidta to 1ha of forest for control of pine caterprllars is reported to be only approx $2-3, depending on the formulation (32). This suggestsa remarkable level ofefficiency, but the extentof the economicanalystsanddependenceon unique Chinese labor and market condttrons ISunclear. Researchersrn Czechoslovaktaand Canadahave alsodeveloped massculture systemstn whtch conrdra areproduced at a liquid-an or solid-air interface (32,33) Thesesystemsarehighly efficient in productrig spores/kgdry wt of nutrients (32) however, surface culture 1stneffictent tn producing spores/Uof fermenter volume. Free-marketeconomtcanalysesof the Chinese and Czechoslovaktanproductton systemsare needed

236

Wralght and Carruthers

Fully automated, commerctal-scale productton of aerral comdia of entomopathogenic fungi in the West has been most aggressively pursued by prtvate mdustry. Mycotech Corporation of Butte, MT, developed a diphastc system, initially for culture of white-rot fungi for soil remediation, which IS now opttmtzed for B basszana production (34). Blastospores, produced m a liquid medium in conventional fermenters, are mcorporated mto a proprietary sohd medmm that 1sloaded mto trays m large chambers wtth forced aeration and computer-controlled envrronment. Profuse sporulatlon IS mlttated throughout the substrate wtthm a few days. After the culture matures, the spores are drted within the chamber at a controlled rate to approx 5% moisture content, and then harvested directly from the chamber, all wtth a mmrmum of labor. The extracted product IS a nearly pure comdlal powder contammg 1.2-l .8 x 10” comdia/g. According to C Bradley (personal commumcation), yields from more than 20 consecutive production runs m a large pilot faclhty during 1996 averaged 1.1 x IO’Ospores/g of substrate (dry wt). Thts translates to a yield of more than 1Or3conidta/kg of substrate occupymg less than 1 L of chamber (fermenter) space (34). Average yields of 2.6 x 1Or3were achieved m a small pilot system (341, suggestmg that the full potential of the large system has not yet been reached. The Mycotech productton technology IS certainly adaptable to other fungal pathogens that can be mass-produced on solid substrates,especially Metarhzzium and Paecilomyces spp, but tt has thus far been optimized only for B. bassiana. The level of B. basszana comdra productton by the Mycotech systemIS approx fivefold higher than the maxtmum commercial-scale production of blastospores generally achtevable m a comparable ltquid fermentation volume, VIZ., approx 2 x 1012/L (6,3.5,36). This does not mean, however, that effictency ofblastospore production m submerged culture could not one day equal that of comdra production on sohd substrates, because blastospores are produced more rapidly than aerial conidla. The productton of 2 x 1012blastospores/Lreported by Fargues et al. (35) occurred within 48 h after moculatton of the hqutd medium. A number of hyphomycete species can be induced to produce true comdra m submerged culture (23,37-39). The submerged comdta of B. bassiana are produced nearly as rapidly as blastospores, and are considerably more stable (40). Studies also indicate that efficiency of submerged comdta production has the potential to equal that of blastospores. Following an extensive screening of B basszana isolates during development of the Boverm@ product (Ukranian Sctentific Research Institute of Plant Protectton, Ktev, Ukraine), Soviet researchers rdenttfied a strain that produced exceptional yields of submerged conidia In a highly enriched medmm, 3-4 x lOI comdta/L were produced wtthm 3-5 d. However, in an economically acceptable medium, ytelds averaged only 5 x 10’ l/L (23). Maximum yields of B. basszana submerged comdta

Mycoinsecticides and Field-Crop Pests

237

reported recently also have not exceeded this level (41,42). However, rates of submerged comdiation are clearly fungal species- and strain-dependent. Jenkins and Prior (38) reported substantially higher yields of Metarhzzzum Jlavoviride-submerged comdia (1.5 x 10r2/L withm 7 d) m an mexpensrve medium, and Lisansky and Hall (43) claimed that concentrattons of Vevtdlzum lecaniz comdia could reach 1013/L. Research and development of submerged spore production 1scurrently focused on improvmg spore yields and stability.

2.2. Product Stabilization The comdia of the Hyphomycetes are thick walled cells with the capacrty to persist in the envnonment under a broad range of condittons. Nevertheless, throughout the history of their development for mrcrobial control, unformulated preparations of these spores exhibited poor stability at moderate temperatures (44-46). This problem was encountered with spore preparations of nearly all entomopathogenic fungal specres, an important exception being comdia of A4 anisoplzae, which are known to survive as long as 24 mo at 26OC. However, this level of stability apparently exists only under condrtions of high humidity (46), and, according to Miller (47), normal 02/C02 concentratrons. Storage of large amounts of product under these condmons would be difficult, requiring special packaging (47) and methods to mhibit growth of microbial contaminants. These storage difficulties constantly interfered wrth field evaluations, and were a major constramt to commercialization. Coincident with the recent advances in fungus mass-productron technologies, however, have come discoveries that make possible the long-term storage of dry B bassiana and A4. anisopliae conidia under moderate temperature condittons (25°C). These discoverres represent a very significant breakthrough, considermg the many difficulties and high costs associated with cold storage. It 1sgenerally accepted that stability for 12-l 8 mo without refrigeration would be required to servtce general agricultural markets, although stability for 3-6 mo would probably suffice for products produced on contract for applications at a specific time (48,49). An important dtscovery came in the 197Os,when researchers reported that Inexpensive clay carriers, commonly used as msectrctde formulants, enhanced stability of B. basszana blastospores stored m so11over a broad range of temperatures (50). At the same time, Sovret scientists were mvestigatmg use of clays to dry B. bassiana spores produced in liquid culture systems (23), and to produce wettable-powder formulations (51). The benefits of usmg clays for formulation of fungal comdra were reviewed by Soper and Ward (52), and, m a notable unpublished report, Ward and Roberts (53) claimed that B. bassiana comdra formulated m attapulgite and kaolinite clays remamed stable for 12 mo at 26°C.

238

Wraight and Carruthers

Despite these dlscovenes, progress in the long-term, moderate-temperature stab&y of dry formulattons was slow and inconsistent over the next 10 yr. Daoust et al. (54) were unable to improve the 20°C stability of M anuopllae comdla by formulating with various clays and other materials. However, Alves et al. (55) substantially increased the room-temperature stability of comdla of A4 anisopliae to 9 mo by formulatmg with rice or corn flour, or with phylhte. An industry-developed formulation of B. basszana conidia produced m 1984 for Umted States pilot trials against Colorado potato beetle was extremely unstable (56). Zhang et al. (57) produced two wettable-powder formulations of B basszana with shelf hves of 8 mo, but this was at cool storage temperatures (lo-20°C). Imttal attempts to exploit the many advantages of anhydrous 011sas conidla formulants (see below) were also frustrated by stability problems (54,58,59) A major breakthrough came during the 1980s For many years, basic researchers, investigating stability of small samples of comdla, studied the effects of moisture expressed as relative humidity m the storage container, controlled through use of saturated salt solutions or desiccants (44-46,60). However, m most of these studies, the drying processes used to prepare the experimental spore preparations and the degree of desiccation achieved were not indicated. A few researchers dtd report that reduction of moisture content to 10% during harvest improved desiccation survival (23,481, and, m a United States patent application submitted m 1983, Jung and Mugmer (28) claimed that polymer-encapsulated B. basszana blastospores could be stabilized by a two-stage process of controlled drying to a water activity (a,) of CO.1. However, unttl recently, the effects of formulation free-moisture on long-term stability of conidial preparations of entomopathogenic fungi remamed little studied and poorly understood. China has perhaps the greatest experience with practical use of B basslana (6), and work with operational-scale quantities of comdial powders led its researchers to investigate moisture content as a more practically measurable and controllable variable. This ultimately resulted m the quantlficatlon of the substantial negative impacts of excess free water on stability of B basszana comdlal powders and formulations stored over a broad range of temperatures (61,62) According to Chen et al. (36), cited by Feng et al. (6), contdla of B bassiana, formulated in attapulglte clay with water content below lo%, showed no significant loss of virulence after storage for 12 mo at 26”C, which replicated the findings of Ward and Roberts (53). Drying of B bassiana to <5% moisture content was consequently established as a production standard by the forestry department of Fujian Province (6), and controlled drymg 1snow a widely utilized and mtenslvely researched protocol for stablllzatlon of hyphomycete comdla (34,47,63-66).

Mycoinsecticides and Field-Crop Pests

239

Followmg these discoveries, researchers reported that addition of dried, nomndicatmg silica gel slgmficantly improved the thermal tolerance and shortterm stability of 011formulations of A4 jlavoviride comdia (63,64). Moore et al. (64) reported that addition of slhca gel to varrous 011formulations increased the stability of M.$avovznde comdla from <2 mo to 30 mo at 17°C Hedgecock et al. (65) demonstrated a marked improvement in short-term, high-temperature stability of A4 jlavowzde, when comdla were dried to 5% moisture content prior to formulation m 011;the formulations retained 70% vlabllity for 70 d at 38°C. Moore et al. (66) confirmed the critical importance of drying spores prior to formulation in 011.Mass-production quantities of undried comdla formulated m oil were not stabilized simply by addition of silica gel. However, even dried comdla stored m 011,or as unformulated powders, were further stablIized by addition of silica gel; survival was increased from 3 wk to at least 3 mo at 28-32X The mode of action of the slhca gel was theorized to involve a contmuatlon of the drying process, maintenance of dry conditions, or absorption of destabilizing byproducts of fungal metabohsm. Sihca gel, mcorporated mto dry powder formulations of comdia of Metarhzzum spp and other hyphomycetes, has not produced stablhzmg effects as pronounced as those reported for or1 formulations (54,660,67). In fact, storing comdla m direct contact with silica gel was detrimental m some cases(54,60). Daoust et al. (Sls), however, reported a slight increase m stablltty when silica gel was added to a kaolimte clay formulation of M anisopliae comdia The amount of silica gel relative to the amount of fungal material and the chemical composltlon of the slhca gel used (especially with respect to mclusion of cobalt chloride as a moisture indicator) may be important factors. Many questions remam with respect to the effects of moisture content and sihca gel on the stability of fungal spores, and addltlonal research is needed. It warrants emphasis, however, that much of the variability in stability of fungal preparations is attributable to other factors, including genetics (stability 1s highly species- and strain-dependent), physlologlcal state arising from culture conditions, and drying rate (66,6&70). Nevertheless, great progress has clearly been made. Although the spore formulation/stabihzatlon methods of industry remam largely proprietary, extremely promismg results have been reported. Unformulated dry comdia of B bassiana strain GHA retained 70% viability for nearly 1 yr at 25”C, and 3 mo at 32’C (70). Mycotech claims that various wettable-powder and 011formulations exhibit similar stablltty (S. Jaronskl, personal communication). Equally encouraging results have been reported from another hne of research More than 30 yr ago, Clerk and Madelin (45) demonstrated that storage under mtrogen or enriched CO1 atmospheres improved the short-term stability of Al. anzsopbae comdla, this beneficial effect was confirmed by Roberts and

240

Wraight and Carruthers

Van Leuken (see ref. 71). Blachere et al. (72) subsequently noted Improved survival of blastospores of B. basszana (= B tenella) stored at 4°C under vacuum or nitrogen atmospheres, and Belova (23) reported improved stability of B. basszana submerged conidia stored m the absence of an. More recently, Andersch et al. (27) found that vacuum storage Improved stabtltty of dry mycehum formulations of A4 anrsoplzae, and Miller (47) reported that removal of oxygen from the storage atmosphere greatly enhanced stabihty of dry comdra of A4 anuopliae, allowmg storage for 12 mo at a htgh temperature of 37°C. In contrast, Jaronskt (70) claims that storage in anoxrc environments IS detrtmental to fungal spores. If the high levels of stability reported by industry prove consistent and applicable to a broad range of hyphomycete spectes and strams, these advances may represent the solutron to the storage problems long associated with formulations of these hyphomycete pathogens Although progress has been made m the opttmizatton of hyphal body and mycelium productton, dtfticulttes with long-term storage have not been resolved. Blastospores and myceha of many spectesof insect pathogenic fungi can survive desiccation and formulatton asmicrobial msecttctdes.However, they typttally begm to lose vtabihty wrthm a few months at room temperature (27,73), and preparations may be unstable, even under refrtgeratton (74). Development of commercral-scale processes for production/formulation of blastospores and myceha, with stabtlmes and mfectton potentials comparable to aerial comdta, would represent a major breakthrough m mycomsectrcrde development. Progress toward this goal has been reported. Jung and Mugmer (28) claimed that blastospores of B. bassiana, formulated m a xanthan gum/carob bean flour gel at 0.07 a,, survtved storage for at least 1 yr at 28°C and Stephan et al. (30) reported 68% viability of a spray-dried M jlavovrnde blastospore preparatton stored for 1 yr at 20°C. Nerther of these technologies, however, has been developed to commerctal scale. Currently, maxtmum stability of dry mycehum formulattons is approx 5-6 mo at 20-22°C (75,76). Storage stabrImes of various fungal entomopathogen preparations are summarized m Table 1. 3. Fungus Formulation Formulation has for many years been recogmzed as an Important key to commercial successwrth the entomopathogemc fungt, and IS undertaken with many specific goals, including formulatton to improve spray coverage, mcludmg mtcrostte targeting, ramfastness, and so on; to increase safety (e.g., reduce dust inhalation, eye n-rttation); to improve and simphfy handling (improve miscibrhty, flowabtlity); to improve storage stabtlity (especially at moderate to high temperatures); to Improve field stability (espectally under exposure to ultraviolet radiation); and to improve efficacy (especially to reduce ambient moisture requrrements).

2

Beauverla bassiana Conidia Comdta Comdra Conidta Blastospores Mycelmm Metarhlzlum amsopllae Conidia Conidta Comdia Mycelmm Metarhlzium flavovinde Conidta Comdia Blastospores Blastospores Paecllomyces fumosoroseus Blastospores

Fungus

Conditions

Freeze-dned powder

108 d at 2832°C 77 d at 2537°C 12 mo at 20°C (oxygen-reduced) 22 wk at 30°C (oxygen-reduced) 75% after 1 wk at 22°C (under vacuum)

after after after after

80% 70% 68% 67%

Diesel oil + Dry powder Spray-dned Spray-dried

silica gel + sthca gel powder powder

63% after 24 mo at 26°C (97% RH) Good viability after 12 mo at 37°C (anoxtc) 95% after 2 mo at 25°C 100% after 20 wk at 20°C (under vacuum)

78% after 12 mo at 26°C No vtabthty loss after 12 mo at 26T 70% after 12 mo at 25°C 93% after 9 mo at 25°C 100% after 12 mo at 28°C 100% after 5 mo at 22°C

Shelf life (vtabilttv after ttme at “C)

Temperature

Unformulated Dry wettable powder Sunflower oil Dry granules

Dry attapulgite clay Dry attapulgite clay Dry, unformulated Paraffimc oil Dry xanthadcarob gel Dry alginate pellets

Moderate

Metarhizium, Under

Formulatton

Table 1 Maximum Reported Stabilities of Beauweria, and Paecilomyces spp Formulations Stored

74

66 66 30 30

46 47 67 76

53 36 70 70 28 75

Ref

242

Wraight and Carruthers

Review of each of these lines of research is not possible in this chapter. Instead, the followmg dtscusston will focus on one of the most sigmficant recent advances, and one that has contributed to multiple formulation obJectives. use of oil diluents. Oils are highly compatible with lipophiltc conidta, as well as with the insect cuticl&eaf cuticle target system. Thts compattbtltty reduces or ehmmates the need for wetting, stickmg, or spreading agents (58). Oils are also more effective carriers for low-volume applications than water, which rapidly evaporates when apphed as small droplets. It was this use of oils as carriers for ultra-low volume (ULV) spray applications that originally stimulated the study of these materials m the formulation of entomopathogemc fungi (S&77). From that pomt, the many other advantages of oils were quickly recognized. According to Jaronski (701, formulation in oil significantly enhanced the high-temperature stability of B. bassiana comdia; however, this was observed only wtth petroleum-based paraffinic 011s.In tests at 40°C paraffinic oil formulations showed no substantial loss m viability for 6 mo; dry conidia and comdia formulated m several vegetable oils, including peanut and cottonseed 011s lost all vtabihty over the same period. On the other hand, Morley-Davies et al. (78) reported that formulation m a light paraffinic oil did not enhance stabthty of dry corndta of A4fluvovzrlde stored over a broad range of temperatures. Jaronski (70) also reported substantially greater stabihty of B. bassuzna comdia stored m paraffmc vs vegetable oils at 25°C In contrast, Marques and Alves (67) reported better stability of B basszana and A4 anisoplzae comdia formulated m sunflower oil than m mineral oil, and Moore et al. (64,66) concluded that botanical 011swere better for storage of A4. j7avovlride than petroleum-based 011s But this was, m turn, contradtcted by Stathers et al. (59), who found kerosene substantially better for storage of M jlavovzride comdia than a variety of vegetable oils. These mconsistencies are probably explamed, at least m part, by the different oils used in these studies. Indeed, there appear to be substantial differences m the spore-stabilizing capacmes of various oils, and the picture IS further comphcated by addition of emulsifymg agents. These problems are the focus of considerable ongoing research. Mycotech is also evaluatmg botanical oils to form the basis of a B. basszana formulatton acceptable to organic farmers (S Jaronski, personal communication) Some dry powder formulations of fungal comdia have proved irrttatmg to the eyes of test animals (79). Persistent irrttatton, even though not the result of ttssue Invasion or pathogen growth, requires a warning or danger label under current U S. Environmental Protection Agency (EPA) regulations. Formulation m oil, however, has been found to reduce this ocular irrttatton problem (79). Oil formulations also circumvent the dust-mhalation problems associated with some wettable powders, espectally simple clay-based powders and powders

Mycoinsecticides

and Field- Crop Pests

243

containing ground silica gel. In particular, reducmg the risk of mhalation reduces the allergenic capacity of a formulation. The extensive recent work with oil formulants 1snow leadmg to important advances m development of nomrritating, nondusty, second-generation wettable powders. The use of 011formulants not only improves handling safety, but also improves handling characteristics m general-a factor of critical importance to successful comrnerciahzatton. Liquid formulations are very easily measured and dispensed under field conditions, and, if supplemented with effective emulsifymg agents, readily mixed m spray tanks. Being nonabrasive, they also extend the workmg life of spray equipment. In 1995, Wright and Chandler (80) patented a mixture of vegetable oils and vegetable proteins and carbohydrates for formulation of hyphomycete comdia This composition was claimed to enhance contact between the fungus and the insect host by acting as an arrestant and feeding stimulant. A B. basszanabased mycomsecttcide (Naturalis @,Troy Brosctences, Inc., Phoenix, AZ) was developed and recently registered in the United States for control of whiteflies and other insects (81). Great expectations for rapid commercial successof oil formulattons of fungal pathogens arose from early laboratory studies, mdtcatmg that this technology could provide a useful level of protection from the lethal effects of UV radiation, and substantially enhance fungal efficacy under dry conditrons. According to Moore et al. (82), comdra of A4.jlavovmde formulated m oil survived exposure to artificial sunlight longer than condta formulated m water. However, this result was not observed m the field by Inghs et al. (83), followmg application of 011formulated conidta to leaves of grass, it was suggested that the protective effect was lost after the oil spread over the leaf cuticle, and was absorbed by mesophyll cells. Prior et al (58) and Bateman et al. (84) recorded substantially greater efficacy of comdia formulated m oil vs water, followmg precise topical apphcations to mdividual insects, They attributed the enhanced efficacy to the greater adhesion and spread of the 011on the insect cuticle, which resulted m the transport of spores mto membranous folds or arttculations on the insect body, where conditions favored spore germmatton and host infection Bateman et al (84) reported that the greater efficacy of an 011formulatton of A4flavovzrzde comdia was most pronounced at low humidities, and that, under these conditions, oil formulation dramatically increased the slope of the dose-mortality regression line. The data reveal, however, that the enhanced efficacy under low humidity condttions was expressed primarily as a faster speed of kill, because the LDsO of the water-formulated conidia calculated at d 6 postapplication (approx 1O4 spores/msect) was equal to the LDS, of the oil-formulated comdia determined at d 5. Even more significantly, the steep slope of the regression lme was ulti-

244

Wraight and Carruthers

mately shown to be a characteristicassociatednot with oil formulattons acting under dry conditions, but with a particular strain of Schzstocercagregana (ss). Utthzing the same assayprotocol (86), Milner et al. (87) recorded more typical slopes of 1.l-l 5 m tests of two locust species In bioassaysin which locusts were fed A4 flavovirz’de-treated foliage, Milner et al. (88) also found that formulating comdta in oil reduced the median lethal time of a simulated field rate by approx 1 d Few studies have directly compared the efficacy of oil- vs water-formulated comdia under field conditions. However, a recent test conducted in Benm showed that oil vs water formulations of submerged comdia of M jlavovmde were equally effective against S gregaria sprayed in field cages and momtored m the laboratory (89). Other tests indicated that oil-formulated comdia were not sigmticantly more efficactous than water-formulated comdia applied m ULVs to crop foliage in the field or laboratory, and then fed to grasshoppers (90) In field tests of B bassmna against Bemzsza whitefhes m Texas, emulsifiable 011formulations of conidia also performed no better than aqueous suspensions prepared with a highly effective wettmg agent (0 03% Silwet@ L-77 Specialties, OS1 Inc., Tarrytown, NY) and applied by a high-pressure hydraulic sprayer (91). Conditions were cool and wet during this trial; however, semilar results were obtamed m a melon trtal during severe drought conditions (S. Wraight and C. Bradley, unpublished data). These results are difficult to explain in light of the greater adheston and spreading of oil-formulated spores on the msect cuticle They suggest that results from bioassays m which the entire dose of spores IS applied m a single, large l-2 pL droplet (l-2 x 1O6pL) of oil or water to the insect body (58,84,86) may not be predictive of effectivenessunder field conditions. In a field treated with an atomizing sprayer, insects are exposed to many small (500 pL) droplets. The results from Bemn and Texas suggest that atomized sprays of water can deliver a lethal dose of spores to the msect as efficiently as droplets of oil under field conditions. In studies of another application strategy, viz., use of fungus-treated baits, Inglis et al. (92) observed significantly less infection among grasshoppers fed lettuce leaf disks treated with B bassiana comdia suspended m water than among those fed disks treated with spores suspended in a relatively large amount of oil (relative to the amount applied to foliage by ULV apphcation). However, this study was conducted m the laboratory, and it is not known, for example, how long 011would be retained by a bait m contact with soil or other field substrates under arid, hot conditions. Many more comparative formulation studies under natural field conditions will be required before any final conclusions regarding the effects of oil formulation on field efficacy of fungal pathogens can be drawn. Even should oils be found to induce no substantial increase m fungal mfectivity under field conditions, the many other advantages of these formulation

Mycoinsecticides

and Field-Crop

Pests

245

matertals (espectally as ULV carrrers) are more than sufficient to ensure continued research and development. An important caveat with respect to this work is that many oils are themselves msecticidal; in fact, various 011sare registered and labeled as insecticides. Also, lake fungal pathogens, the msecticidal efficacy of 011scan be significantly influenced by variations m ambient temperature, humidity, and air movement-all important factors affecting such properties as viscosity and volatihty. This greatly comphcates the processes of development and evaluation of oil-formulated mycomsecticides, making comparison and interpretation of laboratory and field test results more difficult. 4. Use and Delivery of Fungal Pathogens This final section will focus on three field-appltcation systems/technologies that have been rapidly advanced m recent years, and represent examples of control systems that are closest, or have come closest, to commercial implementation 4.1. Broadcast Application of Granular Formulations of En tomopa thogenic Fungi Against Cryptic Pests Cryptic insects, especially larval stagesthat bore mto plant tissues, or tunnel beneath the soil surface, include some of the most important pests of agriculture. However, because these pests do not feed on treated crop foliage and also cannot themselves be directly targeted, spray applications designed to deposit fungal spores more or less umformly over a field or crop row is not necessarily the most efficient use of fungus moculum. In many cases, granular formulations capable of penetrating the crop canopy to reach the ground, gravitating mto specific target sites, such as the bases of leaf whorls, or, if contammg a bait, inducing insects to feed on them or carry them mto their nests, can be considerably more effective. In a laboratory study, Krueger et al. (93) observed that a granular mycelium formulation of A4. anisoplzae was more effective against scarabaeid larvae than unformulated comdia mixed into soil, and hypothesized that spores m aggregates generated by the mycehum parttcles were more infectious than isolated spores. Since cryptic insects cannot be directly inoculated by a spray application, the infectious units must survive until the insect either makes contact inadvertently or is attracted to a bait. In this sttuatton, granular formulations can often provide the protection (especially from UV radiation) necessary to ensure survival (29). Development of granular formulations for application of fungal pathogens has been pursued for many years. Productron has ranged from the coating of dry spores onto bran or gram (94) to the drying and fragmentation of mycelium (25,26) or myceltal pellets (27), to the encapsulation of spores or mycebum in starch or algmate (29,75). Productron of fungus on whole kernels of

246

Wralght and Carruthers

gram, followed by drying and pulverizatton of the substrate, produces an essentially unformulated granular preparation (l&95) Experimental successesand mdustrial development efforts with these formulations have been most promment agamst European corn borer (8,96,97), cockchafers and white grubs (29,95,98), and other coleopteran pests (76). Each of these cited efforts has resulted m advanced development, and, m some cases, registration of mycomsecticide products. Calliope-N.P.P. of France has registered clay granule formulations of B basszana (Ostrmyl R@) for control of corn borers and Beauverza brongniartzz (Betel@) for use m combmation with a chemical msecticide for control of white grubs on Reunion Island (97,98). In Australia, Bio-Care Technology recently developed a granular formulation of A4 anisopbae (BioGreen@) for control of the redheaded cockchafer, Adoryphorus couloni (99) Bayer AG of Germany achieved registration of a pelleted dry mycelmm formulation of M anzsopliae (BIO 1020@)(76). Recent tests have shown that Ostrmyl R applied at the rate of 25 kg/ha IS as effective as chemical msecttcides (97), and development of this product IS contmumg However, Bayer recently discontinued development of BIO 1020, because product costs limited market potential. The current constraints to widespread acceptance and use of these products are many. Fungal products mixed or injected mto soil m the field have exhibited mconsistent efficacy caused by various factors related to so11ecology (76,10&104). Also, applications below the soil surface are generally more difficult than aboveground applications, and require specialized equipment (19,104,105) An additional dtsadvantage is that granular formulations tend to be very bulky. In some cases, application rates of 100-200 kg/ha have been indicated (I9,95). Even at such high rates of application, however, granular moculum drtlled or mixed mto the soil does not provide complete coverage. Infectious units do not immediately come into contact with all or even a large proportion of the target Insects; control thus tends to develop slowly, and full control may not be achieved until the followmg season (19,95). Finally, production of the large number of granules required to provide effective coverage of the sot1 habitat is far less efficient than production of mdividual spores m a given fermentation volume. Not all of these problems are associated with every formulation or target environment, and, taken by themselves, they may not represent serious constraints to commercial development Commercial viability, however, is umversally constrained by cost. There are considerable added costs assoctated with the production, packaging, distribution, storage, handling, and application of bulky products. Agam, this is not, by itself, a serious problem. Many granular formulations of chemical msecticides are successfully applied at rates of many kg/ha However, costs are greatly Increased for volummous, unstable

Mycoinsecticides

and Field-Crop

Pests

247

fungus formulations requlrmg low temperature storage. When these are added to the high costs still associated with production of most fungal pathogens, granular formulations generally become too expensive for routme broadcast appllcatlon against field crop pests (19,95). This problem 1s obviously most severe for formulations of typlcally unstable hyphal bodies or mycelia produced m liquid fermentation systemsat low levels of efficiency. The economICScan be dramatically improved in cases in which the fungus persists in the sol1 to provide control over more than one season (2,29,95); however, control strategies calling for Infrequent or unpredictable appllcatlons provide less mcentlve to commercial developers. Because of this, baited formulations with high densities of active ingredient may ultimately prove most successful. However, targeting of discrete pest habitats (the strategic basis for use of the Ostrmyl product) is clearly another valuable approach. 4.2. UL V Spray Applications Against Rangeland Pests In the wake of destructive outbreaks of grasshoppers and locusts m North America and Africa between 1985 and 1988, numerous IPM projects were organized by agencies of many governments to develop blologlcal control agents for regulation of grasshopper populations (106). Among the most common of grasshopper natural enemies, the entomopathogemc fungi have been the subjects of intensive studies, and development of grasshopper-pathogenic strains of A4jlavovzrzde and B basszana has reached advanced stages m Africa and North America, respectively. Because grasshopper control involves targeting of swarms ranging over very large areas that are often maccesslble by road, a highly efficient apphcatlon method IS required. ULV apphcatlon has been the method of choice. Details of use of ULV technologies for grasshopper control were presented by Mathews (107) and Bateman (108). Strains of M.Jlavovzride proved more virulent against African grasshoppers and locusts than B bassiana in laboratory bioassays (85,109); two isolates were selected for small-scale field trials in Africa (21,11&212). Mass production was carried out on cooked rice in plastic bags, and spores were formulated m keroseneor in mixtures of keroseneand peanut 011.Single apphcatlonsof 2-5 x 1Of2 comdla/ha were applied by hand-held spinning-disk atomizers (Micron Sprayers, Bromyard, UK). Results of several trials m Benin have been promlsmg. Lomer et al. (21) reported up to 95% mortality of Zonocerus varzegatus collected from treated areas of a field and incubated m laboratory cages. Douro-Kpmdou et al. (ZZO) subsequently recorded a 90% reduction m density of free-ranging Z. varzegatus within 15 d after treatment at a rate of 2 x lOI corudla/ha. Kooyman and Godonou (112) also reported high mortahty (100%) of S. gregavza m field

248

Wraight and Carruthers

cages, but were unable to determine treatment impacts on migrating bands. In contrast, Thomas et al. (221) recorded only 62% mortality among Hzeroglyphus daganensis collected from three g-ha treatment plots at various times postapplmatron and held in the laboratory; and, followmg treatment of several large (4-ha) plots, Lomer et al. (113) noted a 70% reduction in densities of freeranging H. daganenszs after 3 wk. Initial field trials of oil-formulated M. jlavoviride against grasshoppers and locusts m Australia have also produced 60-90% control within 3 wk (88; see ref. 113). Although found less pathogenic than M. jlavorvmde m laboratory assays, the existence of commercial-scale production technologies for B bassrana led to more rapid development of this pathogen. In 1995, Mycotech received EPA registration of an oil-flowable formulatton of B basszana developed for appltcations against grasshoppers, locusts, and Mormon crickets. The product, Mycotrol@-GH OF, is composed of 10” conidia of B. basszana strain GHA m 1 qt (0.946 L) of oil. For use, the general recommendation is to dilute 1 qt of product with additional 011and apply at ULV rates of l-5 L/ha. Unfortunately, tests of this product m Africa and North America yielded disappomtmg results (114-117). In some cases,efficacy appeared good when evaluations were made on Insects collected from treated areas and held m field cages or laboratory containers, however, evaluations of impacts on free-ranging populations were not encouraging. Better results were obtained by applymg the same fungus as a granular ban formulation (dry spores loosely adhered to wheat bran). However, even at a high rate of 2 x 1Ot3spores/ha, densities of Melanoplus sangumzpes were reduced by only 60% (94). The consistent discrepancies observed between efficacy estimates from caged vs free-ranging grasshoppers were predictable m view of the studies by Carruthers et al. (118), showmg that basking behavior of Camnula pelluclda could raise hemocoel temperatures to 40°C and strongly mhtbtt infection by a fungal pathogen. Laboratory studies have now shown that elevation of grasshopper internal body temperatures to between 38 and 42’C for as little as 1 h/d substantially reduces pathogemcity of B basszana strain GHA (119). This is a surprismg result likely to have far-reaching implications for the use of B basszana m microbtal control. In contrast, A4 jlavoviride has a higher temperature optimum than B. bassianu, and 1smuch less affected by grasshopper thermoregulatton (120); this characteristtc almost certainly explains the better and more consistent field results obtained with this pathogen, As would be expected, Inglts et al. (120) found B. bassiana more effective than A4 Jlavovmde at lower temperatures, leading them to suggest applying both pathogens m combmation to overcome high- and low-temperature constramts The discovery of behavioral resistance to B basszuna mfectton has led researchers to investigate the potential of combining fungus applications with

Mycoinsecticides and Field-Crop Pests

249

toxic agents that interfere with baskmg behavior, or in other ways enhance the infection process. Sanyang and Van Emden (121) recently reported that a low rate of cypermethrin, combined with A.4 Jlavovzride, inhibited grasshopper feeding by 50% within 1 d after treatment, and advanced mycosis by 48 h. Application of chemical insecticides in conjunction with fungal pathogens is an old strategy m applied mycopathology (I), which is being reexamined as fungal products become more widely available for testing. Recent findings are extremely encouraging (122-124). An alternative solution to the behavioral resistance problem would mvolve development of fungal strains or mutants with greater thermal tolerances. However, this approach has not been pursued, because of the unknown risks associated with increasing a pathogen’s thermal tolerance (especially in the range of mammalian body temperatures). Another important constraint to use of these pathogens for grasshopper control IS the great instability of comdia exposed to solar radiation. This is especially a factor in use of ULV technology, because this application method deposits comdia primarily on the upper, exposed surfaces of plant foliage. Insect pathologists have studied this problem for many years; research has focused primarily on sunscreens for mcorporation into pathogen formulations. Some of the most promising sunscreenstested thus far are stilbene brighteners, including Tinopal LPW (Sigma, St. LOUIS,MO) (125). However, these materials are not soluble in oil, and, even when incorporated at a high rate of 5% m aqueous spore suspensions, Inglts et al. (83) recorded an approx 80% loss in fungus viability within 48 h after application to field plots of crested wheat grass in Canada. Nevertheless, this material did provide a possibly useful level of protection that warrants additional study. Despite the generally poor survival of conidia on treated foliage, recent field tests have indicated that secondary pickup of M, jlavovzrzde spores from foliage or soil can contribute signrficantly to efficacy of spray applications (89,111). Researchers have also suggested that persistence of A4 flavovzride via secondary cycling (sporulation on or in killed hosts) is an important grasshopper mortality factor, and an advantage over chemical msecticldes (211). Although this might be true when grasshopper densities are low and insects are not migrating, it is unlikely that this mode of action could be relied on to protect crops threatened by highly mobile species under outbreak conditions. Beauveria and Metarhizium spp are rarely found at epizootic levels under natural conditrons. Predators such as birds and ants often consume fungus-killed grasshoppers before sporulation can occur (112). ULV application of M flavoviride represents a potentially useful technology for grasshopper and locust control. Realization of this potential will obviously depend on successful development of operational-scale production technologies and stable formulations; however, this is certamly achievable,

250

Wralght and Carruthers

especially m light of recent reports of good efficacy at apphcatton rates as low as 2 x 1Ot2comdia/ha. Ulttmately, an even greater potential for ULV apphcations of mycomsecttctdes may be reahzed against more sedentary pests in other habitats. ULV apphcattons of B. basszanahave been used successfully m China to control pme caterpillars; n-tthe forest environment, the fungus does persist and provide control for extended periods (6). 4.3. Moderate- to High- Volume Spray Applications Against Field-Crop Pests This IS the area m whtch fungal pathogens have known perhaps then greatest practical success.Operational-scale use of A4 anzsoplzae to control sugar cane spittle bug m Braztl, and of B. bassiana agamst Colorado potato beetle m Russra and Eastern Europe, and various msects m China, are the best-known and best-documented examples (2-6). The most extensive current use of fungal pathogens for control of Insect pests of field crops IS m China; however, details of methods and strategtes are not wtdely available (6). These notable successes,however, have not yet led to widespread development and use of these control agents m high-input, short-cycle, row-crop productton systems, even though profitabthty 1s often sufficient to sustain mtenstve use of costly msecttctdes. For the first time since the mceptron of the hundred-year-old hypothesis, however, there are strong indications of a change n-rthis situatton. The many advances outlmed above have stimulated a marked increase in the development and registration of mycomsecttctde products targetmg agricultural pests In thts respect, the entomopathogemc fungi have clearly reached a new level of development. Accordmg to Jaronskt (70), there are 17 mycomsectictde products currently available worldwide. In the Umted States alone, nine products are now available for commerctal or large-scale expertmental appltcations (Table 2). Mycoinsecttctdes are now also betng developed m countries that do not have long traditions tn the use of entomopathogemc fungi. In Colombta, Hoechst Schermg AgrEvo has developed a water-dtspersible granule (WDG) formulanon of B bassiana (Conrdta@), primartly for control of coffee berry borer (226). In Mexico, Agrobrologrcos de1Noroeste has commercralrzed several products contarnmg comdta of M anisopliae (Meta-Sin@), B. bassiana (Bea-Sin@), and Paeczlomyces fumosoroseus (Pae-Sin@) for control of various insects, mcludmg whtteflles and coffee berry borer (127). As m the case of the use of granular formulattons, tt 1s not posstble here to revtew the numerous international efforts m thts area of research Two examples representattve of very different control strategres wtll be the SubJects of the remanung dtscusston.

Mycotech

Troy BroScrences EcoSctence ThermoTrtlogy AgraQuest

BotamGard

Naturalis Bra-BlastTM PFR-97TM LagmexTM

ES WP OF ES WP ES WP WDG AS

Acttve ingredient

in the United

Comdta 2 1 x lO’O/mL Corndia: 4.4 x 1O’O/g Comdra 1 1 x 10r”/mL Comdra 2.1 x lO’O/mL Comdta: 4 4 x lO’O/g Comdta, 2.3 x lO’/mL Corndta, 4 x 109/g Blastospores, 1Og &u/g Mycelium, 1O’O cfi.r/L

for Registration

Formulatron”

or Submitted

B basslana Metarhlzlum anlsopllae Paecllomyces fumosoroseus Lagenidium glganteum

B basslana

Beauverla basslana

Pathogen

and Registered

Prmctpal market Field crops Field crops Rangeland Greenhouse Greenhouse Field and greenhouse Termite control Greenhouse Mosquito control

States

“Emulsrfiable or1 suspensron (ES), wettable powder (WP), or1 flowable (OF), water-dispersible granule (WDG), aqueous suspensron (AS)

Mycotech

Company

in Production

Mycotrol

Product name

Table 2 Mycoinsecticides

252

Wraight and Carruthers

4 3.7. Experiences with lnoculatlve Augmentation and Autodissemination of Fungal Pathogens The work m Europe on fungal control of European cockchafers (Melolontha spp) has spanned a century, and culminated in two effective methodologtes for suppressron of these pests of orchard, field, and horttcultural crops. The earltest strategy, mvolvmg apphcattons of B. brongnzartzz conidta and blastospores m aqueous suspensions and granular formulations to the sol1 as larvlcldes, was cited above (2,951. The other approach utilizes conventtonal hydraulic spraymg of the adult beetles swarming at forest borders. Fresh blastospores m aqueous broth, with skimmed milk added as a sticker and sunscreen, are applied by helicopter during late afternoon (to reduce UV degradation) at rates of 2-3.7 x 1Ot4 spores m a volume of 370 L/ha In a series of trtals, these treatments produced substantial reducttons m beetle fecundity. The treated adults also carried the fungal moculum mto the soil, where larvae were infected. As a consequence, eptzoottcs were mmated that reduced larval densities by 50-80% m the second generation, after treatment at 13 of 15 study sites (128). As yet, no commercially available products have ansen from thts technology. Fungus hasthus far been produced with government support for managementof cyclical pest outbreaks. This work with B brorzgnzartii has stimulated many recent attempts to introduce relatively small amounts of moculum mto pest populattons to augment natural fungus activity, or expand the range of effecttve pathogens (11,129). The strategy of placing devices (traps) m the field that attract msects, inoculate them with a large dose of fungus propagules, and then release them to spread the pathogen into the general populatton, has become known as autodtssemmatton. A number of these inoculattve augmentation strategies have shown considerable promise (130,131), but further discussion IS beyond the scope of this review. 4.3.2. Experiences with lnundative Applications of Mycomsecticides The final discussion focuses on what certainly represents one of the most ambmous efforts to date toward free-market commerciahzatton of an entomopathogenic fungus for microbial control of field-crop pests. Mycotech has registered several wettable-powder and oil-based formulatrons of B basslana strain GHA. Trade-named Mycotrol@, each formulation contams 2 x lOI comdra/lb or qt (4.41 x 10t3/kg or 2 11 x 1013/L). The first products, which recetved EPA approval m March 1995 (including an exemption from residue tolerances), were the first mycomsecttcides regtstered m the United States for control of insect pests of field crops. Development of the original Mycotrol WP mvolved extensive collaboratron between Mycotech and the USDA-ARS Subtropical Agricultural Research Laboratory (SARL) in Weslaco, Texas (232,133) The followmg 1sa brief account of field research and development

Mycoinsecticides and Field-Crop Pests

253

efforts targetmg the stlverleaf whitefly, Bemisza argentifoliz, m the Lower Rio Grande Valley. The objective of this narrow focus IS to relate some of the challenges encountered in developmg a mycomsecticide for field apphcattons, and to famihartze agricultural researchers with the handling of these novel microbial control agents. Results from numerous research trtals presented below have not yet been formally published. 4.3.2.1.

PRODUCT STABILITY AND HANDLING CHARACTERISTICS

The Mycotrol formulatrons were extremely stable under refrigerated storage The materials could be held for several years at 4°C tf kept dry. During these studies, all formulattons were kept tightly sealed and refrigerated. Contamers were warmed to room temperature prior to opening, to prevent moisture gam through condensation. In the field, brief exposures to normal high temperatures were not damaging; however, the products were shaded from direct sunlight at all times. The orrgmal wettable powder (Mycotrol WP) was extremely dusty and hydrophobic, and addition to the spray tank required premixing with a highly active wetting agent. It was discovered that the new organostlicone wetting agents, such as Silwet L77 at 0.01-0.03% and Sylgard 309@(Dow Cornmg, Midland, MI) at 0.03%, were most effective (133-135). An improved secondgeneration wettable powder has been developed, which will require no premixing or additional wetting agent; registration is expected by spring, 1998 (S. Jaronskt, personal commumcation). A new, emulsifiable 011formulation was recently registered, whtch can be applied with low-volume air-assist sprayers or moderate to high-volume hydraulic sprayers. In research trials, the Mycotrol ES (emulstfiable suspension) formulatron proved easier to use than the original wettable powder. This product could be quickly and cleanly measured in the field, and dispensed (poured) directly into the spray tank without premixmg, and without need for added wetting agent. 4.3.2 2. SPRAY APPLICATION AND EFFICACY

Because whiteflies inhabit primarily the lower surfaces of plant leaves, several appltcatton technologtes and various spray parameters were examined to maximize coverage of this target (134-136). Maximum coverage of leaf undersides is obtained by spraying upward from below canopy level. This is normally accomplished by means of nozzles mounted on swivels on vertical tubes (drop nozzles) carried between the rows. In tomato trials, excellent coverage and control was realized using hydraulic sprays at 60 psi (4 2 kg/cm*) (133). Excellent coverage was also achieved in cotton, although, for reasons not yet elucidated, this did not translate mto effective control (133). On the

254

Wraight and Carruthers

other hand, m several small-scale trials, tt was not possible to achieve effecttve coverage of broccolt using either low- or high-pressure hydraulics. Directing sprays from below the canopy ts generally not feastble with recumbent plants, ltke cucurbits cultured in wide beds For cucurbit appltcatrons, a high-pressure hydraulic sprayer, fitted with drop nozzles carried at or slightly above canopy level was ulttmately most effecttve. The booms of the sprayer were fitted with vertical drop tubes spaced 20 cm apart. Each drop tube was approx 25 cm long and included a central 5-cm section of rubber hose, sufficiently rtgid to allow movement only upon contact with a soled ObJect (preventing breakage m the event of ground contact). A ceramtc hollow-cone nozzle (Albuz@ hlac) was mounted on a swivel on the ttp of each drop tube, and the swivels were adJusted to direct the nozzles downward at a 4.5-degree angle to the ground The sprayer was operated at 400 psi (28.1 kg/cm2), at which pressure each nozzle sprayed 0.81 L/min. The tractor was driven slowly (5.6 km/h), with the spray-boom height adJustedto carry the nozzles as close to the crop canopy as posstble. Imttal sprays on small seedlmgs were made wtth three nozzlesper row. The two lateral nozzles were dtrected to spray perpendtcular to the row; the central nozzle was directed forward or rearward, parallel to the row As the plants matured and the bed grew wider, addtttonal nozzles were added to the center of the array. Depending on the row spacmg and the number of nozzles per row, the descrtbed system sprayed volumes of 50-470 L/ha. On level ground, lowering the boom, so that rear-facing nozzles actually penetrated the canopy, provided optimal coverage of leaf undersurfaces (S. Jaronski, personal communication). A 0 5 lb/acre (0.56 kg/ha) broadcast rate of Mycotrol corresponds to a dose of 2.5 x lOI3 comdta/ha, whtch translates to a theoretical concentratton of 2500 comdta/mm2 of planar surface. Concentration of this dose mto a narrow band (approx 35 cm), usmg the above-described sprayer configured wtth three nozzlesper row, consistentlyproduced deposttionsof 1500-3000 corndia/mm2 of the undersurfaces of the leaves of small cucurbtt plants. A spray program of 2-4 weekly applicattons at the 0.5 lb/acre rate in cucumbers and 5-7 apphcattons m cantaloupe melons, mmated at the time of mtttal egg-hatch, consistently provided 65-75% control of first-generation whitefly larvae (15,134). As the plants grew, and addtttonal nozzles were added, coverage of lower surfaces of leaves gradually decreased to a few hundred spores/mm2. Under condttlons of moderate pest pressure, and m short-cycle crops like cucumbers, control was maintained for sufficient time to protect the crop, and yield increases of up to 35% over spray controls were achieved (S. Wrarght, R. Carruthers, S. Jaronskt, and C. Bradley, unpublished data). However, under heavy pest pressures in long-cycle crops, such as melons, control was lost at the end of the season.

Mycoinsecticides

and Field-Crop

Pests

255

Despite the capacity of drop nozzles and high-pressure to enhance spray coverage and efficacy of contact insecticides, these technologies have not been widely adopted, for a number of reasons. Use of drop tubes Increases chances for equipment damage by ground contact (especially at high speeds) and high pressure generates finely atomized sprays that are susceptible to drift. Studies are therefore needed to determine if adjusting other spray parameters might compensate for use of drop nozzles and/or high pressure. Inittal results from cucurbtt trials m Texas suggest that drop nozzles are essential for achievmg effective coverage of leaf undersides with conventional hydraulic sprayers; however, it may be possible to substantially reduce pressure, at least on new seedlings prior to canopy development (S. Wrarght and C. Bradley, unpublished). In Arizona, Jaronskt et al. (137) achieved 65% control of larval whtteflies in melons using drop nozzles and moderate pressures of 80-120 psi. Excellent control of Colorado potato beetle was recently achieved m a trial m Virginia, in which B. basslana was applied at 40 psi using drop nozzles on a hand-held boom (138). Air-assisted sprays represent a promising alternative technology for applymg mycomsecttcides. In the broccoli test crted previously (135), use of portable, air-assist sprayers, to blast an atomized spray upward into the canopy from near ground level, provided sigmftcantly greater control than was achieved with hydraulic sprayers configured with drop nozzles(even at 400 psi). Manual direction of an-assisted sprays laterally into melon and cucumber canopies at a volume of 280 L/ha also resulted in excellent coverage, and up to 95% control of whitefly larvae (133). However, these results were not duplicated with commercially available sprayers, which are generally designed to spray from well above the canopy at low volumes of 45-100 L/ha. Nevertheless, using a sprayer built by Berthoud, Sodus, MI; (model B-N 200), good control was achieved of first-generation whitefly larvae m one cucumber trial, by spraying at the maximum output of approx 280 L/ha (at 5.6 km/h). Nozzles on this sprayer were spaced every 50 cm. The an-streamsfrom two nozzles were directed mto each row at an approx 30-degree angle from 30 cm above the canopy. Additional testing to assessthe potential of this spray technology for apphcation of mycomsecticides is needed. The applications m the aforementioned trials were made without regard to time of day. Early morning was preferred because of calm condttions. Sigmficant reductions m coverage achieved with the hydraulic sprayer were observed at wind speedsgreater than approx 16 kmh. However, in trials with many treatments, applications were usually not completed untrl late afternoon. Researchers working with entomopathogenic fungi often apply durmg late afternoon or evening, to reduce solar degradation (138). The authors are not aware, however, of any studies directly comparing efficacy of morning vs evening apphcattons

Wralght and Carruthers

256

of hyphomycete fungi. Because the comdta of these fungt generally require approx 1O-l 5 h to germinate, spores applied in the evening still must survive daytime conditions to complete the host penetration process There is certainly consrderable potential m such a strategy, however, especially for application of fast-germmatmg blastospores. This also provides an incentive to develop rapid-germmatmg corudia, and to discover formulation additives that induce rapid spore germination. It seemsunlikely that aerial sprays from fixed-wing aircraft could be used to deliver an effective dose of fungal spores to the lower surfaces of leaves m a crop canopy. Upon ejection from the aircraft, the spray droplets are caught m the airstream, which immedtately alters any destred trajectory; ultimately, most droplets settle more or less straight down onto the exposed upper surfaces of the crop foliage 4.3.2 3. INTEGRATION

WITH AGROCHEMICALS

Fruit and vegetable crops are suscepttble to many plant pathogens, and disease outbreaks often elicit heavy usage of fungicides Under these ctrcumstances,there 1spotential for negative interaction between these disease-control agents and beneficial fungi applied for insect pest control. Mycotech 1s supporting studies to provide recommendations for Integrated use of these control agents. As anticipated, prehmmary results of field studies conducted in melons m Texas indicated that not all classes of chemicals are antagonistic to B basszanastrain GHA (Mycotrol). Koctde@,Ridomil@, and Bayleton@ were not mhibitory; Maneb@and combined Rtdomil@-Bravo@ had significant detrimental effects when applied on the same day as Mycotrol. However, weekly applications of these fungicides did not substantially reduce levels of whitefly control when applied 2 d before or 2 d following applicatton of Mycotrol (T. Poprawski and S. Wraight, unpublished data) Investrgators at the Umversity of Maine also reported no significant reduction m B basszana activity against Colorado potato beetle when Kocide, Mancozeb, and Bravo were applied to potatoes 24-72 h after Mycotrol (139). These results suggest that Mycotrol will be compatible with many fungicides under field conditions, if treatments can be applied asynchronously. The authors’ studies mdicate that adult whiteflies are substantially less susceptible to fungal infection than the nymphal stages, Consequently, this technology cannot be used to control adults migratmg mto planted fields, B. bassEana, however, is htghly compatible (can be tank-mixed) with most commonly used chemical adulticides, including the organophosphates and synthetic pyrethrotds. B basslana is not recommended to be tank-mixed with endosulfan; however, these agents are compatible when applied separately. Fall plantmgs are frequently at great risk of being overwhelmed by large populations of

Mycoinsecticides

and Field-Crop

Pests

257

rmgratmg whitefhes. In cucumbers, we were successful m employing an mittal application of endosulfan to reduce adult populattons, followed by weekly appltcattons of Mycotrol to control the nymphs. 4.3.2.4.

COMMERCIALIZATION

Mycotech has recently completed construction of a commercial-scale productton facility with an annual capacity of 5 x 1018conidia (sufficient to formulate 250,000 lb of wettable powder), Mycotrol 1s now commerctally available in some parts of the United Statesat a prtce of $1520/one-half pound or pint, contammg lOi3 comdta (the per-acre rate recommended for most field apphcattons). Under field conditions, Mycotrol has exhibited useful efficacy against a surprtsmgly broad range of pests, including grasshoppers (94), whiteflies (233), Lygus bugs (123), Colorado potato beetles (E. Groden, T. Poprawskt, personal communications), various aphids, and western flower thrtps (J. Vandenberg, B Murphy, personal communicattons); however, the product has not been available long enough to assesscommercial viability. Initial sales of strain GHA have been strongest m the greenhouse market, m which formulations trade named BotamGard@ are now m use against whtteflies, thrips, and aphids. There 1sconsiderable potential for excepttonal efficacy m greenhouse systems, m which the economtcs support highly efficient manual applications of a recommended dose of 2 x 1Ot3comdia (1 lb of WP or 1 qt of ES formulatton) to an area of 10,000 ft*. This translates to a high dose of nearly 1014comdra/acre (more than 2 x 1014/ha). 5. Conclusions Thts discussion has outlined a number of important recent advances m development of mycomsectrctdes for control of field-crop pests. There is considerable promise that these will ultimately lead to the broader uttlizatton of these microbial control agents, but the future remains uncertain. Levels of optimism have run high during many periods of the hundred-year-old hypothesis, only to be met with new challenges. The reason for thts slow progress IS apparent to anyone who has attempted to develop a commerctal mycoinsecttcide. Wtth few excepttons, fungal pathogens have simply not been able to compete directly (economtcally) with the seemingly endless stream of novel, extremely effective, broad-spectrum, synthettc-chemical insecttctdes developed by agricultural chemists. This is especially true m the highly unstable environment of the cultivated field, where severe weather conditions or extremely rapid increases m pest populattons may limit the effectiveness of fungal pathogens. No known strain of B. basszana or M anisoplrae, for example, has the combmed speed, level, consistency, and range of efficacy comparable to many novel msecticrdal chemistries

258

Wraight and Carruthers

Accepting this, it 1snonethelessundeniable that there extstsa growmg demand for btological alternatives to synthetic chemical msecticides, and that the advances in mycomsecticide technologies outlined herem have brought several species of entomopathogemc fungi to the verge of commerctal success.For thts successto be finally and fully realized, however, progress must contmue. Additional research is needed, especially wtth the goals of improvmg product efticacy and conststency under variable field condittons The future looks extremely promismg. Genetic and physiological engineering of fungal pathogens has only begun, but already reveals a great potential to increase speed of kill (140), and to reduce moisture thresholds for spore germmation (141). Studies of synergism between fungal pathogens and low doses of chemtcal msecttctdes Indicate considerable potential for integrated control apphcattons (221424). Addtttonal tmprovements m product stability, applicabthty, and efficacy are certam to come from the rapidly growing field of mtcrobial formulation. The crittcal need for more mtensively replicated field experiments and large-scale trials under commercial production condmons cannot be overstated Because of this need, the development and registratton of various mycomsecticide products stands as the most sigmficant accompltshment of recent years. These products are introducing fungal entomopathogens to agrtcultural researchers worldwide, and enabling them to pursue development and evaluation of novel formulattons, apphcatton strategtes,and delivery technologies on an unprecedented scale. This chapter has shown that, m some mstances,entomopathogemc fungi are already providmg control equivalent to that provtded by available chemical msecticides, at comparable or even lower costs. It further contends that the potential exists for mycomsecticides to, in many cases,provide acceptable pest control at a lower environmental cost than chemical msecttcides, to be mtegrated with synthetic chemical msecticides and thus reduce chemical use and slow development of msecttcide resistance, and to contribute, by sustaining enzootic or eptzootic disease cycles, to the long-term stabtlization of field-pest populations, and, consequently, of insect pest management systems. Note Names are necessary to report factually on avatlable data. However, the USDA neither guarantees nor warrants the standard of the product, and the use of the name by USDA implies no approval of the product to the excluston of others that may also be suitable. Acknowledgments The research and development program cited herein, which culminated m the commerctallzatton of B. basszana (as Mycotrol) for btologtcal control of

Mycoinsecticides

and Field-Crop

Pests

259

Bernlsza whiteflles, would not have been possible wlthout the excellent techmcal assistance of J. Garza, S. Galaml-Wraight, F. De La Garza, P. Wood, J. Bntton, N. Underwood, M. Becerra, M. DeAnda, and J. Riveira We are grateful to Clifford Bradley, Sandra Galaini-Wralght, Stefan Jaronskl, Clayton McCoy, Roberto Perelra, Tadeusz Poprawski, and Donald Roberts for crltlcal reviews of the manuscript.

References 1 Ferron, P. (1985) Fungal control, m Comprehensive Insect Physzology Bzochemzstry and Pharmacology, vol 12, Insect Control (Kerkut, G A and Gllbert, L I., eds.), Pergamon, Oxford, pp. 3 13-346 Ferron, P (1978) Biological control of Insect pests by entomogenous fungi Annu Rev Entomol 23,409-442 Alves, S. B. (1986) Fungos entomopatog&nicos, 111Controle MJcrohzano de Znsetos (Alves, S. B., ed ), Editora Manole, SBo Paulo, pp. 73-126 McCoy, C. W , Samson, R A , and Bouclas, D G (1988) Entomogenous fungi, m CRC Handbook of Natural Pestlcldes, vol. 5, Microbial Insecticides Part A, Entomogenous Protozoa and FungJ (Ignoffo, C M , ed.), CRC, Boca Raton, FL, pp. 15 l-236 5 Roberts, D W (1989) World picture of biological control of Insects by fungi Mem Inst Oswald0 Cruz (RIO de JaneJro) 84,89-100. 6 Feng, M G , Poprawskl, T J , and Khachatounans, G G (1994) ProductIon, formulation and application of the entomopathogemc fungus Beauverla basslana for insect control. current status. Biocontrol SCJ Technol. 4, 3-34 7 DeBach, P (1964) The scope of blologlcal control, in Blologzcal Control of Insect Pests and Weeds (DeBach, P , ed,), Reinhold, New York, pp 3-20 8 Wralght, S P and Roberts D W (1987) Insect control efforts with fungi Dev Ind Mcrobiol 28,77-87 9 Gillespie, A T and Moorhouse, E R (1989) The use of fungi to control pests of agricultural and horticultural importance, m Bzotechnology ofFung1 for ImprovJng Plant Growth (Whlpps, J. M and Lumsden, R. D , eds.), Cambridge Umverslty Press, New York, pp 55-84. 10. Ferron, P , Fargues, J , and Riba, G. (1991) Fungi as mlcroblal msectlcldes against pests, m Handbook of Applied Mycology, vol 2 Humans, Anrmals, and Insects (Arora, D. K , AJello, L., and Mukerji, K. G., eds ), Marcel Dekker, New York, pp 665-706 11 Roberts, D. W and HaJek, A E. (1992) Entomopathogenic fungi as blomsectlcldes, in Frontiers znIndustrJalMycology (Leatham, G. F , ed ), Chapman and Hall, New York, pp. 144-159 12. Bartlett, M. C and Jaronski, S T (1988) Mass productlon of entomogenous fungi for blologlcal COntrd of inseCtS, in FungJ En Blologlcal Control Systems (Burge, R N , ed.), Manchester University Press, New York, pp 61-85. 13 Meynell, G. G. (1957) Inherently low precision of infectivity titrations usmg a quanta1 response BlometrJcs 13, 149-163

Wraight and Carruthers

260

14 Feng, Z , Carruthers, R. I., Roberts, D. W., and Robson, D S (1985) Age-spectfic dose-mortahty effects of Beauverza basszana (Deuteromycotma’ Hyphomycetes) on the European corn borer, Ostrinia nubzlalzs (Lepidoptera Pyrahdae) J Invertebr

Path01 46,259264

15 Wratght, S P , Carruthers, R I., Bradley, C. A , Jaronskt, S T., Lacey, L A , Wood, P , and Galamt-Wraight, S (1998) Pathogemctty of the entomopathogemc fungi Paeczlomyces spp and Beauverza basszana agamst the stlverleaf whrtefly, Bemisza argentzfolzz. J Invertebr Pathol , m press. 16 St. Leger, R J (1994) Mycomsectrctdes an opportunity for genetrc engmeermg, m Proceedings of the VI Internatzonal Colloquzum on Invertebrate Pathology and Mzcrobzal Control, vol 1, Montpelher, France, pp 299-304 17 Alves, S B and Pereira, R M (1989) Productton of Metarhzzzum anzsoplzae (Metsch ) Sorok and Beauverza basszana (Bals ) Vu111 m plasttc trays Ecosszstema14, 188-192 18 Aregger, E. (1992) Comdia productton of the fungus Beauverza brongnzartzz on barley and quahty evaluatron during storage at 2°C J Invertebr Path01 59,2-10 19 Rath, A C (1992) Metarhzzzum anzsoplzaefor control of the Tasmaman pasture scarab Adoryphorus coulonz, rn Use of Pathogens zn Scarab Pest Management (Glare, T R. and Jackson, T A., eds ), Intercept, Andover, MA, pp 2 17-227 20 Qumtela, E D (1994) Producao de Metarhzzzum anzsoplzae(Metsch.) Sor em qurrela de arroz An Sac Entomol. Braszl23,557-560 21. Lomer, C J , Bateman, R. P , Godonou, I., Kpmdou, D , Shah, P A , Paratso, A , and Prior, C. (1993) Fteld infection of Zonocerus varzegatus followmg apphcation of an oil-based formulatron of Metarhzzzumflavovzrzde comdta. Bzocontrol Scz Technol 3,337-346

22 Mendonca,

A. F (1992) Mass productron,

application

and formulatron

of

Metarhzzzum anzsoplzae for control of sugarcane froghopper, Mahanarva postzcata, in Brazil, in Biologzcal Control of Locusts and Grasshoppers(Lomer,

C J. and Prtor, C., eds ), CAB International, Wallmgford, UK, pp 239-244 23. Belova, R. N (1978) Development of the technology of Boverm productton by the submersion method, m Proceedzngs of the First Joznt US/USSR Conference on the Productzon, Selectron and Standardzzatzon of Entomopathogenzc Fungi of the US/USSR Joint Working Group on the Production of Substances by Mzcrobzologzcal Means (Ignoffo, C. M , ed.), Jurmula, Latvia, SSR, pp

102-l 19 24 Lrpa, J J. (1985) Progress m biological control of the Colorado beetle (Leptmotarsa decemlzneata)m Eastern Europe Bull. OEPP/EPPO Bull 15,207-2 11 25. McCabe, D and Soper, R S (1985) Preparation of an entomopathogemc fungal insect control agent US Patent 4530834. 26 Rombach, M C , Aguda, R M., and Roberts, D. W (1988) Productron of Beauverza basszana(Deuteromycotma Hyphomycetes) m dtfferent hqutd medra and subsequent comdiatron of dry mycelium Entomophaga 33,3 15-324 27. Andersch, W , Hartwrg, J., Reinecke, P , and Stenzel, K. (1990) Production of mycehal granules of the entomopathogenrc fungus Metarhzzzum anzsoplzaefor

Mycoirtsecticides

and Field-Crop

Pests

261

biologxal control of soil pests, m Proceedings of the Vth Internattonal Colloqutum on Invertebrate Pathology and Mtcrobtal Control, Adelaide, Australia, pp 2-5 28. Jung, G. and Mugnier, J. (1989) Low-water-activity mocula for biological control US Patent 4886664. 29. Pererra, R. M. and Roberts, D. W. (1991) Alginate and cornstarch mycehal formulattons of entomopathogemc fimgt, Beauverta basstana and Metarhzztum anisopltae. J. Econ. Entomol. 84, 1657-1661. 30 Stephan, D., Welling, M., and Zimmerman, G. (1997) Locust control with Metarhtzium jlavovtrzdae. new approaches m the development of a biopreparation based on blastospores, in New Strategtes tn Locust Control (Krall, S , Peveling, R , and Ba Dtallo, D , eds ), Brrkhauser Verlag, Basel, Switzerland, pp. 151-15s. 31 Pan, W. Y , Tan, D. L., and Yin, F. M. (1988) 011 formulation of Beauverza basstana in ultra-low capacity against the Masson’s pure caterpillar, m Study and Appltcation of Entomogenous Fungt rn China, vol. 1 (Lr, Y W , LI, Z. Z , Liang, Z. Q., Wu, Z. K., and Xu, Q. F., eds.), Academic Periodical, BeiJlng, pp. 70-76 32 Samsmakova, A , Kalalova, S., Vlcek, V., and Kybal, J. (1981) Mass production ofBeauverta basszana for regulation of Lepttnotarsa decemltneata populations J invertebr Path01 38, 169-174. 33 Goettel, M. S (1984) A simple method for mass culturing entomopathogemc Hyphomycete fungi J Mtcroblol Methods 3, 15-20. 34 Bradley, C. A., Black, W. E , Kearns, R., and Wood, P. (1992) Role of production technology in mycomsecticide development, rn Frontters tn Industrial Mycology (Leatham, G. F , ed.), Chapman and Hall, New York, pp. 160-173 35 Fargues, J , Robert, P H , and Reisinger, 0 (1979) Formulatton des productions de masse de l’hyphomyctte entomopathogene Beauvena en vue des applications phytosanitaires. Ann Zool. Ecol. Antm 11,247-257. 36. Chen, C. J., Wu, J W., Li, Z. Z , Wang, Z X., Li, Y. W., Chang, S. H., et al. (1990) Application of microbial pesticides m IPM, in Integrated Management of Pane Caterpzllurs m China (Chen, C. J , ed.), China Forestry, BeiJing, pp. 214-308 37. Van Wmkelhoff, A. J. and McCoy, C. W. (1984) Comdiatron of Hcrsutella thompsonn var synnematosa in submerged culture J Invertebr Path01 43,59-68 38. Jenkins, N. E. and Prior, C. (1993) Growth and formation of true comdia by Metarhizrumjlavovzrtde m a simple liquid medmm Mycol Res 97, 1489-1494, 39 De La Torre, M. and Cardenas-Cota, H. M. (1996) Production of Paeczlomyces fimosoroseus comdia m submerged culture Entomophaga 41,443-453. 40 Hegedus, D D., Bidochka, M J., Miranpuri, G. S , and Khachatourians, G. G (1992) A compartson of the virulence, stability and cell-wall-surface characteristics of three spore types produced by the entomopathogeruc fungus Beauverta basstana Appl Mtcrobtol. Btotechnol 36,785-789 41 Yin, F M , Chen, Q. C , Chen, Y. G., Guo, G. L., and Li, Z. W (1988) Studies on the submerged culture of Beauveria basstana conidia, m Study and Appltcatton of Entomogenous Fungi zn Chtna, vol 1 (Lt, Y W., Li, Z Z , Liang, Z Q , Wu, J. W., Wu, Z. K., and Xu, Q. F., eds.) Academic Periodical, BeiJing, pp. 105-I 10

262

Wraight and Carruthers

42 Hegedus, D D., Brdochka, M J , and Khachatourrans, G G. (1990) Beauverza basszana submerged comdta productton m a defined medium contammg chltm, two hexosamines or glucose Appl Mtcrobzol Btotechnol 33,64 i-647 43 Lrsansky, S. G. and Hall, R A (1983) Fungal control of insects, in Fzlamentow Fungz, vol 4 Fungal Technology (Smith, J E , Berry, D R , and Krtstransen, B , eds ) Edward Arnold, London, pp. 327-345 44 Kawakamr, K and Mrkuna, T (1965) Effects of relative humidity and temperature on the vrabrlrty of comdta of some muscardmes Actu Serzcol 56, 42-46 45 Clerk, G. C. and Madehn, M. F. (1965) The longevity of comdta of three msectparasrtrzmg Hyphomycetes Trans Br Mycol Sot 48, 193-209. 46 Daoust, R A and Roberts, D W. (1983) Studies on the prolonged storage of Metarhtzzum amsopltae cotndra effect of temperature and relative humidity on comdtal vtabthty and vmrulence against mosqurtoes J Znvertebr. Path01 41, 143-l 50 47 Mrller, D W (1995) Commercral development of entomopathogemc fungr formulation and delivery, m Btoratzonal Pest Control Agents Formulatton and Delzvery (Hall, F. R. and Barry, J W , eds ), American Chemical Society, Washington, DC, pp 2 13-220 48 Couch, T. L and Ignoffo, C M (1981) Formulatron of insect pathogens, m Mtcrobtal Control ofPests and Plant Dtseases 1970-1980 (Burges, H D , ed ), Academic, London, pp. 499-5 12 49 McCoy, C W (1990) Entomogenous fungi as mrcrobral pesticides, m New Dwecttons zn Bzologzcal Control (Baker, R R and Dunn, P E , eds ), Lrss, New York, pp 139-159 50 Reismger, 0 , Fargues, J., Robert, P., and Arnould, M -F. (1977) Effet de l’argrle sur la conservation des mrcroorgamsmes I Etude ultrastructurale de la brodegradatron dans le sol de I’Hyphomyctte entomopathogene Beauverta basstana (Bals ) Vu111 Ann Mtcrobtol (Inst Pasteur) 128, 271-287 51 Soper, R. S. (1982) Commerctal mycomsecttcrdes, m Invertebrate Pathology and Mtcrobtal Control, Proceedtngs of the IIIrd Internattonal Colloqutum on Invertebrate Pathology, Brrghton, UK, pp 98-102 52 Soper, R S. and Ward, M G (1981) Productron, formulatton and apphcatron of fungi for insect control, m Btologtcal Control m Crop Productton, BARC Symposzum No 5 (Papavrzas, G C , ed.), Allanheld, Osmum, Totowa, NJ, pp 161-180 53 Ward, M G and Roberts, D W (198 1) Viabrlrty of Beauverza basszana comdra stored with formulation carriers and drluents, m Soctetyfor Invertebrate Pathology Program and Abstracts, Bozeman, MT, pp 30,3 1 54 Daoust, R A., Ward, M. G , and Roberts, D. W (1983) Effect of formulatton on the viabthty ofMetarhiztum anisoplzae comdta J Znvertebr Path01 41, 15 l-160 55 Alves, S B , Srlverra Neto, S , Pererra, R M , and Macedo, N (1987) Estudo de formulacdes do Metarhtztum anrsopltae (Metsch.) Sorok em dtferentes condrcdes de armazenamento. Ecosststema 12,78-87 56 HaJek, A E., Soper, R. S , Roberts, D. W., Anderson, T. E., Btever, K D , Ferro, D N , LeBrun, R A , and Starch, R. H. (1987) Fohar appltcatrons of Beauverta basszana (Balsamo) Vutllemm for control of the Colorado potato beetle,

Mycoinsecticides and Field- Crop Pests

263

Lepttnotarsa decemlzneata (Say) (Coleoptera. Chrysomelidae) an overvtew of pilot test results from the northern Umted States. Can Entomol 119, 959-974 57 Zhang, A , Lm, W., Nong, X., Deng, C , Gou, W., and Jang, B. (1992) A trralproductron of wettable powder of Beauverta basslana. Chznese J Btol Control 8,118-120 58 Prior, C , Jollands, P., and le Patourel, G (1988) Infectivtty of or1 and water formulattons of Beauverza basszana (Deuteromycotma. Hyphomycetes) to the cocoa weevil pest Pantorhytes plutus (Coleoptera. Curcuhonidae). J fnvertebr Path01 52,66-72

59 Stathers, T E., Moore, D , and Prior, C (1993) The effect of dtfferent temperatures on the viabrhty of Metarhtzzumflavoviride comdra stored m vegetable and mmeral 011s J lnvertebr Path01 62, 11 l-1 15 60 Bell, J V (1975) Vtabtlity of entomopathogemc fungi stored outside J Georgta Entomol

Sot 10, 357,358

6 1. Ym, F M. (1983) Storage of conidral powder ofBeauverra basstuna. Guangdong Forestry Newsletter 314, 13-l 6 62 Shr, Z M. (1988) Technology for comdial preparation of Beauverta bassrana, m Study andAppircatton ofEntomogenous Fungt in China, vol. 1 (Li, Y W , Li, Z Z , Liang, Z Q., Wu, J. W , Wu, Z K., and Xu, Q F , eds ), Academic Periodical, BelJing, pp. 114,115 63. McClatchre, G. V , Moore, D., Bateman, R. P., and Prior, C. (1994) Effects of temperature on the vtabrhty of the comdia of Metarhtzzum flavovzrzde m 011formulations Mycol Res 98, 749-756. 64. Moore, D., Bateman, R P , Carey, M , and Prior, C. (1995) Long-term storage of Metarhzztumflavovtrtde comdta m oil formulations for the control of locusts and grasshoppers Btocontrol Scr Technol 5, 193-199 65 Hedgecock, S., Moore, D , Hrggms, P M , and Prior, C (1995) Influence of moisture content on temperature tolerance and storage of Metarhmumflavovzrtde conidia m an oil formulatron Btocontrol Set Technol 5,37 l-377 66 Moore, D , Douro-Kpmdou, 0. K., Jenkins, N E , and Lomer, C J (1996) Effects of moisture content and temperature on storage of Metarhtztum flavovtrtde comdra. Btocontrol Scz Technol. 6,5 l-6 1, 67 Marques, E J and Alves, S B (1996) Ottmizacao de formulacbes na preservacao de conidios de Beauverta basstana (Bals ) Vu111 e Metarhzzzum antsoplzae (Metschn ) Sorok. em diferentes condrcdes de armazenamento. Arq Btol Tecnol 39, 861-877 68 Daoust, R A and Roberts, D W. (1983) Studies on the prolonged storage of Metarhtztum anzsopliae comdia: effect of growth substrate on comdial survival and vn-ulence agamst mosqurtoes. J Invertebr Pathol. 41, 16 l-l 70 69 Hallsworth, J E and Magan, N (1994) Improved brological control by changing polyols/trehalose m corndla of entomopathogens, m Brighton Crop Protectzon Conference Pests and Dzseases, British Crop Protection Council, Farnham, UK, pp 1091-1096 70. Jaronski, S. T. (1997) New paradigms m formulatmg mycomsecttcides, m Pestzctde Formulatrons andAppltcatton Systems, vol. 17, ASTMSTP 1328 (Goss, G R ,

Wraight and Carruthers

264

Hopkmson, M. J., and Collms, H M., eds.), Amertcan Soctety for Testmg and Materials, Philadelphia, pp 99-l 13 71. Roberts, D. W. and Campbell, A. S. (1977) Stability of entomopathogemc fungt. Mtsc Pub1 Entomol

Sot Am 10, 1%76

72. Blachbre, H , Calvez, J., Ferron, P., Corrieu, G , and Peringer, P (1973) Etude de la formulation et de la conservatton d’une preparatton entomopathogene a base de blastospores de Beauverta tenella (Delacr Siemaszko) Ann Zoo1 Ecol Anim 5,69-79.

73 Peretra, R M and Roberts, D. W (1990) Dry mycelmm preparations of entomopathogentc fungi, Metarhmum amsophae and Beauverta basslana. J Invertebr Pathol

56,3946

74. Jackson, M A , McGmre, M R , Lacey, L A , and Wratght, S P (1997) Ltquid culture productton of desiccation tolerant blastospores of the biomsecttcide fungus Paectlomyces fumosoroseus. Mycol Res 101,35-41. 75 Knudsen, G R , Johnson, J. B., and Eschen, D J (1990) Algmate pellet formulation of a Beauveria basstana (Fungi Hyphomycetes) isolate pathogenic to cereal aphtds J Econ Entomol 83,2225-2228 76 Schwarz, M R (1995) Metarhzzzum anzsoplzae for so11pest control, m Btorattonal Pest Control Agents Formulatton and Delivery (Hall, F R and Barry, J W , eds ), American Chemical Society, Washmgton, DC, pp. 183-l 96. 77 Agudelo, F and Falcon, L. A (1983) Mass production, infectivity, and field apphcation studies with the entomogenous fiqus Paectlomycesfarmosus. J Invertebr Path01 42, 124-132. 78 Morley-Davies, J., Moore, D., and Prior, C (1996) Screening of Metarhzzrum and Beauverta spp conidia with exposure to simulated sunlight and a range of temperatures. Mycol. Res 100,3 l-38 79 Goettel, M. S. and Jaronsh, S T. (1997) Safety and registration of mtcrobtal agents for control of grasshoppers and locusts. Mem Entomol Sot. Can , 171,83-99. 80 Wrtght, J. E and Chandler, L. D. (1995) Biopesticide composttion and process for controlling insect pests US Patent 5413784 81 Wright, J E. and Chandler, L D. (1996) NaturahsTM, a biopesticide (Beauverza basszana JW-1) for control of economic insects m field crops, vegetables, ornamentals and greenhouses, with emphasis on control of Bemtsta, m Proceedtngs, XXInternattonal Congress of Entomology, Firenze, Italy, p 698. 82. Moore, D , Bridge, P D., Higgms, P. M , Bateman, R. P., and Prior, C (1993) Ultravtolet radiation damage to Metarhzzrumflavovrride comdia and the protection given by vegetable and mineral oils and chemical sunscreens Ann AppZ Btol 122,605-6 16

83 Inglis, G D , Goettel, M S., and Johnson, D. L. (1995) Influence of ultraviolet light protectants on perststence of the entomopathogenic fungus, Beauverza basstana Btol Control 5, 58 l-590. 84 Bateman, R. P , Carey, M., Moore, D., and Prtor, C. (1993) The enhanced mfectrvrty of Metarhtzzum jlavovzrtde m oil formulations to desert locusts at low humidmes Ann Awwl Bzol 122. 145-152

Mycomsecticides

and Field-Crop

Pests

265

85 Bateman, R., Carey, M., Batt, D., Prtor, C., Abraham, Y., Moore, D , Jenkms, N , and Fenlon, J (1996) Screening for virulent isolates of entomopathogemc fungt against the desert locust, Schzstocerca gregurza (Forskdl) Bzocontrol Scz Technol 6,549-560. 86 Prior, C , Carey, M., Abraham, Y. J., Moore, D., and Bateman, R. P (1995)

Development of a btoassay method for the selection of entomopathogemc fungi vuulent to the desert locust, Schzstocerca gregurza (Forskal) J Appl Entomol 119,567-573. 87 Mtlner, R. J., Staples, J. A., and Prtor, C. (1996) Laboratory susceptibility of Locusta mtgratorra (L ), Austracrzs guttulosa (Walker) and Valanga trregularts (Walker) (Orthoptera: Acrididae) to an oil formulation of Metarhizium flavovtnde Gams and Rozsypal (Deuteromycotma. Hyphomycetes) Awt J Entomol 35, 355-360. 88 Milner, R. J., Baker, G L , Hooper, G. H S., and Prior, C (1997) Development of a mycoinsecticrde for the Australian plague locust, III New Strategtes zn Locust Control (Krall, S , Pevelmg, R , and Ba Diallo, D , eds.), Birkhauser Verlag,

Basel, Switzerland, pp 177-183. 89. Jenkins, N. E. and Thomas, M. B. (1996) Effect of formulatton and application method on the efficacy of aerial and submerged comdta of Meturhzzzum jlavovtrtdae for locust and grasshopper control. Pestic. Set 46, 299-306. 90 Inglis, G D., Goettel, M S., and Johnson, D L. (1993) Perststence of the entomopathogemc fungus, Beauverta busstuna, on phylloplanes of crested wheatgrass and alfalfa B~ol. Control 3,258-270. 91 Wraight, S P., Bradley, C A., and Jaronski, S. T. (1996) Comparattve field efficacy of wettable powder and oil-based formulations of Beauverta basstana applied against Bemista argenttfolu, m Proceedings 29th Annual Meeting ofthe Society for Invertebrate Pathology, Cordoba, Spam, p. 92. 92. Inghs, G. D., Johnson, D. L., and Goettel, M S (1996) Effect of bait substrate and formulation on infection of grasshopper nymphs by Beauveria busstana Btocontrol

SCI Technol 6,35-50.

93. Krueger, S. R., Villam, M. G., Martins, A. S., and Roberts, D. W. (1992) Efficacy of so11applications of Metarhtztum amsopltae (Metsch ) Sorokm conidia, and standard and lyophilized mycelial particles against scarab grubs. J Znvertebr Path01 59, 54-60

94. Johnson D. L and Goettel, M. S (1993) Reduction of grasshopper populattons followmg field application of the fungus Beauveria basstana Btocontrol SCI Technol. 3, 165-175. 95 Keller, S (1992) The Beauverza-Melolontha proJect: experiences with regard to locust and grasshopper control, m Btologtcal Control of Locusts and Grasshoppers, (Lomer, C J. and Prior, C , eds ), CAB International, Wallmgford, UK, pp 279-286 96. Rtba, G (1984) Application en essatsparcellaires de plem champ d’un mutant artiliciel du champignon entomopathogene Beauverza busslana (Hyphomyctte) contre la pyrale du mats, Ostrima nubtlalu (Lep : Pyrahdae). Entomophaga 29,41-48.

266

V/might and Carruthers

97 Labatte, J , Meusmer, S , Migeon, A, Chaufaux, J , Couteaudter, Y , Rtba, G , and Got, B. (1996) Field evaluation of and modeling impact of three control methods on the larval dynamtcs of Ostrnna nub&&s (Leptdoptera Pyraltdae) J Econ Entomol 89,852-862 98 Vercambre, B , Guillon, M , Goebel, R , Neuveghse, C , Robert, P , and Riba, G (1996) IPM strategy for the control of Hoplochelus margznalzs (Co1 Melolonthmae) from field application to mdustrtal process and commercial patent, m Technology Transfer zn Blologlcal Control From Research to Practice (Silvy, C , ed ), IOBC wprs Bulletln/Bulletm OILB srop , 19, 275 99 Bullard, G , Pulsford, D , and Rath, A. C. (1993) BtoGreen-A new Metarhzzlum anzsophae product for the control of pasture scarabs m Australia, m Society for Invertebrate Pathology Program and Abstracts, Asheville, NC, p 38 100. Lmg, A J and Donaldson, M D (1981) Biotic and abiotic factors affecting stabihty of Beauvena bassrana comdta m soil J Invertebr Path01 38, 19 I-200 101 Fargues, J , Retsinger, O., Robert, P. H., and Aubart, C. (1983) Btodegradatton of entomopathogemc Hyphomycetes: influence of clay coating on Beauverza basszana blastospore survival m soil. J Invertebr Path01 41, 13 l-142 102 Gaugler, R , Costa, S D , and Lashomb, J (1989) Stability and efficacy of Beauverla basslana sot1 maculations Environ Entomol l&41 2-4 17 103 Studdert, J P and Kaya, H K (1990) Water potential, temperature, and claycoating of Beauverla basslana comdta. effect on Spodoptera exzgua pupal mortahty m two soil types J Invertebr Pathol 56,327-336 104 Villam, M G , Krueger, S. R , and Nyrop, J. P (1992) A case study of the impact of the soil environment on msect/pathogen mteractions. scarabs m turfgrass, m Use of Pathogens zn Scarab Pest Management (Glare, T. R and Jackson, T A , eds.), Intercept, Andover, MA, pp 11 l-126 105, Jackson, T A , Pearson, J. F , O’Callaghan, M , Mahanty, H K , and Willocks, M. J (1992) Pathogen to product--development of Serratla entomophzla (Enterobactertaceae) as a commercial biological control agent for the New Zealand grass grub (Costelytra zealandzca), m Use of Pathogens znScarab Pest Management (Glare, T. R and Jackson, T A, eds), Intercept, Andover, MA, pp. 191-198. 106 Lomer, C J and Prior, C , eds.(1992) Blologlcal Control of Locusts and Grasshoppers, CAB International, Wallmgford, UK 107 Mathews, G A. (1992) The prmctples of ultra-low volume spraymg m relation to the apphcatton of mtcrobial msectictdesfor locust control, m Bzologzcal Control of Locusts and Grasshoppers(Lomer, C J and Prtor, C , eds ), CAB Internattonal, Wallmgford, UK, pp 245-248 108 Bateman, R P (1992) Controlled droplet application of mycopesticides to locusts, m Blologlcal Control of Locusts and Grasshoppers(Lomer, C J and Prior, C , eds.), CAB Internattonal, Wallmgford, UK, pp 249-254 109 Nowierski, R M , Zeng, Z , Jaronski, S , Delgado, F , and Swearingen,W (1996) Analysis and modelmg of ttme-dose-mortaltty of Melanoplus sanguznzpes, Locusta mrgratonoldes, and Schzstoceragregarla (Orthoptera. Acridtdae) from

Mycoinsecticides

110

111

112

113

114

115

116

117

118 119

120

121

and Field-Crop

Pests

267

Beauverta, Metarhmum, and Paectlomyces isolates from Madagascar J lnvertebr Path01 67,236-252 Douro-Kpmdou, 0 -K , Godonou, I , Houssou, A , Lomer, C J , and Shah, P A (1995) Control ofZonocerus vartegatus by ultra-low volume apphcation of an oil formulation of Meturhzzzumflavovtrtde comdta. Btocontrol Scz Technol 5, 13 l-l 39. Thomas, M B , Wood, S. N , Langewald, J , and Lomer, C J (1997) Persistence of Metarhmumflavovzrtde and consequences for biological control of grasshoppers and locusts Pesttctde Set 49,47-55 Kooyman, C. and Godonou, I. (1997) Infection of Schutocerca gregarta (Orthoptera Acrididae) hoppers by Metarhtztum flavovtrtde (Deuteromycotma Hyphomycetes) comdia m an oil formulation applied under desert conditions Bull Entomol Res 87, 105-107 Lomer, C. J., Prior, C , and Kooyman, C (1997) Development of Metarhlzlum spp for the control of locusts and grasshoppers Mem Entomol Sot Can , 171, 265-286 Lobo Ltma, M L., Brtto, J M , and Henry, J. E (1992) Biological control of grasshoppers m the Cape Verde Islands, m Btologtcal Control of Locusts and Grasshoppers (Lomer, C J and Prior, C., eds.), CAB International, Wallmgford, UK, pp 287-295 Johnson, D L., Goettel, M S , Bradley, C., van der Paauw, H , and Maiga, B (1992) Field trials with the entomopathogemc fungus Beauverta basstana against grasshoppers m Mall, West Africa, July, 1990, m Biologtcal Control of Locusts and Grasshoppers (Lomer, C. J. and Prior, C , eds ), CAB International, Wallmgford, UK, pp 2963 10 Nasseh, 0 M , Freres, T , Wilps, J , Kirkilioms, E., and Krall, S (1992) Field cage trials on the effects of enriched neem oil, Insect growth regulators and the pathogens Beauverta basstana and Nosema locustae on desert locusts m the Republic of Niger, m Btologtcal Control of Locusts and Grasshoppers (Lomer, C J. and Prtor, C., eds ), CAB International, Wallmgford, UK, pp 3 1 l-320 Inglis, G D., Johnson, D L , and Goettel, M S (1997) Field and laboratory evaluation of two comdia batches of Beauverza bassrana (Balsamo) Vudlemm agamst grasshoppers Can Entomol 129,17 l-l 86 Carruthers, R I., Larkm, T S , Ftrstencel, H , and Feng, Z (1992) Influence of thermal ecology on the mycosis of a rangeland grasshopper Ecology 73,190-204 Inglis, G D , Johnson, D L , and Goettel, M. S (1996) Effects of temperature and thermoregulation on mycosis by Beauverza basstana m grasshoppers. Btol Control 7, 13 1-139. Inglis, G D , Johnson, D L., Cheng, K -J , and Goettel, M S. (1997) Use of pathogen combmations to overcome the constramts of temperature on entomopathogemc Hyphomycetes agamst grasshoppers. Bzol Control 8, 143-152 Sanyang, S and Van Emden, H F. (1996) The combined effects of the fungus Metarhtzrum flavovtrtde Gams and Rozsypal and the msecticide cypermethrm on Locusta migratoria migratorioides (Reiche and Fairmane) m the laboratory Int J. Pest Management 42. 183-l 87

268

Wraight and Carruthers

122 Boucias, D. G , Stokes, C , Storey, G., and Pendland, J. C. (1996) The effects of tmtdacloprtd on the termite Retzculttermesflavzpes and Its mteractton with the mycopathogen Beauverta basstana PjTanzenschutz-Nachrtchten Bayer 49, 103-144. 123. Stemkraus, D C and Tugwell, N. P. (1997) Beauverta basstana (Deuteromycotma. Momhales) effects on Lygus lzneolarzs (Hemiptera Mmdae) J Entomol SCI 32,79-90. 124 Qumtela, E D and McCoy, C W (1997) Pathogemcity enhancement of Metarhtzium antsopltae and Beauveria basstana to first mstars of Dtaprepes abbrevzatus (Coleoptera Curculionidae) with sublethal doses of imtdacloprid Envtron. Entomol. 26, 1173-l 182. 125 Shaptro, M. (1992) Use of optical brighteners as radtatton protectants for gypsy moth (Lepidoptera. Lymantrndae) nuclear polyhedrosts vtrus. J Econ Entomol 85, 1682-1686. 126 Morales, E and Knauf, W ( 1994) Comdia@, a formulation of Beauvena bassrana adapted for integrated pest management programs, m Proceedings of the Vlth Internattonal Colloqutum on Invertebrate Pathology and Mtcrobtal Control, vol 2, Abstracts, Soctety for Invertebrate Pathology, Montpelher, France, p. 87.

127 Lorence, A. (1996) Cuadernos de Vigilancta Biotecnologrca 1 Biopestrcldas CamBtoTec, Mextco. 128. Keller, S., Schweizer, C , Keller, E , and Brenner, H. (1997) Control of whtte grubs (Melolontha melolontha L ) by treating adults with the fungus Beauverta brongntarttt Btocontrol Scr Technol 7, 105-l 16. 129 Poprawskl, T J and Wraight, S. P. (1998) Fungal pathogens of the Russian wheat aphtd, m A Response Model for an Introduced Pest-The Russian Wheat Aphid (Quisenberry, S S. and Peatrs, F. B., eds ), Thomas Say Pubhcattons m Entomology, Entomologrcal Society of America, in press. 130 Pell, J K., Macaulay, E D. M., and Wtldmg, N (1993) A pheromone trap for dispersal of the pathogen Zoophthora radzcans Brefeld. (Zygomycetes. Entomophthorales) amongst populations of the dtamondback moth, Plutella xylostella L (Leptdoptera Yponomeuttdae). Btocontrol Set Technol 3,3 15-320. 131 Lacey, L A , Amaral, J. J., Coupland, J., Klein, M G , and Stmdes, A M (1995) Flight activity of Popollta Japonzca (Coleoptera Scarabaetdae) after treatment with Metarhtztum anuopltae. Btol Control 5, 167-172 132 Carruthers, R. I., Wraight, S P , and Jones, W A. (1993) An overvtew of biological control of the sweetpotato whitefly, Bemtsta tabact, m Proceedings of the Beltwzde Cotton Conferences, vol 2 (Herber, D. J. and Richter, D. A., eds.), National Cotton Counctl of America, Memphts, TN, pp 68fl-685 133 Wralght, S P., Carruthers, R I , and Bradley, C A (1996) Development of entomopathogemc fungi for mtcrobial control of whiteflies of the Bemtsza tabact complex m VSICONBIOL, Stmpdsto de Controle Btoldgtco, Anais Conferenctas e palestras, Foz do Iguacu, Braztl, pp. 28-34. 134. Wraight, S. P and Bradley, C. A. (1996) Production, formulatton, and apphcatton technologtes for use of entomopathogenic fungt to control field crop pests,

Mycoinsecticides

and Field-Crop

Pests

m V SICONBIOL, Stmpdsio de Controle Btoldgtco, Palestras, Foz do Iguacu, Brazil, pp 170-177.

269 Anats*

Conferenctas

e

135. Wraight, S. P. and Bradley, C. A (1996) Use of Mycotrol@ (Beauverza basszana) for btologtcal control of Bemzsta whitefltes m field crops, m Memona Stmposzum de Control Bioldgico de Mosquita Blanca, Culiacan, Mexico, pp. 29-33 136 Wrarght, S. P and Bradley, C A. (1996) Use of Mycotrol@’ (Beauverza basstana) for control of Bemzsia argentzfolii (Homoptera: Aleyrodtdae) m field crops, m Technology Transfer tn Btologtcal Control. From Research to Practtce (Silvy, C., ed.), IOBC wprs Bull /Bull OILB srop 19, 15 1 137 Jaronski, S. T., Lord, J C , and Paden, R (1996) Evaluatton of Beauverza basszana (Mycotrol@ WP) with pyrethroids for control of whitefly m spring cantaloupes Arthropod Manage Tests 1996 21, 102,103 138. Poprawski, T. J., Carruthers, R. I., Speese, J., Vacek, D. C., and Wendel, L E. (1997) Early-season applications of the fungus Beauverta basszana and mtroductton of the hemrpteran predator Perillus bzoculatus for control of Colorado potato beetle. Biol Control 10,48-57 139 Jaros-Su, J (1997) Effects of certain fungrctdes on Beauverza basslana-induced mortahty of the Colorado potato beetle (Coleoptera: Chrysomelidae), M. S. Thesis, Univ. of Maine, Orono, ME. 140 St Leger, R. J , Lokesh, J., Btdochka, M. J., and Roberts, D W. (1996) Construction of an Improved mycomsectictde overexpressmg a toxtc protease Proc Natl. Acad Set USA 93,6349-6354.

141 Hallsworth, J E and Magan, N. (1995) Mampulatron of mtracellular glycerol and erythrttol enhances germinatron of conidia at low water availabrhty A4mvbtology 141, 1109-l 115

15 Entomopathogenic Parwinder

Nematodes

Grewal and Ramon Georgis

1. Introduction Entomopathogemc nematodes of the genera Steinernema and Heterorhabdztis (Nematoda: Rhabdltlda) have emerged as excellent Insect biological control agents. This dlsclplme of insect pathology has made enormous strides since Glaser’s discovery more than 60 yr ago of nematodes mfestmg white grubs (1,2). Recent advances in mass-production and formulation technology, and the discovery of numerous isolates/strains, together with the deslrablhty of reducmg pesticide usage, has resulted m a surge of scientific and commercial interest in these Insect-killmg nematodes (3) This has culminated m the commercial availability of many nematode products for use m several medlumand high-value markets (4). This chapter will briefly review nematode biology, mass production, formulation, quality, apphcatlon strategy, and field efficacy. 2. Nematode Biology The parasitic cycle of nematodes IS initiated by the third-stage infective Juveniles (IJs) (Fig. 1). These nonfeedingJuveniles locate and invade suitable (5) host insects through natural body openings (i.e., anus, mouth, and spiracles). Once inside the host, nematodes invade the hemocoel and release the symblotlc bacteria that are held in the nematode’s intestine. The bacteria cause a septicemia, kilhng the host withm 24-48 h. The IJs feed on the rapldly multiplying bacteria and disintegrated host tissues. About 2-3 generations of the nematodes are completed within the host cadaver. When food reserves are depleted, the nematode reproduction ceases and the offspring develop into resistant IJs that disperse from the dead host, and are able to survive m the environment and to seek out new hosts. From a commercial standpoint, the production of a durable infective stage and the symbiotic assoclatlon with a From Methods m Efotechnology, vol 5 Blopestmdes Use and De//very Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

271

Grewal and Georgis

272

-pIj$$

Xe”orhabd”s

Nematodes invade wa natural body openmgs or through the Insect’s Cuticle

\ Infectwe nematodes and search for new

SDD bacterta released Unsecc dtes I” 24-48

Dead host may change colour Paraslles develoo mto ‘qtant’ adults

mlgfata hosts

Nematodes’ food suooly reduces Non-feedmg Infective stages start to be produced

Fig 1 Life cycle of Steznernema nematodes Note that the entire nematode life cycle 1s completed within one host. Only mfecttve juvenile (IJ) nematodes (“dauer larvae”) leave the Insect cadaver to seek new hosts

lethal bactertum are the two most attractive features of stemernemattd and heterorhabdtttd nematodes. The IJs are capable of toleratmg stresses fatal to other developmental stages, and, therefore, can be effectively formulated and preserved for several months and shipped for application. The IJs are produced when conditions instde the insect cadaver are no longer suitable for further reproductton, and are the only stage m the nematode life cycle that leaves the host to infect a new host The ventrtcular portton of the tntestme of the stemememattd IJ ts spectfically modified for storage of symbiotic bacteria, and 1scalled an intestmal vesicle (6,7), In the infective stage of heterorhabditid nematodes, symbiotic bacteria are located m the esophagus and m the ventrtcular portion of the mtestme (8). The symbrottc association of entomopathogemc nematodes with spectfic bacteria facthtates rapid mass productton of the nematodes (bacteria serve as food) and successful pathogemctty. Although axenic nematodes (nematodes without bacteria) may occasionally cause host death, they do not generally reproduce. Furthermore, bacteria alone are incapable of penetrating the ahmentary tract and cannot independently gain entry to the host’s hemocoel.

Entomopathogenic

Table 1 Steinernema and Heterorhabditis and Their Respective Symbiotic Nematode genus Stetnernema

Heterorhabdttts

273

Nematodes Species Bacterial Species

Nematode species affirm

anomalt carpocapsae cubana feltiae glasert tntermedta kushidat krausset longtcaudatum neocurtillis oregonensis puertoricensts rara rtobravts rittert scapterisci Undescrrbed argentmensts bactertophora hawattensts wdlcus

marelatus megtdts zealandtca

Bacteria genus Xenorhabdus

Photorhabdus

Bacterra species bovtenn Undescrrbed nematophtlw Undescrrbed bovtenn potnartt bovtenn Undescrtbed bovtentt Undescribed Undescrlbed Undescrrbed Undescrrbed Undescrrbed Undescrrbed Undescrlbed Undescrrbed beddtngtt Undescrrbed lumtnescens Undescrtbed Undescrrbed Undescrrbed luminescens lumtnescens

Thus, nematodes act as vectors to transport the bacteria mto a host within whtch they can proliferate, and the bacteria create condttrons necessary for nematode survrval and reproductton within the insect cadaver. All species of Steznernema are associated with bacteria of the genus Xenorhabdus, and all Heterorhabdztzs nematode species are associated with Photorhabdus bacteria (9). Each nematode species has a specific natural associatton with only one bacterial species, though any one bacterial species may be associated with more than one nematode species (Table 1). The specrficity of assocratton between nematodes and bacteria operates on three levels (IO): provision of factors that enhance IJ recovery from the dauer phase; provtsron of essential nutrients for the nematode by the bacterium; and the retention of the bacterium within the intestine of the nonfeeding IJ

274

Grewal and Georgrs

Nematode species differ m then foraging behavior (21,12), temperature activity ranges (23), mtrmstc metabolic rates (P. S Grewal, unpublished data), stored energy reserves (14,15; P. S. Grewal and R. Georgis, unpublished data), destccatton tolerance (P, S. Grewal, unpublished data), and sizes(16; see Table 2). These differences have important imphcattons m the development of nematode production systems,storageprotocols, formulattons, and apphcatton strategy 2.1. Strain Discovery Stemernematids and heterorhabdttids are ubtqmtous m dtstrtbutton and have been recovered from soils throughout the world (II). Although some entomopathogenic nematodes have been isolated from insects naturally infected m the field, they are most commonly recovered from sot1by baiting with susceptible msects (17). The wax moth larva Gallerla mellonella IS most commonly used as a bait. Nematode strains can be maintained by repeated subcultures through G mellonella larvae, or can be cryopreserved m liquid nitrogen (l&19) Sixteen species of Stelnernema and SIX of Heterorhabdltls have been described from vartous insects and/or from the soil worldwtde (Table 1) 2.2. Host Range The nematode-bactermm complex ktlls msects so rapidly that the nematodes do not form the intimate, highly adapted, host-parasite relattonshtp characteristic of other insect-nematode assoctattons, e g., mermtthtds This rapid mortality permits the nematodes to exploit a range of hosts that spans nearly all insect orders-a spectrum of activtty well beyond that of any other microbial control agent. In laboratory tests, S. carpocapsae alone infected more than 250 species of insects from over 75 families m 11 orders (20). The nematodes attack a far wider spectrum of insects m the laboratory, where host contact IS assured, environmental conditions are optimal, and no ecological or behavioral barriers to mfection exist (11,21). For example, foliage feedmg leptdopteran larvae are highly susceptible to infection m Petri dishes, but are seldom impacted m the field, where nematodes tend to be quickly inactivated by the environmental extremes (i.e., desiccation, radiation, temperature) charactertstic of exposed foliage. Behavioral barriers also restrict nematode efficacy to a few selected hosts or host groups (21). Some nematode species search for hosts at or near the soil surface (e.g., S carpocapsae and S scapterrscz); others are adapted to search deeper m the soil profile (e.g., H. bacterzophora and S. glaserz) (12). The former group has been referred to as “ambusher,” which remains nearly sedentary while waiting for the mobile surface-dwelling hosts (22). The latter group has been referred to as “cruiser,” which is htghly mobile, responds strongly to long-range host chemtcal cues, and is therefore best adapted to find sedentary hosts (12,23).

43%650 736-950 864-1448 561-701 517-609 512-671 73C800

x20-30 x22-29 x31-50 x2630 x 18-30 x l&31 x27-32

lS2.5 45-65 90-120 45-60 18-32 12-18 45-70

spp

lo-32 S-30 l&37 lo-39 lo-35 10-32 l&35

Infectivity

20-30 IO-25 1532 20-35 2&32 15-30 12-25

Reproduction

Temperature rangec

and Heferorhabdifis

“(pm) Adaptedfrom refs. 16 and126 *mgper mllhonmfectlveJuveniles(P S GrewalandR Georgls,unpubhshed data) ‘(“C) Adaptedwith permisslonfrom ref. 13 dAdaptedwith permlsslon from ref. 22

S carpocapsae S feltiae S glaseri S riobraws S scaptensci H bacterlophora H megidls

Lipid contentb

Species

Length x width0

Steinernema

Table 2 Size, Lipid Content, Temperature Range, and Foraging Strategy of Infective Juvenile

Sit and wait Intermediate Widely ranging Intermediate Sit and wait Widely ranging Widely ranging

Foraging strategyd

Grewal and Georgis

276 3. Mass

Production

Entomopathogemc nematodes can be mass-produced by m vtvo or m vitro methods. In the m viva process, an insect serves as a bioreactor; m the in vitro process, artificial media are used. The m viva process is very simple and requires only mmlmal n-ntial investment. The equipment used IS also simple: trays and shelves. The wax moth larvae are most commonly used to rear nematodes because of their commercial availabihty. The methods of nematode mfection, maculation, and harvesting have been previously described (24-27) Using the m viva process, yields between 0.5 x lo5 and 4 x lo5 IJs/larva, depending on the nematode species,have been obtained (13,24,28). Durmg the past few years, a distmct cottage industry has emerged that utilizes the m vivo process for nematode mass production for sale, especially m the home lawn and garden markets. However, the m viva process lacks any economy of scale; the labor, equipment, and material (insect) costs increase as a linear function of production capacity. Perhaps even more important is the lack of improved quality while increasing scale The m vivo nematode production is increasmgly sensitive to biological variations and catastrophes as scale increases (29). As early as 193 1, Rudolf Glaser recognized the value of developing artificial culture methods for entomopathogemc nematodes, and devised the first such method for S glaserz (I). However, Glaser was unaware of the sigmticance of symbiotic bacteria m the nutrition and pathogemcity of nematodes, which was only recognized much later (30). Therefore, the first successful commercial-scale monoxemc culture was developed by Bedding, and has come to be known as sohd culture (31,32). In this method, nematodes are cultured on a crumbed polyether polyurethane sponge impregnated with emulsified beef fat and pig’s kidneys, along with symbiotic bacteria. Using this method, approx 6 x 1O5- 10 x 1O5IJs/g of medium were achieved (32) Smce then, this method has been commercially used in Australia, China, and the United States In a scale-up model, Friedman (29) reported that the solid-culture method is economically feasible up to a production level of approx 10 x 1012nematodes/ma. Labor costs increase sigmticantly for nematode production beyond this level, making a less expensive method of large-scale production a necessity. Friedman (29) reported the development of a hqutd-fermentation technique for large-scale production of nematodes. In this method, costs of production decrease rapidly up to a capacity of approx 50 x lOi IJs/mo. This method allows consistent production of stemernematids m as large as 80,000-L fermenters. Recent improvements m the nematode fermentation and media formulation processeshave resulted m further improvements m nematode quality and yields (P. S. Grewal, unpublished data) The current yields of S carpocapsae m the hquid culture average about 2 5 x lo5 IJs/g. In addition to S

En tomopa thogenic

Nematodes

277

carpocapsae, S. nobravu, S scapterisci, S feltiae, S glasen, H bacteriophora, and H. megrdis have been produced successfully m large-scale liquid cultures. 4. Bulk Storage Followmg recovery from production substrates,the nematodes can be either stored in bulk for extended periods or formulated immediately. When the nematodes are to be stored as bulk, nematode concentration, temperature, aeration, and contamination control are important considerations for the maintenance of high viability and quality. Differences in storage stability among nematode species can be attributed to their thermal and behavioral adaptations. Each nematode species has a well-defined thermal niche (13) and an optimum temperature for the longest storage stability. For instance, S.feltiae stores better at 5°C whereas S scaptenscl, 5’.riobravis, and H. bacteriophora are more stable at 10°C (P. S. Grewal, unpublished data). The nematode species (e.g., S. carpocapsae and S. scapterisci) that adopt a quiescent posture during storage generally store better than the more active species, such as H. bactenophora and S riobravis The latter species also tend to be more prone to bacterial contamination during storage. Anttmicrobial agents may be used to suppress contamination, and the choice of the antimicrobial should be based on its safety to nematodes and symbiotic bacteria. 5. Formulation Nematode formulations are developed with two main objectives: a means of delivery of live nematodes to the customer and a means of extended product storage. In the simplest type of formulation, the nematodes are impregnated onto moist carrier substrates providing substantial interstitial spaces, leading to increased gas exchange. Such carriers include polyether polyurethane sponge, cedar shavings, peat, vermiculite, and so on. Nematodes held on the sponges need to be hand-squeezed into water before application; from the other carriers, they may be applied directly to the soil as mulch. Because of the laborintensive application, limited economy of scale, and constant refrigeration requirements, these formulations are only applicable in the small home lawn and garden markets. During the past decade, sigmficant progress was made m developing nematode formulations with improved shelf stability, scalabrhty, and ease of use. Most of these formulations were based on the one fundamental principle of conserving IJ nematodes: limited stored energy reserves by either restricting their movement (physical immobihzation) or reducing their oxygen consumption by inducing a state of partial anhydrobiosis (physiological unmobilization). The first of such formulations used activated charcoal to restrict nematode movement (33). Kaya and Nelsen (34) were the first to report on the encapsulation

Grewal and Georgs

278 Table 3

Some Commercially and Heterorhabditis

Available Formulations Containing Steinernema Nematodes with Expected Shelf Life

Formulation Algmate gels” Flowable gels” Attapulgite clay chip&’ Water-dispersiblegranules”

Nematodespecies

Shelf life (mo) Room Refrrgerated

S carpocapsae S feltlae S carpocapsae H bacterlophora

3 a.0

S feltrae

1&15 1 C&l 5

S carpocapsae S carpocapsae S feltiae S rlobravts

OP S Grewal and R Georgls (unpublished hRef. 32

0.5-l 0 1 o-l 5 1 5-2.0

40-50 1 5-2.0 2 O-3.0

6&90 4&50 3&50 0 40-60 3.w 0 90-120 50-70 40-50

data)

of entomopathogemc nematodes with calcium algmate. This discovery subsequently led to the development of a commercial nematode product that used thm sheets of calcmm algmate spread over a plastic screen to trap nematodes (35). For appltcatton, the nematodes are released from the algmate gel matrix by dissolvmg It m water with the aid of sodium citrate. The algmate-based S carpocapsae products were the first to possessroom-temperature shelf life of about 3-4 mo (Table 3), and led to an Increased acceptability of nematodes m the high and medium value niche markets. However, the time-consummg extraction steps, and the problemattc disposal of large numbers of plastic screens and containers, rendered this formulation unsuitable for large-scale use, especially m the professtonal turf mdustry. Although nematodes have been successfully formulated m gel-forming polyacramtdes (35,3/j), flowable gels (4), wheat flour (37), and attapulgtte clay chips (38), none of these formulattons offered any advantage over the algmate gels. A slgmficant advancement was made with the advent of a water-disperstble granule (WDG) m which IJs are encased m lO-20-mm diameter granules consistmg of mtxtures of vartous types of silica, clays, cellulose, hgnin, and starches (39) These granules are prepared through a conventtonal pan-granulatton process m which droplets containing a thick nematode suspenston are sprayed onto fine dry powders on a tilted rotating pan, As soon as nematode droplets come m contact with the powders, the granules start to form, and roll over the dry powders, adsorbing more powder around them, The granules are then sieved out of the powders and packaged mto shipping cartons. The granu-

Erttomopathogenic

Nematodes

279

lar matrix allows accessof oxygen to nematodes held m the center of the granules, which undergo a slow water-loss process. Under appropriate temperature regtmes, the nematodes can undergo a phystologlcal destccatton process and enter into a partial anhydrobiottc state. The development of the WDG formulations has offered several advantages over the extsting formulatrons (Table 3). It extended nematode storage stabrltty to several months at room temperature; enhanced nematode tolerance to temperature extremes, enablmg easier and less-expensive transport; Improved the ease of use of nematodes by eltmmating time-consummg and labor-mtensive preparation steps,decreasedthe contarner size/coverage ratio; and decreased the amount of drsposal material (i.e., screens and contamers). 6. Product Quality Consumer acceptance of nematode products is determined by then ease of use, efficacy, and price. The development of mass-productron technology has led to the availability of nematode products at prices comparable to the standard insecticides m many markets (Table 4). The ease of use of nematode products is constantly being improved through formulatron research. For instance, the easy-to-use WDG formulation has expanded the market potenttal of nematode products sigmticantly. Efficacy is perhaps the most Important factor determmmg quality of nematode products, and depends on many factors, but preservation of high nematode viabihty and vn-ulence durmg large-scale production and formulation are essential components of a quahty control strategy The most important production factors affecting nematode quality are the type of medium, amount and type of antifoam, bacterial phase, deltvery of oxygen, sheer stress, production and storage temperature, and contammatlon. The quality of the nematodes that survive the rigors of the manufacturing process is analyzed by determmmg then shelf life and virulence. Nematode shelf life IS predicted from storage energy reserves (e.g., dry wt or total hptd content) of IJ nematodes; vn-ulence potential IS measured usmg Insect btoassays. Nematode virulence IS the most important component of nematode quahty. Virulence/pathogenlctty can be measured by different methods, mcludmg 1.1 bioassay (40), L&s (41), mvasion or establishment efficiency (42-44), invasion rate (#5), and number of bacteria/IJ (14,&J. One-on-one bioassay is perhaps the most versatile method of virulence assessment (47). Most of the mfectivity assays using multiple nematodes and single or multiple hosts are constdered inappropriate for quahty control purposes because of the host-parasite interactton effects, such as recruitment and overdispersion of natural parasite populations. The one-on-one filter-paper assay and its modificatrons compare virulence of any nematode species with a predetermined standard agamst a very susceptible host This method measures the proportron of mfec-

gel

Water-disperstble

Clay

Flowable

granules

S nobraws

S carpocapsae S carpocapsae S feltiae

S scaptenscl S carpocapsae

S feltlae

S carpocapsae S carpocapsae S carpocapsae S feltlae S feltiae H megldls

Containing

Alginate gel

Products

Nematode species

Commercial

Formulanon

Table 4 Some Available Mioplant Boden-Nutzlmge BioSafe Exhtbn Stealth Nemasys-H LarvaNem Nemasys Entonem Proactant Ss Btosafe Btosafe-N BtoVector Vector TL Helix X-GNAT Magnet BtoVector Vector MC

Product

Sfeinernema Company

Nematodes

Novartts, Vienna, Austria Rhone-Poulenc, Celaflor, Germany SDS Biotech, Minato-Ku, Tokyo, Japan Novartts, Basel, Switzerland Novartis, Macclesfield, Chester, UK MicroBto, Cambndge, UK Koppert B.V., Berkel en Rodenngs, Netherlands MrcroBio Koppert B.V. BtoControl, Gamesvtlle, FL ThermoTnlogy, Columbta, MD ThermoTrilogy ThermoTrtlogy Lesco, Lansing, MI Novartts, Misnssauga, Canada E C Geiger, Harleysvtlle, PA Amycel-Spawn Mate, Watsonvllle, CA ThermoTrtlogy Lesco

and Heterorhabdifis

Entomopathogenic

Nematodes

281

tive nematodes in a populatton, ts sensitive to impatred nematodes, and is appropriate for nematode species that have a lethal level of one IJ/larva. New bioassay methods are required for nematode speciesthat do not ktll G. mellonella larvae at 1: 1 ratio, so that the principal thesis of this method could be applied. The 1:1 method could be used effectively to assessquality of both ambush and crutse foragers by using different bioassay arenas. The 1.1 filter-paper assays are used for ambushing nematodes, such as S, carpocapsae, and 1.1 sand column assaysare used for nematodes that utilize a cruising approach, such as S glaseri and S riobravzs (P. S. Grewal and R. Georgis, unpublished data). Another major aspect of nematode quality is the maintenance of consistent viabthty for a mmimum desired period (i.e., shelf life claim) This aspect of quality is determined by analyzmg the stored energy reserves of nonfeedmg IJ nematodes. Total dry wt and total ltpid content of nematodes are routinely used to determine batch-to-batch variation (see Fig. 2). Total lipid content, which constitutes about 40% of the dry wt of IJs, may be used as a predictor of nematode shelf hfe potential. Genetic drift or inadvertent selection because of passage effect may cause deterioration m nematode quality, In an experiment conducted at Btosys m 1991, no adverse effects on nematode dry wt, liptd content, IJ length, and pathogenicity were found after constant subculture of S carpocapsae for 2 yr. However, a decline in pathogemcity of H bacteriophora following repeated subcultures was observed. Quality cannot be compared between nematode species because of differences m host aftimtres, search behavior, bacterial species, size, and activity of nematodes. For instance, nematodes adapted to parasitize lepidopteran insects cause htgher G meZIonella mortality than those adapted to parasitize white grubs or mole crtckets. Furthermore, nematodes that utilize an ambush approach would cause higher insect mortality in filter-paper btoassays than those utthzing a crutsing strategy, which are more effective m sand columns, The rate of lipid utthzation also differs between nematode species because of differences m size and activity during storage. Therefore, differences m lipid content between species also do not always reflect proporttonate differences in shelf life or field persistence. 7. Application Strategy Entomopathogemc nematodes are used mostly m mundative biologtcal control. They have been most efficacious for insects that reside in sot1or cryptic habitats where there 1s protection from destccatton, UV radratton, and high temperature. Furthermore, best results are usually obtained when nematodes are applied to moist soil in the evening or early morning. Postapplicatton irrigation improves nematode establishment and rinses the nematodes from the

282

Grewal and Georgis

36 -

- 36

34

- 34

32 -

- 32

70

- 30

78 -

- 26 - 26 .- 24 - 22 - 20 - 18 - 16 - 14 - 12 - 10 -8 -6 -4 -2

Nematode

Batch

Ftg 2 Batch-to-batch vartatton In the total lipid content of tnfecttve Juventle (IJ) produced m liqutd culture The regresston line shows the increase m the mean ltptd content obtained wtth several tmprovements made in the fermentanon process

Stemernema carpocapsae

foliage and into then natural reservoir, the soil (48). Although nematodes are generally apphed as curative treatments, prophylactic apphcattons to the soil surroundmg seedlmgs (49-51) and seeds (52,53) have been advocated. Eidt and Weaver recommended an appltcatton of 3 x lo5 IJ 5’ carpocapsae per pure seedling prior to planting for its protection against the seedlmg debarkmg beetle Hylobzus congener (52) Creating a nematode barrier against larvae of the pecan weevil, Curculio caryae, entering the soil to pupate has been suggested (54).

7.1. Soil Application Spraying IJs directly onto the so11surface is the most common method of nematode

appltcation

This method 1s quick, simple, and provides good cover-

age. A spray volume between 750 and 1900 L/ha ts usually sufficient for most nematode species to reach the target msects in soil. Nematodes can be apphed with nearly all commercially available ground or aerial spray equtpment. These include small pressurized sprayers, mist blowers, and electrostatic sprayers, as well as the traditional sprayers used in aerial application via hellcopters (35). Nematodes

are also applied using drop and sprmkler

irrtgatton

systems. Pres-

Entomopathogenic

Nematodes

283

sures of up to 1068 kpa have no detrimental effect on IJ nematodes (55). They can pass through sprayer screens with openmgs as small as 100 pm in dtameter. Cloggmg of nozzles may occur wtth some formulatrons, therefore, manufacturers’ recommendatrons should be followed carefully. 7.2. Foiiar Application Use of nematodes to control insect pests on the fohage presents a considerable challenge, because of rapid destccation, lethal UV light, and, perhaps, difficulty in establishmg attractton gradients (56). Yet, under the right conditions (i.e., high humidity and during the early morntng or evening), S carpocapsae was effective against beet armyworms (56) and serpentine leaf miners (57) on chrysanthemums. In addition, where foliage created a cryptic habitat (e.g., buds, dense canopies, leaf minmg and rolling, and so on), nematode performance increases considerably, compared to that on exposed surfaces, Although the addition of antidesiccants and UV protectants to nematode suspensions has improved the level of insect control, in most cases,economtc control has yet to be achieved (58). 7.3. Application to Trap Crops The use of nematodes to kill insects attracted to trap crops has been attempted. In Hawatt, corn 1s used as a trap crop for the melon fly, Bacfvocera cucurbztae, on tomatoes and cucurbits (59) Chemical msectitides mixed with a protein bait are applied to the corn to kill the attracted melon flies. Unfortunately, corn 1s also a host for the corn earworm, Hellothis zea, a key pest of tomatoes, and its use as a trap crop increases local populations of the corn earworm, resulting in sigmficant damage to tomatoes. Weekly applications, for 9 consecutive wk, of S carpocapsae to silk channels of corn used as a trap crop for tomato pests elicited an 18% increase m the marketable yield of tomatoes over that obtained with an untreated trap crop (60). This labor-intensive application can be economical tn kitchen gardens and small vegetable farms. 7.4. Application to Plant Propagation Material Nematodes have also been used successfully to dismfest plant-propagation maternal. Bedding and Miller obtained 99% control of the sesnd Sjmanthedovl tlpuliformls with S.feltaae in blackcurrent cuttmgs (61). The nematodes were sprayed on the stacked cuttings, which were then placed m large bags. Bari controlled the artichoke plume moth on artichoke root stalks by dipping the sticks m a suspension contammg S. carpocapsae (62). After dippmg, the root stalks were planted m the field, with significantly lower postplantmg shoot infestation.

284

Grewal and Georgis

7.5. Use of Nematodes in insect Traps Nematodes have also been used successfully in traps destgned to lure and kill Insects, m which S carpocupsae and S. scapterlscz have shown particular promise because of their sit and wait foraging strategy. Traps wtth nematodes have been tested against immature and/or adult stages of hemtmetabolous insects, including grasshoppers, Melunoplus spp (63); tawny mole cricket, Scaptenscus vlcinus (64,65); and the German cockroach, Blatteila germanrca (66,67), and adults of holometabolus Insects, such as the house fly, Musca domestica (68,69); the banana weevil, Cosmopolites sordidus (70,71); and the yellowjacket, Vesupa spp (72,73); and larval stages of the holometabolous insect, the black cutworm, Agrotzs lpsilon (64,74). The traps may contain an attractive food source for the target insect (63,64,70-72,741, or a food arrestant and a sex pheromone (69). The traps may also serve as harborage or pupation site (66,75,76), or mating and oviposition site (70,72). The sound traps have also been used to attract flying adult mole crickets to a source of S. scapterisci nematodes (65). Adult mole crtckets, especially females, are attracted to the songs of males (77). 8. Field Efficacy Predictability of biologtcal control, a key objective m the development of sound pest-management strategies, can be achieved usmg nematodes. In strictly controlled environmental condmons that exist m glasshouse mdustrtes, nematodes have replaced aldrm as the most effective method of controllmg rootfeeding black vme weevil larvae (Otiorhynchus sulcatus) m ornamental pot plants (78). However, predictability is far more difficult to achieve in the agroecosystems (79). Progress has been made, so that nematodes are now available for large-scale use m strawberry plantations, cranberry bogs, citrus groves, artichokes, mmt, sugar beet, and turf (Table 5). In agroecosystems, sot1 moisture, temperature, and sun light (UV radiation) are the three most important factors that influence nematode efficacy (80,81); field efficacy of entomopathogemc nematodes has been reviewed recently (82-84). Following is a brief overview of nematode performance in selected markets in which they have been commercially successful, and are routinely used. 8.1. Mushrooms Sctartd flies (Lycoriella spp) are major pests of the cultivated mushroom, Agricus bisporus, worldwide (84,851. Adult flies invade mushroom houses contaming freshly pasteurized or spawned compost, and lay eggs in the compost. Emerging larvae feed on compost, destroying structure and water retention capacity and leading to inhibition of mycelial colomzatton and reduction m mushroom yields. Nematodes provide control of sciarid larvae comparable to

285

Entomopathogenic

Nematodes

Table 5 Commercialization and Heterorhabditis

of Steinernema Nematodes in Different

Segment

Nematode species

Artichoke Berries

S carpocapsae S carpocapsae

Mint

H. bacterzophora H. megldn S rlobravis S carpocapsae S feltlae H. megrdu S. carpocapsae

Mushrooms Sugar beet Pet/vet Turf

S feltiae S carpocapsae S carpocapsae S carpocapsae

Citrus Greenhouses

S nobravis S scapterisci

Market

Segments

Insect common name Artichoke plume moth Black vine weevil, crown borers, cranbeny girdler, strawberry root weevil Black vine weevil, white grubs Black vine weevil Sugarcane root-stalk borer weevil Black vine weevil Sciarid flies Stem borers, black vine weevil Cutworms, mint flea beetle, mint root borer, root weevils Sciarid flies Sugar beet weevil Cat flea Billbugs, bluegrass webworms, black cutworms, armyworms, European crane fly Mole crickets Mole crickets

the commonly used msecttctdes (84,85). Nematode applications to the casing layer are generally more effective than compost treatments (84). This 1s mostly a result of the inadequate moisture, which is rapidly depleted from compost by the growmg mushroom mycelmm. In more than 35 tests conducted at AmycelSpawn Mate facility in California, single apphcattons of 75,000 S feltxzelsq ft. of croppmg area, at, or a few days after, casing, provided 47-90% reduction m fly populations (86). However, if timed accurately to coincide with the late mstar, sigmficant fly control can be achieved with precasmg treatments (86,87). In a large-scale trial, a split nematode treatment at 4 d precasmg and 6 d postcasing produced 80% reduction in fly population (86). Dosage rate IS the most important consideration for the nematodes to have a commercial appeal. The level of fly control varies with the numbers of nematodes applied and the level of pest pressure. In large-scale tests in Pennsylvama and in the UK, single-casing applications of 93,000-140,000 nematodes/sq ft of cropping area provided an acceptable level of fly control (86,881. S. feltme provides long-lasting control of overlappmg fly generations. The third- and fourth-stage fly larvae are most susceptible to attack by nematodes The nema-

286

Grewal and Gem-g/s

todes persrst m the casing material on mushroom beds for the entire cropping cycle, and have been shown to increase mushroom yields (89). Nematodes can also recycle m the parasitized larvae (88), and, therefore, are able to cope with heavy pest pressure. Among nematode species, S feltzae provides the most effective control of mushroom scrarrds (90). S. feltzae 1seffective against all three major specresof the sctartd fly genus, Lycorrella. L. auripila (901, L solanr (91), and L malr (85,921. Phorid and cectd fly larvae, which also occastonally cause stgmficant damage to mushroom crops, are also suscepttble to mfectron by entomopathogenic nematodes (90). Scheepmaker et al (87) observed 74% reductton m the F-2 generation of the phorid fly, Megasella halteratu, with an apphcanon of S feltzae made 1 wk after casmg. 8.2. Glasshouse Crops 8.2.1. Black Vine Weevil Black vme weevil, 0 sulcatus, larvae are major root pests of glasshouse and nursery plants worldwide (93). Entomopathogenrc nematodes have been very successful u-rcontrollmg black vine weevil larvae m ornamental plants m pots, greenhouses, and nurserres A strain of H bacterzophora provided up to 100% control of larvae m freshly potted yew, raspberrtes, and grapes, and over 87% control m cyclamens and strawberries tn Australia (61). Similar results have been obtained m greenhouses in The Netherlands (94,95), United States (9698), Canada (99), England (78), Poland (IOO), and Belgium (101). 8.2.2. Fungus Gnats Fungus gnats m the genus Bradysia are considered major pests of greenhouse ornamentals (102,103). Fungus gnat larvae damage cuttmgs of various ornamentals and reduce root weight and vigor of a wide range of ornamentals Larval feeding IS believed to predispose the plants to attack by pathogemc fungt (104), and the adult scrarids spread several phytopathogens (105). Nematodes have emerged as effcactous and economical replacements for chemical msectrcides for use against fungus gnats in the flortculture Industry m The Netherlands (106), England (107), and Umted States (105,108). Levels of control achieved by the nematodes have usually been equivalent to the most commonly used organophosphate insectrcides (107, S. Gtll, 1993, unpubltshed data; M. Tomalak, 1992, unpublished data). S. feltiae 1s by far the most superior entomopathogemc nematode for the control of fungus gnats in floriculture (205,109,110) and is beheved to be the most adapted species to parasitize btbtomd larvae (86).

En tomopa thogenic Nematodes

287

8.3. citrus At present, nematodes are the only effective means of controllmg the weevll larvae in citrus. The citrus root weevil complex consists of five species: the sugarcane root-borer weevil (Dz’qrepes abbreviatus), the Fuller rose beetle (Asynonychus godmanz), the little leaf notcher (Artq~sf2ondunus), and the cltrus root weevlls (Puchnaeus litus and P opczlus). The hfe cycles of the five species are similar. Adults feed on tree leaves, where eggs are deposited Neonate larvae fall to the ground, enter the soil, and feed on roots. D. abbrevzatus, also called Apoka weevll, ISthe largest and most destructive of the five species (110). Steznernema rzobravis is the most effective nematode species against D. abbrevlatus (111,112), although S. carpocapsae (113-115) and H bacteriophora (116) have also been shown to control the weevlls. Duncan and McCoy (112) reported 77 and 90% decline in D. abbrevlatus numbers at 0 and 45 cm sol1 depths, respectively, m citrus plots 9 wk after treatment with S nobravis. Nematodes are also shown to be effective agamst the Fuller rose beetle. Morse and Lmdegren (II 7) obtained 79 and 82% reduction m the emergence of adult Fuller rose beetles with the Kapow and All strains of S carpocapsae. 8.4. Mint and Berries 8.4.1. Mint Root Borer S carpocapsae provides effective control of mmt root borer, Fumlbotrys fumalzs. The timing of appllcatlon 1svery critical in achieving good control of

larvae because of the limited persistence of nematodes, prolonged emergence of adults, and the formation of resistant hlbernacula. It is therefore preferred to make two appllcatlons of 1 bllhon/acre at pre- and postharvest, rather than a single preharvest apphcatlon of 2 billion nematodes (S. Takeyasu et al., 1992, unpublished data). 8.42. Mint flea Beetle Mmt flea beetles (Longitarsus furruglneus and L. waterhousel) infest peppermint and spearmint fields, and cause significant crop damage. In a field test, H. bacteriophora and S carpocapsae provided 94 and 67% control of L waterhousez, respectively (J. Morris, 1989, unpublished data) 8.4.3. Weevils Strawberry root weevil (Otzorhynchus ova&s) and black vine weevil (0. sulcatus) are slgmficant pests of mint, strawberries, and cranberries. A number of field trials demonstrated that nematodes are efficacious, and may replace chemical insecticides as the preferred approach for suppresslon of these weevds. H bacterlophora tends to be more effective than S carpocapsae for

288

Grewal and Georgis

the control ofboth the strawberry root weevil andblack vme weevil (102,118,119). For example, the NC and HP88 strains of H. bacterlophora applied to a Washington cranberry bog reduced larvae and pupae of the black vme weevil by 70 and lOO%,respectively; the All strain of S carpocapsae caused 75% reduction (119). 8.5. Turf 8.5.1. Mole Crickets The tawny mole crtcket, Scaptenscus vi&us, and the southern mole crtcket, S borellrz, are the two most destructtve crickets, and are dtstrtbuted throughout the coastal plain region of the southeast United States. Mole crickets are considered the most serious pests of turf and pasture grasses.Adult and nymphal mole crickets cause damage by feeding on grass roots and shoots, and by tunneling through the ground. A smgle mole cricket can create 1O-20 ft of tunnel in just one night, drying out the sot1and causing sertous damage to the plant’s root system. Annual costs of controlling mole crickets are estrmated to exceed $50 million in Florida alone. Nematodes have been successful m reducing damage to turfgrass by mole crickets. 5’.scapteriscl, which was originally isolated from mfected mole crickets m Uruguay (2201, showed 75-100% infection of adult mole crickets under laboratory conditions (121). In an inoculative release effort, S scapteriscz was released mto pastures during the summer of 1985 (122). Based on the evaluation of field-collected mole crickets over a 5-yr pertod, the nematodes were established at all the sites, with the mean number of adults infected being 11% for all years (123). More recently, another nematode species,S. riobraws, has shown exceptional potential as a brological insecticide for mole crickets (124). S riobravis, which was ortgmally isolated from sol1 in the Rio Grand Valley in Texas (125), was reported to infect natural populattons of the corn earworm, H. zea, and the fall armyworm, Spodoptera fmgzperda, m corn fields in Texas and Mexico (126). In one test, 6646% reduction in turf qury was observed with a single application of 1 billion S riobravislacre. A commercial product called Vector MCTM (Therm0 Triology, Columbia, MD) 1scurrently marketed by Lesco, for the control of mole crickets in turf. 8.5.2 Billbugs The bluegrass btllbug, Sphenophorusparvulus, 1sone of the most important pests of Kentucky bluegrass and perennial ryegrass. The entomopathogemc nematodes S. carpocapsae and H. bacteriophora have been shown to control billbug larvae and adults very effectively (82,97,12 7). In Japan, S. carpocapsae has provided excellent control of billbugs since 1992 (M. Kinoshtta, personal communication).

Entomopathogenic

Nematodes

289

8.5.3. Black Cutworms The black cutworm, Agrotis ipsdon, is a perennial problem on the close-cut bentgrass turf of golf course greens and tees, and is found throughout North America Cutworms are semisubterranean pests and usually dig a burrow into the ground or thatch, and emerge at mght to clap off grass blades and shoots. S. carpocapsae has been used effectively to manage black cutworm larvae on golf course greens at a rate of 1.0 bullion nematodes/acre (97,127) 8.6. Pet/Vet In the pet/vet sector, the cat flea, Ctenocephalrdes fells felis, is the maJor pest in which nematodes have been successfully used. The cat flea IS a cosmopolitan parasite on dogs and cats, and has also been reported from humans. The flea adults spend most of their time on mammal hosts, where mating and egg laying also occur. The eggs eventually drop off the ammal, and the hatched emerging larvae feed on organic debris in pet beddings on lawns, carpeting, or upholstered furmture. Nematodes have been extremely effective at controllmg flea larvae and pupae in home lawns (128). In tests performed m North Carolina, S carpocapsae applied at 1 billion/acre caused more than 90% mortahty of flea larvae within 24 h. S. carpocapsae also caused 91-97% mortahty of flea pupae in cocoons in a test m Louisiana. Nematodes are most effective against flea larvae in turf and soil, when the outdoor temperatures are above 14°C and the soil is moist. S carpocapsae-based products, InterruptTM and bio Flea HaltTM, were extremely popular in 1994 and 1995. 9. Suppression of Plant-Parasitic Nematodes It has often been observed that commercial applications of nematodes not only control the target insects, but also result m unexpected improvements in plant growth. Analysts of soil samples from the nematode-treated plots usually revealed fewer plant-parasitic nematodes than the untreated plots, Systematic evaluations in the field have provided support for the above observations (129). For example, in one trial, S riobravis applied at 6 x lo9 IJs/acre provided 95-l 00% control of the root-knot nematode, Meloidogyne sp; sting nematode, Belonolamus longlcaudatus; and the ring nematode, Criconemella sp at a golf course (129). In another test, both S. riobravis and S carpocapsae applied at 1 x lo9 IJs/acre provided 86-100% control of the root-knot, sung, and ring nematodes (129). The effect of entomopathogemc nematodes on plant-parastttc nematodes was first reported in 1986 (130,131). Bird and Bird (130) reported reducttons m the reproductton of Melozdogynejavanica on potted tomato plants when 5 x lo6 S glaserz nematodes were inoculated per plant. Ishibashi and Kondo obtained (131) 75-90% reduction in population density of tylenchid nema-

290

Grewal and Georgls

todes by a single applrcatron of 1 x lo4 S carpocupsaell00 cm2 of sandy soil. Work by others (132,133) have corroborated the above reports. The suppressive effects of entomopathogemc nematodes on plant-parasitic nematodes may be a result of several factors. Bud and Bird (130) demonstrated the attraction of S. gluseri to tomato roots, and suggested that the two nematode groups may compete for space. Ishibashl and Kondo (231) attributed the suppressive effects to the increased densrty of predators resultmg from the application of nematode biomass to the soil. Our data demonstrate allelochemical effects of the entomopathogetc nematodes-bacteria complex on Melozdogyne mcugnzta Juvenile behavior (E. E. Lewis and P. S. Grewal, unpublished data). We observed extremely strong repellency of M zncognita from the agar substrate treated with heat-killed IJ S feltiae and cell-free extracts of Xenorhabdus bovleniz bacteria. Further work demonstrated reduction in the root penetration of A4 rncognita when dead IJ entomopathogenic nematodes were added to potted tomato plants. More research1sneeded to elucidate the mode of action of entomopathogenic nematodesagainst plant-parasitic nematodes 10. Conclusions and Future Prospects Progress m nematode commerciahzatron during the 1990s has been phenomenal. Development of large-scale mass-production technology and mnovatrve, easy-to-use formulatrons led to the expanded use of nematodes. Improvements m the quality and mass production of several species, and the development of algmate, clay, and granular formulations were among the maJor milestones accomplrshed that enhanced consumer acceptance of nematode products. These developments led to the use of nematodes for the control of mole crrckets and billbugs m turf, root weevils m citrus, fungus gnats m mushrooms and greenhouses, and fleas in home lawns, This progress was made posstble by the collective efforts of mdustry, universmes, and federal agencies, and a polmcal atmosphere favormg a reduction m the use of chemical pesticides Despite thus progress, the reality 1sthat insecttctdal nematodes are not yet reducing reltance on chemical msectrcrdes to any significant degree This 1sat least partly because of the lack of information on the proper use of nematodes and other biologicals. In the future, discovery of new species and strains, mass productton of new species, and further improvements in formulatrons will be continued The use of genetic engineering techmques will be explored to enhance nematode brologrcal control potential. Research will also be focused on the mechanisms of suppression of plant-parasitrc nematodes by entomopathogemc nematodes. Emphasis needs to be placed on the integration of nematodes m IPM systems because of the mtroductron of new chemical products m the market, and on the trammg of extension agents and public at large in the proper use of nematodes.

Entomopathogenic

Nematodes

291

In the past few years, a noticeable change has occurred m the commerclallzation of nematodes. As opposed to the typlcal chemical pesticide development model followed by Blosys, a distinct small-scale cottage industry has emerged in the United Statesand Europe. These small businessescater mainly to the needs of the high-value home lawn and garden markets, and rely heavily on m vivo nematode production. Decline in the agricultural biotechnology stock markets has provided further impetus to this cottage industry. This trend is expected to continue well mto the early twenty-first century. References 1 Glaser, R. W. (1932) Studies on Neoaplectana glasen, a nematodeparasite of the Japanese beetle (PopllkaJaponzca) New Jersey Department of Agriculture Czrcular No. 2 I 1 2 Glaser, R W , McCoy, E E , and Smith, H. B. (1940) The biology and economx importance of a nematode parasitic in insects. J Parasztol. 26,479-495. 3 Gaugler, R. and Kaya, H K. (1990) Entomopathogenlc Nematodes zn Blologxal Control, CRC, Boca Raton, FL. 4 Georgis, R and Manweiler, S. A (1994) Entomopathogemc Nematodes, a developing biological control technology. Agric Zoo1 Rev 6,63-94. 5 Grewal, P. S , Lewis, E E., and Gaugler, R. (1997) Response of mfectlve stage parasites (Nematoda. Stemernematidae) to volatile cues from infected hosts. J Chem Ecol 23,503-5 15 6 Pomar, G O., Jr. and Leutenegger, R. (1968) Anatomy of the mfectlve and normal third stageJuveniles of Steznernema carpocapsae Welser (Steinernematldae. Nematoda) J Parasztol. 54, 340-350. 7 Bird, A. F and Akhurst, R. J. (1983) The nature of the intestinal vesicle m nematodes of the family Steinernematidae. Int J Parasltlol 13, 599-606. 8. Poinar, G. 0 , Jr., Thomas, G R., and Hess, R (1977) Characteristxs of the specific bacterium associated wlthHeterorhab&is bactenophora Nematologlca 23,97-102 9 Boemare, N E., Akhurst, R. J , and Mourant, R. G. (1993) DNA relatedness between Xenorhabdus spp. (Enterobactenacae) symbiotic bacteria of entomopathogemc nematodes, and a proposal to transfer Xenorhabdus lumlnescens to a new genus, Photorhabdus gen. novo. Int. Syst Bacterlol 43,249-255 10 Grewal, P. S , Matsuura, M , and Converse, V. (1997) Mechamsms of specificity of assoclatlon between the nematode Steinernema scaptenscl and Its symbiotic bactermm. Parasitology 114,483488. 11 Kaya, H. K and Gaugler, R. (1993) Entomopathogenic nematodes. Ann Rev Entomol 38, 18 l-206 12 Grewal, P. S , Lewis, E E., Gaugler, R , and Campbell, J. F (1994) Host finding behavlour as a predlctor of foraging strategy m entomopathogenrc nematodes Parasztology 108,207-2 15. 13. Grewal, P. S , Selvan, S., and Gaugler, R. (1994) Thermal adaptation of entomopathogemc nematodes niche breadth for infectlon, estabhshment, and reproduction J Therm Blol 19,245-253.

292

Grewal and Georgis

14 Selvan, S., Gaugler, R , and Grewal, P. S (1993) Water content and fatty acid compositton of mfectiveJuvemle entomopathogemc nematodes J Parasztol 79, 510-516. 15 Selvan, S., Gaugler, R., and Lewis, E E. (1993) Blochemmal energy reserves of entomopathogemc nematodes. J Parasztol 79, 167-l 72. 16 Pomar, G O., Jr (1990) Taxonomy and biology of Steinernemattdae and Heterorhabdittdae, m Entomopathogenic Nematodes zn Bzologzcal Control (Gaugler R and Kaya, H K , eds ), CRC, Boca Raton, FL, pp 23-61 17 Beddmg, R. A. and Akhurst, R. J (1975) A simple technique for the detection of insect parasitic nematodes m soil Nematologzca 21, 109,110. 18. Poplel, I and Vasquez, E. M (199 1) Cryopreservatlon of Isolates of Steznernema and Heterorhabdztzs spp J Nematol 23,432-437. 19 Curran, J , Gilbert, C , and Butlar, K. (1992) Routme cryopreservatton of isolates of Steznernema and Heterorhabdztzs spp J Nematol 24, 1,2. 20 Poinar, G. 0 , Jr (1979) NematodesforBzologzcal Control oflnsects. CRC, Boca Raton, FL 21 Gaugler, R. (1988) Ecological considerations m the biological control of sotlmhabmng msects with entomopathogenic nematodes. Agrzc Ecosys Envzron 24,3 5 l-360 22 Campbell, J. F. and Gaugler, R. (1993) Nictatton behaviour and its ecologtcal implications in the host search strategies of entomopathogemc nematodes (Heterorhabdmdae and Steinernematidae). Behavzour 126, 3-14. 23 Lewis, E E., Gaugler, R., and Harrison, R (1993) Response of crutser and ambusher entomopathogemc nematodes (Steinemematidae) to host volatile cues. Can J Zool. 71,765-769 24 Dutky, S. R., Thompson, J V., and Cantwell, G. E. (1964) A techmque for the mass propagation of the DD- 136 nematode. J Insect Pathol. 6,417-422 25 Howell, J F. (1979) New storage methods and Improved trapping techmques for the parasmc nematode Steznernema carpocapsae. J Invertebr. Path01 33, 155-l 58 26, Woodrmg, J L and Kaya, H K. (1988) Steinernematzd and Heterorhabdztzd Nematodes A Handbook of Bzology and Technzques. Southern Cooperative Serves Bulletin 33 1, Arkansas Agricultural Experiment Statton, Fayettesvllle, AR 27 Lindegren, J. E , Valero, K A , and Mackey, B E ( 1993) Simple cn VIVOproductton and storage methods for Steinernema carpocapsae mfective Juveniles J Nematol 25, 193-197. 28. Flanders, K L., Miller, J M , and Shields, E. J (1996) In viva productton of Heterorhabdztzs bacterzophora ‘Oswego’ (Rhabdltlda: Heterorhabditidae), a potential blologmal control agent for soil-Inhabiting insects m temperate regions) J Econ. Entomol 89, 313-380. 29. Friedman, M. J (1990) Commerctal production and development, m Entomopathogenzc Nematodes zn Biologzcal Control (Gaugler, R and Kaya, H. K , eds.), CRC, Boca Raton, FL, pp. 153-172. 30. Pomar, G 0, Jr. and Thomas, G. R. (1966) Significance of Achromobacter nematophzlus Pomar and Thomas (Achromobacteriaceae. Eubacteriales) m the

Entomopathogenic

Nematodes

293

development of the nematode, DD-136 (Neouplectana sp , Stemernematldae). Parasztology 56,385-390 3 1, Beddmg, R A. (198 1) Low cost In vitro mass production of Neoaplectana and Heterorhabditu species (Nematoda) for field control of insect pests. Nematologlca 27, lo%114 32 Bedding, R A. (1984) Large scale production, storage and transport of the msectparasitic nematodes Neoaplectana spp. and Heterorhabditls spp. Ann. appl Brol , 104, 117-120. 33 Yukawa, T. and Pitt, J (1985) Nematode storage and transport Int Patent WO85/03412 34. Kaya, H. K. and Nelsen, C E (1985) Encapsulation of steinernematld and heterorhabdltld nematodes with calcium alginate. a new approach for Insect control and other applications Environ. Entomol. 14,572-574 35 Georgis, R (1990) Formulation and application technology, in Entomopathogenic Nematodes VI Bzological Control (Gaugler, R. and Kaya, H K., eds ), CRC, Boca Raton, FL, pp. 173-191. 36. Bedding, R A and Butler, K L (1994) Methods for the storage of entomopathogenic nematodes. Int Patent WO94/05 150. 37 Conmck, W J., Jr., Nlckle, W. R , and Vinyard, B J. (1993) “Pesta”. new granular formulations for Stemernema carpocapsae J Mematol 25, 198-203, 38 Bedding, R A. (1988) Storage of entomopathogenic nematodes. Int Patent WO88/08688. 39 Silver, S. C , Dunlop, D. B , and Grove, D. I. (1995) Granular formulation of biological entities with Improved storage stability. Int. Patent WO 95/05077 40. Miller, R W. (1989) Novel pathogenicity assessment techmque for Stetnernema and Heterorhabditls entomopathogenic nematodes. J Nematol 21,574 (abstract) 4 1 Georgls, R (1992) Present and future prospects for entomopathogemc nematode products. Bzocontrol Scz Technol 2,83-99. 42. Kondo, E. and Ishlbashl, N. (1986) Infectivity and propagation of entomogenous nematodes, Stemernema spp , on the common cutworm, Spodoptera lltura (Lepldoptera: Noctuidae) Appl Entomol Zool 21,95-108 43. Hommick, W M and Reid, A. P (1990) Perspectives on entomopathogenic nematology, in Entomopathogenzc Nematodes in Blologxzal Control (Gaugler R. and Kaya, H K , eds ), CRC, Boca Raton, FL, pp. 327-345 44. Epsky, N. D and Capmera, J. L (1994) Invasion efficiency as a measure of efficacy of the entomopathogemc nematode Steinernema carpocapsae (Rhabdmda: Stemernematidae) J Econ. Entomol. 87,366-370 45. Glazer, I. (1992) Invasion rate as a measure of infectivity of steinernematld and heterorhabditld nematodes to insects. J Invertebr. Path01 59,90-94. 46. Dunphy, G B and Webster, J M (1989) The monoxemc culture of Neoaplectana carpocapsae DD 136 and Heterorhabdltls hellothldzs. Revue Nematol 12, 113-123

47. Grewal, P S , Converse, V , and Georgls, R. (1998) Influence of production and bioassay methods on infectivity of two ambush foragers. (Nematoda: Stemernematldae). J Invertebr Pathol., in press

Grewal and Georgis

294

48 Selvan, S , Grewal, P. S , Gaugler, R., and Tomalak, M. (1994) Evaluatton of stemememattd nematodes agamst PopzlllaJaponzca (Coleoptera* Scarabaetdae) larvae, species, strams, and rmse after applicatton. J Econ Entomol 87,605%609 49, Georgrs, R , Poinar, G. O., Jr, and Wrlson, A P. (1983) Practical control of the cabbage root maggot, Hylemza brasszcae (Dtptera Anthomyndae), by entomopathogenous nematodes IRCS Med SCL 11,332 50. Pye, A. E and Pye, N L (1985) Drfferent appltcatrons of the insect parastttc nematode Neoaplectana carpocapsae to control the larger pme weevtl, Hyloblus abletu. Nematologlca

31, 109-I 16.

5 1, Etdt, D. C. and Weaver, C. A A. (1993) Control damage by seedlmg debarkmg weevil Forestry Canada-Marmmes Regron, Technical Note No 271 52. Pomar, G O., Jr., Evans, J S., and Schuster, E (1983) Fteld test of the entomogenous nematode, Neoaplectana carpocapsae, for control of corn rootworm larvae (Dzabrotxa sp., Coleoptera) ProtectIon Ecol 21, 1441-1447 53, Kaya, H. K , Manmon, C M , Burlando, T M , and Nelsen, C E (1987) Escape of Stelnernema feltlae from algmate capsules contammg tomato seeds J Nematol

19,287-291

54. Smith, M T , Georgts, R , Nyczeptr, A P., and Mtller, R W. (1993) Biologtcal control of the pecan weevtl Curculzo caryae (Coleoptera Curculiomdae), wtth entomopathogemc nematodes J Nematol. 25,78-82 55. Klem, M G and Georgts, R. (1994) Appltcatton techniques for entomopathogenic nematodes, m Proceedmgs of the VI International Colloquium on Invertebrate Pathology and Mxroblal Control, Montpelher, France, pp 483,484 56. Glazer, I. (1992) Survtval and efficacy ofstemernema carpocapsae m an exposed envu-onment Blocontrol Scz Technol 2, 10 l-l 07 57. Broadbent, A B. and Olthof, Th. H. A (1995) Foltar applicatton of Steznernema carpocapsae (Rhabdrttda: Stemernemattdae) to Lzrzomyza trzfolzt (Dtptera Agromyzidae) larvae m Chrysanthemums Environ Entomol 24,43 l-435, 58. Kaya, H K (1993) Contemporary tssues in biologtcal control with entomopathogemc nematodes Ext. Bull. No. 375. Food & Ferttltzer Technology Center, Tatpet Ctty, Republic of China (Taiwan). 59. Nrshida, T and Bess, H A (1957) Studtes on the ecology and control of the melon fly, Dacus (Strumeta) cucurbttae Coqutllett (Dtptera. Tephrnidae) Hawarl Agricultural Exp Statton Bull. 34, 3-4 I

60 Purcell, M., Johnson, M. W , LeBeck, L. M., and Hara, A H (1992) Btologtcal control of Hekoverpa zea with Stemernemacarpocapsae(Rhabdmda Stememematrdae) m corn used as a trap crop Envzron Entomol. 21, 1441-1447 61. Bedding, R. A. and Miller, L. A (1981) Use of a nematode Heterorhabdltzs helzothzdls,to control black vme weevil, Otrorhynchus sulcatus, m potted plants Ann Appl Bzol 99,2 1l-216. 62. Barr, M A. (1992) Drsmfestatron of arttchoke stumps with entomopathogemc nematode against the artichoke plume moth, in PvoceedzngsXIX International CongressofEntomology, Abstracts. Berjmg, China, p 3 19 63. Capmera, J L and Hubbard, B E (1987) Bart formulations of chemical and mrcrobial msectrcrdes for suppresston of crop-feeding grasshoppers J Agrx Entomol 4,337-344

Entomopathogenic

Nematodes

295

64 Georgts, R., WojCtk, W F , and Shetlar, D J. (1989) Use of Stetnernema felttae m a ban for the control of black cutworms (Agrotis rpszlon) and tawny mole crtckets (Scaptertscus vtctnus). Florida Entomol 72,203,204 65. Parkman, J. P and Frank, J. H. (1993) Use of a sound trap to inoculate Stetnernema scapteriscz (Rhabdttrda: Stemernemattdae) mto pest mole crrcket populattons (Orthoptera: Gryllotalpidae). FL Entomol 76,75-82. 66 Appel, A G., Benson, E. P , Ellenberger, J. M., and Manweller, S A. (1993) Laboratory and field evaluations of an entomogenous nematode (Nematoda. Stemernemattdae) for German cockroach (Dtptera. Blattelhdae) control J Econ Entomol 86,177-784 67 Manweller, S A., Appel, A, and Weber, T (1993) Nematode-based btologtcal control of German cockroaches, m, Ftrst International Conference on Insect Pests ln the Urban Enwonment (Wtldey, K B. and Robmson, W. H., eds ), BPCC Wheatons, Exeter, UK, pp 173-180. 68 Renn, N., Barson, G., and Rrchardson, P. N (1985) Prelimmary laboratory tests wtth two species of entomophthc nematodes for control of Musca domestzca m mtensrve ammal units. Ann Appl Bzof 106,229~233. 69 Renn, N. (1990) A comparison of nematode-containing baits with commerctally avatlable msecttcidal bans for the control of adult Musca domesttca mfestatrons of mtenstve animal umts Brtghton Crop Protectton Conference. Pests and Dweases, 1990, vol. 1,26%274 70. Schmitt, A. T., Gowen, S R , and Hague, N. G. M. (1992) Baitmg techniques for the control of Cosmopolztes sordzdus Germar (Coleoptera Curculiomdae) by Stewernema carpocapsae (Nematoda: Stemernemattdae). Nematologtca 22, 159-163. 71 Treverrow, N L. and Beddmg, R. A (1993) Development of a system for the control of the banana weevil borer, Cosmopolttes sordtdus, wtth entomopathogemc nematodes, m Nematodes and Btologtcal Control of Insect Pests (Beddmg, R , Akhurst, R and Kaya, H. K , eds ), CSIRO, Melbourne, Australia, pp. 41-47 72. Poinar, G. O., Jr. and Enmk, F. (1972) The use of Neoaplectana carpocapsae Weiser (Stemernemattdae. Rhabdltotdea) against adult yellowJackets (Vespula spp., Vespidae. Hymenoptera). J. Invertebr Pathol. 19,33 l-334. 73 Wojctk, W F and Georgts, R (1988) Infection of adult western yelLowJackets with desiccated Stemernema feltlae (Nematoda). J Invertebr. Path01 52, 183,184, 74 Capinera, J L , Pelissrer, D., Menout, G. S., and Epsky, N D (1988) Control of black cutworm Agrotts zpsilon (Leprdoptera. Noctutdae), with entomogenous nematodes (Nematoda. Stemernemattdae, Heterorhabdrtidae). J. Invertebr Path01 52,427-435. 75 Kaya, H. K., Joos, J L , Falcon, L A, and Berlownz, A. (1984) Suppression of the codlmg moth (Leptdoptera Olethreutrdae) with the entomogenous nematode, Steinernemafeltlae(tiabdltlda. Stememematrdae) J Econ Entomol 77,1240-1244 76 Reardon, R. C , Kaya, H. K., Fusco, R. A., and Lewts, F. W. (1986) Evaluation of Stemernema feltzae and S btbzonis (Rhabditida. Stemernemattdae) for sup-

Grewal and Georgis

296

77 78 79. 80.

81

82

83

84

pressron of Lymantrza dzspar (Leptdoptera. Lymantrndae) m Pennsylvama, USA Agrzc Ecosys Environ 15,1--g Walker, T. J (1982) Sound traps for sampling mole crrcket fltghts (Orthoptera: Gryllotalpidae: Scapteriscus). FL Entomol. 65, 105-l 10 Rtchardson, P N (1990) Uses for parasitic nematodes m Insect control strategies m protected crops. Aspects Appl Biol 24205-2 IO. Georgts, R. and Gaugler, R (1991) Predictability in btological control using entomopathogemc nematodes. J Econ Entomol. 84, 7 13-720 Kaya, H K (1990) Soil ecology, m Entomopathogemc Nematodes rn Brologlcal Control (Gaugler, R and Kaya, H K., eds.), CRC, Boca Raton, FL, pp 93-l 15, Richardson, P N and Grewal, P. S (1994) Insect parasmc rhabdmd nematodes and the so11envtronment, in Ecology and Bzologv of Soil Organisms (Bhandarr, S C and Somam, L L., eds.), Agrotech Pubhshmg Academy, Udaipur, Indta, pp 106-130 Klem, M G. (1990) Efficacy against sol1 Inhabiting insect pests, m Entomopathogentc Nematodes zn Bzologzcal Control (Gaugler, R. and Kaya, H. K., eds.), CRC, Boca Raton, FL, pp. 195-214. Begley, J W. (1990) Efficacy against insects in habitats other than ~011, m Entomopathogenic Nematodes in Biologzcal Control (Gaugler, R. and Kaya, H K., eds ), CRC, Boca Raton, FL, pp 2 15-23 1 Richardson, P N. and Grewal, P S. (199 1) Comparative assessment of biologrcal (Nematoda Sternernema felttae) and chemical methods of control for the mushroom fly Lycorlella aunprla (Drptera: Scrarrdae) Biocont. Scr Technol, 1, 2 17-228

85 Grewal, P S , Tomalak, M , Keel, C B. O., and Gaugler, R (1993) Evaluatron of

86. 87

88

89

90

genetrcally selected stram of Steinernema feltlae agamst the mushroom sciarid Lycorlella mall Ann Appl B1o1 123,695-702. Grewal, P. S. and Smith, C. (1995) Insect-parasitic nematodes for mushroom pest control Mushroom News 14, 15-25. Scheepmaker, J W A., Geels, F P., van Grrensven, L. J, L. D., and Smits, P. H. (1995) Control of the mushroom sctarid (Lycorzella aurzpila) and the mushroom phorid (Megaseha halterata) by entomopathogemc nematodes, m Science and Cultzvation ofEdzble Fungz (Elliot, T., ed.), Balkema, Rotterdam, pp 491-498, Grewal, P S and Richardson, P. N (1993) Effects of application rates of Steuzernema feltrae (Nematoda. Stemernematidae) on btologmal control of the mushroom sciartd fly Lycorrella auriplla (Drptera: Sctaridae). Blocontrol SCI Technol. 3,294O. Grewal, P S., Rtchardson, P N., Collins, G., and Edmondson, R. N (1992) Comparatwe effects of Steznernemafiltiae (Nematoda: Steinernematidae) and msectmides on yield and cropping of Agarzcus blsporus. Ann Appl Blol. 121, 5 1 l-520. Richardson, P N (1987) Susceptrbrlity of mushroom pests to the msect-parasttrc nematodes Stelnernema feltlae and Heterorhabdltis hekothldis. Ann. Appl Bzol 111,433-438

Entomopa thogenic Nematodes

297

9 1 Tomalak, M. (1994) Selecttve breeding of Steznernema felttae (FihpJev) (Nematoda: Steinernematidae) for improved efficacy in control of a mushroom fly, Lycorzella solam Winnertz (Diptera: Sciaridae). Bzocontrol Sci Technol. 4, 187-198. 92 Ntckle, W. R and Cantelo, W W. (1991) Control of a mushroom-mfestmg fly, Lycortella mali with Stetnernema feitiae J Nematol. 23, 145-147 93. Smith, F. F. (1932) Biology and control of the black vine weevil. US Department of Agric. Tech. Bull. 325. 94. Simons, W. R. (1981) Btological control of Otiorhynchus sufcatus with heterorhabditid nematodes m the greenhouse. Neth J Plant Pathol, 87, 149-I 53. 95 van Tol, R. W. H. M. (1994) Influence of temperature on the control of black vme weevil with strains of some insect-parasitic nematodes. IOBC Bull 17, 116-l 19 96. Stimman, M. W., Kaya, H K., Burlando, T M , and Studdert, J. P (1985) Black vine weevil management m nursery plants. CA Agrzc., Jan., 25,26 97. Georgts, R. and Poinar, G. O., Jr (1989) Field effectiveness of entomophtllc nematodes Neoaplectana and Heterorhabditis, m Integrated Pest Management for Turfgrass and Ornamentals (Leslie, A. R. and Metcalf, R. L , eds.), U S. Environmental Protection Agency, Washington, DC, pp 2 13-224 98. Hanula, J. L. (1993) Vertical dtstributton of black vme weevil (Coleoptera* Curculionidae) immatures and infection by entomogenous nematodes m so11columns and field soil. J. Econ Entomol 86,340-347. 99. Rutherford, T. A., Troter, D., and Webster, J M. (1987) The potential of heterorhabduid nematodes as control agents of root weevils. Can Ent 119,67-73 100. Mracek, Z., Jiskra, K., and Kahounova, L. (1993) Efficacy of stemernematid nematodes (Nematoda: Stemernemattdae) m controlhng larvae of the black vme weevil, Otiorrhynchus sulcatus (Coleoptera Curcuhomdae) in laboratory and field experiments. Eur J Entomol. 90,7 l-76 101. Mtduturi, J. S., De Clercq, R., and De Grisse, A. (1994) Greenhouse and field control of black vine weevil Ottorhynchus sulcatus F. with Stetnernema capocapsae and Heterorhabdttts sp IOBC Bull 17, 148-154 102. Binns, E. S. (1973) Fungus gnats (Diptera: Mycetophtlidae, Sctartdae) and the role of mycophagy in soil* a review Revue d’Ecologte et de Btologue du sol l&77-90. 103. ‘Freeman, M J. (1983) Sciarid flies, Diptera: Sciaridae, in Handbooks for the Identification of British Insects 9 (6), Royal Entomological Society of London 104. Leath, K. T. and Newton, R. C (1969) Interaction of a fungus gnat, Bradysza sp (Sctaridae) with Fusarium spp. on alfalfa and red clover Phytopathology 59, 257,258 105. Hams, M. A., Oetting, R. D , and Gardner, W. A. (1995) Use of entomopathogenic nematodes and a new momtormg technique for the control of fungus gnats, Bradysta coprophtla (Dtptera. Sciaridae), in Floriculture Btol Control 5, 412-418. 106 Nedstam, B. and Burman, M. (1990) The use of nematodes against sciarids in Swedish greenhouses. SROPMPRS Bull 18, 147,148

298

Grewal and Georgrs

107 Gouge, D. H and Hague, N. G. M. (1994) Glasshouse control of fungus gnats, Brau’ysza paupera, on mchsias by Stemernema feltlae. Fundam. Appl Nematol 18,77-80 108. Lmdqmst, R K , Buxton, J , and Peatkoski, J. (1994) Btologtcal control of sclartd fhes and shore fltes m glasshouses. Brighton Crop Protection Conference Pests and Dueases, Brtttsh Crop Protection Counctl, Farnham, UK, pp 1067-l 072 109 Gouge, D H. and Hague, N. G M (1995) The development of Steuzernema feltlae (Nematoda Stemernemattdae) m the sctartd fly Bradysla paupera (Dtptera. Sctartdae). Ann Appl B~ol 126, 395401 110 Woodruff, R. E (1985) Cm-us weevtls in Flortda and the West Indies prehmtnary report on systematlcs, btology, and dtstrtbutron (Coleoptera Curcuhomdae). FL Entomol 68,370-379 111. Schroeder, W J (1994) Compartson of two stemernemattd specres for the control of the root weevtl Dlaprepes abbrewatus J Nematol 26,360-362 112 Duncan, L W and McCoy, C W (1996) Vertrcal drstrrbutron m solI, perststence, and efficacy against citrus root weevil (Coleoptera. Curcuhomdae) of two spectes of entomopathogentc nematodes (Rhabdmda Stemernemattdae; Heterorhabdmdae). Envzron. Entomol 25, 174-l 78 113 Schroeder, W J (1987) Laboratory btoassays and field trials of entomogenous nematodes for control of Dmprepes abbrevzatus (Coleoptera Curcultomdae) m c&us Envrron Entomol 16,987-989 114 Schroeder, W J (1990) Suppression of Dlaprepes abbrevzatus (Coleoptera Curcultomdae) adult emergence with sot1 appltcation of entomopathogemc nematodes (Nematoda Rhabditrda) FL Entomol 73,680-683 115. Schroeder, W J (1992) Entomopathogemc nematodes for the control of root weevtls m citrus FL Entomtol 75, 563-567 116 Downing, A S., Ertckson, C. G., and Kraus, M J (1991) Fteld evaluatton of entomopathogemc nematodes against cttrus root weevils (Coleoptera Curcuhonidae) m Florida m cmus FL Entomol 74,584-586 117 Morse, J G. and Lindegren, J E (1996) Suppression of Fuller Rose beetle (Coleoptera. Curcuhonidae) on cttrus with Steznernema capocapsae (Rhabdmda Stemernematidae). FL Entomol. 79,313-384. 118 Klmgler, J (1988) Investlgatton on the parasmsm of Otzorhynchus salzczcola and 0 sulcatus (Coleoptera Curculionidae) by Heterorhabdztu sp. (Nematoda). Entomophaga 33,325-33 1. 119 Shanks, C H. and Agudelo-Stlva, F. (1990) Fteld pathogemctty and persistence of heterorhabdmd and stemernemattd nematodes (Nematoda) mfestmg black vme weevrl larvae (Coleoptera Curcuhomdae) m cranberry bogs J Econ Entomol 83, 107-l 10 120 Nguyen, K. B and Smart, G C , Jr (1990) Sternernema scapterwcl n. sp (Rhabdttida: Stemernemattdae) J Nematol 22, 187-l 99 121. Nguyen, K. B and Smart, G C , Jr. (1991) Pathogemctty of Steznernema scapterlscl to selected invertebrates J Nematol 23,7-l 1 122 Hudson, W. G., Frank, J H., and Castner, J. L (1988) Btological control of Scapterzscus spp mole crtckets (Orthoptera Gryllotalptdae) m Flortda Bull Entomol Sot Am 34, 192-198

Entomopathogenic

Nematodes

299

123 Parkman, J P., Hudson, W G , Frank, J H., Nguyen, K B , and Smart, G C., Jr (1993) Estabhshment and persistence of Steznernema scaptertsct (Rhabdtttda. Stemernemattdae) in field populattons of Scapterlscus spp mole crickets (Orthoptera Gryllotalprdae) J Entomol. Sci 28, 182-190 124 Gorsuch, C. S (1995) Brological control of mole crtckets. Carolmas Green 31, 18,19. 125 Cabamllas, H E. and Raulston, J. R (1994) Pathogemctty of Steznernema rtobravts against corn earworm, Heltcoverpa zea (Boddte). Fundam Appl Nematol 17,219-223. 126 Rat&ton, J R , Pan, S D , Loera, J , and Cabanillas, H. E (1992) Prepupal and pupal parasttism of Heltcoverpa zea and Spodoptera fiugtperda (Leptdoptera Nocturdae) by Steznernema sp m cornfields m the Lower RIO Grand Valley J Econ. Entomol 85, 1666-1670. 127 Shetlar, D. J (1995) Turfgrass Insect and mtte management, m Managzng Turfgrass Pests (Watschke, T L , Dernoeden, P. H , and Shetlar, D. J., eds ), Lewis, Boca Raton, FL, pp 17 l-342 128 Manwetler, S. A (1994) Development of the first cat flea biologtcal control product employmg the entomopathogemc nematode Stetnernema carpocapsae Proceedings of the Brtghton Crop Protectton Conference Pests and Dtseases, British Crop Protection Council, Farnham, UK, 3, pp. 1005-l 012 129 Grewal, P S , Martin, W R , Miller, R W., and Lewrs, E E. (1997) Suppresston of plant-parasitic nematode populattons m turfgrass by appltcatton of entomopathogenic nematodes. Btocontrol Sci. Technof. 7,393-399. 130 Bud, A F and Bird, J (1986) Observations on the use of insect parastttc nematodes as a means of btological control of root-knot nematodes Int J Parastttol. l&511-516 131 Ishibashi, I and Kondo, E (1986) Steinernema feltiae (DD- 136) and S glaserz Persistence in so11and bark compost and their Influence on native nematodes J. Nematol l&310-3 16 132 Smttley, D R , Warner, W. R , and Bard, G. W. (1992) Influence of trrtgatton and Heterorhabdttu bactertophora on plant-parasittc nematodes m turf Suppl J Nematol 24,637-64 1. 133. Gouge, D H., Otto, A. A., Schirocki, A., and Hague, N G M (1994) Effects of steinernemattds on the root-knot nematode, Meloldogyne Javanlca. Ann Appl Biol 124(Suppl.), 134,135

16 Naturally Occurring Baculoviruses for Insect Pest Control Brian A. Federici 1. Introduction Baculovnuses of insects have been promoted for their pest control potential for more than half a century (I). Despite this, only a few have been successful m blologtcal control, and almost none has proven a commercial success,or 1s used routinely for large-scale Insect control m industrlaltzed countries Baculoviruses have, however, achieved moderate successm some developing countries. Thus, in addition to discussmg the use of naturally occurrmg baculovnuses as pest-control agents, the reasons why these viruses have not been of greater commercial successm industrialized countries will be considered. This requires definition of both the different ways baculoviruses can be used for insect control, and the performance expectations used to evaluate baculovnus success.Such an assessmentidentifies the key features required for a vrrus to be successful as a control agent. Though baculovn-uses are used as the examples here, these principles apply to other pathogens, and many predators and parasites, as well. The guiding principle in pest control remains prtmartly one of economtcs; control strategies and agents used are those that are the most cost-effective m the short or long term. This would appear to be obvious, but too often is overlooked m the literature on viruses and other biological control agents. The effectiveness of a control strategy based on a baculovirus will always be compared with that obtamed wtth other avadable control strategies, especially those based on other pathogens and synthetic chemical insecticides. Cost-effecttveness can vary with such factors as crop, season, the species complex to be controlled, the level of control considered effective, the cost of production of the control agent, geographical location, governmental regulations affecting From. Methods in Blofechnology, vol 5 Bopesbodes Use and Del/very Edited by F A Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

301

302

Fedenci

registratton and use, and the status of economtc development m the country where the vnus 1sused. The latter two factors are particularly relevant, because the costs for development, productron, and use of a baculovnus m developing countries are much less than in highly industrialized countries. This is a result of lower material and labor costs m developing countrres, as well as to the smaller size of most farms, lower levels of agrrcultural mechamzation, and cheaper and less cumbersome registration procedures. For these reasons, baculovu-uses are used more m developmg countries than m the highly mdustrialtzed nations, where they account for CO.1% of operational pest control As a result, positive assessmentsof baculovtrus cost-effecttveness, based on studies carried out in developing countries, cannot be directly translated into correspondmg levels of cost-effectiveness m developed countries 7.7. Strategies for Use The most cost-effective potential strategy for using a baculovnus, as with other biological control agents, 1sas a classical biological control agent. In this strategy, mtroduction of a virus results m outbreaks of disease (epizootrcs) and mamtenance of the vnus m the target populatton. Within a few years, the pest population 1s reduced below the economic threshold on a permanent basis. Baculovnuses may establish and become endemic m a target population withm a few years of mtroduction, yet there is only one good example of a classical biological control successwith a baculovnus. the control the European spruce sawfly, Gilpinia hercynzae, m North America, by its nuclear polyhedrosis virus (NPV) (see Subheading 3.1.). Another strategy is to use a vnus as an augmentattve control agent. In this strategy, a virus endemic m a population, but at a low level, IS applied agamst a pest population at the begmnmg of the season or early m the development of the pest population The vn-us reduces the population below the economic threshold, or reduces pest damage substantially sooner than might occur naturally. The effect usually only lasts one or a few seasons, and must be repeated. The granulosis vuus (GV) of the grapeleaf skeletomzer, Harrzszna bnlllans, and the NPV of the Douglas fir tussock moth, Orgyra pseudotsugata, descrtbed m Subheading 3.2., have been used successfully in this manner. Again, however, this tactic has proven of only limrted utility The most common baculovnus pest control strategy is to use a vu-us as a viral msecticide. In thustactic, a virus, such as the Helzothis NPV, is formulated and applied against a target pest on a periodic basis, as needed, much as are chemical msecttctdes.Dependmg on the target pest and cropping system,apphcattons may be fewer than those requrred with a chemtcal msecticide, because the viruses are quite specific and typically do not krll predatory and parasitic Insects. Natural enemies, therefore, can remain m the ecosystem,retardmg increases m the pest population after the mltial mortality caused by the vn-us. In addttion, viral repro-

Natural

Baculoviruses

for Insect

Pests

303

duction in the target insect could add to the amount of the virus in the crop envlronment, and this can extend control and, thus, cost-effectiveness. In pests with only one or a few generations per season,a single application may yield effective season-long control, where a combmatton of these factors is m operation. 1.2. Performance

Expectations

and Economics

An issue of major importance affecting the adoption of baculovnuses as control agents is performance expectation. In most cases,a successful virus is one that can reduce the pest or vector population to below an economic threshold routmely and rehably, at a cost that is economical m proportion to the value of the crop. Clearly, viruses that are effective as classical biological control agents, or as augmentative agents, would be the most cost-effective, because of the limited number of applications required. But few viruses can be used successfully in either of these strategies. As a result, the efficacy of most viruses is evaluated in terms of their utility as viral insecticides. Smce World War II, the performance of chemical msecticldes has set very high expectations for all alternative pest control strategies. Traditionally, chemicals have been fast-actmg, broad-spectrum control agents with substantial residual activity, which are relatively inexpensive, and easy to produce, formulate, and use. Thus, under most circumstances, baculovnuses are evaluated on the basis of how they compare with chemical insecticides, m particular, how quickly they kill the target msect, and at what cost for an acceptable level of crop protection. This leads to a paradox for microbial control agents. The two properties of chemical insecticides originally considered then best attributes, i.e., a broad spectrum of activity and significant residual activity, are now viewed as detrimental, because they result in the destruction of natural enemy populations and the development of msecticide resistance,though the latter is often the result of overuse. Yet, most baculovnuses have a narrow spectrum of activity and relatively poor residual activity. Though these are now considered desirable properties, until recently, they have discouraged mterest by industry m the development of many potentially useful viruses, becauseof the relatively high costs of development and registration m comparison to the hkely return on mvestment. The costs for development and registration of a naturally occurrmg baculovirus, however, are much cheaper than those for a chemical Insecticide, e.g., approxtmately one-half million dollars for a baculovirus, compared to at least several millions of dollars for most chemicals m the United States,Europe, and Japan. But a company still must see the potential for making its Investment pay off within a few years. For viruses highly specific in host range, unless their target is a pest of a maJor commodity, or a polyphagous pest causing damage to a variety of crops, this is simply not possible given the current regulatory environment and market size in most mdustriahzed countries

Federici

Fig. 1. Electron micrographs of (A) a polyhedron of a typical nuclear polyhedrosis virus, and (B) a granule of a granulosis virus. The NPV illustated is of the multinucleocapsid type (MNPV), with more than 1 nucleocapsid possible per viral envelope. Some NPVs have only a single nucleocapsid per envelope (SNPV type), but all GVs have only one nucleocapsid per envelope.

Industrial interest, whether on a small or large scale, is important, because it is industry that will produce the viral insecticides. This is particularly true in developed countries, where farms tend to be large. In fact, most farmers, whether large or small, want a reliable supply of control agents, and, though willing to change cultural practices, because of numerous other responsibilities, they typically are not willing to manufacture their own insecticides. 2. Baculovirus Types and General Properties The baculoviruses compose a single family, Baculoviridae (2), containing two genera of occluded viruses, Nucleopolyhedrovirus, commonly know as nuclear polyhedrosis viruses (NPVs), and Granulovirus, commonly known as granulosis virus (GV). The occluded viruses are so named because, after formation in infected cells, the mature virus particles (virions) are occluded within a protein matrix, forming paracrystalline bodies that are generically referred to as either inclusion or occlusion bodies (Fig. 1). 2.1. Biological Properties Related to Insect Control A short description of the most important biological properties of NPVs and GVs is given here, to set the stage for discussion of how these viruses are used as insect control agents. Knowledge of these properties for different types of

Natural

Baculoviruses

for Insect

Pests

305

NPVs and GVs provides Insight into the advantages and limltatlons of their use for insect control. NPVs are known from a wide range of insect orders, as well as from crustatea (shrimp), but by far have been most commonly reported from lepldopterous insects, from which well over 500 isolates are known (2). Many of these are different viruses (i.e., viral species), but, even in a general sense, It 1snot known how many viral species are represented among these isolates, because the baculovnus species concept IS not well developed. NPVs are easily transmittedper OSand replicate in the nuclei of cells, generally causing an acute fatal disease, with death usually occurring 5-9 d after infection in larvae infected during the third or fourth mstar. The occlusion bodies of NPVs are referred to commonly as polyhedra, becausethey are typically polyhedral m shape(Fig. 1A). They are large (approx OS-2 pm), and form m the nuclei, where each occludes as many as several hundred vlrions. The NPVs of lepidopterous msects infect almost all host tissues, but produce the most polyhedra m the epidermis, fat body, and trachael matrix. Vinon-containing polyhedra typically are not produced in the midgut epithelium. In other orders of insects, infection and production of polyhedra are typically restricted to the midgut epithelmm. Some NPVs have a very narrow host range, and may only replicate efficiently in a single species; others, such as the AcNPV, i.e., the NPV of the alfalfa looper, Autographa dfornica (Speyer), have a relatively broad host range and are capable of infecting species from different genera (2). The GVs, of which there are now over 100 Isolates, are closely related to the NPVs, but differ from the latter m several important respects.Like NPVs, GVs are highly mfectlousper OS,but are only known from lepldopterous insects (2). Inmally, GVs replicate m the cell nucleus, but repllcatlon involves early lysls of the nucleus (prior to vu-ion formation), which, in NPVs, does not occur until after most polyhedra have formed. After the nucleus lyses, GV replication continues throughout the cell, which consists of a mixture of cytoplasm and nucleoplasm. When completely assembled, the virlons are occluded mdlvldually m small (200 x 600 nm) occlusion bodies, referred as granules (Fig. 1B). Many GVs primarily mfect the fat body (type 1); others have a broader tissue tropism and replicate throughout the epidermis, tracheal matrix, and fat body (type 2). The type 1 and type 2 GVs, like the lepidopteran NPVs, do not produce occluded vlrions m the midgut epithelium (3). One, however-the GV of the grapeleaf skeletonizer, H. brillians-1s unusual because it only replicates in the midgut eplthelium (type 3), where the occluded virlons are produced (3).

3. Use of Baculoviruses as Insect Control Agents In this section, examples will be provided to illustrate the use of baculoviruses as classical biological control agents, augmentative control agents, and

306

Federici

vn-al msectrcrdes.Baculoviruses are most commonly used as viral msectlcrdes, and, thus, this tacttc wrll receive the most attention. 3.1. Classical Biological Control The best example of insect control by a baculovirus IS the use of the NPV of the European spruce sawfly, G. hercyniae, as a classrcal btologtcal control agent to control thts Important forest pest m North America (4,5). The European spruce sawfly was introduced into eastern Canada from northern Europe around the turn of the century, and was a severe forest pest by the 1930s Hymenopteran parasites were introduced from Europe m the mid- 1930s as part of a biologtcal control effort, and, inadvertently, along with these came the NPV, first detected m 1936. Natural epizoottcs caused by the vtrus began m 1938, by whtch trme the sawfly had spread over 31,000 km2. Most sawfly populations were reduced to below economrc threshold levels by 1943, and remam under natural control today, the control being effected by a combmatron of the NPV, which accounts for more than 90% of the control, and the parasttes. Although vnuses, particularly NPVs, are frequently associated with rapid declines m the populations of important leptdopterous and hymenopterous (sawfly) pests, the G. hercyniae NPV is the only virus that proved effective as a classtcal blologlcal control agent (5). Because the G. hercynzae NPV remains the most successful example of pest control by a vnus, It IS worthwhile to consider why tt has been so successful Two mam reasons are apparent: First, the vn-us attacks the midgut eptthelmm As a result, wtthm 48 h of mfectton, larvae begin to defecate newly produced mfecttous vn-al polyhedra. Each infected larva becomes a virus factory for dlstrtbutmg the virus m the envtronment. The second reason IS that the virus can be passedverttcally from adults to larvae. When an mdtvrdual becomes infected in the last instar, it can often survive. The emerging adult then develops an mfectron m the midgut, and also disseminates the vtrus in the envnonment. Another successful use of a vnus as a quasiclasstcal biological control agent IS the use of the so-called nonoccluded baculovnus of the palm rhmoceros beetle, Uryctes rhznoceros, (Scarabaeidae) for control of thusscarab pest in the South Pacific. During the middle of this century, the palm rhmoceros beetle, an introduced pest, became a serious pest of 011and coconut palms m many South Pacific islands, including the FiJi Islands, Western Samoa, American Samoa, and the Tokelau Islands (6). The beetle adult bores mto the heart of the palm, and, aside from damage to the fronds, repeated attacks can cause tree death. Eggs are laid, and larvae develop m decaying palm tissue, as well as m other types of decaying vegetable matter, and in manure. A search for pathogens led to the discovery of a nonoccluded baculovtrus m Malaysra m the mtd1960s (7). This virus apparently did not occur m the islands, where the beetle

Natural Baculoviruses

for insect Pests

307

had become a pest, and it was subsequently developed as a very effective control agent. The virus develops m the midgut eptthelium of larvae and adults, but also m other tissues, such as the fat body. Infection of the gut results m highly mfectious feces, which in turn results in the virus being spread to other adults during the course of mating, and to larvae when infected females visit larval habitats for oviposition. Infection of females also reduces fecundity substantially. In the early stages of the control program, compost heaps and coconut logs were contaminated with virus, which resulted in both infected larvae and adults, with the latter serving to disseminate the virus to other natural habitats When tt was realized that the adults were suscepttble to infection and could dissemmate the virus very effecttvely, the control tactics focused on direct mfection of adults (6). Wild adults were trapped, contaminated with virus by immersing them m aqueous vu-us suspenstons (2-3 freshly ground virus-killed beetles/L) for several mmutes, and then released to carry the virus into natural populations. This program has been very successful m areas where the beetle was a major introduced pest. Damage to fronds m Fiji, for example, ranged from 40 to 90% prior to mtroductton of the virus. Wrthm 4-6 yr m most of these areas, the level of beetle-damaged fronds was <20%, and, in many areas, ~10% (6). Corresponding reductions in adult beetle catches also occurred. In areas where the concentrations of breeding sites were high, periodtc beetle outbreaks can reoccur, necessitatmg reintroduction of the virus With the exception of this requirement for periodic remtroductions of virus m some areas, the use of the O~ctes baculovuus could be considered a true classical biological control success. 3.2. Augmentafive Control Agents Baculovn-uses have not been commercially successful as augmentative control agents, but two examples are worth mentroning: the use of a GV agamst the western grape leaf skeletomzer, and a NPV agamst the Douglas fir tussock moth. They illustrate how baculovnuses can be effective on a long-term basis, yet still do not attract industrial development. The western grape leaf skeletomzer, H. brillians, can be a serious outbreak pest of table and wme grapes in many areas of California. In most cases,tt can be controlled with applications of Kryohte or Bacrllus thurzngiensis (Bt). However, populations ofH. brzllzansusually return m a generation or two, and must be treated again. Treatment of populations at a rate of 8-10 g/ha of GV-diseased cadavers, however, can yield mortality rates of greater than 90%, and reduction of the pest population to well below an economic threshold, for 3 yr or longer (8).

308

Fedeno

This excellent level of long-term control results from a unique combmation of virus and host btology. Grape leaf skeletomzer larvae are gregarious, feedmg side by side during the first three mstars, and follow each other to new feeding sites as leaf tissue is consumed. The virus rephcates m the midgut eprthelium, and, within a few days of infection, a virus-laden diarrhea develops. Thus, once a single larva is infected, all the mdividuals withm the population usually become infected and die within a week. In addition, larvae infected during the last mstar survive and carry the virus over mto the adult stage. Infected adult females generate fresh viral granules in their gut, and contaminate eggs with virus when the abdomen contracts during oviposition. So, like the sawfly and palm beetle viruses, infection of the midgut epithelium is the major factor contributmg to the success of this virus, and this advantageous property is complemented by the gregarious behavior of the larvae (81. Ironically, as successful as thts virus could be as a pest-control agent, it IS not used commercially. The reasons are simple. The market is small, because the virus only has to be applied once or a few times every several years, makmg it difficult for mdustry to justify the registration costs. In addition, broad spectrum insecticides are available to control H brzllians. Methods for limited virus mass production are available that would enable commercial msectaries, which produce parasitic insects and predators, to produce the virus and make a profit. However, the costs for registration of the vn-us by the U.S. Environmental Protection Agency (EPA) and State of Cahforma are too high for these small-busmess insectaries to justify registration. In the case of the Douglas fir tussock moth, 0 pseudotsugata, the OpNPV that attacks this species can also be used very successfully to reduce damage to large areas of Douglas fir forests. The tussock moth is a cyclical problem, capable of causing sigmficant lossesevery several years at the peak of tts population cycle. As populations reach high densities, natural epizootics caused by NPV often reduce the population to well below an economic threshold for several years, but usually only after substantial economtc damage has occurred. Thus, m the Douglas fir tussock moth control program, populations are momtored, and the virus IS sprayed on populations to initiate epizootics before they reach an economic threshold. Tests m the field have shown that one apphcation by aircraft of virus at a rate of 2.5 x 10” polyhedra/ha reduces larval densities by 90%, and damage levels by 4O-55%, with full recovery of foliage the followmg year (9). A range of field trtals m the 1970s made it clear the OpNPV can be used to successfully control the tussock moth. Yet, the periodic nature of the problem, and difficulties in mass production of the virus, have discouraged commercial development. Nevertheless, the OpNPV has been registered by the US Forest Service, and was produced periodically on a contract

Natural Baculoviruses

for Insect Pests

309

basis, most recently by Blosys of Columbia, MD. However, this firm went out of business in 1996. 3.3. Baculoviruses as Viral Insecticides The viruses most commonly used or considered for development as microbial insecttcldes, in industrtahzed as well as less developed countries, are the NPVs (Table 1). GVs are used against pests for which no effective NPVs are known. NPVs have received the most study because they are common and easily isolated from many important lepidopteran pests, productton m their hosts IS straightforward and inexpensive, and the technology for formulatton and apphcatlon IS relatively simple and adaptable to standard pesticide apphcation methods. Typically, however, NPVs have a narrow host range, infecting only a few closely related species. Furthermore, though several can be grown m vitro in small volumes (ca 80-L cell cultures), fermentatton technology IS not currently available for then mass production on a large-scale commercral basis. These two key limitations have proven significant disincentrves for the commercial development of NPVs, especially m industrialized countrtes. The large chemical and pharmaceutical companies that might be expected to take an interest m NPVs, with rare exceptton, have shown httle interest in producing a product with a small market, and which must be grown in live caterpillars or sawfly larvae. This type of technology is more suitable for cottage industries m industrialized or developing countries. But, m industrialized countrtes, where all pathogens used as insecttcides must be registered by one or more governmental agencies, regulatory procedures and assoctated costs impede development by smaller companies, such as commercral predator and parastte msectaries. Even wrth these drawbacks, several NPVs have been registered as mtcrobtal insecticides m industrialized countries (though few are currently marketed), and, whether registered or not, many are used in less-developed countries, particularly for control of lepidopteran pests of field and vegetable crops. Moreover, there is renewed interest m developmg NPVs as insecticides, because of the adverse effects of chemical msectictdes and their Increasing costs, and because recombinant DNA technology offers potential for rmprovmg the efficacy of these viruses. For economrc reasons, as in the past, this interest is still restrtcted largely to NPVs of major lepidopteran pests and a few key sawfly pests.The high host spectfictty of NPVs makes them of most use where a smgle insect species is the only pest, or at least a key one of a particular crop. Utthty is increased tf the target insect 1snot very sensittve to Bt, and 1s resistant to chemtcals, or the latter are too costly, environmentally unacceptable, or unavailable. The viruses listed m Table 1 all meet, or at one ttme met, these criteria. The possibilities for improvement through genetlc engineering are dis-

G

G,

of Insects

Pme forests

NsNPV

NlNPV

Neodlpnon lecontel

Pme forests

Vegetables Douglas fir forests Vegetables, flowers Cotton

Decrduous forests

Cotton, vegetables

Soybeans Frutt orchards Apples, walnuts

Crop

for Control

MbNPV OpNPV SeNPV SINPV

LdNPV

HzNPV

AgNPV AoGV CpGV

VlnlS

Registered

Mamestra brasslcae Orgyla psuedotsugata Spodoptera exlgua Spodoptera Ilttorahs Sawfly larvae Neodlpnon sertlfer

Hekoverpa zea Heliothrs vwescens Lymantrra dupar

Caterpillars Antrcawa gemmatalu Adoxophyses orana Cydla pomonella

Target pest

Table 1 Major Baculoviruses

Canadian Forest Service

Neocheck-S Sentifervirus Lecontivrus

Canadtan Forest Servtce

Therm0 Trilogy Canadian Forest Servtce NPP (Calliope) Therm0 Trilogy Therm0 Trtology NPP (Calliope)

Agrtcola El Sol Andermatt Btocontrol Andermatt Btocontrol AgrEvo NPP (Calliope) Therm0 Trilogy

Producer

Gypcheck Dtsparvuus Mamestrm TM Btocontrol-I Spodex Spodopterm

VPN Capex Madex Granupom Carpovirusme Gemstar

Product name

Pests

Canada

Canada

Columbia, MD Canada France 1 Columbta, MD Columbia, MD France

Brazil Swttzerland Switzerland Europe France Columbra, MD

Country

Natural Baculoviruses

for Insect Pests

311

cussed m Chapter 15; here, the focus is on naturally occurring baculovlruses as viral Insecticides. 3.3.1. Product/on In Vivo and Formulation Baculovlruses are mass produced m larval hosts grown on an artlficlal diet or natural host plants. To maximize production of NPVs, larvae are infected per OSat an advanced stage of development, such as during the late fourth mstar, and reared either m groups, or individually, for species that are canmballstic. After ingesting the virus, the occlusion bodies dissolve m the alkaline midgut, releasing the virions. In lepldopterans, the virus first mvades midgut eplthellal cells, where, during the first 24 h of infection, it undergoes an Initial colonizmg phase of replication in the nuclei of these cells. No occlusion bodies are produced m these nuclei, but rather the progeny vlrlons migrate to the tracheal matrix or through the basement membrane, and invade and colonize almost all other tissues of the host. In these, the occlusion phase of repllcatlon occurs, during which the vlrlons are occluded m polyhedra. Maximum production of polyhedra occurs m tissues that are the most nutrient rich and metabohtally active, such as the fat body, epidermis, and tracheal matrix (3). The occlusion phase of viral disease occurs over a period of 5-l 0 d, and represents several cycles of replication as the virus spreads throughout the tissues and invades most host cells. The actual length of the disease depends on several factors, including the host and viral species, larval mstar at the time of mfectlon, amount of inoculum, and temperature. Near the end of the disease, after most polyhedra have formed, the nuclei lyse. As more and more nuclei lyse, the larva eventually dies, after which the body liquefies, releasing literally blllions of polyhedra. In commercial production, larvae may be harvested prior to liquefaction, to keep bacteria, which quickly colonize dead larvae, at a low level in the final product. Alternatively, antibiotics can be added to the dret to keep bacterial counts low. After the larval production phase 1scomplete, the larvae are collected and formulated. Formulation varies, depending on how the virus will be used (10-12). Both liquid and dust formations of viruses have been developed and tested, with and without additives, such as UV light protectants, antioxidants, gustatory stlmulants, and spreader-stickers. Additives often show improvements in efficacy m laboratory tests, and sometimes in percent larval mortality in field trials, but not usually when the criterion 1scrop yield. Thus, their additional cost generally reduces cost-effectiveness, Recently, however, the addition of stilbene brighteners has been shown to increase the efficacy of the NPV of the gypsy moth, Lymantrza dz’spar (LdNPV; 13). In addition, the inclusion of molasses (5-25%) in liquid formulattons has been shown to improve the efficacy m the field of some NPVs, such as the Op and LdNPVs (9,14).

312

Federici

The productton of leptdopteran GVs and sawfly NPVs is similar to that descrtbed for leprdopteran NPVs (5,11). As noted, however, the sawfly NPVs differ from the lepidopteran NPVs, in that the former only replicate and form polyhedra m midgut eptthehal cells. Polyhedral yields, therefore, are lower than those obtained with lepidopteran NPVs, but field application rates are correspondingly lower (5). 3.3.2. Efficacy and Use of Baculovirus insecticides The extent to which baculovnuses can be useful as msectrctdes depends on several factors, mcludmg the relative importance of the target pest m the pest complex attacking a crop, the amount of virus that must be used to control the pest in both the short term and long term (persistence and carryover), the value of the crop, and the cost and availability of alternative control measures Baculovn-uses are good candidates for use where a single lepidopteran species is the major pest for most of the growing season on a crop with a htgh cash value, when other available pest control methods are not cost-effective Examples include NPVs effective against insecticide-resistant spectes of Helicoverpa, Heliothu, and Spodoptera on such crops as cotton, corn, and sorghum, and, even more so, on tomatoes, strawberrtes, and flowers such as chrysanthemums. The cost-effectiveness of these viruses is determined by the amount of virus that must be applied, and the frequency of application necessary to keep the pest below an economic threshold. As noted earlier, this will vary wrth the vnus, pest, crop, and, more Importantly, among different countries. The amount of virus required to achieve effective control is typically expressed in terms of the number of polyhedra (for NPVs) or granules (for GVs) that must be applied per unit area, e.g., acres or hectares. To correlate this with larval productron costs, the number of applmation rates for different viruses can be translated into the number of larval equivalents (LEs)* required per hectare; illustrative examples are given in Table 2. The number of LEs required to obtain effective control IS crmcal to determlnation of cost-effectiveness, because of the cost of labor and materials that go into vnus productron. The best results with baculovnuses are obtained agamst moderate mfestatrons of early- to mid-instar larvae. Treatment of heavy *A value often used for the larval equivalent IS6 x 1O9polyhedra/larva This value IS based on the number of polyhedra typlcally obtamed from a fifth-mstar larvae of H zea (IO) However,the actualnumberof polyhedraobtamedfrom a larva variesconslderablywith the vuus Isolateand speciesof insectusedfor productlon Therefore,the numberof larval equivalentsrequiredfor effective control ISbestdeterminedempmcallythroughfield trials,andcanthenbeexpressed m termsasLEs, aswell asthe numberof polyhedrarequiredper treatmentfor effective control Regardless of the methodsfor quantlfymgthe virus, Its blologlcalactlvlty mustalsobe establishedthroughbloassays

Natural

Baculoviruses

313

for insect Pests

Table 2 Occlusion Body Yields and Typical for Representative Baculoviruses

Application

Rates

Occlusionbody VlrllS Yteld/larva Application rate/ha Larval equlvalents/haa Helicoverpa zea NPV 6x lo9 2 x 10” 33 Lymantrla drspar NPV 2x 109 5 x 10” 250 Orgyia pseudotsugata NPV 2x 109 3 x 10” 125 Spodoptera exrgua NPV 2X 109 1 x 10’2 500 Neodlprlon sertlfer NPV 3 x 107 7 x 109 200 Cydla pomonella GV 1 x 10” 2 x 10’3 200 OAmount reqmred to a reduce crop loss or pest population below economtc threshold

mfestations of advanced mstars (i.e., late fourth-instars and beyond) IS typltally not effective. Rates agamst moderate mfestatlons can range from 150 LEs/ha/treatment, usmg the H. zea NPV to control Helzothis vzmcem on cotton, to 500 LEs for the Spodoptera exigua NPV on lettuce. The number of LEs required to control a specific pest can vary with the crop, because of differences in plant phenology and chemistry. This type of mformatlon IS essential for evaluating whether a specific NPV merits commercial development, as well as use against a specific insect on a particular crop. There have been many reviews over the years on the use of viruses as control agents (14,16-22). Rather than re-examine this information again in detail, relevant data for key baculovlruses registered in the United Stateswill be used here to illustrate the above economic concepts and rates of application that are considered to be effective. 3.3.2.1.

HELIOTHIS NPV

Larvae of the noctutd moths, H. zea, Heliothis armigera, and H vwescens, continue to be major pests of crops, such as cotton, corn, tobacco, soybeans, and tomatoes, in the United States and many other countries. These species are susceptible to a virus known as the H. zea NPV (HzNPV). Because of the economic importance of the crops attacked and damage inflicted by these pests, especially H. virescens on cotton, the HzNPV was the first virus developed commercially in the United States (IO), and is one in which there remains considerable interest. The HzNPV was orlginally marketed under the trade name Elcar by Sandoz (Basel, Switzerland). However, It came to market at about the same time as the pyrethrold msecticldes, and the lower cost and higher efficacy of the latter essentially eliminated the Elcar cotton market. Over the past decade, high levels of resistance to pyrethroid insecticides has renewed inter-

Federia

314

est m the HzNPV, and Therm0 Trilogy m the United States (Columbia, MD) 1s currently producing the vn-us under the product name Gemstar. The best results with the HzNPV have been obtained on cotton, where, m the Umted States, the major pests are H vzrescens and H zea. Treatment of low to moderate mfestattons at a rate of from 10 x 1Oropolyhedra/O.4 ha, and 45 x 1Oropolyhedra/O.4 ha for heavy mfestattons, resulted m yields comparable to those obtained with chemical insecticides (15). Results obtained on other crops showed the vu-us was not as effective against these noctuid pests as it was on cotton (15). 3 3 2.2.

SPOOOPTERA

EXlGUA

NPV

The beet armyworm, S exlgua, IS a polyphagous insect that has emerged as one of the most important pests of vegetable crops m the world. It is a partlcularly important pest of tomatoes, celery, alfalfa, and strawberries m the United States, a major vegetable crop pest IS Southeast Asta, as well as a pest of chrysanthemums m the United Statesand The Netherlands. Much of Its pest status 1s attributable to tts resistance to most chemical msectictdes. The high cash value of many crops attacked by the beet armyworm led to the development of the S. exagua NPV (SeNPV) as the viral msectictde, Spodex Spodex IS sold m The Netherlands for use agamst the beet armyworm on chrysanthemums, and m Southeast Asia for use on vegetable crops Registration is also being sought m the Umted Statesfor use on vegetable crops, espectally fresh market tomatoes. Field trials of the SeNPV have shown that tt 1smost effective when used at rates of from 2 to 4 x lOI* polyhedra/ha, whether applied m the field on lettuce, or in glasshouses on chrysanthemums (16). Use of Spodex against forage crops, such as alfalfa, IS generally not cost-effective, because of the low value of the crop/ha. 3.3.2 3

LYMANTR/A

DWAR

NPV

The gypsy moth, L dzspar, 1san important pest, periodically, of deciduous trees in the northern hemisphere. Populations are cyclical, but, at their peaks, larvae are capable of total defoltatton of hundreds of square kilometers of forests m many areas of the eastern United States, Canada, eastern Europe, and Russia. The US Forest Service developed and registered a product known as Gypcheck, which contains as its active ingredient the L. dispar NPV (LdNPV), a vnus that 1shighly spectfic for gypsy moth larvae. Larvae can be controlled with chemical insecticides, Bt, and the LdNPV, but, m environmentally sensittve areas, such as most of the deciduous forest in the eastern Umted States, the virus 1s considered by many a better choice, because of tts selectivity. Field trials with Gypcheck have shown that it gives acceptable levels of tree protection when sprayed twice at a rate of 10’ ’ polyhedra/O.4 ha (14). Accept-

Natural Baculoviruses

for Insect Pests

315

able control is considered a reduction of 50% or more in defoliatton, which would have occurred m the absence of treatment. 3 3.2.4. NEOD/PR/ON SERTFER NPV

The European pine sawfly, Neodzprron sertzfer, is a major pest of pme forests m the United States, Canada, and Europe. Because of its importance, the N. sertzfer NPV (NsNPV) is the sawfly virus that has received the most attention as a control agent. It has been registered for use against the European pine sawfly m the United States and Canada by their respective forest service agencies Numerous field trials conducted with the NsNPV have shown that very good control can be obtained when the virus is applied at a rate of 5-9 x 1O9polyhedra/O 4 ha against moderate infestations of early-mstar larvae (5). Higher rates ( lOlo) were required for heavier infestations (5). 3 3 2 5

&D/A

POMONELLA

GV

One of the most important insect pests of apples worldwide IS the codlmg moth, Cydza pomonella. This species also attacks other tree crops, including plums, pears, and walnuts. On apples, larvae burrow directly mto the fruit right after hatching. Thus, they represent a dtfficult target for a mtcrobtal pesticide because of the short period of their exposure. No NPVs effective against the codling moth are known, so the only virus developed to date for control of this important pest is the C. pomonellu GV (CpGV). The CpGV has been registered for use in both the United States and Switzerland. The apple growing conditions in these two countries are very different. In Switzerland, the cooler climate results m lower codlmg moth populations and a shorter pest season. In the United States, especially in northern California, the warmer climate at lower elevations results m larger moth populattons and a pest season that can last for 10 wk. Thus, the total amount of virus that must be applied to get a marketable fresh apple crop m California is much higher than it is in Switzerland For this reason, the CpGV has found a steady market m Swttzerland, but has not found a commercial producer m the United States. However, apple farm cooperatives, especially those that grow orgamcally grown apples (no synthetic chemical pesticides or fertilizers), are currently developing plans to produce the CpGV. Regarding rates of application, control is effective when the virus is sprayed at a rate of lOi granules/O.4 ha. In Cahforma, effective control requires one treatment per week over a IO-wk period. In Switzerland and other cooler chmates, three treatments per seasonwill usually yield effective control of codlmg moth larvae.

Federici

316 3 3 2 6. AUTOGRAPHA CALFORNCA NPV

The viruses discussed above all have a narrow host spectrum, typically only mfectmg the target species and a few other closely related species. From an economtc standpoint, this 1soften considered a disadvantage of viral msectitides. Most of the larger companies interested in viral insecticides would consider a virus that had a broad host range, but one restrtcted to lepidopterans, to be a good candidate for development. For this reason, the NPV ofA. calzjbrnzca NPV (AcNPV), which is capable of infecting more than 50 leptdopteran species, has received considerable attention as a viral msecttcide, and was recently registered by Biosys as Gusano. Because Btosys went out of busmess m 1996, it is not clear whether the naturally occurrmg version of this virus will be developed further and marketed. However, the AcMNPV is the virus that most groups m academia and industry have used as a model for tmprovmg msectitidal efficacy through genetic engineering (see Chapter 15). 3.3.2.7.

ANT/CARS/A GEMMATALE NPV

The velvetbean caterpillar, Antlcarsza gemmatalis, is a major pest of soybeans m many soybean growmg areas, but particularly m Brazil. More than 15 yr ago, the Brazilian government mounted a control program aimed at developing the A gemmatalis NPV (AgNPV) as a viral msecttctde. Though this virus has not been registered in the United States, it is worthy of mention, because it has been one of the most successful documented control programs using a virus as an msecttcide. The AgNPV IS now used to control the velvetbean caterpillar on approx 1 million ha of soybeans m Brazil. The vn-us is produced both by maculation of larvae in production colonies, and by field collectton of larvae that have died from NPV disease (23,24). Rates of applmation vary with the mfestatton, but generally are m the range of l-2 x lOi* polyhedra/ha.

3.4. Use of Baculoviruses

in Developing Countries

Baculoviruses are used m many developing countries to control lepidopteran pests on field and vegetable crops, and even, in some cases,to control forest pests. Examples include the use of various Spodoptera NPVs to control S littoralis and S. lztura of field and vegetable crops m India, Africa, China, and other countries m southeast Asia, the use of the Spodoptera fruglperda NPV to control S.frugiperda on corn m Latin America; the HzNPV to control the H armigera on cotton m southeast Asia, and Heliothls/Hellcoverpa on cotton and vegetables in Latin America and India, and the use of the Pzeris GV to control cabbage worms m China. In most cases, the use of these viruses is much more extensive than in industrialized countries. Thus, on a local basis, these viruses are successful control agents. But, as noted above, these examples

Natural

Baculoviruses

for Insect Pests

377

cannot be used to conclude that baculovn-uses can be used successfully in developed countries, because the economic conditions are so different. In addltlon, the systemsof evaluation differ. For example, cost-effectiveness has typically not been evaluated m a country like China. Methods have been developed for virus production, and the virus is simply sprayed on the crop, frequently or infrequently, to eliminate the pest. Communes and farms responsible for production of vegetable and field crops have their own VII-W production facilities, and ample labor, so the calculation of labor costs attributable to virus production, and corresponding levels of control, are not determined. This will probably change as the economy in China continues to expand and cost effectiveness is more carefully scrutinized. However, even in Latin American countries, the NPVs are considered successful in vegetable crop production. The reason 1sthat labor costs are relatively low, and virus can be produced cheaply m small insectaries, or by collection of virus-infected larvae in the field at the end of the season. Baculoviruses have been used m developmg countries for decades, and this use would probably not have continued If the vu-uses were not considered cost-effective under local economic conditions. 4. Conclusions Viruses are not widely used at present in industrialized countries, because chemical msectlcldes are still readily available and effective, and because viruses have what are considered key limitations. In comparison to chemical insecticides, these limitations include a relatively slow speed of kill, a narrow spectrum of activity, little residual activity, and lack of a cost-effective system for mass production m vitro. On the other hand, these limltatlons have not inhibited the development and use of baculoviruses in several niche markets in the Umted States and Europe. In addition, NPVs and GVs are used in developing countries, especially on field and vegetable crops in China, India, and Brazil, as well as m many smaller countries in Latin America, Africa, and southeast Asia. The latter use results from the high cost of chemical msectlcldes m many of these countries, the development of insecticide resistance, low-to-moderate labor costs for virus production in VIVO,and from the fact that registration for use of viruses IS either not required or is easily obtained. As pressure mounts to reduce the use of synthetic chemical msectlcides, viruses may receive increased attention as alternatives to chemicals, particularly for the control of lepidopterous insects, for which no other effective control agents, such as Bt and parasites, exist. This may lead to an increased effort to develop and use conventional viruses in IPM programs in both industrialized and developing countries. In addition to conventional viruses, the development of recombinant DNA technology, i.e., genetic engineering, offers considerable promise for improving viral efficacy by reducmg or eliminating the maJor disadvantages of vuuses (see Chapter 15).

318

Federici

Some hope for the use of viruses m mdustrtahzed countrtes comes from the relative successof the bactermm Bt. Bt has been successful, even though it IS only 2% of the insecticide market, because it either did not have the major hmitation of viruses, or it overcame them For example, Bt’s performance compares favorably with chemical msecticides in many crop and forest systems in which lepidopterous insects are key or major pests It 1sfast-acting, relatively inexpensive, and easy to produce, formulate, and use It has a relatively narrow spectrum of activity, and its residual activity is rather low Yet the range of msects Bt controls continues to provide a market large enough to Justtfy commercial development. The potential for baculovnuses exists largely for pests in which Bt products are not effective, or m which resistance against these products or Bt transgemc plants is anticipated Still, the commercial success of viruses on a large scale, for crops such as cotton and corn, ~111require the development of techniques for mass production m vitro to meet market demand, and viruses that act more quickly to prevent feeding damage than naturally occurring viruses. On a smaller scale, there are many NPVs that could be used m industrialized countries for extstmg and new niche markets. Last, it should be noted that the window of opportunity for viral msecttcides could be narrowing m mdustrtalized countrtes because of several technologtcal developments. These mclude the development new types of narrow-spectrum synthetic chemical msecttcides, better strains of Bt targeted against H. zea and H. vwescens m the cotton market (the largest potential market for vu-al msecticides), and the development of insecticidal transgenic cotton and corn based on Bt proteins, again targeted against the same above pests as are the viral msecttctdes.The development of stgmficant resistance to any of these new technologies will contmue to provide opportunities for viral msecticides, but this window, for both naturally occurrmg and recombmant baculovu-uses, could well be limited to the next decade. References 1 Stemhaus,E A. (1949) Prznczples oflnsect Pathology. McGraw-Hill, New York 2 Volkman, L E , Bhssard, G W , Friesen, P , Keddie, B. A, Possee,R., and Theilmann, D A (1995) Family Baculovmdae,in Vwus Taxonomy (Murphy, F A , Fauquet,C. M , Bishop, D H. L , Ghabnal, S A, Jarvis, A W., Martelli, G P , Mayo,M A , andSummers,M D , eds.),Springer-Verlag,New York, pp 104-l 13 3 Federicl, B A (1993) Viral pathobiology in relation to insectcontrol, in Patkogens and Parasites of Insects, vol 2, Academic,San Diego, CA, pp 8l-l 0 1 4 Balch, R. E and Bird, F T (1944) A diseaseof the European spruce sawfly, Gzlpznzakercynzae [Htg ] and its place in natural control SczAgrzc. 25,65-80 5 Cunningham,J. C and Entwrstle,P F (1981) Control of sawflies by baculovn-us, in Mzcrobzal Control of Pest and Plant Diseases(Burges, H. D , ed.), Academic, London, pp 379-407

Natural

Baculoviruses

for Insect

Pests

379

6 Bedford, G 0 (198 1) Control of rhmocerous beetle by baculovnus, m Mxrobzal Control of Pest and Plant Diseases (Burges, H. D , ed ), Academrc, London, pp 409-426 7 Huger, A M (1966) A vtrus disease of the Indian rhinocerous beetle, Oryctes rhmocerous (L ) caused by a new type of Insect wrus, Rhabdionvwus oryctes gen. n., sp. n J Invertebr. Pathol. 8, 38-51. 8 Stern, V M and Federici, B. A (1990) Brological control of the western grapeleaf skeletonizer wrth a granulosis virus. CA Agriculture 44,2 1,22 9. Stelzer, M J and Neisess, J. (1978) Field efficacy tests, m The Douglas-Fir Tussock Moth A Synthesis (Brookes, M. H., Stark, R. W., and Campbell, R W., eds.), U. S. Forest Service Technical Bulletm 1585, U. S Department of Agrtculture, Washmgton, DC, pp 149-I 52. 10 Ignoffo, C. M (1973) Development of a viral insecttctde* concept to commerctaltzation Exp Parasztol 33,38@406. 11 Shaptro, M. (1986) In vivo productton of baculovu-uses, in The Biology of Baculovzruses, vol. 2 (Granados, R R. and Federq B. A, eds ), CRC, Boca Raton, FL, pp 3 l-61 12. Young, S Y., III and Yeartan, W C (1986) Formulation and apphcatton of baculovuuses, m Bzology ofBaculovzruses, vol 2 (Granados, R. R. and Federict, B. A , eds ), CRC, Boca Raton, FL, pp 157-179 13 Adams, J R , Shepard, C A , Shapiro, M., and Tompkms, G J (1994) Light and electron microscopy of the htstopathology of the midgut of gypsy moth larvae infected with LdMNPV plus a fluorescent brightener J Invertebr Path01 64, 156159 14 Lewts, F B (1981) Control of the gypsy moth by a baculovuus, m Mlcrobzal Control of Pest and Plant Dzseases (Burges, H D , ed ), Academic, London, pp. 363-377 15 Ignoffo, C M and Couch, T. L. (198 1) The nucleopolyhedrosts virus of Helzothzs species as a mtcrobial msecttcide, m Mlcroblal Control of Pest and Plant Dueases (Burges, H D., ed ), Academic, London, pp 330-362 16 Gelernter, W D , Toscano, N. C , Ktdo, K , and Federrct, B A (1986) Comparison of a nuclear polyhedrosts vu-us and chemical msectictdes for control of the beet armyworm, Spodoptera exigua (Lepidoptera Noctutdae) on head lettuce. J Econ Entomol

79,714-717.

17 Pmnock, D. E. (1975) Pest populations and vu-us dosage m relation to crop productivity, m Baculovwuses for Insect Pest Control Safety Conszderatzons (Summers, M. D., Engler, R , Flacon, L. A , and Vail, P., eds.), American Soctety for Microbtology Press, Washmgton, DC, pp 145-157. 18 Payne, C. C (1982) Insect viruses as control agents. Parasztologv 84,35-77 19. Payne, C C. (1988) Pathogens for the control of insects. where next? Phzl Trans R. Sot Lond B 318,225-248.

20 Huber, J (1986) Use of baculovuuses m pest management systems, m Bzology of Baculovzruses, vol 2 (Granados, R. R. and Federq B. A., eds ), CRC, Boca Raton, FL, pp 18 l-202

320

Fedencl

21 Fuxa, J R. (1990) New dnecttons for Insect control wtth baculovtruses, m Nebv Dzrectzons rn Blologzcal Control (Baker, R. and Dunn, P., eds ), LISS, New York, pp 97-113 22 Cory, J S. and Bishop, D H L. (1995) Use of baculovnuses as btologtcal msectrcrdes, m Methods In Molecular Bzology, vol 39 Baculovlrus Expression Protocols (Richardson, C. D , ed ), Humana, Totowa, NJ, pp. 277-294. 23 Moscardt, F. (1989) Use of viruses for pest control m Braztl The case of the nuclear polyhedrosrs ofthe soybean caterprllar, Antzcarsza gemmatalu Memorlas do Znstztuto Oswald0 Cruz (RIO de Janeiro) 84,51-56 24 Moscardi, F (1990) Development and use of soybean caterptllar baculovirus m Brazil, m Proceedings of the Vth International Colloquzum on Invertebrate Pathology and Mlcroblal Control (Pmnock, D E., ed ), Socrety for Invertebrate Pathology, Adelaide, pp 184-l 87

17 Recombinant

Baculoviruses

Michael F. Treaty 1. Introduction Baculoviridae is a family of occluded, invertebrate-specific pathogens, consrstmg of two genera* the nucleopolyhedroviruses (NPVs) and granulovu-uses. The maJority of basic and applied research efforts, as well as commerctal endeavors, have been focused on NPVs. In addition to vertebrate-mvertebrate selectivity, many NPVs are infectious against only certain species within the insect order Leprdoptera, and impart no direct adverse effects on members of other insect orders, such as Coleoptera, Hymenoptera, Neuroptera, and Diptera (I). Examples of lepidopteran-specific baculoviruses are those isolated from gypsy moth, Lymanma dispar (LdNPV), celery looper, Anagrapha falcifera Kirby (AfNPV), beet armyworm, Spodoptera exzgua (Hubner) (SeNPV), and cotton bollworm, Helicoverpa zea (Boddre) (HzNPV) Target specificity of NPVs make them good candidates for use m integrated pest management systems. Although several baculovnuses have been regrstered as commercial products, they have not gained widespread use m intensive agronomic systems.Two properties of NPVs that limit their utility as fohar insecticides are photolabrlity and the requirement for ingestion by the pest, both of whtch delay or reduce acquisition of a lethal dose by targeted insects Another important limitation of NPVs, and the focus of discussion m this chapter, is the length of time required to kill an infected Insect. Depending on the virus and pest species, it may take nearly a week or longer before an infected insect dies. Further, an NPV-infected larva continues to feed until time of death. Instead of providing rapid curative action, as IS commonly achieved with a synthetic insecticide, a foliar application of an NPV may allow too much crop damage to occur before the pest population is brought under control From Methods m El/otecbnology, vol 5 B~opesbcrdes Use and D&very Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

321

322

Treaty

Recently, genetic engmeermg has been used to enhance pestlcldal properties of NPVs, while mamtaming their desu-able pest-specific characterlstlcs Nucleopolyhedrovlruses have been genetlcally altered to enhance the speed with which they kill pests. Approaches to engineering NPVs as improved blologlcal msectlcides include deletion of genes that encode products prolongmg host survival, and Insertion of genes that express an insectlcldal protein durmg vu-al replication (1 e , utilization of the virus as a vector for carrymg and expressmg genes mto the insect’s body cavity) A nucleopolyhedrovlrus lsolated from alfalfa looper, Autographa callfornlca (AcNPV), has been widely used m both of the aforementioned recombinant strategies. The molecular genetics and pathogemclty of AcNPV has been characterized more thoroughly than any other baculovn-us. Further, host range of AcNPV includes several agriculturally Important pests, such as tobacco budworm, Helzothls vzrescens (Fabncms), cabbage looper, Trzchopluszani (Hubner), and S exzgua 7.7. Gene Deletion Nucleopolyhedroviruses isolated from A calrfornica (2-4), L dlspar (S), Mamestra brasslcae (MbNPV) (6), Spodoptera lrttoralls (SINPV) (7), and Buzura suppressaria (BuNPV) (8) each contam a gene that codes for expression of the enzymatic protein, ecdysterold UDP-glucosyltransferase (EGT) Hu et al (8) hypothesized that the EGT gene 1sublqultous among NPV species. When produced during vu-al rephcatlon, EGT conjugates sugar molecules to ecdysone, and renders it inactive. This process enables the virus to inhibit or delay molting by the infected host Smce an insect stops feeding during a molt, EGT essentially functions to prolong the length of time the insect feeds during viral Infection, which allows for increases m weight gam by the host insect and rephcatlon by the virus. O’Rellly and Miller (3) demonstrated that deletion m the EGT gene of AcNPV causedinfected fall armyworm, Spodoptera fruglperda (J. E Smith), to feed less and die about 30% sooner than larvae Infected with wild-type AcNPV. Slmllarly, Riegel et al. (5) reported that an EGT-deleted form of LdNPV killed larvae of L dispar at a rate approx 20% faster than that of its unmodified counterpart The first genetically Improved baculovirus tested under open field condotlons against natural pest infestations was EGT-deleted AcNPV (9) In these field trials, Treaty et al. (9) found that EGT-deleted AcNPV provided more consistent control of T. nzthan wild-type AcNPV on lettuce and cabbage, but differences m efficacy between the two NPVs were marginal, and usually not statlstlcally significant. Based on results from their field trials, Treaty et al concluded that any further changes that might be made m the baculovn-us genome (e.g , gene insertion), to enhance killing speed beyond that seen with EGT deletion, should produce a more commercially acceptable biopesticlde.

Recombinant Baculoviruses

323

1.2. Gene lnsertion A number of different genes encoding invertebrate toxms, neurohormones, or enzymes have been engmeered into nucleopolyhedroviruses. These msectictdal proteins are fast-actmg, target-spectfic at both pharmacokmetic and pharmocodynamic levels, and brologtcally active at very low concentrations (e g., 1O-l2 44). Briefly, construction of a recombinant NPV requires insertion of a synthetic copy of the foreign gene (may be a mutant or optimized codon vs wild-type codon) into a plasmid-transfer vector. The gene-inserted plasmid IS then propagated m the bacterium Escherzchia coli Following mass production and purification of the transfer vector, the foreign gene is transferred mto the NPV genome by cotransfection with wild-type viral DNA in a susceptible cell line Resultant recombinant viruses can then be separated from nonrecombinant forms by methods such as plaque assay. The recombinant clone is then propagated m permissive larval insects or cell lures (e.g., S.frugzperdu ovarian cell line, designated Sf9). Transcription of the toxin gene during vrral rephcanon 1scontrolled by a promoter gene that may be derived from NPV (e.g., polyhedron gene, p10 gene, and so on) or non-NPV (e.g., fruit fly, DrosophzEa melanogaster, heat shock [hsp70] gene) sources. Additronally, the toxin gene may be fused with a copy of a signal-pepttde coding sequence to facilitate secretion of the toxin from virus infected host cells. In 1988, Carbonell et al. (ZO) reported on their efforts to insert a scorpton, Buthus eupeus, toxin gene into the AcNPV genome Then recombinant (r) AcNPV expressed scorpion toxin m infected host cells, but larval ktllmg speed of the rAcNPV was similar to that of wild-type AcNPV. Smce this imttal recombinant NPV did not contam a signal sequence, the authors hypothesized that toxin could not be secreted from infected host cells and transported to neuronal target sites. Maeda (11) constructed a recombinant stlkworm, Bombyx morz, virus (BmNPV), which expressed hornworm, Manduca sex&, diuretic hormone Srlkworms infected with rBmNPV died 20% faster than larvae infected wtth wild-type BmNPV. Genes encoding the insect enzyme, juvenile hormone esterase (JHE), have been inserted into AcNPV with promrsmg msecticidal results (22,13). AcNPVJHE recombmants have been shown to reduce feeding and survival times of T. ~11larvae by as much as 30% vs wild-type AcNPV. The recombinant baculovnus vectoring system that has received the greatest attention to date is AcNPV, carrying a gene that encodes an insect-specific venom component of the Algerian scorpion, Androctonus australzs Hector (AaIT) This neurotoxm acts as a sodium channel agomst, causing repetitive firing of motor nerves and overstimulation of skeletal muscle (I4,15). Intoxication of insects with AaIT causescessation of feeding, paralyses, and eventual

324

Treaty

death. Studies conducted with muscle tissue, nerve fibers, and synaptosomes have demonstrated that AaIT only exerts rts impact against Insect species and not against crustacean, arachnid, or mammahan species {l6,17). Laboratory studies have demonstrated that various AcNPV-AaIT constructs killed larvae of selected leprdopteran species, mcludmg H. vzrescens,H. zea, S.frugiperda, and T. ni, in less than half the time required by wild-type AcNPV (18-20) Similarly, Maeda et al. (21) reported that BmNPV-AaIT subdued B. mori larvae in a significantly shorter period of time than wild-type BmNPV. In comparing response times of two AcNPV recombinants, Kunimt et al. (22) found that AcNPV-AaIT killed T IZZlarvae at a faster rate than AcNPV-JHE. Two other neurotoxm genes have been widely incorporated mto the AcNPV genome: LqhIT2, which encodes a venom component of the yellow scorpion, Lezrus quznquestrzatus lebraeus, and Tox34, whtch expresses TxP-I toxm derived from the straw itch mite, Pyemotes trztzczLqhIT2 binds to the neuronal sodium channel, but, unlike AaIT (whrch 1s excitatory), LqhIT2 IS a depressant-type toxm. Once produced and transported wrthm the virus-infected insect, LqhIT2 induces progressive suppression of neuromuscular transmission, resulting m complete relaxation of body musculature (i.e., flaccid paralysis). Like AaIT, this depressant-type toxin has been shown to be Insect-specific, rmpartmg no pharmacological effects agamst mammalian or crustacean species (23) The pharmacodynamics of TxP-I have not been fully defined, however, prelimmary evidence suggests that it causes a biochemical lesson at a gated ion channel at the presynapse. Studies m vwo have shown TxP-I to be nontoxic to mice, and to possesspotency against lepidopteran species at levels equal to or greater than scorpion-derived toxins (24,25). AcNPVs carrying either LqhIT2 or TxP-I have been evaluated against lepidopteran larvae, with each found to induce significantly quicker mortality than wild-type AcNPV (2526). Capabilities of isolating genes that encode msectrcidally active proteins and msertmg them into baculoviruses is rapidly evolvmg. Recently, Prikhod’ko et al. (27) reported on a series of rNPV clones that indivrdually expressed toxin genes from the funnel web spider, Agelerzopszsaperta, and two specres of sea anemone, Anemonia sulcata and Stzchodactyla helzanthus Further, Hughes et al. (28) demonstrated that AcNPV clones, carrying toxm genes derived from venoms of the spider, Dzguetza canztes (AcNPV-DTX9.2) or Tegenarza agnestis (AcNPV-TalTXl), reduced survival times m larvae of T nz,H virescens, and S. exigua by up to 33% vs wild-type AcNPV. Even the CryIA(c) endotoxin gene from the bacterium, Baczllus thuringienszs (Bt), has been inserted into, and expressed by, AcNPV (29). The only recombmant baculovnal expression systemsthat have been evaluated for insecticidal activity under field conditions, thus far, are AcNPV-AaIT

Recombinant

Baculowruses

325

and AcNPV-LqhIT2. Field trials were conducted m England and the United Statesduring 1993-1996. Much of the remainder of this chapter will be devoted to reviewing some of the procedures that have been implemented to assess rNPVs for insecticidal performance, benignancy to nontarget invertebrates, and environmental competitiveness relative to wild-type, parental strains. The chapter will conclude with a short discussion on further improvements m target pathogenictty and deployment strategies pertaining to rNPVs. 2. Insecticidal Activity Many of the techniques used to characterize biological activity of rNPVs are similar to those used to evaluate other biological insecticides (e.g., wtldtype baculovirus and bacteria), as well as synthetic msecticides. Some of the more common approaches are descrtbed below. 2.1. Labora tory Assays Bioassaysare mittally designedto quantify inherent improvements m pathogemcity of recombinant viruses, compared with their wild-type counterparts, based on parameterssuch asreduction m amount of virus (median lethal concentrationor dosage, LCjO or LD,,) and/or tnne required to kill the insect (median lethal time, LT,,). Hughes and Wood (30) authored a comprehenstve review pertaining to in vtvo assay of baculovtruses. As stated by these authors, exposure of insects to food contaminated wtth viral occluston bodies (OBs) 1sby far the most widely used means of admmistermg NPVs. Basically, insects are constantly, or temporarily, left on treated plant tissue or artifictal diet. If artifictal diet is used, viral OBs may be evenly spread across the surface or mcorporated throughout the matrix. Advantages of food contamination procedures include stmphcity, their close approximation to natural means of dosage acquisition by insects, and their applicability to any larval age. In using surface contammation methodology with artificial diet, Treaty et al. (20) showed that AcNPV-AaIT and wild-type AcNPV possessed similar LD,as (OB/cm2) against H vu-exerts, based on final mortality ratmgs taken 10 d after placement of third-instars on the diet. However, analyses of mortality data collected on a daily basis m this same study indicated that AcNPV-AaIT killed H virescens at a significantly faster rate than wild-type AcNPV, with calculated LTso values of 2.6 and 5.1 d, respectively. To further compare msecticidal properties of AcNPV-AaIT and wild-type AcNPV, Treaty et al. (previously unpublished data) held neonate H virescens in Petri dish arenas containing mdivtdual cotton leaves, which had been previously immersed (then air-dried) m aqueous suspensions of either virus (at 1 x lo5 OB/mL). In this leaf-dip system, AcNPV-AaIT not only killed H. vzrescens more quickly than wild-type AcNPV, but it also reduced the amount of foliage consumed by nearly 50% vs unmodified AcNPV (Table 1).

326

Treaty Table 1 Rate of Neonate H. virescens Mortality and Food Consumption on Cotton Leaves Treated with Recombinant or Wild-Type AcNPV

Treatment0 Mean % larval mortahty AcNPV-AaIT AcNPV Untreated Mean % consumed AcNPV-AaIT AcNPV Untreated

Days after infestation Three Four 51 a 5b Ob -

73 a 43 b 2c

1oc 19b 24 a

Five 86 a 87 a

5b -

Means within column and test parameter followed by same letter are not slgmficantly different (Duncan’s multlple range test, P = 0 05) %dlvldual cotton leaves were dipped m an aqueous suspension of either vnxs (1 x lo5 OB/mL), treatments were replicated threefold m a randomized complete block design (treatment rephcate = 6 leaves and 40-50 larvae)

A potenttal confoundmg aspect of contammated food bioassays IS that acqutsition of viral OBs is dependent on larval feeding behavior, which can vary wtthm and between species. Additionally, food sources used m these assaysmay directly impact viral infection processes and/or pharmacokmetrcs of the toxin expressed by a recombinant vu-us. For example, Teakle et al. (32) found that LC5,, values of wild-type HzNPV against Helzothlspunct~gera were not only posmvely correlated with increasing larval age, but they also fluctuated when dosing occurred at various periods within an mstar. It was suspected that changes in larval sensittvtty were mediated by factors in/on the midgut that impacted mttratton and spread of viral infection. Followmg btoassaysconducted with rAcNPVs and T nz larvae, Kunimi et al. (22) reported that larval age at time of exposure to OBs had a direct impact on viral pathogemcity, because LT,,s of AcNPV-JHE and AcNPV-AaIT significantly increased with larval mstar. Santiago-Alvarez and Ortiz-Garcia (32) showed that vn-ulence of NPVs can be highly dependent on speciesof plant to which they are applied. In contaminated-leaf assays,these researchers observed that S. 1dtorali.s larvae, fed NPV-treated leaves from Ricznus communes, were srgnificantly less permissive to viral mfectton than those dosed with leaves from cotton or potato. Further, Forschler et al. (33) demonstrated that H. zea larvae, which fed on HzNPV-treated cotton leaves, were about twofold less sensitive to the virus than those that fed on contaminated tomato leaves or artificial diet. Among the conclusrons drawn from their study (33) was that mteracttons mvolvmg host-

Recombinant Baculoviruses

327

plant foliar constituents and larval midgut conditions were probably responsible for reduced mortahty of insects fed and dosed on cotton. Indeed, there have been several reports (34-37) indicating that simultaneous mgestion by noctuid larvae of vu-al moculum with certain phytochemicals can have a profound effect on whether or not the insect develops a lethal infection, and how quickly the insect succumbs to viral mfection. It has been documented, for example, that plant phenohcs or oxidative enzymes are potent modulators of baculoviral disease. Differences in viral pathogemcity among contaminated food sources may also simply be related to palatability of the diets (hence, differential rates of dose acquisition by larvae), rather than complex chemical interactions (38). Recent studies conducted by Hoover et al. (39) indicated that storage conditions and presence of certain antibiotics in artificial diets also impact larval response times to NPVs, which is probably a result of spoilage, nutritional value, and/or palatability of the diets. As alternatives to contaminated food assays described above, Hughes and Wood (30) also described methods for admmistermg liquid suspensions of NPVs directly to lepidopteran larvae. Such methods permit delivery of a measured, known dose of virus to individual test msects. For example, calibrated vu-us suspensions may be delivered to individual larvae via per OSinJection with needles or capillary tubes. Such methodology is applicable only to large, or late-mstar larvae. Since larvae of many lepidopteran species readily drink free water, NPVs can also be presented to these msects by allowmg them to imbibe measured droplets of aqueous viral suspensions from surfaces such as parafilm Unlike oral injection assays, this methodology can accommodate small larvae. In both of these types of assays,care must be taken to ensure that the virus remains suspended throughout the dosing process. In addition to comparing median lethal doses and times among recombinant and wild-type baculovnuses, measurements taken on other biological or physiological responses may provide further insight about which vu-us might make a better insecticide. Jarvis et al. (40) observed that two AcNPV-AaIT constructs (differing m promoter genes) induced lethality m H. virescens at equal rates. However, average larval body weights at time of death differed between the two recombinants, indicating that one of the viruses caused larvae to stop feeding sooner than the other. Using per OSdosing techmque, followed by placement of treated larvae on artificial diet contammg Blue No. 1 stain, Hughes et al (28) were able to monitor both mortality and feeding inhibition (intensity of blue coloration m alimentary canal) induced by two different rAcNPVs m selected lepidopteran species. In this study, AcNPV carrying the spider toxm gene TalTXl killed larvae faster than the AcNPV carrying another spider toxin gene, DTX9.2. Conversely, AcNPV-DTX9.2 caused larval feedmg cessation quicker than AcNPV-TalTXl . Based on quantification of both

328

Treaty

parameters, LTSOsand FTS,,s(median time to feeding cessatton) were calculated for the two rAcNPVs. Hughes et al. concluded that, even though AcNPVTalTXl caused larvae to dte sooner, AcNPV-DTX9.2 might be the better candidate as an insecticide, since it exhibited the lower FT,,, a characteristic that could translate into better overall crop protectton. In whole-plant assays, Hoover et al. (41) observed that larvae of H virescens infected with AcNPVAaIT fell from treated cotton 5-l 1 h before they died, and were unable to climb back onto the plants and resume feeding. However, tf these fallen larvae were subsequently confined to cotton foliage, the were capable of feeding. Because H. vzrescens infected with rAcNPV fell from plants prior to feedmg cessation, there was a greater reduction in feeding damage than that which would be predicted from LT or FT+. From their study, Hoover et al. conceived the median time to knock-off, or KO,c. 2.2. Greenhouse Assays Greenhouse studies conducted with artifictal pest mfestattons on intact plants help determine if improved LC, LD, LT, FT, and/or KOsos exhibited by gene-inserted baculovnuses can indeed translate mto greater plant protection vs their wild-type counterparts. With relatively small quantities of material (formulated or nonformulated), one can use the greenhouse to measure impact of various factors on pesticidal performance of NPVs, such as insect feeding, crawhng, and resting behavior (1 e., dose acquisition), and complexity of plant architecture (i.e., distrtbution of dosage). The field stmulatton studies serve as an excellent intermediate step to larger-scale, outdoor trials. Treaty et al. (20) recently conducted a greenhouse study to compare wettable-powder formulations of AcNPV-AaIT and wild-type AcNPV for efficacy against H. vzrescenson cotton. Plants were mdtvidually grown m 0.25-m diameter pots to a height of about 0.5 m (early flower bud stage) prtor to mmatmg the test. With use of a small paintbrush, five neonate larvae were placed m each cotton terminal about 1 h before each treatment application session Plants were sprayed with aqueous suspensions of formulated virus m a chamber contammg a rotating boom equipped with standard agricultural nozzles, Plants were infested and sprayed 6 times at 6- to 7-d intervals. Treatment efficacy was assessed by mspectmg flower buds for feeding damage by H. vzrexens Results from this study clearly showed that, at equivalent rates of 2.5 x 1Or2OB/ha, AcNPV-AaIT provtded stgmficantly better pest control than wild-type AcNPV. At the end of the 6-wk study, cotton treated with recombinant and wild-type viruses averaged 18.9 and 46.9% damaged flower buds, respectively (untreated cotton had nearly 70% damaged buds). Using methodology similar to that described above, Treaty et al. (previously unpublished data) also compared AcNPV-AaIT and wild-type AcNPV

Recombinant Baculoviruses Table 2 Control of T. nl on Cabbage

329 Under Greenhouse

Mean % plant defohatton Treatment0 AcNPV-AaIT 1 2 x lOI OBiha AcNPV wild-type 1.2 x lOI OB/ha Untreated

Conditions

(1995)

Mean no. feeding holes per plant on 5 July

23 June

5 July

2c

2c

96 B

14 B I9A

17B 39 A

220 A 230 A

Means wlthm columns followed by same letter are not slgmticantly different (Duncan’s multiple range test, P = 0 05) Tabbage plants infested with neonate larvae (15/plant) and sprayed with treatments on 9, 1.5, 20,26, and 30 June.

for efficacy against T. ~11on cabbage. When potted cabbage had 4-cm diameter

heads, each plant was infested with 15 neonates and sprayed with treatments (each at 1.2 x lOI2 OB/ha) a total of 5 ttmes at 4- to 6-d intervals. Similar to findings m the aforementioned cotton study, results from this T. ni experiment demonstrated that AcNPV-AaIT possessedbetter Insecticidal characteristics than wild-type AcNPV. Following the final infestation/treatment-application session,AcNPV-AaIT- and wild-type-AcNPV-treated cabbage averaged 2 and 17% levels of defoliation, respectively (Table 2). Laboratory and greenhouse experimental systemsprovide a means for quantifying the interactive effects of inherent viral pathogenicity and dosage acquisition to/by lepidopteran pests under controlled conditions. Field trials are needed to fully assess the consistency

of msectictdal

performance

of a recom-

binant baculovirus under biotic and abiotrc stresses typically encountered in agroecosystems. Elements that could impinge on performance of a recombinant virus under field conditions include presence of mixed sizes/agesof targeted pest species at time of application, simultaneous infestations of pest species that may exhibit only slight permissiveness to infection by the particular vuus, and loss of viral OBs from foliage as a result of photodegradatton,

rainfall, or wind abrasion. In 1993, Cory et al. (42) conducted a caged-plot field study in England, m which foliar applications of AcNPV-AaIT were evaluated for efficacy agamst artificial mfestations of T. nz on cabbage. Results from this study showed that the recombinant virus killed T. nz at a faster rate, and reduced plant defohatton

by up to 87%, compared to cabbage treated with wild-type AcNPV.

330

Treaty

The first open field release of a baculovirus carrying a toxin gene (another AcNPV-AaIT construct) occurred on cotton m the Umted States during 1995 (20) In this cotton study, AcNPV-AaIT was sigmficantly more effective against H. vzYescenSthan H. zea, and tt killed larvae of both species at rates significantly faster than the noninserted form of the vn-us. However, formulattons of both vtruses used m the study appeared to be short-lived, because neither treatment exhibited significant insecticidal activity 2 d after foliar application During 1996, numerous field trials were conducted m cotton, cabbage, lettuce, and tobacco with an improved wettable-powder formulation of AcNPVAaIT (43) Aqueous dilutions of AcNPV-AaIT, as well as comparative msecticides, were sprayed onto the various crops by using small-plot equipment (e.g., backpack sprayers and tractors) containing hydrauhc booms and standard agricultural nozzles (e.g., hollow-cone types). Number of times treatments were applied for pest control varied by location, but ranged from 2 to 6 at about 3- to 7-d intervals. Seven trials were conducted on cotton, m which treatments were directed agamst mixed populations ofH virescens and H zea Generally, AcNPV-AaIT at 1.2 x lOi OB/ha provided control of the hehothme complex at levels equal to that of various Bt products (e.g., Dipel@). In three of these trials, this dosage of rAcNPV equaled the chemical insecticides, cypermethrin (Ammo@) and chlorfenapyr (Pirate@),m efficacy against this pest complex. Three tobacco field trials were also conducted during 1996. Unlike the aforementioned cotton tests, H vzrescens was the sole hehothme species mfestmg the tobacco sites, thus, performance of the baculovirus was not compromtsed by presence of significantly less AcNPV-permissive H zea In tobacco, AcNPV-AaIT at 0.5-l .2 x 1012OB/ha provided control of H vzrescens at levels consistently equal to those provided by commercial Bt products and chemical standards, acephate (Orthene@) and methomyl (Lannate@) Fmally, three field experiments were also conducted on leafy vegetables (cabbage and lettuce). AcNPV-AaIT at 1.0 x lOi OB/ha was as effective as Bt m controllmg T ni on cabbage. As expected, rAcNPV was not effective m controllmg AcNPV-msensitive species, such as diamondback moth, Plutella xylostella, and imported cabbageworm, Piem rapae In a lettuce trial, AcNPV-AaIT at 0 7-l .5 x lOI OB/ha caused up to 85% mortality in young (second-mstar) S exigua within 4 d after application. 3. Biological Selectivity In assessingthe safety of a recombinant baculovirus, it is important to consider whether the addition of a foreign gene alters host range of the virus, and if the recombmant protein produced m the virus-infected insect will present a hazard to other species. As with determining the potential of this technology

Recombinant

Baculoviruses

331

for pest control, most of the research pertaining to nontarget safety has been conducted with AcNPV-AaIT. Baculoviruses inherently cannot infect vertebrates, and, m order for toxin to be expressed by a recombinant baculovnus, it must first establish a productive systemic infection m the host. If an orgamsm is nonpermissive to infection by the vectoring agent, the foreign gene will not be transcribed. Host range is determined by viral gene products, rather than by those of the introduced gene. Second, m the case of AaIT, when toxin is produced in a host that is permissive to the virus, it too is insect-specific and poses no threat to vertebrate species (see Subheading 1.2.) Considerable research has been conducted to determine both absolute and relative (i.e., vs wild-type AcNPV) impacts of AcNPV-AaIT on survival of nonleptdopteran invertebrates. All studies reported so far indicate that this particular rAcNPV exhibits the same host range as its wild-type counterpart. For example, when provided with a diet consistmg of H virescens or T nl infected with AcNPV-AaIT, larvae and adults of predatory green lacewing, Chrysopa carlzea Stephens, insidious flower bug, Orius insidzous (Say), and carabid beetle, Pterostichus modldus, survived and behaved the same as those predators provided with untreated prey (44,45). Honeybee adults, Apzs mellifera, showed no adverse reaction to hemoceol mJections with budded form of AcNPV-AaIT (45) McCutchen et al. (46) showed that survival of the hymenopterous parasttotd,Microplitzs croceipes, m rAcNPV-treated H vzrescens larvae did not differ from survival rates m untreated hosts. Treaty et al. (47) conducted a series of laboratory assaysin which various predatory, herbivorous, and saprophytic invertebrates were exposed to AcNPV-AaIT via their arena habitats (e.g., soil or agar) and/or food sources (e.g., artificial diet, infected prey, and foliage). Many of these organisms were presented with budded and occluded forms of the virus, as well as with toxin, None of the mvertebrates tested m this assay series exhibited any adverse growth or survival responses to virus or toxin; they were: twospotted spider mite, Tetrunychus urticae Koch; boll weevil, Anthonomus grandls grandzs Boheman; western corn rotworm, Dlabrotzca vzrgzfera vzrgzfera LeConte; Japanese beetle, PopillzaJaponzca Newman; earthworm, Lumbrlcus spp; sp Ixeuticus Chinese mantid, Tenodera aridifolia sinensis Saussure; funnel web spider, Exeuticus spp, and subterranean termite, Reticulotermesflavipes Kollar. The range of virulence exhibited by AcNPV-AaIT, even in Lepidoptera, appears to be unaltered vs wild-type AcNPV. In a larval screening survey encompassing 52 different species of Lepidoptera, Bishop et al (44) showed that dosagemortality responses were similar for AcNPV-AaIT and wild-type AcNPV. Using techniques such aswhole-plant mspection, drop-cloth, sweep-net, and shake-bucket, Invertebrate population surveys were conducted at AcNPVAaIT field trial sites during 1995 and 1996 (20,43) These surveys demon-

332

Treaty

strated that weekly fohar applications of AcNPV-AaIT, at rates as high as 2 x 1Or* OB/ha, had no effects on diversity or density of nontarget arthropods (relative to untreated plots). Members of at least 26 famrltes from the insect orders Homoptera, Heteroptera, Hymenoptera, Coleoptera, Neuroptera, Thysanoptera, and Orthoptera were encountered in these surveys. 4. Environmental Fate Insertion of a gene encoding an msecttcidal protein, such as any of those described earlier, does nothing to alter physical characterrsttcs of the occluded virus Since these foreign DNA sequencesdo not affect structure of the mature occluston body, the perststence of a recombinant baculovnus 1s likely to be similar to that of its wild-type parent. The abiotic factor that ts most responsible for mactivation of baculovnal OBs m natural environments is ultraviolet (UV) light, primarily as earthbound solar wavelengths of about 290-390 nm (48-50) Although viral OBs may persist for months or years when buried m sot1 (51), they are rapidly destroyed when m then btoavatlable state (1 e., as foliar deposits that can be consumed by leaf-chewing Leptdoptera). Ignoffo and Baxter (52) showed that NPVs apphed to cotton foliage lost 70% of then mtttal btological activity following only 4 h of exposure to sunlight. Stmtlarly, Jones et al. (53) demonstrated that sunlight accounted for more than an 80% loss of mfectivtty by NPV over a 4-d period followmg application to cotton. Nucleopolyhedroviruses have been reported to be even more photolabtle than other insect pathogens, such as Bt (54) Sunlight m the UV range causesstrand breakage m viral DNA, and/or generates reactive radicles (e.g., peroxides) durmg irradiation of amino-acid residues, both of which inactivate baculovnus insecticides (55). Presence of free water (56), for example dew deposits on foliage, can accelerate destruction of NPVs by sunlight. Genetic fitness of recombinant baculoviruses (relative to that of wild-type forms) must also be considered when assessingthen envn-onmental perststence. As mentioned earlier, to date, the mam goal of engineering NPVs as biopesticides has been to hasten the speed with which they kill targeted pests. To this end, reduction in the LT,a subsequently leads to a shorter period of time for viral reproduction. Recently, several studies have shown that production of OBs by different AcNPV-AaIT clones m larval T nz and H. vwescens IS slgmticantly reduced relative to replication levels of wild-type AcNPV (nearly 1O-fold reductton, m some cases,based on OB yields measured upon death of larvae) (22,26,42,57-59). Further, m then caged-plot field study, Cory et al. (42) observed that transmission of vtrus between successtve T nl cohorts and/or generations was significantly less on AcNPV-AaIT-treated cabbage vs plots treated with wild-type AcNPV. These researchers attributed the reduction m second-round infection to decreased numbers of OBs produced m

Recombinant Baculcviruses

333

rAcNPV-mfected larvae, and to fewer number of larval cadavers that lysed on foliage, thus making fewer OBs available for consumption by other foliagefeeding individuals, Studies by Goulson (60) indicate that lepidopteran larvae infected with wild-type AcNPV undergo behavioral changes that probably benefit fitness of the vnus. In particular, NPV-infected larvae typically chmb to the top of the plant or apex of foliage, which results m optimal contammation of the plant canopy with viral OBs upon decomposition of the cadaver. Conversely, larvae infected with AcNPV-AaIT become paralyzed and commonly fall from the plant prior to death, without depositing OBs onto foliage (41,42). Cumulattvely, these laboratory and field studies indicate that the AaIT gene 1s of negative value to the vn-us, and probably ensures a competttive disadvantage of the rAcNPV relative to wild-type AcNPV. Usmg contaminated artificial diet methodology, a study was conducted at American Cyanamid (Princeton, NJ) to investigate competitiveness of wildtype AcNPV and AcNPV-AaIT by serial passage of a binary mixture through H. virescens larvae. Briefly, rAcNPV and wild-type AcNPV were mitially mixed in water at a numeric OB ratio of 10: 1, respectively. This mixture was used to establish SIX addmonal virus populattons via serial passage through successive cohorts of H. wrescens larvae (each third-mstar larva from each cohort was exposed to 4000 OB/16 cm* of diet, i.e., approx LD9, dosage). At each passage, the number of insects exhibiting contractile paralysis (indicative of AcNPV-AaIT infection) or body lysis (indicative of wild-type AcNPV infection) was recorded. Also, at the end of each passage, all insects were pooled, and OBs were extracted to make the next viral stock suspension. A portion of each stock was set aside as moculum for the next round of mfectton, and the remainder was used for preparation of viral DNA (DNA was analyzed by restriction enzyme digestion and quantitative DNA blot hybridizatton to determine relative abundance of AaIT-containing viral genomes at each passage). Results from this study showed that, following exposure to the original 10: 1, AaIT:wild-type binary mixture, 97% of H wrescens exhibited contractile paralysis at time of death. However, by the fifth passage (i.e., larvae tested with passage-4 inoculum), the fraction of larvae exhibiting contractile paralysis was 90% of larvae succumbed from apparent wild-type AcNPV infection, which resulted in body lysis). Further, DNA hybridization showed that the AaIT genome effectively disappeared from the viral population by passage-6. Results from this laboratory assayprovides further evidence that the AaIT trait confers a survival disadvantage to AcNPV, rapidly driving the virus to virtual extinction. Another issue pertaining to commercial deployment of a recombinant baculovirus for pest control is whether the foreign gene might be able to move to a virus with a different host range, and therefore affect other species Genetic

334

Treaty

exchange, or recombmation between baculovtruses, probably occurs m nature, but only between closely related viruses. The viruses must share at least one common host, and undergo identical modes of rephcatton in the same cellular compartment. For example, genetic exchange between cytoplasmic polyhedrovnuses (CPVs) and nucleopolyhedrovnuses could not occur, since they replicate m cellular cytoplasm and nucleus, respectively. Addttionally, CPVs undergo RNA rephcatton, but NPVs are DNA viruses In the case of AcNPVAaIT, exchange of its genetic mformation could only occur with another AcNPV-like vnus. Further, as noted earlier, the addition of an AaIT gene to another NPV during a recombmation event would render the recipient virus less environmentally fit, thus putting it at a selective disadvantage vs its wildtype counterpart

5. Additional

Molecular

Design and Deployment

Strategies

Once the viral mfection process IS nnttated, the msecttcidal properties of an rNPV are dictated by a combmation of production rate and potency of the encoded deleterious protein. In addition to opttmrzmg the toxin-gene type and/or codon, improvements m rate of gene transcription and translation can also lead to accelerated killmg speeds imparted by rNPVs The rate of toxm production m a permissive host is dependent on several factors, one of which IS the promoter that controls expression of the toxin-gene Two groups of NPV-derived promoters that have been used m recombinant viral Insecticides are the early and late promoters. Early promoters are activated prior to onset of viral DNA rephcation, and include the immediate-early (e.g., zel and ze2 genes) and delayed-early (e.g., 35K gene) types The DA26 gene also encodes an early promoter, but it has not been clearly determined if rt is an immediate or delayed type. Late vn-al promoters are activated after onset of vu-al DNA replication, and, as such, are dependent on protems produced during viral DNA replication. Examples of late viral promoters mclude 6.9k, and the very late p10 and polyhedron genes, both of which are involved m productton of structural protems used m occlusion body formatton. Lu et al. (62) constructed a series of AcNPV-Tox34 recombmants under control of different promoters, m order to evaluate then influences on toxm expression m lepidopteran larvae. Among the findmgs of this study was that time required for insect paralysis was promoter-dependent, with the late 6.9k being the most effective across different host species. Jarvts et al. (40) found that AcNPVAaIT constructs containing either tel or pl0 promoters killed H. vzrescerzs at equivalent rates m time However, at time of death, H vzre.scenS infected with the iel -containing rAcNPV were significantly smaller than those infected with the plO-containing virus, mdicatmg that expression of AaIT earlier m the infection process enhanced the ability of the rAcNPV to reduce feeding activ-

Recombinant

Baculoviruses

335

ity of the pest. Conversely, m working with AcNPV-Tox34 constructs, Tomalskt and Miller (57) found that a hybrid late promoter was more effective m mducmg paralysis m T. ni larvae than a particular early promoter (designated PETL).It was determined that PETLwas not a strong enough promoter to drive adequate expression of Tox34. As indicated throughout this chapter, much of the research directed at improving pesticidal properties of baculoviruses has been conducted with AcNPV. However, m terms of tts potential to provide acceptable pest control, a rAcNPV might have utility against only the two most AcNPV-permissive pest species, T ni and H virescens. Limited, effective host range would present a barrier to widespread use of rAcNPVs (should one be commercialized). One way to design a recombinant baculovirus with a more desirable target profile is to insert a toxm gene mto a wild-type virus, which naturally exhibits high levels of pathogemcity to the pest complex one needs to control. For example, HzNPV recombinants were recently engmeered to carry either the LqhIT2 or AaIT toxin-gene (62,63). HzNPV-LqhIT2 and HzNPVAaIT have potential to be a more effecttve biopesticide than AcNPV-AaIT for control of the heliothme complex m cotton (i.e., as a vectoring agent, AcNPV is weak m pathogenicity to H. zea relative to H. vzrescens). In the future, it may be possible to genetically manipulate host-range characteristics of baculovu-uses, thus tailoring their pesttcidal spectrum to the needs of key crop markets. Thiem et al. (64) and Du and Thiem (65) inserted a gene from LdNPV (designated host-range factor 1 [hrfl]) into the AcNPV genome, which resulted m producmg a AcNPV recombinant capable of growing m a gypsy moth cell lme normally nonpermissive to replication by AcNPV. The hrfl gene allows AcNPV to overcome a blockage in viral protein synthesis typically encountered by it m the gypsy moth cell line. Insertion of two or more toxin genes mto baculovn-uses may offer an additional approach to enhancmg pesttcidal efficacy of these pathogens. Herman et al. (66) found that binary mixtures of AaIT and LqhIT injected into larval H vzrescens induced 5- to lo-fold levels of potentiation. These authors suggested that simultaneous expression in baculovnuses of synergistic combmations of msecticidal proteins could lead to even more potent, insect-selective biopesticides. Although improvements m recombinant baculovirus technology are currently underway, results from greenhouse and field trials conducted to date with AcNPV-AaIT provide some insight into potential deployment scenarios when such technology becomes commercialized. For example, m row-crop production systems,recombinant baculoviruses could be utilized in any of several situations for control of lepidopteran pest species. as biological control agents m rotational-use strategies, to relax selection pressure exerted by synthetic msecticides, or even Bt on pest populations; for use in binary mixtures

336

Treaty

with certam chermcal msectictdes, and as augmentation of transgemc crop varieties and estabhshment of binary transgemc host-plant resistance/biocontrol systems (43,67). In addition to using transgemc techniques to improve virulence of baculoviruses against targeted pests, research will be required to enhance dosage delivery and availability to pestiferous Lepidoptera. Just as with wild-type insect pathogens, which are used for insect control m cropping systems, proper design of spray-application and formulation parameters could result in a rNPV product that is deposited onto the crop m a manner appropriate with feeding behavior of the pest (1 e., optimum transfer of virus from leaf to pest), and is more resistant to loss from crop foliage because of abiotic factors (e.g., sunlight, wind, or ram abrasion). Improvements m mass-production technology (probably m vitro) will be needed to ensure that price (or cost-effectiveness) of the rNPV product to the end user is competitive with that of currently used insecticides.

References 1 Bishop, D. H L , Entwlstle, P F , Cameron, I. R , Allen, C J , and Possee, R D (1988) Field trial of genetically engineered baculovnus insecticides, in The Release of Genetzcally Engrneered Organisms (Susman, M., Collins, C H., Skmner, F A., and Stewart-Tull, D E., eds.), Academic, New York, pp. 143-179 2 O’Reilly, D. R. and Miller, L K. (1989) A baculovnns blocks molting by producing ecdysteroid UDP-glucosyltransferase. Sczence 245, 111 O-l 112 3. O’Reilly, D R. and Miller, L. K (199 1) Improvement of a baculovlrus pesticide by deletion of the egt gene Bzo/Technology 9, 1086-1089 4. O’Rellly, D. R. (1995) Baculovnus-encoded ecdysterold UDP-glucosyltransferases. Insect Blochem Mel BIOI 25,541-550 5 Rlegel, C I , Park, E., Burand, J , and Slavicek, J (1994) Deletton of the Lymantrza dupar multicapsid nuclear polyhedrosis virus EGT gene enhances viral killing speed General Technical Report NE-188, USDA Interagency Gypsy Moth Research Forum, Beltsville, MD 6. Clarke, E E., Tristem, M., Gory, J S , and O’Reilly, D R. (1996) Characterization of the ecdysteroid UDP-glucosyltransferase gene from Mamestra brasszcae nucleopolyhedrovnns. J Gen Vwol 77,2865-287 1 7. Faktor, 0, Toaster-Achituv, M., and Kamensky, B (1995) Identification and nucleottde sequence of an ecdysteroid UDP-glucosyltransferase gene of Spodoptera llttoralls multicapsid nuclear polyhedrosis vans Virus Genes 11,47-52 8. Ku, Z. H , Broer, R., Westlaken, J , Martens, J W M., Jm, F , Jehle, J A , Wang, L M , and Vlak, J. M. (1997) Characterization of ecdysteroid UDP-glucosyltransferase gene of a single nucleocapsid nucleopolyhedrovnns of Buzura suppressarla Vwus Res 47, 91-97 9. Treaty, M. F , All, J N , and Ghidiu, G. M (1997) Impact of ecdysteroid UDPglucosyltransferase gene deletion on efficacy of a baculovnus against Helzothzs vwescens and Trzchoplusla nz J Econ Entomol., in press

Recombinant

Baculoviruses

337

10. Carbonell, L F , Hodge, M. R., Tomalski, M. D., and Miller, L K. (1988) Synthesis of a gene codmg for an Insect-specific scorpton neurotoxm and attempts to express it usmg baculovirus vectors. Gene 73,409%4 18. 11 Maeda, S. (1989) Increased msecttcidal effect of a recombinant baculovirus carrymg a synthetic dmretic hormone. Bzochem Biophys Res Commun 165, I 177-I 183 12 Hammock, B. D., Bonnmg, B C., Possee, R D., Hanzltk, T N , and Maeda, S (1990) Expression and effects of juvenile hormone esterase m a baculovirus vector. Nature 344,458-46 1 13 Bonnmg, B C and Hammock, B D. (1996) Development of recombinant baculovirus for insect control Ann. Rev Entomol 41, 19 I,2 10 14 Zlotkin, E., Rochat, H , Kopeyan, C., Miranda, F , and Lissitzky S (1971) Puritication and properties of the Insect toxin from venom of the scorpion Androctonus australis Btochemte 53, 1073-1078. 15 Walther, C , Zlotkm, E., and Rathmayer, W (1976) Action of different toxins from the scorpion Androctonus australis on a locust nerve-muscle preparation. J Insect Physzol 22, 1187-l 194 16. Teitelbaum, Z., Lazarovict, P., and Zlotkin, E. (1979) Selective bmdmg of scorpion venom Insect toxin to insect nervous tissue. Insect Btochem 9, 343-346 17. Zlotkm, E. (1986) The interaction of Insect selecttve neurotoxms from scorpion venoms with insect neuronal membranes, m Neuropharmacology and Pestzctde Actron (Ford, M. G , Lunt, G. G., Reay, R. C., and Usherwood, P N. R , eds.), Elhs Horwood, Chtchester, UK, pp. 352-383. 18 McCutchen, B F , Choudary, P. V. Crenshaw, R., Maddox, D., Kamita, S G , Pelekar, N , et al (1991) Development of a recombinant baculovirus expressmg an Insect selective neurotoxm: potential for pest control. Bzo/Technology 9,848-852 19 Steward, L M D., Hurst, M , Ferber, M L , Merryweather, A T , Cayley, P J , and Possee, R. D (1991) Constructton of an improved baculovirus msectictde containing an Insect-specific toxin gene. Nature 352, 85-88 20. Treaty, M. F. and All, J. N. (1996) Impact of insect-specific AaHIT gene insertion on inherent bioactivity of baculovuus against tobacco budworm, Helzothts vzrescens, and cabbage looper, Trtchoplusta nt, m Proceedtngs, Beltwtde Cotton Conferences (Duggar, P and Richter, D. A., eds ), National Cotton Council, Memphts, TN, pp. 911-917 21 Maeda, S , Volrath, S L , Hanzhk, T N , Harper, S A , MaJima, K , Maddox, D. W , Hammock, B. D , and Fowler, E. (199 1) Insecticidal effects of an insect-specific neurotoxm expressed by a recombinant baculovnus. Vtrology 184,777-780. 22. Kummi, Y , Fuxa, J R , and Hammock, B D (1996) Comparison of wild type and genetically engineered nuclear polyhedrosts viruses ofAutographa caltforntca for mortaltty, virus replication and polyhedra production m Trzchoplusta 111larvae Entomol Exp Appl 81,251-257. 23 Zlotkm, E , Eitan, M., Bmdokas, M. E , Adams, M E , Moyer, M , Burkhart, W , and Fowler, E. (199 1) Functional duality and structural uniqueness of depressant insect-selective neurotoxms Btochemistry 30,48 14-482 1.

338

Treaty

24. Tomalskr, M. D , Kutney, R., Bruce, W A, Brown, M R , Blum, M S , and Travis, J (1989) Purification and characterization of insect toxins derived from the mrte Pymotes trrticz Toxzcon 27, 115 l-l 167 25. Tomalskr, M. and Mrller, L. K. (1991) Insect paralysis by baculovnus medrated expression of a mite neurotoxm gene Nature 352, 82-85 26. DuPont Agrtcultural Products (1996) Nottficatton to conduct small-scale field testing of a genetically engmeered mrcrobral pestrcrde (no 352-NMP-U), m Federal Regzster, United States Environmental Protection Agency, Washmgton, DC, 61,24,934-24,935. 27. Prrkhod’ko, G. G , Robson, M., Warmke, J. W , Cohen, C. J., Smith, M M., Wang, P , et al (1996) Properties of three baculovnus-expressing genes that encode Insect-selective toxins cc-Aga-IVB, AsII, and ShI Bzol Control 7, 236-244 28 Hughes, P R , Wood, H A, Breen, J P , Sampson, S F , Duggan, A. J , and Dybas, J A (1997) Enhanced bioactrvrty of recombinant baculoviruses expressing insect-specrfic spader toxins m leptdopteran crop pests. J Invert Path01 69, 112-118. 29 Shih, C J and Hu, Y C. (1996) The msectrcrdal activity of genetically engrneered Autographa calzforruca nuclear polyhedrosis vn-us contammg the Bacillus thurzngzenszs toxin gene to St9 cell line and Spodoptera lltura larvae Zhongua Kunchong 16, l-12 30 Hughes, P R. and Wood, H. A (1986) In-vrvo and m-vitro bioassay methods for baculovnuses, m The Biology of Baculoviruses, Practical Appllcatlon for Insect Control (Granados,R R andFederrcr,B A , eds.),CRC, Boca Raton, FL, pp 1-26 31 Teakle, R. E , Jensen,J M , and Giles, J E (1986) Age related susceptibility of Helzothzspunctlgera to a commercral formulatron of nuclear polyhedrosrs vrrus J Invert Path01 47, 82-92. 32 Santiago-Alvarez, C and Orttz-Garcia, R (1992) The Influence of host plant on the susceptrbrhtyofSpodoptera lutoralzs larvaeto Spodopterallttoralls NPV. J Appl Entomol 114, 124-I 30 33 Forschler, B. T , Young, S Y., and Felton, G W. (1992) Diet and susceptrbrltty of Hellcoverpa zea to a nuclear polyhedrosrsvirus, Environ Entomol 21, 1220-1223 34 Felton, G W and Duffey, S S (1990) Inactivation of a baculovrrus by qumones formed m an insect damagedplant ttssue J Chem Ecol 16, 121l-1236. 35 Hunter, M D and Schultz, J C (1993) Induced plant defensesbreached phytochemrcal mductron protects an herbrvore from drseaseOecologma94, 195-203 36 Duffey, S S , Hoover, K , Bonnmg, B C , and Hammock, B D (1995) The impact of host plant on the efficacy of baculovnuses, m Reviews znPesticide Toxzcology (Roe, M and Kuhr, R , eds), CT1 Toxrcology Commumcattons,Raleigh, NC, pp 137-275. 37 Young, S Y , Yang, J G , and Felton, G W (1995) Inhibitory effects of dietary tanmson mfecttvtty of a nuclear polyhedrosrsvirus to Helzcoverpa zea (Noctmdae: Leprdoptera) Bzol Control 5, 145-l 50 38 Jones, K A (1988) PhD Thesis, Umverstty of Reading, Chatham, Kent, UK. 39 Hoover, K , Schultz, C. M , Lane, S S., Bonnmg, B C., Hammock, B D , and Duffev. S. S 11997) Effects of diet-age and strentomvcm on virulence of

Recombinant Bacdoviruses Autographa calzfornzca nucleopolyhedrovnus Path01 69, M-50.

339 against tobacco budworm J Invert

40 Jarvts, D L , Reilly, L M , Hoover, K , Schultz, C., Hammock, B D., and Guarmo, L. A. (1996) Construction and charactertzatton of Immediate early baculovlrus pesttcrdes Bzol Control 7,228-235. 41 Hoover, K., Schultz, C M , Lane, S. S., Bonnmg, B C., Duffey, S S., McCutchen, B F , and Hammock, B. D (1995) Reduction m damage to cotton plants by a recombmant baculovnus that knocks moribund larvae of Heliothis vzrescens off the plant. Bzol Control 5,419-426 42. Cory, J. S., Hurst, M. L., Wtlhams, T , Haik, R. S., Goulson, D , Greer, B. M , et al (1994) Field trial of a genetically improved baculovirus insecttctde Nature 370, 138-140. 43 Treaty, M F. (1997) Efficacy and non-target arthropod safety of an AaIT geneInserted baculovuus. results from field and laboratory studies conducted durmg 1995-1996, m Bzopestzczdes and Transgenzc Plants, Internattonal Busmess Commumcations, Southborough, MA, m press. 44 Bishop, D. H. L., Hnst, M L , Possee, R. D , and Coty, J. S (1995) Genettc engtneermg of microbes* virus msecticides-a case study, in Fzf@ Years ofAntzmzcrobzals Past Perspectzves and Future Trends (Hunter, P. A., Darby, G. K , and Russell, N. J., eds ), Cambridge University Press, Cambridge, UK, pp 249-277 45 Heinz, K M , McCutchen, B F , Hermann, R., Parella, M P , and Hammock, B D. (1995) Direct effects of recombinant nuclear polyhedrosts viruses on selected nontarget organisms. J Econ Entomol 88,259-264 46 McCutchen, B F , Hermann, R., Heinz, K M., Parella, M P , and Hammock, B D (1996) Effects of recombinant baculovuuses on a nontarget endoparasttoid of Helzothzs vzrescens Bzol Control $45-50.

47 Treaty, M F., All, J N , and Kukel, C. F (1997) Invertebrate selectivity of a recombinant baculovtrus* case study on AaHIT gene-Inserted Autographa calzfornzca nuclear polyhedrosis virus, m New Developments zn Entomology (Bondart, K , ed ), Research Stgnpost, Trtvandrum, India, pp 57-68. 48. Jaques, R P (1968) The Inactivation of the nuclear polyhedrosts vzrus of Trzchoplusza nz by gamma and ultravtolet radiation. Can J Mzcrobiol. 14, 116 l-l 163 49 Jaques, R P (1977) Stabthty of entomopathogemc viruses, m Envzronmental Stabzlzty ofhrlrcrobzal Insectzczdes (Ignoffo, C. M. and Hostetter, D. L., eds ), Mtsc Publ. Entomol. Sot Am 10,99-l 16. 50 Jones, K A., Moawad, G , McKinley, D. J., and Grzywacz, D. (1993) The effect of natural sunhght on Spodoptera lzttoralzs nuclear polyhedrosts vuus. Bzocontrol Scz Technol 3, 189-197 5 1. Jaques, J. P (1969) Soil stability of baculoviruses. J Invert Pathol. 13,256263 52 Ignoffo, C M. and Batzer, 0. F. (1971) Microencapsulation and ultraviolet protectants to increase sunlight stability of an insect VU-US J Econ Entomol 64, 850-853 53 Jones, K A and McKinley, D J (1987) Persistence of Spodoptera lzttoralzs nuclear polyhedrosls virus on cotton in Egypt Aspects Appl Bzol 14, 323-334

Treaty

340

54 Ignoffo, C. M , Hostetter, D. L , Sikorowski, R. P., Sutter, G., and Brooks, W M (1977) Inactivatron of respective species of entomopathogemc viruses, a bacterium, fungus, and a protozoan by ultraviolet light source. Envzron Entomol 6,4 114 15 55 Ignoffo, C. M , Rice, W. C., and McIntosh, A H. (1989) Inactivation of nonoccluded and occluded baculovnuses and baculovn-us-DNA exposed to simulated sunlight. Envzron Entomol 18, 177-183 56 Ignoffo, C M and Garcia, C (1992) Combmations of environmental factors and simulated sunlight affecting activity of inclusion bodies of the Helzothzs nucleopolyhedrosis vu-us Environ Entomol 21,2 1O-2 13 57. Tomalski, M. and Miller, L. K. (1991) Expression of a paralytic neurotoxin gene to improve insect baculovnuses as biopesticides. Bzo/Technology 10, 545-549 58 American Cyanamid (1994) Notification to conduct small-scale field testmg of a genetically altered baculovtrus (no. 241-NMP), m Federal Regzster, United States Environmental Protection Agency, Washmgton, DC 59 Ignoffo, C. and Garcia, C (1996) Rate of larval lysis and yield and activity of mcluston bodies harvested from Trzchoplusza nz larvae fed a wild or recombinant strain of the nuclear polyhedrosis virus of Autographa calzfornzca J Invert Path01 68, 196198. 60 Goulson, D. (1997) Modtlication of host behavior during baculovual mfection Oecologla 109,2 19-228 61 Lu, A , Seshagm, S , and Miller, L K. (1996) Signal sequence and promoter effects on the efficacy of toxm-expressmg baculovuuses as biopesticides Blol Control 7,32&332.

62 DuPont Agricultural Products (1997) Notification to conduct small-scale field testing of genetically engineered mtcrobtal pesticides (no. 352-NMP-004), m Federal Register, United States Environmental Protection Agency, Washmgton, DC, 62,23,448-23,449. 63 American Cyanamid (1997) Notification to conduct small-scale testing of a genetically engineered microbial pesticide, m FederaZ Regzster, United States Envtronmental Protection Agency, Washmgton, DC, 62, 39,5 18 64 Thiem, S M , Du, X , Qentin, M. E., and Berner, N M (1996) Identification of a baculovnus gene that promotes Autographa calzfornzca nuclear polyhedrosis virus replication m a nonpermissive insect cell line J Vzrol 70,2221-2229 65. Du, X and Thiem, S. M. (1997) Characterization of host range factor 1 (hrf-1) expression in Lymantrla dlspar nucleopolyhedrovnus- and recombmant Autographa calrfornxa nucleopolyhedrovnus-Infected IPLB-Ld652Y Vzrology 227, 420-430 66 Hermann, R., Moskowitz, H , Zlotkm, E., and Hammock, B D (1995) Positive cooperativity among insecticidal scorpion toxins Toxzcon 33, 1099-l 102 67 All, J N and Treaty, M. F. (1997) Improved control of Helzothzs vlrescens and Helzcoverpa zea with a recombinant form ofAutographa calzfornzca nuclear polyhedrosts virus and interaction with Bollgard@ cotton, m Proceedzngs Beltwzde Cotton Conferences (Duggar, P and Richter, D. A., eds.), National Cotton Council, Memphis, TN, pp 12941296.

Joint Actions of Baculoviruses and Other Control Agents William F. McCutchen and Lindsey Flexner 1.

Introduction

Baculoviruses are arthropod-specific viruses that have been utthzed as biological control agents since 1930 (1). Several advantages are associated with the use of baculovnuses for pest control, including the host specificity and environmental

compatibility

attributes that offer major advantages over clas-

sical insecticides (2,3). Because of these attributes, baculoviruses may be readily integrated mto an IPM program because they do not adversely affect beneficial insects and other nontarget organisms (2,4,5). However, the success of baculovtruses as biological control agents of insect pests has been widely variable. Over the past two decades, baculovn-uses have been commercialized for the control of codling moth (Cydia pomonella), gypsy moth (Lymantria dispar), corn earworm (Helzcoverpa zea), tobacco budworm (Helzothzs vzrescens), beet armyworm (Spodoptera exigua), and the cabbage looper (Trzchoplusia ni) in the United States; the rhinoceros beetle (Oryctes rhznoceras) in the Pacific; the velvetbean caterpillar (Antacarsia gemmatalis) m Brazil (6-8); and Spodoptera exigua and Cydia pomonella in Europe (9,lO). In addition, large scale control programs for periodic forest pests, such as the Douglas-Fir tussock moth (Orgyia pseudotsugata), Eastern Spruce Budworm (Chorzstoneura fumzjkana), and European pme sawfly (Neodiprion sertzfer), have been conducted in countries worldwide, including Canada, the United States, the United Kingdom, Finland, Norway, Sweden, Austria, Italy, Poland, and the Soviet Republic (former USSR) (11). Although these materials have been used sporadically in the past, biological control agents currently account for
347

342

McCutchen and Flexner

Table 1 Lethal Times of Recombinant AcMNPV Expressing LqhlTP on First lnstar Larvae of Heliothis wirescens Under Early (IEI) and Very Late (~10) Viral Promoters

vnus Wild-type AcMNPV (WT) AcLqhIT2 (p 10) AcLqhIT2 (IEl)

LT,n’ (h) 91 0” 65 0” 44 4”

% Reduction LT from WT b 28 5 51.2

OLTSO lethal time for 50% of the infected population h Percent difference compared to insects killed by wild-type virus c d e Times with different letters m the same column are statlstlcally different

cant msectlclde market share prlmarlly because of their mstablllty m the field and their relatively slow time to kill (13,14). Recently, the Nucleopolyhedrovlruses, such as Autographa calzfornzca (AcMNPV) from the family Baculovlndae, have been genetically modified for an increased klllmg speed. The most promising of these genetically modified vu-uses encode and express insect-selective toxins (25-19) and have resulted m an approximate 20- 40% reduction m the klllmg time of insect hosts. Further improvements have resulted in a >50% reduction m time to kill (Table 1) by utlllzmg different toxins and baculovuus promoters (24,20,21) Although recombinant Nucleopolyhedrovuuses are now considered to have increased msectlcldal properties and are thus more competltlve with the classlcal msectrcides, these modified viruses will encounter other impediments before they are readily adopted for insect control (22). In addition to overcoming formulation, registration, and production obstacles, these new recombinant baculoviruses must still effectively compete with classical insecticides to gain a share of the market. This chapter will investigate one strategy for the use of natural or recomblnant baculoviruses in integrated pest management systems. We believe that baculoviruses and especially recombinant baculovlruses modified for increased insecticidal properties hold great promise for the control of selected insect pests m major cropping systems. A successful launch of these new products ~111 likely come m combmatlon with the use of existing insect control strategies. Specifically, recombinant vu-uses will be combined with low rates of classical insecticides to potentlate efficacy. This interaction may be dose- or tlmedependent. In other words, the interactlon could synergize the dose required to kill ( lethal dose = LD,) or the time required to kill (lethal time = LT,). We will review the reports of several mvestlgators who have documented both positive and negative Interactions between classical insecticides and baculovlruses.

Joint Actions

of Baculoviruses

343

Earlier studies showed potentiation between wild-type viruses and certain msectrcides (23-28) More recently, laboratory studies evaluated several classical msecticides m combination with a recombmant and wild-type AcMNPV (29-31). These studies document potentiation between an insect-selective neurotoxm (AaIT) expressed by AcMNPV with representative compounds from different classes of msecticides. In the final section of this chapter we will briefly discuss the statistical tools best suited for evaluating the mteracttons between microbial and synthetic insecticides. 2. Interactions with Wild-Type Baculoviruses Many studies over the past four decades have investigated mteracttons between wild-type baculovnuses and synthetic msectrcrdes. These early studies were summarized by Benz m 1971 (24). Many of these early studies did not show any clear patterns m terms of positrve or negative mteractions when baculoviruses and chemicals were combmed. Interactions ranged from “synergistic” to “antagomstic” based on the classifications described by Benz (24). Most of these early studies mvolved organochlorine and organophosphate compounds. More recently, results with photostable pyrethroids have shown positive interactions with several different baculoviruses. Aspirot et al. (32) document potentratton between the nucleopolyhedrovtrus from Spodoptera lzttorah (SpllNPV) and Mamestra brasszcae (MbMNPV) and a pyrethroid (deltamethrm) on several economically important Leprdopterous species but did not find potentiation when they combined the nucleoployhedroviruses of H. zea (HzSNPV), L. dzspar (LdMNPV), or Helicoverpa armzgera (HaSNPV) with the same pyrethroid. This study also found that MbMNPV was potentiated by deltamethrm when tested on Spodoptera frugzperda, S exlgua, and H. vwescens, but showed no potenttation when tested on H. arrmgera, Spodoptera littorahs, Ostrtnla nubialu, and L dispar. A similar study by Blanche (27) showed that combinations of MbMNPV with a different pyrethroid (fenvalerate) exhibited potentiation on another economically important pest, the diamondback moth (Plutellu xylostella). Peters and Coaker (33) working with another photostable pyrethroid (permethrm) found a stgmticant reduction m the LC& of the granulovirus of Pzerzs brassicae (PbGV) when combined with a low concentratton of permethrin. Although the results with pyrethroids appear fairly consistent (generally showing positive mteractions), combmattons of vu-uses with other classes of chemicals (i.e., organochlormes, organophosphates, and carbamates) have been more variable. Kompolpnh and Ramakrishnan (34) looked at the Joint action of five rates of DDT, lmdane, malathion, and pyrethrin with a single rate of the nucleopolyhedrovtrus of Spodoptera litura (SpltNPV). Only pyrethrin showed potentiation at the lowest rate tested (i.e., 5 ppm); all other mixtures were either

344

McCutchen and Flexner

independent or addmve. Chaudhari and Ramakrtshnan (35) tested I3 different insecticides at three or four rates m combmation with SpltNPV and found that the mteracttons with the chlorinated hydrocarbons, carbamates, and pyrethrum were generally either additive or synergistic. Interactions with the organophosphates were antagonistic with the exception of drazmon, which showed synergism at the lower doses. Both of these studies showed that the type of mteractron between the vnus and msectrcrdeswas very sensitive to the insectrctdal dose. In fact, another study by Mathat et al. (36) combmmg five different msectictdes with the Nucleopolyhedrovn-us of Spodoptera mauntza (SpmaNPV) found that for some compounds (i.e., phosphamidon and phosalone) mteractions were synergistic at low doses and antagonistic at high doses of the msecttctde. However, when the SpmaNPV was combined with femtrothron, mteractrons were antagomstic at the low and medium doses and synergistic at high doses. In crrttcally evaluating these studies tt becomes dtfficult to determine if these dose-dependent changes m mteractions are real or simply a result of the ltmitatrons of the bioassay and/or the quantitative measures used to characterize the mteractions. Quantifying synergrsm between microbial and chemical msecttcrdes will be discussed later in the chapter. Temporal synergism (mteracttons affecting the LT,) involving Joint actions between wild-type baculoviruses and insectrcides has also been discussed m the literature. Chaudhari (37) evaluated the effect on LT to Diacrisza oblzqua by combming the nucleopolyhedrovrrus of D obliqua with four doses of four different msecttcrdes. He found that both DDT and pyrethrin potenttated the LT,s at hrgh rates and were additive at the low rates. Carbaryl was additive at all rates and femtrothion was antagonistic at all but the lowest rate (5 ppm), which was additive. Shapiro et al. (38) found that the combination of Neem extract plus the nucleopolyhedrovirus of Lymantrza dzspar (LdMNPV) srgmficantly reduced the LTsOfor the gypsy moth (P > 0.05) compared to the LdMNPV-plus-water treatment. McCutchen et al. (31) investigated temporal synergism with SIX synthetic msecttctdes in combination with the nucleopolyhedrovtrus of Autographa calzfornica (AcMNPV) on larvae of Heliothn vlrescens. Only one of the stx msectrctdes (methomyl) significantly decreased (P < 0.05) the median time to death when combmed with the wild-type AcMNPV. A physiological explanatton for at least some of these positive interactions may be found in the work by Shreesam et al. (39). Hrstopathology was performed on Mythzmna separata when the nucleopolyhedrovn-us of M separata was combined with either endosulfan or ferntrothron. They found that when the virus was combined with the msectrcrdes there was severe damage to the gut eptthelmm. The epithelmm was greatly disorganized and fragmented compared to insects infected with vnus alone.

Joint Actions

of Baculoviruses

345

Studies of virus-insecticide combinations have also been conducted m the field. Luttrell et al. (40) found that reduced rates of permethrin or methomyl plus the recommended rate of HzSNPV targeted at Heliothis spp. produced cotton yields no different than the HzSNPV. However, reduced rates of methyl parathion plus HzSNPV were antagonistic and produced yields less than the HzSNPV alone. Combinations between baculoviruses and insecticides may be desirable even in the absence of true synergistic effects. Jaques (42) conducted field trials with AcMNPV plus PrGV, the granulovirus of the imported cabbageworm (Pieris rupae), and a low rate of permethrm. He found that crop protection against the cabbage looper and Imported cabbageworm on cabbage was similar to protection from the full rate of permethrin or methomyl (although the corresponding low rates of insecticide alone where not used to directly confirm dosage potentiation). This combmation might suggest a feasible procedure for reducing the amount of chemical insecticide without sacrificing crop protection. Finally, it is worth notmg that several authors have described potentiatton of the LTs and LDs of several different baculoviruses by the addition of noninsecticidal fluorescent brighteners. The fluorescent brighteners appear to have been first combined with baculoviruses over a decade ago by Martignoni and Iwai (42). They combmed Tinopal DCS with the baculovirus of the Douglasfir tussock moth (Orgyla pseudosugata) and found that the virus appeared to persist longer in the field. They made the assumption that this was a result of enhanced UV protection. In 1986 Chundurwar et al. tested several adJuvants m combination with the nucleopolyhedrovirus of Spzlosoma obliqua on Sunflower (43). They found that a combination of egg albumen, soybean flour, sugarcane Jaggary, and the stilbene oxide Tinopal when combmed with the virus of S. obliqua gave significantly better control than the virus alone. Although they tested several combinations of adjuvants, they did not test any of these ingredients in single combmations with the virus or the effect of the adjuvants alone on S. obliqua. In 1990 Shapiro et al. disclosed that the activity of several different types of insect viruses (i.e., baculovnuses, entomopox viruses, and cytoplasmic polyhedrosis viruses) may be significantly enhanced when combined with one of several chemicals known to be fluorescent brighteners. These mixtures could increase activity up to almost 2000-fold depending on the virus-brightener combmation and the insect that was being tested (44). However, the brighteners had no activity when tested on the insects alone. Smce this disclosure this mteraction has been well documented m the laboratory and the field with a variety of baculovnuses and brighteners (45-48). Very recently, Washburn et al. have described a possible mode of action for these compounds when combined with viruses in the insect gut (49). They suggest that these compounds may block sloughing of infected primary target cells in

346

McCutchen and Flexner

the insect midgut, thus countermg the insect’s developmental resistance and stgrnficantly mcreasmg mortality. Currently, several laboratortes and mdustries are mvesttgatmg the potential of fluorescent brighteners to enhance baculovtrus products. 3. Interactions with Genetically Modified Baculoviruses To date recombinant viruses have not been utilized as insect control agents m any integrated pest management scheme. However, DuPont (50) and Amertcan Cyanamid (51,52) have recently tested recombinant NPVs m selected cotton and vegetable growing areas of the United States. In anticipation of the wide-scale use of recombinant NPVs for the control of Leptdopterous pests, effective control strategies are being formulated and tested to ensure the successand acceptance of this new technology by producers and consultants. Since the baculovtruses are very amenable to current control strategies and appltcatton technologtes, it 1slikely the recombinant viruses will be utilized m conJunction with both classical (e.g., chemicals) and emerging insect control technologtes (e.g., Insect-resistant, transgemc plants). Recently, it has been suggestedthat a strategy of combmmg low rates of classlcal synthetic msecttcideswith recombmant baculoviruses may provide increased effectiveness for control of insect pestsin the field (29,31,53). The results of laboratory and field tests show an unexpected benefit m the form of synergistic or additive responses among several classical msectrcides m combmatron with a genettcally modified AcMNPV expressing an insect-selective neurotoxin. In a laboratory experiment (3Z), six classical insecticides representing several modes of actton were tested alone and m combmation with wild-type AcMNPV or recombinant AcMNPVs Neonate tobacco budworm, Heliothzs vzrescens(F.), larvae were exposed to films of chemicals on the inner surfaces of 20-mL glass liquid scintillation vials (54-56) and >LCggs of recombinant or wild-type virus were added simultaneously using a diet plug. Interactions of the NPVs and mdtvidual msecticides were analyzed and determined. Estimates of LTsas for insects treated with the mixture of NPV and a classical msecticide were faster than those with the NPV alone. However, the decrease m LT,o was significant only in btoassays where classical msectictdes were combined with the recombinant vnus, AcAaIT, constructed m a previous study (ZS). In fact, the LTSOfor all combmattons of AcAaIT and an msecttctde were stgmficantly faster than the vnus alone The recombinant vn-us 1sgenetically modified to express the insect-selective neurotoxm, AaIT, which was isolated from the scorpion Androctonus australzs (57). To date, AaIT is the most thoroughly characterized protemaceous insect toxin and IS known to be a sodtum channel agonist eliciting effects in the insect nervous system stmtlar to that of the pyrethrotd insecticides (58).

Jomt ActIons

of Baculovmses

347

McCutchen et al. (31) observed synergism between cypermethrm (a type II pyrethroid) or methomyl (a carbamate) and AcAaIT. Both cypermethrm and methomyl m combmation with AcAaIT killed insects quicker than would be expected assuming independent or similar action additive models. AcAaIT in combmatton with the msecticides allethrm, DDT, endosulfan, or profenofos also showed a positive interaction, and these data are consistent with either the independent or the similar action additive models. Although these msecticide combinations were not synergistic by definition, they did result in significantly quicker rates of kill compared to the virus alone. Response curves for the combinations of wild-type AcMNPV and subsequent msectictdes were marginally higher than the expected responses assuming the hypotheses of independent or similar additive action (31). Another recombinant vn-us, AcJHE.KK, was also combined with low rates of cypermethrm and screened for interactions; however, no evidence of synergism or antagonism was detected with the combmatton. AcJHE.KK expresses a modified version of juvenile hormone esterase, an insect-derived enzyme important in the regulatory development of many Lepidopterous insects. The modified JHE has been shown to be insecticidal to several Lepidopterous msects (59,60). Although the mode of action of JHE.KK is not yet fully understood, we can assume it does not interfere with the nervous system of insects. Since AcJHE.KK has been shown to kill insects at a rate comparable to that of AcAaIT, these data indicate that the pyrethroids do not enhance speed of kill mdiscrimmately with any recombinant baculovn-us. Just recently another study (30) has corroborated many of the findings disclosed above. In this study two recombinant viruses were utilized, including a recombinant AcMNPV expressing AaIT and an AcMNPV contammg a deletion in the gene encoding ecdysteroid UDP-glucosyl transferase (EGT). O’Reilly and Miller (61-63) have thoroughly characterized an AcMNPV with a deleted EGT gene. This recombinant vnus, AcMNPV EGT-, mcreases the speed of kill of host insects by disrupting the natural tendencies of the wildtype AcMNPV vnus to slow or retard larval molting or pupation, which extends the duration of viral mfectton Thus, as is the casewith the recombinant viruses expressing AaIT, the modified AcMNPV EGT- significantly enhances the rate of kill. In the study by Black et al. (30) the second instar larvae of H vzrescens and Helicoverpa zea were assayed in a diet overlay method whereby vtrus, insectttide, or virus and insecticide combinations were tested for activity. Several chemical and vtrus combmations were studred and found to be synergistic or additive m nature. For example, H. zea larvae were tested using a formamtdme, an arylpyrrole, diacylhydrazme, synthetic pyrethroid, and benzoylphenylurea m combmation with the two recombinant vn-uses or wild-type AcNPV Syner-

348

McCutchen and Flexner

gism was observed with the recombinant AaIT virus m combination with the formamidme, arylpyrrole, diacylhydrazine, and the synthetic pyrethroid (cypermethrm), but was not observed with the insect growth regulator, benzoylphenylurea. In contrast, only diacylhydrazine in combination with the recombinant virus AcMNPV EGT- significantly hastened the speed of kill, whereas the other msecticides did not significantly increase the bioactivity of this vn-us. In the same study, larvae of H. vzrescens were treated with a combination of the pyrethroid and AcMNPV AaIT, AcMNPV EGT-, or wild-type AcMNPV. Both recombmant vu-us and pyrethroid mixtures produced synergism and were superior to the combmation of cypermethrin and wild-type AcMNPV, which did not generate synergism, as was also previously shown by Aspnot et al. (33) and McCutchen et al. (31). Recent field studies by Treaty (53) would indicate combmations of certain classical msecticides, and the recombmant virus AcMNPV AaIT does result in a positive interaction. In cotton field studies carried out m Louisiana and Georgia, both the combination of AcMNPV AaIT and an arylpyrrole or cypermethrin resulted in a significant decrease in fruit damage when compared to either the virus or chemical alone at the same rates. Decreases m lethal times from a combmation of an NPV and an msecticide could occur based on a number of different scenarios. Enhancement of speed of kill could be expected if the insecticides and the NPV were acting independently (independent mode of action model) or if the two compounds were additive m dose (similar action additive model). These two models would result m a decrease in lethal times even though they assumeno synergisms or antagonisms between two compounds. Another explanation for the enhanced speed of kill might result from a positive interaction between the insecticide and the NPV because of interaction at the same target site and/or tissue. In addition to possible interactions at the tissue and/or target site, other physiological interactions could provide for the enhanced effect, such as the NPV infection facilitatmg absorption or distribution of insecticide, the NPV infection suppressing detoxication mechanisms, and viral mfection of nerve cells resulting m increased sensitivity of nervous tissueto msecticides.Certainly, synergism could also result from a combmatron of one or more of these possible scenarios. These data suggest that regardless of the target site of the pesticide, simultaneous action of the peptide toxin, the insecticide, and/or viral mfection can result m positive mteractions, some of which may be synergistic and some of which may be additive. It IS evident that the efficacy of all recombinant viruses will not be enhanced indiscriminately by combining them with an Insecticide, as suggested by the combmations of cypermethrm with the recombmant virus AcJHE.KK (31), AcMNPV AaIT with the insect growth regulator benzylphenylurea (30), and AcMNPV EGT- with several classical insecticides (30)

Joint Actions

of Baculoviruses

349

Based on these two most recent studies, predictmg the interaction cannot be based simply on our knowledge of the insecticide, but must also be based on our knowledge of the pepttde expressed by the recombinant virus. In fact, there is more precedence for the phenomenon of positive cooperativity as demonstrated by Herrmann et al. (64). In this study it was demonstrated that the insecticidal activity of neuorotoxic peptides was increased 5 to 1O-fold when combinations of toxins were injected into insects. It is difficult to account for the augmentation made by the virus to the overall reduction in the time to kill host insects. It should be noted that NPVs can infect insect nerve cells, potentially perturbing the peripheral nervous system, and thus leading to neuromuscular abnormalities. Most likely the combined action of the virus, neurotoxm, and insecticide would enhance the overall insecticidal activity. Simply stated, prior to the wide-scale release of combinations of recombinant insect viruses and classical chemistry, thorough studies should be conducted to aquire a better understanding of this control strategy. 4. Statistical Methods for Evaluating Chemical and Microbial Interactions Synergism of microbial insecticides with other insecticides has been reviewed by Benz (24) and Harper (65). The authors recommend that anyone interested in this topic should read these articles because they give clear definitions with examples of the different types of interactions that can occur between microorganisms and synthetic insecticides. It is also important to note that some statistical analysis designed for mteractions based on mixtures of xenobiotics may not be appropriate for mixtures of microbial insecticides and synthetic chemistry. However, the statisticsdescribed m these articles for quantification of the various mteracttons are somewhat outdated. For an updated general review of testing insecticide mixtures we recommend Chapter 8 in Robertson and Preisler (66), Green and Streibig (67), and Laska et al. (68). An example of analysis using the conditional likelihood function and testing two hypotheses (i.e., the independent action model and the similar action additive interaction model) to measure mteractions between baculoviruses and synthetic insecticides may be found in McCutchen et al. (31). 5. Commercial Potential for Virus Chemical Interactions Slow rate of kill and resulting crop damage is often cited as a major limitation to the commercial successof viral insecticides, because 24-48 h mortality IS a common target for classical insecticides. Given the emphasis on developmg agents that kill target pests more quickly, many of the strategies discussed above concentrated on studying speed of kill. However, lethal doses or effective concentrations of viral and chemical combmations will be of major eco-

McCutchen and Flexner

350

SeNPV+

Pretreatment

Methomyl

7 Day Days After First Appl!catlon

Fig. 1. Efficacy of methomyl and SeMNPV alone and in combination on carbamate-resistantSpodoptera exigua in Thailand, February 1992. nomic and environmental importance as well. In practice, a variety of factors would influence the rates of insecticidal agent(s) used, including the cost and availability of the active ingredient, environmental factors, potential or perceived levels of resistant pests in the field, as well as other pest management considerations. Another potential benefit of baculovirus and insecticide mixtures could come in the form of new resistance management strategies, as evidenced by two recent studies (Flexner and Marsden unpublished results; 31). In the laboratory study by McCutchen et al. (31), it was clearly demonstrated that a pyrethroid-resistant strain of H virescens was significantly more sensitive to the recombinant virus, AcAaIT, compared with a susceptible strain. In field studies in Thailand, Flexner and Marsden found that methomyl-resistant Spodoptera exigua on grapes was controlled significantly quicker by a combination of SeMNPV plus methomyl than either the virus or methomyl treatment alone (Fig. 1). In fact, the methomyl alone treatment gave no control and larval populations were actually higher in these plots than in the untreated control plots. As new wild-type and recombinant viruses are identified, developed, and commercialized, it is important to evaluate them in combination with classical pesticides as well as transgenic crops, especially insect-resistant varieties. Such

Joint Actions

of Baculoviruses

351

studies should uncover potentially useful interactions as well as circumvent potential failures or pitfalls of the virus as well References 1, Bird, F T. and Whalen, M M. (1953) A virus disease of the European p ne sawfly, Neodtprion serlfer (Geoffr ) Can Entomol 85,433-43’7 2. Miller, L. K., Lingg, A J., and Bulla, L A. (1983) Bacterial, viral and fungal insecticides. Science 219, 7 15-72 1. 3. Carbonell, L. F and Miller, L. K. (1987) Baculovnus mteraction with non target orgamsms a virus-borne reporter gene is not expressed m two mammalian cell lines Appl Environ Mtcrobtol 53, 1412-1417 4. Summers, M D., Engler, L , Falcon, A, and Vail, P (1975) Baculovtruses for Insect Control Safety Constderatrons American Society for Microbiolog),, Washington, DC, pp 186. 5. Flexner, J L , Lighthart, B , and Croft, B. A (1986) The effects of n icrobial pesticides on non-target beneficial arthropods Agrzcult Ecosystems Environ 16, 203-254. 6. Bedford, G 0 (1981) Mzcrobzal Control ofPests and Plant Diseases (Burges, H D , ed ), Academic, New York, pp. 409-426 7 Entwistle, P F and Evans, H F (1985) Comprehenstve Insect Phystology, Btochemistry and Pharmacology, vol. 12 (Kerku, G A and Giblert, L 1 , eds ), Permagon, Oxford, UK, pp 347-412 8. Moscardi, F (1988) Btotechnology, Btologtcal Pesttctdes and Novel Phznt-Pest Resistance for Insect Pest Management (Roberts, D W and Granados, R R , eds ), Cornell Umversity Press, Ithaca, NY, pp. 5360 9 CuiJpers, T A M M , Steeghs, N W F , Dimock, M. B , and Smits, P H (1994) Spod-XTM GH, recently registered m the Netherlands as a viral msecticide for control of beet armyworm Spodoptera exrgua m cutflowers and potted plants m greenhouses Med Fat Landbouww Umv Gent 59,393-403 10 Kolodny-Hirsch, D. M., Warkentm, D L., Alvarado-Rodrigues, B., and Knkland, R (1993) Spodoptera extgua nuclear polyhedrosis virus as a candidc’te viral msecticide for the beet armyworm (Lepidoptera. Noctuidae) J Econ E’ntomol 86,314-321. 11. Cunningham, J C (1982) Mtcrobral and Viral Pestzczdes (Kurstack, E., ed ), pp 335-386 12. Jutsum, A. R. (1988) Commercial apphcation of biological control status and prospects Phil. Trans Roy Sot Lond Ser. B 318,357-373. 13 Yearian, W C and Young, S Y (1982) Microbial and Vtral Pesttctdes (Kurstack, E , ed ), pp. 335-386. 14 Black, B C , Brennan, L. A , Dierks, P M , and Gard, I E (1997) The Baculovrruses (Miller, L. K., ed.), Plenum, New York, pp 341-387. 15. Maeda, S , Volrath, S L., Hanzlik,T. N., Harper, S A , Maddox, D W , Hammock, B D , and Fowler, E. (1991) Insecticidal effects of an insect-specific neurotoxm expressed by a recombmant baculovirus. Vtrol 184,777-780

McCutchen and Nexner

352

16. McCutchen, B F , Choudary, P. V., Crenshaw, R., Maddox, D Kamita, S. G , Palekar, N., Volrath, S , Fowler, E., Hammock, B. D., and Maeda, S. (1991) Development of a recombinant baculovirus expressing an insect-selective neurotoxin. Potential for pest control BzoTechnology 9, 848-852. 17. Stewart, L M. D , Hurst, M., Ferber, M. L , Merryweather, A. T., Cayley, P. J., and Possee, R. D. (1991) Construction of an improved baculovnus msecttcide containing an insect-specific toxin gene. Nature 352, 85-88. 18 Tomalski, M D and Miller, L. K. (1991) Insect paralysis by baculovirus-mediated expression of a mite neurotoxm gene. Nature 352, 82-85 19. Tomalskt, M. D. and Miller, L K (1992) Expression of a paralytic neurotoxm gene to improve insect baculovnuses as biopesticides. Bzotechnology 10, 545-549 20. Lu, A., Seshagiri, S., and Miller, L. K. (1996) Signal sequences and promoter effects on the efficacy of toxin expressing baculovuuses as biopesttcides Bzol Control 7,32&332

21, Presnail, J., Flexner, J L , Jarvis, D L , Davis, D , Greenamoyer, C , Wilhams, C., and McCutchen, B F (1997) Comparison of four modified Autographa calzfornzca baculovuuses expressing insect specific toxins from the scorpions Lezurus quznquestrzatus hebraeus and Androctonus australzs under early and late baculovuus promoters. Nature Bzotech , submitted 22. McCutchen, B F. and Hammock, B D (1994) Natural and Derzved Pest Management Agents (Hedm, P A., Menn, J. J., and Hollmgworth, R. M., eds.), ACS Symposium Series 5 1, American Chemical Society, Washmgton, DC, pp 348-367 23. Ignoffo, C M and Montoya, E L (1966) The effects of chemical msectrcides and insecticidal adJuvants of a Helzothzs nuclear polyhedrosis wus J Inv Path 8,409-412. 24 Benz, G (1971) Mzcrobial Control ofInsects and Mztes (Burges, H D and Hussey, N W , eds.), Academic, London, UK, pp. 327-355 25 Jaques, R. P. (1985) Vzral Insectzczdes for Bzologzcal Control (Maramarosch, K and Sherman, K E., eds.), Academic, New York, pp. 285-360 26. Bianche, G. and Guillon, M. (1990) Process for combating vineyard cohylis and Eudemia, utihzmg a nuclear polyhedrosis baculovu-us. European Patent Apphcation 0,392,928. 27 Blanche, G (1991) Method for the biological fight against the crop ravaging insect, Plutella xylostella, using a nuclear polyhedrose and at least one synthetic pyrethrmoid United States Patent No 5,075,lll 28. Flexner, J. L. and Marsden, D. A. (1995) International Patent Application No wo 95/0574 1 29. Hammock, B D and McCutchen, B F (1996) Insect control method with genetically engineered biopesticides International Patent Apphcation No WO 96/O 1055 30. Black, B. C., Kukel, C. F., and Treaty, M. F. (1996) Mixtures of genetically modified insect viruses with chemical and biological msecticides for enhanced insect control International Patent Apphcatton No WO 96/03048 31 McCutchen, B F , Hoover, K., Preisler, K , Betana, M D , Herrmann, R., Robertson, J L , and Hammock, B D (1997) Interactions of recombinant and

Joint Act/ens of Baculoviruses

353

wild-type baculovlruses with classtcal msectlcldes and pyrethrold-reslstant tobacco budworm (Lepldoptera: Noctuldae). f. Econ Entomol 90, 1170-l 180 32 Aspirot, J., Blache, G , Delattre, R., and Ferron, P (1987) Process for the blological control of insects which destroy crops, and insecticidal composltlons. United States Patent No. 4,668,s Il. 33 Peters, S E. 0 and Coaker, T. H. (1993) The enhancement of Pzens brusszcae (L.) granulosls virus mfectlon by mlcroblal and synthetic insectlcldes. J Appl. Ent 116,72-79.

34. Komolpith, U and Ramakrishnan, K. (1978) Joint action of Spodoptera lltura (Fabriclus) and msectlcldes J Ent. Res 2, 15-19 35 Chaudhari, S and Ramakrishnan, K. (1983) Effect of msecttcldes on the actlvlty of nuclear polyhedrosls virus of Spodoptera lrtura (Fabnclus) m laboratory bloassay tests J Ent. Res 7, 173-179. 36. Mathai, S., Nan, M R G. K , and Mohanas, N (1986) Jomt action of nuclear polyhedrosls virus (NPV) and msectlcides against the rice swarming caterplllar Spodoptera maurltla (Bolsduval). Indzan .I Plant Prot 14, 13-l 7 37. Chaudhan, S. (1987) Combmatlon effect of baculovirus of Dzacrlsw obllqua Walker with low doses of msectlcldes and fertilizers on larval virosis Indzan J Ent. 49,5 15-5 19 38 Shapiro, M , Robertson, J L., and Webb, R E (1994) Effect of neem seed extract upon the gypsy moth (Lepldoptera. Lymantriidae) and Its nuclear polyhedrosls vwus J Econ Entomol

87,356-360.

39. Shreesam, S , Mathad, S. B , and Savanurmath, C. J. (1986) Hlstopathology of the armyworm Mythzma (Pseudaleha) Separata (Walker) treated with combinations of nuclear polyhedrosls vnus and some chermcal insecticides. J Karnatak Umv Scz 23, 176-l 80 40. Luttrell, R. G., Yeanan, W. C., and Young, S. Y. (1979) Laboratory and field studies on the efficacy of selected chemical insecticide-Elcar (Baculowrus helcothis) combmations against HellothIs spp J. Econ Entomol 72, 57-60. 4 1 Jaques, R P (1988) Field tests on the control of the imported cabbageworm (Lepldoptera. Pleridae) and the cabbage looper (Lepldoptera. Noctuidae) by mixtures of microbial and chemical insecticides. Can Ent. 120,575-580 42. Martlgnoni, M. E. and Iwal, P. J (1985) Laboratory evaluation of new ultraviolet absorbers for protection of Douglas-fir tussock moth (Lepidoptera*Lymantrlldae) baculovnus. J Econ Entomol 78,982-987. 43. 43 Chundurwar, R. D , Pawar, V. M., et al. (199 1) Bioassay and efficacy of some adJuvants for NPV of Spilosoma obliqua (walker) on Sunflower. J Maharashtra Agnc. Untv 16,369-371

44. Shapiro, M., Dougherty, E., et al. (1990) Compositions and methods for blocontrol using fluorescent brighteners. United States Patent No 5,124,149 45. Shapiro, M. and Robertson, J L. (1992) Enhancement of gypsy moth (Lepldoptera, Lymantrtidae) baculovirus activity by optical brighteners. J Econ Entomol 85,1120-l 124 46. Shapiro, M and Vaughn, J. L. (1995) Enhancement in activity of homologous and heterologous baculoviruses infectious to cotton bollworm (Lepldoptera. Noctuidae) by an optical brightener J Econ. Entomol. 88,265-269.

McCutchen and Flexner

354

47 Dougherty, E. M., Guthne, K , et al (1995) In vitro effects of fluorescent brtghtener on the efficacy of occlusion body dissolution and polyhedral-derived vmons Btol Control 5383-388.

48. Argauer, R. and Shaptro, M (1997) Fluorescence and relative activmes of stilbene optical brighteners as enhancers for the gypsy moth (Lepldoptera Lymantrudae) baculovnus J. Econ Entomol 90,416420. 49 Washburn, J O., Kirkpatrick, B A , et al (1998) Evidence that stilbene-dertved optical brightener, M2R, enhances Autographa caltforntca M Nucleopolyhedrovnus mfection of Trtchoplusra no and Helrothts vtrescens by preventing sloughing of infected mtdgut epithehal cells Biol. Control, m press 50 DuPont (1996) Notification to Conduct Small-Scale Field Testing of a Genetically Altered Baculovirus, EPA No 352-NMP-4 5 1 Amertcan Cyanamid (1994) Notification to Conduct Small-Scale Field Testing of a Genetically Altered Baculovuus, EPA No 241-NMP-2, EPA-OPP Publtc Docket No 50799. 52 American Cyanamid (1996) Notification to Conduct Small-Scale Field Testmg of a Genetically Altered Baculovlrus, EPA No 241-NMP-3, EPA-OPP Public Docket No 508 16 53. Treaty, M (1997) Efficacy and non-target safety of an AaIT gene-Inserted baculovnus IBC Blopesttctdes and Transgentc Plants 2”d Annual Conference, Washington, DC 54. Plapp, F W , Jr and Campanhola, C. (1986) Proceedtngs, Beltwtde Cotton Productzon Research Conferences, Dallas National Cotton Council of America, Memphts, TN, pp. 167-169 55 Campanhola, C and Plapp, F W , Jr. (1989) Toxicity and synergism of msectitides against susceptible and pyrethroid resistant neonate larvae and adults of the tobacco budworm (Lepidoptera. Noctuidae). J. Econ Entomol 82, 1495-1501 56 McCutchen, B F., Plapp, F. W , Jr., Nemec, S J , and Campahnola, C. (1989) Development of diagnostic larval pyrethroid resistance momtormg techniques for Heltothts spp m cotton J Econ Entomol. 82, 1502-1507. 57 Zlotkin, E , Rochat, H , Kopyean, C , Miranda, F., and Lissitzky, S (197 1) Puntication and properties of the msect toxin from the venom of the scorpion Androctonus australts Hector. Btochtmte 53, 1073-1078 58 Zlotkm, E (1986) Neuropharmacology and Pesttctde A&on (Ford, M G , Lunt, G G , Reay, R C., and Usherwood, P N R , eds ), Ellis Hot-wood, Chichester, UK, pp 352-383 59. Bonnmg, B and Hammock, B. D (1994) Natural and Derived Pest Management Agents (Hedm, P A, Menn, J J., and Hollmgworth, R M , eds ), ACS Symposium Series 5 1, American Chemical Society, Washington, DC, pp 369-383 60. Bonnmg, B C., Hoover, K., Booth, T F., Duffey, S S., and Hammock, B. D. (1995) Development of a recombinant baculovnus expressing a modified Juvemle hormone esterase with potential for insect control. Arch Insect Btochem. Phystol 30, 177-l 94. 6 1. O’Reilly, D. R. and Miller, L. K (1989). A baculovn-us blocks molting by producing ecdysterotd UDP-glucosyltransferase Sczence 245, 1110-l 112.

Joint Actions of Baculoviruses

355

62 O’Reilly, D. R and Miller, L. K. (1990) Regulation of expression of a baculovuus ecdysterotd UDP-glucosyltransferase gene. Virology 64, 132 1-1328 63. O’Retlly, D. R. and Mdler, L. K (199 1) Improvement of a baculovnus pesttcide by deletion of the EGT gene. B~technology 9, 1086-1089. 64 Herrmann, R , Moskowitz, H., Zlotkm, E , and Hammock, B D (1995) Positive cooperatlvity among msecticidal scorpion neurotoxins Toxlcon 33(8), 1099-l 102 65. Harper, J D. (1990) Microbial pesticides- synergism and integration with other pesticides, m Vth Internatronal Colloquium on Invertebrate Pathology and Mzcrobzal Control, SIP Publrcation ISBN # 0 646 00549 9, pp. 482-485 66 Robertson, J L and Preisler, H. K (1992) Pestzcrde Bioassay8 wzth Arthropods. CRC, Boca Raton, FL, 127 pp 67 Green, J M. and Streibig, J. C. (1993) Herblclde Bloassays (Streibig, J C and Kudsk, P , eds.), CRC, Boca Raton, FL, pp 118-l 34 68 Laska, M , Morris, M , and Siegel, C. (1994) Simple designs and model-free tests for synergy Blometrx 50,834841.

IV BIOHERBICIDES

19 Mycoherbicides Alice L. Pilgeram and David C. Sands 1. Status of Microbial Weed Control The potential of fungal pathogens to control unwanted plant species has often been underestimated, primarily because the impact of a pathogen on its host plant withm a given region is usually subtle, and only the final populatton equihbrmm is observed (1). Moderate levels of pathogemcny, m combination with host tolerance or resistance, facilitate survival of both the pathogen and the host. As a rule, few pathogens risk completely decimating their host population, and therefore eliminating their ecological niche. However, potato late blight, Dutch elm disease, chestnut blight, and coffee rust are just a few examples of devastating plant disease epidemics that have resulted when this fragile equilibrium has been disrupted. These epidemics all resulted from the accidental introduction of an exotic pathogen mto areas where the host plant had been cultivated m the absence of the pathogen. In the absence of this selective pressure, little resistance or tolerance was present m the host population, favoring development of severe disease in the plant population. Thus, the challenge of biological control is to artificially shift the ecological balance m favor of the pathogen through the introduction of new strains, or by artificially increasing the concentration or virulence of an mdigenous pathogen within a given area. 7.7. History Weeds are a major constraint to the realization of optimum yields, as well as optimum profits, in agricultural systems. In natural ecosystems, exotic plants can displace native plant species and negatively impact the animals dependent on them. During the past four decades, chemical herbicides have dommated the arena of weed control; but these herbtcides, although effective m many From Methods m Botechnology, vol 5 B/opestrodes Use and Del/very Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

359

360

Pilgeram and Sands

instances, may also have deleterious effects on humans, nontarget plants, and the environment (2,3). Nonselective herbicides are active against both desnable and nondesirable plant species. More selective herbrcrdes can control a narrower spectrum of weeds, but, because many field situations require combating several different weed species,application of multiple chemicals is then required. Such herbicide combmations can compound environmental hazards, as well as have mhibitory effects on nontarget plant species.Years of herbicide use have also resulted in the increased Incidence of herbicide-resistant weeds (4). Finally, chemical control is often impractical or noneconomical m wellestablished weed stands in rangeland and forest situations (5). There are many pathogenic organisms, such as insects, fungi, bacteria, nematodes, and viruses, that will kill weeds. Two distinct strategies for biological weed biocontrol, classical and augmentative (or “bioherbicidal”), have been successfully employed. In classical biocontrol systems,pathogens are isolated from regions where both the pathogen and the weed species are Indigenous, and where the pathogen exerts some degree of control on the weed population. These pathogens are then released into areas where the weed has become problematic, in the hope that the pathogen will establish as a sustained population and suppress further expansion of the weed population. Pathogens used in this classical approach are generally rusts and other fungi capable of self-disseminatron by airborne spores, which then Induce additional infections on neighboring, as well as distal, target plants. Since there is no attempt to delimit the spread or replication of the introduced biological control agent, such agents must be proven to be environmentally safe and host-specific (6). This approach works well m areas of low economic return per unit, such as rangeland and timberland, because the same level of inoculum can be used for weed control on 1 ha as on 1000 ha (7). The natural spread of the organism, coupled with its ability to remain in the environment, allows for long-term control of noxious weeds. Biological control of rush skeletonweed exemplifies the potential of the classical approach. Rush skeletonweed (ChondrillaJuncea), native to Eurasia and the Mediterranean, was introduced into Australia m the early 1900s and is widespread throughout wheat and fallow fields m southeastern Australia. Of the various insects and pathogens discovered on the weed m the Mediterranean region, the rust fungus Puccma chondrillzna was found to be the most damagmg (s). In 1971, Australian scientists introduced a strain of P. chondrillzna into the wild, and withm a short time the rust spread over the entire area infested by the weed (9). The subsequent suppression of the weed population resulted in a 79% reduction of wheat yield loss in 1979-1980 (10). Similar successes were found using P. chondrillina against rush skeletonweed in northern California (II), Washington (12), and other western states (13). A major drawback to the overall successof rush skeletonweed eradication was that the rust was

Mycoherbicides

361

highly specific to only one biotype of rush skeletonweed, and did not infect the other two forms of the weed. As the population of the susceptible biotype decreased, the abundance of the resistant biotypes greatly increased (8,14). In augmentative (or bioherbicidal) biological control systems, indigenous populations of a microbial biocontrol agent are greatly supplemented, m order to shift the pathogen-host equilibrium in favor of the pathogen (15). This approach utilizes native pathogens occurring on the weed in the infested area, eliminatmg much of the expense associated with international collectton and testing of pathogens, and decreasing the perceived risk of mtroductton of exotic organisms. Successful augmentative applications of mycoherbicides include control of yellow nutsedge in the United States with Puccznia canalzculata (16,17), and control of northern Jointvetch m rice m Arkansas with Colletotrichum gloeosporioides fsp. aeschynomene (18,19). In contrast to classtcal biological control, bioherbicide agents may not spread rapidly on their own, and may not cause epidemics the following year (20). Unfortunately, because mycoherbicides must be applied directly to the entire area of intended control, they have a constant cost per acre. 7.2. Hunting and Gathering

of Potential Bioherbicides

Numerous plant pathogenic organisms have been isolated from noxious weeds. Once isolated, time and effort has been directed toward pathogenictty screenmg of these pathogens in growth chamber studies. As a result, researchers now possessan extensive library of potential bioherbictdes. However, few pathogens have shown adequate control of the target weeds in field situations, and, as a result, few have emerged as commercially vtable products. For example, Canada thistle (Cirsium awense) is a noxious perennial weed mfesting rangeland, turf, and cropland in the northern plams. Its wind-blown seed and extensive perennial root system make it a difficult weed to control, and, therefore, a likely candidate for biocontrol. During surveys of endemic Canada thistle diseases m Montana, six genera of plant pathogens were collected, including Alternarza, Fusarzum, Septoria, Puccinia, Sclerotznza,and Pseudomonas (21). Two of these pathogens, Puccinia punctzformis and Sclerotinia sclerotiorum, are being investigated aspotential bioherbicides. P. punctzformis, an endemic host-specific rust on Canada thistle, causes severe ettolatton and necrosis of infected tissues, and prevents flowering and, therefore, seed set. Unfortunately, different ecotypes of Canada thistle showed differmg suscepttbilities to the screened isolates of P punctzjbrmis (22). Rust isolates effective against a larger spectrum of Canada thistle ecotypes could provide more complete control of Canada thistle. Many host-specific pathogens, such as the rusts, lack the lethality needed for effective btological control of weeds. Conversely, many lethal pathogens

362

Pigeram and Sands

attack large numbers of plants, often mcludmg crops and beneficial native species. S sclerotzorum is a lethal pathogen of Canada thistle. The potential of this organism for weed control was demonstrated in limited field trials, m which strains of the fungus reduced populattons of Canada thistle and Spotted knapweed by up to 80% (21). A subsequent reduction m weed densities was also observed the followmg year. Bourdbt et al. (23) provided additional evidence that the pathogen moves from diseased thistle stems mto propagated roots, and that, with repeated annual apphcations of the pathogen, shoot density will declme However, S. sclerotiorum does have Its limttations as a potential biocontrol agent. First, It is one of the most nonhost-specific pathogens known, attacking over 400 plant species (24). Second, the majority of pathogens that have been evaluated or developed as mycoherbictdes produce mfective spores. Sclerotznza does not produce comdia, and ascospore production is uneconomical and timeconsuming. Because of these hmitations, the use of Sclerotznia for biocontrol of Canada thistle will require tmplementation of mnovative strategies for containment of the pathogen, as well as development of a commercially acceptable formulation of fungal mycelmm (25). 1.3. Registered Mycoherbicides Although many bioherbicides have shown great promise, only four bioherbicides have been registered m the United States or Canada (2627). In 1981, DeVme@, a formulation of the plant pathogen Phytophtora palmwora (Abbot Laboratortes, Chtcago, IL), was regtstered for control of strangler vine (Morrenza odor&a) m Florida citrus groves (28). CoIlego@, a formulation of C gloeosporzozdes f. sp. aeschynomene (Pharmacta & Upjohn Company, Kalamazoo, MI; Encore Technologies Inc , Mmnetonka, MN), was registered shortly thereafter to control northernJomtvetch m Arkansas (29). BioMal@ (C gloeosporlozdes f sp. malvae) was registered by Phtlom Bias (Saskatoon, Saskatchewan, Canada) for round-laced mallow (M&a puszlla L.) control m Canada and the United States (30,31) and Dr. Biosedge@ (Pucc~y1rucanafzculurtu) (Ttfton Innovation Corporation, Tifton, GA) was registered for control of yellow nutsedge (Cyperus esculentus L.) (17) Two other bioherbicides have recently been registered m other parts of the world. In Japan, Camperico@ (Xunthomonus cupestrzs pv. poae) was registered for control of annual bluegrass by Japan Tobacco Inc. (Yokohama, Japan) (32) and Chondrostereum purpureum (Biochon@) was marketed in the Netherlands by Koppert Biological Systems (Berkel en Rodenrtjs, Netherlands) for suppression of stump sproutmg in black cherry (Prunus serotzna) (33). Numerous impediments have hindered the mycoherbtcide registration process.The case of Collego provides an interesting case study as to why registration of mycoherbicides has been limited (29). Collego was origmally registered

Mycoherbicides

363

by the Upjohn company for the control of northernlomtvetch in Arkansas, and was marketed as a complex formulatton of dried comdta of C. gloeosporioldes f.sp aeschynomene. The formulation was rehydrated m aqueous solutions, and applied aerially to the surface of rice paddies, when stemsof northernlomtvetch begin to exceed the height of the rice. Within 2 wk, lesions form on the stems of the weed, followed raptdly by weed death. Collego occupied a very small market mche, had a limited area of applicability, and provided long-term control of northern Jointvetch, curbing the need for periodic reapplicatton. BtoMal was shelved because of Its small market niche and expensive formulatton technology after only 2 yr. 2. Barriers to Weed Control 2.1. Mythology vs Practice: Efficacy and Host Range In theory, biologtcal control of weeds can replace and/or augment the use of chemical herbicides with environmentally benign plant-pathogenic bacteria and fungt. In practice, however, btologtcal control of weeds has yet to achteve the levels of successneeded for commercial acceptance. Efficacy levels are measured as the ability to reduce a weed’s population to a level below the economic or social threshold at which tt would be considered noxtous (6). Success in the marketplace depends on many other factors besides its ability to destroy the target weed, mcludmg the magnitude of the effective host range, the dlstrtbutton of the target weed, whether the mycoherbiclde can be produced economtcally, and whether the mycoherbtcide ~111 pose any threat to the environment or to nontarget plant species The host specifictty of mycoherbictdes 1sa double-edged sword If the host range of a pathogen is narrow, the potential market will be limtted by the distrtbution of the susceptible weed. However, these hrghly spectfic mycoherbicldes pose mnnmal risk to other plant species. If the host range of the biocontrol agent is broad, a single product can be used to control a battery of unwanted weeds. Consequently, the potential market will be much larger. Unfortunately, wild-type strains of these pathogens also pose substantial risk to desirable plants m the environment. Thus, the choice between broad- and restricted-host range pathogens is very similar to the choice between a broad range herbicide, such as Roundup@ (Monsanto, St. LOUIS,MO), or a more selective herbicide, such as Hoelon@ (AgrEro USA, Wtlmington, DE) for control of wild oats 2.2. Ecological Restraints Most pathogens have exacting envnonmental parameters that need to be met before infection or symptoms of disease can occur. In many Instances, environmental restraints, such as adverse temperatures, sot1pH, and humidity, are responsible for reduced disease incidence and severity (9,15) Furthermore,

364

Pilgeram and Sands

environmental factors are ever-changing, and are difficult to predict or duphcate m growth-chamber studies. However, a pathogen’s ability to overcome environmental constraints will greatly improve its ability to be an effecttve biological control agent under a variety of conditions The search for an effective agent, therefore, requires a better understanding of the complex mteraction between the environment, the pathogen, and the host plant (6). Specific formulations may provide a means to overcome some of these environmental limitations (3435). 3. Enhancement of Mycoherbicides Low virulence of mycoherbicides is commonly overcome by the application of extremely high levels of fungal moculum. Unfortunately, this practice is expensive. Formulation and delivery systems can greatly improve the field performance of a given dose of a mycoherbicide. The effectiveness of a mycoherbicide could also be increased by enhancing the pathogenicity of the biocontrol agent, or by coappltcatton of the mycoherbicide with other pathogenic organisms (3638), or with low levels of chemical herbicides (31,39,40). 3.7. Pathogen Marking Once a mycoherbicide is released into the environment, it is difficult to differentiate tt from similar organisms present m the soil or on the plant. It may also be difficult to confirm that the released mycoherbicide is the causative agent of the observed symptoms on the target weed, rather than some mdigenous pathogen Genetic marking simplifies the reisolation and positive identification of a biocontrol agent. Two types of genetic markers have been used to label mycoherbicides: auxotrophic (25) or nitrate nonutilizing (nit) mutants (41,42); and herbicide-resistance mutants (42). Genetic marking also facilitates assessmentof the risk of gene transfer from biocontrol agents to related fungi. Nutritional mutants were used to assessthe risk of asexual gene exchange m the mycoherbicide C gloeosporiozdes f.sp. aeschynomene (41). Heterokaryosis was demonstrated between paired C gloeosporloldes f.sp. aeschynomene strains m vitro, but conidia recovered from the colonies were composed of only parental phenotypes. In host maculation studies, only parental phenotypes were recovered from co-colonized lesions on northern jointvetch. Thus, asexual gene exchange and mitotic recombmation appear unlikely in C. gioeosporzozdesf.sp. aeschynomene. However, m a subsequent study, sexual outcrossing was detected when stems of northern jointvetch were co-inoculated with an isolate of C. gloeosporioides f.sp. aeschynomene pathogenic to northern jomtvetch and an isolate of C. gloeosporloides f.sp. aeschynomene pathogenic to pecan (43).

Mycoherbicides

365

3.2. Synergy with Herbicides Plant defenses (physlcal/chemlcal barriers and biochemical responses) protect against attack from essentially all microorganisms. Plant pathogens possessinfection mechanisms that allow them to either evade or break down one or more of these defenses. Several chemical herbicides reduce a plant’s ability to prevent pathogen invasron and establishment. The efficacy of a mycoherbicide could be enhanced when applied in combination with the mlcroblal agent (39,4#-46). Such combmatlons could result in reduced apphcatlon levels of both the chemical herbicide and the mycoherbicide required for economical control of noxious weeds. Sicklepod seedlings infected with the mycoherbiclde, Alternarla cassiae, have elevated levels of a phytoalexin toxic to fungi (45). Sublethal concentrations of glyphosate inhibit phytoalexm production, and, therefore, render the sicklepod plants more susceptible to Alternaria infection Thus the herbicide acts synergistically with the pathogen to kill the target plant. Although several herbicides have been shown to predispose plants to fungal invasion, this synergistic effect is negated if the herbicide also compromises the pathogen (30,31,42,47,48). In such cases,herbicide resistance m the pathogen would permit coapphcation of the mycoherbiclde and the chemical For example, blalaphos 1sa nonselectlve, broad-spectrum herbicide produced by Streptomyces hygroscopicus (49). Blalaphos-reslstant strams of C gloeosporzoides f.sp. aeschynomene were obtained by transformation with the bar gene (phosphinothricm acetyltransferase) (46,50). Coapplication of bialaphos and sublethal concentrations of the bar-resistant transformant of C. gloeosporioides f,sp. aeschynomene resulted m a significant increase m disease severity on northern jomtvetch (46). Decreased disease severity was observed when jomtvetch was co-inoculated with the wild-type strain of C. gloeosporzoides fsp. aeschynomene and bialaphos. Thus, coapplication of the blalaphos-reslstant mycoherblclde and bialaphos resulted in increased efficacy against the target weed, northern jointvetch, as well as increased virulence against a closely related weed, Indian jointvetch. 3.3. Genetic Manipulation of Virulence and Host Range In attempts to improve the efficacy of mycoherblcides, several research groups are studymg ways to genetlcally alter the vuulence and host-speclficlty of pathogens. Miller et al. (25) obtained an auxotrophic mutant of S. sclerotzorum that required cytosine for growth. This cytosme auxotrophy rendered the pathogen avirulent on four of seven susceptible species of plants, m the absence of exogenous cytosine. However, if cytosme was present in the moculum or coapplied at the inoculation site, the auxotroph was virulent on all seven susceptible plant species. Isolates of this fungus obtained from nature attack

366

Pilgeram and Sands

numerous beneficial crop and native plants, but the auxotroph was limited to the area of cytosmeapphcatron, thus reducing the threat to beneficial plants. Thus, the mutatton increasedthe host-specrficrty of this broad-host-range pathogen, and provided a means to limit its drssemmattonafter Its release mto the environment. Another approach to increasing the vnulence of bioherbtctdes IS to transform the brocontrol agent with genes encoding vtrulence factors, such as productron of aphytotoxm. The genesencoding productton of the natural herbicide btalaphos were cloned from S. hygroscopzcus@I), and transferred mto the plant pathogenic bactermm, Xanthomonas campestris pv. campestrzs(47) The transformed bacteria retained pathogemctty to then- natural hosts (broccoh and cabbage), and showed an altered hypersensttrve response on four nonhost plants. This altered hypersensitive response suggests that the effecttve host range of the plant pathogemc bacterium was broadened. Transfer of this technology to mycoherbtctdal strains of plant pathogenic fungi may be limited by the number of genes and sizeof the gene cluster necessaryfor btalaphos productton (4649). Enzyme production by many plant-pathogenic organisms facrhtates plant penetration and establishment (52-54). The impact of overexpressron of vnulence factors, such as enzymes, IS yet to be reported. However, a mutant of S sclerotlorum, which constrtutrvely produced high levels of several cellulytrc enzymes, was highly vuulent (Sands, personal commumcatton). However, thus massive productron of enzymeswas counterproductrve to its long-term survtval. 3.4. Containment of Mycoherbicides Chlamydospores are generally formed by F. oxysporum m response to starvatron or an adverse envtronment. The authors inadvertently obtained a mutant of F oxysporum fsp. papaverzs that does not produce normal numbers of chlamydospores m response to unfavorable conditions, but continues to produce vast quantmes of microcomdra (55). Such mutants concervably are more pathogenic, because they produce more mrcrocomdia, and the energy normally diverted to form chlamydospores can be channeled to mlcrocomdra and/or enzyme productron Survival of thuspathogen m the envu-onment m the absence of chlamydospore productron would be severely restricted An additional mutant of S. sclerotzorum was characterrzed, and was mcapable of producing sclerotia, but did retam wild-type pathogenicity on eight host species (56). Lack of sclerotia formation renders this mutant incapable of producing ascospores or survrvmg adverse envnonmental conditions, such as freezing or desiccation. 4. Conclusion Mycoherbtcide development has been hmdered by competitron from highly efficacrous and profitable chemical herbicides, the perceived risk and potentral

Mycoherbicides

367

liability of btologlcal releases, restricted market size because of the host speclficity of potential bioherblcides, environmental constraints imposed on their pathogemcity, and to the regional distribution of most target weeds, and, finally by the cost m time and money reqmred for bioherblcide development and registration. The authors believe these obstacles are slowly being overcome. The potential market size continues to increase, as the population becomes more envlronmentally conscious and demands alternatives to chemical weed management. Environmental risk IS being prioritized by the development of containment strategies that limit survival and dissemination of released bloherblcldal species. Registration of products is likely to be expedited as It becomes more routine. These changes should increase funding for biologlcal control research, and promote bIologIca products m the marketplace. References 1 Cullen, J. M and Hasan, S. (1988) Pathogens for the control of weeds Phil Trans R Sot London 318,213-224. 2 Hoar, S K , Blair, A., Holmes, F A, Boyson, C D., Robel, R. J., Hover, R., and Fraumem, J. F (1986) Agricultural herbicide use and risk of lymphoma and softtissue sarcoma JAMA 256, 1141-l 147. 3. Templeton, G. E , TeBeest, D. 0 , and Smith, R. J (1984) Biological weed control m rice with a strain of Colletotrlchum gloeosporzozdes (Penz.) Sacc used as a mycoherblclde Crop Protection 3,409-422. 4 Gressel, J (1996) Btotechnology of weed control, m Bzotechnology In Agrrculture (Altman, A , ed ), Marcel Dekker, New York, pp. l-3 1 5 Nyvall, R F. and Hu, A. (1997) Laboratory evaluation of mdlgenous North America fungi for bIologIca control of purple loosestnfe. Bzol Control 8,37-42, 6 Wapshere, A. J., Delfosse, E S., and Cullen, J. M. (1989) Recent developments m blologlcal control of weeds. Crop Protection 8, 227-250 7 Charudattan, R and DeLoach, C J , Jr (1988) Management of pathogens and insects for weed control m agroecosystems, in Weed Management in Agroecosystems* Ecologzcal Approaches (Altien, M. A. and Llebman, M., eds.), CRC, Boca Raton, FL, pp 245-264 8. Hasan, S. (1981) A new strain of the rust fungus Puccznza chondrzllwza for blological control of skeleton weed m Australia Ann. Appl Bzol. 99, 119-124 9 Cullen, J. M , Kable, P. F , and Catt, M (1973) Epldemlc spread of a rust Imported for blologlcal control. Nature 244,462-464. 10 Marsden, J S., Martin, G. E., Parham, D J., Ridsdill Smith, T. J , and Johnston, B G (1980) Returns on Australian agricultural research. 84-93 The Joint Industry Assistance Commlsslon-CSIRO benefit-cost study of CSIRO Dlvlslon of Entomology, Canberra, Australia 11 Supkoff, D M., Joley, D B., and Marols, J J. (1988) Effect of Introduced blologlcal control orgamsms on the density of ChondnllaJuncea m Cahforma. J Appl Ecol 25, 1089-1095

Pilgeram and Sands

368

12 Adams, E. B and Line, R. F (1984) Eptdemiology and host morphology m the parasitism of rush skeleton by Puccmza chondrtlhna Phytopathology 74,745-748. 13. Emge, R. G , Melchmg, J. S., and Kmgsolver, C. H (1981) Epidemiology of Pucctma chondrtllzna, a rust pathogen for the btologtcal control of rush skeletonweed in the Umted States. Phytopathology 71, 839-843 14. Holm, L., Doll, J., Helm, E., Pancho, J., and Herberger, J., eds (1997) World Weeds, Wiley, New York, 1129 pp 15 Auld, B. A and Morm, L (1995) Constraints m the development of btoherbicldes Weed Technol 9,638-652 16 Bruckhart, W L and Dowler, W. M (1986) Evaluation of exotic rust fungi m the United States for classtcal control of weeds Weed Sci. 34, 1 l-14. 17, Phatak, S C., Sumner, D R., Wells, H. D , Bell, D K., and Glaze, N C. (1983) Biological control of yellow nutsedge with the indigenous rust fungus Puccznza canaliculata. Science 219, 1446,1447 18. Daniel, J. T., Templeton, G E., Smith, R J., and Fox, W. T (1973) Biological control of northern Jomtvetch in rice with an endemic fungal disease. Weed Scz 21,303-307.

19. Charudattan, R (1991) The mycoherbictde approach with plant pathogens, m Mzcrobzal Control of Weeds (TeBeest, D. O., ed ), Chapman and Hall, New York, PP 24-67 20 Mortensen, K. (1986) Btologtcal control of weeds with plant pathogens Can J Plant Path01 8,229-23 1. 2 1 Brosten, B. S and Sands, D. C. (1986) Field trtals of Sclerotznza sclerotzorum to control Canada Thistle (Czrstum arvense). Weed Set 34, 377-380. 22 Turner, S K , Fay, P. K , Sharp, E L , and Sands, D C (1981) Resistance of Canada thtstle (Ctrsrum arvense) ecotypes to a rust pathogen (Puccrma obtegens) Weed Set 29,623,624. 23. Bourdot, G. W , Harvey, I. C., Hurrell, G. A., and Saville, D J. (1995) Demographic and biomass productton of Ctrstum arvense wtth Sclerotmta sclerotrorum Btocontrol Set Technol 5, 1 l-25 24 Boland, G. J and Hall, R (1996) Index of plant hosts of Sclerotmza sclerotiorum Can J Plant Path01 I&93-108 25 Miller, R. V., Ford, E J , Ztdack, N J., and Sands, D C. (1989) A pyrtmtdme auxotroph of Sclerottnla sclerottorum for use m btological weed control J Gen Microbial 135,2085-209 1, 26 Cook, R J , Bruckart, W L., Coulson, F R , Goettel, M. S , Humber, R A, Lumsden, R. D , et al (1996) Safety of Microorgamsms intended for pest and plant disease control. a framework for sctenttfic evaluation, Bzol Control 7, 333-35 1 27. TeBeest, D 0 , Yang, X B , and Ctsar, C. R (1992) The status of biological control of weeds with fungal pathogens. Ann Rev Phytopathol. 30, 637-657 28. Kenney, D S. (1986) DevmeO-The way tt was developed-an mdustrialist’s view Weed Scz. 34(Suppl. 1), 15,16 29 Bowers, J (1986) Commerciahzation of Collego-an industrialist’s view Weed Sci 43(Suppl.

l), 24,25

Mycoherbicides

369

B, 30 Grant, N. T , Prusmkiewicz, E , Makowski, R. M. D , Holmstrom-Ruddick, and Mortensen, K (1990) Effect of selected pesticides on survival of Colletotrtchum gloeosponoldes f. sp. malvue, a btoherbicrde for round-leaved mallow (Malva pusdla) Weed Technol 4,701-7 15 31 Grant, N. T , Prusmktewtcz, E., Mortensen, K., and Makowskt, R. M. D. (1990) Herbicide mteractions with Colletotrlchum gloeosporloldes f. sp malvae, a bioherbtcide for round-leaved mallow (Malva puszlla) control Weed Technol. 4, 716-723 32 Imaizumi, S., Ntshmo, T., Miyabe, K , FuJimori, T , and Yamada, M (1997) BIOlogical control of annual bluegrass (Poa annua L.) with a Japanese isolate of Xanthomonas campestrzs pv poae (JT-P482). Bzol Control 87-14 33 Jong, M. D , de Scheepens, P C , and Zadoks, J C (1990) Risk analysts for biological control. a Dutch case study m btocontrol of Prunus serotzna by the fungus Chondrostereum purpureum Plant Du 74, 189-194 34 Weidemann, G J , Boyette, C D , and Templeton, G E (1995) Utilization criteria for mycoherbicides, m Bzoratzonal Pest Control Agents (Hall, F. R and Barry, J. W , eds.), American Chemical Society, Washmgton, DC, pp 238-25 1 35. Amsellem, Z , Sharon, A , Gressel, J., and Quimby, P. C , Jr. (1990) Complete abolition of high inoculum threshold of two mycoherbtctdes (Alternarla casszae and A crassa) when applied m invert emulston Phytopathology 80,925-929 36 Hallett, S. G , Paul, N D., and Ayers, P G. (1990) Botrytzs cznerea kills groundse1(Seneczo vulgarzs) infected by rust (Puccrnza lagenophorae) New Phytologzst 114,105-109. 37. Morin, L., Auld, B. A , and Brown, J. F. (1993) Synergy between Puccznla xanthzz Colletotrichum orblculare on Xanthlum occidentale. Blol Control 3,296-3 10. 38. Cother, E J (1992) Commensabsm to synergism the potential role for biological combmations in bioherbtcides Plant Protection Q 7, 157-l 59 39 Khodayari, K and Smith, R J (1988) A mycoherbicide integrated with fungicides m rice, Oryza satlva Weed Technol 2,282-285 40. Beste, C. E , Frank, J R , Bruckart, W. L., Johnson, D R., and Potts, W E (1992) Yellow nutsedge (Cperus esculentus) control in tomato with Pucclnla canallculata and pebulate Weed Technol. 6, 980-984 41. Chacko, R J , Weidemann, G. J., TeBeest, D 0 , and Correll, J. C. (1994) The use of vegetative compatibility and heterokaryosis to determine potential asexual gene exchange in Colletotrzchum gloeosporioldes B~ol Control 4,382-389 42 Yang, X B and TeBeest, D 0 (1995) Compettttveness of mutant and wild-type isolates of Colletotrlchum gloeosporloldes f. sp aeschynomene on northern jomtvetch. Phytopathology 85, 705-7 10. 43 Cisar, C R., Thornton, A B , and TeBeest, D. 0. (1996) Isolates of Colletotrzchum gloeosporroldes (Telomorph Glomerella czngulata) with different host specificities mate on northern Jointvetch Blol Control 7, 75-83 44. Kremer, R J and Schulte, L. K. (1989) Influence of chemical treatment and Fusarmm oxysporum on velvetleaf (Abutzlon theophrastz) Weed Technol 3, 369-374

370

Pilgeram and Sands

45 Sharon, A., Amsellem, Z , and Gressel, J. (1992) Glyphosate suppression of an elicited defense response Increased susceptibilny of Cassza obtuslfolta to a mycoherbicide. Plant Physzol 98, 654-659. 46 Brooker, N L., Mischke, C F , Patterson, C D , Mischke, S , Bruckhart, W L , and Lydon, J (1996) Pathogemcity of bar-transformed Colletotrzchym gloeosporlotdes f sp Aeschynomene. Blol Control 7, 159-l 66 47 Prasad, R (1994) Influence of several pesticides and admvants on Chondrostereum purpureum-a bioherbicide agent for control of forest weeds Weed Technol 8,445%449 48 Bruckart, W L , Johnson, D R., and Frank, J R. (1988) Bentazon reduces rustmduced disease m yellow nutsedge, Cypresus esculentus) Weed Technol 2, 299-303. 49 Charudattan, R , Prange, V. J , and Devalerto, J. T. (1996) Exploration of the use of the “bialaphos genes” for improving bioherbicide efficacy Weed Technol 10, 625-636 50 Upchurch, R G , Meade, M J , Hightower, R. C., Thomas,R S., andCallahan,T M (1994) Transformation of the fungal soybean pathogen Cercospora kzkuchzzwith the selectablemarker bar. Appl Envu-on Mlcroblol 60,4592-4595 5 1. Murakami, T., Anzai, H., Imai, S , Satoh, A., Nagaoka, K., and Thompson, C J (1986) The bialaphosbiosynthetic genesof Streptomyces hygroscoprcus. molecular clonmg and characterization of the genecluster Mol Gen Genet. 205,42-50 52. Shimosaka,M, Kumehara, M , Zhang, X. Y , Nogawa, M , and Okazaki, M (1996) Clonmg and characterization of a chitosanasegene from the plant pathogemc fungus Fusarlum solant J Fermentation Bloeng 82,42&43 1 53. Battling, S., Van den Honbergh, J. P. T W., Olsen, O., von Wettstem, D., and Visser, J. (1996) Expression of an Erwrnla pectate lyase in three species of Aspergdlus Curr Genet 29,47448 1. 54 Pietro, A and Roncero, M I G (1996) Purification and characterization of a pectate lyase from Fusarlum oxysporum f. sp. lycoperslcl produced on tomato vascular tissue. Physrol Mol Plant Path01 49, 177-l 85 55 Pilgeram, A P , Tiourebaev, K., Weaver, M B., Morgan, C T., Anderson, T W , McCarthy, M K., Tepe, M., and Sands,D. C (1996) Characterization of a chlamydospore mutant of Fusarlum oxysporum f sp papaver American Phytopathological Society, Indianapolls, IN. July 27-3 1. 56 Miller, R V , Ford, E J , and Sands, D C. (1989) A nonsclerotial pathogenic mutant of Sclerotmra sclerotlorum. Can J Mzcroblol 35, 5 17-520.

20 Formulation and Application of Plant Pathogens for Biological Weed Control Nina K. Zidack and Paul C. Quimby, Jr. 1. Introduction The public’s demand for safe, biologically based weed controls has created an impetus to commercialize biological weed-control agents. Numerous plant pathogens have shown excellent potential as biological herbicides over the years, but there has been very little commercial success.For bioherbicides to succeed in the marketplace, products must be efficacious, consistent, easy to use, economical, and have adequate storage hfe. The satisfaction of these criteria depends on the development of formulation technology tailored to the mdustries that will develop the bioherbicides, and to the consumers who ~111 use the products. Traditionally, biological control has been divided into two mam areas: classical and augmentative Classical biocontrol involves the mtroduction of a weed pathogen from the weed’s area of origin. Ideally, the establishment of a natural enemy against the mvasive weed brings the introduced weed mto a more balanced positron in the plant community. This method is primarily used on widespread infestations, often on such areas as forests and rangelands. Augmentative or inundative biocontrol involves the use of endemic plant pathogens as btological herbicides. The ultimate goal of this approach IS to replace or reduce chemical herbicide use and/or provide controls for weeds that have no alternative controls, or are closely related to the crop in which the weed is a problem. The type of biocontrol agent, whether classical or augmentative, dictates the criteria for the development of appropriate formulation and application technology. For classical control, cost may not be as important, because long-term economic benefit will be realized if the agent is successful. Also, optimization From Methods m B/ofechno/ogy, vol 5 EOopestmdes Use and De//very Edited by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

371

372

Zidack and Quimby

of control m the apphcatton area is not necessarily required m the first year, because successful agents will be self-perpetuatmg. The formulatron requuements for classical agents are more flexible, because the applications will most likely be performed by someone tramed in application techniques for biologtcal weed control. For augmentative control, the goal is to produce a commercially viable, consistently efficacious product. Formulation and application methods must be adaptable to conventional equipment, and have adequate shelf life for marketing purposes. Once the product is commercialized, it will be applied by people who have little or no experience working with microbes and their unique environmental sensitivities. Therefore, the formulatron and apphcation method needs to be as msensmve as possible to environmental fluctuations and reasonable variations m application protocol. Shelf life is much more important, because the products will be stored for marketing over a long period of time, often m factlmes with variable environmental control. This chapter will describe biological weed control formulatton technology in use m the laboratory and at the commercial level. Addmonally, literature on formulation of biological insecticides and brologtcal disease control is addressed when the technology is transferrable to btoherbicides. 2. Driving Factors for Formulation and Application 2.1. Biology of the Plant The biology of the target weed, as well as its habitat, will dictate which types of formulations and apphcatron are most appropriate. Annual weeds may be adequately controlled with fohar plant pathogens applied m a sprayable formulation; perennial weeds may require a systemic, soilborne type-pathogen, formulated as a granule. Weed habitat also influences formulation requnements. Xertc weeds will present different challenges than plants growing m mesic or aquatic environments. 2.2. Biology of the Pathogen The biology of the pathogen is also very important. In the case of fungi, the ability to sporulate in submerged culture is desirable, because liquid fermentation 1smore practical and economical than other methods (1,2). Solid-state fermentation methods have been developed for many fungi, but they are typically more difficult and expensive (3). Conrdia of some fungi have proven to be difficult to stabilize n-rdry formulatrons, and more resistant structures have been chosen for formulation. Jackson et al. (4) showed that microsclerotia formulated with corn flour retained viabihty for 18 mo at 4°C. Earher studies, were not able to achieve adequatesurvival of conidta, even immediately after formulation (56) Also, dew requirements for fungal spore germination have necessitateddevelopment of technology to reduce the critical period when moisture is required

Plant Pathogens for Weed Control

373

Plant pathogenic bacteria offer unique challenges for formulation. The majority of plant pathogenic bacteria are more environmentally sensrtive than most fungi, which often have pigmented spores or structures that are naturally adapted to withstanding environmental stresses,such as UV radiation and desiccation. Also, bacteria are not able to penetrate leaves directly, and require wounds or free water on the leaf surface to enter natural openings, such as stomates, hydathodes, or lenticels (7).

3. Materials and Methods for Formulation (Historical) The simplest formulations are fresh preparations of the biocontrol agent. One of the most successful bioherbicide products to date IS formulated m this manner. DeVme@ (Abbott Labs, Long Grove, IL; Phytophthora palmivora), a product to control strangler-vme m citrus, is formulated as a fresh liquid with a 6-wk shelf life. This shelf life is adequate, because of the spatially limited market area for this product (8). Another technology under consideration by Mycogen (San Diego, CA) is the use of on-site fermenters to produce a bacterial bioherbicide, Xanthamonas campestrzspv poae, for annual bluegrass (Poa annua) on golf greens, Fresh inoculum would be produced as needed (9). Infested organic materials have been used extensively for fungal inoculum, especially at the experimental level. This method IS suitable for mycelial fungi, or for organisms that do not sporulate readily in liquid culture However, a large amount of inoculum IS often required for weed mortality in the field. For instance, to achieve sigmficant kill of Canada thrstle (Circzum arvense) m the field with Sclerotinia sclerotlorum, 830 kg/ha of infested wheat seed was applied (10). Pre-emergent herbicidal activity of the saprophytic, phytotoxmproducing fungus Gliocladzum virens was ftrst tested using Infested rice. Ultimately, formulations were developed utilizing peat, which was chosen because its acidic nature enhances the stability of the phytotoxm viridiol. The addition of sucrose and ammonmm nitrate significantly enhanced production of the phytotoxin. Although this formulation was effective, very high levels of inoculum (815 kg/ha) were required for weed control (21,22). Mmeral-based formulations have been developed for agents targetmg weeds and sotlborne drseases.Diatomaceousearth granules tmpregnatedwtth 10% molasseswere effective for the growth and delivery of Trichoderma harzianum for control of Sclerotzum rolfszi m peanuts (13). Lignite granules, amended with stillage (a biproduct of sorgum fermentation), were used to culture and drssemrnate T. harzianum and G virens (14). Clay matrices have been used to stabilize the plant pathogens Colletotrlchum gloeosporloides f.sp. clldemlae and Septoria passijlorue. Conidial preparations of each fungus were mixed directly with kaolm, then air-dried, pulverized, and stored at -18, I, and 22°C. Optimal storage for each organism was -18 and 1°C respectively, with greater

374

Zidack and Quimby

than 84% survtval for the Colletotrzchum and 95% for the Septorza over a 4-mo period (15). Encapsulation of microbes m sodium algmate and kaolin clay was first described m 1983 for the fungi Alternaria macrospora, A. casszae,Fusarium lateritzum, Colletotrzchum malvarum, and a Phyllosticta sp (16). Algmates are a family of copolymers contaimng 1,4-lmked /3+mannuronrc and a-L-galuromc acid residues m varying proportions and sequential arrangements (27). Algtnate has been used extensively m formulations of biological weed-control agents, and also m fungal preparations for biological control of sotlborne diseases (18,191. In general, moculum is prepared by dropwise addition of homogenized mixtures of sodmm algmate, kaolin clay, and mycelmm mto a 0 25 M CaCl, solution (20). Weidemann (21) showed that conidial production in sodium alginate granules and field efficacy could be improved when nutritional amendments were added. Production of comdta by Fusarium solani f.sp. cucurbztae was sigmficantly increased by adding 2% oatmeal, cornmeal, or soyflour. Texas gourd control was >80% with oatmeal and soyflour-amended granules at 220 kg/ha; nonamended granules gave only 3% control. The survival of bacteria, fungi, and nematodes m alginate formulations has been improved by coating granules with an inverting oil, followed by an oil adsorbent (22). The enhanced survival is putatively a result of a retardation of evaporation of water from the granules during the drying process. After 6 h of drying at room temperature, coated granules were at -40 bars; uncoated granules were at-74 bars. In the caseof Colletotrzchum truncatum, coated granules had three times as many colony-forming units/g as uncoated granules after 42 h of drying. The nematode Subanguzna picrzdu, a btocontrol candtdate for Russian knapweed, survived for 9 mo at -20°C with no significant loss of mfectivity when formulated with this method (23) Conmck et al. (24) recently developed granules formed by encapsulation of mycoherbictdes in a wheat-gluten matrix, using a pasta-like process, which they have named Pesta. Pasta-like dough was prepared with semolma flour, kaolm, and fungal biomass grown m shake culture. The dough was then kneaded and passed through a pasta maker repeatedly, producing an homogenous sheet. Sheets were allowed to dry until they could be broken and ground. This process produced stable moculum of Fusarzum oxysporum for sicklepod, coffee senna, and hemp sesbania (25). Through controlled humidity experiments, they have shown that the water activtty of the formulation has a profound effect on the survival of the organism entrapped in the matrix. In Pesta formulations of C truncatum (for control of hemp sesbama), they determmed that, when the water activity reached a level at which free (unbound) water was available to the fungus (a, > 0.35), fungal metaboltsm was possible, and long-

Plant

Pathogens

for Weed Control

375

term survtval of the fungus was adversely affected (26,27). Recently, new technology has been adapted for preparation of Pestagranules, whtch mvolves twm screw extrusion and fluid-bed drying, which can be used for commerctal quantities of mycoherbrctde formulatrons (28). More recently, a granulatton procedure has been developed utrhzmg a waterabsorbent starch, sucrose, unrefined corn oil, and silica (29) The method mvolves suspending fungi or bacteria in a sucrose solution and mtxmg wrth water-absorbent starch (Waterlock @,Muscatme, IA) and unrefined corn 011, which forms a cohesive dough. The dough is then granulated by mixing with hydrated silica (Hr-Sil 233@,Pittsburgh, PA) and air-dried. The granulation process 1svery simple when compared to preparation of alginate granules, and has been used successfully for a variety of fungi, bacteria, and a plant-pathogenic nematode (Qurmby and Zrdack, unpublished results). Thus formulatton can be used either as a granule applied to the soil, or, after sieving, as a wettable powder for spray formulatrons. Preliminary data indicate that the mclusron of sucrose m the formulatton is very important for enhancing survival of the organisms m the formulation. The sucrose putatively stabilizes the membranes during the drying process by replacing water molecules m the lipid btlayer. This has been shown for bacteria that have been freeze-dried (30) 4. Materials and Methods for Foliar Application (Historical) Several adjuvants have been shown to increase both survival of mycoherbrcrdal fungi and broherbicide efficacy. Addmon of sorbitol to moculum of Colletotrzchum coccodes for velvetleaf control increased numbers of viable spores recovered from maculated leaves 20-fold, but dew was still reqmred for infection (32). In contrast, the addition of gelatin, sorbrtol, or Bond@(Loveland Industrres, Loveland, CO), a latex product, did not increase mfectron under subopttmal moisture condmons, when used as spray additives for Phomopsis coy1voIvu1us,a broherbicide candidate for field bindweed (32). Dangle and Cotty (33) showed that a variety of additives influenced germination of Alternarza casszae comdra m vitro, and that this could be correlated with performance of the bioherbicrde in greenhouse bioassays. One percent potato dextrose broth, 0.1-l% Tween-80, and 0.02 M potassium phosphate buffer promoted germrnation m vitro, and 100% of the plants were killed after a 24-h dew period, compared to 25% kill for spores in buffer alone. Boyette and Abbas (34) reported that the addition of fruit pectin and plant filtrates to spray solutrons of Alternarza crassa actually altered the host range of the fungus. The addition of pectin to comdial suspensions resulted in 100% mortality of the weeds hemp sesbama, showy crotalaria, and eastern black nightshade. Apphcatron of the comdta in filtrates of the host plant, jimsonweed, also caused normally resrstant plants to become susceptrble. Some crop plants, including tomato, were

Zidack and Qumby also susceptible when sprayed with the amended comdta. The authors postulate that apphcatron of the spores with the amendments induced enzymes, such as pectin esterase,which may have enhanced pathogenests. They also constder the inhibition of antifungal metabohtes (phytoalexms) to be a possibility. Humectants also have the capacity to increase the infection rate of plant pathogens on the leaf surface. Psyllmm hydrophlhc mucilloid (Metamucrl@) at 0.5%, augmented disease levels when applied with mycelial moculum of Alternarza eichhornza, a pathogen of waterhyacmth (35). Dtsease caused by powdered algmate formulattons of the same pathogen was augmented by use of a hydrophilic polyacrylamide (36). The dew requirement for infection by fungal spores was reduced wtth the development of water-in-o11 invert emulsions. Original experiments showed that a mixture of mineral oil, paraffin, and lecnhm, mixed in a 6 5 ratio with the aqueous phase, drastically reduced evaporatton of water from the preparation When A. cassza spores were sprayed m the Invert emulsion wrthout dew, 88% sicklepod mortahty was achieved, compared to no mortality when spores were sprayed m the aqueous carrrer alone (37). The original, Invert emulsion formulations were extremely viscous and required spectahzed,air-assist atomlzmg spray nozzles (38). Connick et al. (39) reported on an improved invert emulsion with reduced viscosity and high water-retention capabthty. This formulatlon utilized an unsaturated monoglycertde as an emulsifier, instead of lecithin. Application of A. casszaeand A. crassa in an invert emulsion resulted in mfectton, usmg extremely low levels of moculum when compared to a stratght aqueous application. Only one spore m a 2-uL droplet was required to infect plants of Casszaobtusifolia and Datura stramonium (40). Similar results were realized when Ascochyta pteridzs was applied in an invert emulsion for control of bracken (Pterzdzumaquzlznum)(41). Amsellem et al. (42) went on to show that application of A cassiae and A crassa in invert emulsion abolished their selectivity and caused them to attack eight other plant species tested. Nonpathogenic fungi were also able to colonize plants, when they were applied m the invert emulsron The authors hypothesized that the invert emulsion may have caused cutrcular damage that allowed penetration by the fungi. Schrsler et al. (43) reported enhancement of disease by C truncatum m hemp sesbarna (Sesbania exaltata) by comoculating with eprphyttc bacteria. They identified a number of bacterial isolates that stimulated appressorla formatton and enhanced diseasesymptoms on hemp sesbania.Fernando et al. (44) reported enhanced efficacy of Colletotrlchum coccodes on velvetleaf when the fungus was comoculated wrth phylloplane bacteria. Although the bacteria decreased comdial development, appressorta formatton was stimulated and germ tube length was decreased. Thts suggested that disease may have been enhanced by reducing the saprophyttc, premfectlon mycehal growth.

Plant Pathogens for Weed Control

377

Application of bactertal pathogens to plants requires special formulatron consideratrons. Because bacteria are not able to penetrate plants directly, then entry to the plant must be artrfictally facilitated. This has been done successfully in two ways. The first is through the use of a nonionic organosilicone surfactant, which lowers the surface tension of aqueous soluttons to the point that stomates are penetrated (45). The potential of this method was illustrated m field experiments on kudzu with the bacterial pathogen Pseudomonas syrzngae pv. phaseolzcola. Field applications were made with Log 8 CFU/mL Pseudomonas m 0.2% Stlwet L-77 at 745 L/ha. Control of kudzu was not achieved in these experiments, but excellent mfectron levels were attained (46). A second form of apphcation for plant-pathogemc bacterra IS mechamcal woundmg. The pathogen Xanthomonas campestris pv. poannua was applied to golf course greens for control of annual bluegrass in conjunctron with mowing (47). The mowing wounds the plant, providing a mode of entry for the bacteria. This has proven to be an effective control for annual bluegrass, and IS being developed for commercial use in Japan (48). An example of biological weed control with bacterial plant pathogens that does not require facrhtated mfectton is the use of rhizobacterla to suppress weed growth. Kennedy et al. (49) showed an inhtbttion of downy brome and an increase m winter wheat yield m field solIs where deleterious rhizobacterra had been applied to the sot1surface, either as an aqueous spray or m infested straw. 5. Integration with Chemical Herbicides The efficacy of biological herbicides may be synergtzed by applying them with reduced rates of chemical herbicides. Sharon et al. (50) showed specific suppresston of a phytoalexin derived from the shiktmate pathway in Cassza obtustfolza L. by a sublethal dose of glyphosate. This concurrently increased the susceptibility of the weed to the mycoherbicide A. cassia. The amount of moculum needed to cause dtseasesymptoms was reduced fivefold when apphed wtth glyphosate. Synergy of bacterial plant pathogens with sulfosate and glufosinate was demonstrated by Christy et al. (51) in greenhouse and field trials. Bacterial strains that caused no symptoms when applied alone dramattcally increased the activity of sublethal rates of the herbicides when applied together. They named this approach the “X-tend” bioherbicide system.Although results were encouraging, they did not achieve commerctal levels of control. 6. Research Needs Sigmficant advances have been made m formulation and apphcatron technology, but the technology stall has not been developed that is necessary to propel a number of brological weed control products to the forefront of the marketplace. Therefore, concerted research efforts are required m a number of

378

Zidack and Quimby

areas. First, marketability of btocontrol products would be greatly enhanced by mcreasmg the shelf life of formulattons. A lofty goal would be 2 yr of storage at room temperature. This would greatly stmplify mventory management for retatlers, and increase the likelihood that they would stock btologtcal products Second, technology for pre-emergence treatments of pathogens that attack weed seeds, both dormant and germinating,

must be developed. This will require strat-

egtes that deliver moculum below the so11surface. Third, reducing the amount of moculum required for adequate mfectton would make products easier to use, and more economtcal Thus point is critical for reducing productton, transportatton, and applrcatton costs. Fmally, stgrnticant opportumty resides m the areas of pathogen vuulence enhancement and ameltoratton of plant defense response. A

recent review by Hoagland (52) describes abundant research into both plant and pathogen btochemtstry that could be exploited to enhance the efficacy of btologtcal weed control agents. The mclusron of novel synerglsts m btoherbtctde

formulations could take them past the point of research, and mto the development of efficacious, reliable, and economtcal

products for the marketplace.

References 1 Bowers, R. C (1982) Commerclabzatron of mtcrobtal btological control agents, m Brologlcal Control of Weeds wzth Plant Pathogens (Charudattan, R and Walker, H L., eds.), Wrley, New York, pp 157-I 73 2 Bowers, R. C (1986) Commerclahzatton of CollegoTM an industrralrst’s view Weed Scl 34s, 24,25 3 Boyette, C D , Qurmby, P C , Jr, Connick, W J., Jr , Dangle, D J , and Fulgham, F E (1991) Progress m the productron, formulatron, and apphcatlon of mycoherbrctdes, m Mlcroblal Control of Weeds (TeBeest, D 0 , ed ), Chapman and Hall, New York, pp 209-222 4 Jackson, M A., Shasha, B S., and Schrsler, D A. (1996) Formulation of Colletotrlchum truncatum mlcrosclerotta for improved biocontrol of the weed hemp sesbama (Sesbanla exaltata). BIOI Control 7, 107-l 13 5 Jackson, M A. and Schisler, D A (1992) The composrtion and attributes of Collectotrlchum truncatum spores are altered by the nutritional environment Appl Envzron Mlcrobrol S&226&2265 6 Stlman, R W , Bothast, R. J , and Schrsler, D A (1993) Productron of Collectotrlchum truncatum for use as a mycoherbiclde: Effects of culture, drying and storage on recovery and efficacy Blotech Adv 11,56 l-575 7 Agnes, G N (1988) Plant diseases caused by prokaryotes, m Plant Pathology Academic Press, San Drego, CA, pp 5 1O-586 8 Kenney, D S (1986) DeVme@-The way tt was developed-an mdustrtallst’s view Weed Scl 34(Suppl. l), 15,16 9 Zomer, P. S (1996) Mycogen Corporation, San Dtego, CA. Personal commumcatlon IO. Brosten, B S and Sands, D C (1986) Field trials of Sclerotmza sclerotzortlm to control Canada thistle (Clrszum arvense) Weed Scl 34,377-380

Plant Pathogens for Weed Control

379

11. Jones, R. W and Hancock, J. G (1987) Conversron of vtrtdm to vtrtdtol by vtrtdmproducmg fungi. Can J Mzcrobiol 33, 963-966 12. Jones, R. W , Lanmr, W. T., and Hancock, J. G. (1988) Plant growth response to the phytotoxm viridtol produced by the fungus Glzocladzum virens Weed Scz 36, 683-687. 13 Backman, P. A and Rodrtquez-Kabana, R. (1975) A system for growth and delrvery of biologrcal control agents to the sot1 Phytopathology 65, 8 19-82 1. 14. Jones, R W., Pet& R. E , and Taber, R A. (1984) Lignite and sttllage: carrter and substrate for applicatron of fungal brocontrol agents to so11 Phytopathology 74, 1167-1170 15 Norman, D J. and TruJlilo, E. E. (1995) Development of Colletotrzchum gleosporzozdes f. sp. clzdemzae and Septoriapasszflorae into two mycoherbrctdes with extended vtabtltty Plant DES 79, 1029-1032. 16 Walker, H L , and Conmck, W. J., Jr. (1983) Sodium algmate for productron and formulation of mycoherbtcrdes. Weed Sci 31,333-338 17 Martmsen, A., SkJak-Braek, G., and Smrdsrod, 0 (1989) Algmate as immobrbzatron matenal. I Correlation between chemical and physical properties of algmate gel beads Bzotechnol Bzoeng 33,79-89 18. Marois, J. J , Fravel, D R , Conmck, W J , Jr, Walker, H. L., and Qunnby, P C. (1989) US Patent 48 18530 19 Papavtzas, G. C , Fravel, D. R , and Lewrs, J A (1987) Prohferatron of Talaromycesflavus m sot1 and survival in alginate pellets Phytopathology 77, 131-136. 20. Boyette, C D and Walker, H L (1985) Productton and storage of moculum of Cercospora kikuchzz for field studies. Phytopathology 75, 183-l 85 21. Werdemann, G. J (1988) Effects of nutrmonal amendments on cotndtal production of Fusarzum solanz f. sp. cucurbztae on sodmm alginate granules and on control of Texas gourd Plant Dzsease 72,757-759. 22. Quimby, P C., Jr, Birdsall, J L , Caesar, A. J , Connick, W. J , Jr., Boyette, C D , Caesar, T C., and Sands, D C (1994) US Patent 5358863 23 Caesar-Tonthat, T C., Dyer, W. E., Quimby, P C., Jr, and Rosenthal, S S (1995) Formulatton of an endoparasitrc nematode, Subanguzna pzcrzdis Brzeskt, a btological control agent for Russian knapweed, Acroptzlon repens (L) DC. Bzol Control 5,262-266. 24 Connick, W J , Jr., Boyette, C D , and McAlpine, J R (199 1) Formulation of mycoherbtcrdes using a pasta-like process. Bzol Control 1,281-287. 25. Boyette, C D., Abbas, H. K., and Connrck, W. J , Jr. (1993) Evaluatton of Fusarzum oxysporum as a potential bioherbrcrde for sicklepod (Cassza obtuszfolza), coffee senna (C occrdentalis), and hemp sesbania (Sesbanza exaltata) Weed Scz 41,678-68 1 26. Connick, W J., Jr, Dargle, D. J , Boyette, D D , Wrlliams, K S , Vmyard, B T , and Quimby, P C., Jr. (1996) Water activity and other factors that affect the viability of Colletotrzchum truncatum conidia m wheat flour-kaolin granules (‘Pesta’). Bzocontrol Scz Technol 6,277-284

380

Zidack and Quimby

27. Connlck, W , Jr., Dangle, D., Williams, K , Vinyard, B., Boyette, D., and Qmmby, P , Jr (1996) Shelf life of a bioherblcide product. Am Bzotechnol Lab 14, 34,35 28. Dangle, D J , Conmck, W J , Jr., Boyette, D D , Lovlsa, M. P., Wllhams, K. S , and Watson, M (1997) Twm-screw extrusion of ‘Pesta’-encapsulated blocontrol agents. World J Microblol Blotechnol., in press. 29 Qulmby, P C., Jr, Caesar, A J , Birdsall, J. L , Conmck, W J , Jr, Boyette, C D., Zldack, N. K., and Grey, W E. (1996) Granulated formulation and method for stablhzing blocontrol agents US Patent Application 08/695249 30 Leslie, S B , Israeli, E , Lighthart, B., Crowe, J H , and Crowe, L M (1995) Trehalose and sucrose protect both membranes and proteins in intact bacteria durmg drying Appl Environ Mcroblol 61,3592-3597 31. Wymore, L. A., and Watson, A. IS (1986) An adJuvant Increases survival and efficacy of Colletotrlchum coccodes, a mycoherblclde for velvetleaf (Abuflon theophrustl) Phytopathology 76, 1115,1116. 32 Morm, L., Watson, A K., and Relleder, R. D (1989) Effect of dew, moculum density, and spray additives on infection of field bmdweed by Phomopsis convolvulus. Can J Plant Path01 12,48-52 33 Dangle, D. J and Cotty, P. J (1991) Factors that influence germination and mycoherblcidal activity of Alternarla cassiae Weed Technol 5, 82-86 34. Boyette, C D and Abbas, H. K. (1994) Host range alteration of the bloherblcldal fungus Alternarla crassa with fruit pectin and plant filtrates Weed Scl 42,487-491 35 Shabana, Y M., Charudattan, R , and Elwakll, M. A (1995) Identification, pathogenicity, and safety of Alternarra erchormae from Egypt as a bloherblclde agent of waterhyacmth. Biol Control 5, 123-l 35 36 Shabana, Y M , Charudattan, R , and Elwakll, M A (1995) Evaluation of Alternarla eichhorniae as a bloherblcide for waterhyacmth (Elchhornia crasszpes) m greenhouse trials. Brol Control 5, 136-144 37. Qulmby, P. C , Jr., Fulgham, F E., Boyette, C D , and Connick, W J., Jr (1988) An invert emulsion replaces dew in biocontrol of sicklepod-a preliminary study, m Pesticide Formulations and Application Systems (Hovde, D A. and Beestman, G B , eds ), American Society for Testing and Materials, Philadelphia, PA, vol. 8, pp 264-270 38 Qulmby, P C., Jr, Fulgham, F E., Boyette, C. D., and Hoagland, R E. (1988) New formulations nozzles boost efficacy of pathogens for weed control Proc Weed Scl Sac 28,52. 39 Conmck, W J., Jr, Dangle, D. J , and Quimby, P. C , Jr (1991) An improved invert emulsion with high water retention for mycoherblcide delivery. Weed Technol 5,442-444 40. Amsellem, Z., Sharon, A , Gressel, J., and Qulmby, P. C., Jr (1990) Complete abohtlon of high moculum threshold of two mycoherblcides (Alternana casslae and A crassa) when applied m invert emulsion. Phytopathology 80,925-929 41 Womack, J G. and Burge, M. N (1993) Mycoherbicide formulation and the potential for bracken control. Pestwde Sci 37,337-34 1.

Plant

Pathogens

for Weed Control

381

42. Amsellem, 2 , Sharon, A , and Gressel, J (199 1) Abolition of selectivity of two mycoherbtcidal orgamsms and enhanced vn-ulence of avtrulent fungt by an invert emulsion Phytopathology 81, 985-988. 43 Shisler, D A., Howard, K M., and Bothast, R. 3 (199 1) Enhancement of dtsease caused by Colletotrichum truncatum in Sesbania exaltata by comoculatmg wtth eptphyttc bacterta. Blol Control 1,26 l-268 44. Fernando, W. G. D., Watson, A. K., and Paulttz, T C. (1994) Phylloplane Pseudomonas spp enhance disease caused by Colletotrwhum coccodes tn velvetleaf. Blol Control 4, 125-l 3 1. 45 Ztdack, N K , Backman, P. S , and Shaw, J J (1992) Promotion of bacterial mfectron of leaves by an organosihcone surfactant. tmphcattons for btologtcal weed control. Bzol Control 2, I1 1-l 17 46. Zrdack, N. K. and Backman, P A. (1996) Biological control of kudzu (Puerarza lobata) with the plant pathogen Pseudomonas syrmgae pv, phaseolzcola. Weed Sa 44,645-649 47 Johnson, B. J. (1994) Biological control of annual bluegrass wtth Xanthomonas campestrw pv. poannua m bermudagrass. HortSclence 29, 659-662. 48. Savage, S (1996) Formerly of Mycogen Corporatton, San Diego, CA Personal commumcation. 49 Kennedy, A. C., Elhot, L. F., Young, F L., and Douglas, C L. (199 1) Rhtzobacteria suppresstve to the weed downy brome. Sol1 Scl. Am. J 55,722-727 50. Sharon, A., Amsellem, Z., and Gressel, J (1992) Glyphosate suppresston of an ehctted defense response. Plant Physiol. 98, 654-659 5 1 Christy, A L , Herbs& K A , Kostka, S J., Mullen, J P., and Carlson, P S. (1993) Synergizing weed biocontrol agents with chemrcal herbrcides, in Pest Control wzth Enhanced Envzronmental Safety (Duke, S. O., Menn, J J., and Plmmer, J R , eds.), ACS Symposmm Serves524, American Chemrcal Society, Washmgton, DC, pp 87-100 52. Hoagland, R. E. (1996) Chemrcal mteractrons with bioherbictdes to improve efficacy. Weed Technol 10,651-674.

V OTHER BIORATIONALTECHNOLOGIES

21 Pheromones

for Insect Control

Strategies and Successes D. R. Thomson,

L. J. Gut, and J. W. Jenkins

1. Historical Perspective There IS general agreement among government agencies, research mstttuttons, industry, grower organizations, and the public that there is a need to reduce reliance on broad-spectrum msectrcides by acceleratmg efforts to incorporate ecologically sound technologies mto agrrcultural pest-management programs. The development and implementation of pest control technology based on behavior-controllmg chemicals, or semlochemicals, offers a umque opportunity to move m this direction. Semrochemrcals are chemical messages that organisms use to communicate with each other. Among the semtochemrcals, insect sex pheromones have probably recerved the most attention from the scienttfic, regulatory, and agrtcultural commumties. Sex pheromones are chemical messagesbetween individuals of the same specres,whrch facrhtate mating. By their nature, pheromones are highly specific and then use for insect control would not disrupt other brological mteractrons wrthm a cropping system. The first published rdentrfication of an msect sex pheromone was that of the silkworm moth Bombyx morz L. (I). During the 1960s the pheromones of 11 other insects were rdenttfied (2) Research on sex pheromones increased m subsequent years, as efforts were made to utilize these materials for managmg insects. Currently, the pheromones of more than 1600 species have been rdentified (3,4). The productron of synthetic copies of sex pheromones has led to the development and widespread commercial use of sex-pheromone traps for monitoring and trapping insect pests m many sectors, mcludmg agriculture, forestry, government detection and quarantine programs, and consumer protection, From Methods VI Botechnology, vol 5 B~opeshndes Use and De//very Edited by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

385

Thomson, Gut, and Jenkins

386 Table 1 Mating-Disruption

Products

Pest Pink bollworm Codlmg moth Tomato pinworm Oriental fruit moth Gypsy moth Peachtwig borer Peachtree borer Tufted apple bud moth Grape berry moth Omnivorous leafroller Leafrollers Other (aphids, mites) aObllquebanded,

Registered

Abbreviatron PBW CM TPW OFM GM PTwB PTB TABM GBM OLR LRa

by USEPA,

1997

Year of first registratron Registrations Companies 1978 1991 1982 1987 1983 1995 1995 1994 1990 1996 1997

9 5 5 3 2 2

1 1

4 4 3 3 2 2 1

1

1

1

1

1 1

1 2

2

blackheaded fireworm, pandemls

The potential of using sex pheromones to control insect pests was first demonstrated 30 yr ago (5). This and other early research (6-9) demonstrated that the release of large amounts of synthetic sex pheromone mto the atmosphere of a crop could interfere with mate locatton, thereby controlling the pest by delaymg or preventing matmg. Despite the successdemonstrated m these studtes, the first commerctal mating-disruption product registered m the Urnted States for the control of an insect pest was not until 1978, when the pheromone of the pmk bollworm (PBW) Pectinophora gossyplella (Saunders), was registered (10). Currently, there are over 30 mating-disruption products registered by the US Environmental Protection Agency (EPA) for control of more than 12 pests (Table 1). Several factors have enhanced the development and commerctal use of mating-disruption technology for the control of insects In the United States and elsewhere, regulations governing the registration and field application of conventional insecticides have become more restrictive. The cost and time to register these materials has increased substantially, and is viewed as a barrier to the development of new products (22). Stmtlar concerns associated with new data requtrements for reregistration have made many products economtcally marginal to their registrants, resulting m then removal from the market. For the manufacturers of mating-dtsruptton products, the registration process was also considered an impediment to the development of technology (12). The time and the costs to register these narrow-spectrum products could not be Justified by small market size.

Pheromones for Insect Control

387

The EPA has implemented many changes in the regulatory process to accelerate the development and registration of pheromone-based control technologies (13,111). The EPA established a toxicology-based, tiered testing requirement for pheromones and other biochemical products. Longer-term toxicology studies are required if adverse results are obtained m the initral acute tests. Further, the EPA exempted from the requirement of tolerance all inert Ingredients of pheromone products formulated m dispensers made of polymeric matrix materials. This exemption applied to dispensers large enough to be retrieved from the field, and enabled companies to contmuously improve their formulations without the need to seek EPA approval (14). In 1994, the EPA issued a general exemption from tolerance requirements for all arthropod-pheromone active ingredients, again in large, polymeric materials, and when applied at ~370 g active mgredient/ha/yr. In addition, the EPA increased the amount of area that could be treated wtth pheromone products without an Experimental Use Permit (EUP) from 4 to 100 ha. In 1995, the EPA expanded regulatory relief to include sprayable pheromone formulations, by issumg tolerance exemptions for all lepidopteran pheromones, and increasing the acreage cutoff to 100 ha for these formulations as well. In order to better manage the registration of pheromones and other biochemical technologies, the EPA established the Biopesticide and Pollution Preventton Division, Together, these changes have provided pheromone products with a distinct advantage in the registration process,and facilitated and accelerated then introduction into the marketplace. In 1993, the EPA published the Worker Protectlon

Standard for Agricultural

Pesticides (40 CFR, Part 170). Mandatory compliance with all the requn-ements became effective April 15, 1994. The new regulations were designed to protect farm workers and pesticide handlers from health hazards associated with pesticide exposure. The new regulations increased the restricted entry mtervals, enhanced training, required greater notification and posting, and provided for the establishment

of decontamination

sites. Compliance

with the new

regulations has made farm worker management more difficult and expensive m agricultural crops m which conventional insecticides are routinely applied. Fortunately, the EPA exempted pheromones from these regulations. As a result, m agricultural crops in which mating-disruption technology is used alone, or in conJunction with the limited use of insecticides, the management of agricultural workers is simphfied and less expensive, giving mating-disruption technology another advantage in the marketplace. It is well over 50 yr since the discovery and commerciallzatlon

of DDT. The

success of DDT and other insecticides in killing insect pests was followed closely by control problems associated with insecticide resistance (15). Currently, there are over 504 insect pests known to be resistant to msecticrdes, and the number continues to increase (16). Mating disruption, with its unique mode

388

Thomson, Gut, and Jenhrns

of action, may be able to slow or prevent the development of resistance by reducing exposure to insecticides. Resistance to pheromones has not been documented in the field, and, at least m some species, is unlikely, given the broad response to blend ratios (17). Therefore, mating-disruption products should have longer life expectancy, and may help preserve the dwmdlmg supply of effecttve conventional insecticides. Mating-disruption technology controls the target pest by manipulating certain aspects of sexual behavior (18). Pheromones are nontoxic to natural enemies. In contrast, conventional insecticides are generally broad-spectrum, killing both pest and nonpest species Greater reliance on pheromone technologies m agricultural crops will increase the potential for the biological control of secondary pests by allowmg for crop environments that can sustain high populations of predators and parasitoids. Enhanced biological control often corresponds to a reduction m the number of insecticide applications for control of secondary insects (I9-21), resulting in savmgs for the grower (22). The above discussion reviewed some of the factors that have favored the successof mating-disruption technology in the marketplace. The focus of the remainder of this chapter is to discuss how specific advances in critical areas have enhanced the development, implementation, and adoption of mating-disruption technologies m agricultural systems. It does not review the principles of mating disruption, nor does it attempt to discuss the many successful usesof this technology. These have been carefully reviewed in other publications (l&23-26). The followmg critical issueswill be discussed regarding the use of mating-disruption technology in two agricultural systems:biology and pest status; identification, formulation, and delivery system; and applied research, extension, and economics. We have selected two pest management systems, PBW m cotton and codlmg moth (CM) Cydiapomonella, m pome fruits, to illustrate how these critical issues have been addressed, and to discuss their overall tmpact on mating disruption product development and commercial adoption. PBW and CM represent two very different challenges for matmg-disruption technologies. Worldwide, mating disruption is used most extensively for the control of PBW, a serious pest of an annual row crop that is grown on large acreages. In 1996, tenders were awarded m Egypt to supply matmgdisruption products for approx 200,000 ha of cotton (24). The CM system provides a model for the successful development and adoption of this technology m horticultural crops, and provides an assessmentof the technology m a much different croppmg situation, with the added challenges of a large crop canopy, larger pest complexes, and lower damage thresholds because of high crop value. Mating-disruption technology is also extensively used to control CM in pome fruit in the United States (20,27), Italy (28), and South Africa (29). Worldwide in 1997, it is estimated that mating-disruption tech-

Pheromones for insect Control

389

nologies were used to control CM on between 24,000 and 28,000 ha of pome fruit (Thomson, personal communication). 2. Case Studies 2.1. Pink Bollworm 2.1 7. Biology and Pest Status The PBW is the most serious pest of cotton, Gossypzum hzrsutum L. and G. barbadense L, worldwide, including North and South America, Spain, Greece, Egypt, Pakistan, India, China, and Australia. In the desert southwestern United States, this pest infests approx 200,000 ha of cotton, mcludmg the highest yielding areas of Arizona and the Imperial Valley of Cahforma (30) In 1997, there were an additional approx 54,000 ha of cotton infested Just south of the United Statesborder m Mexico (R. Staten,personalcommunication) Throughout its range, PBW is usually the primary economic pest of cotton. Cotton bolls are the preferred site for ovtposition. Upon hatching, firstmstar larvae quickly enter squares or bolls. Injury is caused when larvae cut and stain fiber and feed in seeds wtthm the developmg cotton bolls. Larval damage also permits the development of decay from microorganisms. Infestations can lower quality of lint and seed, and yield reductions of 30% or higher are not uncommon (31), Quarantine programs, cultural practices (32), sterile moth releases (33,34), and conventional chemical msecticides have been used to manage PBW populations. Control with conventional msecticrdes 1sdifficult, because larvae are well protected within cotton squares and bolls. Insecticide applications are therefore normally targeted at nocturnally active adults. Insecticide costs for PBW control are high. Gonzales (35) reported that msectlcide costs during the period 1978-l 988 averaged $640/ha/yr. Furthermore the reliance on conventional chemical insecticides has led to outbreaks of secondary pests (36) and development of resistance (37). Because of the problems associated with traditional chemotherapy, PBW has been a prime candidate for development of alternative control methods. Development of mating-disruption technology for this pest began more than 25 yr ago (38). 2.1.2. Critical Issues in Identification, Formulation, and Delivery System PBW sex pheromone, gossyplure, was first identrfied by Hummel et al. (39) and Bier et al. (40). Compared to many insect pheromones, gossyplure (Z:Z and Z:E 7,11 hexadecadienyl acetate 5050) is relatively simple to synthesize, and is stable in the environment (41). Soon after its tdentification, work began on formulations and delivery systems (7,10,42). In 1978, the first matmg-disruption product was registered by EPA for PBW control (10).

390

Thomson, Gut, and Jenkins

The earliestresearchon PBW mating disruption was conductedwtth hand-applied devices (7,38). Commercial successcomcided with development of mechanically applied formulations. These mvolved mixture of slow-release dispensers,such as plastic microtube fibers and lammated flakes, with stickers. The mixtures were applied by specializedequipment attachedto aircraft or tractors (20,43,44). These formulattons had greater commercial acceptancecompared to hand-applied dispensers,especially among cotton growers in the southwesternUnited States. A series of studies led to the development of an attracticide formulation. Nocturnal observations of PBW moth mating behavior, m fields treated with mechanically applied formulations, demonstrated that males approach and contact the pheromone-laden dispensers. Furthermore, moth scalescould be found m the stickers surrounding the dispensers (45). These observations led to mcorporation of small amounts of insecticide m the stickers used to adhere the pheromone dispensers to crop foliage. Control was an outcome of males contacting the sticker and absorbing a lethal dose of msecticide (46,47). Butler and Las (48) demonstrated that apphcations of attracticide for PBW control did not adversely impact beneficial insects m the field. This attracticide approach is thought to be more robust than mating disruption alone, resulting m greater efficacy under higher population pressure (49). Even though some researchers have not seen significant advantage of this approach over classical mating dtsruption (50) use of attracticide has been adopted by growers, because it allows reduced rates of synthetic pheromone, as low as 100mg active ingredient (at)/ha/d, and lower costs. Presently, most commercial applications of PBW mating disruption m the Umted States use the attracttctde approach. In certain markets, hand-applied dispensersare acceptable or even preferable These include areas where labor is readily abundant and inexpenstve, where mechanical application equipment is not available, and in crops where proper dtspenser placement is difficult to achieve with broadcast application equipment. For PBW and other pests,long-lived pheromone dispensershave been developed that protect labile pheromone-active ingredients, and provide effective release for extended periods of field exposure (51-53). These formulations can be easily applied to plant foliage, and may provide season-longpopulation suppressionfrom a single apphcatton. Furthermore, they eliminate possible gaps in coverage durmg which mating may occur. This is a frequently observed problem with shorter-lived formulations. Season-long dispensersare loaded with higher rates of pheromone, typically 147-165 mg, and are applied at 250-500 dispensers/ha.Recent measurements, using field-portable electroantennograms of pheromone movement from high-load dispensers, indicate concentrations sufficient to affect PBW mating behavior may be transported by wind up to 100 m from the emitting source (54). These observations suggestpheromone dispenserscould provide effective mating disruption when placed at much more widely spacedintervals.

Pheromones for Insect Control

391

The major disadvantage of all types of mechanically or hand-apphed formulations IS the dlffculty of apphcation or the need for specialized mechanical equipment. The equipment required for mechanical applications 1snot avallable in most foreign markets, and hand-apphcation, although possible, is often not preferred. Hand labor may be scarceor expensive, and apphcation of sticky formulations can be undesirable, especially If insecticides are added. These drawbacks helped encourage the development of sprayable, mlcroencapsulated formulations for PBW mating disruption (55,.56). Commercial microencapsulated formulations are now avallable for several important insect pests m the United States. For PBW, microencapsulated formulations are applied at 5.0-12.5 g/ai/ha, and may last up to 14 d per appllcatlon, depending on temperature. They are most often used during periods of little rainfall. At lower rates, mlcroencapsulated formulations are often applied concurrently with standard rates of conventional msectlcides, as a biom-ltant strategy designed to increase pest exposure to the msectlclde (57). 2.1.3. Critical ksues in Implementation: Appljed Research, Extension, and Economics The many contributions to pheromone identification and formulation development discussed in the previous section have resulted in the establishment of mating-disruption technology as an effective and economical method for PBW management (18,31,33,58,59). One of the most successful demonstrations of the use of mating disruption for PBW control has been conducted by the Anzona Cotton Research and Protection Council (ACRPC). During 1990-l 996, ACRPC conducted a pheromone-based control program rn cotton planted m the Parker Valley along the Colorado River (60). Acreage ranged between 9575 ha and 11,430 ha. PBW control consisted of mating disruption or attracticide, and the judicious application of conventional insecticides. Control measures were applied based on intensive momtormg using pheromone-baited traps (1 trap/4 ha) and plant inspection. In 1989, m the year prior to the program, average boll infestation m the valley was more than 23% (Table 2). After the first year of the program, PBW-infested bolls were reduced by more than 50%. By the fourth year, 1993, no larvae were detected m more than 25,000 bolls inspected, at a cost to the grower of only $55.60/ha. PBW infestation increased slightly m 1994 and 1995, then m 1996, boll infestation increased substantially, to 2.63%. This caused concern among growers and program organizers. Success of PBW mating disruption is inversely density-dependent. Many studies illustrate the importance of early application while populations are still low (7,61-64). The decline m efficacy in the Parker Valley Project might be attributed to high population densities during 1996 (L. Antilla, personal com-

392

Thomson, Gut, and Jenkins

Table 2 Pink Bollworm Infestation and Control Cost in Parker Valley Pink Bollworm Pheromone Program Year

1989 1990 1991 1992 1993 1994 1995 1996

Bolls inspected

Larvae found

% Infestation

Cost/ha ($)

26,879 34,726 35,477 30,064 25,200 16,109 16,520 45,597

6282 3442 507 261 0 32 63 1200

23 35 9.91 1 42 0 86 0 00 0 20 0 38 2 63

17 20 1720 22 00 9 00 11 52 13 10 20 36

SourceArizonaCotton Research and Protection Couml munication). Higher-than-normal populations entered diapause at the end of 1995, and mild overwmtermg conditions contributed to larval survival. Furthermore, population pressure was aggravated m early 1996 by uncontrolled PBW infestations coming from volunteer cotton plants which had grown undetected m nearby wheat fields. As a result, population pressure was uncommonly htgh in 1996, and approx 10% of the cotton fields withm the program experienced infestation of at least lo%, despite repeated mating disruption and insecticide applications, at a cost of more than $120/ha. However, 90% of the fields m the program were relatively clean, and the average infestation of 2.63% should not indicate a failure of mating disruption to control the pest. Although the Parker Valley Program may be considered a successful example of mating disruption, it also illustrates the tenuous nature of areawide programs. Although mating disruption has been biologically and economically effective, the organized program did not contmue in 1997. Instead, growers reduced pheromone-protected acres and increased use of transgenic Bt-cotton. This change may be attributed to some of the problems experienced m the pheromone program during 1996, and to the relatively mexpensive and low-maintenance transgenlc cotton technology. Grower cost for Bt-cotton is approx $80/ha, significantly lower than the cost of PBW control to program growers m 1996. The use of transgemc cotton more than doubled m Arizona between 1996 and 1997. It 1sestimated that approx 144,000 ha of cotton were planted m Arizona during 1997, and that nearly 50% of this area was planted to transgenic Bt-cotton varieties. At present, transgenic cotton appears to be very effective in controllmg PBW infestations, and, compared to mating disruption, easier to implement. However, there is serious concern that widespread use of the technology will place extreme pressure on insects to develop resistance to

Pheromones for Insect Control

393

Bt toxin (65). Furthermore, research mdicates that the resistance factor 1san

incompletely dominant trait (A. C. Bartlett, personal communicatton). Unfortunately, the greater the degree of dominance, the less likely a high-dose strategy as presently employed is going to succeed (T. J. Dennehy, personal commumcation). Despite these concerns, the planting of trangenic cotton has reduced the use of PBW mating disruption in the desert southwest United States.Mating disruption is now limited to refugia fields, or, m a much smaller amount, to supplemental application on top of Bt-cotton fields. Termination of the Parker Valley Pheromone ProJectIS unfortunate, becausethe area-wide program employed many of the crtteria necessaryfor successful implementatron of mating dtsruptton 1 2 3. 4. 5. 6 7.

Early initiation at pinsquare stage cotton, while populatrons are relatively low Intensive monitoring with pheromone traps and frequent plant mspections, Use of economic thresholds; Careful adherence to effecttve rates and uniform pheromone dtstrrbution, Proper timing to avoid gaps between application; Reduction of immtgratton of mated females; and Judictous use of other control measures, including standard msecttcrdes.

Unfortunately, these criteria are often not used in less organized programs. There are now various methods available for management of PBW m the desert southwest United States. Mating disruption has been an effective and commercially viable method for controlling this pest for nearly 20 yr. Although recent mtroduction of transgenic cotton has decreased the use of mating disruption, in the future this technique could be combined with other methods,

such as sterile-male releases, short-season cotton and transgemc varieties to combat infestations, and manage resistance development to Bt-toxins. PBW mating disruption is likely to increase in other countries, where transgemc cotton is not yet available and PBW remains a serious pest. 2.2. Codling Mofh 2.2.1. Biology and Pest Status CM 1s a key pest of pome fruits in North and South America, South Africa, Australia, and Europe. Eggs are laid on the twigs or leaves adjacent to the fruit, or directly on the fruit. Upon hatching, larvae tunnel into the fruit. Failure to control thus pest can result m very high levels of crop loss, with at least 80% wormy fruit at harvest (66). CM is primarily controlled throughout the world by one or more applications of broad-spectrum, primarily organophosphorous, insecticides. These compounds, if used correctly, generally provide commercially acceptable levels of control, with damage to fruit kept below 1%. Unfortunately, organophosphorous compounds are highly toxic to natural enemies

394

Thomson, Gut, and Jenkins

of most pests, and are a maJor factor limiting the successof btological control m pome frutt. CM resistance to organophosphate msecticrdes has been detected m parts of the western United States(67,68), and in South Africa (M. Addison, personal communication) Up to 12 sprays per season are common m South Africa, and, despite this intensive program, some orchards suffer at least 30% fruit inJury at harvest (29) Increasmg difficulty m controllmg this pest has certainly provided a sense of urgency in research efforts to develop new CM-control technologies. In North America, considerable research has been directed toward the development of more selective controls for CM, mcludmg the insect-growth regulators (IGRs; 69-72). In Europe, IGRs have been used commercrally for years. They are generally considered to be nontoxic to natural enemies, and have become an important component of apple pest management programs (28). However, Riedl and Zelger (72) have reported high levels of resistance m some regions to one class of growth-regulating compounds: the chmn synthesis inhibitors. In addition to the potential for resistance, field trials conducted m North America and Europe have indicated that IGRs are not as efficacious as organophosphates in controlling CM (J. Brunner, personal commumcatton), Therefore, it is unlikely that IGRs will be an effective and reliable stand-alone replacement for organophosphates. Other tactics tested m North America, including CM granulosis vu-us (73) and mass release of sterile males (74,75), have been demonstrated to be relatively soft on beneficial insects and mites. However, these have either been too expensive, remam unregistered, or have not been effective enough to obtain widespread commercial acceptance m the United States. The use of sex pheromones for mating disruption has long shown promise as a control for CM (7679), but only recently has rt become widely adopted (2 7,28,80,81). The total area of pome fruit production around the world treated with various mating-disruption technologies has grown from an estimated 1500 ha m 1991 to at least 24,000 ha in 1997. The results have generally been good. However, control problems still occur that may be related to issues m pheromone identification, formulation, and delivery system Control problems have also been attributed to environmental conditions, high population densities, immigration, or improper use, and extensive research and education is still required to improve the level of control and enhance use of the technology 2.2.2. Critical Issues in Identification, Formulation, and Delivery System Lab and field research on mating disruption by mdustry and university SCIenttsts m Australia, Canada, Europe, Japan, and the United States over the past 20 yr has provided the foundation for the commercial development of pheromone technology for the control of CM (79). Roelofs et al (82) identified (E,E)-8,1O-

Pheromones

for lnsect

Control

395

dodecadien- l-01 (codlemone) as the major component of CM sex pheromone Evidence for secondary components m the CM pheromone was reported 10 yr later by Bartell and Bellas (83). Subsequent research supported this findmg, with dodecanol-1 and tetradecanol-1 identified as the important secondary components (84-86). Rothchild et al. (87) conducted research on a blend of codlemone, dodecanol, and tetradecanol for mating disruption of CM, and subsequently patented its use. On the basis of this new mformatton, a major manufacturer incorporated the three components, codlemone, dodecanol, and tetradecanol, in their formulation. The identification of secondary compounds, and their subsequent use m a commercial formulation, was important, because many of the early trials employmg dispensers loaded only with codlemone had demonstrated inconsistent and often poor results (79). It was a widely held assumption that mating disruption worked best when the complete blend was used (88). However, many products appeared on the market m both Europe and the United States with only codlemone in the formulation. To clarify the importance of secondary components, McDonough et al. (89) looked at the behavioral responses of CM males m a wind tunnel to sources emitting codlemone, or a mixture of codlemone, dodecanol, and tetradecanol, and were unable to show differences. In small orchard plots, McDonough and colleagues further demonstrated that Inhibition of male attraction to virgmfemale-baited pheromone traps was the same whether traps were m an environment containing elevated levels of codlemone alone, or codlemone plus dodecanol and tetradecanol (90). McDonough et al. (91) concluded that male CMs were sensitive only to the major component, E,ES-lo-dodecadlen-l-01. He suggested that control problems often seen m pheromone-treated orchards were probably related to the photodegradation of the pheromone m the dispenser, and not the result of an mcomplete blend. The weight of the scientific evidence indicates that dodecanol and tetradecanol are not of critical importance to the efficacy of CM mating disruption. However, their exact role remains unclear. There is still an active search for additional components in the CM pheromone (G. Gries, personal commumcanon). If additional components are eventually identified, perhaps their inclusion m commercial formulations will enhance control of CM. McDonough et al. (91) reported that, in small plots, dispensers loaded with an equilibrium blend of codlemone and its geometric isomers resulted m a higher level of disorientation of males to virgm-female traps than dispensers loaded with htghpurity codlemone. These findings have not yet been thoroughly tested m the field to determine their importance to the efficacy of CM mating disruptton. CM pheromone has been formulated and tested for efficacy m both broadcast and retrievable dispensers. Sprayable formulations have mcluded microcapsules (92) and chopped hollow fibers (77). Retrievable formulations have

396

Thomson, Gut, and Jenkins

included hollow fiber tapes (93), rubber tubing (94) laminate flakes (94), and polyethylene-tube dispensers (95). Smce 1991, the EPA has registered four dtspensing systems for the control of CM, mcluding polyethylene tubes, bagtype membranes, plastic coils, and laminate flakes. No broadcast formulations have been registered. Results from early research trials were often poor, and probably related to formulatton and delivery problems (79). Formulations and dehvery systems have improved, but control problems stall occur wtth currently regtstered commercial formulations. If codlemone is the sole component of CM pheromone, then control problems related to formulation and delivery system are probably caused by photochemical degradation (9697) and/or longevity of release from the dispenser (97). The impact of photochemical degradation on dispenser performance was investigated for the polyethylene tube formulation. McDonough et al. (96) reported that as much as 61% of the pheromone was lost through photochemically induced degradation, and only 39% through evaporation. Codlemone is a conjugated diene alcohol, and, like other similar chemicals, is prone to photochemical degradation via exposure to heat, light, and oxygen (98,99). Degradation occurs via isomerization of the double bond (ZOO),and oxidation to peroxides and furans (97,99) The amount of codlemone degradation in polyethylene-tube dispensers, first used commercially, effectively decreased the longevity and necessitated two applications of dispensers per season (96). Unfortunately, the cost of the dispensers precluded a second apphcation (22), and many attempted to get season-long control with a smgle application of dispensers. This approach often resulted in control problems late in the season. In response, new formulations of the polyethylene-tube dispenser have been developed, with substantially less photodegradation. The effective field life has increased by close to 50%, from 75-90 d to 120-140 d, depending on temperature (D. Thomson, personal communication). A field life of 140 d can provide season-long control m more temperate pome fruit-growing regions. Overall, the improved performance of polyethylene-tube dispensers has improved the efficacy of CM-mating disruption later in the season,resulting m a better rate of success,leading to enhanced adoption by growers. A related problem with current delivery systems has been inadequate field life, because of factors other than chemical degradation. Field life ts a function of the pheromone load and its rate of release from the dispenser. The loading rate is easily controlled m the lab; however, release rate in the field (assummg no photodegradation) is complex, and is affected by the physical and chemical characteristics of the dispenser, and environmental factors, such as temperature and wind. Ideally, dispensers should show a flat release rate at a constant temperature over the expected field life. However, release rates change strongly, relative to temperature, and weakly, relative to wmd velocity (M. Suckling, per-

Pheromones for insect Control

397

sonal communication). Thus, dispensers are likely to have shorter field lives m areas with high temperatures, or when conditions in a cooler region are unusually warm. This variation m release rates caused by temperature difference makes it difficult to determine the effective field life of a dispenser. Many commercial formulations have shown dramatic decreases m release rate following field exposure (101). The result has been inadequate amounts of pheromone released into the orchard later in the season. However, some manufacturers have improved their products, resultmg in substantial improvements in performance (101). Currently available commercial formulations employ either a single- or multiple-application strategy. These strategies are designed to ensure the adequate release of pheromone durmg the mating period of up to three CM generations. Control problems have occurred when dispensers have run out of pheromone earlier than expected, leavmg gaps when there is no pheromone dispensed, Dispenser manufacturers must carefully determine the expected field life of their products, to ensure proper use and performance m the field. 2.2.3. Critical Issues in lmplemen ta tion: Applied Research, Education, and Economics We believe CM matmg disruption has succeeded because the research, extension, and agricultural communitres have tried to mtegrate pheromones into pest management programs, rather than srmply adopt another technology for CM control. By implementmg a pheromone-based pest management strategy, many of the limitations to the efficacy and acceptance of pheromones imposed by the orchard-croppmg system have been mrtigated. These include highly vartable environmental conditions, a low tolerance for fruit damage, a complex CM mating system, and a diverse arthropod commurnty. Early on, it was recognized that the best opportunittes for CM control were m orchards where physical characteristics and environmental conditions, mcluding topography, size and shape, canopy structure, and wmd allowed for uniform distributton of pheromone. Relatively flat and even canopied sites have served as the primary candidates for CM mating disruption, sites with steep slopes or large numbers of missing trees have generally been avoided The borders of disrupted orchards have been especially vulnerable to CM (2478). Two processes have contributed to the development of border mfestattons. Mated females inumgrate from adjacent orchards that are not treated wtth pheromone. In addition, it is suspected that pheromone concentrations are lower on the borders than the interior, thus increasmg the likehhood of males locating females and mating on the borders. Growers have Judiciously protected the orchard perimeter by treating with insecticides, a 2x rate of pheromone, or a combination of the two. Border problems can also be remedted by

398

Thomson, Gut, and Jenkins

the area-wide application of mating-disruption technology. Recently, an areawide approach to control CM with mating disruption and other suitable technologies was mitiated m western North America. In 1995, the USDA, m associatton with cooperatmg universities m California, Oregon, and Washmgton, initiated the Codling Moth Area-wide Management Program (CAMP). The excellent results achieved at all CAMP sites has increased awareness about the benefits of the area-wide application of mating-dtsruption technology The CM mating system has posed some special challenges to achievmg control with mating-disruption technology. Adults mate shortly after emergence m the spring, and mating activity is concentrated m the upper-third of the canopy (102). Thus, both the timmg of application and the posittonmg of dispensers withm the canopy can dramatically affect the efficacy of mating disruption. For example, Weisslmg and Knight (103) demonstrated that significant levels of mating occurred m the upper-half of the tree, when dispensers were placed at a mid-canopy height of 1.8 m However, little or no mating occurred m the tree when dispensers were placed high m the canopy, at 3 6 m (trees were about 4.2 m tall). The best control of CM with mating disruption has been achieved when dispensers are placed high m the orchard canopy (0.6 m from the top of the tree). In the absence of an education or trammg program, dtspensers have frequently been placed too low in the canopy (approx 1 8 m) Commercially, it is the experience of the authors and others (28,29) that low dispenser placement has resulted m many CM control problems. Improvements m methods for applying CM disruption products has greatly facilitated the use of this technology. In orchards with canopy heights >3 m, proper placement of dispensers during the first few years of commercial use could not be accomplished from the ground. Applymg dispensers with the assistance of ladders was time-consuming (up to 12.5 h/ha), and added considerably to the already high cost of control. Superior Ag (Yaktma, WA) recently introduced a very good method for applying tube and lammate dispenser types, using a combmation of a telescoping pole and a clap to whtch the dispenser is attached. Some manufacturers have engineered then dtspensers with a clip already attached Apphcation entails pushmg a clip holdmg a dispenser onto a selected branch and leaving it there when the pole 1stwisted and pulled away. It takes <5 h to treat 1 ha of apples with this technique. The population density of CM prior to treatment has been ctted as a key factor determining the efficacy of mating disruptton (77,80,95). Commercially acceptable levels of control, m which pheromones are used without supplemental msecticides, have consistently been achieved only when mltial CM densities are very low. The strong interaction between density and efficacy has encouraged some crmcal modificattons m use patterns for this tactic Gut et al. (20) and Thomson (27) reported that, under condittons of low pest pressure, a

Pheromones

for Insect

Control

399

tube-type dispenser, applied at rates of 1000 dispensers/ha or 500 dispensers/ha, provided the same level of control. Growers have adopted the cost-savmg strategy of lowermg applicatron rates rf CM densities are known to be low. Moderate or high mitral pest densities has resulted in serious control failures (20,27-29). High population densities occur when and where the control of CM wtth conventional insecticides has not been effective because of resistance or poor application techniques, Concerns about using mating-disruption technology under condittons of high pest pressure have been mitigated by adopting a broadbased pest-management program. Many organic growers have implemented a program that includes pheromone, botanical msecticrdes, mineral 011,and samtation. An area-wide project mtttated in 1993 has used a combinatron of matmg-disruption technology and the judrctous use of msectrctdes to combat CM resistance to azmphosmethyl in northern Cahforma pear orchards (204). Matmg-dtsruptron technology has been supplemented with msecttcrdes only when extensive monitoring indicates it 1snecessary. As a result, the use of organophosphate msecticrdes decreased dramattcally. The impact of reduced exposure to azmphos-methyl to resistance IS stall being examined. A stmrlar approach has been the normal practice in Washington apple orchards, partrcularly during the first year of a mating-disruption program Essentially, earlyseason sprays, in conjunctron with mating-disruption technology, are used to ensure that CM densities are low and, thus, controllable. Even orchards with a history of poor control with msectrcrdesand, therefore, high populattons, have become primary targets for pheromone-based management. Evidence shows that a combination of mating-disruption technology and a few msecticrde treatments IS more efficacious than a conventional program of season-long application of insecttctdes. Over the past 2 yr, many orchards with known high populattons of CM have applied tube-type dispensers at 500 dispensers/ha instead of the label rate of 1000 dispensers/ha and then supplemented wrth l-3 applications of msecticrdes. To date, the results have been excellent. For the past 20 yr, monitoring CM with pheromone traps has been a standard management practice in orchards throughout the world. Trapping systems can be used to determine when to apply msectrcrdes, and whether populatton densities are htgh enough to warrant treatment (105). Monitoring CM adults has proved to be more difficult in orchards treated with mating-disruptron products. Failure to capture moths m a pheromone trap barted with a standard lure contammg 1 mg of codlemone has been found to be an unreliable mdtcator of successful drsruption (20). Charmillot (78) showed that the sensitivity of CM pheromone traps in disrupted orchards could be increased by usmg lures containing higher amounts of codlemone. Further research m the western United States has led to the adoption of a 10-mg red septa-batted pheromone trap as an important component of pheromone-based CM control programs Gut and

400

Thomson, Gut, and Jenkins

Brunner (20) determined that the effective use of this high-load trapping system required replacing lures every 3 wk during the first CM generation, and every 2 wk during subsequentgenerations. CM pheromone traps used m conventtonally managed orchards have typically been placed at midcanopy or lower, becausethis position has resulted m good moth capture, and they are easier to mamtam than traps placed higher in the canopy. However, when high-load lure-baited pheromone traps have been used in mating-disrupted orchards, they have been more sensitive when placed in the upper part of the canopy (106). Trapping programs have been most effective when they are used m conJunction with visual mspection of fruit for CM damage. Routme exammations of fruit in the upper canopy, along orchard borders, in susceptible varieties, on the tops of slopes, near prop or bm piles, and near fruit-packing operations has proved invaluable for early detection of CM fruit damage when a disruption program is faihng. The most important impediment to the commercial acceptance of pheromone-based integrated pest management (IPM) IS the higher costs relative to insecticide-based pest management. Thu-ty-rune people employed m the field of insect pest management m Washington and Cahfornra were surveyed for their opinions on the cost of CM mating-disruption technology relative to the cost of conventional insecticides (107). Fifty-one percent stated that they found the cost of matmg-drsruption technology to be high compared to conventional insecticides; 3 1 and 8% indicated that they found the cost to be very high and extremely high, respectively. Only 8% of the respondents indicated that the cost of CM mating disruption was reasonable. In small plot trials conducted m pear orchards over a 6-yr period m California, the average cost of a pheromone-based IPM program was 40% higher than a standard msecticide program (108). Two applications of mating disruption dispensers are required to control CM in California, thereby substantially increasing costs. A study conducted m Washington State found that the cost of a pheromone-based IPM program, when adjusted for material, labor, and machinery costs, was $133/ha higher than a conventional insecticide program (22). The disparity m these costs provides a strong dismcentive to adopt a pheromone-based IPM system approach Orchard systems are highly complex, with several hundred arthropod species having the potential to reach pest status (109,120). The successful deployment of CM mating disruption, and the subsequent reduction or ehmmation of insecticides to control CM, often creates orchard environments where some of these species can rapidly increase to economically damaging levels (20) In 1993, a 485ha apple orchard m central Washington State was successfully treated in its entirety with tube-type dispensers. Supplemental msecticides were only applied to border areas. Forty-eight sex pheromone traps captured over 6000 obliquebanded leafroller, Chorzstoneura rosaceana, males. In 1994, the orchard was again treated with tube-type dispensers and, again, supplemental

Pheromones for Insect Control

401

msecticides were only applied to local border areas. Trap captures of obhquebanded leafroller increased fourfold, to over 24,000 Increased applications of msecticides were required to prevent serious economic injury to the fruit. This example clearly illustrates how the successful use of CM matmg disruption can increase the control problems of certain secondary pests, and thus provide a dismcentive to the adoption of the technology. The costs associated with the increased use of insecticides for leafroller control, or the economic losses because of fruit injury can outweigh the benefits and savings derived from the biological control of other secondary pests. The aforementioned impediments to the adoption of a pheromone-based IPM systems approach present serious obstacles to the commercial successof mating-disruption technology for CM. Yet, despite these obstacles, CM mating disruption is mcreasmgly being adopted by apple and pear growers around the world, because of a combmation of good results and the increased difficulty of using broad-spectrum insecticides for the control of CM. Using CM mating disruption in a large, contiguous area IS considered a better strategy than m small, individual orchards. The initiation of the USDAsponsored CAMP has enhanced the status of mating-disruption technology as viable alternative to conventional insecticides, especially when pheromones are applied in an area-wide approach The objective of the program ISto reduce the use of organophosphate insecticides to control CM by 80% over a 5-yr period (111). The CAMP adopted an IPM approach utihzmg matmg-disruption technology, judicious and timely apphcations of insecticides, biological control, and sanitation. In 1995, five project sites were selected in Califorma, Oregon, and Washington. The size of each site and number of growers mvolved varied between locattons. The results have been quite promising durmg the first 2 yr of the project, with the number of insecticides applied for CM control and fruit injury at harvest declining substantially. For example, at the Howard Flat CAMP site in north central Washington, cullage because of CM fruit injury averaged 0.8% m the year prior to the project, and dropped to 0.01% after 3 yr of area-wide mating disruptton. Seventy percent of blocks showed no detectable damage. Organophosphate sprays targeted against CM at Howard Flat have decreased from an average of 2.8 to 0.8 for the season. The number of CAMP sites, and their size, has grown constderably, from five sites, totaling 1300 ha m 1995, to 10 sites, totaling 3970 ha in 1997. It is hoped that the successof the area-wide approach will encourage growers to work cooperatively, ensurmg the application of mating-disruption technology over wide areas to improve efficacy. 3. Future Needs and Trends Since the first isolation and identification of an insect pheromone more than 35 yr ago, pheromones have been used successfully to control several impor-

402

Thomson, Gut, and Jenkins

tant pest species. However, then use has not advanced as far as many have hoped or predicted. The reasons for this are varted and often debated. There has been considerable progress over the past decade m pheromone tsolatton and identtficatton, and formulattons and delivery systems.Accurate tdentiticatton, increased stability, and longevtty, in addition to more uniform release rates following field exposure, have enhanced the performance of dispensers. However, the mechanisms of how mating disruptton controls insects is not well understood m many msect/croppmg systems,even where the technology 1sused successfully. Therefore, the amount of pheromone needed per hectare, and the best method to deliver it, are seldom accurately known, As a result, tt IS dtfficult to design and commerctahze cost-effecttve technology, and most end users consider mating-disruption technology to be both risky and expensive. The development of the portable electroantennogram (112) has enabled instantaneous measurement of pheromone concentrattons wtthin crop canopies (213,114). Previous technology was only able to measure time-averaged concentrattons (115), whtch did not take mto account the instantaneous fluctuations m concentrattons or plume structures relative to changes m temperature or wmd speed However, the new sampling apparatus 1s complex, expensive, and dtffcult to use, restricting the number of samples that can be taken at a point in time. With limited samples, the dynamics of pheromone movement, relative to the physical structure of the canopy and environmental conditions, is dtfticult to elucidate. Advances m this area, in conJunctton with a better understanding of how pheromone concentrations within the canopy impact the flight and matmg behavior of the target pest, should enhance understanding of how the mating-disruption technique works. Armed wrth this mformatton, more efficacious technology should be designed and delivered tt to the end user m a more economtcal manner. The effectiveness of mating-disruption technologies has largely been determined by the biological limitations imposed by the pest and cropping system. It 1sapparent that pests vary m then susceptibility to pheromone-mediated control. Some species, such as PBW, tomato pmworm (116), and peach tree borer (117), seem to be highly sensitive to then pheromones, and control can be achteved even under relatively high population density. Other spectes, such as CM, appear to be moderately responstve to synthetic pheromone and more dtfficult to control by mating dtsruptton. In this group of species, there seems to be a strong mteraction between populatton density and efficacy. Mating dtsruptlon often can only be used successfully over the long term tf pertodtcally combined with one or more msectictde treatments, Currently, commerctal disruption products are most cost-effective m situattons m which conventtonal msecttcides are not performmg well, or then avatlability is restricted In these cases,the high cost of using mating disruption can

Pheromones

for Insect

Control

403

be Justified. On many farms, however, pests are easily controlled with one or two applicattons of fairly mexpenstve insecttctdes. These farms are less likely to switch to a more expensive control program until these msecttcides are lost, or the indirect benefits of using mating-disruption technology can be shown to outweigh the increase in costs High costs, difficulty of application, and the need to control several pest species have limited the acceptance of hand-applied dispensers. In response, sprayable formulations are now being tested or used commercially for a number of insect pests, including obltquebanded leafroller (J. Brunner, personal communication), oriental fruit moth, Cydia molesta (L. Gut, personal communication), and PBW (5556). Recently, new dispenser technology termed “puffers” (118) or “metered semiochemrcal timed release system (MSTRS)” (119) has been tested for the control of a number of insects, These devices are applied at very low rates of 2-5 dispensers/ha. If efficacious, these new technologies could greatly reduce the costs of application, and thereby enhance adoption of mating-disruption technology. The use of mating-disruption technology for control of key pests must be compatible with management strategies for the rest of the pest complex. Frequently, the use of mating-disruption technology, and the subsequent reduction m insecticides for a primary pest, results in outbreaks and necessitates control of secondary pests. Growers are keenly aware of the potential added costs. The pheromone industry must respond wrth technology that is capable of disrupting several species In addition, other control tactics must be developed that are selective and compattble with mating dtsruption. Efforts to develop and commerctalize mating-disruption products have largely focused on pests with a high market potential. This IS the same strategy that has effectively led to the development of most conventional insecticides. However, it may not be the best scenario for the pheromone mdustry. The chemical mdustry has largely been able to provide controls for minor pests by spinning off broad-spectrum chemistries that were primarily developed for control of economically important pestsin maJor crops. Generally, mating-disruptton products control only a single pest species, thus precluding the spin-off approach The uniqueness ofmating-disruption technologies is that the release device is ofparamount importance. Once an effective device is developed, it potentially can be used to control any target pest that IS amenable to disruptron. Manufacturers of dtsruptton products should consider developing and providing products for both maJor and minor pest species.This approach would be similar to one taken by another segment of the pheromone industry-manufacturers of lures for traps, which provide many products based on only a few release devices. The implementatton of pheromone-based pest-management programs IS often accompanied by reductions in the sales of conventional msectictdes The

404

Thomson, Gut, and Jenkins

sales of mating disruption products do not provide the same financial returns on investment as conventional insecticides, and may explain, m part, why so few large agrochemtcal compames have developed mating-disruption technology, and why many suppliers of agricultural chemicals have not actively promoted the use of mating-disruption technology Pheromones have been referred to as the chemical equivalent of parasites and predators Like beneficial organisms, pheromones are target-specific, exploit certain aspects of Insect behavior, and often result m reduced use of msecticides. In general, they have been developed by small companies, or by industries not normally associated with agricultural chemicals. Iromcally, a major impediment to the acceptance of pheromones may be their successful use in IPM programs. In summary, the adoption by growers of pheromone-based IPM programs ~111depend on how well the systems can meet grower concerns about efficacy and cost. A better understanding of how and why the technology works will enable the design of more cost-effective technology. The integration of mating-disruption technology into reliable and predictable pest-management programs will enhance adoption. Therefore, the development of momtormg and sampling techniques, in conjunction with economic thresholds, will be essential to accurately assessthe biological relationships between key and secondary insects and their natural enemies, to ensure cost-effective control of all pests in the system. Acknowledgments We thank the Arizona Cotton Research and Protection Council for permission to use their data from the Parker Valley Program. Thanks also to Tim Dennehy, Robert Staten, Larry Antilla, Alan Bartlett, and Hasan Bolkan for providmg mformation and making suggestions. References Butenandt, A , Beckman, R , Stamm, D., and Hecker, E (1959) Uber den Sexuallockstoff des Seidenspmner Bombyx marl, Retdarstellung und Konstitutton. Zeztschrlft Naturforschung B 14,283,284 Tamaki, Y. (1988) Pheromones of the lepidoptera, m CRC Handbook of Natural Pestzczdes, vol 4 Pheromones (Morgan, E. D. and Mandava, N. B , eds ), CRC, Boca Raton, FL, pp. 35-93. Am, H., Toth, M., and Priesner, E , eds. (1992) List of Sex Pheromones of Lepldoptera and Related Attractants, 2nd ed , International Orgamzation for Biological Control, Montfavet, France. Mayer, M. S and McLaughlm, J. R , eds. (1991) Handbook of Insect Pheromones and Attractants, CRC, Boca Raton, FL. Gaston, L. K , Shorey, H H., and Sarto, S A. (1967) Insect population control by the use of sexpheromonesto inhibit onentationbetweenthe sexesNature 213, 1155

Pheromones for insect Control

405

6 Mitchell, E R., Copeland, W W , Sparks, A N , and Sekul, A A (1974) Fall armyworm: disruptton of pheromone communicatton with synthetic acetates. Envtron Entomol 3,778-780 7 Shorey, H. H., Kaae, R. S., and Gaston, L. K. (1974) Sex pheromones of Leptdoptera Development of a method for pheromonal control of Pecttnophora gossyptella in cotton. J. Econ. Entomol. 67, 341. 8. Taschenberg, E F., Garde, R T., and Roelofs, W. L. (1974) Sex pheromone mass trapping and mating disruption for control of redbanded leafroller and grape berry moths in vineyards Environ Entomol 3, 239-242 9. Rothchtld, G H. L. (1975) Control oforiental fruit moth (Cydzamolestu [Busck] [Leptdoptera, Tortncldae]) with synthetic female pheromone. Bull. Ent. Res. 65,473-490. 10 Brooks, T W , Doane, C. C., and Staten, R. T (1979) Experience with the first commerctal pheromone communication disruptive for suppression of an agrtcultural pest, m Chemical Ecology Odour Communtcation zn Antmal (Ritter, F J , ed ), Elsevier/North Holland, Amsterdam, pp. 375-388 11. Ktrsch, P. K. (1988) Pheromones: then potential role m control of agricultural pests Am J. Alternattve Agrtculture 3,83-97. 12. Jellmek, S. D. and Gray, E. C. (1992) Opportumttes for expediting the US. regulatory process for pheromones and other semiochemicals, in Insect Pheromones and Other Behavtour-Modtfytng Chemicals (Ridgeway, R. L., Inscoe, M., and Am, H , eds.), BCPC Monograph 5 1, Surrey, UK, pp 116122 13. Hutton, P. 0. (1993) New regulatory strategies for pheromones. EPA pohctes, interpretations and rules, m Pest Management* Btologtcally Based Technologzes (Lumsden, R. D and Vaughn, J L., eds.), American Chemical Society, Washington, DC, pp. 388-390 14 Kirsch, P. K. (1997) Amazing daze. confusion brings change, m Technology Transfer znMating Dtsruption (Witzgall, P and Am, H , eds ), IOBC wprs Bulletin vol 20 (l), Montfavet, France, pp 249-256. 15 Georghtou, G. P. (1994) Principles of insecticide resistance management. Phytoprotectton 75,s l-59 16 Georghiou, G. P. and Lagunes-TeJeda, A. (199 1) The occurrence of resistance to pesticides m arthropods. U.N. Food and Agriculture Orgamzatton, Rome, AGPP/ MISC/91-1, p 318. 17 Haynes, K F and Baker, T. C (1988) Potential for evolutton of resistance to pheromones J Chem. Ecol 14,1547-1560. 18. Card& R. T. and Mmks, A K. (1995) Control of moth pests by mating dtsruption. successes and constraints. Ann Rev Entomol 40,559-585. 19 Van Steenwyk, R. A., Toscano, N C., Ballmer, G. R , Kido, K , and Reynolds, H T. (1975) Increase of Hellothis spp. in cotton under various insecticide treatment regimes. Envtron. Entomol 4,993-996. 20 Gut, L. J. and Brunner, J. F. (1994) Pheromone-mediated control of codlmg moth m Washington apple orchards Good Frutt Grower 45,35-48 2 1 Trumble, J. T. and Alvarado-Rodriguez, B. (1993) Development and economtc evaluation of an IPM program for fresh market tomato production m Mexico Agrtc Ecosyst. Envtron 43,267-284.

406

Thomson, Gut, and Jenkms

22 Wilhamson, E. R , Folwell, R J , Knight, A. L , and Howell, J F (1996) Economics of employmg pheromones for mating dtsruptton of the codlmg moth, Carpocapsa pomonella. Crop Protection 15,473-477 23 Bartell, R J (1982) Mechanisms of commumcatton dtsruption by pheromone m the control of Leptdoptera: a review. Physzol Entomol 7, 353-364 24 Garde, R T. and Mmks, A. K (1997) Insect Pheromone Research New Dzrectzons Chapman and Hall, New York. 25 Jutsum, A R. and Gordon, R. F S (1989) Insect Pheromones zn Plant Protectzon Wiley, Chtchester, UK 26 Wyatt, T (1997) Puttmg pheromones to work paths forward for dn-ect control, m Insect Pheromone Research New Dzrectzons (Card& R. T. and Minks, A K., eds ), Chapman and Hall, New York, pp. 445459 27 Thomson, D. (1997) Confusion amongst codling moth fellows contmues a commercial perspective on the Implementation of codlmg moth mating disruption m North Amertca, m Technology Transfer rn Matzng Dzsruptzon (Witzgall, P and Am, H , eds ), IOBC wprs Bulletin, vol 20 (l), Montfavet, France, pp 57-63 28 Waldner, W. (1997) Three years of large-scale control of codlmg moth by matmg disruption m the South Tyrol, Italy, in Technology Transfer zn Mating Dzsruptzon (Witzgall, P. and Am, H , eds.), IOBC wprs Bulletin, vol. 20 (l), Montfavet, France, pp. 35-44 29 Barnes, B N and Blometield, T L (1997) Goading growers towards matmg dtsruptton the South African experience wtth Grapholztha molesta and Cydza pomonella (Leptdoptera, Tortrtctdae), m Technology Transfer zn Mating Dzsruption (Wltzgall, P and Am, H , eds ), IOBC wprs Bulletm, vol 20 (1), Montfavet, France, pp 45-56 30 Wtlliams, M R (1996) Cotton Insect losses 1995 Proceedzngs, Beltwzde Cotton Production Research Conference, Nashville, TN, pp 670-689. 31 Doane, C C , Haworth, J K , and Dougherty, D G (1983) NoMate PBW, a synthetic pheromone formulatton for wide area control of the pmk bollworm, m Proceedzngs ofthe 10th Internatzonal Congress ofPlant Protection, Brighton, UK 32. Chu, C. C., Henneberry, T J , Weddle, R C., Natwtck, E., Carson, J., Valenzuela, C., Budsall, S., and Staten, R. (1996) Reduction of pmk bollworm (Lepldoptera Gelechndae) populattons m the Imperial Valley, Cahforma, followmg mandatory short-season cotton management systems. J Econ. Entomol 89, 175-182 33 Staten, R T , Anttlla, L , and Walters, M. L (1995) Pmk bollworm management prospects for the future, m Proceedzngs, Beltwzde Cotton Conferences, vol 1, San Antorno, TX, 4-7 January, pp 153-156 34 Walters-Geyer, M., Sequena, R. A., Paradice, D B., Fowler, G C., Miller, C , and Staten, R T (1994) Cotton pest management a knowledge-based system to handle information mput overload AI Applzcatzons 8, l-20 35. Gonzales, R. (1990) Compartson of pest control costs m Mexicah Valley and Imperial Valley, m Proceedzngs, Internatzonal Cotton Pest Work Commrttee, Mazatlan, Mexico, 8-9 November, pp. 13 1-138. 36 Werner, F. G., Moore, L., and Watson, T. F (1979) Arizona cotton Insects. Umverslty of Arizona Cooperattve Extension Service Bulletin, A23 R

Pheromones

for insect

Control

407

37. Burrows, T M., Sevachenan, V , Moffit, L J , and Bantelle, J L (1984) Economlcs of pest control alternatives for Imperial Valley cotton CA Agrzc 38, 15,16 38. McLaughlin, J R., Shorey, H H , Gaston, L K , Kaae, R S , and Stewart, F D (1972) Sex pheromone of Lepidoptera. XXX1 Disruption of sex pheromone communication m Pectmophora gossyplella with hexalure. Enwon Entomof 1,645-650 39. Hummel, H E., Gaston, L K , Shorey, H H., Kaae, R. S , Byrne, K J , and Silverstein, R. M. (1973) Clarification of the chemical status of the pmk bollworm sex pheromone. Sczence 181,873-875. 40. Blerl, B. A , Beroza, M., Staten, R. T., Sonnet, P. E., and Adler, V. E (1974) The pmk bollworm sex attractant. J Econ Entomol 67, 2 1l-2 16 41 Yamamoto, A, and Ogawa, K. (1989) Chemistry and commercial production of pheromones and other behavlour-modifying chemicals, m Insect Pheromones rn Plant Protection (Jutsum, A. R. and Gordon, R F. S., eds.), Wiley, ChIchester, UK, pp 128-148 42. Gaston, L. K , Kaae, R. S , Shorey, H H., and Sellers, D. (1977) Controllmg the pmk bollworm by disrupting sex pheromone communication between adult moths Science 196,904,905 43 Henneberry, T J , Gillespie, J. M , Barlola, L A., Flint, H. M., Lmgren, P D , and Kydomeus, A F (198 1) Gossyplure m lammated plastic formulations for mating disruption and pmk bollworm control J Econ Entomol 74,376-38 I 44. Kydomeus, A F., Glllesple, J. M , Barry, M W , Welch, J., Henneberry, T J , and Leonhardt, B. A. (198 1) Formulations and equipment for large volume pheromone applications by aircraft, in Insect Pheromone Technology Chemutry and Appllcatlons (Leonhardt, B. A and Beroza, M , eds ), ACS Symposium Series 190, 175-191 45 Miller, E., Staten, R T., Nowell, C., and Gourd, J (1990) Pmk bollworm (Lepldoptera. Gelechlidae). pomt source density and tts relationship to efficacy m attractlclde formulations of gossyplure. J Econ Entomol 83, 132 l-l 325 46. Haynes, K F. and Baker, T C (1985) Sublethal effects of permethrm on the chemical communication system of the pink bollworm moth, Pectznophora gossyplella. Arch insect Blochem Physlol. 2,283-293. 47 Haynes, K. F , LI, W -G., and Baker, T C. (1986) Control of the pink bollworm moth (Lepldoptera: Gelechiidae) with insecticides and pheromones (attractlclde) Lethal and sublethal effects. J Econ Entomol 79, 1466147 I 48 Butler, G D. and Las, A. S (1983) Predaceous insects: effect of adding permethrm to the sticker used m gossyplure applrcations. J Econ Entomol 76, 1448-145 1 49. Conlee, J. K and Staten, R T (198 1) US Patent 467 10 10 50 Beasley, C. A and Henneberry, T J. (1984) Combinmg gossyplure and msectltides m pmk bollworm control CA Agrzc 38,22-24 51. Flint, H. M , Merkle, J R., and Yamamoto, A. (1985) Pmk bollworm (Lepldoptera. Gelechllddae). field testmg a new polyethylene tube dispenser for gossyplure J Econ Entomol 78, 1431-1436

408

Thomson, Gut, and Jenkins

52 Staten, R. T , Flint, H M., Weddle, R C., Hemandez, E., and Yamamoto, A (1987) Pink bollworm (Lepldoptera: Gelechndae): large-scale field trials with a high-rate gossyplure formulation. J Econ Entomol 80, 1267-1271. 53. Jenkins, J W., Antllla, L., and Miller, E (1995) Mating disruption of the pmk bollworm with three commercial formulations Proceedzngs, Beltwzde Cotton Conferences, vol. 2, San Antonio, TX, 4-7 January, pp 989,990. 54 Farbert, P., Koch, U. T., FBrbert, A., Staten, R. T , and Card& R T. (1997) Pheromone concentrations measured with EAG m cotton fields treated for mating dlsruption of Pectinophora gossypzella (Lepldoptera Gelechlldae). Environ Entomol, 26, 1105-l 116. 55. Flint, H. M. and Merkle, J R. (1984) Studies on dlsruptlon of sexual communication m the pink bollworm, Pectmophora gossyprella (Saunders) (Lepldoptera Gelechiidae), with mlcroencapsulated gossyplure or Its component ZZ-Isomer Bull Entomol Res 74,25-32 56. Hall, D. R and Marrs, G J (1989) Mlcrocapsules, m Insect Pheromones in Plant Protectzon (Jutsum, A R. and Gordon, R F. S , eds.), Wiley, Chichester, UK, pp. 199-248 57 Weatherston, I (1990) Principles of design of controlled-release formulations, m Behavior-Modlfyuzg Chemicals for Insect Management Appllcatlons of Pheromones and Other Attractants (Rldgeway, R. L , Sllverstem, R M , and Inscoe, M N , eds.), Marcel Dekker, New York, pp 93-l 12 58. Baker, T C., Staten, R T , and Flint, H. M. (1990) Use of pmk bollworm pheromone m the southwestern United States, m Behavior-ModlfLmg Chemicals for Insect Management Applications of Pheromones and Other Attractants (Ridgeway, R L , Silverstein, R. M., and Inscoe, M N , eds ), Marcel Dekker, New York, pp 417436. 59 Camplon, D G , Cntchley, B R., and McVelgh, L J (1989) Mating dlsruptlon, m Insect Pheromones in Plant Protection (Jutsum, A R. and Gordon, R. F S., eds.), Wiley, Chichester, UK, pp. 89-119. 60 Antllla, L., Whitlow, M., Staten, R. T., El Lissy, O., and Myers, F. (1996) An integrated approach to areawide pink bollworm management m Arizona, in Proceedings, Beltwlde Cotton Conferences, Nashville, TN, pp 1083-1085 61 Doane, C C., and Brooks, T. W. (1981) Research and development of pheromones for Insect control with emphasis on the pmk bollworm, m Management of Insect Pests with Semlochemrcals (Mitchell, E. R , ed.), Plenum, New York, pp 285-303 62. Cntchley, B. R , Campion, D. G., McVelgh, L. J., Hunter-Jones, P., Hall, D. R., Cork, A , et al (1983) Control of pink bollworm, Pectlnophora gossyplella (Saunders) (Lepldoptera. Gelechildae), m Egypt by mating disruption using an aerially applied microencapsulated pheromone formulation Bull Ent Res 73, 289-299. 63. Stone, N. D , Gutlerrez, A. P , Getz, W. M , and Norgaard, R (1986) Pink bollworm control m southwestern desert cotton. III. Strategies for control: an economic slmulatlon study. Hllgardia 54,42-56.

Pheromones

for Insect

Control

409

64. Staten, R. T., Miller, E., Grunnet, M., and Andress, E. (1988) The use of pheromones for pmk bollworm management m cotton, m Proceedzngs, Beltwzde Cocton Production Research Conference, New Orleans, LA, pp, 206-209. 65. Tabashnik, B. E. (1997) Seeking the root of insect resistance to transgemc plants Proc Natl. Acad Sci. USA 94,3488-3490.

66. Myburgh, A. C. (1980) Infestation potenttal of the codlmg moth Dectduous Frutt Grower

10,368-377.

67. Varela, L. G., Welter, S. C , Jones, V. P., Brunner, J. F., and Rtedl, H. (1993) Monitoring and characterization of insecticide resistance in codling moth (Lepidoptera: Tortrtcidae) m four western states J, Econ Entomol 86, l-10. 68. Knight, A. K , Brunner, J F., and Alston, D. (1994) Survey of azmphosmethyl reststance m codlmg moth m Washington and Utah. J Econ Entomol 87,285-292 69. Butts, E. C. (1983) Effectiveness of a soft-pesticide program on pear pests. J Econ Entomol

76,936941

70. Westigard, P. H (1979) Codlmg moth: Control on pears with diflubenzuron and effects on non-target pest and beneficial spectes. J Econ. Entomol. 72,552-554. 71. Westigard, P. H , Gut, L J., and Liss, W. J. (1986) Selective control program for the pear pest complex m southern Oregon. J. Econ Entomol 79,25O-257 72 Rledl, H. and Zelger, R ( 1994) Erste Ergebmsse der Untersuchungen zur Reslstenz des Apfelwtcklers gegentber Diflubenzuron. Obstbau Wembau 31, 107-109 73 Westigard, P. H and Hoyt, S. C. (1988) Codlmg moth (Lepidoptera Olethreutidae). Evaluatton of the granulosts vnus for control in Pacific Northwest apple and pear orchards Melanderta 46, 14-l 8 74. Proverbs, M. D., Newton, J. R., and Logan, D. M. (1966) Orchard assessment of the sterile male technique for control of codlmg moth, Carpocapsa pomonella (L.) (Lepldoptera* Olethreutidae) Can Entomol. 98, 90-95. 75 Dyck, V A., Graham, S. H., and Bloem, K. (1992) Implementation ofthe sterile Insect release program to eradicate the codling moth, Cydtapomonella (L ) (Lepidoptera: Tortrmtdae), in Brittsh Columbta Symposium. Management of Insect Pests Nuclear and Related Molecular and Genettc Techniques. International Atomic Energy Agency, Vienna, p. 23. 76 Rothschild, G. H. L. (1982) Suppression of matmg m codling moths with synthetic sex pheromone and other compounds, m Insect Suppresston wzth Controlled Release Pheromone Systems, vol. 2 (Kydomeus, A. F and Beroza, M , eds ), CRC, Boca Raton, FL, pp. 117-l 34. 77. Moffitt, H. R. and Westigard, P. H. (1984) Suppression of the codling moth (Lepidoptera. Tortrrcidae) population on pear m southern Oregon through matmg disruption with sex pheromone. J, Econ Entomol 77, 15 13-15 19. 78. Charmillot, P. J. (1990) Mating disruption technique to control codlmg moth m western Switzerland, m Behavior-Modtjjytng Chemtcals for Pest Management (Ridgeway, R. L , Silverstem, R. M , and Inscoe, M N., eds.), Marcel Dekker, New York, pp. 165-182. 79. Vickers, R. A. and Rothchild, G. H. L. (1991) Use of sex pheromones for control of codlmg moth, m World Crop Pests, vol 5, Tortrtctd Moths Then Btofogy,

410

Thomson, Gut, and Jenkrns

Natural Enemzes and Control (Van der Geest, L P S , ed ), Elsevter, Amsterdam, pp 339-370. 80 Barnes, M M., Mlllar, J. G , Ktrsch, P A., and Hawks, D. C (1992) Codlmg moth (Leptdoptera. Tortrtcidae) control by dissemination of synthettc female sex pheromone. J Econ Entomol 85,1274-1277 81. Knight, A. K (1995) The impact of codling moth (Leptdoptera. Tortrtctdae) mating dtsruption on apple pest management m Yaktma Valley, Washington J Entomol Sot Br Columbia 92,29-38. 82 Roelofs, W. A , Comeau, A , Hill, A., and Mtlicevic,

G (1971) Sex attractant of the codlmg moth. charactertzatton with electroantennogram technique Sczence

174,297-299 83 Bat-tell, R. J and Bellas, T E ( 198 1) Evidence for naturally occurrmg secondary compounds of the codlmg moth female sex pheromone. J Aus Ent Sot 20,

197-199. 84 Emhorn, J , Wttzgall, P , Audemard, M., Bomface, B., and Causse, R. (1986)

85

86.

87

88 89

90

91.

92

93

Constttuants secondaries de la pheromone sexuelle du carpocapsae des pommes, Cydzapomonella L (Leptdoptera Tortrtcidae) II Premiere approche des effets comportementaux C R Acad Set Parts 302,263-266 Am, H , Guerm, P M., Buser, H. R , Rauscher, S., and Mam, E (1985) Sex pheromone blend of the codling moth, Cydza pomonella+ evidence for a behavioral role of dodecan-l-01. Experzentza 41, 1482-1484 Bat-tell, R. J , Bellas, T E., and Whtttle, C. P. (1988) Evidence for blologlcal activity of two further alcohols m the sex pheromone of female Cydzapomonella (L ) (Lepidoptera. Tortrictdae) J Aust Entomol Sot 27, 11,12 Rothchild, G. H L , Whtttle, C P , Bellas, T E , and Vtckers, R A (1988) Improved pheromone formulatton. International Patent World Intellectual Property Organization. International Publication Number, W/88/03755 Minks, A. K and Card&, R T (1988) Disruptton of pheromone commumcatton m moths: is the natural blend really most efficacious? Entomol Exp Appl 49,25-36 McDonough, L H , Davis, H G , Chapman, P. S , and Smtthhtsler, C (1993) Response of male codlmg moth (Cydrapomonellu) to components of conspecificfemale sex pheromone glands m flight tunnels J Chem Ecol 19,1737-1747 McDonough, L H., Davts, H G., Chapman, P. S , and Smithhtsler, C (1994) Codling moth (Cydia pomonella) Dtsruptants of sex pheromonal commumcanon. J Chem. Ecol 20, 171-18 1 McDonough, L H., Davis, H. G., Chapman, P S., and Smithhtsler, C (1995) Codlmg moth (Cydza pomonellu), (Leptdoptera. Tortrtctdae) Is tts sex pheromone multicomponent? J Chem Ecol 21, 1065-1071 Moffit, H R (1974) Confusion with a sex pheromone as a means of controllmg codlmg moth IOBC Working Party Meeting on Btological Control of Codlmg Moth and Adoxophyses, Wadenswtl, 26-29 November Mam, E., Am, H., and Wtldbolz, T (1978). Inhtbitton of moth catch by evaporating the pheromone of the codlmg moth. Mztteilungen der Bzololschen Bundesanstalt fur Land-und Fortwirtschafi (Berlin-Dahlem) 180. 50-52.

Pheromones

for insect

Control

411

94 Charmillot, P. J and Bloesch, B (1987) La technique de confusion sexuelle un moyen specifique de lutte contre le carpocapsae Cydza pomonella L Revue Suzssede !&culture, Aborlculture et Horticulture 19, 129-138 95 Howell, J. F., Knight, A. L., Unruh, T. R , Brown, D. F , Krysan, J. L , Sell, C R , and Ktrsch, P A. (1992) Control of codlmg moth m apple and pear wrth sex pheromone-mediated mating disruption. J Econ Entomol 85,9 18-925 96. McDonough, L H , Aller, W. C., and Knight, A K. (1992) Performance characteristics of a commercial controlled release dispenser of sex pheromone for codlmg moth control by mating disruption. J Chem. Ecol. 18,2 177-2 189 97 Millar, J G. (1995) Degradation and stabilization of E8,ElO-dodecadienol, the major component of the sex pheromone of the codling moth (Lepldoptera. Tortrtcidae) J Econ Entomol 88, 1425-1432 98 Ideses, R., Sham, A , and Klug, J T (1982) Cychc peroxide. an isolable mtermediate m singlet oxidation of pheromones to the furan system Chem Ind 12, 409,410 99 Ideses, R and Sham, A (1988) Chemical protectron of pheromones containing an internal conjugated drene system from isomerization and oxidation J Chem Ecol 14, 1657-1669 100. Brown, D. F , Knight, A. K , Howell, J. F., Sell, C R., Krysan, J L., and Weiss, M (1992) Emission charactertsttcs of a polyethylene pheromone dispenser for mating disruption of codlmg moth (Lepidoptera: Tortricidae) J Econ Entomol 85,910-917. 101 Millar, J G , McElfresh, J S., and Rice, R E. (1997) Technological problems associated wrth use of insect pheromones m insect management, m Technology Transfer znMuting Dwuptzon (Wttzgall, P and Am, H , eds.), IOBC wprs Bulletin Vol. 20 (1), Montfavet, France, pp 19-26 102 Howell, J F. (1990) Reproductive biology, m World Crop Pests, vol 5 Tortrzcid Moth Their Biology Natural Enemies and Control (Van der Geest, L P S , ed ), Elsevrer, Amsterdam, pp 157-l 74. 103. Wetssling, T. J. and Knight, A. K. (1995) Oviposition and callmg behavior of codling moth m the presence of codlemone. Ann. Entomol Sot Am. 89, 142-147 104. Welter, S (1995) Areawide management, msecttcide resrstance, codlmg moth pheromone mating disruptron. Research Project for California Bartlett Pears, California Pear Advisory Board, Sacramento, CA 105 Riedl, H., Croft, B. A , and Howitt, A J. (1976) Forecasting codlmg moth phenology based on pheromone trap catches and physlologlcal time models Can Entomol. 108,449-460. 106 Barrett, B A. (1995) Effect of synthetic pheromone permeatton on captures of male codling moth (Leptdoptera: Tortrictdae) m pheromone and vngm female moth-baited traps at different tree heights in small orchard blocks Envzron. Entomol 24, 1201-l 206 107 Weddle, Hansen, and Associates (1993) Barriers to the implementation of matmg disruption for the control of codling moth ResearchReports for Cahforma

Bartlett Pears California Pear Advisory/Pear Pest Management Research Fund

412

Thomson, Gut, and Jenkins

108 Weddle, Hansen, and Associates (1994) Management of codlmg moth m Bartlett Pears m Caltforma* A prelimmary analysis of the relative costs of insecttctdeand pheromone-based IPM strategies. Research Reports for California Bartlett Pears California Pear Advisory/Pear Pest Management Research Fund. 109 Meszaros, Z. (1984) Results of faunistical and flortsttcal studies in Hungarian apple orchards (Apple ecosystem research no 26). Acta Phytopathol Acad Scl Hungar 19,91-176 110 Gut, L J , LISS, W. J , and Westtgard, P. H (199 1) Arthropod community organization and development in pear. Environ Manage. 15,83-l 04. 111 Codlmg Moth Areawtde Management Program (1995) Pilot test implementation and research progress reports, Year l-1995 Oregon State University, Corvallis, OR, pp l-36 112. Sauer, A E., Karg, G., and Koch, U T. (1990) Measurement ofpheromone concentrations for the improvement of matmg disruption. OILB-SROP/IOBCWPRS Congress* Pheromones in Mediterranean Pest Management. Granada, Spain, 1O-l 5 September 113. Karg, G., Suckhng, D M., and Rumbo, E. R (1992) Measurement of sex pheromone m orchard air using electroantennogram apparatus, in Proceedzngs ofthe 45th New Zealand Plant Protectroh Conference, pp 304-306 114 Karg, G and Sauer, A. E. (1995) Spatial dtstrtbutton of pheromone m vineyards treated for mating disruption of the grape vme moth Lobesra botrana measured with electroantennograms. J. Chem. Ecol. 21, 1299-1314. 115. Otahshi, M., Uchqima, Z., and Yamamoto, A. (1991) Relationship between aerial concentratton of a synthetic sex pheromone and matmg m the dtsruptron field of the smaller tea tortrtx, Adoxophyes sp. (Lepidoptera: Tortrtctdae). Jpn J Appl. Ent Zoo1 35,207-2 11. 116 Jenkms, J. W , Doane, C. C , Schuster, D. J., McLaughlin, J. R., and Jtmenez, M. J. (1990) Development and commercial application of sex pheromone for control of the tomato pmworm, in Behavzor-ModifLmg Chemicals for Insect Management Applications of Pheromones and Other Attractants (Rtdgeway, R. L., Silverstem, R M., and Inscoe, M. N., eds.), Marcel Dekker, New York, pp. 269-280. 117 McLaughlin, J R , Doolittle, R E , Gentry, C R , Mitchell, E R., and Tumlmson, J. H. (1976) Response to pheromone traps and disruption of pheromone communication m the lesser Peachtree borer and the Peachtree borer (Leptdoptera Sesiidae) J Chem Ecol 2,73-81 118 Shorey, H and Gerber, R. G (1996) Use of puffers for drsruptton of sex pheromone commumcatton among navel orangeworm moths (Lepidoptera* Pyrahdae) m almonds, ptstachtos, and walnuts. Environ Entomol 25, 11541157 119 Baker, T. C , Mafra-Neto, A , Dtttl, T., and Rice, M. E. (1997) A novel controlled-release device for dtsruptmg sex pheromone communication m moths, in Technology Transfer rn Matmg Duruptlon (Witzgall, P. and Am, H., eds.), IOBC wprs Bulletin, vol. 20 (l), Montfavet, France, pp 141-l 50

REGISTRATIONOF BIOPESTICIDES

22 The Federal Registration Process and Requirements for the United States J. Thomas McClintock 1. Introduction The US Envtronmental Protection Agency (EPA) regulates pestictdes under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) and the Federal Food, Drug, and Cosmetic Act (FFDCA). Under FIFRA, the EPA is authorized to regulate pesticides to ensure that their use does not cause unreasonable adverse effects to humans and the environment. Under FFDCA, EPA has the responsibility to establish tolerances for pesticide residues on food crops. A tolerance is the maximum allowable residue of a pesticide on food. Such regulatory oversight is designed to minimize risks while allowmg the public to benefit from pesticide use. Prior to lawful use in commerce, a pestttide must be registered by the EPA, unless specifically exempted by regulation Registrants of pesticides are responsible for submittmg specific data to the agency to support the conclusion that the pesticide ~111not significantly mcrease the risk of adverse effects to humans or to the environment. Once a pesticide is registered by the EPA, it may be sold and distributed m the United States and used as specified on the approved label. Under FIFRA, a pesticide is broadly defined as “any substance or mixture of substances intended for preventing, destroying, repelling, or mitigatmg any pest, or intended for use as a plant growth regulator, defoliant, or desiccant” (40 CFR 152.15). Two broad classesof pesticides are generally recognized by the EPA* conventional chemical pesticides and biological pesticides. Chemical pesticides Include synthetic compounds, such as the synthetic pyrethroids, carbamates, and organophosphorus ester msecticides, herbicides, rodenttctdes, fungicides, and fumigants. Biologtcal pesticides or bropesticrdes From Methods III Biotechnology, vol 5 Bopestrodes Use and Dehery Edted by F R Hall and J J Menn Q Humana Press Inc , Totowa, NJ

415

416

Mcc/intock

are divided mto three categories: mlcroblal pesticides, blochemlcal pesticides, and transgemc plant pesticides. Microbial pesticides contam microorganisms (e.g., bacteria, vu-uses, fungi, protozoa, or algae) as the active Ingredient. The active microbial ingredient may be naturally occurring or may be altered via manipulation of the genetic material. Blochemlcal pesticides contam pheromones, hormones, natural insect or plant growth regulators, repellents, and enzymes as the active pestlcidal Ingredient. Transgenic plant pesticides, which may come under the regulatory oversight of the Office of Pesticides Program (OPP), can be defined as plants genetically altered via mtroductlon of genetic material for the purpose of imparting or increasing the production of a pesticide. The active pestlcldal ingredient could be considered the pestlcldal substance produced from, or modified by, the introduced genetic material. During the past few years, there has been renewed interest m the use of blologlcal pesticides as effective pest control agents, primarily because of the development of resistance to conventional chemical pesticides, and because of adverse human health and environmental effects This interest has been reflected m a significant increase m the number of applications for expenmental use permits (EUPs) and registrations submitted by registrants for blologlcal pesticides. Because of the unique characterlstlcs of biological pesticides, it IS generally recognized, and supported by data submitted by registrants and data presented in the open scientific hterature, that these pesticides present lower overall risk than most conventional chemical pestlcldes. Since blologlcal pesticides are typically found m nature, and are usually effective at low concentrations, the resulting exposure to humans and the environment IS extremely low. Consequently, because of the nature of these biological pesticides, the EPA currently provides some incentives for registration and commerclallzatlon of such active pestlcldal ingredients and products. The objective of this chapter is to discuss the current registration process of biopesticides by the EPA, the kind of data and mformatlon appropriate for the evaluation of human health and environmental risks associated with the wldespread use and distribution of biopesticides, and the existing mechanisms and incentives that encourage the development and registration of reduced-risk pesticides. 2. Data Requirements for Microbial Pesticides The generic and product-specific data requirements for mlcroblal pesticides appears m Title 40, Part 158, of the Code of Federal Regulations (CFR) Such data requirements or tests to be completed to support an EUP or registration of a microbial pesticide are determined based on the proposed use pattern. A complete description of all data requirements and study protocols for mlcroblal

Federal

Registration

Process

417

pesticides ISpresented m the Pesticide AssessmentGuidelines, Subdivision M: Guidelines for Testing Biorational Pesticides (1). The information and/or data required by the EPA for experimental use, or for registration of microbial pesticides, includes a thorough taxonomic characterization of the active microbial ingredient, as well as a description of the manufacturing (growth) process, including those measures taken to munmize the presence of contammating organisms. Newly prepared batches or lots of manufactured microbial pesticides are required to be screened for the presence of certain types of human pathogens, and, with Bacillus thurzngzenszs (Bt) products, by subcutaneous injection of newly prepared batches into rodents (as described m 40 CFR 180.1011). In addition, the potential for toxicity of protem components is determined by testing the active microbial ingredient, together with fermentation medium, m laboratory animals and nontarget orgamsms. 2.1. Product Identity/Analysis The product identity/analysis requirement for a microbial pesticide requires submtssion of detailed informatron on the identity and characterization of the active and inert ingredients, a description of the manufacturing process, mcludmg any unintentional ingredients formed, and, if appropriate, specification of the analytical method used. The product analysis requirement should include data and/or information to identify, to the extent possible, the taxonomic position, serotype, composition, and strain of the microorganism, and the unique nature and composition of the active microbial ingredient. The data/information should mclude the origin of the microorgamsm, the relationship to other species, unusual characteristrcs, biological activity and properties, and, if the microorganism is genetically altered, the method of alteration, as well as the identity of the inserted or deleted genetic material, and regulatory regions or sequences, if known. This information should also include a description of the phenotypic traits gained or lost, and the genetic stability of the altered genetic region. In addition, there are certain microorgamsms that are not readily amenable to adequate characterization from standard taxonomic procedures, either because they cannot be grown in pure culture (i.e., contammating microorganisms are impossible to remove), they can only be grown m assoctation with a particular host organism, or the system of taxonomy used is based on morphological characteristics and the microorganism under consideration has few to no unique morphological structures. Therefore, because historical experience often is lacking on adverse effects that might occur when humans are exposed to high numbers of environmentally isolated microorganisms, the agency requires a battery of acute pathogenicity/ toxicity studies m laboratory animals.

A&C/in tack

418 2.2. Description

of Manufacturing

Process

The taxonomic data and the acute mammalian toxtcity studies provtde mformatton useful m assessing toxicity of protein components of the active mtcrobtal ingredient, but tt is mformatton on the manufacturing process that addresses the ltkelthood of toxrctty that might occur from the presence of contaminating organisms. Particular attentton 1sgiven to the measures that pesttctde manufacturers use to mmimize the potential for growth of contaminating organisms A descriptton of the quality control procedures used to ensure a uniform or standardized product should m&de analyses of stock and seed cultures for biological purity, descrtptton of sterrlrzatlon procedures of growth media and of fermentation vessels; monitoring of appropriate physical condtttons during fermentation, and analysis of lots when fermentation is completed. The agency requests that the pesticide manufacturer present this mformation, because it provides a framework for a dtscussron on the likelthood of the presence of toxic or sensitizing materials artsmg from growth of contaminatmg microorganisms m the pesticide product. If the standardtzation technique(s) includes methods of a bioassay, then these methods should be described. EPA 1sparticularly interested in the ingredients that may be toxic or sensittzmg to humans and other nontarget organisms. This mformatton allows for a conclusion whether the microorgamsm is a recogmzed human pathogen. Each newly produced batch of microbtal pesticide can be analyzed for certam human pathogens (e.g., Shigella, Salmonella, and Vzbrzo), and for unexpected toxins vta injection into laboratory animals. If the production method can support growth of human or animal pathogens, then each production batch should be tested for then presence. The appllcattons also should stateproposed methodologies for detecting these pathogens, and/or then elimination from the productron batch, tf not discarded. For Bt fermentation batches, each lot 1stested “by subcutaneous mjectlon of at least 1 million spores mto each of five laboratory test mice.” The test results should show “no evidence of infection or injury m the test animals when observed for 7 d followmg mlection” (40 CFR 180.1011). For reregistration of active Bt ingredients, an mtraperttoneal mjectton screen 1srequired m whtch mice are injected with 106, 107, and lo8 units of Bt. As with the subcutaneous test, animals are observed for toxicity and mortality for 1 wk (2). The agency currently is reevaluating whether or not the subcutaneous injection test should be replaced with an mtraperitoneal mjectron test. 2.3. Toxicity Testing of Microbial Pesticides in Laboratory Animals The data and information obtained from the product characterization can be used to establish the mammalian toxicology data necessary to determine the

Federal

Registration

Table 1 Mammalian Toxicology for Microbial Pesticides

Process

419

Data Requirements (40 CFR 158.740)

Kurd of data required Tier I studies (acute toxtctty studies) Acute oral toxtctty/pathogenicity (rat) Acute dermal toxtctty (rat/mouse) Acute pulmonary toxicny/pathogemctty (rat/mouse) Acute mjectton toxtctty/pathogemclty (rat/mouse) Primary eye irritation (rabbit) Reporting of hypersensitivity incidents Cell culture tests with viral pest control agents Tier II studies Acute toxicity Subchronic toxicity/pathogenicrty Tier III studies Reproductive and fertility effects Oncogenicity Immunodefictency Primate mfecttvtty/pathogenlclty

Guidelme ref. no a 152-10 152-11 152-12 152-13 152-14 152-15 152-16

(885.3050) (885.3100) (885 3150) (885.3200) (885.3400) (885 3500)

152-20 (885.3550) 152-21 (885 3600) 152-30 (885.3650) 152-3 1 152-32 152-33

“Revised guideline numbers are listed in parentheses risks associated with human and domesttc ammal exposure. The current mammalian toxicology data requirements are structured m a tiered testing system, to provide focus only on those studtes considered necessary for an adequate human health risk evaluation (Table 1). Studies that are usually required in Tier I for registration of a mtcrobral pesticide for use on a terrestrial food crop mclude acute toxicity tests (oral, pulmonary, and tv injection [or mtraperttoneal for larger microorganisms]) and mammalian cell-culture studies. After dosing, test animals are evaluated by determining mortality, body weight gain, making cagesrde observattons for clinical signs of toxictty, performing a gross necropsy, and by evaluating the pattern of clearance of the mtcroorgamsm from the animals. For the latter end point, the mtcroorgamsm 1s periodtcally enumerated from appropriate organs, ttssues, and body fluids of test animals, to ensure lack of pathogemctty/mfecttvity or persistence m mammals. These studres would also be required at the EUP stage, if the treated food crop is not to be destroyed. The information from these acute toxictty studies allows an assessment for the potential of the microorgamsm to be pathogenic to, or toxic to, mammals. In most cases, lack of adverse effects allows for the reasonable concluston that the protein components of the mtcroorganism are not toxic to mammals. How-

McC/if?tock ever, if toxicity IS observed m the test animals, m the absence of signs of pathogenicity, then the toxic components m the test material are to be identified, and, to the extent practical, Isolated. Further testing m laboratory animals with the toxic components usually will be required to provide an estimation of the amount of material needed to elicit toxic or lethal effects. The potential toxicity of proteins m the growth or fermentation medium can be evaluated by mcludmg the growth/fermentation materials m the dosmg material for the acute oral, pulmonary, or intraperitoneal studies. It is important to enumerate the number of microbial units (e.g., colony-forming units, plaque-forming units [PFU]) in the dosing material. It is often inappropriate to include significant amounts of fermentation ingredients when dosing rodents via the intravenous route, since lethality from nonspecific toxicity may occur. For example, particulates m the fermentation material may result in mechamcal blockage of capillaries. On some occasions, nonspecific toxictty may result from reaction to tnjection of significant amounts of foreign protein into the bloodstream. Also, it should be expected that intravenous injection of large numbers of Gram-negative bacteria would cause acute mortality from shock reaction to the lipopolysaccharide (endotoxin) component of cell wall material Hypersensitivity (i.e., dermal sensitization) studies are generally waived as a requirement for registration of microbial pesticide products, since mjectton mduction and challenge with foreign protemaceous components of microbial pesticides into the commonly used laboratory ammal (i.e., guinea pig) would be expected to yield a positive response. Conversely, topical mduction and challenge with the active microbial ingredient would most likely lead to a negative response. This, coupled with the historical experience with fermentation products, has allowed for the concluston that reporting of observed allergic responses to microbial pesticides during manufacture and use should be adequate to address the potential for risk. However, registrants must submit to EPA any information/data on incidents of hypersensitivity, including immediate-type and delayed-type reactions of humans or domestic animals that occur during the production or testing of the technical grade of the active ingredient, the manufacturing-use product, or the end-use product. For reporting of mcidents occurring after registration, refer to the requirements in connection with section 6(a)(2) of FIFRA. Cell-culture tests are required to support the registration of products whose active ingredient is a virus (e.g., baculovirus). These studies provide information on the ability of these viral agents to infect, replicate in, transform, or cause toxicity in mammalian cell lines. Using the most infectious form or preparation of the vnus that gives optimal infection in a susceptible insect cellculture or insect (if a cell line is not available), human or mammalian cell lines are challenged and observed daily for appearance of cytopathic or cytotoxic

Federal

Registration

Process

421

effects, and the inability of the virus to infect or replicate in the host cell Cytopathic effects include such end pomts as morphologrcal or biochemical, and include, but not limited to, cell growth, attachment, morphology, nucleus size and shape, and cellular processes,such as macromolecular synthesis. Toxtcity evaluation focuses on the ability of the virus to inflict injury or damage to host cells when infection by, and/or replication of, the vu-us are not necessarily required. Toxicity can also be the ability of nonviral components of a preparation to inflict injury or damage to the host cell(s). Prior to viral challenge, the inoculum should be tttered by the most sensitive assay available, When a plaque assay for the virus 1savailable, a minimum of five PFUs/cell is required If a plaque assay is unavailable, a tissue-culture infectious dose (TCID&ell can be used when the TCIDSO value results m cytopathic effects, or seven times the LD,, unit, from the permissive Insect host system. For each series of tests, the vu-al inoculum should be tested in the permissive cell line or host organism as a positive control, and for direct reference to the data obtained from the vertebrate cell lmes. Current protocols for these studies are found in Subdivision M: Guidelmes for Testing Microbial and Biochemical Pest Control Agents (I). Additional mformation and procedures describing assaysof insect virus for toxic effects m mammalian cells are described elsewhere (3). For genetrcally engineered microorganrsms, the health-risks assessmentprocess is not sigmticantly different than for nonengmeered, naturally occurrmg microorganisms. The only addmonal data/information required for a recombinant microorganism includes a description of procedures and/or methods used to modify the microbe, mcluding source and descrtption of vectors; a description of the identity and location of the rearranged, inserted, or deleted gene segment(s); a description of the gene(s) and the regulatory control region(s), mcludmg new traits or characteristics that are expressed; information on the potential for genetic transfer and exchange with other microorganisms, and the genetic stability of any inserted sequence; and the competitiveness of the recombinant microbe, compared to the parental wild-type strain. 2.4. Non target Organism Data Requirements The data and information required to assesshazards to nontarget organisms are derived from tests to determine pesticidal effects on birds, mammals, fish, terrestrial and aquatic mvertebrates, and plants. These tests include short-term acute, subacute, reproduction, simulated, and/or actual field studies arranged in a tier system that progresses from the basic laboratory tests to the applied field tests (Table 2). For genetically altered microorgamsms, the toxtcity of the pesticidal substance produced, or modified, as a result of the genetic insertion, would be required, as well as the fate and effect of the inserted genetic

422

McClintock

Table 2 Nontarget Organism Data Requirements for Microbial Pesticides (40 CFR 158.740) Kmd of data required Tier I studies Avian oral toxicrtylpathogemcity (bobwhtte quail/mallard duck) Avtan respiratory pathogemcity (bobwhite quail/mallard duck) Wild mammal toxrcity/pathogemcny Freshwater fish testing (rainbow trout) Freshwater aquatic invertebrate Estuarme and marme animal test Nontarget plant studies Nontarget insect testing Honeybee testing Tier II studies Terrestrial environmental testing Freshwater environmental testing Marme or estuarme environmental expression Tier III studies Terrestrial wildlife and aquatic organism testing Chronic avian pathogenicity and reproductton test Aquatic invertebrate range testing Fish life cycle studies Aquatic ecosystem test Nontarget plant studies Tier IV studies Simulated and actual field tests (birds and mammals) Simulated and actual field tests (aquatic organisms) Simulated and actual field tests (insect predators and parasites) Simulated and actual field tests (insect pollinators)

Guideline ref. no a 154-16 (885 4050) 154-17 (885 4100) 154-18 154-19 154-20 154-21 154-22 154-23 154-24

(885.4150) (885.4200) (885 4240) (885.4280) (885 4300) (885.4340) (885.4380)

155-18 155-19 155-20

54-25 54-26 (885.4600) 54-27 (885 4650) 54-28 (885 4700) 54-29 (885 4750) 54-3 1 154-33 154-34 154-35 154-36

ORevisedguideline numbers are listed in parentheses

and the resultmg recombmant to nontarget organisms and the envlronment. A purpose common to all data requirements IS to prowde data that determine the need for precautionary label statements to mmunlze the poten-

material

tial adverse effects to nontarget organisms.

Federal

Registration

Process

423

In the acute toxlclty/pathogeniclty test currently required, avian wlldhfe are exposed through diet or the respiratory tract. The avian acute oral toxiclty/pathogenicity study provides data on any toxic effects to avian wlldhfe from exposure to the naturally occurring microorganism, or any toxins that may be produced from a genetically modified mlcroblal construct. This test would also provide data on pathogenic effects followmg an acute exposure by either the oral or injection route. The avian respiratory pathogeniclty test provides information on the pathogenic effects of the active mlcroblal mgredient on birds following exposure caused by drifts or aerosolation. The duration of both the avlan acute oral and respiratory studies should be about 30 d, to permit time for Incubation, infection, and manifestation of effects m the test organism. In both the acute oral toxicity/pathogemcity and respiratory pathogemcity tests, the test animals are evaluated by noting mortality, changes m behavior, pathogenic or toxic effects, gross necropsy, and hlstopathologlcal examination, including culture and isolation of exposure-site tissues and other organs showing anatomical or physiological abnormalities. If no toxic or pathogenic effects are observed after exposure via oral and respiratory routes, then no further testing in birds 1srequired. If effects are observed, then Tier II, envlronmental expressions tests, would be required. Data on wild mammal toxicity/pathogemcity are required on a case-by-case basis when data mdlcate that there is conslderable variation in the sensitivity of different mammalian species to the effects of a microbial-based insecticide, or when wild mammals would be expected to be exposed to a high dose under normal use. However, the toxiclty/pathogenicity submltted to evaluate hazards to humans are normally adequate to indicate potential hazards to wild mammals. If no toxicity/pathogemcity effects are observed in these tests, no further testing of wild mammals would be required. For mlcroblal pesticides applied m terrestrial-use patterns, when direct aquatlc exposure is not antlclpated, one freshwater fish and one freshwater aquatic invertebrate should be tested to assesstoxicity and pathogeniclty. These tests should be conducted as 30-d (for fish) or 21-d (for aquatic invertebrate) static renewal bioassays, m which the microbial inoculum is administered as a suspension m water, m diet in the form of diseased host insects or treated feed, or as a combination of both exposure routes. These tests should be designed to assessboth toxlclty and pathogemclty, as well as to detect and quantify the mlcroorgamsm m the test animal. Indlvldual test ammals are also removed periodlcally durmg the course of the study, and upon termmatlon of the study, to assesspathogenicity. If mortality is observed during the course of the studles, the cause of death (toxicity, pathogeniclty) should be determined, and, If possible, reisolatlon of the mlcrobe from test organism tissues.

McClintock Assessment of potential risk to nontarget insects from the use of naturally occurring and/or recombinant microorganisms is a primary environmental concern. For recombinant microbes, several issues or concerns need to be considered prior to field trials and wtdespread commercial use. Such issues include modification of host range, stability and persistence of the mtcrobial construct m the environment, which could increase its potential for uncontrolled spread, and the potential for genetic exchange, particularly, of the foreign insecticidal gene with other naturally occurrmg mtcrobes. Despite the factors cited above, the nontarget organism tier-testing scheme is adequate to address some of these issues and concerns The tter-testing scheme IS based on a fairly extensive first tier, which assessestoxicity and pathogenicity of the microbe to the honeybee, and to three species of predaceous and parasitic insects. Selection of the predator/parasitic species should be representative of groups that will be exposed under the conditions of proposed use, and which have some relationship to the target pest. Tier I testing also mcludes toxictty/pathogemcity testmg with Daphnza, or another aquatic insect species, or both, depending on use pattern. Data derived from the Tier I tests are used in conjunctron with available information on use patterns, specificity of host range, fate, and other factors, to assess potential for adverse effects. If the results indicate no adverse effects, no further testing is required. By contrast, if toxicity or pathogenic effects are observed, then Tier II testing, envu-onmental expression, would be required. It should also be noted that the best routes of exposure in the Tter I tests will depend on the developmental stage and locatton of the nontarget msect. 3. Data Requirements for Biochemical Pesticides Biochemical pesticides are distinguished from conventional chemical pesticides by their natural occurrence and nontoxic mode of action to the target pest. In contrast, conventtonal chemical pesttcides usually control the target pest(s) by a toxic mode of action, and are broadly toxic to other organisms. Many biochemtcal pesticides are often limited to a narrow range of target species, and are effective at low application rates. However, low use volume and target-species specificity are not criteria that are used m determining whether a substance is appropriately defined as a biochemical pesticide. Likewise, the definition of a btochemical pesticide does not presume a lack of mammahan toxicity, although many of the substancescurrently classified as btochemical pesttctdesare known to be, or have been shown to be, of low toxicity to mammals 3.1. Classification of Active Ingredients as Biochemical Pesticides In cases in which an active pesticidal ingredient is isolated from a natural source, and 1sknown to be nontoxtc to the target pest, the classtfication as a

Federal

Registration

Process

425

biochemical pesticide is obvious. For example, insect pheromones and plantgrowth regulators, such as auxins, gibberellins, and cytokinins, are by detimtion biochemical pesticides. Likewise, active pesticide ingredients that are recognized as common food sources or components, such as garlic and cmnamon, are also by definition biochemical pesticides. Plant-extracted materials, although of natural origin, are not necessarily always pesticidal by a nontoxic mode of action. For example, pyrethrins mitigate target pestsvia a toxic mechanism of action, Some plant extracts (e.g., capsaicm from red pepper) are repellents for certain pests, perhaps becauseof the irritating characteristics of the extract. Pestcontrol resulting from irritation has thus far been considered equivalent to a nontoxic mode of action for this type of plant-derived substance.Control of a pest by simple suffocation (e.g., by vegetable oil) also would be considered equivalent to a nontoxic mechanism of activity. Antibiotics from microorganisms, if used as pesticides, would not be considered biochemical pesticides, because, by defimtion, these substancesact via a toxic mode of action to the target pest. Although natural occurrence is a criterion for classification as a biochemical pesticide, a number of active biochemical ingredients have been chemically synthesized. If synthesized,then the active ingredient must be structurally semilar to, and functionally identical to, a naturally occurring counterpart. For example, the active ingredient mdole-3-butyric acid is classified as a biochemical pesticide, smce the synthetic plant-growth regulator is a structural analog of indole acetic acid (auxin), and also mimics the function of the natural plant hormone. In some instances, the synthesis of a biochemical pesticide or a structural analog, rather than isolation from naturally occurrmg material, may be preferred, because sufficient quantities of the material can be generated economrcally and in a more highly purified form (e.g., P-farnesene, an aphid alarm pheromone), and may yield products with increased efficacy and longevtty m the environment (e.g., modified forms of the neem seed extract, azadirachtin) In some cases,the precise mode of action of an active pestictdal ingredient against a target pest may not be readily apparent, and, consequently, the determination of a nontoxic mode of activity cannot be precisely elucidated. In these cases, the best available scientific information and knowledge are applied to make the most appropriate decision on the candidate material. It is possible to conclude that a pesticidal substance is best classified as a btochemical pesticide, even though the precise mode of action against the target pest is not known. The active pesticidal ingredients that have thus far been classified as biochemical pesticrdes and registered by the EPA are listed in Table 3. 3.2. General Guidance for Classification If an active pesticidal ingredient meets the criteria for classification as a btochemical pesticide, then the registrant can request that the agency make such a

McClin tuck

426 Table 3 Biochemical

Active Ingredients Target pest(s)

Chemical name

Pheromones Dodecenyl acetates, aldehydes, alcohols, and isomers Isomers of trlmethyl dodecatrlene Hexadecanyl acetates, aldehydes, alcohols, and isomers (R,Z)-5-( I-Decenyl)dlhydro2-(3)-furanone (Z,E)-7,9,11 -Dodeceatrien- 1-yl formate Octadecadlenyl acetates Periplanone B Trldecenyl acetates, aldehydes, and isomers Tetradecenyl acetate and alcohols (Z)-9-Tncosene (E)-5-Decenol (E)-5-Decenyl acetate Grandlure Musculure Czs-7,X-epoxy-2-methyloctadecane (dlsparlure) Methyl cyclohexenone

Grape berry moth, western pme shoot borer, codlmg moth, orlental frtut moth Tetranychld mite, aphids Pink bollworm, artichoke plume moth Japanese beetle Carob moth Peachtree borer American cockroach Tomato pmworm, tobacco budworm, cotton bollworm Grape berry moth, tufted apple bud moth Housefly Peach twig borer Peach twig borer Cotton boll weevil Housefly Gypsy moth Pine bark beetle

Plant growth regulators Streptomyces

fermentation byproduct

Various ornamental plants and food crops

N-6-Benzyladenme Natural plant extracts contammg Glbberellms, zeatins, IAA Cytokmm (6-furfural(ammo)purme) Ethylene Glbberellins and salts Indole-3-butync acid Keatme 5-Nltrogualacolate Ortho-mtrophenol Para-nitrophenol Pelargomc acid (contrnued)

Federal

Registration

427

Process

Table 3 (continued) Target pest(s)

Chemical name

Ornamentals, apples Fruits, vegetables, gram crops Sprout mhibitoi Potato sprout inhibitor Fruit ripenmg, storage stability of fruits, vegetables, cut greens/ flowers Herbicide

Aminoethoxyvmylglycine Biozyme Dnsopropylnaphthalene 1,4-Dimethylnaphthalene Lysophosphatidylethanolamme

Acetic acid Floral lures/attractants/repellents 4-Allylanisole Benzenepropanol Capsaicin Castor oil Cedar leaf oil Cedarwood oil Cmnamaldehyde

Southern pme bark beetle Corn rootworm Insects, dogs, birds Moles Rodents Fleas and moths Corn rootworm, spotted cucumber beetle Roaches Corn rootworm, spotted cucumber beetle Rabbits, dogs

Cmnamon Cmnamyl alcohol Dried blood Eucalyptus oil Eugenol(2-methyl-4-(2-propenyl) Garhc Indole Isoborneol Lemmongrass oil Marigold extracts Meat meal Methyl anthramlate 4-Methoxybenzenethanol 4-Methyl cmnamaldehyde 4-Methyl phenethyl alcohol 1-Octene-3-01 Oil of citronella Oil of geramum (geramol) Oil of pennyroyal

phenol)

Japanese beetle Birds Corn rootworm, spotted cucumber beetle Ants Moths Flying and stinging insects Deer, rabbits, raccoons, birds Birds Corn rootworm Corn rootworm, spotted cucumber beetle Corn rootworm, spotted cucumber beetle Mosquitoes, biting flies Mosquitoes, ticks Japanese beetle Japanese beetle

McClintock

428 Table 3 (continued)

Target pest(s)

Chemtcal name 3-Phenyl propanol Piperonal Putrescent whole egg solids Red pepper Rhodmol Terpmeol 1,2,4 Trimethyoxybenzene Natural insect regulators Azadirachtm Dihydroazadnachtin Tetrahydroazadtrachtin Trimethyl-dodecadienoates Hydroprene Kinoprene Methoprene Trypsm modulatmg oostatic factor Nematicides Fermentation solid and solubles of Myrothecium verrucarla Fungicides Clartfied hydropholic extract of neem oil Neem oil Phosphonic acid Volatile sulfur components Other Calcium sulfate Cellulose gum (sodmm carboxymethylcellulose) Potassmm bicarbonate Sodium bicarbonate JoJoba 011 Vegetable (soybean) oil Lactic acid

Corn rootworm, spotted cucumber beetle Head lice Big game animals Deer, rabbits, raccoons, birds Not specified Insects Corn rootworm, spotted cucumber beetle Insects Insects Insects Roaches Whiteflies, aphids, scales, gnats Mosquitoes, hornflies Mosquitoes, sand fleas, houseflies, fleas Nematodes

Fungi and insects Fungi Phytophtora, pythlum, and downy mildew Fungi Fleas Insects, mites Not hsted Not listed Silver leaf whitefly Insects and mttes Antimicrobial

determination. A formal request, containing the basic mformatton that supports the claim of natural occurrence and nontoxic mode of action to the target pest, can be submitted to Biopesticides and Pollution Prevention Division (BPPD).

Federal

Regis tra t/on Process

429

The final decision for or against classification as a biochemical pesticide is then conveyed back to the petitioner through BPPD. If warranted, a registrant can contact BPPD directly for preliminary guidance on classificatton issues. The kmd of information and data essential for classification of active mgredients as biochemical pesticides include documentation by citation to, and submission of, references from the published literature that support the natural occurrence of the substance.If the active ingredient is chemically synthesized, then the molecular structure of the substance, and its structural relationship to a naturally occurring substance, should be submitted, along with a brief description of the manufacturmg process. If the active ingredient is extracted as a mixture of substances from biological material(s), a description of the manufacturmg process should include the nature of the source substance(s) to be extracted, extraction materials, any subsequent purification process and materials used, and a characterization of the extracted substance(s), using appropriate analytical methods. The registrant should also submit a list of pests to be controlled, and any available information that would allow for the reasonable conclusion that control of the target pest(s) is achieved by a nontoxic mode of action. 3.3. Classes/Uses of Biochemical from Regulation Under NFRA

Pesticides

Exempted

Under specified conditions of use, the agency has determined that pheromones and certain plant-growth regulators have been exempted from all provisions of FIFRA. As stated m Part 40 of the CFR, 152.25 (b), Subpart B (July 1, 199l), pheromones, and identical or “substantially similar” compounds, produced by arthropods, and used only m traps, are exempt from regulation, as long as the substance traps individuals of the same arthropod species and achieves pest control solely by removal of the target pests from the environment via attraction to the trap. The pheromone trap also cannot result in increased levels of pheromones or identical compounds over a sigmficant portion of the treated area. For the purposes of this exemption, “substantially simrlar” means that “the only differences between the molecular structure are between the stereochemical isomer ratios of the ...compounds” (40 CFR 152.25 (b) [2]). The EPA, however, may determine that certain synthetic substances used in traps may possessmany characteristics of a pheromone, and thus meet the criterion of a substantially similar compound. A subset of natural plant-growth regulators, delineated as “vitamm-hormone” products, are also exempt from regulation under FIFRA. These horticultural products, which consist of a mixture of plant hormones, plant nutrients, moculants, or soil amendments, must be nontoxic (to humans/mammals) m the undiluted package concentration at which they are distributed and sold, and

430

MCClhtOCh

not used on food-crop sites. Finally, products considered as “foods,” which attract pests, but do not contain actrve pestictdal mgredtents, also are exempt from regulatron by EPA under FIFRA. 3.4. Data Requirements for Biochemical Pesticides Because of the unique characteristics of brochemical pesticides, OPP recognized that appropriate, and, m some instances, reduced data requtrements were Justified, to adequately evaluate the safety of these pest control agents The advantage to registrants of having an acttve mgredtent classttied as a btochemlcal pestrcrde, vs as a conventional chemical pesticide, resides in the reduced data requirements for the former group (drscussron under Subheading 3.6. below, for human health effects). Part 158 of 40 CFR specifies the kmd of data and mformatton appropriate for the evaluatton of human health and envnonmental risks associated with the widespread use and dtstribution of btochemtcal pesttctdes. The fundamental mformatton necessary to evaluate such products for human health hazards includes product analysts mformatron and data on the toxicity of the active ingredient to laboratory mammals The key mformatton is summarrzed below; however, for a complete descrtptton of all data requtrements and study protocols for blochemtcal pesttctdes, refer to the Pesttctde Assessment Guidelines, Subdivision M: Guldelmes for Testing Btorattonal Pesttcldes (4) 3.5. Product Identity/Analysis Data Requirements The product tdenttty/analysrs data for btochemrcal pesticides closely parallel those for conventtonal chemical pesttctdes. Subdrviston M solicits detailed mformatton by whtch the active ingredient is produced, and the techniques used to ensure a uniform or standardized product. Product tdentity/analysts mformatton encompasses three general areas: product Identity and compostnon, analysis and certified limits, and physical and chemical characterrsttcs Product identity data and mformatton are used to determme whether an active ingredient IS “identtcal or substantially simtlar” to another active Ingredient, or a naturally occurring substance Data on product composmon include both the active Ingredient and any intentionally added inert materials. Each product to be registered must be analyzed for the upper and lower concentrattons (certlfied hmtts) for both the acttve Ingredient and any mtentionally added inert substance. In addition to composmon of the final or end-use product, the product charactertzatton data mcludes a descrrptron of starting materials, production and formulatton process, and a drscussron of the posstble formatron of impurities Data on physical and chemtcal charactertstrcs of the pestrcrdal active mgredient and end-use products include, when appropriate, mformatton on then

Federal

Registration

431

Process

Table 4 Mammalian Toxicology Data Requirements for Biochemical Pesticides (40 CFR 158.890) Guidelme ref no n

Kind of data required Tter I studies (acute toxtcity studies) Acute oral (rat) Acute dermal (rat/mouse) Acute mhalation (rat/mouse) Primary eye irritation (rabbtt) Primary dermal nritation (rabbmgumea pig) Dermal sensmzatron Hypersenstttvity incidents Genotoxicity studies Ames assay Forward gene mutation assay In vivo cytogenetrcs assay Subchromc studies Immunotoxicity (1 spp) 90-day feeding, dermal, inhalation (1 spp) Developmental toxictty (1 spp) Tier II studies Immune response (rodent) Tier III studies Chronic exposure (rodent) Oncogemcity (rodent)

81-1 81-2 81-3 81-4 81-5 81-6 84-2

(880 3550) 82- 1, 82-3, 82-4 83-3

83-l 83-2

“Revised guldelme numbers are listed m parentheses

color, odor, physical state, stabtlity, oxidizing and reducing potential, storage stability, and corrostveness. 3.6. Mammalian

Toxicology

Data Requirements

The current mammalian toxtcology data requirements are set forth m 40 CFR 158.690, and are listed m Table 4. Specific guidance on methods and procedures for conduct of these studies 1sdescribed in Subdtviston M of the Pesttctde Testmg Guidelmes (4). The toxicology data requirements are structured m a tiered testing system to provide focus only on those studies constdered necessary for an adequate health-risk evaluation, Studies that are usually required m Tier I for registration of a biochemical pesticide for use on a terrestrial food crop include acute toxicity tests (oral, dermal, and mhalatton), a primary eye and a dermal trrttation study, a battery of genotoxicity studies, an tmmunotoxtctty study, a 90-d feeding study, a teratogenicity (developmental toxicity)

432

McClintoch

study, and reporting of hypersensitivity mcidents. These studies would also be required at the EUP stage, if the treated food is not to be destroyed. Specific conditions, qualtfications, or exceptions to the designated tests are provided m Part 158.690 (a)( 1). For example, the acute oral and dermal toxicity study would not be required if the test material is a gas, or is sufficiently volatile so as to render performance of a test impractical If the test material is corrosive to skin, then the acute dermal toxicity study and the primary eye and dermal trritation studies would not be required. A dermal sensitization study is required at registration, if there IS repeated contact with human skin under the conditions of use. Although no specific tests are requtred, all incidents of hypersensitivity must be reported to the agency immediately followmg then occurrence. However, the requirement for allergemc incident reports, and specific lack thereof, could be suggested as a basrs for requestmg a waiver for the dermal sensitization study. Studies to determme genotoxicity/mutagemcity are required to support any foodlnonfood use, if the use is likely to result in significant human exposure, or if the active pestictdal ingredient is structurally related to a known mutagen, or belongs to a class of chemical compounds containing known mutagens. The genotoxicity battery of studies includes those currently found most useful for evaluating mutagenicity potential of chemical pesticides, namely, the Salrnonella typhimurium reverse-mutation assay (Ames assay), mammalian cells m culture forward-gene-mutation assay, and an m vtvo cytogenetics assay. Current protocols for these studies are found m the EPA’s Office of Pesticides and Toxic Substances Health Effects Testing Guidelines (40 CFR Part 158, Subpart F-Genetic Toxicity). If repeated human exposure to the pesticide 1sexpected to occur, subchromc studies (90-d feeding, dermal, and/or mhalation) are required. As with the acute toxicity studies, there are specific conditions, qualifications, or exceptions to the designated subchronic test requirements as described in Part 158.690 (a)( 1). For example, the 90-d feeding study IS conditionally required for nonfood use, but is required if the use of the product results in repeated human exposure by the oral route, or the use requires a food tolerance determmation. If repeated contact with skm occurs, then a 90-d dermal study m the rat is required. Likewise, if there is repeated pulmonary exposure to the pesticide at concentrations that are likely to be toxic, as indicated from the acute inhalation study, then a 90-d inhalation study would be required. Although not specifically indicated, the oral and dermal subchronic studies requirements should be significantly reduced if the test material is of a nature to render performance of a test impractical (i.e., the material is a gas at room temperature). Data on alterations of immune response are conditionally required to support the registration of a pesticidal product, but essentially becomes required

Federal

Registration

433

Process

when there is a requirement for any of the subchromc studtes reflecting, again, sigmficant human exposure situations. Protocols for the immunotoxicny study are avatlable from the agency, and have been summarized elsewhere (5). Briefly, the study employs either the rat or the mouse as the test animal, and assays are performed after 30 d of dosing, to evaluate effects of the test substance on humoral, specific cell-mediated, and nonspecific cell-mediated immumty. It should be noted that a developmental toxicity (teratogemcity) study (Tier I) is required for food use and is conditionally required for nonfood use when the use of the product is expected to result in significant exposure to females. If significant adverse effects are observed at the Tier I level, then a Tier II study may be needed to provide an estimate of risk. To assesspotential hazard resulting from prolonged and repeated exposure, a chronic-exposure study (Tier III) would be required, tf the potential for adverse effects is indicated in any of the Tier I subchronic studies, coupled with the frequency and level of human exposure expected to occur under the use pattern. A carcinogemcity study, also in Tier III, is required, if the active ingredient (or any metabolites, degradates, or rmpurmes thereof) causes morphological effects (i.e., hyperplasia) m the subchronic study test animals indtcative of neoplastic potential, or if carcinogemc potential is indicated m the mutagenicity and/or immunotoxicity studies.

3.7. Nontarget Organism Testing As with nontarget organism testmg with microbial pesticides, the purpose is to develop data necessary to assesspotential hazard of biochemical pesttctdes to terrestrial wildlife, aquatic animals, plants, and beneficial insects The agency bases the hazard evaluatron of biochemtcal pesticides on tests stmtlar to those required to support registration of conventional chemical pesticides. However, recognizing the nature and nontoxic mode of action of most biochemical pestictdes, the agency has structured the data requirements m a tiertesting scheme. The use of tiered data requirements allow regulatory dectsrons to be made with fewer teststhan for conventional chemical pesticides,and results m much lower coststo the registrant, and lesstime for the registration process. In general, biochemical pesticides control behavior, growth, and/or development of target organisms. Ideally, Tier I tests should be capable of detecting adverse effects resulting from the primary mode of action on the nontarget organisms. The followmg criteria are used to determine the need for further testing of biochemtcal pesticides beyond the first tier: 1 If signsof abnormal behavior are reported in Tier I testsat levels equal to or less than the maximum expected environmental concentrations,

or

2. If detrimental growth, developmental, or reproductive effects can be expected, based on Tier I test data, available fate data, use-pattern mformatlon,

results of

McClm tack

434

testing, and the phylogenetlc slmllarlty between target pest and nontarget orgamsm(s),or

the mammalian toxicology

3. If the maximum expected enwronmental concentration IS equal to or greater than one-fifth the LC,, valuesestablishedin Tier I terrestrialwildlife studies,or equalto

or greater than one-tenththe LC,, or EC,, valuesm Tier I aquaticanimal studies In addition, both Tier I and Tier II tests would be required if the pesticide IS to be applied directly to water, high use rates are proposed, and if the biochemical agent is not volatile. Tier II testing involves environmental fate testmg for blochemlcal pesticides, to estimate envn-onmental concentrations of pesticides after application. Tier III consists of further acute, subacute, and chronic laboratory testing on nontarget organisms, and Tier IV consists of applied field tests encompassing both nontarget organisms and envn-onmental fate. The results of each tier of tests must be evaluated to determine whether further testing 1snecessary. Representative test species are dosed at high rates and may constitute a maximum challenge situation to evaluate adverse effects. Normally, if the results of Tier I testmg indicate toxicity, further testing at a higher tier level 1s required. The data requirements, as outlined m 40 CFR 158.740, are outlined m Table 5

4. Transgenic Plant Pesticides Since the early 1980~3,the mtroductron and expression of chlmerlc genes m plant cells has been possible, especially through the use of Agrobacterzummediated transformation. Such technology has been used to genetlcally englneer plants to express pestlcldal substances.The most Important or recognized examples mvolve transgemc plants engineered to confer insect resistance (Bt &endotoxin) and tolerance to viral infections (tobacco mosaic virus). The proposed definition of transgemc plant pesticides includes those plants genetically altered via the mtroductlon of genetic material for the purpose of imparting or increasing the production of a pesticide. The pesticide active ingredient is the pestlcldal substance(s) produced from, or modified as a result of, the direct mtroductlon of genetic material. The pestlcldal product includes the active ingredient and any substance(s) directly produced from, or modified as a result of, the mtroductlon of genetic material. The appropriate focus for a risk determination is on the pestlcldal product, Including the active ingredient, other information, as described below, 1sneeded to effectively evaluate potential risks associated with human exposure. The agency recognizes that there are substances m plants imparting reststance to insect or microbial damage, and that some are involved in herbicidal actlvlty against other plant species. Such plant-pestlcldal substances occur m many food crops, and are currently consumed without presenting a human dietary hazard. However, some pestlcidal traits from microbes, animals, or even other

Federal

Registration

435

Process

Table 5 Nontarget Organism and Environmental Expression Data Requirements for Biochemical Pesticides (40 CFR 158.890) Kind of data required Tier I studies Avian acute oral toxrcity (bobwhite quail/mallard duck) Avlan dietary toxicity (bobwhite quail/mallard duck) Freshwater fish LCso testing (rainbow trout) Freshwater aquatic invertebrate LCsO testing Nontarget plant studies Nontarget insect testing Tier II studies Volatility Dispenser water leaching Absorptton-desorption Octanol/water partition UV absorption Hydrolysis Aerobic soil metabolism Aerobic aquatic metabolism Sol1 photolysls Aquatic photolysis Trer III studies Terrestrial wildlife testing Aquatic animal testing Nontarget plant studies Nontarget insect testing

Guideline ref no a 154-6 154-7 154-8 154-9 154-10 154-11 155-4 155-5 155-6 155-7 155-8 155-9 155-10 155-11 155-12 155-13 154-12 154-13 154-14 154-15

plants, when introduced mto another plant spectes, may represent a novel exposure, and perhaps a new risk for human health or the environment. It should be noted that the agency has proposed not to examine the plant per se,

but rather the pestictdal substance produced m the plant, and the novel exposure that plant may provide for the plant-pestictde substance. The agency has identified a regulatory systemthat spectfically exempts those compounds that are least likely to present a risk to human health or the envtronment. The exemptions from FIFRA, as proposed, mclude plant-pesttctdal substancesthat are derived from plant speciessexually compatible wtth the plant m question, pesticidal traits that act primarily as physical barriers and receptors responsible for the hypersensmve response;and the coat proteins from plant pathogenic vn-uses.In addition to theseFIFRA exemptions, the agency has proposed to exempt from FFDCA requirements those pesticidal traits that are derived from food plants, and do not present a new dietary exposure in the modtfied plant.

436

A&C/r-dock

New dietary exposure for a plant-pesticide acknowledges that some food plants contain toxic constttuents present m vartous plant parts, or that are expressed at specific periods during maturation. Although the new exposure(s) may not require a toxicrty evaluatton of the novel pestictdal trait, the agency believes rt is prudent to examine the toxtctty of the trait prior to Its registration for use m a new food crop. Fundamental mformatron and/or data needed for a risk assessmentby the agency IS a thorough description of the source and nature of the inserted genes or gene segments, and a descrtptlon of the novel products (e.g , protems) encoded for by the genetic material. Presuming that the encoded products have been characterized adequately, this information would allow for a reasonable predtctton of toxicology issues and for the type of data essential to the evaluanon of potential risks EPA has divided the pesttctdal active ingredients into two categories. protemaceous pesticides and nonprotemaceous pesticides. This approach IS based on the fact that plant proteins, whether characterized or not, are stgmticant components of human diets, and are susceptible to acid and enzymatic dtgestton to amino acids prior to assimtlatton. Presuming that the new protemaceous products are adequately characterized, minimum human health concerns would exist unless the proteins have been tmphcated in mammalian toxlctty, exposure of the protein, although never tmphcated m mammahan toxicity through the different routes of exposure, has not been documented; or novel proteins are created via modification of the primary structure of the natural protein pesticide. The nonproteinaceous plant pesttctdes, which have not been submitted to date, may be evaluated separately, or in a manner analogous to that for conventional chemical or biochemtcal pesticides. Product characterizatton IS crittcal for assessing potential rusks resultmg from exposure of humans to pesticide-contaming plants. Charactertzatton embraces four basic areas: identification of the donor organism(s) and the gene sequence(s) to be inserted into the recipient plant; ldenttficatton and description of the vector or delivery system used to move the gene mto the recipient plant; rdentificatron of the recipient organism, including information on the insertion of the gene sequence; and data and mformation on the level of expression of the inserted gene sequence. Thts information IS crlttcal for assessing potential risks to humans and domestic animals when exposed to pesticide-contammg plants. Spectfic data/mformatton that is necessary for a risk evaluatron by the EPA have been previously described (6). The product characterization data/mformatlon can be used to establish the level of mammahan toxtcology data necessary to determme the potential risks associated with human and domestic animal exposure to transgemc plant-pesticide products. Key factors determining the extent of data requirements would

Federal

Registration

Process

437

include the nature of the pesticidal product (i.e., proteinaceous or nonprotemaceous), and whether or not the use pattern will result in dietary and/or nondietary exposure. Since dietary consumption is presumed to be the predominant route of exposure for food and feed crops engineered to express pesticidal substances, the potential toxicity of these unique substances could be assessedby oral toxicity studies (acute, subchronic, or chronic feeding studies). An in vitro digestibility assay may provide mformation about the potential for a protein to produce food allergy. If plants were engineered to produce volatile pesticide components, pulmonary exposure might be significant, even without a food use. The assessment of dermal irritation/toxicity might be addressed through reporting of any adverse reactions from skin contact during product development or use.Reporting of dermal effects resultmg from Incidental exposure may also suffice for nonfood uses,depending on the extent of exposure. Environmental fate (movement m the environment) and effects (toxicity) end pomts for transgenic plants are often quite different than for conventional chemical pesticides. Fate end points address the movement of the gene trait to other crops and/or noncrop plants (btological fate) and movement of the pestttidal product in the envrronment (chemical fate). Toxicity end points address the ability of the pesticide to cause adverse effects to nontarget organisms. Such effects could occur following consumption of the transgenic plant contaming the pestrctdal product by nontarget organisms. In general, environmental fate and effects end points include, but may not be limited to, gene-product persistence and movement in the environment, weediness; unplanned productron of the pesticidal product offsrte, leading to exposure to a new group on nontarget organisms; disruption of the ecosystem by the establishment of a new trait in wild relatives; and the effects on nontarget organisms and fate in the environment. For persistence, the end points, and the potential of the pesticidal product to leach to groundwater and run-off mto surface waters, IS similar to that considered for conventtonal chemical pesticides. Although many crop plants are highly domesticated, and do not survive outside cultivation, the potenttal for transgenic plant pesticides to survive is an end point not applied to conventional pesticides. If outcrossing occurs, whereby the novel trait IS transferred to a related wild plant species, the pesticidal product may be produced in unintended areas, leading to exposure to a new group of nontarget organisms. If the engineered tract is transmitted, and is able to become established m wild relatives, the newly acquired trait may provide the weld relatives a selective advantage within the natural plant community and disrupt the ecosystem. The ability of the gene and the expressed product/trait to be acquired and persrst are end points to consider for transgenic plants. Finally, effects of the gene/trait on nontarget organisms, and fate m the environment, are addressed by basic data requirements similar to those studies

438

McClrntock

required for the registratton of other btological pesticides. Such studies would include an avian dietary study; an avian reproduction study, if chrome exposure exists, an acute fish study, a freshwater aquatic invertebrate acute study, and/or a fish life-cycle study, if aquatic and chronic exposure is a possibihty, currently, a Collembola and earthworm study, if crop-residue exposure is expected; host range mformation; a honeybee study; an assessmentof outcrossmg potential; an evaluation of competitiveness of the novel trait m the plant community; an assessmentof the ability of the gene and/or product to degrade and persist in the environment; and momtormg of adverse effects 5. Encouraging the Development of Reduced-Risk Pesticides In July 1992, the EPA announced its interest m developmg new policies in the area of “lower risk pesticides,” and solicited pubhc comment via a Federal Register notice (57 FR 32140) and a public workshop. In January 1993, OPP published a Federal Register notice that provided an update of the reduced risk pesticides program, with an overview of plans for the future and the agency’s short-term and long-term strategies(58 FR 5854). A Pesticide Regulation Notice (PRN) was published m July 1993, which implemented the reduced-risk pesticide mmative (PRN 93-9), and provided gutdance on EPA’s voluntary, mterim process for identifying pesticides that may be eligible for priority treatment as lower risk products. In addmon to announcing this progressive mmative, PRN 93-9 set forth the basrc requirements that registrants must meet on order to be considered a reduced-risk candidate. The agency also issued a draft PRN expanding the mmal voluntary reduced-risk PRN to include new uses of those pesticides that have been granted reduced-risk status. The primary goal of this draft PRN was to grant expedited reviews to those new-use applications of reduced-risk pesticides OPP is currently developmg a PRN that ~111provide mformation to the regulated communrty on the specific requirements for achievmg reduced-risk status under the Food Quahty Protection Act (FQPA). The agency interprets FQPA as requirmg the current reduced-risk program to include consideration of new active ingredients, new usesof previously identified reducedrisk pesticides, and amendments to all uses considered reduced-risk. In essence, FQPA ISviewed as codifymg current practices and OPP intends to contmue usmg its weight-of-evidence approach m applying the criteria specified m the law. The purpose of the reduced-risk program was to encourage the development, registration, and use of newer chemical pesticides that present lower risks to public health and the environment when compared to existmg alternatives. In addition, the agency encouraged the adoption of pest-control strategies, such as integrated pest management (IPM), that would impede pest resistance to pesticides and achieve lower overall risk. To this end, the agency proposed to give special treatment to registration applications of potentially

Federal

Registration

Process

439

lower-risk pesticides by expediting the review of new active ingredients that met the agency’s reduced-risk criteria: reduces pesticide rusksto human health; reduces pesticide risks to nontarget organisms; reduces the potenttal for contammatton of valued, environmental resources; or broadens adoption of IPM, or makes rt more effective. The average amount of time required to register a new conventtonal pesticide m 1995 was 38 mo; the new reduced-risk/safer pesticides took, on the average, only 14 mo. Since 1993, 29 new chemical submissions were received by the agency as reduced-risk pesttctde candidates. Of the 29 submrssrons, 14 met the reducedrisk criteria for expedited review, of which 8 have been regtstered. 5.1. Existing Mechanisms and Incentives for the Development of Biological Pesticides A srgmficant mrttattve to encourage the development of btologtcal pestttides was the establishment of the BPPD in November, 1994, to manage the registration and reregistration of bropesticrdes. BPPD approved 14 new active ingredients of brological origin m 1995 and 10 m 1996, which represents more than one-third of the new active mgredtents (36 and 41%, respectively) registered in those years. BPPD also issued reregistration ehgrbihty documents for three biopestrcides m 1995 and eight m 1996. It should be noted that the agency currently provides incentives for the development and commercralrzatton of btological pesticides Such mcentrves include tiered data requirements, reduced data requirements based on the nature and hrstortcal use of the product, expedtted review times, and tolerance-fee waivers. OPP also developed a set of gurdance crrteria for a subset of pestrcrdes that are naturally occurrmg food components/food additives, which would limit the kinds of hazard evaluatton data required for both human health effects and nontarget organisms. In fact, if the pesticrdal food component met certam crttena, then the required databasemight be minimized or even totally waived. The rationale was based on the historical use and available safety informatron that suggestedthat such food components did not pose any additional or unreasonable risks to humans and the environment when used as a pesticide. Other exrstmg mechanisms or incentives to registration and commerctahzanon Include, but are not hmrted to, a 250-acre limit before an EUP IS required for many pheromones; tolerance exemptions for certain classes of pesticides, rather than for each compound; specific exemptions based on formulatrons (e.g., pheromones in traps are exempt under 40 CFR Section 25[bJ), and 25(b) exemptton for certain, historically safe naturally occurring compounds. 6. Summary The OPP at the EPA has recognized that certain categories of pesticides are of such a nature that less data and/or mformatton IS required to support a find-

440

McClintock

mg of no sigmficant adverse effects to humans and the environment. BIOpesticides (i.e., microbial and blochemlcal pestlcldes) are included m this category. The requn-ements for registration are intended to generate necessary data and information to address concerns relating to the identity, composltlon, potential adverse effects, and envlronmental fate of a specific pestlclde. Specifically, when applying for an EUP, a registration, or an amended registration, the registrant must satisfy a generic set of data requirements that are associated with the active pesticldal ingredient. Additional product-spec~fic data are usually required to address potential concerns associated with the end-use formulation. However, in some instances, the data requirements may be mappropnate, or such data and/or mformatlon may not provide useful information to fully evaluate the pestlclde, and, therefore, additional data would be required

If a small-scale field test IS planned, and if the new pesticide contains certain types of genetically modified or nonmdigenous microorganisms, then specific data must be submitted with the notlficatlon. The data submitted allows EPA the opportunity to evaluate unregistered blopestlcldes, and to determine whether an EUP is necessary prior to conducting small-scale field tests. However, prior to submlttmg such data, the agency recommends consultmg with all parties Involved, to resolve questions relating to protocols and data requirements for notlficatlons, EUPs, reglstratrons, and amended registrations, before

undertaking extensive laboratory testing to support the application. Determining data requirements at the outset 1s the key to preparing biopesticlde application.

a satisfactory

References 1. US EnvIronmental Protection Agency, Office of Pesticides and Toxic Substances (1989) Subdwslon M of the Pestwde Testzng Guldelwes Mzcroblal and BLOchemzcal Pest Control Agents. Document No. PB89-2 11676 National Technical Information Service, U. S Department of Commerce, Springfield, VA. 2 US Environmental Protection Agency, Office of Pesticides and TOXIC Substances (1988) Guidance for the Rereglstratlon of Pestxlde Products Contarnzng Bacillus thurlnglensis as the Actzve Ingredzent Document No 5401 RS-89-023. Natlonal Techmcal Information Service, US Department of Commerce, Springfield, VA 3 Hartlg, P C., Chapman, M A., Hatch, G G., and Kawamshl, C Y (1989) Insect virus assays for toxic effects and transformation potential m mammahan cells. Appl Environ Mlcroblol 55, 19 16-l 920. 4. US Envu-onmental Protection Agency, Office of Pesticides and Toxic Substances (1983) Pestrclde Assessment Gwdelmes Subdwlslon M Bloratronal Pestlcldes EPA540/9-82-028. Document No PB89-2 11676. National Technical Information Service, US Department of Commerce, Sprmgfield, VA.

Federal Registration Process

441

5 Sjoblad, R. D (1988) Potentral future requtrements for umnunotoxtcrty testmg for pesttctdes. Tox~ol Ind Health 4,391-394 6 McClmtock, J T., SJoblad, R. D , and Engler, R. (1991) Potential data requtrements for the toxtcological evaluatton of genetically engineered plant pesticides, in Food Safety Assessment (Finley, J. W., Robinson, S. F , and Armstrong, D. J., eds.), ACS Sympostum Series 484, ACS Pubhshers, Washmgton, DC, pp 41-47

23 IR-4 Biopesticide Christina

Program for Minor Crops

L. Hartman and George M. Markle

1. Introduction In 1962, the State Agricultural Experiment Statlon Du-ectors recogmzed the needs of growers and requested the then US Department of Agriculture’s Cooperative State Research Service (CSRS) to mitlate an interregional research proJect that would coordinate the agricultural commumty’s efforts, and assist growers m obtaining registrations of agricultural products for minor-use needs. The proJect, known as IR-4, was estabhshed m 1963, and has developed into a program that involves the US Department of Agriculture (USDA), the US Envlronmental Protectlon Agency (EPA), State Agricultural Experiment Stations (SAES), the Cooperative State Research, Education, and Extension Service (CSREES), agricultural chemical compames, commodity orgamzatlons, and mdlvldual growers located from Puerto Rico to Hawall In 1977, the objectives of IR-4 were expanded to include the registration of pesticides needed for the protection of commercially grown nursery and floral crops, forest seedlings, and turf grass. The program was further expanded m 1982 to include the registration of biological pest-control agents, including mlcrobials and blochemlcals. An IR-4 headquarters faculty/staff provides the leadership and overall coordination for the diverse components of this national program. The IR-4 headquarters is located at Rutgers, the State University of New Jersey. The objectives of the IR-4 minor-use program are to obtain minor and specialty use pesticide clearances, and assistm the maintenance of current reglstrations; and to further the development and registration of biopestlcides for use in pest-management systems. The IR-4 program is administrated by two separateagencies,USDAGSREES and USDA-AR& both of which report to the undersecretary for research, From Methods /n Botechnology, vol 5 Blopestudes Use and De//very Edited by F R Hall and J J Menn 0 Humana Press Inc , Totowa. NJ

443

Hartman and Mark/e

444 SECRETARY

OF AGRICULTURE

UNDER SECRETARY RESEARCH, EDUCATION ECONOMICS

AND

I

I

USDA-CSREES *Administrator

l

USDA-ARS Administrator

IR-4 ADMlNlSrRATIVE ADVISORS I

I

I IR-4 HEADQUARTERS *National Director

IR-4 I’ECHNICAL COMMlTl EE

REGIONAL LEADER LABORATORIES * Ihntlor

I USDA-AR.3 ANALYTICAL LABORATORIES

---------

. Field Coordinator * Lnharntoty Coardmtor

USDA-ARS MINOR USE OFFICER I

------

I , USDA-ARS SCIENT IS1 S

1 I

I STATE LIAISON RIPRESENTATIVES

I I --------

Fig. 1

J

Secretaryof agriculture

education, and economics(seeFig. 1) Admimstratlve advisers, who are deansand directors of state agricultural expenment stations,and the USDA admmistrators provide administrative oversight. An IR-4 techmcal committee (recently renamed “project management committee”), composed of the directors of the four IR-4 regional leader laboratones, USDA-ARS and CSREES representatives, chair of the admmistratlve advisors, and the IR-4 natlonal dlrector (recently renamed “executive director”), provide technical guidance to the program. The regional laboratory director provides oversight of an IR-4 staff at each laboratory, which consists of a regional field coordmator, reglonal laboratory coordinator, and support staff. There also is a quality assuranceoffice located at each regional laboratory. Regional laboratorles are located at the Umverstty of Cahforma-Davis, Michigan State University, University of Florida, and Cornell Umversity. These laboratones interact with IR-4 liaison representatives from each state,who are normally faculty members of then-respective institutions, and serve without compensation. A companion minor-use program IS administered by USDA-Agricultural ResearchService. This program IS headquarteredat the Agricultural ResearchCenter m Maryland, and interfaces with three USDA-ARS analytical laboratorles and USDA-ARS scientistswho serve as federal liaison representatives to IR-4.

IR-4 Biopesticide Program Table 1 M-4 Biopesticide

Project

Successes

445

on Food Crop9

Crop

Biopesttctde

Methyl eugenolplus malathion Grape berry moth pheromone Codling moth granulostsvirus Pseudomonasfluorescens strain NCIB 12089

All Strawberries Blueberries Pastures All Grape Apple, pear, walnut, and plum Mushroom

Lagenldlum

Rvze

Bacillus thuringtensls

Gibberellic acrd Grbberellic actd Bacillus popilliae

glganteum

aProJects prior to 1994

There has been general agreement in the Umted Statesthat a minor use is any use of a pest-control product for which the volume IS msufficient toJustify the cost by acommercial registrant to obtam a regtstratron. Thts may relate to the general or frequent use of a product on a low-volume crop, or it may apply to the infrequent or localized use of a product on a higher-volume crop. In either case,the problem of obtaining clearancesfor the minor-crop/mmor-use market IS primarily one of economics. The Food Quality Protection Act of 1996defines “minor use” as follows. “The term ‘minor use’ means the use of a pesticide on a commercral agrrcultural crop where: the total United Statesacreagefor the crop is lessthan 300,000 acres,or * The Administrator of the US EPA determtnesthat the use doesnot provide sufficient economic mcenttve to support the initial or contmumgregistration ” (7) l

The definition further states that the administrator may determine that a minor use exists if there are insufficient alternatives available for use on the crop; that the alternatives pose greater risk to the environment or human health; or that the minor-use pesticide plays a significant part in the management of pest resistance, or in integrated pest-management (IPM) systems, Although the scope of the IR-4 program was expanded in 1982 to include research leading to the registration of biopestrcides, there was little progress in this area until the availability of additional funding in 1994 (see Table 1). Although btotechnology offers many opportunities for developing new, safe, and effective pest-control products, rt is unlikely that there will be an accelerated biopestrclde registration effort on minor crops m the United States wrthout pubhc assistance. So rt was in late 1994 that the IR-4 program hired a manager for the bropesticide program, and set aside additional earmarked funds to be used to support btopesticide registrations.

Hartman and Mark/e

446 2. IR-4 Biopesticide

Program

The IR-4 chemical program has a long history of success in obtainmg pestcontrol materials for minor crops, but the biopesticide program clearly could not be modeled after it. First, the data needs for biopesticides are very different from chemicals, second, the system for finding and selecting the IR-4 chemical projects was not appropriate for biopesticides. The first priority for the new program was to establish a network of mdividuals in the biopesticide research/production area who could be relied on to provide up-to-date information on new products, as well as current status of extstmg products This was no small task, considermg that the term “biopesticides” covers biochemicals like pheromones, attractants, and neem, microbials varied in then own right; and transgemcs in the microbial world, not to mention transgemc plants Add to this the division of pest-control products into insecticides, fungicides, and herbicides, and the ability to pull a cohesive group becomes rather futile. Instead, it was found that certain meetings and organizations could provide mformation on each of the groups, and that information could be synthesized to form the base for the program. The establishment of a base group of customers was only the begmnmg. Much needed to be done, and resources were limited. To meet some of these needs, the IR-4 Btopesttcide Grants Program was established. This program sought to consolidate requests to a once-a-year call for grant proposals, and to formalize the requu-ements for submitting a proposal. Proposals solicited in 1995, 1996, and 1997 resulted m the funding of nine, ten, and eight prolects, respectively. Proposals funded in 1995 are listed here* 1 Pepper extract trials for minor crops to control insects. 2 Alternarla spp and Fusarlum tricwctum for the control of dodder on cranberries 3 Beauverza basstana for control of citrus root weevil larvae

4 Development and evaluation of recombinant viruses as biological 5 Entomaphaga malmaiga for the control of gypsy moth 6. Baczllus cereus for the control of alfalfa seedling diseases 7 Flavobacterlum balustmum and Trlchoderma hamatum fortified for control of damping-off and root-rot pathogens of vegetable and omamentals 8 Paectlomyces jiimosoroseus for control of insects on greenhouse 9 Pseudomonasfluorescens PRA25 and Burkholderla cepacla for borne dtseases of peas, snap beans, and sweet corn

insecticides

potting mixes bedding plants cuttings. control of soil-

Proposals funded in 1996 are as follows: 1 Nonaflatoxm cotton

producing Aspergrllusjlavus

to reduce aflatoxm contamination

2 F balustlnum and T hamatum as disease suppressive agents in potting mix. 3 Biopesticides for use in specialty mushrooms.

in

IR-4 Biopes ticide Program 4. Macleaya plant extract for control of trisects on greenhouse ornamentals. 5 P jluorescens PRA25 and B cepacla AMMD as a seed treatment for disease suppresslon m pea, snap bean, sweet corn, and super sweet corn 6 B. basszana for control of citrus root weevll. 7 F. truxnctum and Alternarla conpncta/mfectorza for dodder control m cranberry 8 MCH pheromone bubble caps for control of spruce beetles. 9. Trzchoderma harzianum for control of botrytls on grape and strawberry 10 B. bassiana for control of western flower thnps. Although the funding program was designed to obtam the data needed for an EPA registration, the registration process itself remains a mystery, and often a frustration

to many people, including

small compames

trying to market a few

or even one product. IR-4 had a long-standing relatlonshtp with EPA concernmg the submission of residue data, but the submission of complete registration packages for blopestlcide registration was less familiar territory. To Increase IR-4 expertise m this area, the blopesticide manager was sent to EPA for over a year to work as a part-time visiting scientist in the newly formed Blopestlcide and Pollution Prevention Division of EPA. This training, combined with the existing knowledge m-house at IR-4, enabled the program to greatly increase the number of experimental use permits and full registration packages submitted by IR-4 to the agency. In 1995, EPA approved tolerance exemptions for the following: 1. Methyl anthrandate as a bird repellent on blueberries, themes, and grapes 2. Codling moth granulosls virus for apple, pear, walnut, and plum 3 Cmnamaldehyde on mushrooms for the control of Vertxlllwm spot and dry bubble disease

Also in 1995, IR-4 submitted four petitions to EPA, three proposmg temporary tolerance exemptions for microbial pest control agents, as follows: 1. Flavobacterlum

balustwum 299 and T hamatum 382 XI or on vegetable bedding

plants

2 Asperglllusflavus

AF36 on cotton 3 One petltlon exemption from tolerance for formic acid m honey and beeswax

In 1996, EPA approved temporary tolerance exemptions for: 1 A. jlavus AF36 on cotton to prevent aflatoxin contamrnation of cottonseed 2 T. hamatum and Flavobactenum balustmum on vegetable bedding plants for control of damping-off

and root rot

Also m 1996, IR-4 submitted two petitions to EPA proposmg’ 1. Temporary tolerance exemptions from the requirement of a tolerance for the mlcroblal pest control agents Burkholderza cepacia AMMD and P fluorescens PRA-25 m or on peas, snapbeans, sweet corn, and supersweet corn, to control

448

Hartman and Mark/e

seedling diseases The petition also supported B cepacza AMMD as a fohar application in or on American ginseng,potatoes,carrots,tomatoes,and turf 2 Temporary tolerancefor kaolin on various crops Registration of a biopesticide is never the same story twice. In 1976, the IR-4 program submitted a petmon “to exempt from the requirement of a tolerance residues of the microbial msecticide Bacillus thuvingzensis (Bt) Berliner m or on beeswax and honey and all other raw agrtcultural commodmes when it IS applied either to growing crops, or when it is applied after harvest in accordance with good agricultural practices” (5). This broad tolerance exemption allowed the proliferation of Bt products, and eliminated the need for a company to petition the agency every time a new crop needed to be added to the label. This petition was supported by existing mdustry data, and did not require further generation of data by IR-4. In contrast, the use of a nonaflatoxm-producing strain of A. flavus m Arizona cotton fields required the gathering and submission of over 1500 pages of literature references, as well as toxicology testing. Although the data requirements for biopesttcides are greatly reduced, the costs of a full Tier I (see Table 2) can still be unreachable for many small compames The health testing must be done under Good Laboratory Practices (GLP), and should always be done by a reliable, certtfied contract lab familiar with EPA protocols. Almost every microbial must be subjected to some health tests, and, if the pesticide is new to the agency, then more data 1s required. Microbials have their own set of requirements, and these are the most defined; biochemicals follow the chemical protocols, and transgenics still wait for a final rule from EPA Waivers for toxicology testing can be requested, and are sometimes the most appropriate course of action. For example, most bacteria are too heavy to be considered a problem via the pulmonary route; therefore, a waiver of this test is an option. Use pattern may also ltmtt exposure, or labelmg may prevent eye or dermal exposure from occurring The literature is a valuable source of mformation Occurrence m nature can be verified, and growth curves can be cited; in fact, some companies have discovered that the application of then microbial pest-control agent does not mcrease the population level over that already existmg m nature. Sometimes, even toxicology data can be found m the literature. In particular, for biochemical products that have food or cosmetic uses, the toxicology data is readily available m the public literature. Although waivers have been used, and continue to be an aid to biopesticide registratton, it should be remembered that some testing is needed to set a confidence level. The Tier I testing for environmental data can be addressed at all stages of field efficacy testing. In particular, nontarget insect and honeybee effects can be monitored m the field, and this data can be used for waiver rationale. Plant studies can also be addressed by gathering data on the effects of the product on

IR-4 Biopesticide

449

Program

Table 2 Tier I Toxicology Requirements for Microbial Pest Control Agents Guidelines ref. no Health effects 152A-10 152A-1 I 152A-32 152A-13 152A-14 152A-15 152A-16

Nontarget organtsms 154A-16 154A-17 154A-18 154A-19 154A-20

Test Acute oral toxtcity/pathogemctty study Acute dermal toxicity study Acute pulmonary toxtctty/pathogentctty test Acute tv toxicity/pathogenicity study Prtmary eye nrttatton/tnfectron study Hypersensmvtty mctdents Cell culture tests with viral pest control agents Avian oral Avtan respiratory pathogentcity test Wild mamma1 toxlclty and pathogeniclty

testing0

Freshwater fish toxtclty and pathogenrcny testing Freshwater aquattc invertebrate toxicity and pathogemcrty testing Estuarme and marme animal toxtctty and pathogenrcrty tests” Plant studies Nontarget insect testing for toxtctty/pathogentcny to Insect predators and parasttes Honey bee toxtctty/pathogenlcity test

154A-2 1 154A-22 154A-23 154A-24

“The Tier I toxicology tests reqwredunderItems 152A-10 through 152A- 14 above are usually

adequate “Required

when direct apphcatlon, or runoff because of use pattern, IS made to the estuarme

andmarineenwronment

nontarget plants. Greenhouse studies looking for infectivity or phytotoxicity can easily be incorporated into the early research. If the microbral agent does not grow at bird body temperatures, then this can be the basis for a waiver request for the required bud studies. If the use pattern does not impact surface or dramage water, then this will be important for fish and aquatic invertebrate requirements. The more a researcher or company knows about the product m terms of how much 1sout there, where it goes, and who it affects, then the more comfortable EPA can be in making dectsions. In working with EPA, rt is important to remember that the registrant 1s responsrble for providing all necessary information. For a regrstratton, or even

an experimental use permit, a preregistration meeting with the agency ISa must. The purpose of that meeting is to provide EPA with as much mformatlon as possible about the active ingredient. Since the meetmg trme 1soften brref, product identity information, manufacturmg process, quality assurance, and infor-

Nartman and Mark/e

450

mation/data on health and environmental effects (mcludmg rattonale for any data waivers you may be requesting) must be presented in a concise manner Agreements made at this meeting can be verified by sendmg a letter to EPA detailing the agreements and askmg for concurrence. Thts does not guarantee that EPA will not later ask for other data, but IR-4 has had much greater success with submissions to EPA by utilizing preregistration meetings. Once a submtssion has been sent to the EPA, thts does not mean that the work 1sover. The sequence should be as follows: 1. All EPA pesticide registratron submrssions go first to the central document processmg desk, where the data is cataloged, assigned trackmg numbers, microfilmed, and so on 2. If a letter from document processmg is not received within 4-6 wk of submission, then it IS time to call and try to locate the documents 3 If the submission has made it over to the biopesticide office, then it is time to call the mtcrobtal branch chief (also for transgenics) or biochemical branch chief, to see who has been assigned as the regulatory action leader (RAL) 4. The RAL ~111 be m charge of the product, and will be the chief point of contact Keep in touch, find out where the petition is located m the review process, and call in for updates at appropriate mtervals 5. Do not be a pest, but do not get lost m the shuffle, either Many people who have sent mformation to EPA do not follow up on it, and sometimes wait over a year A regrstrant must be proactive.

3. The Future The biopesticide market, hke most busmesses,ts constantly changmg. There was a time when Bt products were the biopesticide market Today transgemc plants are important m pest control, and expectattons are that they will contmue to be so, at least tn the corn, cotton, and soybean markets. Bt 1sstill alive, and new products, mcludmg genetically modified Bts are now on the market. Products like B. basszana and T. harzzanum struggle to become part of the agricultural ptcture. Interest m biochemical-type products, such as baking soda, kaolin,

sucrose esters, and plant extracts, has increased

The degree of manu-

facturing and marketing of biopesticide products 1s several degrees more sophisticated than tt was even 5 yr ago. Some smaller companies are gone or have been bought out by larger companies. The picture has changed, but It IS not bleaker. EPA can and does make changes. A company can watt for EPA, or the company can move forward m a concerted effort to present its vtews and Ideas on

regulatory change. EPA is not equipped to deal with mdividual requests, but has worked successfully with groups in the past. An important example IS the regulations set for leprdopterous pheromones. The pheromone producers as a group sat down with EPA and drew up regulations that both producers and

IR-4 Biopesticide

Program

451

EPA could accept. The results have been a much easier pheromone registration path. Other regulatory changes can be brought about m the same manner. The future IS bright for blopesticldes under the new FQPA. FQPA will have a positive Impact on blopesticides, with the need for more registered uses. Reduced-risk pest-control agents will continue to have a favorable position m an expedited EPA registration review process. Typically, new blopesticlde active ingredients are registered m 11 mo; conventional pesticides can take 2-3 times longer. EPA believes that blopesticldes pose less risk than other pest-control agents. Less risk means less exposure, which would result m reduced risk to human health and the environment. Many pest control tools, e.g., organophosphates, may be lost under the FQPA provision for tolerance reassessment of all pesticides wlthm 10 yr of the FQPA enactment. With this potential loss, more alternatives will be required to fill the pest management voids, especially in IPM and resistance management. Blopestlcides should be more compatible with the area-wide IPM concept of USDA, which will help to address the USDA’s goal of 75% of the United States acreage under IPM by the year 2000. More compatible pest management tools, especially blo-pestlcldes, will be required. Conventional pesticides ~111still be needed for United Statesagriculture now and mto the future, but the benefits of reduced-risk pesticides make them potential replacements or alternatives for uses at risk. Therefore, the FQPA and EPA will positively Impact the registrations ofbiopesticldes, because of reduced risk to the environment,

reduced human exposure, expedited

registration and review, and reduced data requirements, and because there are potential alternatives to conventional pestlcldes under FQPA reassessment provisions that are more compatible with area-wide IPM programs for pest reslstance and management The new law requires that tolerances be safe, defined as a reasonable certainty that no harm will result from aggregate exposure.

IR-4 will continue to work with anyone Interested in moving a biopesticide product toward commerciahzation. Resources are limited, but can be used wisely to help as many products as possible. IR-4 will continue to stay current

with regulatory requirements, as well as with the changmg products themselves.

References and Suggested Reading I. Hartman, C. L (1995) The IR-4 blopestlcldes program registration of biopesticides in the mmor uses proJect Proceedings of the Society of Invertebrate Pathology, Cornell Umverslty, July 16-2 1, p. 26. 2 IR-4 (1996) Annual Report for the Interregional Research ProJect no 4 Office of IR-4, NJAES, Cook College, Rutgers, the State Umverslty, New Brunswick, NJ 3 IR-4 (1993) IR-4 ProJect Statement for the Period 1 October 1993 to 30 September 1998. Office of IR-4, NJAES, Cook College, Rutgers, the State Unlverslty, New Brunswick, NJ.

452

Hartman and Markle

4 Markle, G M (1982-l 997) IR-4 Newsletter, NJAES Pubhcattons, Office of IR-4, NJAES, Cook College, Rutgers, the State Umverstty, New Brunswtck, NJ, Volumes 13-28 5. Markle, G. M. (1976) Tolerance exemption for Bacillus thurznglenszs on all raw agricultural commodities Federal Register 41, 24,885 6 US EPA (1997) Guidelines for expedtted review of conventional pesttctdes under the reduced-risk mtttative and for btologtcal pesttcldes Pesttctde Regulation Notice 97-3, Office of Pesticide Programs, US EPA, Washmgton, DC 7. US EPA (1997) The FIFRA and FFDCA as amended by the Food Quality Protection Act (FQPA) of August 3, 1996, OPP, US Envtronmental Protectton Agency March 1997,73OL9700 1, 8 US EPA (1996) Tttle 40-Protection of Environment Parts 150 to 189, Code of Federal Regulattons, U S. Environmental Protection Agency, Office of the Federal Register, US Governmental Prmtmg Office, Washington, DC 9 US EPA (1996) Residue Test Gutdelines, OPPTS 860 1000 to 1900. US Governmental Prmting Office, Washmgton, DC

24 Registration/Regulatory

Requirements in Europe

Mike Neale and Phil Newton 1. Introduction

A range of btopestictdes, including bacteria, vnuses, fungi, nematodes, protozoa, and benetictal insectsas active ingredients, are now commerctally available in Europe for control of insect pests,fungal and bacterial diseases,and weeds. The regulatory environment has been generally favorable to blopesticrdes: less stringent m demands for studies, and less costly procedures than that requned for chemicals. However, the regulatory sttuatton in Europe for biopesticides is in a state of flux. Europe follows the OECD defimtion of bropestrcides, i.e , including pheromones, insect- and plant-growth regulators, plant extracts, transgemc plants, and macroorganisms, as well as microorganisms (I) However, the chief emphasis in the European regrstratron arena has been the development of registration requirements for microbial pesticides and the notification procedure for macroorganisms Although the case IS clear for mtcrobral pesticides, the jury still seems to be out on the macrobiologicals-living products capable of self-renewal. Though exempted under the EPA regulation, the EU has no mechanism in place for such notrfication, and has yet to make a clear recommendatron. A license arrangement IS in place m the United Kingdom, but few restrrctlons are used m the Netherlands. A registration scheme, mostly concerned with efficacy and quality, 1sin its infancy in France, and currently operates on a voluntary basis. Until now, pheromones, plant extracts, and so on, have been treated in Europe as chemical pesticides. Several papers (24) have attempted to review and discuss the current status of registration requirements for bropesttcrdes. Although all authors agree that progress IS being made toward achieving harmonization, differences in detail and in interpretation may undermme these From Methods m Botechnology, vol 5 Bopestrcrdes Use and Del/very Edlted by F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

453

454

Neale and Newton

efforts, and contmue to raise the hurdles against the development of new biopesticides. Regulatory control of biopesticides m Europe has been based on precedents and standards established for chemicals. However, m recent years, a number of countries have introduced their own individual registration requirements to address the distinct characteristics of microorgamsms. This has led to higher regulatory demands in individual countries, and, hence, lack of a harmomzed approach both for registration requirements and for interpretation of the data obtamed. For example, the following data requirements have been requested for registration of a new Bacillus spp. France, 90-d toxtcity studies and mutagenicity studies; Denmark, 90-d toxicity studies, with particular emphasis on persistence, Germany, sensitization studies; and UK, analytical methods and residue studies, including fate m the environment. The resulting economic pressure has inhibited the future commercialization of certain biopesticides. With the development of the European regtstration Directive 9 l/4 14/EEC, covermg requirements for both chemicals and microorgamsms, attempts have been made to harmonize requirements and mterpretation of registration data throughout Europe. Precedence has been given to the chemical guidelmes, and, as a consequence, those for biopesticides are only now being considered. It is very important to be aware of the historical background of the new directive and its aims, and the impact of recent OECD proposals for harmomzatlon of regulatory requirements that may shape the future requirements for biopestictdes m Europe.

2. EU Registration Directive 91/414/EEC: Background The environmental pohcy of the European Community was, from the early 1970s to 1987, without a proper legal basis in the Treaty of Rome. To a large extent, this pohcy developed, within the framework of the international markets, in order to avoid barriers to trade and distortion of competition between European countries that would result from differing environmental legislation. It was not until the mid-1980s, when amendments to the Treaty of Rome were discussed, that the environment m its own right became part of the treaty of the Community. This happened through the adoption of Article 130 A, which represents the basis of the Commumty envtronmental policy. The article states: “Actlons by the Commumty relatmg to the environment shall have the followmg objectives: 1. To preserve, to protect and improve the quality of the environment 2 To contribute toward protecting human health. 3 To ensure a prudent andrational utilisatlon of natural resources

Registration

455

in Europe

Action taken by the commumty relatmg to the envtronment shall be based on the prmctple that preventative action should be taken, that environmental damage should as a priority be rectified at source,and that the polluter should pay. Envtronmental protectton requtrements shall be a component of the other Commumty polwes ”

This means that other policies, such as agricultural policies, must include a component of envtronmental protectton. 2.1. Pesticides

and Community

Policy

In 1976, the European Commtssion proposed a Councrl Dn-ectrve for the harmonizatron of national procedures for the regulatton of those pesticides used for plant protectton purposes. This failed to acquire the necessary degree of acceptance to ensure Its adoption. The general policy of the EU Commission regarding pesticides was set out in its 1988 Commumcatton on Environment and Agriculture, m which it recognized an increasmg concern about the environmental impact of pestrcrde use, and the ehmmatton of the agricultural surplus of the early years of the Communtty and a growing culture, and tn other forms The general objective of mum the use of pesticides,

interest in the deintenstficatton of Commumty agrtof alternative agrrculture Communrty policy 1s now to reduce to a strict mtntand to explore a range of possible measures for the

future, includmg harmomzatron m the field of pesticide registration, 2.2. Amended Proposal for a Council Directive Concerning the Placing of EC-Accepted Plant Protection Products on the Market, COM (89) 34 The main features of the proposal were as follows: 1 The establishment of a Community posttwe ltst of active substances,whose use m formulatrons may be considered as safe for human and ammal health, and for the environment. 2 Indtvtdual Member States ~111retam powers m respect of the national authorrzations of products, based on the active substances listed 3. Mutual recognitton by Member States of registrations granted by other Member States would be ensured, to allow the free movement of plant protectton products within the commumty. 4 A system of provrslonal authortzatton of formulatrons of new actrve mgredrents by Member States would be pernutted, pending Community dectstons on ltstmg m the postttve list 5. A IO-yr review program for evaluatmg the older active substances currently on the market 6. Harmonized rules concerning data protectton and confidenttahty 7. Harmonized labelmg and packaging provrstons. 8 Off-label uses.

456

Neale and Newton

9 Improved mformation exchange between Member States’ registration authormes

Followmg the publication of these amended proposals in 1989, detailed negotiations led to the adoption m July 1991 of the Council Directive 9 l/414/ EEC, “Concerning the placing of plant protection products on the market.” The purpose of this directive 1sto establish harmonized procedures for the authorization of plant protection products within the Community, removing barriers to free trade. 3. Directive 91/414/EEC: Basics/Description The purpose of the directive is to establish harmonized procedures for the authorization of plant protection products within the Community. Under the provisions of Directive 91/414/EEC, the followmg actions will be taken within the Community: 1, A postttve list of active substances which may be authorized for use within the community ~111 be established (Annex I). 2 Only substances which can be shown not to pose unacceptable rusks to human and animal health and to the envtronment will be mcluded 3 Indtvtdual Member States ~111retain powers m respect of the national authortzatton of plant protection products containing the active substances listed 4 The criteria to be used when authorizing products are specified m the Uniform Principles (Annex VI of the Directive) 5 The Directive also confirms the prmctple of mutual recognition of national authortzations in order to ensure the free movement, and enable free trade, of plantprotection products, treated plants and plant products wlthm the Commumty

The Directive provides derogations to allow a tiered approach to the authorization of new active substancesand transitional measures allowing the continued use of existing plant protection products. In addition, provision is also made for official and scientific users to request extension of the authorized uses In addition to Annex I, which lists active substances authorized for use within the Community, and Annex VI, the “Uniform Principle” to be followed by member states when authorizing plant protection products, annexes to the Directive providing detailed guidance on the data required to support apphcations for the authorization of active substances(Annex II) and plant protection products (Annex III) are under development. Annexes IV and V will eventually contain Risk and Safety Phrasesto be incorporated mto the label of authorized plant protection products. See Table 1 for a summary of annexes.

3.7. Procedures Specified in EU Directive 91/414/EEC Schematic descriptions are given below that outline the procedures for inclusion of a new active substance m Annex I of EU Directive 91/414/EEC. The procedure for review of an existing active substance for inclusion in Annex

Registration in Europe Table 1 Summary

457

of Annexes

Annex I: Active substances authorized for mcorporation m plant protection products Annex II: Requirements for the dossier to be submitted for the inclusion of an actwe substance in Annex I Annex III: Requirements for the dossier to be submltted for the authorlzatlon of a plant protection product Annex IV: Risk phrases Annex V: Safety phrases Annex VI: Uniform principles for the evaluation of plant protectlon products

I according to Regulation 3600/92 is stmilar (Fig. 1). It IS important to note that the detailed procedures, the structures of dossiers, and the guidelines are

still undergomg evolution. 3.1. I. Preparation and Submission of Dossier Gmdance on the wrltmg of the summary parts of the dossiers (the complete dossier 1sthe summary dossier, plus the actual study reports) has been under development within the EC for some time. The current version is Working Document 1663/VI/95 Rev 6, 3 1. January 1995: Guidelines and crlterla for the preparation and presentation of complete dossiers and of summary dossiers for the mcluslon of active substancesin Annex I of directive 9 l/4 14/EEC ) 3.1 2. Choice of Member State When submitting a new active substance for inclusion in Annex I (the European positive list), the initial choice of the Member State for the submlsslon of a dossier is up to the Applicant. Many factors can influence the chotce, mcluding size of the market, length of time for review, bias of country for agriculture, acceptance of reviews by other Member States,and good communication between applicants and other Member States. Although this has not yet happened in practice, the Commission is entitled to reallocate dossiers to another Member State. There is therefore no guarantee that the final choice of the Rapporteur Member State will be the same as that chosen by the Apphcant. 3.7.3. Completeness Check by Member State The Member State initially chosen as Rapporteur by the Apphcant conducts the initial completeness check on the dossier, according to Commumty Gmdehnes. If the dossier IS not complete, the Applicant IS requested to update his dossier. Once the Rapporteur is satisfied that the dossier is complete according to Commumty standards, he informs the Commission and the other Member

Neale and Newton

458 Ao~rox

T/me Scale Start I 3. Member State checks for comoleteness

dossle

1 3-6 Months Yes

countries

and sends

report

c 3. Applicant sends copy of dossier to other 14 Member States and to the Commwon

on

J

and

SCPH

examine

dossier

for

7 12 Months

Fig 1 Procedure for inclusion of a new active substance In Annex I of EU Dlrective 91/414/EEC

States of the results of his completeness checks, and requests the Applicant to send copies of his dossier to the Commission and the other Member States 3.1.4. Completeness Check by Working Group and Standing CommIttee on Plant Health (SCPH) This is a checkup on the work done by the Rapporteur Member State, to ensure that a uniform standard of appltcatlon IS achieved throughout the EU. When the SCPH IS satisfied that the dossier IS complete, a notice to this effect ~111be pubhshed in the official journal. At this point, other countries will begm

459

Registration in Europe Fig 1 Continued

1 year

7 + 6. Other Member States evaluate dossier far Provlslonal Regrstrabon

5. Detailed evaluation of the dossier by the Rapporteur Mem State in collaborabon with the applicant. possibly suggesting Improvements to be made to the summary dossrer

6

4 Member State may gran> ProvIsional Aeglstratlon for 3 years, renewable once for another 3 years

2 years

v 7. Appkant re-submits ‘improved’ summary dossier to all Member States and to the Commission

f

I

Peer

Review

1 7. Rapporteur submits his evaluatton (Monograph) to the Commission together with a proposed declsron and addrtlonal data required

I

of Monographs

Toxicology

I

by Commission Expert Hearing of the Applicant? Environmental

9. Proposed

Fate

Groups

(5 Member

States)

Residues

Ecotoxtcology

Recommendatton

I

1

v (

9. Voting

In SCPH

I

v 9. Recommendation

3-4 years longer

to Commission

or

Fig 1. Contmued. theu- own evaluation of the dossier for provlslonal authonzatlon, when this has been requested by the Apphcant 3.1.5 Choice of Rapporteur Member Stare/Evaluation As mentioned earlier, there IS no guarantee that the choice of Rapporteur ~111 be the same as the original choice made by the Applicant. The decision IS made by Commission and Member States,depending on the workload of the countries

460

Neale and Newton

3.1.6. Provisional Authorization for Plant-Protection

Products

1I Member States may grant Provisional Authorization for up to 3 yr after the Commission has confirmed that the dossier IS complete 2. However, Member States may dectde not to start evaluation until after ratilication by SCPH, I e., after monograph preparation 3. Member States may also be reluctant to take mulateral dectsions during early stages of mtroduction of Directive 91/414/EEC. 4 Provisional authorization may be very limited at first

3.1.7 Detailed Examination and Preparation of Monograph by Rapporteur (Dot 1654/V//94) The objectives are’ 1 To provide a full description of key points identified in assessing the database evaluated 2. To ensure a consistently high standard m the documentation and decision-making process. 3 To standardize the format, to ensure efficiency and economy in use of resources 4. To facilitate decision-making by Standmg Committee on Plant Health (SCPH) and by the Commission 5. To ensure transparency with respect to the basis for decisions made

3.1.8. Review of Monograph/Outcome

of Review

Proposals for decision by the Commission included m the monograph. 1. To include the active substance m Annex I of the Directive, stating the conditions of its inclusion, or 2 To postpone any decision on possible mclusion pending the submission of the results of additional trials or information specified, or not to Include the active substance in Annex I. Followmg such a decision, any existing provisional registrations would have to be revoked and the product would have to be withdrawn from the EC market.

3.1.9 Annex I Listing The proposed recommendations must be accepted by SCPH, on the basis of voting by a qualified majority. Politics of voting on mclusions plays a key role. Active substancescan be included on Annex I for a maximum period of 10 yr This procedure will cover all new biopesttcrdes to be registered tn the EU, as well as all existing biopesticides that will be the subject of the review process.

Those active substancesthat were registered m Europe prior to the pubhshmg of the Directive can continue to be marketed until they are reviewed, mcludmg the followmg: Ascherersonza aleyrodis; Agrotis segetum granuloszs virus; Bacillus sphaericus, Bacillus subtzlis, Bacillus thurzngienisz, Including; subspp azzewai; subspp israelensis; subspp kurstaki, Beauveria basszana,

Registration

in Europe

461

Cephalasponum, Dacnusa sibinca, Digglyphus isaea: Meturhizlum amsopllae: nuclear polyhedrosts virus; Phleblopsls gigantea; Phytoseiulus persimllu; Steinernema feltiae, tomato mosaic wrus, Trichoderma sp; Veltlclllium dahlzae, Verticillium lecanii; vtrus granulose carp.; vnus noctuelles. 4. Data Requirements for Biopesticides as Active Ingredients Annex II, Part B, details the requirements for the mclusion of mtcroorganisms and viruses into Annex I as active ingredient, as shown in Table 2. 5. Data Requirements for Biopesticide Products Annex III, Part B, details the requirements for the dossier to be submitted for authorization of a plant protectton (formulated) product containing mtcroorganisms and viruses (not applying to genetically manipulated organisms, when tt 1s already covered under a prevtous Directive 90/220/EEC), as shown m Table 3 6. Latest Status In early 1996,EPPO (EuropeanPlant Protection Organisatton)/CABI (Commonwealth Agrtcultural Bureau Intematlonal) held an International Workshop on “Safety and Efficacy of Blocontrol in Europe.” In addition to thesebodies, representativesof the EU Commtssion, European and US government oftictals and regulators, and industry attended. The workshop recommended that, given the potential value of microbials in IPM, and the constraintsfacing small-scaleproducers of mtcrobtals, efforts should be made to make the registration of microbials more efficient and rapid. To achieve this, the workshop recommendedthat a tiered approachfor microbial testing be developed to reduce unnecessarytesting,and that changesbe made m processing time, m order to speed the process, i.e., a fast-track approach. However, in May 1996, the results of preparatory discussion in several five-expert groups on Annexes IIB and III B were provided to industry for comments. Many of those requirements listed earlier, t-e., as for chemicals, were still mcluded

The major concerns are as follows:

1. tiered structure includes 90-d toxicity studies m Tier I; 2 no flexibility m the study requirements. lack of harmonization with other OECD countries; 3, testmg of both technical and formulations; 4. residue data required m crops and environment; 5 preharvest intervals and re-entry intervals; 6. certain environmental and ecotoxlclty studies not relevant, and 7 efficacy data requirements not clear treated as “chemical ”

6.1. Uniform Principles: Delays in Directive 91/414/EEC The European Parliament issued a lawsuit before the EU Court of Justice m Luxembourg against the Uniform Principles, which establish Annex VI of the

Neale and Newton

462 Table 2 Data Requirements

for Biopesticides

as Active Ingredients

1 Identity of the organism 1 1 Apphcant (name, address, and so on) 1.2 Manufacturer (name and address, mcludmg location of plant) 1 3 Common name, or alternattve and superseded names 1 4 Taxonomtc name and strain for bacterta, protozoa, and fungi; mdicatlon whether tt IS a stock variant or a mutant strain, for vu-uses, the taxonomrc designation of the agent, serotype, strain, or mutant 1 5 Collectton and culture reference number where the culture 1s deposited 1.6. The appropriate test procedures and criteria used for tdentification (e g , morphology, biochemrstry, serology) 1 7 Composttion. microbiological purity, nature, Identity, properttes, content of any impurities, and extraneous organisms 2 Btologtcal properties of the organism 2 1 Target organism Pathogemcny or kmd of antagonism to host, infective dose, transmrsstbthty, and mformatron on mode of action 2 2 History of the organism and its uses Natural occurrence and geographical dtstrtbutton 2 3 Host-spectfictty range and effects on species other than the target harmful organism, including species most closely related to the target species To include mfecttvtty, pathogemclty, and transmtsstbtlity 2.4 Infectivity and physical stability when used according to the proposed method Effect of temperature, exposure to air radtatton, and so on Persistence under the likely environmental conditions of use 2.5 Whether the organism is closely related to a plant pathogen or to a pathogen of a vertebrate species or a nontarget invertebrate species 2.6 Laboratory evidence of genetic stab&y (i.e., mutation rate) under envnonmental condittons of proposed use 2 7 Presence, absence, or production of toxins, as well as their nature, identity, chemical structure (if appropriate), and stability 3 Further mformation on the organism 3 1 Function e.g , fungicide, herbicide, msecticrde, repellant, growth regulator 3 2 Effects on harmful organisms, e g , contact poison, mhalation poison, stomach potson, fungttoxtc or fimgistattc, and so on; systemic or not m plants 3 3 Field of use envisaged e.g , field, glasshouse, food or feed storage, home garden 3 4 When necessary, m the light of the test results, any specific agricultural, plant health, or environmental condttrons under which the active substance may or may not be used 3 5 Harmful organisms controlled, and crops or products protected or treated 3 6 Method of productton, with descrlpttons of the techniques used to ensure a umform product, and of assay methods for Its standardtzatton In the case of a mutant, detailed informatton should be provtded on its production and tsolatlon, together with all known differences between the mutant and the parent wild strains

Registration

in Europe

463

Table 2 (continued) 3.7 Methods to prevent loss of virulence of seed stock 3 8. Recommended methods and precautrons concernmg handling, storage, transport, or fire 3.9. Possrblhty of rendering the organism unmfectrve 4. Analytical methods 4.1 Methods for establtshmg the identity and purity of seed stock from which batches are produced and results obtained, including tnformatton on varrabrlrty 4.2 Methods to show microbrological purity of the final product, and showmg that contaminants have been controlled to an acceptable level, results obtamed and mformatron on variability 4.3 Methods used to show that there are no human or other mammahan pathogens as contaminants m the acttve agent, includmg, m the case of protozoa and fungr, the effects of temperature (35Y and other relevant temperatures) 4 4. Methods to determme vtable and nonvrable (e.g., toxins) residues m or on treated products, foodstuffs, feedingstuffs, animal and human body flutds and trssues, soil, water, and an, when relevant 5 Toxtcological, pathogemcity, and infectivity studies 5 1 Bacteria, fungi, protozoa, and mycoplasma 5.1 1 Toxicity and/or pathogenicity and infectrvtty 5.1.1.1 Oral single dose 5 1 1.2. In cases m which a single dose IS not appropriate to assess pathogemcrty, a set of range&ding texts must be carried out to reveal highly toxic agents and mfecttvtty 5.1.1.3 Percutaneous single dose 5.1 1.4. Inhalatron single dose 5.1 1.5 Intraperitoneal single dose 5 1 1.6 Skm and, where necessary, eye nritatron 5 1 1.7 Skin sensitrzation 5 1 2. Short-term toxrcrty (90-d exposure) 5.1.2.1. Oral administration 5.1.2.2. Other routes (mhalatton, percutaneous, as appropriate) 5 1.3 Supplementary toxtcologrcal and/or pathogemcity and mfecttvuy studies 5 1.3 1. Oral long-term toxtcrty and carcinogemcny 5 1.3.2. Mutagenicity (tests as referred to under point 5.4. of part A) 5 1.3.3 Teratogenictty studies 5 1.3.4 Multrgeneratron study in mammals (at least two generations) 5 1 3.5. Metabolic studies. absorptton, dtstrtbutron, and excretion m mammals, mcludmg elucrdatton of metabohc pathways 5 1.3.6. Neurotoxictty studies, mcludmg, when appropriate, delayed neurotoxrcrty tests m adult hens

464

Neale and Newton

Table 2 (continued) 5.1 3.7 Immunotoxicity, e.g , allergemcity 5 1.3 8. Pathogemcity and infectivity under immunosuppression 5.2. Vu-uses, viroids 5 2.1 Acute toxicity and/or pathogemcity and mfectivtty. Data as outlined under point 5.1 l,, and cell-culture studies using purified infective vnus and primary cell cultures of mammahan, avian, and fish cells 5 2 2. Short-term toxicity Data as outlmed under point 5 1.2 , and tests for mfectivrty carried out by bioassay, or on a suitable cell culture at least 7 d after the last administration to the test animals 5 2 3 Supplementary toxicological and/or pathogemcity and mfectivity studies, as outlined under pomt 5.1 3. 5.3. Toxic effects on livestock and pets 5.4. Medical data 5.4 1. Medical surveillance on manufacturing plant personnel 5 4.2 Health records, from both mdustry and agriculture 5 4 3 Observations on exposure of the general population, and epidemiological data, if appropriate 5 4 4 Diagnosis of poisoning, specific signs of poisonmg, chmcal tests, if appropriate 5 4 5 Sensrtization/allergemcity observations, tf approprtate 5.4 6. Proposed treatment first aid measures, antidotes, medical treatment, if appropriate 5 4 7 Prognosis of expected effects of poisoning, if appropriate 5 5 Summary of mammalian toxicology and conclusions (mcludmg NOAEL, NOEL, and ADI, if appropriate) Overall evaluation regarding all toxicological pathogemctty and infectivity data, and infectivity and other mformatron concerning the active substance 6. Residues in or on treated products, food, and feed 6 1 Identification of viable and nonviable (e g., toxms) residues m or on treated plants or products, the viable residue by culture or bioassay, and the nonviable, by appropriate techniques 6 2. Likelihood of multiphcation of the active substance m or on crops or food, together with a report on any effect on food quality 6 3 In cases m which residues of toxins remain in or on an edible plant product, data as outlmed under points 4.2 1 and 6. of part A are required 6 4. Summary and evaluation of residue behavior resulting from data submitted under points 6.1-d 3. 7 Fate and behavior in the environment 7.1 Spread, mobility, multiplicatton, and persistence m an, water, so11 7 2 Information concerning possible fate m food chains 7.3 In cases m which toxins are produced, data as outlined under part A, point 7 , are required, when relevant

Registration

in Europe

465

Table 2 (continued) 8. Ecotoxlcologlcal studies 8 1 Birds. acute oral toxlclty and/or pathogeniclty and mfectlvlty 8 2. Fish: acute toxicity and/or pathogemcity and infectivity 8 3. Toxicity, Daphnza magna (if appropriate) 8 4 Effects on algal growth 8 5. Important parasites and predators of target species; acute toxicity and/or pathogemclty and mfectivlty 8.6. Honeybees* acute toxicity and/or pathogenicity and infectivity 8 7. Earthworms acute toxicity and/or pathogemcity and infectivity 8 8 Other nontarget organisms believed to be at risk acute toxicity and/or pathogemclty and mfectlvlty 8 9 Extent of indirect contamination on adjacent nontarget crops, wild plants, soil, and water 8 10 Effects on other flora and fauna 8 11 In cases m which toxins are produced, data as outlined under Part A, points 8.1 2 ,8 1 3., 8 2 2 , 8.2 3., 8 2.4., 8.2.5 , 8.2.6., 8.2 7., and 8 3 3 are required, when relevant 9 Summary and evaluation of points 7. and 8. 10 Proposals mcludmg Justification of the proposals for the classification and labelmg of the active substance m accordance with Directive 67/548/EEC Hazard symbol(s) Indications of danger Risk phrases Safety phrases 11 A dossier, as referred to m Annex III, Part B, for a representative plant protection product

Plant Protectlon Dtrecttve 91/414/EEC. The European Parliament is reasoning that, for formal reasons, it was not able to check if the Unlform Prmclples adhere to the EU environmental policy, and was not heard in the leglslatlve process. This should have been mandatory, because the Uniform Prmclples have departed from the framework of the Plant Protection Directive (through derogation of the ground water standards), and violate the Drinking Water Directive (derogation from 0.1 pg/L possible). On April 30, 1996, the Advocate General proposed the annulment of the Uniform Prmclples. If the court follows the advice of the Advocate General (which It usually does), the Uniform Principles will be declared null and void. The Judgment on this case was given on June 18, 1996, In favor of the European Parliament. The matter IS still under discussion

Neale and Newton

466 Table 3 Data Requirements

for Biopesticide

Products

1 Identity of the plant protection product 1 1. Appltcant (name, address, and so on) 1 2 Manufacturer of the preparation and the acttve agent(s) (names, addresses, and so on, Including location of plants) 1 3 Trade name, or proposed trade name, and manufacturer’s development code number, or the plant protectton product, tf appropriate 1 4 Detailed quantitative and qualttatlve mformation on the composition of the plant protection product (active orgamsm[s], met-t components, extraneous organisms, and so on) Physical state and nature of the plant protectton product (emulstfiable concen15

trate, wettable powder, and so on) 1 6. Use category (insecticide, fungicide, and so on) 2 Technical properties of the plant protection product 2 1 Appearance (color and odor) 2 2. Storage stability stabtltty and shelf life. Effects of temperature, method of packaging and storage, and so on, on retention of btologtcal activity 2 3 Methods for estabhshmg storage and shelf-life stability 2 4 Technical characteristics of the preparation 2 4 1. Wettabihty 2 4 2 Persistent foaming 2.4 3 Suspenstbthty and suspension stability 2 4 4 Wet sieve test and dry Steve test 2 4.5 Particle-size distributton, content of dust/fines, attrition, and friability 2.4.6. In the case of granules, sieve test and mdtcations of weight dtstributlon of the granules, at least of the fraction with particle sizes bigger than 1 mm 2 4.7 Content of active substance m or on ban particles, granules, or treated seed 2 4 8. Emulsttiabihty, re-emulsifiability, emulsion stability 2 4.9 Flowabtlrty, pourabthty, and dustabtlity 2 5 Physical and chemical compatrbiltty with other products, mcludmg plant protection products with which its use IS to be authorized 2 6. Wetting, adherence, and distribution to target plants 3. Data on appltcation 3 1 Field of use, e g , field, glasshouse, food or feed storage, home garden 3.2 Details of intended use, e g , types of harmful organism controlled and/or plants or plant products to be protected 3 3 Application rate 3 4 Where necessary, because of test results, any specific agricultural, plant health, and/or environmental conditions under which the product may or may not be used 3.5 Concentratton of acttve substance m material used (e.g., % concentration m the diluted spray)

Registration

in Europe

467

Table 3 (continued)

4

5

6.

7

3.6. Method of application 3.7. Number and ttmmg of applications 3 8. Phytopathogemcity 3.9 Proposed mstructtons for use Further mformation on the preparatron 4 1 Packagmg (type, matertals, size, and so on), and compatibthty of the preparation with proposed packagmg materials 4 2 Procedures for cleaning application equipment 4 3 Re-entry periods, necessary wartmg penods, or other precautions to protect humans and animals 4 4. Recommended methods and precautions concerning handlmg, storage, transport 4 5 Emergency measures in case of an accrdent 4 6 Procedures for destructron or decontammatron of the plant-protectton product and its packaging Analytical methods 5 1. Analytical methods for determinmg the composttron of the plant-protectron product 5.2 Methods for determming residues m or on treated plants, or m or on plant products (e g., btotest) 5.3. Methods used to show microbiologrcal purity of the plant-protectron product 5.4. Methods used to show the plant-protectton product to be free from any human and other mammalian pathogens or, rf need be, from honeybee pathogens 5.5 Techniques used to ensure a umform product and assay methods for Its standardization Efficacy data 6 1. Preliminary range-findmg tests 6.2 Field experimentatron 6 3 Information on the possible occurrence of the development of reststance 6 4. Effects on the quality and, when appropriate, on the yield of treated plants, or effects on the qualny of treated plant products 6 5. Phytotoxrcity to target plants (including different culttvars), or to target-plant products 6.6 Observations concernmg undesirable or unintended side effects, e.g., on beneficial and other nontarget orgamsms, on succeeding crops, other plants, or parts of treated plants used for propagation purposes (e g., seeds, cuttmgs, runners) 6 7. Summary and evaluation of data presented under points 6.1.-6 6 Toxrctty and/or pathogeniclty and mfectivrty studres 7.1 Oral single dose 7 2 Percutaneous single dose 7.3 Inhalation 7 4 Skm and, where relevant, eye irrttation (continued)

468

Neale and Newton

Table 3 (continued)

7 5. Skm sensrttzatton 7.6 Available toxrcologrcal data relating to nonactrve substances 7 7 Operator exposure 7 7 1. Percutaneous absorptron 7 7.2 Lrkely operator exposure under field conditrons, including, where relevant, quantitattve analysrs of operator exposure 8 Residues in or on treated products, food, and feed 8 1 Residue data concerning the active substance, including data from supervrsed trials m crops, food, or feedingstuffs for which authorrzatron for use is sought, grvmg all experimental condmons and details Data should be available for the range of different clrmatrc and agronomic condmons encountered m the proposed area of use It IS also necessary to identify viable and nonviable residues m treated crops 8 2 Effects of industrial processing and/or household preparation on the nature and magnitude of residues, rf appropriate 8.3 Effects on taint, odor, taste, or other quality aspects because of residues on or m fresh or processed products, if appropriate 8 4 Residue data m products of animal origin, resulting from mgestron of feedmgstuffs or contact with bedding, rf appropriate 8 5 Resrdue data m succeeding or rotatronal crops, where presence of residues might be expected 8 6. Proposed preharvest intervals for envisaged uses or wrthholdmg periods, or storage perrods, m the case of postharvest uses 8 7 Proposed maximum residue levels (MRLs) and theJustification of the acceptability of these levels (for toxms), if appropriate 8.8 Summary and evaluation of the restdue behavior on the basts of the data submttted under points 8 1 -8 7 9 Fate and behavior m the environment 9.1. In cases in which toxins are produced, data as outlined under Part A, pomt 9 , are required, rf appropriate 10. Ecotoxrcologrcal studies 10 1. Effects on aquatrc organisms 10.1 1. Frsh 10.1.2. Studies m Duphnza magna, and m species closely related to the target organisms 10 1 3. Studies m aquatic mrcroorgamsms 10.2 Effects on beneficial and other nontarget organisms 10 2 1. Effects on honeybees, rf appropriate 10.2.2. Effects on other beneficial organisms 10.2 3. Effects on earthworms 10 2.4. Effects on other soil fauna 10.2.5. Effects on other nontarget organisms believed to be at risk 10 2.6. Effects on soil mrcroflora

Reglstratlon m Europe

469

Table 3 (continued) 11. Summary and evaluation of pomts 9. and 10. 12. Further mformation 12.1 Information on authorizations tn other countries 12 2. Information on established MRLs in other countries 12.3. Proposals, mcludmgJustification for the classification and labeling proposed in accordance with Directives 67/548/EEC and 78163 l/EEC Hazard symbol(s) Indications of danger Risk phrases Safety phrases 12 4. Proposals for risk and safety phrases In accordance wtth Article 15(l)(g) and (h) and proposed label 12 5. Specimens of proposed packagtng

7. The Next Steps to Harmonization of Data Requirements and Registration of Biopesticides One of the conclustons of the paper commtssroned m 1994 by the Pesticide Forum (1) was that Industry felt that the OECD’s most useful role would be m assisting m the development of public policy by providing information and possibly an international forum to promote international harmomzation of data requirements, test procedures, and criteria for interpreting the results. The OECD survey (I) has gone a long way toward this. The next step, to convmce Member Countries to accept a common approach and mterpretation of results, is difficult. However, action has recently been taken by the Canadian and United States authorities on the harmomzation of guidelines on semtochemtcals and pheromones. This is a very posttive approach, but tt must be taken on a global basis if true harmonization of regulations is to be achieved. Based on the experience of the United States, industry proposed that the tiered system used by the EPA serve as a model for all OECD countries. European regtstration requirements and registration of btopesticides, including genetically modified plants and organisms, would benefit from thts approach (for comparison, see Table 4). However, the draft Directive 91/414/EEC does not incorporate these proposals fully. Industry expressed Its disappomtment to the Commtsston and sent its comments. The next meeting of the OECD Pesticide Forum on the harmonization of test guidelmes was to take place in Autumn 1997. Unless there is a fundamental change in the recommendations and requirements listed in the Annexes IIB and IIIB, little progress 1senvisaged on real harmonization on registration of biopesticides

Residues m food or environment Biological control agents, e g , nematodes Pheromones

Ease of regulations

Exempt

Established m 1984, modified in 1994 to ease notificatron Exempt

IO (mcludmg genetically engineered Bt)

Number of new microorganisms registered m last 24 mo Regulations for GMOs

Time scale for reviews

Tier I sufficient for most bropesticides Attempting to reduce from 8 to 6 mo

Ttered testing procedures

European Union

of Biopesticides

Under review, questions on nematodes No clear policy---considered under 91/414/EEC as chemical

EU system expected 12 mo for acceptance m Brussels, plus addrtronal 12-18 mo for evaluation 0 several m the “queue,” but cannot be registered until system is in place) 9012 19lEEC 90/220/EEC 91/414/EEC Proposal for mclusion in Tier 1

Comments

Canadian requirements also lessened

EU countries are not able to register new organisms independently EU Directives under revrew

Still under discussion with Commission Registration time m EU coun varies from 15 to 24 mo

No specific regulatory body for biopesticides Annex II B / III B of 91/414/l under development

for Plant Protection

EU Standing Commntee/country’s regulatory authontres Currently covered by EC Directive 91/414/EEC, but country requirements differ EU Proposed directly mto Tier II

Union Registration

United States

States and European

EPA’s Biopestictdes and Pollution Prevention Division Subdivisron M and 40 Code of Federal Regulations Part 158

of United

Organizational/ registration unit Gmdelmes/reqmrements

Area

Table 4 Comparison

Registration

in Europe

471

InternatIonal harmonlzatlon, and the provlslon of a uniform set of rules, would not only lower costs without increasing risks, but would also give encouragement to the development of new technologies for rational and sustainable agriculture. References 1 Draft OECD Monograph No 106-Data Requtrements for Reglstratlon of Blopestlcldes m OECD Member Countries Survey Results Paris 1995 2. Neale, M. C (1996) Registration of Blopestlcides m the EU update of the OECD Harmomsatlon Program for Blopestlcides, IBC Conference, London 3 Betz, F. S (1995) Regulation of blologlcal control agents in the Umted States, m Mzcrobzal Control Agents zn Sustaznable Agrzculture, Aosta. 4 Neale, M C (1995) Mlcroblal Pestlcldes wlthm the EC Reglstratlon dlrectlve 91/ 414/EEC-the need for Harmomsatlon, m Mzcrobzal Control Agents zn Sustaznable Agrmlture, Aosta 5. Bode, E. (1995) Authorlsation of blologlcal plant protectlon agents m Germany present status and future prospects, in Mcrobzal Control Agents in Sustatnable Agrzculture, Aosta 6 Llsansky, S G (1994) International harmomsation m blopestlclde reglstratlon and Ieglslation, m Brighton Crop Protection Conference, Farnham, UK

25 Environmental

and Regulatory

Aspects

industry View and Approach Joseph D. Panetta 1. Introduction Since 1947, when the Federal Insecticide Fungicide and Rodentictde Act (FIFRA) was first passed, almost 20,000 pesticide products have been registered, mitially by the US Department of Agriculture (USDA), and, more recently, by the US Environmental Protection Agency (EPA). These products are composed of some 1000 active ingredients, but, of these, only about 150 can be defined as biopestictdes (I). Biopesticides are defined as a group of products that are distmgutshable from conventional chemical pesticides by their nontoxic mode of action, their specificity to target pests, and, most tmportantly, by the fact that they are, or can be, produced by living organisms (2). For the first 50 yr of FIFRA regulation, there were two basic categories of btopesttcides. Microbial pestictdes, including algae, bacteria, fungi, protozoa, and vtruses, are biopesttcrdes that express activity through various modes of action, such as the production of a toxin or by pathogemcity. Biochemical pesticides are not as clearly distmgmshable from conventtonal chemicals as are the microbials, but, in general, are of natural origin and act through a nontoxic mode of action. Some examples of biochemicals are natural plant-growth regulators, which inhibit, modify, or stimulate plant growth; semiochemicals or pheromones, which are emitted by a plant or animal, and effect the behavior of similar or other species; hormones that are synthesized m one part of an organism and translocated to another, where they have controlling or regulating activity; and enzymes or protein molecules that control gene expression or catalyze reactions. From Methods m Botechnology, vol. 5 B/opestmdes Use and Del/very Edlted by. F R Hall and J J Menn 0 Humana Press Inc , Totowa, NJ

473

474

Panetta

Recently, the followmg new categories of btologtcal pesttcrdes have been defined. Nonviable mtcrobtal pesttctdes, first registered m 1991, are produced as live mtcrobtals, and are inactivated prior to then formulatton as biopesticides. The only examples of these actrve ingredients that are currently registered are &endotoxms of Baczllus thuringzenszs (Bt) encapsulated m killed Pseudomonasfluorescens; and plant pesticides, first registered m 1995 m the United States,which are pestrctdal materials produced m plants To date, there are SIX registered active-ingredient plant pesticides in corn, cotton, and potato, all of which are expressed as Bt toxms. The first commercial btopestrcide was registered in the Umted States m 1948, very shortly after FIFRA was passed; this was Bacillus poplllrae, for control of Japanese beetle larvae (3). Thirteen years elapsed before commerctaltzatton of the next mtcrobtal pesticide, Bt var thurzngienszs (later replaced by var kurstakz). The real interest m microbial biopesttcides began m about 1980; smce then, 45 or so mrcrobtal pesttcides have been registered. This figure includes the most widely used active ingredients: various Bts for control of leptdopteran, coleopteran, and dtpteran larvae. The United States market for Bt products for control of pests m fruits, vegetables, and field crops has been estimated to be about $24 mrlhon (4). The successof Bt and Bt-based btopestrcrdes has far exceeded that of other btopesttctdes m terms of use and acceptance. Recently, use of Bts has increased even more as the result of several factors: the discovery of new strams with novel msect activity, such as Bt var tenebraonu, for control of Colorado potato beetle; improvements m fermentation methods, controls, and equipment, Improvements in formulation technology; use of btotechnology to improve potency, spectficrty, and persistence m the field; and development of reststance to synthetic chemicals. Registration of btologicals for control of insect pests has been far more common than that of other apphcattons, such as fungicide or herbtctde use. Recently, new types of mrcrobrals for control of insect pests have been commerctaltzed. For example, the fungal pathogen Beauvaria basszana is now being sold for control of several arthropods, including Mormon crickets, grasshoppers, locusts, and soft-bodied Insects, as well. The first nuclear polyhedroSISviruses (NPVs) for control of insects were regtstered m 1975 for control of hellothis m cotton. Smce then, NPVs have been registered for control of Douglas fir tussock moth, gypsy moth larvae, beet armyworm larvae, and alfalfa looper. These products have experienced fairly limited success,because of several factors, mcludmg a rather slow mode of action, and overlap m target pest activity with Bt products and chemical pesticides. New vn-us-based products are currently being tested in the field in the United States for control of crop pests. Recently, products containing NPV became commercially available for

industry View of Regulatrons

475

control of beet armyworm, celery looper, and codling moth. NPV products have been used for some time in Central and South America, Asia, and Africa, for control of caterpillar pests in vegetables. Current efforts to use genetic engineering techniques to enhance virus activity may lead to greater acceptance in the future (5). In the btochemtcals area, insect pheromones are being used for the most part m cotton, fruit, and vegetables, as mating disrupters. Accordmg to a US Office of Technology Assessment study m 1995, pheromone use in 1995 for control of the pink bollworm in cotton m Arizona was on about 80,000 acres, or onefourth of the state’s cotton crop. Pheromones are also bemg used successfully for control of the apple codling moth-acreage on which one such product, Isomate (Oregon State Umverstty, Corvalis, OR), has been used on about 500025,000 acres from 1991 to 1993, prmcipally because of the development of resistance to tradtttonal chemical pesticides, and as a tool m a mass suppression program (5). Biologtcal herbtcides have tradtttonally presented challenges m development and commerctahzatton, because they do not persist well m the envn-onment, and can be used only m niche markets, makmg research and development costs dtfficult to justify. Mycogen (San Diego, CA) has been undertaking research m biological herbicides since 1985, with tmprovmg success,as these factors are addressed through better formulatton and improved potency One product showing much promise contains Pseudomonas syringae for control of Canada thistle m soybeans and other crops. Regardless of the past limitations of biologtcal pesticides, they are becommg a more sigmficant force m crop protectton. Btopestictdes have been given a major role in efforts to adopt integrated pest management (IPM), because farmers, consumers, and the EPA, concerned about the potential health and environmental effects of conventional chemtcals, are m search of alternatives EPA has recently encouraged the registration of reduced-risk pesttctdes, and biological pesticides m particular, because of then typically low mammalian toxicity and envtronmental impact. The renewed interest in these products has spurred the growth of biopesticide development programs by established producers of conventtonal chemicals, and the formatton of new companies, such as Mycogen, whose sole focus 1sresearch, development, and commerctaltzatton of btological pest-control products. In 1993, the Clinton Admmistratton announced tts support of btopesticide development and commercralization, in committing to intensifying efforts to reduce the use of higher-risk pesticides and promoting IPM, mcludmg btological control. Progress in the registration of biopesticides 1sdirectly related to efforts by the biopestlcide industry and EPA’s Office of Pesticide Programs (OPP), working together to obtain a more effective and efficient product-registration pro-

476

Panetta

cess.The focus thus far by industry has been to identify market niches m which there is a need for a biopesticide because there are no acceptable means of control. In some cases,m which resistance has developed to a synthetic chemrcal pesticide, a biopesticide can fill the need for a control product. Biopesticides have also been used as alternatives to some more acutely toxic chemical products. EPA shares responsibility for field release of biopesticides with USDA’s Animal and Plant Health Inspection Service (APHIS), which is responsible for permittmg field release of nonmdigenous organisms with the potential to be plant pests. While significant progress has been made in streamlining the review processes of both EPA’s OPP and the USDA’s APHIS, but there are opportunities for the biopesticide industry and the two government agencies to work together to develop and more efficiently review data for approval of products. The formation of the Biopesticides and Pollution Prevention Division within OPP/EPA has provided a more effective focus on the uniqueness of btopesticide registration, moving away from what was in the past a process of treatmg btopestictdes m the review process generally as chemicals with fewer data requirements At the same time, there are areas in which improvements are needed, in terms of legislation and regulations, in particular. These aspects are discussed m the following sections m greater detail. 2. Industry’s Progress in Product Development Before movmg into a discussion of the progress of the regulatory process, it is tmportant to consider advances in product development and the current state of the art in terms of available biopesticide products. Why are more biopesticides being used now than in the past, aside from the obvious fact that more are available now? As mentioned, biopesticide products present a number of advantages to users. Chief among these is the fact that bropesticides are typically of low toxicity to mammals and other nontarget organisms. This results m direct and indirect advantages. For example, a direct advantage is that fieldworker exposure concerns are minimized by the application of a biopesticide. An example of an indirect advantage ts that a biopesticide restdue on a crop is normally inconsequential from the standpoint of human dietary exposure, because most biopesticrdes and all microbial pesticides are exempted from the requirements of crop-residue tolerances. When comparing all of the available biopesticide products to available conventional synthetic chemical products, one fact clearly stands out: biopesticide products are not subjected to the complex battery of toxicity and crop-residue tests required to evaluate potential long-term effects, such as carcinogemcity, chronic toxicity, and reproductive effects. In addition, most biopesticides are not subjected to the extensive battery of tests to evaluate persistence and other long-term effects m the environment. These studies can cost $10-20 million,

hdustry View of Regulations

477

depending on the number of crops on which the product is to be applied, making the development of a conventional chemical pesticide a significant cost factor. A positive result m any one test can result in severe restriction of uses or failure to obtain registration. Biopesticides, in contrast, are not nearly as exposed to registration costs or potential failures m the registration process. Mycogen has incurred biopesticide registration costs of <$500,000 to complete the entire tier of tests required by EPA. In terms of the ttme required for safety testing, biopesticide testing can be completed m Cl yr, compared to 3-4 yr for conventional products. The second clear advantage of usmg biopestrcides is that they are specific to target pests. In contrast to broad-spectrum synthetic products, biopesticides typically control a limited spectrum of pests, and have negligible effects on other pests m the field. Although still viewed to some extent as a drawback to the use of biopesticrdes, this has recently been identified as a positive point in terms of maintaining high levels of beneficial insect populations m fields treated with Bt products, and an added factor to be considered in the implementation of IPM programs (6). The broad spectrum of activity of certain synthetic products can cause a flare-up of secondary pest populations as a result of elimmating all Insect predators and parasites from the field. This can result in the need for use of additional application of pestrcides. For example, leafmmer and whitefly species in vegetable crops are typically kept under control by natural predator populations, and these pests are often resistant to many synthetic pesticides because they have been exposed repeatedly to the same products as the target pests. Because they reproduce more rapidly, resistant populations have become established. Biopestictdes are being used to attack thts problem successfully, although resistance to Bt has also become a concern, because it has been demonstrated in the field. For many years, one of the disadvantages of Bt products was the fact that field persistence was so short, and the products typically degraded prior to providing an acceptable level of control. The advantage of environmental compatibility of biopesticides has also worked to their disadvantage because of the short persistence in the field. The development of Mycogen’s CellCap@ encapsulation process has enabled farmers to obtam more reliable results as a result of the enhanced persistence of the Bt toxin contained within the microcapsule (7). By far the greatest boost to improving persistence, while maintaining environmental compatibility, is the mtroduction of Bt toxin genes directly mto crops, where they are able to express the toxin either selectively or throughout the plant, depending on the type of promoter sequences used. This advance in biopesticide technology has provided to biopesticides the opportunity to compete on a level playing field with conventional chemicals in this respect (8).

Panetta

478 Table 1

EPA-Approved CellCap@ Products Product MVP@ biomsecticlde

Actwe ingredient

Pest/crop

F-endotoxm of Bt var

Lepldopteranslcotton, vegetables, tree fruits, vines Coleopteranslpotato, tomato, eggplant, omamentals

kurstab fluorescens

M-Trak@ blomsectlclde

&endotoxm

of Bt var

san dzego encapsulated In kllled Pseudomonas fluorescens

M-Peril@ blomsectlclde

&endotoxm kurstah

MVP@II blolnsectlclde

Mattch blomsectlclde

M/C blomsecticlde

of Bt var encapsulated

European corn borer/corn

m killed P fluorescens 6-endotoxm of Bt var kurstakr encapsulated m killed P jluorescens Blend of CryIA(c) and CryIC &endotoxms encapsulated m killed P jluorescens CryIC 6-endotoxin encapsulated In P jluorescens

Lepldopterans/cotton, vegetables, tree fruits, vines Lepidopterans/vegetables, tree fruits, vines, nursery crops Lepidopterans/vegetables, tree fnuts, vmes, nursery crops

Several examples of biopestlcide registrations follow, including genetically engmeered microbes and transgemc pestlcldal plants Mycogen pioneered encapsulation of Bt toxms with the mvention of the CellCap encapsulation

system in 1985, and has used the process since then to develop advanced Bt formulations. Since 1991, SIX products contammg various Bt toxms produced using the CellCap system have been registered by EPA and foreign regulatory agencies. Most recently, in 1995, EPA approved Mycogen’s Mattch@ biomsecticide, containing a blend of encapsulatedCryIA(c) and CryI &endotoxins for control of leptdopteran pests in vegetable crops. Table 1 lists EPAapproved CellCap products. The Bt toxins in CellCap products are expressed by Bt genes mtroduced mto the P J!uOreScenScell on a plasmid using rDNA technology. The recombinant cells are grown m aerobic, submerged-culture

fermentatton,

and induced to

express the toxin protein at an optimum stage m the fermentation. A chemical fixative IS added after the fermentation IS completed, to rapidly kill and stablhze the cells. The fixative strengthens the cell wall by causing crosslmkmg of components, forming a protective matrix around the Bt toxin. This prolongs breakdown m the field, while providing consistent performance and effective

Industry

View of Regulations

479

control of pests. The CellCap system offers a tremendous range of flexibility m delivering novel Bt toxins to the field. Insertion of Bt and herbicide reststance genes into crops 1sby far the area of greatest potential to the biopesticide industry. It 1sestimated that wtthm 10 yr as much as 50% of all corn seed and 80% of cotton seed will contam Bt genes, but industry efforts are not limited to corn and cotton. Work is m progress to insert Bt genes for control of various insect pests m vegetables, alfalfa, canola, soybean, and sunflower. Work IS also m progress with potato, melon, and squash (9). There has been stgmficant progress in the registration of plants engineered to express pesticidal proteins. Advances m biopesticide efficacy have been made m the areas of better formulations, with prolonged and more consistent field performance, greater specificity to target pests, and field research provldmg more understandmg of pest-host interactions and product-pest mteractions. There are still hmrtations m biopesticide efficacy, which will require research as next-generation products are brought through research and development to registration and commercialization. These limitations include msuffictent understanding of mechanisms of insect resistance, little mformation on the mode of action of microorgamsms, and an inability to formulate products for optimum field activity. Still more understanding is needed to explain interactions between beneficial insects and pests, relative to combmed effects with biopesticide use, and the use of IPM strategies combmmg conventional chemicals, btopesticides, and other control methods (10). 3. A Perspective on Registration 3.1. The Role of EPA Past efforts by EPA to regulate the risks of biopesticldes have focused on estabhshmg a different, abbreviated set of data requirements for field testmg and regtstration of bropesticldes, compared to conventional chemical pesticides. To a large extent, the strategy has been to treat biopesticide data requirements as a subset of chemicals, with added consideration of mfectivlty, pathogemcity, and field persistence. The data requirements for microbial and biochemical pesticides were first established m the late 1970s; they are presented m Title 40, Part 158, of the Code of Federal Regulations. These requirements have evolved to form a tiered approach to registration requirements, the requirement for a higher tier of tests depending on the potential shown at a lower tier for an adverse effect to occur. Guidance for testing of blopesticldes has been provided by the EPA m Subdivision M of the FIFRA Pesticide Assessment Guidelines, and m the Reregtstration Guidance Document for Bt, both issued in 1989. Subdivtsion M has gone through a number of changes, as the types of tests required have been

480

Panetta

determined to be more or less demonstrative of potential effects. Most stgmficantly, the testing has become more relined, to deal with the umque activity of btopesttcides, compared to chemical pesticides. But both the biopesticide registration requirements and the Subdivision M Guidelines are now outdated VISa-vis the new breed of biopesticides. There are no requirements established for killed mtcrobtal pesticides or for plant pesticides. The Bt reregistration document is rapidly becoming antiquated becauseof the many new strains that have been, and are being, registered. In September 1994, EPA issued a Final Rule for Microbial Pesticides; Experimental Use Permits and Notifications. This rule was an amendment of the EUP requirements under 40 CFR Part 172, and was placed mto regulatton pohties that the agency had adopted over several years to deal with the new generation of genetically engmeered and nonindigenous microbes. The essenceof this rule is asfollows: Small-scale (
lndus try View of Regulations

48i

Pesticidal protems expressed m plants represent a novel class of active Ingredients that the agency has not found reason to evaluate m the past. Data requirements must be proposed to permit EPA to address the influx of registration applicattons for these products, and to provide a uniform standard under which data can be developed and evaluated. There are several reasons to approach the risk of such products as bemg relatively low when compared to chemicals and microbes: They are expressed at extremely low levels; they are very target-specific; they are contained withm the plant; and they can be expressed selectively within the plant. Future successm the commercialization of pesticidal plants wtll be m large part determmed by the degree of flexibility adopted by EPA in its regulatory approach to transgemc plants. In October, 1994, OPP established the Biopesticides and Pollution Prevention Division as a pilot program focused on addressing the registration of biopesticides outside of the traditional registration process. The formation of this division was the result of a concerted effort by the biopesticides industry and EPA to provide a streamlined pathway for the registration of biopesticides, and m conjunction with EPA’s effort to encourage the use of such products. This Division has since become a permanent one within OPP. Its successhas been remarkable, having registered at least 20 new active ingredients in the first fiscal year of its existence, and many more since. Further improvements may be achieved by evaluating current regulatory requirements for biopesticides, to determine if the volume of regulatory activity can be reduced by allowing certain amendments and testing to be conducted under notification processes. This would permit limited resources to be focused on more significant regulatory actions. 3.2. The Role

of USDA

Naturally occurring native and nonindigenous microbes are also regulated by USDA/APHIS under the Plant Pest Act and regulations under 7 CFR Part 330. Transgemc organisms are also regulated under the Plant Pest Act and regulations under 7CFR Part 340. Prior to conductmg field release of a microbe or plant, it is necessary to obtain a determination under these regulations as to whether notification or permitting is necessary. It is noteworthy that Part 340 of the regulations is more user-friendly than Part 330. This is because Part 340 of the regulations was written more recently, to address the new breed of transgenic microbes and plants; Part 330 of the regulations was written several decades ago to deal with the potential mtroduction of plant pests mto United States agriculture. Part 340 of the regulations works quite effectively, and has evolved to allow field testing of most transgenic plants to be conducted by notification, and for

482

Panetta

exemptions from regulated status to be granted to allow for commerctaltzatton, but there are no such processes under Part 330. A recent, long-awatted proposal by APHIS to amend Part 330 was rescinded after comments were received that were almost 100% in opposttion to the proposal, which actually increased the universe of organisms encompassed by the regulatton. Part 330 of the regulations is unnecessarily burdensome to researchers, and 1sa waste of resources within USDA. The latter should focus on developing a more streamlined and less bureaucratic method for authortzmg field release of potential btologtcal control mtcroorganisms. 3.3. hprovemenfs in the Registrafion Process EPA and USDA should take a number of stepsto promote future progress m the regtstratton of biopesttcides.The following actions are needed in the short term. 1. More responsibthtyfor evaluation should begiven to the researcher,via the issuance guidelines for self-evaluation of risk 2 Greater streamlining should be built into the regulatory process 3 The EPA should tinahze its Plant Pesticide Policy and establish data requuements 4. There must also be an amendment of USDA/APHIS Part 330 regulattons to correspond more to the mtroductton of nonindtgenous mtcrobes as btocontrol products, rather than potential plant pests, and requtrements for permits based on movement across state boundaries must be changed 5. There must be contmued, increased cooperation between EPA and USDA to make dectstons based on shared data, and based on submtsstons made to one at the two agencies

In the long term, there must also be greater mcenttves offered by EPA for the registration of biopesttcides. These should include the following constderations: 1 An mcreased focus on the evaluabon of benefits of btopesttcldes, so that thts mformatton can be passed on to the user by EPA. 2 More opportunittes to expedtte regtstratton of products of mmtmal rusk, such as exempttons from tolerance and data requirements 3 Consoltdatton of databases, and rehance by EPA and USDA on prtor data to extrapolate to new and stmilar products 4 Incentives for product use, includmg green or safer labelmg to hrghhght products that are of reduced risk. 5. Reduced fees for registratron of biopestictdes, and expansion of the refundmg of tolerance fees for products that meet crtterra for uses that are m the pubhc Interest

4. Future Expectations Perspecttves withtn the blopestlcldes

industry on the future of biopestrctde

product commerctalization are very optimistic. It IS expected that there will be a significant increase in registratton activity for transgemc plants, particularly

Industry View of Regulations

483

m the use of Bt genes, within a IO-yr period. After the process of placmg Bt m the cornmodrty crops is refined, we will see Bt genes in vegetable crops and vines. Beyond this, several companies are working with new insect-resistance genes other than Bt, such as the cholesterol oxidase gene to control weevils and caterpillar pests (12). Vast improvements in formulation technology and fermentation can also be expected. Mycogen is currently improving granular and spray-dried formulation methods; fermentation efficiency has improved more than 100% m the last 5 yr, as well, Improvements will be seen m product potency through the use of genetic engineering to increase expression in both plants and microbes. Combmations of toxins would potentially be engineered into single microbes, rather than blended as they are now, thus providing better pest-control tactics. It is likely that the broad range of microbial products currently under development throughout industry, and in government or university research programs, will provide an increasingly expanding array of biological products for control of various pests.There will be an expansion from Bt products for insect control, to greater use of fungi for control of such pests as crickets, grasshoppers, locusts, termites, and whitefhes. In the plant diseases area, three bacteria are well mto the development stages: Pseudomonasfluorescens and Erwznla herbxola for control of fire blight in apples and pears, and Gllocladlum vu-ens for control of damping-off diseasesin greenhouse seedlmgs for vegetables and ornamental bedding plants (12). In the registration arena, we hope to see more reliance on previously generated data to better handle the increased volume of registration apphcations. Resistance management is also beginning to play a major role in product registration, and we anticipate that this will continue, perhaps not as a formal data requirement, but as a key consideration m registration, Concurrent with the focus on resistancemanagement, we anticipate more focus on IPM, to ensure that both chemrcal pesticides and biochemicals are mamtamed as effective tools, and used at rates that are both safe and that delay the potential for resistance to develop. Finally, as more experience is gamed m the evaluation of rDNA-derived products, we expect to see more flexrbility within the regulatory process, the implementation of an exemption process, and less of a process-based focus, The line between the present and the future is becoming blurred as biopesticide technology moves forward at an ever-increasing rate. Although EPA 1s perfornnng admirably tn the registration of new products, there is a need to keep up with the fast pace that has been mitiated through the development of even more efficient processes of registration. The biopesticide industry will increase the number and types of products available; regulators ~111 face an ever-increasing workload of applications for testing and approval. Cooperation among all concerned parties will be the key to successin the future.

484

Panetta

References 1 Envtronmental Protection Agency, Office of Pesttctde Programs (1997) EPA OPP Pulse, June, 4 2. Matten, C., Schneider, W., Slutsky B., and Mtlewskt, E (1993) Brologlcal Pestztides and the US Envtronmental Protectton Agency, in Advanced Engineered Pestzczdes (Ktm, L., ed.), Marcel Dekker, New York, p 324 3 Matten, C., Schneider, W., Slutsky, B , and Milewskt, E (1993) Btologzcal Pesttctdes and the US Envtronmental Protectron Agency, m Advanced Engineered Pesticzdes (Kim, L., ed.), Marcel Dekker, New York, p 325 4 Mycogen Corporation (1997) Sales of B.t. Products-Products for Agriculture m U S A (Farm Price), Internal Communication. 5 Wood, H. and Hughes, P. (1993) Environmental and commercial evaluations of genetically engineered baculovuus pesticides, m Advanced Engineered Pesttctdal (Kim, L , ed ), Marcel Dekker, New York. 6 U. S Congress, Office of Technology Assessment (1995) Bzologtcally Based Technologzesfor Pest Control, US Government Printing Office, Washmgton, DC, pp. 44-48. 7. Zalom, F and Fry, W., eds. (1992) Food, Crop Pests and the Environment, APS, St Paul, MN. 8. Panetta, J (1993) Engineered microbes. the CellCap@ system, m Advanced Engrneered Mcrobes (Kim, L., ed.), Marcel Dekker, New York, pp 380,38 1 9 Dornbos, D , Jr. (1996) Field assessment of B.t. corn performance. challenge and opportunity, in Commerctaltztng Btopesttctdes, Applred Products and Transgentc Plants-Conference Presentations, International Business Commumcations, Washmgton, DC 10. Biotechnology Industry Orgamzatton (1997) Edttors and Reporters ’ Gurde to Btotechnology, 1997-1998, Washington, DC 11 University of Cahforma, Division of Agriculture and Natural Resources (1992) Beyond Pestictdes, Executive Summary, ANR, Oakland, CA, pp. 33-36 12 PBI Bulletin (1997) National Research Council/Plant Biotechnology Institute, May 13 National Research Council, Board on Agriculture (1996) Btologtcally Based Pest Management, National Academy, Washington, DC

VII MANAGEMENT PROTOCOLS

26 Formulations

of Biopesticides

Susan Boyetchko, Eric Pedersen, Zamir Punja, and Munagala Reddy

1. Introduction A large number of factors can potentially affect the economic feasibility of any given biological control product. These include the impact on the target pest, market size and spectrum of pests affected by the biocontrol agent, variability of field performance, costs of production, and a number of technological challenges, including fermentation, formulation, and delivery systems (I-4). Selection of the appropriate formulations that can improve product stability and viability may reduce inconsistency of field performance of many potential biological control agents (2,5,6). It has been indicated that slow progress in research on formulation and delivery systems is a major hurdle to the development of biopesticide products (I, 7). This chapter summarizes the efforts and successes toward formulation of biocontrol products for use against diseases (biofungicides), weeds (bioherbicides), and insect pests (bioinsectitides). The discussion emphasizes the use of bacteria, fungi, and viruses as the agents. Information on formulation of other important biocontrol agents, such as nematodes, can be found elsewhere (8). Since growers may not be willing to invest in new equipment to apply biological control products, microbial agents must be sold in a product form that is compatible with existing equipment and farm management practices. Compatibility with cultural and chemical control methods, as well as field application systems, are important requirements for the success of biocontrol products (3,7), as well as the need for the agricultural industry to accept and adopt the new technology. The establishment of a modest market share for biopesticide products in the future

From:

Methods F. R. Hall

in Biotechnology, and J. J. Menn,

vol. 5: Biopesticides: eds. 0 Humana Press,

487

Use and Totowa,

Delivery NJ

488

Boyetchko et al.

should prompt the development of new and mnovattve technologies m formulation and dehvery in addition to those already available (9). A number of challenges are encountered in the formulation of brocontrol agents, including good market potential, ease of productron and appbcatron, adequate product stability and shelf life during transportation as well as m storage, and guaranteed propagule viability and efficacy over the long term (7) Some reasons why biocontrol agents have met with limited commercial success are drfticulty of productron, sensitivity to UV light and desrccatron, requirement of high humidity for infection, msufficrent performance over a wide range of environmental condrtrons, and lack of appropriate formulatron (IO). Formulatrons should be used to alter the mrcrobral product to improve product stabthty, btoacttvtty, and delivery (I e , ability to mix and spray the product) as well as to integrate the bropesticrde into a pest management system (II) Other important characteristics of a successful formulatron are convenience of use, compattbility with end-user equipment and practices, and effectiveness at rates consistent with agrrcultural practices (12). For foliar biocontrol agents, envrronmental factors that influence plant mfectton and disease development are temperature, free moisture or dew period, and protection against UV irradiation and desiccation (12,13). For soil-applied biocontrol agents, physical and chemical characterrstics of soil, moisture, and temperature regimens, as well as microbial competmon can all influence efficacy. All of these parameters need to be taken mto consideration when developing an appropriate formulation. 2. Formulation of Bacterial Biopesticides 2.1. General Requirements Bacteria are generally mass-produced using a deep-tank liquid fermentation process, although m some cases they may be more amenable to semisolid or solid-state fermentatron. Nutrient components of the fermentation medium and growth condmons are crrtrcal to both biomass and secondary metabohte production (14). Components of the growth medium should be mexpensrve and readily available. Developing the final formulatron usually requires processmg of the ferment and addition of further components. The end-product can be a solid, hqurd, slurry, powder, or granular. The formulatron should mamtam bacterial viability during transit from manufacturer to retailer and long-term storage (a minimum of 4 mo) (14). Formulation of the biopestrcide also plays a major role m consistency of performance by improvmg or mamtammg bacterial survival following application, A suitable formulatton should provrde a protective habitat for the introduced bacteria, thereby mcreasmg then potentral for survival and successful colomzatron (2.5).

Blopestude

Formulation

489

Bacteria may be formulated m either a dormant or a metabolically active state (14); the former tend to have a longer shelf-life and are more tolerant to temperature fluctuations and chemtcal pestictdes. However, these formulations may be more expensive to produce and require a lag per-rodbefore they become metabohcally active and express beneficial effects. On the other hand, formulatrons containing active cells may be less tolerant to temperature fluctuations, less compatible with chemrcal pesticides, have a shorter shelf-life, and requne specific packaging for gas and motsture exchange, but the bacteria are active at the time of application.

2.2. Formulations Developed Formulations for bacterial biopestrcides may be either liquid or dry Lrqutd formulattons include those that are oil-based, aqueous-based, polymer-based, or combinations thereof. Aqueous-based formulattons reqmre few steps other than fermenting bacteria in a liqutd medium and adding components, such as stabilizers, stickers, surfactants, coloring agents, antifreeze compounds, and additional nutrients (9,164 8) (Table 1). Alternatively, the ferment can be processed (e.g., concentrated or dried) and then resuspended in a liquid medium. The fluid properties of the formulation can be altered by the addition of polymers (e.g., polysacchartdes or derivatives of polyalcohols). O&based formulatrons typically mvolve blending a processed ferment with a mineral or vegetable-based 011carrier and emulstfiers to allow dilution m water. On-based formulattons reduce evaporation of droplets and allow for ultra-low-volume aerial application. Dormant propagules are generally formulated in oil-based and polymer-based hqurds, whereas dormant or metabolically active propagules can be formulated in aqueous-based liquids. Dry formulations, wettable powders, dry flowables, and granulars (mcluding wettable granules) can be produced through such processes as spray drying, freeze-drying, or air drying either with or without the use of a flutdized bed. Wettable and dry granulars are produced by adding binder, dispersant, wettmg agents, and water to the dry powdered ferment in a granulator. The extra processing steps in producing a dry formulation mcrease manufacturing cost, but reduce shipping cost because of the reduced weight. Most dry formulatrons mclude an inert caner, such as fine clay, peat, vernncuhte, alginate, or polyacrylamide beads.The carrier facilitates delivery of the necessary concentration of viable cells in the correct physrologrcal state.Among all of the components that make up a formulation, the carrier occupies the greatest volume and, therefore, often functions as an extender. Effective carriers are mexpensive, easily sterihzed, nontoxic, and consistent in physical characteristics. Moreover, the carrier must ensure both adequatedispersal of the bacteria and performance by protecting the bacteria from adverse envnonmental condttrons.

of Registered

K84

Burkholderza cepacla type Wlsconsm M36 Streptomyces grlseowndls K6 1

P jluorescens NCIB 12089 P qwzgae ESC 10 P syringae ESC 11 Banllus sub& B sub& GB03 B subtlhs GB03

Pseudomonas jluorescens A506 P jluorescens NCIB

Agrobactenum radrobacter A radlobacter

Blocontrol agent

Table 1 Examples

Dry powder Dry powder Dust Peat carrier or liquid Powder or spray

Blo-save lOO/lOOO Blo-save 110 Epic Kodiak, Kodiak HB System 3

Blue Circle

Mycostop

Aqueous suspension of fermentation broth Wettable powder

Post harvest drench, dip or sol1 applied Added to a slurry, mix Added to a slurry, mix Seed treatment m planter box Seed treatment or drip rrrlgatlon Drench, dip, or spray

SPmY

Spray

NA

Conquer

VlCtLlS

Spray, drench

Wettable powder

BlightBan A506

Root dips, drench

Root dips, drench

and Delivery

Petri plate with pure culture Washed plates, culture suspension

of Formulation Dehvery

and Methods Formulation

Biofungicides

Nogall, Dlegall

Galltrol-A

Trade name

Bacterial

Kemna Agro Oy, Pokkalankatu, Fmland

AgBloChem, Inc , Ormda, CA Blo-Care Technology Pty, Ltd , Somersby, NSW, Australia Plant Health Tech, Boise, ID Mauri Foods, North Ryde, Austraha Sylvan Spawn, Klttannmg, PA EcoSclence Corp , Orlando, MA Gustafson, Inc , Dallas, TX Gustafson, Inc Helena Chemical Co, Memphis, TN CCT Corp , Carlsbad, CA

Company

Biopes hide

Formulation

491

Other materials that have been added with the bacteria are diatomaceous earth, adhesive clays, such as talc and vermiculite, cellulose derivatives (e.g., carboxy-methyl-cellulose), and other polymers, such as xanthan gum (29). Techniques for the immobilizatron of bacteria with polymers such as polyacrylamide and sodium alginate, are available (5,19,20). However, slow hydration and release of the active ingredient are major impediments to this technology. Alginate has been successfully used to formulate a variety of bacteria, includmg Pseudomonas spp. (7). Carriers, such as Pyrax or powdered wheat bran, that provide a food base have been mcorporated with the bacterial biomass and alginate. Digat (21) described a new encapsulatton method for bactertal moculants that resulted m a high concentration of bacteria ( lo7 cfu) in a 6 mm granule. The bacteria were suspended in a nutrient broth that caused less nutrttronal stress, and It was suggested that the system enabled the formulatron of several microbial agents or strams (I.e., a cocktail mix). Strmgent quahty control at all stages of manufacturing is necessary to produce a high-quality product. Any variability in the manufacturmg process, whether the result of contamination or inconsistent procedures, can reduce the reliability of the end-product. For example, Bacillus thurmgiensls IS easily produced m liquid fermenters, but production condttions strongly Influence potency of the final product (22). 2.3. Formulation of Bacterial Bioherbicides One of the challenges confronting the use of phytopathogemc bacteria for biological weed control is the requrrement of free water for dispersal and the need for wounds or natural openings for entry of the bacteria mto the plant (23,24). Researchers investigating Xanthomonas campestns pv. poae for control of annual bluegrass (Pea annua L.) have demonstrated that cutting or mowmg of turfgrass will permit the bacteria to enter into the plant (25). In additron, bacteria applied at a rate of log cfu/mL at high water volumes (400 mL/m2) showed over 90% disease severity in the annual bluegrass. One formulatron that has facilitated the penetration and entry of bacterra mto plant stomata and hydathodes IS the organosilicone surfactant Stlwet L-77 (0.2%) (24). To dehver liquid into the stomata of a leaf, a low surface tension of 30 dynes/cm or lower IS required; Stlwet reduces the water surface tension to 20 dynes/cm Application of Pseudomonas syrvzgae pv. tagetis with this surfactant facilitated the penetration and entry of the bacterra mto stomata and hydathodes, resulting in significant increases in disease severrty and incidence m Canada thistle when compared to plants sprayed wrth the bacteria mmus the surfactant (23). It has also been suggested that delivery of the bacteria into these natural openmgs protects them from UV trradtatron and desrccatron. Research on P syrzngae pv. phaseolicola (Psp) for btocontrol of kudzu

492 (Puerarza

Boyetchko et al. lobata [Wllld.] Ohwi) has also demonstrated that formulation with

Silwet L-77 led to higher disease severity in the field (26). Bacteria may be delivered to the soil in a granular form and banded at plantmg or applied as a liquid, but the type of delivery system will depend on the type of crop and farmmg practice m use (Table 2). 2.4. Formulation of Bacterial Biofungicides Bacteria that are registered for use as biofungicides have been recently reviewed by several authors (7,27,28) (Table 3). Several Bacillus-based products are currently being used for disease control and yield enhancement. In China, Bacillus spp. are used to enhance yield of wheat, rice, corn, sugarbeet, rapeseed, turmp, and Chinese cabbage (29). In the United States, the products Epic@,Kodiak@, and Kodiak HB (Bacillus subtilis GB03) are available for use on cotton, legumes, vegetables, and ornamentals to control diseasescaused by Rhizoctonia and Fusarium species. The products are formulated as wettable powders and are compatible with several seed treatment fungictdes (Table 1). Agrobactenum tumefaciens is commercially available in Australia, the Umted States,and New Zealand, and is formulated as a concentrated liquid or a moist peat-basedproduct, or is supplied asa nonformulated agar culture (28,30). Followmg suspensionm water, the bacteria can be applied to seeds,cuttings, roots or root wounds of susceptible orchard and ornamental plants as a dip, spray, or drench. Mycostop@is abiofungicide basedon the bacterium Streptomyces grlseovirtdis K6 I, which is formulated as a wettable powder and is registered in various countries for control of damping-off and root and basal rot diseasesof omamentals and vegetables caused by Fusanum, Phomopsu, and Pythium. The product contams mycelium and spores and can be applied to seed as a dry powder or suspended m water and used as a dip, spray, or drench, and is compatible with a range of insecticides, fungicides, and herbicides. Three products based on Burkholderza cepacza strains are Blue Circle@ (type Wisconsm M36), Deny@(type Wisconsin Iso J82), and Intercept. They are formulated as either lrquids or peat-basedproducts for control of the fungi Fusarzum, Phytophthora, and Pythium, and the nematodes Globodera rostochzensis, Heterodera glycrnes, and Hoplolalmus Columbus. Other bacterial-based biofungicide products are listed in Table 1. 2.5. Formulation of Bacterial Bioinsecticides Most of the commercial bioinsecticides m use today are based on formulations of B thuringiensls Berliner (Bt), an aerobic gram-posmve spore-forming bacterium (31,32) (Table 3). The principle mode of action of Bt biopesttctdes is based on target insect ingestion of the toxic delta-endotoxm protein, which causes feeding inhibition and eventual toxemia to the mid-gut of susceptible

agent

Canada thrstle Annual bluegrass

NA

Campenco

Sicklepod

Pseudomonas syringae pv tagetis Xanthomonas campestns pv poaea

Dodder Silky hakea

LUBOA II NA

CASST

Yellow nutsedge

Dr BtoSedge

Altemaria cassla

Round-leaved mallow

BioMal

Velvetleaf Hemp sesbama

Northern jointvetch

Collego

Velgo COLTRU

Stranglervme

Target weed

DeVine

Trade name

of Formulation

C. coccodes C truncatum

Phytophthora palmwora MWV Colletotnchum gIoeosponoides f sp aeschynomene C. gloeospotioides f sp. malvae Puccmla canabculata ATCC 40199 & C gloeosporioides u

Biocontrol

Table 2 Examples of Registered and Unregistered Fungal and Bacterial Bioherbicides and Methods

NA

Water + sot-bit01 (0 75%) Fungus-infest wheat gluten (PESTA) Water + nonoxynol surfactan~ pan&in wax, mineral 01, soybean oil, lecithin Stlwet L-77, Silwet 408

Granular mixture Granular, wheat bran

NA

Wettable powder

Dry powder

Ltqutd

Formulatton

Japan Tobacco, Kanagawa, Japan

Encore Technologres

Mycogen Corp., San Diego, CA

Ningxia Region, Chma Plant Protection Reasea Instttute, Stellenbos South Africa

Philom BIOS, Sasaktoor Canada NA

Abbott Labs, Chicago, IL Encore Technologies, Mmnetonka, MN

Company

Boyetchho et al

494 Table 3 Examples of Bacillus thoringiensis Bioinsecticides Pathotype of Insect target

B thurmgrenw

Leptdopteran

B thurmgzensu var kurstakz

Dtpteran

B. thurmglensu var waelensis

Coleopteran

B. thurmgzenszs var. sun dlego B thurmglensls var. tenebnoms

Example of commertctal product

Company

Dipel

Abbott Labs

Javelrn, Thuricide Foray, Novo Btobtt, Bactospetne MVP Teknar Skeetal Vectobac M-Track

Ecogen, Langhome, PA Therm0 Trilogy, Baltimore, MD Mycogen Ecogen

Tndent Novodor

Ecogen Therm0 Tnlogy

Abbott Labs Mycogen

larva. Only a few spectes of Insects m the families Leptdoptera, Coleoptera, and Dtptera are suscepttble to the protein. Thus, these btopesttctdes have a relatively narrow msecttcrdal spectrum. Examples of commerctally available Bt-based btopesttctdes are listed m Table 3. These products are formulated as concentrated liquids, otl-based flowables, wettable powders, water dtsperstble granules, and dusts. Several new Bt-based products have been developed using recombmant DNA technology Two products currently available m the United States, MVPTM and M-Trackr”, have been developed using Mycogen Corporatton’s CellCap@ encapsulation process. Thts ts a process whereby a gene encoding the delta-endotoxm protein IS removed from Bt, mcorporated mto a plasmtd, and Inserted mto an isolate of Pseudomonasfluorescens (33,34). The recombtnant cells are grown in aerobic culture and induced to express the delta-endotoxin before being killed through heat and chemical treatment. The dead bacterial cells m the aqueous formulanon serve as mtcrocapsules that protect the fragile Bt toxm from environmental degradation.

3. Formulation of Fungal Biopesticides 3.1. Mycoherbicides Environmental major limitations ease development

condittons, such as temperature and moisture regimes, are to the efficacy of mycoherbtctdes. Moisture required for dtsIS often dependent on the amount of dew period “DeVine,”

Blopes tlcide Formulation

495

the first registered mycoherbtcrde, IS a hquid formulatron of chlamydospores of Phytophthorapalmzvora for control of stranglervine (35,36). The product IS not very stable and there is only 6 wk of shelf life when the product is refrtgerated. “Collego” (Colletotrzchum gloeosporioides f.sp. aeschynomene) IS formulated as dried spores in a wettable powder (37). For control of stcklepod, “CASST” 1sformulated as spores of Alternaria casszaem emulsifiable paraffinic or1 (II) (Table 2). Several adjuvants and other amendments can be used to enhance spore germination, improve pathogen stabihty, and modify the environmental requn-ements or expand the host-range of various mycoherbictdes (II). For example, Colletotnchum truncatum is a host-specific and highly virulent pathogen on hemp sesbama,but requtrement for free morsture has lrmrted its bioherbrcidal potenttal(3 7). Formulating the biocontrol agent usmg unrefined corn 011as an adjuvant significantly enhanced its bloactivity and reduced its dew period requrrement from 12 to 2 h and reduced spray volume requirements from 500 to 5 L/ha (3 7). Surfactants have been explored as ingredients m formulattons because they help to wet the plants by reducing surface tension and they may improve drspersal of the fungal spores m the spray droplet mix. Several surfactants that have been used are Tween 20 with Fusarlum lateritum, nonoxynol wtth Alternaria macrospora, and A. cassiae and sorbttol with Colletotrlchum coccodes (II). Selectton of approprtate surfactants must first Include an evaluation of then- inhibttory or sttmulatory effects on spore germinatton, mfectton, and other aspects of disease development. Use of Invert emulstons (water-in-oil) with foliar fungal btocontrol agents has provided a favorable environment for germination and infection (I, IZ,38,39). The efficacy of C truncatum was significantly improved when apphed with an invert emulston (40) Research wtth Alternarza cassiae indicated that the level of spore inoculum per droplet could be dramattcally reduced from 10-I 00 to 1 spore per droplet to achieve effective control of sicklepod when formulated with an invert emulsion (41-43). However, Invert emulsions are very vrscous and may demonstrate phytotoxicity in some target plants. Conmck et al. (44) developed an invert emulsion with improved water-retention properties that was less viscous. Also, vegetable oils can be used to enhance efficacy of mycoherbtctdes, such as Colletotrzchum orbzculare, for control of spmy cocklebur (45). No phytotoxtctty and improvements m spread of the invert emulsion were observed. Although ltqutd formulations have been primarily used for post-emergence mycoherbictdes, solid-based formulattons have been developed for those mycoherbictdes that infect the weeds at or below the sol1 surface, a system more approprtate for preemergence mycoherbrcrdes (11,46). These

Boyetchko et al.

496

formulatrons can provide a food-base for the pathogen, act as a buffer m environmental extremes, and retain moculum so rt may not be easily washed away A wheat-gluten matrix (liquid moculum, semolina wheat flour, and kaolm) has been used to formulate fungal agents, such as C. truncatum, A. crassa, and Fusarzum lateritwm (47). This formulation has been termed “PESTA” and can be applied aspreemergent and soil-mcorporated treatments. Shelf-life of the product can be improved by manipulatmg the water activity (moisture content of the granule) and sucrosecontent (48). Other solid substratesused to formulate mycoherbicides are bran, wheat kernels, cornmeal/sand, and vermiculite (1,11). For example, mycelmm, micro- and macroconidra and chlamydospores of Fusarlum solanr f.sp. cucurbitae were formulated m cornmeal-sand for control of Texas gourd (49). This pre-emergent granular formulanon provrded 96% control of the weed.

3.2. formulations

of Fungal Biofungicicfes

The environmental conditions drscussedpreviously that limit the efficacy of mycoherbicides, namely temperature and moisture, also affect growth and survival of fungal biofungicrdes. The orgamsms researched as biocontrol agents are primarily filamentous fungi, e.g., Ghocladwm wrens and Trichoderma harzzanum, but there are also examples of some yeast-like fungi, e.g., Pseudozymajlocculosa (50) and Tllletiopsis pallescens (51). The applications of these biofungicides are for control of root-infecting pathogens, e.g., Pythlum, Rhizoctonia, and foliar fungal pathogens, e.g., powdery mildew (50-52) and Botrytzs (28). The formulations that have been developed include granules, pellets, dusts or wettable powders containing spore inocula that are applied directly or as a suspension m water (Table 4). The granular formulations protect against desrccation as well as provide a food base for the fungus, whereas the powders are easily amenable to spraying and provide coverage of large areas. Treatment of seedswith liquids or dusts IS an alternative method of application of these biocontrol agents. In addition, formulatrons of spores m invert emulsions have been tested for yeasts,such as Tilletlopsls (51). The use of alginate prrll was successfully developed to formulate Glzocladium virens (Soil Gard) as a granular formulation for control of root-infecting fungi in pottmg media (7). Similarly, powder or dust formulatrons contammg Trzchoderma with pyrophyllitte clay (Pyrax) have been successfully deployed. Biomass production is generally achieved in large-scale deep tank fermenters contammg appropriate nutrient substrates, and then either used wet or dried prior to formulation (7). Most of the factors that affect product development are similar to those discussed under Subheading 2.2. for bacteria. 3.3. Formulations of Fungal Bioinsecticides Several fungi have been studied as potential biological control agents of insects, and the most well researched include Verticzllrum lecaniz for control of

Fungal

Biofungicides

SoilGard, GhoGard Contans

Aspire

CandIda oleophda l-l 82

T-22G, T-22 HB Btnab T

T. harzanum Rtfat KRL-AG2 T harzianum ATCC 20476 T polysporum ATCC 20475 Gliocladrum wrens GL-2 1 Comothyrmm mwatans

Trichodex

Rotstop Bto-Fungus

Phlebla gzgantea Trlchoderma spp.

6Y.J T. harzianum

Granule or powder

Polygandron

Wettable powder

Granules

Granules

Wettable powder and pellets

Granules or dry powder

Wettable powder

Spores in inert powder Granular, wettable powder

Drip to rock wool

Spores, rmcrogranule

Drench, dtp or spray

Incorporated in soil, soilless mix Soil application

Granules added m furrow broadcast Spray, rruxing with pomng medtum

SPraY

Seed treatment or sot1 incorporated Spray, cham saw 011 Spray or injected

SPraY

and Delivery

Water-dispersible granule

Delivery

of Formulation

AGlO Btotinrgdde Fusaclean

Formulation

with Methods

Ampelomyces qulsqualls MI0 Fusanum oxyspontm (nonpathogenic) @thrum ohgandron

Trade name

of Registered

Biocontrol agent

Table 4 Examples

Prophyta, MalchowPoel, Germany Ecogen

Therm0 Triology

Bio-Innovatton AB, Algaras, Sweden

Natural Plant Protectton, Nagueres, France Vyskumny ustav rasthnnej, Ptestany, Slovak Republic Kemtra Agro Oy, Finland Grondortsmettmgen, St.-Katelijne-Waver, Belgmm Makhteshun Chemical Works Ltd., Beer Sheva, Israel Bioworks, Inc , Geneva, NY

Ecogen

Company

498

Boyetchko et al

aphids, Beauvaria basszana for whiteflies, locusts and beetles, Metarhzzzum jlavovwde and M anlsopllae for locusts, and Lagenldwm gzgantem for mosquito larvae control (53,54). The fungi may be applied directly to the msect as wettable powders, emulsrons or dusts, amended into baits or traps, or added to so11 (55-59). Formulattons are essential to protect against envrronmental extremes of moisture and temperature, as well as to provide protectton from UV damage and desiccation. For example, sunlight, espectally the UV-B component (280-320 nm), was one of the most tmportant factors hmmng survival of B bassiana comdta on foliage (60). Entomopathogens can be applied under tield condmons in or1 at ultralow volumes to increase then efficacy and to protect against UV damage (57,61-63). Sunlight blockers (clay) and UV-Babsorbing compounds (Tmopal) can be added to moculum formulatrons or starch encapsulation (64-66) to increase survival and shelf-life 4. Formulation of Viruses 4.1. General Requirements Baculovtruses have been investigated for control of insect pests belongmg to the Leprdoptera, Hymenoptera, and Coleoptera (67) Thetr advantages are that they are highly specific, do not attack beneficial insects, and can persist m the envrronment, making long-term control of insect pests possible. Examples of baculovnuses are the nuclear polyhedrosrs vu-uses (NPVs) and granulosis vn-uses (GVs). Some limitations of these biocontrol agents are the slow speed of btological activity, their low stabrhty under UV light, and difficulties of production (IO). Stability of the baculovnuses, whrch IS often a function of then viability, IS not a stgmficant problem for small-scale field trials smce the viruses are collected from macerated larvae and mrxed with water and can be stored for short periods through refrtgeratron (67). However, these systems do not lend themselves to large-scale production and applrcatron Formulation of these viruses is an important aspect of product development but has not received as much attention by researchers as the bacteria and fungi. 4.2. Formulations Developed The majortty of the baculovnuses are formulated as concentrated wettable powders (67). The corn earworm (Helicoverpa zea) NPV biocontrol product, “Elcar,” IS either spray- or an-drred after bemg diluted with an mert carrrer Such products as the gypsy moth (Lymantria dispar L.) NPV are freeze-dried etther wtth a carbohydrate or by acetone prectpttatton. Such factors as UV rrradiation, particularly wavelengths of 290-320 nm, can inactivate the virus. Some UV protectants, etther reflectants or absorbers, can be added to formulatrons to stabilize the baculovtruses. Several effective dyes, such as hssamme green, acrrdme yellow, alkali blue, and mercurochrome have been used as UV pro-

Biopes ticide Formulation tectants, especially to absorb UV-A Irradiation (68). Optical brighteners (fluorescent brighteners), such as those commonly used m soaps, detergents, and fabric softeners, also absorb UV light and have been shown to slgmficantly reduce photodegradatlon of NPVs and enhance viral activity (68-72). The precise mode of action of optical brighteners is not known, but research suggests that they interfere with the chltin mlcrofibrils m the paratropic membrane lmmg of the midgut of insects, which ards m the protection of invasion by microbes, such as baculovnuses m Insects. The brighteners act m the insect mldgut and thus affect the host susceptlbllity to the baculovn-us. Optlcal bnghteners may therefore increase the host-range spectrum of baculovlruses 5. Methods for Delivery of Formulated Biocontrol Products Dehvery of products must be easy, economlcal, effective, timely to the appropriate site of action, and compatible with current agronomic practices and equipment. Formulated microbes can be delivered to seed, seed pieces, tubers, cuttmgs, seedlings, transplants, mature plants, or soil; these delivery methods are discussed m more detail below in Subheadings 51.43. 5.1. Seed Treatment For optimal protection of germinating seeds and seedlings against disease, the blofunglcides need to be delivered in a manner that allows the organism(s) to colonize the spermosphere and the developing rhizosphere at a density that IS high enough to suppress the pathogen (73). Biocontrol agents can be precoated or encapsulated onto the seed, mixed with the seed at planting, apphed m-furrow, or incorporated into the soil-mix or seed bed (7676) Precoating of seed usually involves formulations of dry powders or oil- and polymer-based liquids wtth dormant microbes that are capable of survlvmg a period of desiccation (14). Additives, such as xantham gum and gum arable, are sometimes used to increase adhesion of the mlcroblal product to the seed. A speclallzed seed-coatmg process, termed seed encapsulation, involves enveloping the seed, microbe, and possibly other components, such as pestlcldes or mlcronutrients, m a gelatinous or polymer gel-matrix, thereby prolongmg survival of microbial agents on seed. An example of a seed encapsulation product IS GEL-COATTM, which is an alginate hydrogel preparation patented as a delivery system for entomopathogenic nematodes. The seed encapsulation method of delivery has the distinct advantage of user safety and reduced environmental hazard, since the active mgredtents are effectively sealed until they are released during seed germination. Factors to consider m selecting a formulation for coating seeds include inoculum density achievable on the seed, stability of the coating, both for microbe viability and coat integrity, and the feaslblhty and cost of production (12).

500

Boyetchko et al.

Formulations conststmg of fine dusts or powders, wettable powders, or hquids can be applied to seed wtth or without sticker materials at the time of planting. Delivery at the time of planting usually ensures a high number of viable mtcrobes and may allow growers to apply the product directly into the planter box. Drawbacks to this delivery method include possible vartabthty m efficacy resulting from a rehance on the grower’s ability to apply the seed treatment correctly and the extra task for growers. 5.2. Soil Treatment If seed treatment is not a practical option, e.g., if dnect moculatton onto seed 1s harmful to the microbe due to desiccation, or presence of mhibitmg compounds (77), biocontrol agents can be applied to sotl. So11treatment is most effective when the agents are applied as a post-fumigation treatment or at time of planting In sterile soil or growth mixes, colomzatton by pathogens may be reduced by estabhshmg a high populatton of the btocontrol agent. This creates a “suppressive soil,” making subsequent colonization by other less beneficial organisms difficult (7). Dust, powder, and granular formulations can be broadcast and mcorporated into soil, whereas wettable powder, water-dispersible granular, and liquid formulations can be delivered in furrow (14,74,75) So11application may also be a useful method for controllmg overwmtermg pathogen propagules m soil. For example, the product CONTANS@, a water dispersible granular formulation of the hyperparaslte Comothyrium mwutans, can be incorporated mto sot1to reduce the number of sclerotia of Sclerotznuz sclerotiorum.

In greenhouse crops, a simple yet effective method of dehvermg biocontrol agents to soil or growth medium is by direct injection into an irrigation system, such as overhead boom or spaghetti systems.This type of delivery is advantageous in that it allows precise control of the concentration and total volume of mtcrobtal suspension being applied, and requires mmtmal labor to treat large numbers of plants. Multiple treatments of a crop are also possible when existing irrigation equipment is utthzed. The one drawback to this type of deltvery system is that it requires specialized injection equipment and therefore IS not cost-effective unless the grower has equtpment for injectmg liquid fertlhzers Root-colonizmg fluorescent pseudomonad bacteria have been demonstrated to grow on wheat and barley straw, suggesting the posstbility of using crop residues retained under minimum and zero-tillage as a method for delivering them as microbial moculants (78-80). Populattons of lo6 cfu/g straw applied onto barley residues were recovered the following year and were capable of colonizing roots of winter wheat in the year of application (80). Bactertal populations were greater in no-till seeded crops than in conventionally seeded crops, mdtcatmg that cropping systems can influence the acttvity and survival of the

Biopes ticide Formulation

501

soil microbial inoculants. Some factors that should be considered if apphcatton onto crop residues 1spursued are addition of UV protectants and antidesiccants to the formulatton, and application of the inoculum mto the crop residue to maximize the benefits of these residues, which would protect the bacteria from extremes in temperature and motsture.

5.3. Treatment of Plants Biocontrol products can also be applied to plant roots, wounds, and foliage by drenchmg, dippmg, or spraying. Formulated bacteria can be applied directly to roots as a dip or drench (81). Spores of the biofungicide PheIbza gigantea in an aqueous suspenston can be brushed onto freshly cut stumps of pme to prevent entry of Heterobasidzon annosum (82), thereby protecting exposed wounds. Alternatively, spores can be incorporated into cham saw oil so that they are delivered at the same time the tree is harvested. Formulations of bacterta or fungi used as foliar sprays vary accordmg to the crop to be treated, the pest to be controlled, and the anticipated delivery system. The two formulations most commonly used for fohar sprays are liquids and slurries, with the slurries usually reconstituted from either dry or motst carrier-based formulations. Emulsifiers, stickers, spreaders,and other adjuvants and additives aid in application, dispersal and adhesion of the microbes on plant surfaces, and protect the microbes from adverse environmental condmons, such as desiccation, unfavorable pH, and UV radiation (32,83). A broad range of spray application equipment and techniques is available for applying chemical pesticides, mcludmg high volume (1000 L/ha), medium volume (350 L/ha), low to very low volume (3-l 50 L/ha), and ultra low volumes (0.5-3-l L/ha), controlled droplet application, and electrostatic spraymg (58). If biocontrol agents are to be applied using the same techniques, formulations must have the necessary physical properties. Stemke and Akesson (84) found that surface tension and viscosity of the suspension to be sprayed are important factors in reducing droplet size and maintaining the necessary dispersion and control of droplets. Density of the suspension was not an important factor. Successful application of blocontrol agents using different spray techniques has been achieved. For example, Bt-based products have been applied to numerous crops using conventional ground or aerial spraying methods. Highly concentrated ultra-low volume hqutd formulations of Bt-based products have also been used to control insect pests on such crops as cotton and banana (32) and to control spruce bud worm over large areas of coniferous forests (18,32). A low-volume electrostatic rotary atomizer has been used to apply Verticillium lecanii, an entomopathogenic fungus, to successfully control the aphid Aphzs gossypzz.In addmon, ultralowvolume equipment, such as spinning disk sprayers, are now commonly used for application of baculovnuses in forests (67).

502

Boyetchho et al

6. Conclusions and Future Research Although extensive research has led to the tdenttfication of numerous strams of bacteria, fungi, and viruses that can act as potenttal biocontrol agents, one of the major factors that has limited their commerctal successis adequate biomass scale-up and formulation technology. With each mrcrobial agent, there are mherent obstacles to formulation that need to be addressed, including the effect on viabihty of propagules, microbial stability, competitive ability after application, and activity under various environmental conditions. The formulations developed should also be compatible with crop production practices and equipment. Future research efforts m formulation technology should emphasize processesthat will yield optimal mfectivtty of the agent and bioactivlty, as well as achieving viable and stable biological products. Little research has been conducted on methods to promote efficacy of the product after it has been applied to the target pest. In addmon, delivery and application technology of the formulated product need to be addressed. Timing of application as well as placement of the formulated product onto the target pest must be considered to obtam a highly effective btopesticide product. Future research should attempt to develop systematic approaches for selecting formulations based on the characteristics of the biocontrol agent. The potential successof a btocontrol product during the discovery phase should include an evaluatton of the ability to massproduce and formulate the agent. Screenmg of all formulations currently available to select the most effective one should give way to developing criteria that enable researchers to rapidly select classes of formulations based on their determined characteristics and the desired characteristics of the biocontrol agent. Adequate funding of research by the private and public sectors to develop new formulatton technologies and the dissemmation of research findings should considerably enhance the rate at which future developments are made m this area. Collaborations between biologists and chemists, parttcularly m the area of food chemistry and preservatton, would facilitate development of new formulations and apphcations. References 1 Auld, B A and Mot-m, L (1995) Constramts in the development of btoherbtctdes Weed Technol 9,638452 2 Boyetchko, S M (1996) Impact of sot1 microorganisms on weed btology and ecology. Phytoprotectzon 77,41-56 3. Jacobsen, B J. and Backman, P. A (1993) Btologtcal and cultural plant disease controls: Alternattves and supplements to chemicals in IPM systems Plant Dzs 77,311-315 4. Retchelderfer, K (1984) Factors affecting the economic feasibtlrty of the blologtcal control of weeds, m Proceedings of VI International Symposium

Blopesticlde Formulation

503

Control of Weeds (Delfosse, E. S , ed ), Agrrc Can Bull , pp. 135-144 5 Glass, D. J (1993) Commercralizatron of sot1 mrcrobial technologies, m Sozl on Btological

Mtcrobtal Ecology Appltcattons tn Agricultural and Envtronmental Management (Blame Mettmg, F., Jr , ed ), Marcel Dekker, New York, pp 595-6 I8 6 Greaves, M P (1993) Formulatron of mrcrobral herbrcrdes to Improve performance m the field, m Proceedtngs of 8th EWRS Sympostum “Quantttattve Approaches tn Weed and Herbicide Research and Their Practtcal Appltcatton, ”

Braunschwerg, Germany, pp. 2 19-225. 7 Lumsden, R. D , Lewis, J A , and Fravel, D R (1995) Formulatron and delrvery of brocontrol agents for use agamst soilborne plant pathogens, m Btoratzonal Pest Control Agents Formulatton andDelivery (Hall, F. R and Barry, J W , eds ),

ACS Symposrum Series 595, Washmgton, DC, pp. 166182. 8 Georgis, R., Dunlop, D B , and Grewal, P S. (1995) Formulatron of entomopathogemc nematodes, m Biorattonal Pest Control Agents Formulatton and Delzvery (Hall, F R and Barry, J W , eds ), ACS Symposium serves595, Wash-

ington, DC, pp 197-205 9 Boyetchko, S. M (1996) Formulating bacteria for brological weed control, m Proceedings of Expert Commtttee on Weeds, Victoria, Canada, pp 85-87

10 Powell, K A , and Jutsum, A. R (1993) Techmcal and commercral aspects of biocontrol products Pesttcrde Set 37, 3 15-32 1 11 Boyette, C. D , Qmmby, P C , Jr., Caesar, A. J , Brrdsall, J L , Conmck, W J , Jr., Dangle, D. J , Jackson, M. A , Egley, G. H , and Abbas, H K (1996) AdJuvants, formulations, and spraying systems for improvement of mycoherbrcides Weed Technol 10,637-644.

12 McIntyre, J L. and Press, L S (1991) Formulatron, delivery systems and marketing of brocontrol agents and plant growth promotmg rhrzobacterra (PGPR), in The Rhtzosphere and Plant Growth (Kerster, D L. and Cregan, P B , eds ), Beltsvrlle Symposia m Agrrcultural Research, Beltsvrlle, MD, pp 289-295. 13 Barley, K L., Boyetchko, S M., Mortensen, K , and Wolf, T. M. (1997) Brologrcal control of weeds using plant pathogens, m Proceedings of Souls & Crops Workshop, Saskatoon, Saskatchewan, Canada, pp. 205-2 10 14 Paau, A. S (1988) Formulations useful in applying beneficial mrcroorgamsms to seed. TtbTech 6,276279 15 van Elsas, J. D. and Heqnen, C. E. (1990) Methods for the introductron of bacteria into soil: A review. Btol Ferttl Sods 10, 127-133 16 Harman, G E. (1991) Production and delivery systems for biocontrol agents, m New Approaches tn Btologtcal Control of Soil-Borne Diseases (Schoenbeck, F , ed ), Copenhagen, Denmark, pp. 201-205 17 Fages, J (1992) An industrial view of Azosptrtllum moculants formulation and applrcatron technology Symbtosts 13, 15-26 18 Bryant, J E. (1994) Commercral productron and formulatron of Bacrllus thurmgtensts. Agrtc Ecosys Envtron 49, 3 1-35

504

Boyetchko et al.

19 Dtgat, B. (1989) Strategies for seed bactertzatton. Acta Hortzc. 253, 12 l-130 20. Jha, P. K., Nan, S., and Babu, S. (1993) Encapsulation of seeds of Sesbanza sesban with polyacrylamtde and algmate gel entrapped rhtzobia leads to effective symbtotic mtrogen fixatton Znd J Expt Btol. 31, 161-167. 21 Dtgat, B. (1991) A new encapsulation technology for bacterial moculants and seed bactertzatton, m Plant Growth-Promottng Rhzzobacterta. Progress and Prospects Internattonal Workshop on Plant Growth-Promotmg Rhizobacterta Interlaken, Swttzerland, Bulletin SROP, No 14, pp 383-391 22 Stafford, C J (1995) Productton and formulatton of Bt, m Btopestzczdes * Opportutunes for Australian Industry (Monsour, C. J , Reid, S., and Teakle, A E., eds ), Proceedmgs of the 1st Brisbane Symposium, Umverstty of Queensland, pp 78-83 23. Johnson, D R., Wyse, D L., and Jones, K J. (1996) Controllmg weeds with phytopathogemc bacteria Weed Technol lo,62 l-624. 24 Zidack, N K , Backman, P A , and Shaw, J J (1992) Promotion of bacterial mfection of leaves by an organostltcone surfactant tmplications for biological weed control Bzol. Control 2, 11 l-l 17 25 Imaizumi, S , Nishmo, T , Mlyabe, K., FuJlmort, T , and Yamada, M (1997) Btologtcal control of annual bluegrass (Poa annua L.) with a Japanese isolate of Xanthomonas campestrts pv, poae (JT-P482) BIOI Control 8, 7-14 26 Zidack, N K. and Backman, P A (1996) Biological control of kudzu (Puerarta lobata) with the plant pathogen Pseudomonas syrzngae pv. phaseoltcola Weed Scz 44,645649 27 Cook, R. J. (1993) Making greater use of introduced microorganisms for biological control of plant pathogens Annu Rev Phytopathof 31,53-80 28 Elad, Y and Chet, I (1995) Practical approaches for biocontrol implementation, in Novel Approaches to Integrated Pest Management (Reuvem, R , ed ), Lewis, London, pp. 323-338 29 Shouan, Z Weimin, X., Zhinong, Y , and Ruhong, M (1996) Research and commercialization of yield-increasing bacterta (YIB) m China, in Advances of Bzological Control ofPlant Diseases (Wenhua, T , Cook, R J , and Rovira, A , eds ), Proceedings of the International Workshop on Biological Control of Plant DISeases, BeiJmg, Chma, pp 47-53 30 Kerr, A. (1980) Btological control of crown gall through production of agrocm 84 Plant Du

64,25-30

31. Cannon, R. J. C (1993) Prospects and progress for Bactllus thunngzenszs-based pesticides Pestic. Set 37, 33 1-335 32 Shteh, T R (1995) Btopesticide formulatrons and their applications, m Proceedings of American Chemrcal Society (Ragsdale, N. N , Kearney, P C , and Plimmer, J R., eds.), Eighth Internattonal Congress of Pesticide Chemistry Options 2000, Washmgton, DC, pp 104-l 14 33 Gaertner, F. H., Quick, T C., and Thompson, M. A. (1993) CellCap: an encapsulation system for msecttctdal btotoxin proteins, m Advanced Engzneered Pestrczdes (Kim, L., ed.), Marcel Decker, New York, pp. 73-83. 34. Panetta, J. D (1993) Engineered microbes: the CellCap system, in Advanced Engrneered Pesttctdes (Kim, L , ed.), Marcel Decker, New York, pp. 374-382

Biopesticide 35 36 37

3x.

39

40

41 42

43

44.

45 46 47. 48.

49. 50.

Formulation

505

Burnett, H C , Tucker, D P H., and Rldings, W. H. (1974) Phytophthora root and stem rot of milkweed vine. Plant Du Rep 58,355-357. TeBeest, D. 0 and Templeton, G. E. (1985) Mycoherblcide: progress m the blologlcal control of weeds. Plant Du 69,6-10. Boyette, C. D. (1994) Unrefined corn oil improves the mycoherbicldal activity of Colletotrlchum truncatum for hemp sesbama (Sesbanza exaltata) control Weed Technol 8,526-529 Conmck, W. J., Jr, Lewis, J A., and Quimby, P C , Jr. (1990) Folmulatlon of blocontrol agents for use in plant pathology. UCLA Symposmm m Molecular and Cell Biology, pp. 345-372. Daigle, D J , Connick, W. J , Jr., Qulmby, P C , Jr., Evans, J., Trask-Morrell, B , and Fulgham, F. E (1990) Invert emulsions* carrier and water source for the mycoherbiclde Alternarla casslae. Weed Technol 4,327-33 1a Boyette, C D , Qutmby, P. C., Jr., Bryson, C. T., Egley, G. H , and Fulgham, F, E (1993) BIologIcal control of hemp sesbania (Sesbanla exaltata) under field conditions with Colletotrzchum truncatum formulated m an invert emulsion WeedScz 41,497-500 Walker, H L. and Boyette, C. D. (1985) Biocontrol of sicklepod (Cassla obtusifoba) m soybeans (Glyczne max) with Alternarra casslae. Weed Scl 33,2 12-2 15. Amsellem, Z , Sharon, A, Fressel, J., and Qulmby, P. C., Jr (1990) Complete abolition of high inoculum threshold of two mycoherbicldes (Alternarza casslae and A crassa) when applied in invert emulsion Phytopathology 80,925-929 Dangle, D. J. and Conmck, W J , Jr (1990) Formulation and application technology for microbial weed control, m Microbes and Mlcroblal Products as Herbicrdes (Hoagland, R. E , ed.), ACS Symposmm series 439, American Chemical Society, Washington, DC, pp 288-304. Conmck, W. J , Jr, Dangle, D J , and Qulmby, P. C , Jr (1991) An Improved invert emulsion with high water retention for mycoherblclde dellvery. Weed Technol 5,442-444 Auld, B. A (1993) Vegetable 011suspension emulsions reduce dew dependence of a mycoherblclde. Crop Prot. 12,477-479 Dangle, D. J and Cotty, P. J (1992) Production of comdla of Alternarza casslae with algmate pellets Blol Control 2,278-28 1 Conmck, W. J., Jr., Boyette, C. D., and McAlpine, J R (1991) Formulation of mycoherblcides using a pasta-like process. Bzol Control 1,281-287. Conmck, W J., Jr., Dangle, D. J., Boyette, C. D., Williams, K. S., and Vmyard, B. (1996) Water activity and other factors that affect the vlablhty of Colletotrzchum truncatum comdia in wheat flour-kaolin granules (‘pesta’). Biocontrol Sci Techno1 6,277-284 Boyette, C. D., Templelton, G. E., and O&r, L. R. (1985) Texas gourd (Cucurblta texana) control with Fusanum solam f.sp. cucurbitae. Weed Scz 32,649-654 BClanger, R. R. and Benyagoub, M (1997) Challenges and prospects for mtegrated control of powdery mildews in the greenhouse. Can J. Plant Path01 19, 310-314

Boyetchko et a/

506 51 Urquhart,

E. J and PunJa, Z. K (1997) Eplphytlc

growth and survival of

Tlllettopsu pallescens, a potential blologlcal control agent of Sphaerotheca jiilzglnea, on cucumber leaves. Can J Bot. 75, 892-90 1. 52 PunJa, Z. K (1997) Comparative efficacy of bacteria, fungi, and yeasts as bIologIcal control agents for diseases of vegetable crops Can J Plant Path01 19,3 15-323 53 Lacey, L A and Goettel, M S. (1995) Current developments m mlcroblal control of Insect pests and prospects for the early 2 1st century Entomophaga 40,3-27 54 Bateman, R (1997) The development of a mycomsectlcide for the control of locusts and grasshoppers. Outlook Agrlcult 26, 13-l 8. 55 Feng, M G , Poprawskl, T J , and Khachatounans, G G (1994) ProductIon, formulation and appltcatlon of the entomopathogemc fungus Beauverza basslana for insect control: current status. Blocontrol Scl Technol 4,3-34 56 Goettel, M S , Johnson, D. L , and Inghs, G. D. (1995) The role of fungi m the control of grasshoppers. Can J Bot 73,571-575 57 Inglls, G D , Johnson, D L., and Goettel, M. S (1996) Effect of halt substrate and formulation on mfectlon of grasshopper nymphs by Beauverla basslana Blocontrol Scl Technol 6,35-50 58 Auld, B. A (1992) Mass productIon, formulation and apphcatlon of fungi as blocontrol agents, m Biological Control of Locusts and Grasshoppers CAB

International,

Wallmgford,

UK, pp 2 19-229

59 Caudwell, R. W. and Gatehouse, A G (1996) Laboratory and field trial of halt formulations of the fungal pathogen, Metarhzzlzlmf2avovw~de, agamst a troplcal grasshopper and locust Bzocontrol Scl Technol 6, 56 l-567 60 Daoust, R A. and Perena, R M (1986) Stability of the entomopathogemc fungi Beauverla basslana and Metarhlzzum anisopllae on beetle-attractmg tubers and cowpea fohage in Brazil. Envwon Entomol 15, 1237-I 243 61 Moore, D , Bridge, P D , Higgms, P M., Bateman, R P , and Prior, C (1993) Ultra-violet radiation damage to Metarhlzlumflavovwide comdla and the protection given by vegetable and mineral 011sand chemical sunscreens Ann Appl Blol 122,605-615. 62 Bateman, R P , Carey, M , Moore, D., and Prior, C (1993) The enhanced mfect&y of Metarhzzrum flavovwlde m 011 formulations to desert locusts at low humidities. Ann Appl Blol 122, 145-152 63 Moore, D., Bateman, R. P , Carey, M., and Prior, C (1995) Long-term storage of Metarhmumflavovwlde comdla m 011formulations for the control of locusts and grasshoppers. Blocontrol Scl Technol. 5, 193-199 64 McGmre, M R. and Shasha, B. S. (1992) Adherent starch granules for encapsulatlon of insect control agents J Econ Entomol 85, 1425-1433 65 Perelra, R M. and Roberts, D W (199 1) Alginate and cornstarch mycehal formulatlons of entomopathogemc fungi, Beauverla basslana and Metarhlzlum amsoplzae. J Econ Entomol 84, 1657-1661 66 Caudwell, R. W. and Gatehouse, A G (1996) Formulation of grasshopper and locust entomopathogens in halts using starch extrusion technology Crop Prot 15,33-37. ‘*

Blopes trcrde Formulation

507

67 Cory, J. S. and Bishop, D H. L (1995) Use of baculovnuses as btologtcal rnsectrcldes, m Methods in Molecular Biology, vol 39: Baculovu-us Expression Protocols (Rtchardson, C D., ed ), Humana, Totowa, NJ, pp. 277-294 68 Shapiro, M. (1995) Radtation protection and activity enhancement of vuuses, m Blorattonal Pest Control Agents. Formulation and Delzvely (Hall, F R. and Barry, J. W., eds.), ACS Symposium Series 595, Washmgton, DC, pp 153-164. 69 Shapno, M. (1992) Use of optrcal brighteners as radiation protectants for gypsy moth (Leptdoptera.Lymantriidae) nuclear polyhedrosts VKUS J Econ Entomol 85, 1682-1686 70 Shapn-o, M. and Robertson, J. L (1992) Enhancement of gypsy moth (Lepldoptera: Lymantrndae) baculovirus actrvtty by opttcal brighteners. J Econ Entomol 85, 1120-I 124 71 Dougherty, E. M , Guthrte, K., and Shapuo, M. (1995) In vrtro effects of fluorescent brightener on the efficacy of occluston body dtssolutton and polyhedralderived virions Bloi Control 5,383-388 72. Erlandson, M A and Moore, K. C. (1994) Enhancement of bertha armyworm baculovuus acttvrty by an optrcal bnghtener compound, m Proceedings of 42nd Annual meeting of the Entomological Society of Alberta, Canmore, Alberta, Canada, p. 11 73 Cook, R J and Baker, K R (1983) The Nature and Practice ofBlologlca1 Control ofPlant Pathogens American Phytopathologtcal Society, St Paul, MN, 539 pp. 74, Kommedahl, T. and Wmdels, C E. (198 1) Introductton of mtcroblal antagomsts to specific courts of mfectton seeds, seedlmgs, and wounds, m Bzocontrol zn Crop Production (Papavlzas, G C., ed.), BARC Symposium 5 Allanheld and Osmun, Totowa, NJ, pp 227-248 75 Lewis, J A (1991) Formulatron and delrvery systems of brocontrol agents with emphasis on fungi, m The Rhzzosphere and Plant Growth (Keister, D L and Cregan, P B , eds ), Kluwer, Dordrecht, The Netherlands, pp 279-287 76 Thomashow, L S and Weller, D. M (1990) Applrcatron of fluorescent pseudomonads to control root diseases of wheat and some mechamsms of drsease suppression, m Blologlcal Control of So&Borne Plant Pathogens (Hornby, D , ed.), Redwood, Melksham Wrltshire, UK, pp 109-122 77 Gindrat, D. (1979) Biocontrol of plant diseases by moculatron of fresh wounds, seeds and so11with antagomsts, m So&Borne Plant Pathogens (Schrppers, B. and Grams, W., eds ), Academic, New York, pp 537-55 1. 78 Elhott, L F and Lynch, J. M (1985) Plant growth-mhrbrting pseudomonads colomzmg winter wheat (Tntrcum aestlvum L ) roots Plant Sod 84, 57-65. 79 Fredrrckson, J. K , Elliott, L F , and Engibous, J C (1987) Crop residues as substrates for host-specrfic mhrbltory pseudomonads. Soul Blol Blochem 19, 127-l 34 80 Stroo, H F , Elliott, L. F., and Papendtck, R I (1988) Growth, survrval and toxm production of root-mhtbttory pseudomonads on crop restdues. Socl Blol Blochem 20,201-207 81 Funk, L M , He, D N , Pedersen, E A., and Reddy, M. S (1997) Opttmizatton of product delivery for a mrcroblal inoculant, Burkholdena cepacza, for commercral use m the forest Industry (Abstr ). Can J Plant Pathol 19, 108

508

Boyetchko et al.

82 Rtshbeth, J (1975) Stump moculatron: A brologrcal control of Fomes unnosus, m Btology and Control of Soil-Borne Plant Pathogens (Bruehl, G W , ed ), Amertcan Phytopathologxal Society, St. Paul, MN, pp. 158-162 83. Harvey, L T. (1991) A Guide to Agrzcultural Spray AdJuvants Used In the United States Thompson Publicattons, Fresno, CA 84 Stemke, W. E. and Akesson, N. B (1993) Atomrzatron of btopesttcide formulatrons, m Pestlcrde Formulations and Applzcatlon Systems, Volume 12, ASM STP 1146 (Devlsetty, B N., Chasm, D. G., and Berger, P D , eds.), American Soctety for Testmg and Matertals, Phrladelphra, pp. 257-27 1

27 Delivery Systems and Protocols

for Biopesticides

Roy Bateman 1. Introduction I, I. Conditions for Successful Biopesticide Development Biopesticides have little raz~on d’&tre unless they are blologlcally specific. Their perceived advantage of mammalian safety over chemicals has been eroded by new developments in pesticide chemistry (I) and with possible rare exceptions, blopesticldes ~111be targeted at “niche markets.” Research and development, therefore, will be reliant to a greater or lesser extent on public support, but unfortunately microbial research is usually funded piecemeal. Multidisciplmary teams are uncommon, but where opportunities have arisen to form such teams the results have often been rewarded with success, where “success” could be defined as: scientific novelty or elegance (with outputs in the scientific literature) or lmplementatlon with commercial products available for use. Although a successful outcome will be biology driven and must depend on sound science, the ultimate test must be the latter: we are in the business of providing technical solutions, not the production of “better mousetraps.” The opportumties for blopesticide development will be described in detail m this section of the volume; the greatest scope appears to be in the following broad categories (2): Treatment of natural and seminaturalhabitatsin which conservationof blodiversity IS important (e g., pestmanagementin forestsand rangelands), Crops subJectto public pressurefor high ecologlcal and toxicological standards (e.g., organic food crops, reduction of pesticideresidues); Substitution for chemicalapplicationsdeemedto be unsatisfactory(e g , msecticlde resistancemanagementstrategies); Situationsm which very low mammaliantoxlclty is crucial (e.g., storagepests); From

Methods

m EQotechnology,

F R HallandJ

J Menn,eds

vol 5 B/opesbodes OHumanaPress,Totowa,

509

Use and Dekvery NJ

Bateman

510 S

a componentof integratedpestmanagement(IPM) strategieswhere preservation of natural enemiesis of known importance, or where legislation has been enactedto reducepesticideuse. As

Many pest-pathogen interactions have been known for a long time (e.g., more than a century m the case of Metarhizum 131) and new assoctattons continue to be discovered. The development of a new btopestrcide product may follow a “break-through” m, for example, cost efficiency m mass production (4) or formulatton (5) combmed with an identifiable niche m the market How‘ever, the key component of a bropestrctde must be the brological mteractron between the target pest and a virulent, appropriately stable pathogen propagule. Many microbial pesticide projects are now publicly funded for lrmtted time pertods (typically 3 yr), wrth further support contingent on adequate progress having been made. The major elements of a three-phase program are shown in Fig. 1, with a shaft m emphasis from sctentific research to product development. An mmal techmcal idea leads to a “divergent” stage of thmkmg (usually accompanied by the development of new techniques), whtch may be followed by a consoltdatton or “horizontal expansion” phase (6). Projects often fat1 to go beyond this stage because msufftcrent progress 1smade to show that laboratory data can be translated to field efficacy. Further success will result m a third implementation phase charactertzed by focusmg efforts on the development of product(s) often m collaboration with commerctal partners. From a screnttftc point of view the third “convergent stage” may seem like a “dead end”. Dtfferent skills (especially commercial expertise) are required for product development Biopesticides will only be of interest to agro-mdustry if development costs are very low and commercial partners may also require preltmmary mammalran toxicity data at an early stage m development (7). Several key aspects can make or break a biopesticide project during the thtrd phase, including the development of economic mass productton of the pathogen and an application strategy. 1.2. Scope of, and Definitions

Used in, This Chapter

The crucial gap between microbial research and product development has to be bridged by sctenttsts and technictans involved m “practical vertficatton” the turmng of scientific ideas mto (prototype) practical solutions This chapter attempts to focus on appropriate development of formulation and apphcatton techntques, referred to here as the “delivery system.” I will describe specrfitally some cost effective, step-by-step techmques that constitute an important transition between laboratory assaysand small-medium scale field trials (highlighted m the center of Fig. 1). Such work appears to constitute a “no man’s land” m the scientific literature.

_

Identtficatron of other target pests

Biology of pest,

Stram - tsolatton - screenmg - selectton 1 Isolate charactertsatton/ j Understandmg pathogentctty

Identtficatron of pest -+ pathogen mteracttons

Field rates

Environmental studies

Large scale field tests

Medium scale field tests

Pilot productton plant

Application strategy

Dose trmfer

l3utc3type fomulatian

Early productton methods

Practical verification

Further formulation development Market study Integration with other control strategies Mass production Final formulatton Confirmatton of field apphcatton rates Complete safety testing (formulation) Demonstration/partxtpatoty trials Regutratton Cotnmercialisatron

Inmate mrcrobtal storage tests Mammalran safety tests (isolate)

Product development

Ftg 1 Some major elements m the development of a btopesttctde (highlighted central area shows the mam theme of this chapter)

Phase 3 (convergent)

Phase 2 (horizontal expansion)

Phase 1 (divergent)

Biological/scientific research

512

Bateman

Many of the examples used in this chapter have been gamed from expertence m the LUBILOSA* project: an international, collaborative, multidonor, and multtdisctplinary program dedicated to the biological control of locusts and grasshoppers using mycopathogens (especially Metavhizzum sp ) This will be perhaps of greatest interest to those working on pest management projects for natural and semi-natural habitats (see Subheading Ll.), where an tmportant technical challenge is to treat large areas of land at a high work rate, hence the use of ultra-low volume (ULV) spraying techniques at >3 L of formulated product per hectare. This chapter will deal mostly with true biopesticides; products that contam living organisms that have the capacity to reproduce once delivered mto the environment, and require a delivery system that maintains the viability of the pathogen Other “biorationals,” including Bacdlus thunngzenszs products m which the bacterium has been killed, can be tested m the same way as chemtcal formulations, and are outside the scope of this chapter. Maintaining pathogen virulence in true biopesticides mvolves careful mass production and formulation techniques that are described m this book and elsewhere (8), but some issues will be mentioned briefly in Subheadings 2. and 3. 2. Biological Agents, Their Formulation, and Application Many btopesticide delivery systemsuse techniques originally developed for conventional chemicals, so first we need briefly to examine some of the ways m which these agents are similar to, and contrast with, one another 2.1. Differences Between the Delivery of Chemicals and Biopesticides All hvmg agents at least have the potential for multiplying in the environment and ecologists have attempted to quantify this horizontal transmission or secondary cycling with the aid of models (!Wl). The ways m which biopestitides act m the field may constitute a contmuum between activity similar to that of a slow-acting chemical active ingredient (e.g., B. thunngiensis) on one extreme and a true biological control agent on the other (2,169. The effective mode of action in the field will profoundly affect the preferred delivery system. Hails (11) has classified these as: 1 (Conventional) blopestudes, m which the agent IS apphed lrke a chemical for (relatively) immediate action One of the major priorities for the delivery system may be to enhance speed of kill and an obvious solution 1s to maximize dose transfer to the target (see Subheading 3.2.) *LUtte BIologlque centre les LOcustes et les SAutereaux a research and development program funded by the governments of Canada, the Netherlands, Switzerland, and the United Kmgdom

Delivery Systems and Protocols for Biopesticides

513

2. Blopesticldes with a longer term action. immediate control followed by longer term (wlthm season) actlon provided by secondary cycling 3. ClassIcal biological control, which attempts to achieve long-term suppresslon of pests over several generations (and across seasons) In this case a dehvery system that achieves a high transfer of propagules to target insects by direct impact may be unnecessary or even deleterious to the estabhshment of the pathogen (9) Blopesticides of types 2 and 3 therefore increasingly contrast with chemical pesticides with accompanying advantages and disadvantages. On the one hand biological insecticides may persist in cadavers, and propagules can have much greater persistence in the field than would be indicated by laboratory bioassays; on the other, they are subject to environmental influences, such as temperature and humidity It IS important to note that these three categories represent points on a contmuum and that certain biopesticlde products may act m more than one way under different environmental conditions; thus the same o&based formulation of Meaarhzzium may act like a conventional pesticide m arid conditions but have a longer term action against acridids in moister, riverine, or tropical

environments

(2,10),

2.2. What Can We Learn from Chemical Application Techniques? In the commercial world, the implementation of biopestlcides is most likely to come about with the development of products that will be applied to a greater or lesser extent with the aid of conventional application equipment. In the short term we must assume that efforts to improve dose transfer and work rate will be rewarded by improved efficacy and cost effectiveness of these products Much useful information can be gained from understanding chemical dose transfer m the field, which will be helpful for selecting a delivery system; for

example whether the mode of acquisition of a contact insecticide 1sby du-ect contact with spray droplets or by secondary pickup of spray residues on vegetation. Direct contact effects may be quite similar to those conventional insect]-

tides where, for a given amount of active ingredient, msecttcide effectiveness 1s inversely proportional to drop size (12,13); this has been shown to apply to B tlauringienszs(1415). Secondary pick-up of spray residues is important with contact insecticides against locusts and It may be beneficial further to develop formulations that contain a viscous nonevaporative component (16). The presence of a nonevaporative component will limit the reductions m the droplet size (which would reduce impactlon efficiency on leaves), and may enhance secondary pickup from leaf surfaces if the final deposit 1sviscous; however, the interactions involved are complex. Biological agents that act after mgestion (i.e., equivalent to stomach poisons) will be influenced by attractant or repellent properties m the formulation and by “coverage” (a term that 1soften used loosely, but here meaning the probablhty that a pest will encounter a pes-

514

Bateman

ticide). Efforts are now bemg made to model the relatronships between apphcation systems and biologtcal results wtth bropestlcides and btorationals (17). A notable characteristic of biological pestictdes ISthen slow speed of action relative to chemicals, which often is perceived as a disadvantage. One solution IS to enhance speed of action by genetic engineering (I&, but this is controversial and still does not create the very fast knock-down perceived by some to be necessary for certain insect pests. Genetic engineering of entomopathogemc baculovnuses has recently received Interest from certain large insecttctde manufacturers, since then intellectual property investment can be protected readily and then efficacy IS considered to be most like chemicals (19). However, not all chemicals are fast acting and much could be learned about testing, deployment, and marketing from the manufacturers of new, slow acting, specific molecules, such as the substituted ureas (Insect growth regulators). 3. A “Step-by-Step” Approach to Biopesticide Product Development It is clear, therefore, that certain adaptations may be necessary to standard techniques for chemical pesticide development to also be used for btopestrclde products. The special problems of evaluating biopesticides are often assoctated with thetr slow speed of action (e.g., how to monttor mobile target pests), very high virulence and complications arising from their capacity to reproduce (e.g., handling control mortality). Testing techniques must also be cost effective m the agrrcultural sector because of the limited market for any final products. In contrast, large, publicly funded projects (often for pest management m semi-natural habitats such as forests) provide opportunmes for useful developments m sctentlfic techniques and detailed examination of the complex abiotlc and btotic mteractions that take place m biopesticide mterventions. For example, Evans (20) describes the research required on the pest, crop, environmental condrtions, pathogen, and its application m order to provide an effective and robust “control window”; this work included improvements to spray apphcatton techniques for entomopathogemc viruses 3.1. Laboratory Assays Selection of isolates IS usually carried out using relatively simple laboratory screening assays,where a single dose is applied. Because of the variable nature of pathogens rt is advisable to screen against a standard isolate as well as including controls. B. thurzngienszsis the only pathogen for which standardrzed systems for assessmentof pathogenicity have been developed (21). Dose-response analyses can be carried out for specific days after moculanon, and can reveal significant differences with various formulations and isolates However, much work needs to be done to relate such differences to effects

Delivery Sys terns and Protocols for Biopes ticicies

515

m the field. With mycomsectrcides, for example, differences m dose of less than an order of magnitude only have a very small biological effect (522). Btopesticide assaysusually give no clear end point, so recent techmques have included the fitting of a two-dtmenslonal model (for time as well as dose) to explain mortality data (23,24). The selection of an isolate for development will often be based on a comblnation of properties, including vnulence, productron characteristics, mammalian toxicity, and environmental stability (usually simple tests for heat and UV tolerance). However, m practice, research groups working on naturally occurring pathogens have found it only rarely worthwhile to swnch from standard tsolates, these are identified at the beginning of a project (25). A comphcatmg factor IS that samples isolated from a smgle diseased insect can have remarkably different virulence (II) and production characteristics (26). 3.2. Choice of Application 3.2.1. Ground Rules

Equipment

and Spray Parameters

If novel application equtpment is a prerequisite for the use of a novel biopestictde, then the hkehhood of that product being widely used is greatly reduced (27). Although the ubiquitous hydraulic nozzle has many shortcomings, there are no wtdely used, acceptable alternatives for field crops. For example, the use of techniques to manage glass-househumidity for the effective use of Verticilliunz lecanii (28) has proved more acceptable than promising developments m electrostattc application of this pathogen (29) (m a relatively specialized and high value market). “Novel” dehvery systemsbecome acceptableonly if special factors, such as inoculative release or cessattonin the use of key chemtcals,create a special niche m the market. The latter might be causedby political (envn-onmental) pressure or technicalnecessity(e.g.,reststance,secondarypestresurgence).An extremeexample of this IS the useof Beauvena brongnlartii in Switzerland, where aenal applications of aqueousblastospores(as a raw fermentation product with adJuvants)are sprayed onto swarming adult (egg laying) Melolontha melolontha beetles by helicopter at >lOO L/ha (30). This very expensive form of application can beJustified in order to achteve timely moculatron of a pathogen that develops epizootlcs tn semmatural environments of a mountamouscountry with an ecologically aware populace 3.2.2. Rotary Atomizers: Appropriate Deployment and Tools for Research Hydraulic spraying IS notortously inefficient (31) and many appltcatton research efforts have attempted to use less pesticide by applying it more efficiently (32). Of all the controlled droplet application (CDA) techniques, the use of rotary atomizers has been most widely adopted. One of the most conspicuous exceptions to the use of hydraulic sprayers has been m ULV applica-

516

Baternan

tion of bropesttcrdes for the treatment of natural and semmatural ecosystems (e.g. forestry [20,33,34J and locust control [35]). Lowermg volume application rates has necessitated detailed analysts of formulatton, droplet size, and recovery (36). Parkm and Merritt (3 7) emphasized the need to establish protocols for evaluating such issues as spray drift and to build a comprehenstve database, but unfortunately, despite maJor initiatives, such as the Spray Drift Task Force in the United States,little of this information 1sm the pubhc domain. Although the final ObJectivemust be to adapt btopesttctdes to existing applrcation practrce, small spinning disk sprayers have been used for a wide range of laboratory and “pre-field” tests (see Subheading 3.3.). Equipment for the assessmentof pesttcides has been reviewed by Matthews (36); a particularly useful 25toothed rotary atomrzer developed by J. S. Clayton creates relatively small numbers of droplets that simulate field dosages in a confined space (38). Rotary atomizers have several advantages, includmg: 1, They produce of droplets that are more commensurateto field apphcattonsthan spray residuescreatedwrth standardequrpment,suchasthe Potter tower (36), 2. They enableeasyspray apphcation of small quantitiesof experimental products (that may also block narrow Pottertower or “air-brush” nozzles);and 3 Theyproduceanarrow dropletsizespectrumwrth aVMD that easilycanbeadJusted by regulatmgthe apphedvoltage. The residuesproducedthereforeconsistof more succmct“doses” thatsnnplify the processof pathogenacqulsrtronby the targetpest Although rotary atomrzers are very useful as research tools, rt is important to be aware of the differences between preliminary tests and field trials using hydraulic sprayers. For example, the shearmg action on a formulation by pumps and nozzles has been shown significantly to affect suspended particle dtameters, especially m formulations containing emulsified oils (27) There IS a much more substantial decrease (by a cubic function) m the volume of matter these particle diameters represent; this mathematical relationship must also be remembered when interpreting droplet size spectra, 3 2.3. Measuring and Interpreting Droplet Size Spectra Although further researchISalways needed, the droplet sizebands that are most likely to achieve satisfactory dose transfer have been published by several authors (32,39,40). A certain amount of fundamental work has also been done with chemtcal rnsectrcrdesthat relates “optimum” droplet srzeswrth application rate and field concentratton (e.g., 41). There 1sa wide range of techniques available for measuring droplet sizes,either in flight after leaving the nozzleor after collection on arttficral surfaces in the target zone (32,42). Laser particle size analyzers provrde a raprd meansof measurmgthe droplet sizespectraof spraysas they leave the nozzle; the data is processedelectronically and can therefore easily be entered mto databasesfor further analysis. Estimatesof the numbers of particles m each stzeclass

Delivery

Systems

and Protocols

for Biopesticides

517

can be deduced from this data (43), which 1suseful when developmg a formulatron and estimating an appropriate dilutron for the final tank mix. Figure 2 shows some droplet srze spectra produced by a range of both experimental and operattonal nozzles. As well as rotary ULV sprayers, pathogens have been applied at low volume apphcation rates with both thermal (#dj and cold (45) fogging equipment. Although flat fan (or hollow cone) hydrauhc nozzles are most commonly used, some biopesticrde researchers have preferred anvil tips smce they are less prone to blockage with experimental (and surprtsingly some commercial) formulattons. The spectra illustrated were measured with a “Malvern 2600” particle size analyzer (using its “model Independent” analysts) and should be constdered stmply as examples for tllustratton. This 1sa spattal sampling technique that gives similar results to temporal sampling with CDA sprays; however, some authorrties prefer to adJustfor droplet veloctty when mterpretmg hydraulic nozzle data (46-&I). Blank formulatrons have been used m these tests (water + 0.1% Agral 90 for the hydraulic nozzles), but adJuvants (which may offer the greatest scope for enhancing a pathogen’s delivery /49/arid efficacy in hydraulic systems)can have a profound effect on droplet spectra (50) Juxtaposed with the droplet sizegraphs in Fig. 2 are secondary x axes showing the numbers of colony forming units (cfu) or similar Infective partrcles (of any reasonable size) that can be expected to occur in each size class,assuming a random distrrbutron m the spray tank. These have been calculated by convertmg droplet diameters to volumes (m prcohters or 10-t* L) and multiplying by the numbers of cfu per unit volume in the tank mix; microbtologists usually work tn terms of cfu/mL so a factor of IF9 must be used. The volume scalesare simple arithmetic transformatrons from the diameters and end with a vertical lme representing the point at which there IS<50% chanceof a droplet containmg a partrcle. At very low concentrations,the probabtlitres of cfu being contained m a droplet actually follow a Porsson dtstrrbutron (51). The concentrattons selected are similar to those described in the literature: Relatively dilute suspensronshave been sprayed wtth mycoherbtcides (5I,52) and certain virus applications, medmm/high volume hydraulic sprays of entomopathogens tend to be in the region of 1O’O-10’ t &t/L (81and from 1 to 5 x I O1* conidra/L for ULV apphcatton of mycomsectrctdes(35) 3.3. From Laboratory to Field (“‘Prefield Trials’3 Jones and Burges (52) give a useful overview on the principles of biopesticrde formulatron, which include: 1. Stabthzattonof biological agentsfor dtstrtbution and storage; 2. Aidmg handling and apphcation of the product; 3 Protectton of the agent from harmful environmental factors, thus increasing persrstence,and 4 Enhancing the activtty of the agent at the target site.

518 20

Bateman Mlcrogen E 10 (deodonsed paraffin)

Black zones represent powble optlmum droplet sizes (i) for aerosols (small flying insects

‘;; 20. (Ii) for msectlcldes !j (and fhnglcides) z on plant surfaces t

Micron Ulva+ 7000 RPM (UL, blank) -

droptet volume (pl) Pi I . .-.-I10 1 .-’loo (eqmvalent to the expected number of#amcles pe! droplet m a formulahan conhwung 10 parhcles 1 ) (m) for fohar herblclde deposition

CI 20. ,!j 3 z

.-...’ 1000

Spraymg Systems 8002 fan nozzle 300 kPa

formulation contiunmg F.i 5 x IO” parhclesI ’

..‘-’ 10

$’ 2o ’ Lurrnark AN2 anwl nozzle, 100 kPa % 5 3

2 R 5 5 0. s

fomtulatlon contaming IO’ particles I ’

Fig 2 Theoretical particle distrlbutlons (see fed)

pi

...-I IO

.

’ 100

m the droplet Size spectra of various nozzles

Havmg produced a prototype formulation of a virulent pathogen, there may follow an iterative process to check that the apphcatlon techmque IS truly appropriate. The rate of active mgredient (dosage) and volume apphcation rate (likelihood of encounter with residue) must be balanced against certain practl-

Delivery

Systems

and Protocols

for Biopesticides

519

cal considerattons (cost of pathogen production, improvements to the formulanon, operator acceptability, work rate, and so forth). However, m field crops the apphcatton method is more or less fixed (27) and it 1simportant that prelimtnary trtals represent the normal method of application as closely as possible. The use of “prefield trials” has become a well-established concept for relating laboratory broassaysto field performance (53). These may take the form of spray chambers or track sprayers, glass-house or cage tests, and other techniques that attempt to recreate field spray depositsand assesstheir effect on target pestsbefore going to the expenseof carrymg out full-scale field trials. Standardtechniques have been reviewed for msectrcides(36) and herbicides (54). Prefield trials are also valuable for assessmgpestictdal efficacy on highly mobile targets where conventional field trials are impracticable. “Arena” testsconstitute a srmple way of srmulatmg apphcatronof pathogensto insectsat ULV rates(55) andare useful for assessingnew formulations and compartng strains.Figure 3 showsan experimental layout designed to provide a varrety of mean droplet numbers at various sampling stations,from which rt has beenpossibleto correlate mortahty with numbers of droplet unpactions (56). The distancesfrom the spray lure have no particular stgnificance in themselves, apart from provrdmg a range of spray deposits,as illustrated the bottom of the figure. The position and magnitude of the initial peak is dependenton droplet size,ermssron height, flow rate, wind-speed, an-turbulence, and formulatton charactertsttcs,this can be assessedusing well-known spray recovery techmques(42,57). Although large cagesare expensive to construct,they can be useful for testing slow-acting formulattons on adult insectsthat would otherwise fly away. They permit teststo contmue under “near field conditions” at expernnental stationsand outside the normal field season.Several practical problems accompanying the use of cagesrequire seriousforethought. They include: 1. Obtaining a “typical” spraydeposit m the cagefor representativedosetransfer to the target pest 2, Hughbackground mortahty resulting from restrictedmovement(e.g., tnabrhty to rind shade,depletion of food source)and “handlmg” mortality 3. Change of mlcrochmate. Increasedtemperature and humidity can causearmicrally high insectmortality 4 Excessrvepredation andremoval of cadavers(that areneededfor assessment)by ants and other predators 5 Damageby animals, theft, andvandalism. 6 Decontammatronof cagesafter use This must be carried out thoroughly with blopestrcldes: Prolonged exposure to sunlight usually may be effectrve, but mrcroorganrsmswill persist rn the soil. 3.4. Field Trials It is only possible to give a few genera1 notes on field trials, because each one will be different depending on terrain, target insect, available resources,

Ba teman

520

I

5m

10m

20m

30m

Indwldual

arena wth

poup of target samples (e g locustson potted plants and spray cards)

Typical pattern of spray deposits

0

5 10 20 30 distance downwind of spray line (m)

Fig. 3. Arena testdesign. and the quantity of biopesticide available. In any case,larger scale field trials are similar in their execution to those carried out on slow-acting chemicals. A solution designed for the fast moving brown locust in South Africa is the use of field enclosures made from plastic sheetson a collapsible metal frame (58); unfortunately these do not work universallyqhey are only appropriate for insects that are incapable of climbing plastic walls (such as brown locust nymphs). 3.4.1. Trial Design and Spray Deposition With ULV drift spraying, a lethal dose must be delivered to the target by a relatively small number of spray droplets containing a high concentration of active ingredient. The risk of contamination to downwind plots is especially severe with application of very virulent pathogens, so wide separation of plots is necessary. Table 1 is for guidance only and based primarily on ULV spray

in Biopesticide

1 ha 9-l 00 ha 10 km2 >lO km2

1 krn2b 1okm*

0.25 ha

100 m2

Preferred plot size

Only)

0.25 hab 4 ha

Minimum plot size

Trials (for Guidance

10~1000

50-200

l-5 3-10 15-50

Track spacing (typical range)

‘Irery low volumeappllcatlon(often meaningwater-based sprayingat ~20 L/ha wltb rotary sprayers). %mallerplotsmay beexcusableunderexceptionalcmumstances, but should be avolded wherever possible

Hand-held hydraulic, VLVa and so forth Hand-held dnfi sprays Vehicle, aerial @ ~3 m off ground Aerial @ = 5 m off ground Aerial @ l&30 m off ground

Treatment type

Table 1 Plot Sizes and Spacing

10m 100m 200 m 500 m lkm

Mmlmum downwind distance to next plot

522

Bateman

deposItIon. The choice of plot size must, of course, take pest behavior mto account, and there may also be “trade off’ between maxlmlzmg plot size and mimmlzing the size of the total trial (or block) in order to treat reasonably homogeneous target populations. In the absence of sophisticated analytical equipment, spray deposition must be estimated from droplet numbers usmg formulation-sensltlve cards or other means of tracing spray deposits onto natural surfaces A well-known technique for appllcatlon research 1sthe use of pigments mixed mto the formulation that show the presence of droplets with portable UV lamps (#2,59). Optical bnghteners, such as “Tmopal” and “Uvltex,” can be used m a slmllar way and their addition to operational formulations could confer addltlonal benefits, mcludmg UV protection (60) and determmatlon of batch viability for mycomsectlcldes (61). An assessmentof fohar coverage 1sneeded for blologlcal herblcldes and msectlcides that have a stomach actlon (e.g., B thuringzensu, viruses). There may only be time for very approximate assessmentsof droplet size recovery under field condltlons, but these can be espectally useful if a large number of samples are taken (43). It can also be useful to assessthe level of direct contact to target insects; this may be very variable, depending on the application technique, meteorologlcal condltlons at the time of spraying, and crop architecture (2). 3.4 2. Assessing Modes of Pathogen Act/on in the field Measuring the relative importance of direct contact and secondary pick-up from fohage IS important with pathogens that have a contact action (such as mycomsectlcldes). Prefield trial studies are valuable, but should be followed by an assessmentof the acqulsltlon of pathogen propagules in the field; this may require approaches that are different from chemicals (for example, searches for locust cadavers after spraymg mycoinsecticides are often fruitless) Thomas et al. (62) used open-bottomed field cages to assessthe relative importance of the Initial contact and secondary acqulsltlon of the residue With the aid of models, such cages also can be used to assessthe decay of the spray residue and any increase m the level of moculum as a consequence of subsequent sporulatlon of cadavers. However, insects placed in cages are subject to an artificial microenvironment, especially when the pathogen-host interaction 1s sensitive to temperature and humidity (63). Cross-contammatlon and high control mortality can also be a problem. Besides accounting for direct acqulsltlon of applied blopestlcides and secondary cyclmg (horizontal transmlsslon), models may also be used to estimate. 1 Immlgratlon and emigration of mdlvlduals from trial plots, 2 The lag time between mfectlon and pest mortahty under field condltlons, and 3 Effects of the environment on pest-pathogen mteractlons (e.g , thermoregulatory behavior for combatmg infection)

Delivery 3.4.3.

Systems

and Protocols

Data Collection

for Biopesticides

523

and Analysis

Further guidelines on assessing mortality of slow-acting pestlcldes used against locusts are suggested by FAO (64). It is most important to record carefully all relevant data during a trial: The FAO Pesticide Referee Group have set minimum standards of reporting. Here again, locust control provides us mternatlonally recognized protocols; unfortunately this 1sunusual m agricultural research. Cage samples are almost invariably taken to establish that dose-transfer and infection have taken place, and to ensure that at least some mformatlon 1s recovered even if other assessments fall. Data from cage mcubatlons can be treated similarly to bioassay results with results presented graphically or summarlzed with a few key statistics. One of the simplest statlstlcs to calculate IS the median lethal time (MLT): the number of days to achieve an accumulated 50% mortality (using linear interpolation of cumulative, dally mortahtles). The slgmold mortality curve may not be symmetrical, so such statistics as the average survival time (AST) are more appropriate since they incorporate all the data. Results obtained with basic AST formulas are affected by sample size and the end point of the experiment if not all the msects die Medical statistical techniques, such as Kaplan and Meier’s survival analysis method (65) overcome this problem, and have been used m the analysis of mycopesticlde trials The most substantial measure of success will always be demonstration of pest population reductions m the field, the length of time populations are suppressed, and ultimately yield increases for crops; there are several standard texts for the analysis of such data. Percent efficacy can be simply estimated using the formula of Henderson and Tilton (66). Statlstlcal procedures now exist to analyze data on a day-by-day basis using “repeated measures” techniques or split-plot ANOVA designs when the level of auto-correlation between sampling dates is small (67,68).

4. Conclusions Most successful blopestlclde development projects are achieved with multidisciplinary efforts. They start with a promising pest-pathogen mteractlon, pass through production, formulation, and application development (the “delivery system”) m conJunctlon with field testing, then contmue reglstratlon and commerciahzatlon. 1 The most likely scenario for lmplementatlon will be the development of commercial products by small- to medium-scale compames and ~111involve a “package of technology” (developed at least m part with the ald of public funds) 2 One of the most crltlcal stages in development 1s the passage from laboratory to field This IS an essential element of the practical verlficatlon process that provides a link between sclentlfic research and product development. The estabhshment of low environmental impact 1s especially Important with blopestlcldes and

524

Ba teman these studies are now rightfully being well supported Unfortunately, the thorough practical vertficatton of baste delivery systems 1sstall margmahzed, smce It is not considered to be “cutting edge” sctentific research and IS too expensive for smaller commercial enterprises There 1s a great temptation to develop a package of technology that involves novel formulation and/or appltcatton techniques. However, if biopestictde products are to have a role in broad acre agrtculture, they should be adapted to existmg delivery systems as much as possible. Farmers and growers are unhkely to adopt btopestlctdes rf they are obliged to radically alter existing practtce The examples described here in greatest detail represent an exceptional apphcatton scenarto that proves the rule: ULV spraymg is the normal method of application for locust msecticides. However, m many cropping systems such techniques as CDA are underresearched, pressures to develop alternattve methods of btopesttctde application may be given greater impetus with the genuine lmplementatton of IPM The development of models can be very useful for improving the understandmg of infection processes, but a step-by-step empirtcal approach to testing is also rigorous and provides essential data for the registration of products In the foreseeable future btopestictdes will occupy mche markets, therefore, fundmg for development research for mdtvtdual products will be hmited The refinement of protocols that maximize the cost-effecttveness of testing will Improve the chances of tmplementmg these valuable tools for integrated pest management

References 1. Georgts, R (1997) Commerctal prospects of mtcrobial msecttctdes m agriculture, m Mcroblal Insectlcldes Novelty or Necessity? British Crop Protection Council ProceedwgsMonograph Series No. 68, pp. 243-252. 2 Bateman, R P. and Thomas, M (1996) Pathogen application agamst locusts and grasshoppers: insecttctde or btological control? Antenna 20, 10-15. 3 Petch, T (1925) Entomogenous fungi and thetr use in controllmg insect pests Bulletin of the Department of Agrrculture, Ceylon, No 7 1 Government Prmter, Colombo, 40 pp 4 Jones, K A (1994) Use of baculoviruses for cotton pest control, m Insect Pests of Cotton (Matthews, G. A. and Tunstall, J P , eds.), CAB International, Wallmgford, UK, pp 477-504 5 Prior, C , Jollands, P., and Le Patourel, G (1988) Infecttvtty of 011 and water formulattons of Beauverla basslana (Deuteromycotma, Hyphomycetes) to the cocoa weevil pest Pantorhytes plutus (Coleoptera. Curculionidae) J Invertebrate Path01 52,66-72 6 Dent, D. R (1997) Integrated Pest Management and mtcrobtal insecttcides, m Mlcroblal Insecticides. Novelty or Necessity? British Crop Protection Council ProceedrngslMonograph Serves No 68, pp. 127-138 7. O’Connell, P. J. and Zoschke, A. (1996) Limitations to the development and commerctahsatton of mycoherbicides by industry. 2nd International Weed Control Congress. 6. ! Cooenhaeen. I “1“ DD. 1189-l 195

Delivery

Systems

and Protocols

for Biopesticides

525

8 Burges, D H. (ed ) (1997) Formulatton of Mcrobtal Btopestuxdes, Benefictal Micro-Organtsms and Nematodes. Chapman and Hall, London, m press. 9 Anderson, R M and May, R. M. (198 1) The population dynamics of microparasttes and then Invertebrate hosts Phil. Transact Roy Sot B 291,45 l-524. 10 Thomas, M B., Wood, S. N., and Lomer, C J (1995) Btologtcal control of locusts and grasshoppers using a fungal pathogen: the Importance of secondary cyclmg Proc Roy Sot Lond B 259,265-270. 11 Hails, R S (1997) The ecology of baculoviruses towards the destgn of viral pest control strategies, m Mzcrobtal Insecticides Novelty or Necesstty? Brttzsh Crop Protectton Councrl ProceedlngslMonograph Series No. 68, pp. 53-62 12 Munthah, D C and Scopes, N. E A (1982) A techmque for studymg the btologrcal effctency of small droplets of pestrcrde solutrons and a consrderatron of the lmphcatrons Pest Set 13,60-62. 13 Adams, A J , Chapple, A C., and Hall F. R. (1989) Droplet spectra for some agricultural fan nozzles, with respect to draft and brologtcal efficiency, m Pestztide Formulattons and Applicatton Systems 10th Volume ASTM STP 1078 (Bode, L. E., Hazen, J L., and Chasm, D. G., eds.), Amerrcan Society for Testmg and Materials, Philadelphta, pp 156-169 14 Bryant, J. E and Yendol, W. G. (1988) Evaluation of the influence of droplet size and densrty of Bacrllus thuringrensis against gypsy moth larvae (Lepidotera. Lymantriidae). J Econ Entomol. 81(l), 130-134 15. Maczuga, S. A. and Mrerzejewskr, K. J (1995) Droplet size and denstty effects of Bacillus thurmgiensrs kurstaki on gypsy moth (Lepldoptera: Lymantrndae) larvae J Econ Entomol 88(S), 1376-1379. 16. Ford, M G. and Salt, D W. (1987) The behaviour of pesttcrde deposrts and then transfer from plant to insect surfaces, m Crttical Reports on Applted Chemutry, vol 18: Pesttctdes on Plant Surfaces (Cottrel, H. J., ed.), Wiley & Sons, pp 268 1, 17. Hall, F. R. Chapple, A. C., Taylor, R. A. J., and Downer, R. A. (1994) Dose transfer of Bactllus thurrngtenszs from cabbage to the dramond back moth a graphical srmulator J Environ Set Hlth B29 (4), 661-678 18 Cory, J. S., Hnst, M. L., Williams, T., Hails, R S., Goulson, D., Green, B M , Caley, T. M., Possee, R. D., Cayley, P. J , and Bishop, D. H. L. (1994) Field trial of a genettcally improved baculovirus insecticide. Nature 37, 138-140 19 Hans, J. G. (1997) Microbial insectxides-an industry perspectrve, m Mcrobtal Insecttctdes Novelty or Necessity? Brtttsh Crop Protectton Counctl ProceedzngslMonograph Serves No 68,41-50. 20 Evans, H. F. (1994) Laboratory and field results with vn-us for the control of msects, m Comparing Glasshouse and Field Performance Ii (Hewttt, H. G., Casely, J., Copping, L. G., Grayson B. T., and Tyson, D , eds ), Brmsh Crop Protection Councrl Monograph No. 59, pp. 285-296. 21. Dulmage, H. T., Boenmg, 0. P., Rehnborg, C S , and Hansen, G. D. (197 1) A proposed standardized bioassay for formulations of Baczllus thurzngzenszs based on the international unit J Invertebrate Pathol. 18, 240-245.

526 22

23. 24

25

26

27

28 29.

30

31

32 33 34

35 36

37

Bateman Bateman, R. P , Carey, M., Moore, D , and Prior, C (1993) The enhanced mfectivrty of Meturhzzzum jlavovznde m 011 formulations to desert locusts at low humtditres Annul Apple Bzol 122, 145-152 Robertson, J L and Pretsler, H K (1992) Pesticide Bioassay wzth Arthropods CRC, Boca Raton, FL. Nowierski, R M , Zeng, Z , Jaronskr, S , Delgado, F , and Swearmgen, W ( 1996) Analysis and Modelmg of time-dose-mortality of Melanoplus sanguznzpev, Locusta mrgratorza mlgratorloldes and Schzstocerca gregarca (Orthoptera Acrididae) from Beauvena, Metarhzzzum and Paecdomyces isolates from Madagascar J Invert Path01 67,236-252 (1996) Fungus stram selection for mycopestictdes. Semmar held at the Society of Invertebrate Pathology Meetmg, Cordoba, Spain, September, co-ordmated by C J Lomer Bradley, C. A , Black, W E , Kearns, R , and Wood, P (1992) Role of production technology m mycomsectrcide development, m Frontiers in Industrial Mycology (Leatham, G F , ed.), Chapman and Hall, London, UK, pp 16&173 Chapple, A C and Bateman, R P (1997) Application systems for mtcrobial pesttctdes necessity not novelty, m Mrcroblal Insectlcldes Novelty or Necessity? Bntuh Crop Protection Councd ProceedzngslMonograph Series No 68, pp 18 1-l 90 Helyer, N., Gill, G , and Bywater, A. (1992) Control of chrysanthemum pests with Vertrcillmm lecann. Phytoparasrtrca 20, 5-9 Sopp, P I , Grllespie, A T., and Palmer, A. (1990) Compartson of ultra-lowvolume electrostatic and high volume hydraulic apphcatron of Verticillium lecann for aphrd control on chrysanthemums. Crop Pro&&on 9, 177-l 84 Keller, S (1992) The Beauvena-Melolontha proJect. experiences with regard to locust and grasshopper control, m Biological Control of Locusts and Grasshoppers, (Lomer, C J and Prior, C., eds ), Pub1 CAB Intematronal, Wallingford, UK, pp 279286 Graham-Bryce, I J. (1977) Crop protectron. A consideration of the effectiveness and disadvantages of current methods and of the scope for improvement Phzlos Transact Roy Sot Lond B281, 163-179. Matthews, G A (1992) Pestzclde Applrcatlon Methods, 2nd ed Longman Scientific and Techmcal, Harlow, Essex, UK Reardon, R (1991) Aerial Spraymgfor Gypsy Moth Control A Handbook of Technology Urnted States Department of Agrtculture Forest Servtce, NA-TP-20, 167 pp Entwtstle, P F , Evans, H F , Cory, J. S , and Doyle, C (1990) Questions on the aerial application of mrcrobral pestrctdes to forests Proc Vth Znternatzonal Colloquium on Invertebrate Pathology, Adelaide, Austraha, 159-163 Bateman, R. P ( 1997) Methods of application of microbial pesticide formulations for the control of grasshoppers and locusts Memoirs Entomol Sot Cunadu 171,698 1 Matthews, G A (1997) Techmques to evaluate msecttcide efficacy, m Methods m Ecologzcal & Agrzcultural Entomology (Dent, D R and Walton, M P., eds ), CAB International, Wallmgford, UK, pp. 243-269 Parkm, C S and Merritt, C R (1988) The measurement and prediction of spray draft Aspects of Appl Blol 17,351-361

Delivery

Sys terns and Protocols

for Biopes ticides

527

38. Bateman, R P (1994) Performance of myco-insecttcides importance of formulation and controlled droplet apphcatton BCPC Monograph 59,275-284 39 Himel, C M. (1969) The optimum drop size for msecticlde spray droplets J Econ Entomol 62,919925. 40. Doble, S. J., Matthews, G. A., Rutherford, I., and Southcombe, E. S. E. (1985) A system for classrfymg hydraulic nozzles and other atomizers mto categories of spray quality. Proceedmgs BCPC Conference-Weeds, 1125-l 133 4 1 Johnstone, D. R. (1973) Insectlctde concentration for ultra-low volume crop spray application. Pest. Set 4,77-82. 42 Cooke, B K and Htslop, E C (1993) Spray tracmg techniques, m Applzcatzon Technology for Crop Protection (Matthews, G A and Hislop, E. C , eds.), CAB International, Wallingford, UK, pp 329-347 43 Bateman, R P (1993) Simple, standardised methods for recordmg droplet measurements and estlmatton of deposits from controlled droplet apphcattons Crop Protection 12,201-206 44. Jarrett, P. and Burges, H. D. (1982) Use of fogs to dissemmate pathogens. Proceedings Iiird Internattonal Colloquwm on Invertebrate Pathology, Brighton, UK, pp. 49-54 45. Falcon, L A. and Sorensen, A, A. (1976) Insect Pathogen-U 1 v combmation for crop pest control PANS 22 (3), 322-326. 46 Arnold, A C (1987) The dropsize of the spray from agricultural fan spray atomizers as determined by a Malvern and the Particle Measuring System (PMS) instrument. Atomtsatton Spray Technol 3, 155-167 47 Arnold, A C (1990) A comparative study of drop sizing equipment for agrtcultural fan-spray atomizers Aerosol Science Technol. 12 (2), 43 l-445. 48. Chapple, A C., Taylor, R. A. J., and Hall, F. R. (1995) The transformation of spatially determined drop stzes to then temporal equivalents for agricultural sprays. J Agrtcult Engineer Res. 60,49-56. 49. Chapple, A. C. (1996) Application of biological control agents. some theoretical considerattons of dispersal, m Proceedtngs of the 5th European Meettng of the Internattonal Organisatton of Btological Control, West and East Palearcttc Regions Mtcrobtal Control of Pests. Poznan, Poland, pp 24-28 50 Hall, F R , Chapple, A C , Downer, R. A, Kirchner, L. M., and Thacker, J R M. (1993) Pesticide application as affected by spray modtfiers Pesttctde Set 38, 123-133 5 1 Amsellem, Z , Sharon, A , Gressel, J , and Quimby, P C (1990) Complete abohtion of high inoculum threshold of two mycoherblctdes (Alternarza casszae and A crassa) when applied m invert emulsion. Phytopathology 80,925-929 52 Lawrie, J , Greaves, M P , Down, V M , and Chassot, A. (1997) Some effects of spray droplet size on distribution, germmatlon of and mfection by mycoherbictde spores Aspects Appl B1o1 48, 175-182 53. Jones, K A and Burges, H. D. (1998) Prmctples of formulation, m Formulatzon of Mtcrobial Biopesttcides, Beneficial Mtco-Organums and Nematodes (Burges, H D , ed ), Chapman & Hall, London, UK, in press

528

Bateman

54 Brutish Crop Protection Council (BCPC) (1994) Comparzng Laboratory and F&d Pestzclde Performance 0 BCPC Monograph No 59 (Hewitt, H. G., Casely, J , Copping, L. G., Grayson, B. T., and Tyson, D., eds.), 323 pp 55 Qmmby, P C and Boyette, C. D (1987) Productron and appltcatlon of biocontrol agents, m Methods ofApplymg Herbzczdes (McWhorter, C. G and Gebhardt, M R , eds ), Monograph No 4, Weed Society of America, Champagne, IL, 358 pp 56 Bateman, R. P., Godonou, I , Kpmdu, D., Lomer, C J , and Paralso, A (1992) Development of a novel “field bioassay” technique for assessing mycopestlcide ULV formulations, m Brologzcul Control ofLocusts and Grasshoppers (Lomer, C J and Prior, C , eds ), Pub1 CAB International, Wallmgford, UK, pp 255-262 57 Bateman, R P, Douro-Kpmdou, 0 K , Kooyman, C , Lomer, C , and Ouambama, Z. (1998) Some observations on the dose transfer of mycomsecttcrde sprays to desert locusts Crop Protectzon, 17, 151-158. 58. Prckm, S R (198 1) in Manualfor TestzngInsectzczdeson Rice (Hemrichs, E. A , Chelhah, S., Valencia, S. L., Arceo, M. B., Fabellar, L T., Aquino, 0. B., and Pickin, S R , eds ), IRRI, Los Banos, Philippines, pp 52-66 59 Bateman, R P , Price, R E , Muller, E J., and Brown, H D (1994) Controllmg brown locust hopper bands m South Afrtca with a myco-msecticide spray. Proceedings of the Brighton Crop Protectton Conference-Pests and Diseases,

60 61

62

63.

64

65 66 67. 68

November 1994, pp 60996 16 Stamland, L N. (1959) Fluorescent tracer techniques for the study of spray and dust deposits J Agrzcult Engtneer Res 4,42-81 Shapiro, M and Robertson, J L (1992) Enhancement of gypsy moth (Lepldoptera. Lymantrndae) baculovrrus activity by optical brighteners J Econ Entomol 85(4), 1120-l 124. Jimmez, J and Gillespie, A T (1990) Use of the optical brightener Tmopal BOPT for the rapid determmatton of conidtal viabilities in entomophagous deuteromycetes Mycological Res 94,27!%-283 Thomas, M B , Wood, S. N., Langewald, J , and Lomer, C J (1997) Persistence of Metarhizzum j7avovzrzde and consequences for brological control of locusts and grasshoppers Pestlclde SCI 49,47-55 Price, R E , Bateman, R P , Brown, H D , Butler, E T , and Muller, E J (1997) Aerial spray trials agamst brown locust (Locustuna pardulrna, Walker) nymphs in South Africa using oil-based formulations of Metarhlzlum jlavovzrzde Crop Protect 16,34 l-35 1 Food and Agriculture Organizatron of the United Nations (199 1) Guide-lines for pesticide trials on grasshoppers FAO Booklet, comptled by Dobson, H., Rome, 16 pp Kaplan, E L and Meter, P. (1958) Nonparametrtc estimation from mcomplete observations J Am Statzst Assoc 53,457-481 Henderson, C. F. and Tilton, E. W (1955) Tests with acartcides agamst the brown wheat mite. J Econ Entomol 48, 157-161 Perry, J N (1997) Statistical aspects of field experiments, m Methods zn Ecologzcal & Agricultural Entomology (Dent, D R and Walton, M P , eds ), CAB International, Wallmgford, UK, pp. 171-20 1.

28 Analysis, Monitoring, and Some Regulatory

Implications

Jack R. Plimmer 1. Introduction Pesticide analysis is generally conducted with one of two objectives: product analysis to determine *the quantity of active ingredient in a manufactured product or formulation, or resrdue analysis to determine amounts of material resulting from application or use. In addition, the analysis may mclude elucrdation of the composition and confirmation of the identity of the active mgredient, or its metabolites or alteration products. This summary of methods of characterizatton, analysis, and identification of biopesticides is based on examples that range in composition from homogeneous macromolecules to whole organisms. Such diversity demands a variety of analytical approaches, and presents a challenge to the ingenuity of the analyst, often requiring the use of sophisticated mstrumentation. Some regulatory topics have also been included in the discussion, because requirements for analytical data are specttied by regulatory authorities when new materials are to be used in pest management. Such analytical data is essential for risk assessmentand ensurmg that adequate safeguards to protect human health and the environment are mcorporated m applications of new technology.

1.1. Regulatory Issues The use of a material as a pesticide entails compliance with the requirements of the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA; 2) and the Food Quality Protection Act (FQPA; 2). The risk-assessment process undertaken by the U.S. Environmental Protection Agency (EPA) as part of its decision to register a pestictde relies on data obtained by measurements of residues in commodrties, environmental samples, crops, or other substratesthat From Methods IR Bofechnology, vol 5 Bopestm%s Use and De//very Edited by F R Hall and J J. Menn 0 Humana Press Inc , Totowa, NJ

529

530

Phmmer

may contam residues. The term “residues” applies not only to the parent pesticide, but also to toxicologically significant metabolites, or other products that might arise by alteration of the parent molecule. The regulations provide substantial discussion of the data required by the EPA for the risk assessmentand approval process. Not least important IS the necessity to conduct studies to obtain residue data withm the legal framework of Good Laboratory Practices (3). To accommodate classes of pesticides that differed substanttally m their mode of action from conventional chemical pesticides, the EPA prepared new gmdelmes to cover the so-called “brorational pesticides” (Subpart M; 4) From the regulatory standpomt, this group of pesticidal agents was divided into biochemicals and microbials. The term “microbials” embraces both naturally occurrmg or genetically engmeered orgamsms. The guidelmes were updated m February, 1996 (5). The purpose of these new guidelines was to harmomze testing requirements described m earlier FIFRA, Toxic SubstancesControl Act (TSCA), and Organization for Economic Cooperation and Development (OECD) publications. Not only did the new classes of pesticides differ substantially m mode of actron, but they were frequently not susceptible to the methods of analysis generally applicable to conventional chemical pesttcides and their residues. Brological control agents other than some mtcroorganisms were exempted from the requirements of FIFRA (6). However, the proposal by EPA to regulate some transgenic plants as “plant pesttcides,” under FIFRA has generated controversy, because of the potentially heavy burden of additional data that might be required for approval (7), Because there is limited experience of the ecological effects of introducing organisms, or self-rephcatmg agents contammg genetic mformation, into the environment, their potential for adverse effects has drawn regulatory attention. Risks associated with the release of genetically engineered microorganisms (GEMS) may be linked to potential for pathogemcity, and for colomzmg the environment and drsplacmg existing species @I. Conventional methods of residue chemistry may often be applicable to the analysis of biochemicals (i.e., identification, detection, and quantttation). This applies to well-defined molecular species,generally synthetic products, used as pest-control agents. 1.2. Regulation of Transgenic Plants Incorporation of genes that confer resistance to herbicides mto crop plants may extend the range of pesticides currently m use, and thus prolong the economic life of herbicides that have a demonstrated history of safe and effective use. Monsanto (St. Louis, MO) has applied this approach to soybeans and cotton, and experiments are being conducted on glyphosate-tolerant canola in Canada.

Analysis of Biopesticides

531

EPA is now registering genetically engineered plants capable of expressmg proteins similar to those contained m Bacillus thurzngzenszs(Bt). These are described as “plant pesticides” and are registered under FIFRA. Implicit m such approval is the requirement to analyze the products of gene expression, and monitor the environmental impact of the engineered organism. Recent patents have been issued for techniques to modify genes from Bt to optimize insecticidal protein expression m plants. Several major crop plants (including corn, cotton, canola, potatoes, and tomatoes) have been transformed with synthetic Bt genes (9). EPA has approved the sale of hybrid seed corn that incorporates B&based resistance to European corn borer (JO). A plant pesticide, targeting the European corn borer, was recently registered. It is a field corn containing the gene, pCIB443 1, responsible for the productton of the S-endotoxin protein, Bt CryIA(b), an Insect toxin. The gene produces a truncated version of the naturally occurring insecticide. In May 1995, Monsanto’s request for the first full registration of a plant pesticide was approved by the EPA. This is a Colorado-beetle-resistant potato carrying the genettc maternal required for production of Bt CryIII(a), a &endotoxin. 1.3. Nocontrol Agents and Genera/ Analytical Approaches Biocontrol agents include a variety of organisms, and the most important class of these is Bt Berliner, which also provides a source of genettc material for many products now in use. Safety evaluations of Bt following different expoSure routes showed that these entomopathogens were virtually nontoxic to mammals, provided that high dose levels were not used (11). Currently, assessmentsof acute toxicity/pathogenicity utilize colony forming units (CFU) for assessmentof exposure. A CFU ts defined as a single, viable propagule that produces a single colony (a population of cells visible to the naked eye). The analysis of Bt preparations and some of the regulatory implications are discussed later (see Subheading 5.). The analysis of microbiological pest control agents (MPCAs) 1s discussed by EPA in, the Subdivision M guidelines. Analytical methods are required for data collecTion to support tolerances, and for enforcement of the regulations (12). A monitormg method is required for all MCPAs that are exempted from the requireG>entsof tolerance. Conventional analytical procedures, such as gas chromatography (GC), mass spectrometry-(MS), or high-pressure liquid chromatography (HPLC) (or combinations thereof), are typically used for many biopesttcides. If the MPCA per se, a mutant, OF viable recipient of MCPA genetic material is a residue of toxicological concern, then various immunological methods (such as enzymelinked immunosorbent assay and dot-immunoassay) or molecular probe methods (such as dot hybridization, Southern hybridization procedure, or

532

Plimmer

restriction endonuclease mapping) may be used for identification and/or quantitatron. Since the above methods do not necessarily determine vtable MCPAs, culturing of trssues (maceration followed by dilution plating) or infectivity assayswill frequently be necessary.Methods useful for detectton of microorganisms are discussed m Subheading 4. (13).

2. Analysis 2.1. Biochemicals and Natural Products EPA guidelmes describe four major classes of biochemical agents: semiochemicals, hormones, natural plant regulators, and enzymes. Current mstrumentation combines MS, nuclear magnetic resonance spectrometry (NMR), infrared (IR), ultravrolet (UV) spectroscopy, and X-ray crystallography wrth powerful separation techntques, such as high-pressure hquid chromatography (HPLC) and capillary electrophoresis (CE), and incorporates the capabilittes of computer-based data processing to rapidly acquire information concerning not only details of primary structures of complex natural products, but also their spatial conformatron and interactions with macromolecules, Natural products, such as pyrethrotds, rotenords, nicotine, and so on, have long been in use as pesticides, or have served as models for synthetic modifkation. The realization that the genetic potential of a plant (or other organism) may be transferred to an appropriate host and utihzed to generate commercially valuable, genetically engineered products has revitalized the quest for natural sources of biologrcal acttvity, and strmulated the Investigation of natural products and the molecular biology mvolved in thetr biosynthesis. The incorporation of genes capable of expressing toxins in baculoviruses has generated interest in natural toxins such as sprder and scorpton venoms. For example, an engineered form of the baculovn-us, Autographa calzjbmzca Speyer NPV, which expresses a scorpion toxin, was tested against the cabbage looper, and other genetic modifications are in early testing stages. New fermentanon products include Naturalyte, a spmoside derived from soilborne actmomycetes, which is produced by DowElanco (now Dow Agrochemrcals, Indianapolis, IN) at a fermentation facility. The material 1sacttve against lepidopteran pests (tobacco budworm, cotton bollworm), and rt also shows activity against termites, Colorado beetle, diptera, and other pests (14). Thts was tested on cotton m 1995-1996, and subsequently on vegetables, trees, and vines. The spinosads are a novel class of macrocyclic lactones produced by the soti actmomycete, Saccharupolyspora spuzosa (IS). The two most active insecticidal factors in the mixture are spinosads A and D. Immunoassay techniques have been developed, and are available for analysis. However, at the present ttme, EPA appears strongly committed to a preference for confirmation of residue analysesby conventional methodology.

Analysis of Hopesticides

533

Products of microbial fermentation are often complex m structure, but analysis by HPLC may present dlfficultles when conventional UV detectors are used, when chromophores (aromatic or conjugated systems) are absent. In such a case, pre- or postcolumn derlvatization with a chromogenic or fluorogemc reagent may be useful to enhance the level of detection above that of the refractive index detector system. A typical example of methodology 1sthe method of analysis of gentamicin, an aminoglycoside antibiotic, which may be used as a pesticide for control of plant disease in fi-mt crops. The product is a complex mixture containing three major products and several minor components Gentamycin is formed by the attachment of two ammosugar residues (garosamine and pupurosamine) to 2-deoxystreptamine (an aminocyclitol) through glycosidlc lmkages. Several methods of analysis have been recommended for pharmaceutical preparations, including MS. A recent method employs HPLC separation, combined wtth derivatlzation with o-phthalaldehyde, and such methods can be used for determination of residues on crops (16). Derivatlzation with N-methylimidazole IS used to generate a fluorescent derivative of moxidectin (I 7), a macrocyclic derivative of nemadectin, produced by Streptom-yes sp. The derivative IS analyzed by HPLC, and the method has been used to study plasma levels. Neem extracts (Azadirachta indica A. Juss. [syn. Melra azadirachta L.]), like many crude natural-product extracts, contain a variety of compounds. The extracts of the seedspossessa wide spectrum of biological activity, Including pesticidal activities. The major insecticidally active component of neem extracts 1sazadirachtin (18). Generally, the biological activity of neem 011sIS highly correlated with their azadirachtm content (19), which may be determined by HPLC (ZO), using an HPLC spherisorb (Phase Separations, Franklin, MA) column with acetomtnle/water gradient and a variable wavelength detector at 2 10 nm. Neem extracts generally have low oral toxicity to laboratory mammals, but recently (21) it was shown that a component of the extract of seeds, nimbolide, IS cytotoxlc. Nimbohde could be isolated from neem extracts by chromatographic fractlonatlon on silica columns, followed by TLC on silica gel, and final purification by reversed-phase HPLC on a Merck 1OORP-18 column, and detection with a photodiode-array detector at 220 nm. 2.2. Analysis of Pheromones and Other Semiochemicals Semiochemicals are defined as naturally occurring or synthetic substances, or mixtures of substances,emitted by one species, which modify the behavior of receptor orgamsms of other individuals of like or different species (4). Of these, insect sex attractant pheromones are used in pest management for several purposes: in insect traps for momtoring or survey purposes; in a mass trapping program to reduce insect populations; in combination with msectl-

534

Phmmer

ctdes to attract insects to an area treated with insecticides, or to a device contaming insecticide; and to permeate the an and suppress insect populatton by dtsruptmg mating or aggregation (in this mstance,pheromones may be regarded as btopesticides). The above usesof pheromonesrequtreextremely smallamountsof matertaland the targetsare quite specific.Many semtochemicalsare tdenttcalto, or closely resemble, other naturally occurring matenalstn then chemicalcomposttton,are generally readily degradedtn the environment, and show low toxic@ to nontargetspecies. From a regulatory standpomt, lepidopteran pheromones have recetved special constderatton, and there 1sgrowing expertence of then practical appltcatton (22). In establishing an exemption from requirement of a food tolerance for residues of certam leptdopteran pheromones (independent of formulation or mode of application), m cases m which annual appltcatton was ltmlted to 150 g active ingredient per acre for pest control, or in all raw agrtcultural communities, the EPA took a number of factors into account (23). Lepidopteran pheromones were defined as “naturally occurring compounds (or identical or substantially similar synthetic compound), designated by the unbranched altphattcs (carbon chain between 9 and 18 carbons) ending m an alcohol, acetate, or aldehyde functional group and containing up to 3 double bonds rn the altphattc backbone” (23). This defmttton included the malortty of leptdopteran pheromones, and the EPA considered that, “although other chemical structures have been demonstrated to be leptdopteran pheromones or pheromones of other arthropods, there 1s insufficient toxicity data and exposure mformatton to merit their exemption from tolerances” (23). The decision was made on the basis of the absence of significant toxtctty associated with the structural features of leptdopteran pheromones (compounds from six to 16 carbon unbranched alcohols, acetates, and aldehydes). In addtnon, subchronic toxicity of an tsomeric mixture of tridecenyl acetates mdtcated no significant signs of toxicity other than those associated wtth exposure to a hydrocarbon. Published studies indicate no stgmficant health effects from subchronic exposures to this group of chemicals. Studies of volatihzatton from a mtcrocapsule show that about 70% of the pheromone remains after 30 d. These results indicate that a considerable portion of the total pheromone 1snot capable of being released, which suggests a potential for residues to occur m the absence of any biologtcal or envtronmental factors. However, in a submitted field study, residue analyses from fieldtreated plants indicated no stgmticant amounts of pheromone could be detected on the fruit. Detectable residues of tomato pmworm pheromone on unwashed fruit ranged from 21 to 72 ppb on the day of apphcatton, decreasedto 0.948 ppb on d 15, and 0 29-l .2 ppb on d 30. Washing the fruit brought all residues below the level of detection.

Analysis of Biopesticrdes

535

A field study was conducted and residues were determmed on apples that had received treatment wrth pheromones(52 g/ha of dodecenyl alcohol and 10 gha of tetradecenyl alcohol) (24). Application, weathering, and other envuonmental degradation processescauseda reduction in the active ingredient to a level that approached the system lrmrt of detection in the expected 3-wk hfetlme of the raw agricultural product. The literature contains many descriptions of techniques for isolation and structural elucidation of pheromones (for example, Ostrovsky and Bestmann (251 provide a useful summary). Typical approach to the isolation and Identification of a lepidopteran pheromone mvolves extraction of active mgredients, which may be achreved by analystsof solvent extracts of the sex glands, or by rinsing the gland region with a solvent. The latter method affords a cleaner extract Alternattvely, collection of the volattles secreted by the insect to elicit mating responses provides material that corresponds more closely to the composition of the pheromone blend. For further exammation, extracts may be purified by gel permeation chromatography, column chromatography, HPLC, thin-layer chromatography (TLC), or gas chromatography (GC). Identification of pheromones 1sfrequently based on spectroscoptcdata. UV, IR, and NMR (particularly FTIR and FTNMR) may provide some useful information, but MS and, most frequently, GUMS IS the method of choice. It may not be a simple matter to elucidate the geometry and posmon of the double bonds in long-chain aliphatic compounds, but reactions, such as ozonolysis, may be conducted on nanogram samples and yield fission or other products that can be characterized.Addtttonal information may be obtained by using biological detectors that respond to very specific stirnull. The major hurdle m determining pheromone structures 1susually that of accumulatmg sufficient material for investigation. Smce lepidopteran pheromones may be a blend of two or more components, tt is important to elucidate the quantitative and qualitative composition of the blend that will elicit the desired response. Trace components play a significant role m eliciting behavioral responses (26). This influences the design of effective pheromone dispensers, which must be constructed to emu vapor corresponding in composition to the natural stimulus. 3. Analysis of Biological Macromolecules Biopesttcide mvestrgattons may call for the elucidatton of structure of macromolecules important to molecular biology that are involved m the processof gene expression: nucleotides, proteins, and peptides. Attention has also focused on other macromolecules, such as the carbohydrates, n-rthen role as bioregulators. There has been substantial recent progress m adapting mass spectrometry to the investigation of macromolecular structure. It IS a preferred technique, m

536

Plimmer

combmation with ammo acid analysis and HPLC, for the characterrzatron of peptides. Interfaces between MS and liquid chromatography systemshave been greatly improved. Electrospray ionization and matrix-assisted, laser-desorption/romzation, time-of-flight (MALDI-TOF) mass spectrometry permits determination of protein mass in excess of 100 kDa, utilizing quantities of material, sometimes as low as subprcomoles. Reducttons m cost of such mstrumentation, and improved ease of its operation, have brought the techniques within economrc reach of many laboratories, and closer to routme operation. There are limtts to the applications of MALDI, but it IS currently an extremely active field of investigation. In the MALDI-TOF technique, the analyte is desorbed from a crystallme lattice by a laser beam, and Ions are accelerated through a short distance by a high voltage (up to 30,000 V) Ions are accelerated down a linear fbght tube and their time to arrive at the detector is determined. Their mass IS a function of the square of the flight time and may be measured very accurately (better than 20 ppm at mass 1200), rf internal standards are included Very large charge-to-mass ratios can be measured by this method, and singly charged molecules up to 200 kDa can be analyzed. However, the method 1snot readily applicable to molecules of C900 amu. TOF analyzers are very sensitive and requue 4 to very few prcomoles of material for analysis.However, a relatively high sample concentration (1O-20 ClM) may be necessary.The efficiency of desorptron and detection of an ion is dependent on amino acid composition. Basic residues particularly affect peak heights. Patterson and Aebersold (27) have reviewed application of gel electrophoresis coupled with mass spectrometry for protein identification. A combination of Edman sequencing and MALDI-TOF MS provides a powerful approach to identification of peptides and proteins. Typical procedures mvolve the rsolatton of protems and peptides by gel electrophoresis, and hydrolysis by dtgestron with trypsin or a proteinase. Peptides are separated by HPLC and masses are determined. MALDI-TOF technique is widely used for studies of whole proteins in many laboratories. Fast atom bombardment (FAB) and MALDI techniques are applicable to protem mapping, and may be used to mvestigate protem-protein mteraction and tertiary structure 3.1. Characferizafion of Macromolecules Caprllary zone electrophoresis (CZE) is a method of choice for analysis of polypeptide samples. A comparatrve study of the method (on Bio-Rad-HPE m 20 cm x 25 mm capillaries) showed that tts performance was superior to TLC and paper electrophoresls, and equivalent to that of reversed-phase HPLC chromatography (29). Other separation techniques include electrophoresrs on sodium dodecyl sulfate-polyacrylamide gel electrophorests (SDS-PAGE), size-

Analysis

of Biopesticides

537

exclusion chromatography (SEC), reversed-phase HPLC, and centrifugatton. Capillary electrophoresis output combined with electrospray mass spectrometer (CE + ES). ES-MS and MALDI-MS are applicable to the determmation of accurate mol wt and purity of proteins. 3.2. Sequencing Peptide primary structures may be investigated using tandem mass spectrometry (MS/MS). Primary structure may be determmed at the pmol level, and peptides obtained from purified proteins may be sequenced by MS/MS techniques. However, it has been stated that sequencing by MS/MS is not yet a routine technique, and may not be applicable to determination of the complete sequence of every peptide. Interpretation programs may not yield single unambiguous results, and it does not replace Edman degradation or make it obsolete. MS/MS may be useful for primary structure determmation at the pmol level, and it has been reported that collisron-induced dissociation (CID) processes may be useful for resolvmg details of sequencing. A modified detector, which rapidly records the CID spectrum, increases the sensmvtty of the technique, which is also enhanced by elimination of transfer losses,if the digestion sample is directly introduced from a packed-capillary HPLC column interfaced with the first mass spectrometer (30). Automated sequencersare used to analyze small quantities of proteins or DNA samples.The structure of ohgo- and polynucleotides may be investigated by mass spectrometry, but they do not give good results under conditions normally employed for peptides and proteins. Desalting of DNA is important to avoid the presence of sodium and potassium ions, which bmd to DNA and gave rise to complex spectra. A variety of matrices have been evaluated for oligonucleotide mass spectrometry, and 6-aza-2-thiothymine, dissolved m 50% acetomtrile wrth 20 mMdrammonium citrate, appears to work well for modified oligonucleotides (32). DNA fragments of up to 426 base pairs have been analyzed, and the ultimate goal of such studies IS the potential application to DNA sequencmg. MS/ MS cannot be used for sequencing of DNA. A current limitmg factor in the utility of mass spectrometric techniques for DNA sequencing is the poor efficiency of electron multiplier detectors for detecting large tons; research is being conducted to develop detectors of Improved efficiency (31). Laser vaporization and FTMS are under investigation as a technique for DNA sequencing. Traditional techniques of structural investigations of carbohydrates involved lengthy derivatizatton procedures of derivatization and identification of individual mono- or oligosaccharide units obtained on hydrolysis or degradation. This field has been revolutionized by the applications of GUMS, FT NMR, and other newer instrumental techniques, such as FAB-MS of derivatized carbohydrates, which may be useful for sequencing.

538

Plimmer

4. Microorganisms

4.1. Genetically Engineered Microorganisms As gene transfer techniques have rapidly developed, there has been a corresponding growth in experimental and commercial production of genetically englneered microorganisms (GEMS) and plants. The fate of GEMS hasbeen amatter of concern, stnce the technology has become widespread. Although techniques of gene alteration may be applied to plants, animals, or mlcroorganlsms, technical dlfficultles associatedwith containment and monitoring the consequencesof their releaseshave become more complex. Imtlally, there were few releasesof GEMS into the environment, and they received intensive scrutiny. Much debate has taken place over the potential consequencesof releasing mlcroorgamsms. Few genes are added or deleted from the parent strain in the construction of a GEM, and the GEM may be expected to behave in the same way as the parent, unless the ecological properties are modrfied Major concerns over the release of microorganisms are that the organisms might be pathogenic or they might displace species that already exist in the environment. The need to monitor GEMS has been emphasized (8). Their ldentlficatlon m the natural environment presents problems of extraction and recognition. Assessment of the survival and persistence of GEMS m the sol1 environment has been dlscussed by Edwards (32). An accumulated decade of experience of the consequences of release of engineered organisms into the environment suggests that there do not appear to be greater risks associated wrth engineered organisms than with other organisms, either naturally occurring or genetically modified by mutation processes (chemical or irradiation), and so on. As a safeguard, a risk analysis must be built mto the process of regulatory approval, and this requires that procedures be available for detecting and monitoring GEMS in the environment

4.2. Monitoring Methodologies 4.2. I. Reporter Genes Reporter genes, or screenable markers, may be useful in ldentlfymg genetltally engineered cells, and such genesare routinely incorporated mto transgemc plants, animals, and microorganisms. Antibiotic resistance markers have frequently been used, because they allow the selection of recombinant organisms by plating onto a selective medium containing the antibiotic on which only resistant organisms will grow. An lmportant crlterlon in the selection of such a marker sequence is that it confers reslstance only to antibiotics of limited pharmaceutical use. Reporter genes most commonly used in detection and momtormg have been the GUS (P-glucuromdase: u&A) and 1acZ (P-galactosldase) genes from

Analysis of Biopestlcides

539

Escherichia colz, and the luciferase gene from firefly or bacteria that encodes biolummescence. The 1acZ system encodes /3-galactosidase production, and has been used in tracking engineered Pseudomonas strains in sods Strains incorporatmg this gene can be recognized by a characteristic blue color produced by incubation with a specific substrate. Other genes that encode for chromogenic reaction potential (such as production of catechol2,3-oxgenase) have also been used. Potential usesand improvements m the utility of the Green Fluorescent Protem gene have been discussed (33). Cells transformed by this reporter gene, which is obtained from the jellyfish, show bright fluorescence under UV illummation. It can be detected by noninvasive techniques, and has potential application for monitoring organisms released into the environment Genetic modifications of maize containing a gene expressing insect resistance have been developed. A marker gene for antibiotic resistance was also mcorporated in some of these plants, but there is a concern that release of such an antibiotic resistance gene m the environment might be associated with the risk of increased resistance of bacteria to antibiotic drugs. The presence of such a gene seems likely to delay regulatory approval in Europe, and other varieties of maize, which do not mcorporate the antibiotic resistance gene, are being developed. 4.2.2. Techmques for Recognition of Organisms The three major categories of detection methods mclude culture and metabolic techniques, genetic techniques, and mnnunological techniques (13). 4.2.3. Culture of Organisms Traditional techniques of culture and metabolic techniques will be useful for confirmatory procedures. Conventronal cultural methods can be used with selective or nonselective media, and require an moculant contaming at least 100 bacteria/ml. The method of counting colonies on agar plates has been used extensively, and the data are subject to statistical analysis The orgamsm must be culturable, and, in the case of an engineered organism, an inserted marker must be present, which is stable, and without influence on metaboltsm or ultimate survival of the organism. Accepted FDA methods require enrichment m broth, followed by enrichment m agar, and culture on nonselecttve agar. This is a time-consuming procedure. Since this method provides amplification, it merits more research. If the organism is not culturable, staining may afford an alternative. Fluorescent antibodies may also be added to the plate. The sample medium may present a source of problems. The PCR technique requires 1O4organrsms/mL, and an additional hmttation is that polymerase may

540

Pllmmer

be mhtbited by media. Immunoassay requires about lo4 orgamsms/mL. The generation of fluorescent bacteria may be very useful for microscopy of organisms that are viable, but nonculturable. 4.2.4. Whole Cell Analysis Helm et al. (34) have described a computer-aided procedure for identifying bacteria based on FT-IR data. They utilized a library based on 97 strains, and indicated the need for suitable databases.Pyrolysis under controlled condmons 1scapable of providing much data, but tts interpretation will depend on patternrecognmon techniques and the avatlabtlity of adequate reference libraries (35”. 4.2.5. Metabolism Analysis of spent media, or characteristic products of bacterial metabolism, will provide information that can be used to classify or identify bacteria, Data may be obtained by GLC or HPLC analysis, and, additionally, these techmques may be coupled with mass spectrometric analysts. Much mformatton is obtained that IS characteristic of the organism, but recognmon and identtficatton depends on pattern recognmon, and matching the patterns with those m existing reference databases. Profilmg of lipids has been investigated as a recognition technique: Profiles of triglycerides, wax esters, fatty acids, or phospholiptd ester-lmked fatty acids may be useful for characterization or recognition of microorganisms (36,37). Fatty acid profiles have been used to characterize Bt strains from ancient samples of ambers. Cloned colonies were grown under standardized conditions, the cell mass was saponified, and free acids were methylated and analyzed by GLC The dtstrtbutton of fatty acids was subjected to multivariate analysis to assessthe distribution of fatty acid variation m populations of similar bacteria (38). Other techmques that have been employed Include macromolecular profiles, DNA fingerprinting, electrophoretic polymorphism of enzymes or total protems DNA fingerprinting, and electrophorettc polymorphism of enzymes or total proteins (39). 4.2.6. Immunological

Methods

Immunologtcal procedures, such as the use of fluorescent-labeled, monoclonal anttbodtes, may complement macroscopic methods of identification. Immunoassays are rapid, convenient, and adaptable to field or laboratory sttuattons They reqmre antibodies specific for the gene product or microorganism of interest. A wide variety of antibodies is readily available commercially, and are in routme use for diagnosis of pathogens, mycotoxms, and so on. Detection techniques using nnrnunologtcal approaches are epifluorescent microscopy (requires observation of about 50 fields/sample), dot-blot meth-

Analysis of Biopesticdes

541

ods, and agglutination (requires antiserum and 30-60 s reactton time, and can be used at log organisms/ml level [coagglutination may increase sensitivity to lo6 orgamsms/mL]). A species-specific antiserum and a fluorescent dye are required, and the specimen is viewed with an epifluorescent microscope, which reveals the cell as a fluorescent green band beneath the cell wall. Several formats may be used for immunoassays. In the simplest form, polyclonal antibodies are bound to a solid surface. A solution contammg the target analyte (antigen) is added, together with a polyclonal antibody linked to an enzyme. If an antigen is present, this binds to the polyclonal antibody at the solid surface, and the enzyme-linked polyclonal antibody also binds to antigen. If enzyme IS present, the addition of an appropriate substrate generates a colored product that can be determined spectrophotometrically. In a variatton of this method, the antigen is bound to the well of a microttter plate. A specific polyclonal antibody is added. This binds to the antigen, and the antigen-specific antibody binding is detected by the addition of another antibody conjugated with an enzyme. A chromogenic substrate is added to the wells, and the presence of antigen-antibody-enzyme complex is determined colonmetrically. Immunofluorescence techniques employ specific polyclonal antibodies that will bind to a spot of the antigen bound to a glass slide. The binding site is detected by adding an antibody conjugate labeled with a fluorescent reagent (fluorescein isothiocyanate), and viewing the plate under UV light For the dot-binding assay,antigen is bound directly to a nitrocellulose membrane, and sites that are unbound are covered with a blocker (buffer-milk powder). Spectfic polyclonal antibodies are added and bind to the antigen. The specific antibody bound to the antigen is detected by addition of a proteinenzyme conjugate, and an appropriate enzyme substrate is used to visualize the binding site. Organisms cultured in media may be concentrated by immunocapture, by adding magnetic beads coated with antibody. Beads are then added to fresh media as an inoculant. The advantages of tmmunologtcal methods are: they can detect specific bacteria within highly mixed populations; they detect total cells; sample preparation may be simpler; and nnmunological methods are less expensive than DNA-based techniques. 4.2.7. Genetic Techniques Immunological and genetic techniques are applicable to samples of large size. Genetic techmques rely on the specificity of gene probes, which detect sequencesof nucleic acids. Gene probes allow tracking of the genome, and this method can be used without culturing and without the necessity of having a specific selectable marker. Gene probes have the advantage that a gene can be detected, even if it has been transferred to another organism.

542

Phmmer

Colony hybrrdization is the simplest molecular approach for detection of GEMS, which can be combined with conventional environmental microbiological sampling and analysis; its advantages have been summarized (40). Bacteria are grown on solidified agar media to form colonies. Specific-target nucleic acid sequences are then detected by gene probes and nucleic acid hybridization. After transfer to hybridrzation filters, colomes are lysed and hybridization is conducted. PCR-based technologies appear promising for wild-type and recombinant viruses. Complementarity is the term given to the binding of two macromolecules on the basis of sequence-specific or steric-molecular recognition. It may be used to describe the bmdmg of the two strands of the double helix of DNA, or the binding of an antibody to a protein by specific steric recognition. On this basis, a probe molecule can bmd with a target molecule to form a hybrid. If the probe is radioactively labeled, hybridization can be used to obtain information about the target molecule. The hybridization of two complementary nucleic acid sequencesprovides a specific technique for recognmon of particular sequences. This IS useful for detection of organisms that cannot be cultured, or do not have a detectable marker gene (such as resistance to antibiotics) It also has the advantage that a gene can be tracked, whether it is expressed or transferred to another organism. The technique is capable of higher sensitivity than can be achieved by other methods Three classesof genes have been used as probes (23). These are probes that target ribosomal RNA, randomly cloned sequencesof DNA unique to the organism of interest, or the engineered sequence. For a GEM, the sequenceof interest is already available, and this often determines the most convenient choice. Dot-blot, slot-blot, or Southern-blot techniques may be used for detection. The target organism usually contains many nucleic-acid sequences present m genes, mRNA, and so on. In the Southern blot procedure, the DNA is cut with restriction enzymes.The double-stranded DNA fragments have an extended-rod conformation that can be separated by electrophoresis on a solid gel support, on a mol wt basis. Bands are stained with a suitable stammg agent (such as ethldmm bromide for nucleic acids), and mol wt are estimated by comparison with standards. The DNA bands to be examined must be adsorbed on a solid support (usually nitrocellulose paper) for the hybridrzation process (“blottmg”). DNA IS transferred to the support, and, since it is double-stranded, it must be converted to the single-stranded form before hybridrzation (“denaturation”); conversion is accomplished by conductmg the transfer m a strongly alkaline buffer. The labeled probe (DNA labeled with 32P)is added to the support (after blocking sites on the filter that might adsorb the probe) and nonhybridized DNA is washed away. The amount of radroactrvity remaining is determined by autoradiography or LSC. 1O4organisms could be detected m Southern blots (41).

Analysis of Biopesticides

543

Probes may be prepared from plasmids contammg the gene of interest. Fragments obtained by treatment of the plasmrd with restriction enzymes are separated by gel electrophorests. The polymerase chain reaction (PCR) technique allows the target sequence to be magnified by adding a primer unique to the target, followed by cycles of synthesis of complementary strands and denaturatton. PCR allows gene sequences to replicated m vttro and amphfied exponenttally, thus enhancing the probabthty that specrfic sequences present in GEM may be detected. The method, which requires only picogram amounts of DNA, has been used for detecting GEMS m complex environments. To perform gene probe analyses without culturmg microorgamsms by direct recovery of DNA from environmental samples, such as water or soils, cells may be collected from water samples by filtration or by centrifugatlon of ~011s. Subsequently, cells are lysed by chemical or phystcal methods. DNA may then be extracted and purified (42). Gene probes and nucleic acid hybridtzatton procedures are then used to detect DNA sequences. The combmatton of PCR technique and restriction endonuclease analysis has been used to detect baculovnuses A group of baculovnuses m which nuclear polyhedrosts virus was embedded (Autographzca callfornzca multtple-embedded nuclear polyhedrosis VU-US (MNPV), Anticarszagemmatalls MNPV, Bonzbyx mori MNPV, Origia pseudosugata MNPV, Spodoptera frugiperda MNPV, S Exigua MNPV, Anagrapha falclfera MNPV, and Hellothzs zea smgle-embedded nuclear polyhedrosis vnus) was selected for mvesttgatron Distinct profiles were obtained for each virus by amphfymg a highly conserved DNA coding sequence, and analyzing the PCR products by restrtctron analysts (43). 5. Analysis of Bacillus thuringiensis Bacillus thuringiensis (Bt) IS a major mtcrobtal msecticlde and a source of genes encodmg several proteins toxic to insects. Consequently, there IS now considerable experience with its analysts (44). Bt product IS produced by fermentation process; the broth contams a mtxture of proteins, bactertal metabohtes, spores, and growth media components. These components, whtch act synergistically, vary m their toxtctty to different insects. 5.7. Quantification of Active Components The principal toxins of Bt are the S-endotoxms or msectictdal crystal proteins (ICPs), a group of structurally related proteins that are present as crystalline inclusion bodies m sporulated cultures. 5.1.1. Standardized Methods for Potency and Analysis of Endotoxins Regulatory methods have been based on btoassay methods, and methods for the quantitative analysis of &endotoxm of Bt preparations have been revtewed

544

Plimmer

(45). The insecticidal spectrum of each Bt subspecies differs; these have been divided mto four major classes,depending on the &endotoxins produced. Each of the four classests spectfic for insect orders. An official regulatory notice (PR 7 l-6), stating that no quantitative analytlcal procedures had been developed for &endotoxins, led to the development of a standardized bioassay to determine the active ingredient. A primary standard E61 (Bt subspp thuringiems) was adopted as international primary standard for analysis of Bt spore-crystal complexes, and was assigned a potency of 1000 IU/mg. The test insect was the cabbage looper (Trichoplusza ni [Hubner]) Product labels were required to list the active ingredient percentage, based on the assumption that a 100% product would contain 500,000 IU/mg. Newer standards of higher potency have been developed (46). A new registration standard (47) replaced the requirements of the earlier regulatory notice (PR Nottce 71-6), and required that the label specify the active mgredient in terms based on percentage by weight of insecticidal proteins determmed by analytical methods. The percentage of &endotoxms was to be specified for each order of msectsaffected. Potency umts remamed on the label as an option. Doubts have been expressed that m vitro methods of analysis ~111provtde a viable alternative to m vtvo methods for quantttatton of toxm btoactivtty (48). The retention of bioassay methods has been suggested, specifically, the flytoxicity test (49), to guarantee the absence of exotoxins that may not be detected by standard HPLC procedures. It has been proposed that ELISA-based assays may provide a convenient technique for the assayin msecttcidal protems. However, caution is required in extrapolation of the results to activity, because the proteins may be highly spectfic, and the biologtcal acttvity of combmattons of proteins occurring in Bt isolates may not be that expected from the sum of the component proteins (50). The msect toxins produced by Bt kurstaki have been detected by nnmunoassay technology developed by Strategic Diagnostics (Newark, DE) (51) The technology is apphcable to testing for transgene verification m a variety of crops A raptd, field-portable nnmunoassay test with a nonenzymattc visual detection system should be readtly adaptable to screening and monitoring the presence and tdenttty of transgenic products. Procedures have been developed for the immunoassay of msecttcidal protems, and an ELISA protocol for the total analysts of CryIA has been described (52). Results were in good agreement with those obtained by protein gel assay. ICPs toxic to lepidopterans are grouped m the class designated CryI, Cry II, and so on, and there are many related genes encoding different ICPs. Among the important questtons concerning the use of Bt is to determine how many ICP genes are present m parttcular Bt strains, and the levels of expression of

Analysis of Biopesticides

545

individual genes. These questions were investigated by Masson et al. (53) by gene cloning, oligonucleotide probes, and cyanogen bromide cleavage. The &endotoxins are generally mixtures of proteins of mol wt from 27 to 140 kDa. The composttion of the mixture depends on the strain of the organtsm, but chromatographic techniques have proved inadequate to separate the proteins sufficiently to distinguish Bt subspecies. Yamamoto (549has described the isolation of the crystalline proteins from a Bt culture. Endogenous proteinaseswere removed by sodium chloride treatment, and the crystal was solubihzed at high pH. Sporesremained msoluble. The supematant was applied to a Sephacryl S-300 column, and eluted with 50 mM Tris-HCl, pH 8.0, containmg 0.1% 2-mercaptoethanol and 1mM EDTA, monitoring at 280 nm. Proteins were separatedby size.The principal components were a major protein, PI (135 kDa), and a second protein, P2 (65 kDa), responsible for mosquitoctdal activity. Then purities were assessedby SDS-PAGE or immunoelectrophoresis. Subsequently, the isolated proteins were digested with trypsm, and the complex peptide mixtures formed were separated by HPLC, using a C- 18 reversedphase column. Peptides were eluted with phosphoric acid, with a gradually increasing concentration of acetonitrile. The chromatogram provides a charactenstic fingerprint, and over 20 strains were selected for pepttde mapping. To purify protem Pl more rapidly than by Sephacryl column chromatography, an antiserum against PI from Bt kurstaki HD73 was prepared. The antibody was purified and immobilized on CNBr-activated Sepharose to provide an affinity column for chromatography. Peptide mapping provides information on the type of crystal proteins encoded by the genes present in individual strains. There are three genes termed cry1A in the HDl and some other strains of Bt kurstakz. Differences m the amino acid sequences in the protems encoded by these genes may cause significant differences in the spectrum of biological activity. Several cry genes are present in commercral Bt strains, and individual genes may be highly expressed. Information concerning which genes are expressed and the level of expression is important in the development control strategiesutilizing Bt toxins. 5.7.2. The Presence of Exotoxins In addition to the &endotoxins, some Bt strains produce a heat-stable, insecticidal, adenine-nucleotide analog, known as p-exotoxin or thuringiensm. This showed a broad range of toxicities to organisms, including mammals (moderate-to-high toxicity, oral acute LD,, rat approx 170 mg/kg). Its toxicity may be caused by its ability to inhibit DNA-directed RNA polymerase by competing with ATP, and, because mammalian mRNA polymerases are sensitive to /3-exotoxin, Bt active ingredients must be tested to show the absence of p-exotoxin as a condition of registration for use on food in the United States(55).

546

Plimmer

A housefly, Musca domestzca,bioassay of autoclaved supernatantsIS used to detect the presenceof exotoxms.HPLC has beenused m attemptsto develop raptd, sensitive methods for detection and quantitative analysis of @exotoxm. Standard methods may require modification to reveal the complete spectrum of p-exotoxms present. A study of location of genes mvolved m p-exotoxin production showed that p-exotoxin production was plasmtd-encoded m sex strains. In the caseof the HD-12 strains, no /3-exotoxm peak was observed on HPLC. However, housefly toxtcity was observed, and a new toxic factor, termed P-exotoxm, Type II, was isolated and partly characterized.The p-exotoxms were purified on HPLC (using a Vydac 218TP54 wide-pore, reversed-phasecolumn with 57 &acetic acid mobile phase,pH 3.0) by modifymg previously developed HPLC methods. 5.2. Methods for Quantitation of Bt Preparations and Residues in Field Tests 5.2.1 ELBA Enzyme-linked tmmunosorbent assay can be used to determine the total &endotoxm proteins m residues in the field. This techmque IS useful for the investigation of the fraction of toxins deposited and their effects on nontarget organisms. 5.2.2. Total Protein Method This method is rapid and low cost, compared with ELISA, but ignores toxicity caused by spores. Determmation of total protein has been used to determine residual amounts of Bt on foliage of a spruce fir forest. Spray applications of Bt berlmer subsp kurstaki were applied to control spruce budworm (Chorzstoneura fumiferana Clem). Total protein 1smeasured by the bicmchommc acid method (56). 5.2.3. Bioassay Methods Bioassay is relatively easy to conduct. The cabbage looper has been used as the test organism m the standardized bioassay for potency. Sundaram et al. (57) have described bioassays for testing persistence of residues, using the spruce budworm, C fumzferana Clem. Determination of levels of protein expressed m transgenic plants is necessary to determine their agronomic potential, and may be accomplished by bioassay (58) 6. Analytical Requirements in Field Testing 6.1. Formulation of Active Ingredient Btopesticide efficacy is highly dependent on methods of delivery that adequately protect the agent against the action of sunlight and oxidants in the

Analysis

547

of Biopesticides

envtronment. An analytical and/or bioassay process should be included m field evaluatton, to guarantee that the active ingredient remains viable. The half-life of Bt is l-4 d on the surface of bean leaves m sunlight; other microbtals are also rapidly inactivated m sunlight. Design of delivery systems must overcome these hmitations. It is necessary also to deliver droplets, each contammg a lethal dose of Bt toxin, because the toxin rapidly causes cessation of feeding. A sublethal dose may only have a temporary effect, and the Insect may recover and resume feeding.

6.2. Baculovirus Field-Test Approval by EPA Field testsof engineered organisms are subjectedto a variety of reqmrements to ensure that there are no adverse environmental effects. Small-scale field trials to assessthe impact of release of a baculovu-us AaIT strain, modified to express the insect control protein from a North African scorpion, are typical. This involved trials m 12 statesagamsttobacco budworm and cabbage looper on cotton, tobacco, and leafy vegetables, at a dose rate of 100 g active ingredient on 7.4 acres. In similar field testsconducted m 1995, EPA requested sot1sampling data to evaluate survival and persistence of the organism. In the 1996 tests, sot1 samples were to be taken at specified points during the course of the release: prior to the final application of the recombinant virus, following apphcatton, just prior to spraying with wild-type virus, and after allowmg sufficient time for dispersal of wild-type vn-us m the soil. The following assays were to be conducted: bioassay with highly sensitive insect to detect infectious polyhedra, and PCR assayto detect the recombinant gene construct. Finally, lime was to be applied to the sot1to raise pH, thus inactivating residual virus 7. Conclusion Methods of analysts are essential for effective and safe application of biopestictdes. The quantity of active ingredient applied must be known, and, to guarantee thts, the quantity present in the formulatton to be used must be measured accurately. Residues that remam m or on treated materials, and in the environment, must also be measurable, because such data is needed to assess risks entailed in using the product. Analyses of formulated products and the residues generated during their use usually require different approaches. In the latter case, higher sensitivities are required. The identity of terminal residues is important. Good analytical practices require frequent calibration of equipment and use of standards during data acquisition. Additional confirmatton of identity by spectroscoptc methods and other instrumental techniques is essential to add confidence to the data. Sensitivity m methods of pesticide analysts for quantttation of btologtcally active ingredient(s)

has increased

by orders

of magnitude

in recent decades.

548

Plimmer

However, for many biopesticides, the question of analysts ISoften less straightforward. Biopesticides include many different types of agents, which range from complex macromolecules to whole organisms. Questions of identity and homogeneity may present difficulties, when, for example, fermentation broths are to be used. Methods of analysis may range from conventional chemical approaches to biological techniques Quantitative analysts of complex molecules, particularly those of btological origin, is becoming a routme matter, but the need for expensive, sophtsttcated mstrumentation often restricts some analyses to a limtted number of laboratories. The scope of analytical methods requires, depending on the nature of the problem, not only a sound background m analytical chemistry, but also skills that are essential to biotechnology: a comprehensive combmatton of biochemical and microbiologtcal expertise. Because approaches to the analysis of biopesticides have originated m a variety of disciplinary areas, it is no simple task to summarize their current status, in view of the rapid progress in applying new technology to the diverse areas under consideration. In addition to the problems of analysis, btologtcal matertals, such as plants or mtcroorganisms, that are capable of self-rephcation present different challenges, and are usually not amenable to classical analytical techniques. Implications of their environmental release must be investigated by appropriate biological techniques. Methods are available for the identification of genetic materials, but the assessmentof the significance of gene transfer, beyond confirmation of identity and determination of the extent to which it has occurred, lies outside the scope of the present discussion. References 1 US Statutes at Large (1910) Federal Insectzczde, Fungzczde, and Rodentzczde Act. vol 1, pp 331-335 2 US House (1996) Public Law 104-170 (HR 1627) (7 USC 136 note) Food Quality Protection Act of 1996. 110 Stat 1489. 3 (1989) Title 40 Code of Federal Regulatzons Part 160 Good Laboratory Practice Standards (EPA), (1979) Final Rule (1989). 4 Anonymous (1989) EPA Pestzcide Assessment Guzdelznes Subdzvzszon M Part A (Mzcrobzal) Series 153A. 5 EPA (1996) Mzcrobzal Pestzcide Test Guzdelznes, OPPTS 885, EPA 712-C-96280, February 1996 6 Federal Regzster 47,23928 7. Wtlkinson, C. F. (1996) When is a plant a pesttctdev Pestzczde Ozctlook 7,40,41 8 Beringer, J E and Bale, M. J (1988) The survival andpersistenceof geneticallyengineered mtcroorgamsms, m Release of Genetically-Engineered Mzcro-Organzsms (Sussman, M , Colms, C H , Skmner, F. A , and Stewart-Tull, D. E , eds ), Academtc, London, pp 2946.

Analysis

of Biopesticides

549

9. (1995) Chemistry and Engtneertng News, Jan 30 (1995) US Patent 538083 1 10 (1995) New York Times, 1 I Aprtl 11. McClmtock, J. T., Schaffer, C. R., and Sjoblad, R T. (1995) A comparative review of the mammalian toxicity of Bacrllus thuringzensu-based pesticides Pesttctde SC1

45,95-105.

12 (1986) Title 40 Code of Federal Regulations Part 180. Tolerances and Exemptions from Tolerances m or on Raw Agricultural Commodities. 13. Kearney P. C and TiedJe, J M. (1988) in Biotechnology for Crop Protectton (Hedin, P A., Menn, J. J , and Hollmgworth, R. N., eds ), ACS Symposium Series, 379,352-358. 14. Anonymous (1995) Agrow 225, 10 15 DowElanco ( 1994) Spinosad Technical Guzde 1994 16. Calderon, L., Brunetto, R., Leon, A., Burguera, J. L., and Burguera, M. (1996) HPLC determination of gentamicm m pharmaceutical dosage forms by postcolumn derivatization with o-phthalaldehyde Am Lab 56-59. 17 Alvinerte, M., Sutra, S. F., Badri, M., and Galtier, P. (1995) Determmatton of moxtdectm m plasma by high-performance liquid chromatography with automated solid-phase extractton and fluorescence detection. J. Chromatog. 674, 119-l 24. 18. Nakamshi, K. (1975) Structure ofthe Insect anttfeedant, azadnachtm, m Recent Advances m Phytochemutry, vol. 9 (Runeckles, C., ed.), Plenum, New York, pp. 283-298 19. Isman, M B., Koul, O., Lowery, D. T., Arnason, J. T., Gagnon, D., Stewart, J G., and Salloum, G. S. (1990) Development of a neem-based insecticide m Canada, m Neem’s Potenttal tn Pest Management Programs, Proc. USDA Neem Workshop, Beltsville MD April 16-l 7, USDA, ARS-86, 32-39. 20 Warthen, J D , Jr., Stokes, J B , Jacobson, M., and Kozempel, M. P. (1984) Estimatton of azadirachtm content m neem extract and formulations J Lzqutd Chromatog 7,59 l-598. 2 1 Cohen, E , Quistad, G B , Jefferies, P. R , and Casida, J E. (1996) Nimbohde is the prmcipal cytotoxic component of neem-seed insecticide preparations. Pesttcrde Set. 48, 135-140 22. Plimmer, J R. (1996) Regulation of natural pesticides, m Crop Protectzon Agents from Nature, Natural Products, and Analogues (Coppmg, L. C , ed.), Crtttcal Reports on Apphed Chemtstry 35, Royal Society of Chemistry, Cambridge, UK, pp 468-489. 23. Anonymous (1995) Lepidopteran pheromones: tolerance exemption, Federal Regzster. August 30,1995 (Vol. 60, No. 168) EPA, OPP 40 CFR Part 280, pp. 45,06045,062 24. Spittler, T D. (1994) Effect of regulation of pheromones as chemical pesticides on their viability in insect control, in Natural and Engineered Pest Management Agents (Hedm, P. A, Menn, J J., and Hollingworth, R N , eds.), ACS Symposium Series 55 1, ACS Washington, DC, pp. 509-5 15. 25 Bestmann, H. J. and Vostrowsky, 0. (198 1) Chemistry of insect pheromones, m Chemte der Pjlanzenschutz-und Schaedlingsbekaempfungsmittel, vol 6 (Wegler, R., ed.), Springer-Verlag, Berlin, pp. 29-164

550

Plimmer

26 Klun, J A , Phmmer, J R , Bierl-Leonhardt, B A , Sparks, A N , and Chapman, 0 L (1979) Trace chemicals the essence of sexual communication systems m Heltothts spectes Science 204, 1328-l 330 27 Patterson, S. D and Aebersold, R (1995) Mass spectrometric approaches to the identification of gel-separated proteins Electrophorests 16, 1791-l 8 14 28 Frenz, J , Battersby, J., and Hancock, W S (1990) An exammation of the potential of capillary zone electrophoresis for the analysis of polypeptrde samples, m Proceedtngs of the Eleventh Pepttde Symposium 1989 (Rivier, J. E and Marshall, G R , eds ), ESCOM, Amsterdam, pp 430-432 29 Btemann, K , Btller, J E , Hill, J A, Johnson, R. S , Martin, S A., Pappanopolous, I A , and Vath, J. E. (1990) Determmation of the sequence of peptrdes and proteins by tandem mass spectrometry, m Proceedings of the Eleventh Pepttde Sympostum 1989 (Rivier, J E and Marshall, G R , eds ), ESCOM, Amsterdam, pp 426429 30 Lecchi, P , Le, H. M T , and Pannell, L. K. (1995) 6-Aza -2-thiothymine a matrix for MALDI spectra of ohgonucleotides Nucleic Acids Res 23, 1276,1277 3 1. Murray, K. K (1996) DNA sequencing by mass spectrometry J Mass Spectrom 31, 1203-1215 32 Edwards, C (1993) The significance of zn sztu activity on the efficrency of momtormg methods, m Monttortng Genettcally Mantpulated Organisms tn the Envtronment (Edwards, C., ed.), Wiley, Chrchester, UK, pp. l-25 33 Prakash, C S (1996) Green Fluorescent Gene. A New Reporter Gene m Transgemc Research, ISB News Report, NBIAP, VPI, Blacksburg, VA, April, pp 2-4 34 Helm, D., Labischmskt, H., Schallehn, G., and Naumann, D. (1991) Classrfication and identification of bacteria by Fourier transform infrared spectroscopy J Gen Microbial 137,69-79. 35. Magee, J T , Hmdmarch, J. M., Duerden, B. I., and Mackenzie, D W. (1988) Pyrolysis mass spectrometry as a method for inter-strain discrimmation of Candada albtcans J Gen Mtcrobtol 134,2841-2847 36 Shaw, N., and Stead, D (1970) A study of the hprd composttton of Mtcrobacter-turn thermosphactum as a guide to its taxonomy J Appl Bactertol 33,470-473 37. Snyder, A. P., McClennen, W. H., Dworzanski, J. P , and Meuzalaar, H L (1990) Characterization of underivatized lipid bromarkers from mtcroorgamsms with pyrolysis short-column gas chromatography/ion trap mass spectrometry Anal

Chem 62,2565-2573 38 Cano, R (1996) Characterizing ancient bacteria Anal Chem 68,609A--611A 39 Ricard-Pasquier, N , Ptcard, B , Heeralal, S., Krishnamoorthy, R., and Goullet, P (1990) Correlabon between rtbosomal DNA polymorphrsm and electrophoretic enzyme polymorphism in Yersinia J Gen Mzcrobiol 136, 1655-1666 40 Atlas, R M , Sayler, 0 , Burlage, R S , and Bej, A. K (1992) Molecular Approaches for environmental monitoring of organisms Bio/Technzques 12,70&7 11, 41. Holben, W. E , Jansson, J K , Chelm, B. K , and TiedJe, J M (1988) Appl

Environ Mtcrobtol

54,703.

42. Atlas, R. M. (1992) Molecular methods for environmental monitoring and containment of genetically engineered microorganisms Bzodegradatzon 3, 137-146

Analysrs

of Blopesticides

551

43 de Moraes, R R and Marumak, J E. (1997) Detection and tdenttfication of multiple baculovuuses using the polymerase chain reaction and restriction endonuclease analysts J Vzrol Methods 63,209-2 17. 44. Htckle, L A. and Fitch, W. L., eds. (1990) Analytzcal Chemzstry of Bacillus thurmgiensis, ACS Symposium Series 432, ACS, Washmgton, DC 45. Beegle, C. C (1990) Bioassay methods for quantification of Baczllus thurzngzensls &endotoxm, m Analytzcal Chemrstry ojBacil1us thuringiensts (Hickle, L A and Fitch, W. L., eds.), ACS Symposium Series 432, ACS, Washington, DC, pp 14-21. 46. Tompkins, G., Engler, R., Mendelsohn, M., and Hutton, P. (1990) Historical aspects of the quanttftcation of the active ingredient percentage for Bacillus thurzngzenszs products, m Analytical Chemistry oj’Bacrllus thurmgiensis (Hickle, L A. and Fitch, W. L., eds ), ACS Symposium Series 432, ACS, Washmgton, DC, pp. 9-13 47 Anonymous (1988) Reference registration standard for the re-registration of pesticides containing Bacdlus Thuringzenszs as the active ingredient. Case Number 0247 EPA, OPP 1988. 48 Schwab, G. E. and Culver, P. (1990) In vitro analysis of Baczllus thurlngzensls &endotoxm action, in Analytical Chemzstry of Bacillus thuringiensis (Htckle, L. A. and Fitch, W L , eds ), ACS Symposium Series 432, ACS, Washmgton, DC, pp 3645. 49 Levmson, B. L. (1990) High performance liquid chromatography analysts of two (p-exotoxins produced by some Bacillus thurmglenszs strains, m Analytical Chemzstry ofBacillus thuringtensts (Hickle, L A and Fitch, W L , eds ), ACS Symposium Series 432, ACS, Washington, DC, pp 114-136. 50 Mtlne, R , Ge, A. Z , Rivers, D., and Dean, D H (1990) Specificity of msecttctdal crystal proteins: implications for industrial standardization, m Analytzcal Chemutry ofBacillus thurmgiensis (Htckle, L A. and Fitch, W L., eds ), ACS Symposium Series 432, ACS, Washington, DC, pp 22-35 51 (1997) ZSB News Report, National Biological Impact Assessment Program, VPI, Blacksburg, VA, September 1997, p 3. 52. Groat, R. G G., Mattrson, J W., and French, E J. (1990) Quantitative immunoassay of msectictdal proteins in Bacillus Thurlnglensls products, m Analytical Chemistry ofBacillus thuringiensis (Hickle, L A and Fitch, W. L , eds.), ACS Symposium Series 432, ACS, Washington, DC, pp. 88-97 53. Masson, L., BOSS& M , Prefontame, G., Peloqum, L., Lau, P C K., and Brousseau, R (1990) Characterization of parasporal crystal toxins of Baczllus thurznglenszs subspecies, kurstakr strains ED-1 and HD-2, in Analytical Chemzstry ofBactllus thurmgiensts (Hickle, L A and Fitch, W. L., eds ), ACS Symposium Series 432, ACS, Washmgton, DC, pp 6169. 54 Yamamoto, T (1990) Identificatton of entomocidal toxins ofBacillus thunngzensu by high-pressure liquid chromatography, m Analytical Chemistry of Bacillus thurmgiensis (Hickle, L. A and Fitch, W. L , eds.), ACS Symposium Series 432, ACS, Washmgton, DC, pp. 46-60. 55 Anonymous 40 CFR Section 180. 1011.

552

Plimmer

56. Sundaram, A , Sundaram, K. M S , Leung, J W., and Sloane, C L. (1994) J Envzron Scl Health 24,615-703. 57 Sundaram, A, Sundaram, K M S , and Sloane, C L (1996) Spray deposttton and persistence of a Baczllus thurmglensu formulation (Foray@ 76B) on spruce foliage, followmg aerial apphcatton over a northern Ontarto forest J Environ Health SCL 31,763-813. 58. Fuchs, I. L. L., Mackintosh, S. C., Dean, D. A, Greenplate, J T , Perlak, F. J , Pershing, J C , Marrone, P 0 , and Fischoff, D A. (1990) Quantttation of Baczllus Thurzngzenszs insect control proteins as expressed m transgenic plants, m Ana1ytzcaEChemrstry ofBacillus thurmgiensis (Hickle, L A. and Fetch, W. L., eds ), ACS Symposium Series 432, ACS, Washington, DC, pp 105-l 13 59 (1966) ZSB News Report, National Biological Impact Assessment Program, VPI, Blacksburg, July 1966, pp. 1,2

29 Principles of Dose Acquisition for Bioinsecticides Hugh F. Evans 1. Introduction Although dose acquisition is a process common to all pesticides, including btopestictdes, the mformation on use of mtcrobial agents is dominated by bioinsecticides, reflecting the need to manage the many insect pests m all sectors of crop production. This chapter, therefore, deals with aspects of the dosetransfer process for bioinsecticides, concentrating on the key interactions between hosts and microbial pathogens. In this respect, dose acquisitton is considered from the three fundamental aspectsof the target host, the microbial agent, and the matching of dosage to target in the field. Discussion then concentrates on how quantitative btological information on these three aspects enables dose requirement and tank mix for the given biomsecticide to be calculated. 1.1. The Nature of Microbial Bioinsecticides Although microbial bioinsecticides are often considered to be direct analogs for chemical pesticides, particularly when they are merely substituted for them during spray applications, this disguises the many fundamental differences that must be understood to allow microbial agents to be used successfully. These dtstmctions are certainly a key theme in many of the papers m Evans (I), m which the authors discuss the biological, technological, and environmental Issues affecting use of microbtal insecticides. Evans (2) has discussed some of the principal features of microbial biomsecticides, emphasizing their biological and physical attributes, as well as the major constraints on then use for pest management. Routes of infection are determined primarily by the behavior and ecology of the target host, and, in the majority of cases,require ingestion of the infectious unit (IU), thereby placing From Methods m Botechnology, vol 5 Blopestmdes Use and Del/very Edled by F R Hall and J J Meno 0 Humana Press Inc , Totowa, NJ

553

554

Evans

emphaseson host feeding habits. The ecology of the pathogen itself 1salso an important attribute m determmmg the roles of secondary inoculum m further mfection and disease expressron. In fungal pathogens, however, the prmctpal route of infection is direct contact between the fungal spore and the host integument (3). By contrast, the majority of chemical insecticides have a combination of contact, ingestion, and/or fumigant effects. Successful mfection results in etther replication, in the case of fungal, protozeal, and viral pathogens, or toxemia (possibly with rephcatron), m the case of bacterial mfections. The ability to replicate, however, is a fundamental difference between microbtal and chemical pesticides, because it introduces both a time delay in host mortality, and also results m multiplicatton of the pathogen. The latter attribute gives rise to secondary moculum and the possibility of pathogen perststence, contrrbutmg to the potential for further infection in susceptible host populations. These attributes provide opportumties, as well as constraints, m the use of mtcrobial btomsecticides, and they must be assessed quantitattvely, tf these agent are to be used effectively for pest management. The essential features of microbtal agents as pesticides are considered by a number of authors (&9J. Knowledge of the biological attributes of microbial btomsecticides 1s not only of fundamental importance to then- potential for pest management, but must also be extended to methods of application and delivery to the target hosts, The process of dose acquisition has both biologtcal and physical charactertstics, the latter being dominated by the particulate nature of mtcrobial agents, which confers limits on the degree of dilution of the pathogen that can be achieved m preparations formulated for spray application. Below a certain concentration, depending on droplet size, a point is reached at which droplets will not contain any active ingredient, and thus will not contribute to the process of dose acquisition. Particulate entitles in liquid suspenston wtll reach thts point sooner than active ingredients m solutton. The purpose of this chapter, therefore, is to bring together the biologtcal and physical processes that must be considered m designing a pathogen-apphcation system to optimize dose acqutsttion by the target organism. The prmctpal variables of the target population (Subheading 2.) and the microbial agent (Subheading 3.) are nntially considered separately, and then brought together m matching delivery to the target (Subheading 4.). 7.2. Biology and Ecology as Determining Factors in Practical Use: Historical Perspectives Eptzoottcs are the most obvious mamfestatrons of diseasesand are expressed as abnormally high levels of disease incidence, often resulting m extensive host mortality. Observattons of epizootics have been noted from early history,

555

Dose Acquisition for Bioinsecticides

when reference was made by the Chinese to diseasesm silkworm culture. References to diseases, including descriptions of fungal, bacterial, and viral dlseases m insects, date back to the sixteenth century, even though the causal organisms themselves were not always recognized or described (IO). Fungal diseases, m particular, attracted much attention, leadmg to production of Metarhizwm anisopliae for control of Anisoplia austriaca (11). The most slgnificant observations were those lmked to pathogens that caused obvious symptoms and significant mortaltty of affected hosts. This led to the search for new pathogens, and, more importantly, to a better understanding of ecology and eplzootlology of the causal disease agents. Steinhaus (12) was one of the ploneer scientist who recogmzed this need. Although many diseases were recorded, it IS the discovery of Bacillus thuringzenszs(Bt) m 1915 (13) that brought a major breakthrough m microbial pest control. The Kurstakl strain has since become the benchmark Bt isolate for use against lepidopterous pests, although new Isolates with activity against hosts m other insect orders are being discovered regularly 2. The Target Microbial pathogens are generally unable to search actively for their potential hosts. The encounter between pathogen and host that leads to infection IS, therefore, driven mainly by host determinants, rather than by particular attributes of the pathogens themselves. Fundamental to the process of encounter between pathogen and host are the two key attributes of host susceptibility and host biology. 2.7. Host Susceptibility:

The Basis for Determining

Dosage

Successm crop protection IS determined by the ability to reduce pest populations to a level below an economic threshold. The stringency required to achieve this economic threshold will vary enormously between different crops: extremes range from no cosmetic damage, typically associated with high-value horticultural crops, to a considerable degree of damage that might be tolerated m forest crops. However, m all cases,the process of pest management aims to target the most susceptible stage of the host, relative to its potential to cause damage to the crop With few exceptions, the first-mstar larva tends to be the principal target stage, reflecting its small size and lower capacity to cause damage. For each organism, the relationship between dosage of the pathogen and susceptlbihty of the host must be determined as the basis for calculating field dosage, 2.1.1. Dosage-Mortality

Relationships

Laboratory studies of the relationships between dosage and mortality are the first steps m assessinglikely field dosage requirements. For viruses and bacte-

556

Evans

95% kill (probit 6.64)

50% kill (probit 5.0)

50% kill

.

Log dose Fig 1 The influence of slope value in dosage-mortality regressionson the calculated LD,, and LDg5dosagerequirements.

rta, moculum must be delivered orally and the responses assessedas mortality over time (14). Fungal pathogens of insects requn-e topical application to the cuticle, reflecting the route of entry of the pathogen to the host (15). In all cases,the quanta1 response of interest is mortahty of the host, and tt is essential that the assays are rigidly quantified, in order to describe the data as a regression of proportionate mortaltty (usually converted by probit or other transformation) against log dosage ingested. Ideally the prectse dosage ingested in a given time should be determined, so that the result can be expressed as LD50 (the actual numbers of IUs required to kill 50% of the test population), rather than LC50 (the concentration of IUs that provides only an approximation to the precise dose ingested by each mdivtdual). Mortality levels in excess of 90% are the minimum requirement for fieldpopulation reduction, and, therefore, the slope of the dosage-mortal@ regression line must be determined accurately, to allow the required dose to be derived from the calculated regression. The tmportance of thts reqmrement is illustrated m Fig. 1, which shows, schematically, how pathogens with identical LDSovalues can have widely differing LDa5 requirements. Indeed, the mformatton gained at this stage can have far-reaching tmphcations when field dosages are calculated. In relation to relative susceptibility of different stages of msect development to applied pathogens, the great majority of studies have indicated a stg-

Dose Acquisition for Bioinsecticides

557

nificant increase m dosage requirement with larval age, although this pattern may not be consistent for all pathogen groups (16). For example, LI et al. (17) demonstrated that susceptibility of obliquebanded leafroller, Chorrstoneura rosaceana, to Bt was greatest in the fourth instar and least in the sixth mstar. In this case, the early larval stages were approx one-half as susceptible as the fourth mstar, but, as the authors point out, the application of lethal concentration to the early stages, rather than precise dosage, may have obscured their true susceptibility to the bacterium. This illustrates the importance of knowmg how much food, and, hence, amount of moculum, is Ingested m a given time (see Subheading 2.2.4.). By contrast to the variability in age-related responses of larvae to Bt, larval suscepttbllity of baculovnuses tends to be strongly linked to age, although this 1smore accurately expressed as a response to mcreasmg larval weight. Evans (16) assessedthe published results for a number of studies, and calculated slope values for relatlonships between log,, LDSO/mg and log,, larval weight. For granulosls viruses (GV), slopes ranged from 0.17 for Pieris rapae to 1.77 for P. brassicae (18); for nuclear polyhedrosis viruses (NPV), slopes ranged from 0.08 for Operophtera brumata (19) to 1.02 for Hyphantrza cunea (20). The greater the value of the slope, the greater the dosage required to Infect and kill the later instars of the pest. In many cases,this can mean mcreases of several thousand-fold in IUs, which can have major cost implications for calculating field dosages. 2.1.2. Replication and Production of Secondary /now/urn Fungal and viral pathogens are characterized by massive Increases m IUs following replication within the host. Information on the rate of increase, and on the total quantity of inoculum produced on death of a given host stage, can guide design of dosage rates for field use. This knowledge can be particularly important if basic dosage-mortality responses indicate that LD,, requirements are unlikely to be economically viable. Mortahty arlsmg from the acqulsltion of primary moculum can lead to the release of secondary inoculum that is many orders of magnitude greater than the mltially applied dosage. Although this may be locally dlstrlbuted on the host substrate, it 1slikely that any hosts, mespective of their innate susceptibiltty, will succumb to the massive quantity of inoculum, should they encounter it. Several studies on baculovirus growth have indicated that peak productivity of up to 1.5 x 1Ol” polyhedral mcluslon bodies (PIBs) per individual larva can be achieved, representing around 3 x lo7 PIBs/mg body wt (NPV of Heliothis = Helicoverpa zea [21/). In relation to LDsOvalues of under 10 PIBs for first-instar larvae of most Lepidoptera, this represents a source of inoculum that can have considerable influence on the outcome of a spray operation employing baculoviruses. Similarly, fungal diseaseswill also

558

Evans

yield massive quantities of secondary moculum, with potential increases m the order of 200,000-fold (22) 2.2. Host Biology: The Fundamental Basis for Dose Acquisition Drscussion in this section ~111provide the detailed basis for linking host biology to the precise point of delivery of the required dose. Knowledge of the biology of the host is fundamental to optimizing the use of microbial agents for pest control, but this 1s a relatively neglected component of the design of management programs, and one that has the potential to increase efficacy and reltabtllty, without necessarily requnmg major changes m absolute dosage requirements. 2 2 1. Recruitment and Vultinism Both the rate of recruitment (egg hatch) and voltnnsm (number of generations per year) can have profound influences on the results of dose acquisition. At its simplest level, a pest with a single generation per year ~111have a period of recruitment that, unless egg hatch occurs over a very short time period, can lead to a mix of mdividuals that range widely m susceptibtlity. Overlappmg generations add a further level of complexity that can result in virtually all stages of larval susceptibihty bemg present m the same population Therefore, sound knowledge of the stage structure of the pest populatron is essential. Such knowledge can point to methods of samplmg, such as egg counts and observed hatch rates, that will mdicate the presence of the most susceptible mdivtduals. Although simple m concept and practice, direct assessmentof hostpopulation development in real time does not allow prediction, and, therefore, offers relatively little time to make informed decisions on dosage rates and application parameters. A more sophisticated decision-support system would be based on models that use sample data to predict egg hatch and larval development rates for a particular region. Such an approach has been used for modeling both the population dynamics of target hosts and the effects of pathogens on those populations (23,24). By concentrating on the actual or predicted rates of recruitment to different life stages, and on the requirement to kill the leastsusceptible life stage present, a quantity of moculum for the whole population can be calculated. This 1sdiscussed m more detail m Subheading 2.2.3. 2.2.2. Distribution Patterns Invertebrate populations have a tendency to change dlstrrbution patterns m space, depending on the density of that population The degree of aggregation, and various mathematical mdtces that can be used to describe these changes, have been discussed by a number of authors (e.g , ref. 25). Information on the spatial distribution of a population is an important attribute that 1snot always

Dose Acquisdion for Biomsecticicies

559

considered m assessing dose acquisition m the field. This is especially true if the distribution also changes with the age of the target host, so that, for example, the first-mstar stage, although being the most susceptible to the pathogen, may actually be hidden, or, indeed, not feed at all. Such is the case with spruce budworm, Choristoneuru funziferana, which is one of the most damagmg pests of forestry in Canada and the northern United States. Timing of spray applications and placement of droplets are difficult, because the first-instar larvae do not feed, and either remain wrthin a silken hibernaculum, or, if they are disturbed or crowded, will be dispersed by wind on silken threads (26). Secondinstar larvae appear in April or May, having wintered in hibernacula, and further wind dispersal takes place. The first realistic stage of a spray operation is, therefore, the third mstar, and even this stage IS difficult to target, because of its tendency to feed cryptically on newly expanding needles, which, as they grow, will tend to dilute applied moculum per unit area. Further complication can arise when an insecticide application changes target-host density, thereby changing the distribution of the population. This factor must be taken mto account m developing appropriate sampling regimes to assesspre- and postspray pest populations. 2.2.3. Population Susceptibility Over Time Combinmg knowledge of host susceptibihty at different larval stages with information on the rate of recruitment into each of those instars provides the basis for assessingpopulation susceptibility. The need for this information will depend on the extent of damage reduction required; the mformation will also enable decisions to be made on costsof active ingredient that will be required. Figure 2 ~llustrates schematically the prmciples involved in assessingpopulation susceptibihty. Changes in the susceptibility of different larval stagesto the inoculum are lmked to the instar distrtbution pattern. Clearly, damagewill increasewith increasmg sizeof larvae, and so damage reduction ~111require targeting of theseolder larvae, with a consequent rise in inoculum requirement. In the case of insect baculovirnses, the resultmg increase in moculum costs may rule out optimal targeting of the most damaging stages.By contrast,the relatively flat dosageresponsecurves for Bt tend not to rule out its use against older larval stages. 2.2.4. Host Feeding Rate Over Time Larvae Increase feeding rate as they age. For example, the leaf area consumed by Pieris rapae increases 11S-fold from the first to the fifth mstar (27). Such increases are typical of lepidopterous larvae, and are an accurate reflection of the weight increase observed as larvae age, so that approx 90% of total food may be consumed by the final feedmg instar (28). Increasing consumption of leaf area with age has several consequences for dose acquisition.

560

Evans

Pathogen Activity (% of original)

window of opporhmity for G/ successful

Damage (feeding rate)

Time

______)

Fig. 2. Schematic representation of the three principal components of dose acquisition for application of microbial insecticides during spray operations. 1. A larger area of food consumption by a given target-insect stage provides a greater available surface for deposition of inoculum, thus potentially easing the task of droplet deposition during spray operations. This is a positive attribute that increases the likelihood of encounter between the target host and the droplets delivered to the feeding sites. 2. There is a decrease in susceptibility as larvae age, which requires delivery of much greater quantities of inoculum to feeding sites, despite the fact that increased feeding has a concomitant increase in encounter frequency between host and inoculum.

Dose Acquisition

for Bioinsecticides

561

3 Increasedareaconsumednormally equatesto increaseddamageto the crop being protected.This may beintolerable m high-value crops in which cosmeticdamage can result in seriouslossof value Pestmanagementwill normally aim to reduce damage,which, consequently,reducesperiods of feedmg andthe areaconsumed, thus limltmg the probablhty of encounterbetweenhost and inoculum A balance must, therefore, be struck between increasing the likelihood of acquisition of inoculum by increasing the area of food consumed, and the Increased dosage requirement and potential damage that could result from thts strategy. Other attributes, such as attrition of applied inoculum arising from ultraviolet (UV) light, rainfall, and so on, must also be considered in this process (see Subheading

3.1.).

3. The Microbial Agent Irrespective ofthe type ofmlcrobial agent employed asa bioinsecticlde,the fate of apphed inoculum, and Its ultimate effectiveness against target hosts,depends on a sequenceof critical events. Delivery of inoculum to the target areais the first step m this sequence,followed by the degreeof persistenceof that moculum, which determines duration of a lethal doseat the host feeding site.The final stagemay anse from the impact of secondaryinoculum on target hoststhat survived the imtlal treatment. 3.1. Field Persistence of Primary lnoculum All blomsecticides are subject to attrition once they are applied m the field The principal factor m loss of moculum IS undoubtedly UV light (29-31) Damage occurs when the microbial agents are exposed to UV light in the waveband 290-380 nm, although most absorption of UV occurs at the lower end of this scale (5). Effects can be severe, with over 50% loss of activity after a few hours of exposure of the pathogen to UV light in the field. For example, Bt, applied to vines in Australia for the control of light brown apple moth, Epzphyas postvittana, lost over half its activity within 24 h of application to fully exposed leaf surfaces (31). By contrast, Bt applied to shaded leaves still retained over 60% activity after 2 d exposure to sunlight. In a study of the relative survival of various pathogens when exposed to UV under laboratory conditions, Ignoffo et al. (32) demonstrated that the order of stabthty was Bt > Nomuraea rileyz (fungus) > entomopox virus > NPV = cytoplasmic polyhedrosis virus (CPV) > Vairimorpha necatrix (Protozoa) > granulosis vrrus (GV). However, the wavelength used in this study resulted in very rapid inactivation (half-life of 4 h or less) of all pathogens, but, nevertheless, provided useful guidance on relative rates of inactivatlon of different pathogen groups. Interactions between UV damage and temperature have also been noted, so that higher temperatures tend to increase the speed of mactlvation when pathogens are exposed to a given amount of UV. This was demonstrated m studies

562

Evans

of stability of Metarhiziumflavoviride spores, in which UV exposure resulted in 20% reduction in germination at 20°C compared with 80% reduction at 50°C (33). The net effect of these two factors is to decrease the potency of the applied inoculum, sometimes at extremely rapid rates, thereby reducing the effectiveness of treatment. Awareness of the dynamics of these effects is a central component of dose acquisition, and can point to ways of alleviating the problems. 3.2. The Potential Contribution of Secondary lnoculum The majority of pathogens reproduce in their hosts, and this results in higher inoculum loads than originally delivered to the crop, although this is rarely the case with Bt. Production of significant quantities of secondary inoculum can be important in determining infection levels in hosts that may have escaped the inoculum applied in the original spray operation. The importance of secondary inoculum can be illustrated by reference to a hypothetical lepidopteran host in the family Noctuidae (in which larvae have a weight range from approx 5 mg in the first instar to around 1000 mg in the sixth instar). Infection within each instar is related to larval weight by a constant of around 1 x 1O7PIBs/mg. Thus, production of secondary inoculum will increase from 5 x 1O7PIBs for larvae dying in the first instar to 1 x 1Oi” PIBs for larvae dying in the sixth instar, representing massive multiples of a lethal dosage that may be < 100 PIBs for an early-instar larva. Likewise, an infected fourth-instar gypsy moth, L. dispar, produces around 2 x 1O5conidia on death, demonstrating the potential for secondary inoculum production (22). In terms of potential to infect further hosts, the distribution of these massive increases in inocula must be considered both spatially and temporally. Clearly, the distribution of secondary inoculum will reflect, to a very great extent, the site of death of the infected individual and the spatial structure of the population. Behavioral changes often accompany infection for baculoviruses (3435) and fungi (22,24), and are regarded as evolutionary adaptations to increase the probability of further infection and growth of the pathogen. In most cases,there is a tendency for aggregation of infected individuals, thus concentrating inoculum at specific sites in the host ecosystem. Fuxa (36) showed that populations of soybean looper (Anticarsia gemmatalis) larvae, infected by the fungus Nomuraea rileyi, were more aggregated than uninfected larvae, reflecting larval behavior and the impacts of secondary inoculum. 4. Matching Delivery to the Target The concepts outlined in Subheadings 2. and 3. provide the foundation for determining field dosage on an apriori basis, thus reducing the need for ad hoc trials, in which the optimal dosage rates are arrived at through a process of

Dose Acquisition for Bioinsecticides

563

iteration. The more detailed the information, the higher the confidence in the assumptions that can be made, and the greater the potential reliability of the field dosage rate that is calculated. This subheading deals with ways of bringing separate data together in assessing the spray window for optimal use of microbial insecticides (37). 4.7. Determining Field Dosage Rates Bateman (see Chapter 27) has discussed the choice of spray equipment and spray parameters in relation to microbial bioinsecticides, especially fungi. This information is fundamental to the calculation of concentrations of active ingredients and carrier fluids in tank mixes for the delivery of microbial agents, both to increase precision in application and to increase overall efficacy of the applied inoculum. It is assumed in the following discussion that the parameters outlined by Bateman have been included in assessing spray equipment and droplet generation characteristics. 4.1.1. Dose Acquisition Over Time: Calculation of Tank Mix for Delivery of the Microbial Agent At its simplest, the use of a microbial control agent must satisfy a single basic concept: At the host feeding sites,the distribution of droplets containing the lethal dosage must match the distribution of the susceptible host population. However, within this simplistic approach, the attributes of the susceptibility of the host population, the stability of the applied pathogen, and the host feeding rate (which equates directly to crop damage, and, potentially, economic loss) must be considered together over time. Figure 2 indicates how these parameters change together over time, so that the probability of a given dosage of pathogen inducing mortality below an economic damage threshold decreaseswith increasing time after application. This occurs through decreasing susceptibility, decreasing pathogen activity, and increasing damage over time. In essence,this results in only a limited window of opportunity to match pathogen and host at realistic cost. In most cases,the earlier the application, the greater the likelihood of delivering dosage to host within economic limits, and, thence, to acceptable mortality. Optimization of the parameters in Fig. 2 proceeds through a series of criteria derived from quantitative information on several fundamental attributes of both host and pathogen biologies. The purpose is to match the dosage requirement for a field LD9, to the deposition of droplets, and to the area consumed by the host, so that, where the three overlap, the host will consume at least an LD9, in a predetermined time (usually a low number of hours postspray). This is illustrated conceptually in Fig. 3. It is difficult to determine precisely what the field LDg5 should be for use within a spray operation, but it is possible to use the data from laboratory assays

564

Evans

area consumed

Area consumed

is time limited (usually

<24 hours)

Fig. 3. The relationship among droplet deposition pattern, host feeding area, and placement of the LD,, dosage during application of microbial insecticides.

to provide a realistic measure of dosage for each larval stage (37). As outlined in Subheading 2.2.1., further refinement of dosage must be made in relation to the distribution of susceptible host stagespresent at the time of spray. This will typically consist of a number of different instars, so that the highest effective dose should, ideally, be determined by the LD,, of the least susceptible stages present (Fig. 4). In Fig. 4, this would be represented by the LD,, of the fifth instar, which may be many orders of magnitude greater than that for the first instar. Rate of lossofpathogen over time (seeSubheading 3.1.) must also be included in initial calculation of concentrationswithin the tank mix. The conceptcan be summarized by a simple equation that illustrates the approachand derives a basic relationship between dosagerequired at time of spray and that necessaryto ensuremortality at the end of the defined feeding period. The formula for this relationship is: Di = d/ar

(1)

where Dj = initial dose required (IU/mm2 before loss factors, such as UV, and so on); d, = LD,, (IWarget host stage); a = proportionate loss of activity over time required for host to acquire target number of droplets; and r = host feeding rate in mm2 over time interval, t. In essence,Eq. 1 statesthat the area consumed by the target host in a predetermined time t must contain a LD9s dosage at the end of the feeding period, thus allowing for any attrition of pathogen (at rate a) over that time. Substituting in Eq. 1 for a hypothetical example, let d = 500 IU (LD,, for target host

Dose Acquisition for Bioinsecticides

565

by value

for oldest

ii’;‘ii’/

Pathogen dosage (Log scale) Fig. 4. Schematic representation of the increase in dosage requirements with increasing age of the target-host population. Field LDg5 is, ideally, determined by the response of the least susceptible stage to challenge by the microbial agent.

individual); a = 0.5 (i.e., 50% of activity after time t has elapsed); Y = 4 mm2 over time t; then Q = 250 IU/mm2. Consideration must then be given to how this should be delivered in relation to the numbers of droplets targeted per unit feeding area. In the example above, the question is, should the inoculum be delivered in one droplet containing 250 IU, or should 10 droplets, each containing 25 IU, or some other combination be used? The solution to this problem will be determined by spray equipment, carrier fluid, droplet size,wind speed,leaf area, and so on. It is essentialthat a realistic measure of achievable droplet coverage is made and linked to the distribution of droplet sizesin the spray cloud. Such data will then aid the decision on how many IUs should be present per droplet within the host feeding area. The aim is to determine the number and proportion of emitted droplets captured per unit target area, converted, where appropriate, to number per unit ground area (the latter conversion is not essential,but aids easycomparison between different spray operations). 4.1.2. Calculation of Tank Mix: Worked Example The final calculations of tank mix can now be made using simple relationships between pathogen, host, and droplet parameters: N = Numbers of drop-

566

Evans

lets emitted by atomizer/L: for convenience, use Volume Median Diameter (VMD) (if span 1s small); and CE = Capture Efficiency defined by the number of droplets required to ensure at least one droplet per host feeding area, expressed in terms of droplets per unit ground area, based on Leaf Area Index (LAI), loss to ground, and so on.

Determine feeding rate (r), LDg5 (d), and vtrus acttvtty loss (a), to gave mlteal dose D,, expressed as IU/mm2.

Dose per ha, and final concentration of the tank mix, 1sdetermined by. CE = (1 x 10”) LAZ/lI(s x r.) droplets per ha

(2)

where 1 x lOto = area of 1 ha in mm2; LAZ = Leaf Area Index, a multtpher

to

express surface area of leaves m units of ground area, s = Loss of spray fluid to nontarget area. V = CEIN L/ha

(3)

Dha = CE x D, expressed in III/ha

(4)

D, = N x D, expressed in W/L

(5)

The prmctple of tank mtx calculation can be illustrated by use of hypothettcal data for the above equations. In this case, the imttal dosage, D,, 1s 250 IUs. Using Eq. 2, CE = 4 x lOi droplets/ha (assume r = 4 mm*, LAI = 8, loss to nontarget area = 50%) Let N = 1 53 x 10” droplets/L (assume 50 ym VMD) Let D, = 1000 (assume d = 500, a = 0 5 [50% loss]) Using Eq. 3, V= (4 x lO’O/1.53 x lOlo) = 2.6 L Using Eq. 4, Dha = (4 x lOi x 1000) = 4 x lOi III/ha Using Eq. 5, D, = (1 53 x lOi x 1000) = 1.53 x lOi IU/L

Equation 5 determines the actual tank concentration total volume of spray fluid in the tank, including mulation products.

of IUs, which relates to any volume attributed to for-

4.1.3. Optimization of Tank Mix Parameters Although the btologtcal parameters of host feeding behavior and suscepttbtltty of the pathogen are fixed, the physical parameters of droplet stze and dtstrtbu-

tron can be modtfied to assessthe potential effects on total dosage/ha. This is parttcularly important tf the maximum effective dosage calculated by the above method appears to indicate an unacceptably high cost of active ingredient. 4 1

3 1. COST-BENEFIT ANALYSIS OF CALCULATED DOSE RATE

The key question arising from the calculations carrted out m Subheading 4.1.2. 1swhether the cost of the tank mtx 1s low enough to Justify tts use to reduce damage below the economtc threshold. Here, a balance has to be struck

Dose Acquisition for Boinsecticides

567

between the unit cost of the pathogen (often at a considerably higher cost than a conventional chemical Insecticide) and the dosage necessary to achieve target-host mortality and consequent reduction m damage. Optimtzatton can proceed along a number of pathways that can be used alone, or, more likely, m combination. These are hsted in Table 1. The development of a spray control program for use of a baculovtrus to control pine beauty moth, Panolisflammea, m Scotland, illustrates many of the attributes described above, and also serves to emphasize that a number of opttons can be considered m achieving the desired level of protection (37-40). Data on larval distribution and on dosage requirements for mstars I and II mdtcated the need to deliver virus to the top 30% of the canopy wtth a mmtmum droplet density of 5 droplets/cm length of needle. Imttal calculattons, based on laboratory bioassay results, mdtcated that dosage per feeding area (r) for the first-instar target stage was between 10 and 100 PIBs to mduce >90% mortality (37). The questton then was whether to deliver this in a single droplet or m multiple droplets at the feeding site. Based on experiments with dtfferent VMDs for delivery of droplets, rt was concluded that more effective coverage of the upper 30% of the canopy would be achteved with the smallest droplet sizes m the range of 40 to 50 pm m diameter (38). Thts required the dtstrtbutton of the LD,, dosage across several droplets. Results confirmed that thts strategy gave more effective control than placing the lethal dose m a single droplet. Further refinement was provided followmg assessment of larval behavior that indicated that moculum was encountered at both prtmary feeding sites (the newly expanding needles), and on the older needles, thus mcreasmg the apparent feeding area, r, and reducing the dosage requirement/ha (42). Thts enabled effective control to be achieved with dosages as low as 2 x 10” PIBs/ha, even though mittal calculattons had indicated a dosage requirement of up to 1 x lOI* PIBs. 4 1.3 2. FORMULATION TO REDUCE COSTS AND INCREASE EFFICACY Two key requirements of formulation are to protect the pathogen from macttvatton on the host substrate, and to enhance physical retention through the use of stickers and spreaders (42). However, the addittonal benefits of formulation must be carefully assessed against the costs of the process and the potenttal changes to droplet generation that may result from mapproprtate formulation. For baculovn-uses, recent advances have been made m UV protection through the use of stilbene optical brighteners, which also appear to increase virulence (by up to 2 14-fold), even in the absence of UV (43,441 A switch from aqueous to oil-based carrier fluid can also increase performance from a given quanttty of inoculum. An example of this approach IS descrtbed by Bateman (see Chapter 27) for the LUBILOSA project.

Droplet losses to nontarget areas may increase required dosage/ha

CE (capture efficiency of emitted droplets)

for Microbial

Half-life of activity may be only hours, mcreasmg mttial dose requirement consrderably

Constraint

Dose Acquisition

a (rate of attrition of pathogen m field)

Potentialfor

The calculated value may be very high, leading to unreahstrc costs

and Optimization

d V’95)

Parameter

Table 1 Constraints potential

1 Alter droplet spectrum and/or volume delivery rates to improve targeting. 2 Formulate to avotd evaporation 3 Use natural wind to improve selectivity of targeting.

1 If the dosage-mot-t&y slope 1s relatively low, a considerable reduction m dosage can be tolerated before mortality declmes to suboptimal levels 2 Target most susceptible stages only, acceptmg some increased damage to the crop 3 Assess potential of secondary moculum that may contribute to overall mortality. 1. Formulate to protect against UV losses, or to increase adhesion to target surface. 2 Time spray to avoid main UV daily peaks (espectally In troprcs).

Optimization

Insecticides

Dose Acquisition

for Bioinsecticides

569

4.2. Transgenic Plants: A Special Case of Dose Acquisition Dose acquisttlon may also be achieved directly at the feeding site by incorporation of genes, usually Bt, within the target plant Itself (45-47). This allows more directed dose acquisition precisely at the feeding site, thus reducing the need for spray application. However, the continuous expression of the toxin gene may increase the risks of resistance developing, and, consequently, strategies for resistance management are being developed to address this possibihty (#8,49). 5. Future Requirements and Conclusions Although there have been significant advances in spray technology, and m the understanding of precise targetmg of the host feeding site, there has been a distinct reluctance to embrace this technology for control of pests m broad acre crops (50). Although it must be accepted that existing technology has a sound track record for the purpose for which it was designed, namely apphcatlon of chemical pesticides, the attributes of microbial insecticides point to the need for further appraisal of dose acquisition for these agents. This chapter has considered the interface between technology and pathogen-host interaction, concentrating on quantlficatlon of a series of key parameters that offer the potential for enhancing the efficacy of mlcroblal insecticides. A structured approach indicates areas in which improvements in application technology can be matched to the biology of both host and pathogen. Other factors, such as the need for a full lethal dose per feeding area when applying Bt, must also be considered (51). Van Frankenhuyzen et al. (51) showed that a LDg5 dosage must be ingested m one or two droplets, otherwise the larva stopped feeding and eventually recovered from the sublethal dose, effectively escaping the applied inoculum. Such considerattons can only be discovered by detailed knowledge of the parameters discussed in this chapter, and confirm the value of such information in design of management regimes employing microbial insectlcldes. Acknowledgments My grateful thanks go to Dr. Richard Jinks (Forest Research) for his constructive comments and suggestions for improvements to layout and readability. Thanks also to an anonymous reviewer for further helpful suggestions. References 1. Evans, H. F., ed. (1997) Microbial insecticides.novelty or necessity?BCPC Symposium Proceeduzgs, No. 68. British Crop ProtectionCounctl, Farnham, UK 2. Evans, H F. (1997) The role of microbial Insecticides in forest pest management, in Microbial Insectrcldes Novelty or Necessity? (Evans, H F , ed ), BCPC

570

3. 4. 5 6 7. 8 9

10 11 12. 13

Evans Symposmm Proceedings No 68, Brltlsh Crop Protection Council, Farnham, UK, pp. 29-40 HaJek, A. E. and St Leger, R J. (1994) Interactions between fungal pathogens and insect hosts Ann Rev. Entomol 39,293-322. Burges, H D , ed. (1981) Mtcroblai Control of Pests and Plant Diseases 197& 1980, Academic, London. Entwlstle, P F and Evans, H. F. (1985) Viral control, m Comprehenszve Znsect Physzology, Biochemistry, and Pharmacology (Kerkut, I. and Gilbert, L I , eds ), Pergamon, Oxford, pp 347-4 12 Granados, R R. and Fedenci, B A., eds (1986) The Bzology ofBaculovzruses,vol 1, Blologlcal Properties and Molecular Bzology, CRC, Boca Raton, FL, 275 pp Granados, R. R. and Fedencl, B. A , eds (1986) The B1oIog-y of Baculovzruses vol 2, Practical Application for Insect Control, CRC, Boca Raton, FL, 276 pp. Fuxa, J R and Tanada, Y , eds (1987) Epzzootzology of Insect Dzseases, Wiley, New York Lacey, L. A , ed (1997) Manual of Techniques zn Insect Pathology, Biologzcal Technzques Serzes, Academic, San Diego Stemhaus, E. A. (1975) Dzsease zn a Manor Chord, Ohio State Umverslty Press, Columbus, OH Stemhaus, E. A. (1956) Microbial control. The emergence of an idea A brief history of insect pathology through the nineteenth century Hilgardza 26, 107-I 60 Steinhaus, E A. (1946) Insect Mzcrobiology, Hafner, New York. Berliner, E ( 19 15) Uber die Schlaffsucht der Mehlmottenraupe. Zeltschrzft fur angewandte

Entomologie

2,29-56.

14 Hughes, P. R and Wood, H. A (1987) In vlvo and m vztro bioassay methods for baculovnuses, in The Bzologv of Baculovuxses, vol 2 Practical Appllcatlon for Insect Control (Granados, R R and Federici, B A, eds ), CRC, Boca Raton, FL, pp. l-30. 15 Vandenberg, J D (1996) Standardized bioassay and screening of Beauverza basszana and Paeczlomyces fumosoroseus agamst the russian wheat aphid (Homoptera. Aphldidae). J Econ. Entomol. 89, 1418-1423 16 Evans, H F. (1986) Ecology and epizootiology of baculoviruses, m Bzology of Baculovlruses, vol 2 Practccal Applrcatlon for Insect Pest Control (Granados, R R and Federicl, B. A., eds ), CRC, Boca Raton, FL, pp 89-132 17 Ll, S. Y , Fitzpatrick, S. M., and Isman, M. B. (1995) Susceptlblllty of different mstars of the obliquebanded leafroller (Lepidoptera. Tortncrdae) to Bacillus thuringzenszs var kurstakl. J Econ En tomol 88, 6 1O-6 14. 18. Payne, C. C , Tatchell, G. M , and Williams, C F (198 1) The comparative susceptibilities of Plerrs brasszcae and P rapae to a granulosis vu-us from P brasszcae. J lnvertebr

PathoI 38,273-280

19 Wlgley, P. J (1976) The eplzootlology of a nuclear polyhedrosls virus disease of the winter moth, Operophtera brumata L at Wlstman’s Wood, Dartmoor Unpublished D Phil. Thesis, Oxford University 20 Bouclas, D G. and Nordm, G L. (1977) Intermstar susceptibility of the fall webworm, Hyphantrla cunea, to Its nucleopolyhedrosls and granulosis vnuses J Invertebr Path01 30,68-75

Dose Acquisition

for Blomsecticides

571

21 Ignoffo, C. M and Hmk, W. F. (1971) Propagation of arthropod pathogens m living systems, m Mtcrobial Control of Insects and Mttes (Burges, H. D. and Hussey, N. W., eds.), Academic, New York, pp. 541-580 22. Hajek, A. E., Larkin, T S., Carruthers, R. I., and Soper, R. S. (1993) Modeling the dynamICS of Entomophaga matmatga (Zygomycetes, Entomophthomles) epizootics in gypsy moth (Lepidoptera, Lymantriulae) populations. Environ Entomol. 22,1172-l 187 23 Elkmgton, J S , Dwyer, G., and Sharov, A. (1995) Modellmg the eptzootiology of gypsy moth nuclear polyhedrosis vwus. Comput Electron Agrtcult 13,91-102. 24. Carruthers, R I and Soper, R. S (1987) Fungal diseases, in Eptzootiology oflnsect Dzseases (Fuxa, J. R and Tanada, Y., eds.), Wiley, New York, pp. 357-416. 25. Taylor, L. R. (1984) Assessing and interpretmg the spatial distributions of Insect populations. Ann. Rev. Entomol 29,321-357 26. Mattson, W. J , Simmons, G A., and Wetter, J. A. (1988) The spruce budworm m eastern North America, m Dynamtcs of Forest Insect Populattons Patterns, Causes, Impltcattons (Berryman, A. A., ed.), Plenum, New York, pp 309-330 27 Tatchell, G M. (198 1) The effects of a granulosis vnus infection and temperature on the food consumption of Pteris rapae (Lep.:Pieridae). Entomophaga 26,291-299 28. Harper, J. D. (1973) Food consumption by cabbage loopers infected with nuclear polyhedrosts virus. J Znvertebr. Pathol. 21, 191-197. 29. Moore, D., Bridge, P D., Higgins, P M., Bateman, R P., and Prior, C. (1993) Ultra-violet radiation damage to Metarhzziumflavovzride conidia and the protection given by vegetable and mineral oils and chemical sunscreens Ann Appl Btol 122,605-6 16 30. Ignoffo, C. M. and Garcia, C (1994) Antioxidant and oxidative enzyme effects on the inactivation of mclusion bodies of the Heltothts baculovnus by simulated sunlight-UV. Environ, Entomol 23, 1025-1029. 3 1. Bailey, P., Baker, G., and Caon, G. (1996) Field efficacy and persistence of Bacrllus thuringtensrs var kurstaki against Epiphyaspostvtttana (walker) (Lepidoptera: Tortricidae) m relation to larval behaviour on grapevine leaves Aust J Entomol 35,297-302 32. Ignoffo, C. M., Hostetter, D. L., Stkorowslu, P. P., Sutter, G., and Brooks, W. M (1977) Inacttvation of representative species of entomopathogemc viruses, a bacterium, fungus, and protozoan by an ultraviolet light source. Environ Entomol. 6,411415. 33. Moore, D., Higgins, P M., and Lomer, C. J. (1996) Effects of simulated and natural sunlight on the germination of conidia of Metarhtztum jlavovtrtde Gams and Rozsypal and interactions with temperature. Btocontrol Set. Technol. 6, 63-76. 34. Evans, H. F and Allaway, G. P (1983) Dynamics of baculovn-us growth and dispersal in Mamestra brasszcae L. (Lepidoptera: Noctuidae) larval populations introduced into small cabbage plots Appl Envtron Mtcrobtol 45,493-50 1 35 Goulson, D. (1997) Wipfelkrankheit. modtfication of host behaviour during baculoviral infection Oecologta 109,219-228. 36. Fuxa, J. R. (1984) Dispersion and spread of the entomopathogemc fungus Nomuraea rtleyt (Momliales: Moniliaceae) in a soybean field. Environ. Entomol. 13,252-258.

572

Evans

37. Evans, H. F (1994) Laboratory and field results with vuuses for the control of insects In BCPCMonograph No 59 Comparing Glasshouse and Field Pesticide Performance II (Hewitt, H. G., Caseley, J., Copping, L. G , Grayson, B. T., and Tyson, D., eds.), BCPC, Farnham, UK, pp. 285-296 38. Entwistle, P F., Evans, H F , Cory, J, S., and Doyle, C J. (1990) Questtons on the aerial application of microbial pesticides to forests. Proceedings of Vth Internatzonal Colloquium on Invertebrate Pathology, (Pinnock, D E , ed.), Adelade, Australia, pp. 159-163. 39. Evans, H. F , Stoakley, J. T., Leather, S. R., and Watt, A. D. (1991) Development of an integrated approach to control of pine beauty moth in Scotland. Forest Ecology Manage 39,19-28.

40. Evans, H F (1994) The control wmdow: a conceptual approach to usmg baculovlruses for forest pest control, in Proceedwgs VI International Colloquium on Invertebrate Pathology andMtcrobza1 Control (Bergoin, M., ed ), Montpellier, France, pp. 380-384. 4 1. Gory, J S. and Entwistle, P. F. (1990) The effect of time of spray application on mfectlon of the pine beauty moth, Panolis jlammea (Den. and Schiff.) (Lep , Noctuidae), with nuclear polyhedrosis wus. J. Appl. Entomol 110, 235-241 42 Jones, IS. A., Cherry, A. J., Grzywacz, D., and Burges, H. D. (1997) Formulation: Is it an excuse for poor application? m Mwroblal Insecticrdes: Novelty or Necesszty7 (Evans, H. F., ed.), British Crop Protection Council, Famham, UK, pp. 173-l 80 43 Dougherty, E. M., Guthne, K. P , and Shapiro, M. (1996) Optical brighteners provide baculovnus actlvlty enhancement and uv radiation protection. Biol Control 7,7 l-74 44. Evans, H. F. and Shapiro, M. (1997) Viruses, in Manual of Techniques zn Insect Pathology (Lacey, L A., ed.), Academic, London, pp. 17-53, 45. Kozlel, M. G., Carozzl, N B , Desal, N., Warren, G. W , Dawson, J., Dunder, E , Laums, K , and Evola, S. V (1996) Transgemc maize for the control of European corn borer and other maize insect pests. Engineering plants for commercial products and applications Ann NY Acad Scl 792, 164-l 7 1 46 Sims, S R., Pershing, J. C., and Reich, B. J. (1996) Field evaluation of transgemc corn contammg a Bacdlus thuringlensls berlmer msectlcldal protein gene against Hellcoverpa zea (Lepldoptera Noctuldae). J Entomol Scz 31,340-346. 47 Stewart, C. N , Adang, M. J , All, J. N., Raymer, P L., Ramachandran, S , and Parrott, W. A. (1996) Insect control and dosage effects m transgenic canola containing a synthetrc Bacdlus thurzngrenszs CrylAC gene Plant Physlol. 112, 115-120 48. Roush, R T (1994) Managing pests and then resistance to Baczllus thurzngienszs Can transgemc crops be better than sprays? Biocontrol Scl Technol 4,50 l-5 16. 49. Snow, A A. and Palma, P. M. (1997) Commerclalizatlon of transgenic plants: potential ecological risks. BzoScrence 47, 86-96. 50 Chapple, A C and Bateman, R. P. (1997) Application systems for microbial pesticides* necessity not novelty, m Mxroblal Insecticides Novelty or Necesszty7 (Evans, H. F., ed ), British Crop Protection Council, Farnham, UK, pp 18 l-l 90

Dose Acquisition

for Bioinsecticides

573

51. Van Frankenhuyzen, K and Payne, N J. (1993) Theoretical optlmlzatlon of Baczllus thurzngzenszs Berliner for control of the eastern spruce budworm, Chorzstoneura fumijkrana Clem--(Lepidoptera: Tortricldae)-Estimates of lethal and sublethal dose requirements, product potency, and effective droplet sizes. Can Entomol

125,473-G%

Strategies for Resistance Management Richard T. Roush

1. Introduction: Potential for Resistance Resistance to pesticides has evolved in more than 500 species of insect pests and more than 70 species of weeds (1,2). It was once believed by some that resistance would be unlikely for blopestmides because they were natural, already exposed to eons of evolution, and of short persistence However, because of the pervasiveness of resistance to other pesticides, few entomologists have ever agreed with that view. In the case of Bacillus thurwzgzenszs (Bt), the most commercially important biopesticrde currently, the myth of mvincrbility was challenged in 1985 with the relatively easy selectron in the laboratory of resistance m the Indian meal moth (Plodla interpunctella, a pest of stored gram) (3) and was truly demolished startmg in 1990 by the appearance of resistance m the diamondback moth (Plutella xylostella, a pest of cabbage and other cole crops) in Hawaii, Asia, the continental United States,and Central America from the use of Bt sprays +8j. Subsequent laboratory experiments have selected resistance in several other species (7). Resistance has also been selected to Baczllus subtilu, a blocontrol agent of a fungal disease of plants (9). 1.1. Specificity of Mode of Action Will other biopestrcides be any less likely to suffer from resistance? Key mdicators of the potential for resistance are specificity of mode of action and past history of resistance m the targeted pests. Prior to the introduction of modern synthetic insecticides, reststancewas a rare event, wrth no more than about a dozen cases prior to 1945. Resistance emerged as a common problem for insects and mites only after the introduction of DDT and subsequent msectitides that acted on particular sites in the nervous system (20). Srmllarly, resisFrom

Methods F R Hall

In Rotechnology, and

J J Msnn,

vol 5 Elopesbodes eds

0 Humana

575

Press,

Use and De//very Totowa,

NJ

576

Roush

tance to herbicides was uncommon until the mtroduction of triazmes (attacking photosystem II), and first emerged as a problem for fungicides with benomyl (disrupts beta-tubulm). In all of these cases,reststance could be conferred by single gene changes that either altered the target site or increased the degradation and excretion of the pesticide (II). Old morgamc pestrcides, such as lead arsenate and Bordeaux mixture (and even some more modern protectant fungtctdes), apparently have multiple modes of action and are difficult to overcome with single gene changes. In effect, there was little usable genetic variation for resistance to inorganic pesttctdes, but relatively abundant genetic variation for resistance to synthetic pesticides. Although increased specificity was also probably generally associated with increased efficacy and therefore increased selection pressure (ZO), specificity of mode action 1sclearly a major predictor of resistance risk. In the case of Bt toxms, at least one route to resistance appears to be a single major gene that confers resistance to a limited group of Bt toxins (Cry 1A and Cry 1F) through reduced binding at a target site m the insect mid-gut (7,12-16) Thus, we might expect that any biopesticide with a specific mode of action would be at risk. At least some biofungtcides and mycoherbicides probably achieve selectivtty by attacking particular physiological target sites. Resistance has already been found to insect viruses (17). Of particular note is the apparent resistance to Heliothis nuclear polyhedrosis vu-us in Heliothzs subjlexa, an insect so closely related to the tobacco budworm (Ffelzothis virescens, a major pest of cotton) that it was possible to use crosses with the budworm to demonstrate that resistance seemed to be under control of a single major gene (28). Depending on the stage at which resistance occurs (e.g., preventing replication), resistance may render irrelevant any other genetic mampulattons to the vn-us (such as the mcorporatton of venoms that attack the Insect nervous system); any manipulation that improves potency might nomtally even increase the risk for resistance by increasing efficacy and therefore use and selection pressure. It is even possrble that insectscould evolve resistance to the venom toxins themselves by a mutation at the target site m the insect. 7.2. History of Resistance in Targeted Pests In addition to genetic variatton, selection pressure 1srequired for resistance to evolve. Selectton pressure is a function of the mtensity (especially frequency) of pesticide use, the proportion of the population that is treated each generation, and other features of pest biology. Absolute numbers of pests exposed each generation may also play a major role at least at the local level because this may increase the likelihood that resistance genes can be present at any given trme, as appears to be the case for herbicide resistance m annual ryegrass (2). Perhaps the most readily available mdicator of the potential for

Strategies for Resistance Management

577

selectlon pressure m any given species and envlronmental circumstance is its past hlstory of resistance. For example, major insect and mite pests of glasshouses routmely evolve resistance very rapidly. Because of the high value of the crops, they are sprayed frequently m a largely futile attempt to maintain zero Infestation levels. Further, the pest populations are often isolated, with little dilution of resistance from unselected mdividuals. Even where not strictly contained within glasshouses, the pests are usually concentrated on suitable host plants in and around the glasshouse complex (19,20). In many comparisons of similar or closely related pests(even with the same speciesin different habitats), resistanceevolves quickly where a high proportion of the population is exposed each generation (21,22), i.e., when there are only small refuges of untreated mdlviduals. Beyond the obvious prediction that species and circumstances that have suffered resistance clearly have the requisite ecological characteristics for reslstance and are likely to evolve resistance again, such examples also help to predict problems in novel circumstances. For example, the cattle horn fly, Huematobia irritans, had not suffered serious resistance problems until the introduction of pesticide impregnated ear tags, a slow release device (23). Because the ear tags controlled flies so effectively, they prevented the dilution of resistance on tagged animals even where there were refuges of flies on nontagged cattle (24). Thus, it should not have been a surprise that resistance evolved. This example also illustrates the importance of persistent formulations on the evolution of resistance, a feature that will be discussed m more detail in Subheading 2.2. 7.3. Registration Requirements for Resistance Management Governmental regulatory agencies have shown increasing mterest m reslstance and resistance management as a consideration m the registration of new products. Within the European Union, a very non-bureaucratic proposal for resistance risk assessmentis being developed within the European and Medlterranean Plant Protection Organization for possible adoption as early as 1998. The evaluation system relies on expert opmion for answers to a maximum of ten questions that focus on selection pressure (25). Resistance management IS also a major regulatory issue for Bt-transgenic crops, with at least the US EPA and Australian National Registration Authority placing restrictions for resistance management purposes on the sale and use of transgemc crops. The initial frequency of resistance alleles prior to the first use of a pestlclde IS rarely known, but population genetics theory and some experimental work suggest that the frequencies range from 1Oe3to 1Oe9depending on resistance mechanism (2,10,26,27). Resistance will, of course, evolve more quickly if the frequency of resistance alleles is higher rather than lower, so this Includes

578

Roush

one element of resistance risk. Although it would be impracttcal to try to estimate the frequencies of resistance alleles that are extremely rare, it is feasible to test whether reststance alleles are present at hrgh frequencies, as has been attempted for Bt (27) and fungicides (28). In sum, resistance management should be a major concern for most tf not all biopesticides, not only m terms of tradmonal commercial consideratrons, but also potentially as a regulatory issue,espectally whenever the pestictde can be considered important to the public good. Resistancerisk assessmentwill ltkely remam an Inexact science,but still useful and inexpensive compared to pesticide resistance. 2. Tactics for Resistance Management Resistance management plans have been characterized as proactive and reacttve. Proactive plans anticipate the potential for resistance and attempt to delay it before resistance IS ever detected. Reactive plans respond to a resistance crisis that has already occurred in the field, often with such tactics as pestictde mixtures and higher rates that desperately aim only to control the resistant pests. Because the frequency of resistance is already high, these tactics have httle likelihood of actually slowmg the evolutton of resistance, as will be outlined later in this chapter. The focus of this chapter will be on proactive resistance management. It IS commonly argued that resistance managers rarely recommend more than reducmg the use of a particular pesticide, that is, to simply apply good integrated pest management (IPM). Reducing the overall number and area of applications through good IPM is critically important, but there are also other key resistance management tactics that are not obvious from general prmctples of IPM. For example, as discussed m Subheading 2.7., rotational use of pestrtides across generations 1ssuperior to mosaics or shorter term rotations of pesticide use. This is true even rf the same amount of each of the pesticides 1s applied. Because many pests (such as the cotton bollworm Hellcoverpa zea) affect several crops (e.g., cotton, corn, tomatoes, soybeans), Integrated management of pests and then resistance often requires a unified cropping system approach rather than focusing on specific crops. Given the diversity of potential biopesticides, rt would be impossible to describe m any detail a complete resistance management strategy for all products and circumstances, but I will attempt to describe as specifically as possrble which tactics seem most appropriate for particular kinds of pesticides. In particular, the tactics that are most appropriate for transgenic crops are often very different than those for sprays The tactics to be considered here include: refuges (reduced numbers of apphcatrons and areas treated, taking advantage of alternate controls), low persistence formulations, high doses, low doses, targetmg most sensitrve life stages,improved

Strategies

for Resistance Management

579

spray coverage, pesticide rotations, pesticide mixtures, and selective vs broad spectrum products. 2.1. Refuges: Reduce Number and Area of Applications As noted above, one of the most obvious and consistentinfluences on the rate at which evolution evolves IS the proportion of the population that escapespesticide exposure each generation m “refuges.” The refuge may include, for example, noncrop plants or crop plants that are not treated. Refuges may also include plant parts in which insectsare protected from exposure (say a codlmg moth larva m an apple), but it is critically tmportant to realize that not all susceptible insects that survive treatment are necessarilyin a refuge. In many cases,the insectsthat survive failed to receive a lethal exposure rather than were totally unexposed. Peststhat most often evolve resistanceare generally concentrated on high value crops or, in the caseof medical or veterinary pests,are closely associatedwith humans or hvestock. Refuges have been legally mandated as part of the resistancemanagement strategy for Bt-transgenic crops m the United Statesand in Australia (29). One can gain some quantttattve insight into the potential impacts of refuges through the use of simulation models. Many readers wtll be correctly skeptical about the predictive power of models, but the ones used here are sample,make few assumptions, and aim to highlight which influences are most likely to be most important for resistance.The models just quantify In more detail some basic arithmetic. For example, if 90% of the population is exposed to selection that kills all susceptible (SS) individuals but none of the heterozygotes (RS), we would expect that resistance would increase by 1O-fold each generation. If there were 10 resistant heterozygotes in a population of 10,000 eggs, 100% mortality of the 90% of the larvae exposed to selection would leave only 1000 larvae, where 10 were still resistant heterozygotes, and a lo-fold increase m the frequency of the reststance allele. Two more generations of such selection would push resistance to a frequency approaching 50% (each heterozygote has one R and one S allele) with control failures imminent. When the simulation calculates this more precisely (accounting for the resistant homozygotes), the actual value is a frequency for the resistance allele of 34% after three generations of selection, and 56% after four generations; four is the value actually graphed m Fig. 1 (10% refuge at 0% heterozygous mortality). Even where the initial resistance frequency is fairly high (1Oe3),a large refuge causes a srgmficant delay of resistance (Fig. 1). In the field, control failures typically occur within a few generations (before or after) of when the resistance allele frequency exceeds 50%, depending on population growth rates and what constitutes acceptable levels of control m the field. All other things being equal, resistance evolves slower for models that assume there is more than one gene involved in resistance.

580

Roush

I

0.00

.

I

0.20 Mortality

I

I

0.40 of

*

.

.

t

I

9

.

1

.

.

0.60 0.80 Heterozygotes

*

1 .oo

Fig 1 Effect of mortality of RS heterozygotes and refuges on the evolution of resistance, as measured when the frequency of the resistance allele [R] exceeds 50% Results of a simulation model assuming a single locus, random mating, no selective mortality of reslstant homozygous larvae, that some fraction of the population escapes exposure (refuges of 5, 10,20, or SO%), and mltial resistance allele frequency @) of 10e3. Data pomts Include 0,50, 75,90,95,96,97,98,99,99.5, and 100% mortality For most pests that regularly evolve resistance, the only way to preserve refuges 1s usually to reduce the frequency or dlstnbutlon of sprays. These pests are generally already heavily concentrated on valuable crops or are otherwise closely associated with humans. In other words, for pests that are at high risk for resistance, we can only afford to use pesticides when they are absolutely required to control the pests, I.e., when the pests exceed economic (or action) threshold densities It 1sarguable that the most important contribution to resistance management has been the development of improved thresholds, samplmg schemes (including the use of “presence/absence”

techniqueslike insect pheromone traps), and predictive models. There are many examples control pests without treating refuge. For example, early (Leptinotarsa decemlzneata)

in which “spot treatment” is often sufficient to the entire population and thereby increasing the season infestations of Colorado potato beetle and spider mites (Tetranychus species) are often

concentrated along the edges of fields or orchards. Weeds are often patchy within

fields and can be controlled

by post-emergent

herbicides.

Spot treat-

ments only where the pest densities have exceeded the threshold means that the rest of the crop contributes to the refuge (e.g., 30). In such cases, spraying the entire field or orchard will not generate economic returns in the short term and wtll only worsen resistance.

Strategies for Resistance Management

581

For at least most insects, so many factors can control pest populations that allowmg some to survive this generation will not necessarily mean that their offspring will be damaging in the near future. For weeds and fungal diseases, there is perhaps an even greater need to integrate biopesticides with other control tactics, such competltlve or resistant cultlvars, crop rotation, and mechamcal controls. Such tactics have resulted in a significant delay of resistance in some of the most recalcitrant insect pests, such as the use of crop rotation for the Colorado potato beetle (31). 2.2, Low Persistence

Formulations

The persistence of a pesticide 1s a two-edged sword. Users want enough persistence to control the pest, but excessive persistence can continue to select for resistance long after the pest has been suppressed below damaging numbers. A persistent pesticide can have the effect of regular prophylactic treatments: rapid resistance.Thus, the advice to avoid persistentpesticidesor formulations has been among the oldest and most widely accepted in resistancemanagement (32), and is well documented by experiment (e.g., 33) and theory (34). Most blopestlcldes probably will have short persistence. Perhaps the most notable exceptions are insect-resistant transgemc crops, including those using Bt but also those using other toxms (e.g., 35). In this case, persistence is required as a matter of necessity m some crops; without expression m the plants, the toxins are simply not sufficiently effective to be economically competitive m cotton or potato crops. Further, for some of the targeted pests, the expression of these plants 1smaintained at such a high level (more than 10 times the dose needed to kill all susceptible insects) throughout the growmg season (36) that they may actually manage resistance more effectively than if the same toxms were intensively used as sprays (37), as discussed in the next section. In contrast to transgemc crops, most pesticides cannot be consistently and economltally maintained at such high residue levels. Nonetheless, because of the persistence of exposure, resistance management for transgenic plants requires the use of neighboring refuges of nontransgemc host plants (36-39), as has been mandated for Bt-cotton (291, High levels of persistence of biopesticides are probably not desirable outside transgemc crops, and even then, transgenic technology 1s not always appropriate. Especially where pests exceed damaging densities only occaslonally, such as for soybeans m the United States, it is probably be much more sensible to use Bt sprays (where they are effective) or other technologies (e.g., baculoviruses) for control. Although it is now techmcally feasible to express Bt toxins in such organisms as algae for control of mosquitoes and other medically important pests, it must be questionable whether such delivery systems will be counterproductive by accelerating resistance.

582

Roush

2.3. High Doses Contrary to popular myth, there 1s no general advantage to applymg high doses of pesticides, and indeed high doses may even inhibit the biological control component of integrated management programs for insects and mrtes. Experiments with fungicides have consistently failed to show an advantage to high application rates and have often showed that lower rates selected for resistance more slowly (28). For msecticides, acaricides, and herbicides, there are fewer experiments, but neither theory or experiments support the necessary use of high rates in the field (3#,40,41). To explore this more fully, I will focus on the general case for msecttcides, where the arguments are very similar to those for other pesticides. When selected m the field, resistance to insecticides is typically caused by a smgle major gene (10,26), whrch provides a convenient but not essential assumption for the discussion that follows. Whether resistance is generally simply mherrted for herbicides and fungicides IS less clear, but it is well establtshed that single major genes can also confer high levels of resistance to both of these groups (2,11,28). For any gene locus with one resistance allele, there would be three genotypes: SS susceptible homozygotes, RS heterozygotes (which may be either resistant or susceptible depending on the intrmstc charactertstics of the mechanism and the dose applied), and RR resistant homozygotes Resistance is often described as dominant (heterozygous individuals show senstttvtty that is most like that of the resistant homozygotes) or recessive (heterozygotes tending to be susceptible), The expected frequencies of the various genotypes is a simple binomial probabmty function (first developed independently by Hardy and Wemberg). wherep represents the frequency of the resistance allele, and q the frequency of the susceptible allele, the frequencies of RR, RS, and SS are. p*, 2pq, and q2, respecttvely. While resistance is still rare (as when a pesticide is first introduced), the most common carriers of a resistance allele should be the heterozygotes. For example, if the frequency of resistance is 10m3(which IS only modestly rare), the frequency of heterozygotes will be approx 2 x 10-3, whereas the frequency of resistant homozygotes ~111be I 04, about 2000-fold less common. The high-dose strategy 1sbased fundamentally on the twm assumptions that essentially all heterozygotes will be killed at the doses of pesticide used and resistance alleles are stall so uncommon that resistant homozygotes will be greatly outnumbered by and will mate only with susceptible homozygotes immigratmg from refuges (3437-40). As a practical matter, these assumptions can be rarely met for pesticides that must be sprayed or drenched on a crop or pest breeding site. To achieve a significant delay of resistance, the mortality of the heterozygotes must exceed 90% unless refuge sizes are very large, i.e.,

583

Strategtes for Resistance Management $50 10

A 40 E s 3o z = 20 ii .E IO ti

Resistance

Recessive

Resistance Dominant , I 1 0.85 0.90 0.95 Mortality of SS Homozygotes

1.0 0

Fig 2 Effects of low to hrgh doses on selectton for reststance in terms of mortaltties of heterozygotes. The imttal resistance gene frequencres were lOA and 20% of the populatron was assumed to escape exposure each generation. “Reststance Recessive” assumes that resistance 1svery similar in expression and inherttance to Bt resistance m the diamondback moth, that IS, the heterozygotes are almost as sensitive as the susceptible homozygotes (16). “Resistance Dominant” assumes that heterozygotes are never killed by the toxtn within the ranges of doses used. When resrstance IS semrdominant (“Semldom”), mortality of heterozygotes reaches 70% at doses that kill 100% of the susceptible homozygotes

>20% (37-39, see Fig. 1) Even doses that kill 100% of a susceptible population will confer no benefit if resistance ts somewhat dominant, that is, the heterozygotes largely survive (Fig. 2). Given that many tf not most reststance mechanisms confer at least lo-fold resistance to heterozygotes, doses needed to achieve sufficiently high mortality would have to be at least 10 times higher than are needed to control the initially susceptible pest populatron, whtch ts generally unacceptable environmentally or economtcally. Two other factors hmtt potential for consistently high mortality of heterozygotes: Most apphcatton methods fat1 to provide uniform coverage of the crop (thereby allowing some mmimally treated heterozygotes to survtve) and sprays generally expose a range of lifestages, some of which will be less susceptible to the toxrcant (e.g., 42). The high dose strategy also falls to delay resistance significantly once the resistance allele frequency exceeds 10e2 (37) unless the refuge stze 1s very large (20% or greater). Thus, the use of higher doses 1sespecially mappropriate once resistance has been found m the field (40), which 1sgenerally difficult to do before the resistanceallele frequency exceeds1% (as discussedm Subhead-

584

Roush

ing 3.). The high dose strategy also assumesthat the nnmtgration of susceptible migrants from refuges and then matmg with resistant survtvors will not be affected by the pesticide, an assumption that is usually not met by chemical sprays (34,37,40), The level of expression of Bt toxins found in transgemc cultivars 1s often high enough to constitute a high dose for some pests when they feed as recently hatched larvae (27,37,43), but this may be one of the few places where the high dose strategy can be successfully applied. 2.4. Low Doses As noted above, experiments with fungicides show that at least occasionally, lower concentrations of pesticides select for resistance more slowly than full label rates. Fewer experiments have been run for herbicides or msectictdes, but simulation models suggest that low doses that allow up to 20% of susceptible Individuals to survive will at least slightly slow the evolution of resistance (Fig. 2). Of course, the resulting poorer short-term control may also cause greater pest damage, and would probably be unacceptable m the absence of alternative control tactics. 2.5. Targeting Most Sensitive Life Stages In many cases,even resistant pests can be ktlled when they are most sensitive, generally when they are young (e.g., 42) Thus, we often try to target appltcations to the most sensitive life stages and to avoid exposure of the less sensitive ones for which resistance genes are more likely to provide a selective advantage. In the case of Bt transgemc crops, this leads to a recommendation against seed mixes of transgenic and nontransgenic plants as a refuge strategy (37), and even against the umntentional impurity of lines. The problem is that heterozygous larvae can grow on nontransgenic plants until they are no longer very susceptible to the Bt toxin, then move to transgenic plants and survive even when then susceptible siblings are still killed, resultmg in a sigmticant fitness advantage to resistance. 2.6. hproved Spray Coverage To the author’s knowledge, the impact of spray coverage on resistance evolution has not been thoroughly investigated m either experiments or simulation modeling. However, simple models suggest that doses causing a range of mortality from 80-100% of susceptible individuals have little effect on the rate of selection for resistance unless resistance is recessive (Fig. 2). One might expect that very sloppy coverage would allow more susceptible individuals to survive any given apphcation, thereby slowing the rate of resistance selection, whereas neither thorough or sloppy coverage would be likely to kill a high proportion of resistant mdtviduals. As described m Subheading 2.3., poor coverage of the

Strategies for Resistance Management

585

crop is but one reason that an insufficient proportion of heterozygotes can be controlled to make a high dose strategy effecttve. However, poor coverage can also mcrease the need for repeated applications, which seems likely to mdirectly increase overall selection pressure, not to mention control costs. Thus, improved spray coverage seemsa sensible goal for both pest control and reststance management.

2.7. Pesficide Rotations Rotatmg the use of pestictdes over an entire area m a so-called “window strategy” tied to the calendar has proven to be a very effective resistance management tactic both m terms of both adoption and efficacy (3444). The use of different compounds at roughly the same time in netghbormg fields creates a mosaic of treatment patterns and should be avoided. Mosatcs are stmply the worst way to deploy a set of pesticides (34). Similarly, the rotation of pestttides within a generation IS not desirable. The problem for both is that you have simultaneous selection with several pesticides, resultmg in much lower pesticrde durability (3445). Consider a case m which selectton for resistance is so strong that you get resistance m a single generation. If you have two pesticides and select resistance to the first pesticide, you at least have the second pesticide to fall back on, for a total of two generattons of control. If instead you splat the population in half, depending on what assumptions you make about the dominance of resistance, roughly half of the populatton will be resistant to each of the pesticides in the next generation, which is essentially a failure after one generation. Make a more elaborate model (34), or do the experiment (49, and you find that rotation across generations 1sroughly twice as good as field-to-field mosaics or rotation within a generation when two or three pesttcides of differing crossresistance (i.e., do not share a common resistance mechamsm) are available. When there are overlapping generattons, one should aim for a cycle longer than the mean generation time. Most pesticides differ in efficacy within a season and in effects on beneticial species. This provides a rational basis for allocatmg pestictdes during a year to specific windows to optimize their use for both pest and pestictde resistance management (e.g., 27). For example, Bt products often have lower efficacy than other insecttcides, and may therefore be more appropriate to use early in the cropping cycle, when the crop can withstand higher denstties of larvae and preservatton of benetictals is especially important. In the case of insecticides, each new mode of action might be allotted a 1-3 mo period depending on local condttions (e.g., duration of the cropping season, which months had the fastest generations). This system has proved to be extremely successful for cotton m Australia and Zimbabwe (34,41,44).

586

Roush

The advantages are not specifically dependent on fitness costs to resistance. Such fitness costs improve the durability of a pesticide and the effectiveness of any resistance management program (381, but the genetic conditions that favor rotations over other tactrcs on the basts of fitness costs appear to be rare (3446) 2.8. Pesticide Mixtures Contrary to another popular myth, it is not necessarily true that mixtures of pesticrdes will delay resistance. It has long been clear that mixtures of msectitides do not necessarily delay resistance compared to the rotational use or sequential mtroduction of the same msecticides. Experimental studtes have failed to consistently find any advantage to mixtures (47,48), and theoretical models showed that mixtures will significantly delay resistance only when several condttions are met (34,38,39,49). Even in the case of fungicide mixtures, whtch are widely considered to have been effective m delaymg resistance, the effective mixtures have used combmations of protectant fungicides to which resistance apparently rarely evolves (presumably because of multiple sites of action) with systemic fungicides of more specific modes of action (28). To be most effective, mixtures require low mitral frequencies of the resistance genes, refuges (as with the high dose strategy, such that resrstant genotypes are rare and can be diluted), high mortal&y from each of the pesticides when used alone, a high spatial correlation of residues (not just equal decay rates), and a lack of crossresistance between the toxins. In sum, the key is that almost all individuals resistant and exposed to one pesticide must be killed by the other for mixtures to be highly effective (34,38,39) These conditions, especially a high spatial correlation of residues, are probably rarely met for sprays. For example, experiments with mixtures of Bt serotypes, applied at doses that did not provide high levels of control when used Individually, failed to delay resistance in Indianmeal moth (50). Not only is this experiment a good model for the field (where control with Bt sprays probably rarely exceeds go”/,), the results are just as would be predicted from the models outlined above, On the other hand, significant delays of resistance might be achieved with transgemc crops, where highly effective concentrations of toxin might be maintained (37-39). As an example of the problems that can result from mixtures of toxins, Bt sprays often include a mixture of specific toxins that appear to attack different bmdmg sites within the insect gut. The resistance of the diamondback moth to Bt has resulted from the use of B thuringiensis subspecieskurstukl (Btk) (4,6-$), which produces CrylA and Cry2 toxins (51). Resistanceappearsto be caused by reduced binding of the Cry 1A (and Cry1 F) toxins to the insect’s mid-gut membrane, with relatively ltttle crossresistance to toxins from other families, especially CrylC (7,23-15). Subsequent to widespread resistance to Btk, B

Strategies for Resrstance Management

587

thuringlenszs subspecies aizawai (Bta) was marketed. Bta produces Cry1 A, Cry lC, and Cry 1D proteins (.51), where CrylC is apparently the toxm with sufficient activity to control resistant larvae. In at least some diamondback moth populatrons, resistance to Btk appears to dechne m the absence of continuing sprays (7,16). However, given that Bta Includes CrylA toxins, it seems likely that use of Bta would maintain enough exposure to CrylA toxins to retard the decline m resistance to Btk To mvestrgate this, Btk-resistant diamondback moth larvae from Florida were divided mto four treatment groups and selected with (1) Btk, (2) Bta, (3) purified Cry1 C toxm, or (4) left unselected. When tested with 100 pg/mL CrylA(b) m leafdip assaysafter four generations of selection, the unselectedand Cry 1C selectedcolonies showed 58-70% mortality, but the Btk and Bta colonies both showed only 4% mortality. Thus, there was enough CrylA toxin in the Bta product to mamtam resistancefor Cry1 A-resistance gene(s) (52). In the field, this would eliminate the possibihty of even occasional reuse of Btk products relying on Cry IA. Mycogen Corporatron has recently developed a Cry 1C-specific product from transformed Pseudomonas, which seems to be the desirable alternative to Bta for resistance management. In general, products with shared toxins, especially for those to which resistance 1salready widespread m the targeted pests, should be avoided. 2.9. Specific vs Broad Spectrum Products All other things being equal, selective products are probably less likely than broad spectrum insecticides to select for resistance, for two ecological reasons. First, a pesticide that is so broad m spectrum that it eliminates the natural enemies or competitors of a pest will probably be sprayed more often, simply because the natural enemies are less able to suppress the pest. Second, when there are multiple pest targets for the same pesticide, applications against either pest will select for resistance to both pests simultaneously. For a hypothetical example, consider a pesticide that is effective against the Colorado potato beetle and the European corn borer in potatoes. Applications against corn borers could select for resistance in the Colorado potato beetle even when it was at such low densities that control was not required. Resistancewould be selectedm potato beetles even when there was no economic benefit to their control.

3. Resistance Monitoring Resistance monitoring can be very important for determining tf a resistance management strategy is failing and whether improvements are required. Unfortunately, resistancemonitoring efforts in the past have often failed to go beyond documenting failures and have rarely predicted failures before they occurred; momtoring will be uselessif there is not also some follow-on management!

588

Roush

However, a key problem for routine momtormg efforts (m contrast to determining background frequencies, as discussed m Subheading 1.3.) for some species is the large sample sizesneeded. Hundreds or thousands of mdividuals must be tested to detect resistance at a 0.1-l% frequency at any given location before a crisis, especially where bioassays are difficult and do not neatly distinguish between susceptible and resistant genotypes (53). Given the dtfficulties of btoassays with biopesticides, and the limited knowledge of their modes of action, it seemsmost prudent to invest in highly proactive resistant management plans designed to delay resistance before it is ever detected. 4. Example of a Resistance Management Strategy: Diamondback Moth As noted before, it would be impossible to outline a general resistance management strategy for all pests and biopesticides, even Just for those described m this book. However, it seemsworthwhile for illustrative purposes to outline a strategy for just one pest, The diamondback moth is a useful example because it was the first pest to evolve resistance to a btopesticide in the field, contmues to be economically important, is widely resistant to other pesticides, and seems hkely to continue to be a target of biopesticides. Perhapsthe single most important feature of any insecticideresistancemanagement effort is a samplmg scheme that eliminates unnecessarysprays,and targets spraysonly to thoseareaswhere spraysare truly needed.To do this Mollywill require the development of alternate controls. One tactic that has been used effecttvely m Mexico (A. Shelton, personal commumcatton) and Australia is a break in crucifer crop production during oneseasonof the year, which starvesthis host specific insect. In addition, the available msecticidesm Australia have been assigned to one of two 6 mo long annual windows on the basrsof presumed patternsof crossreststance and modes of action (R. Roush, P. Buerger and S. Jones, unpublished). Bt sprays can be used m either window, but arerecommended only agamstsmall larvae early in the growth of the crop. If suitable products become available m the future, Cry I A-specific products will be used in one window and Cry IC products m the other. Growers are being urged to use insecticides that are “soft” on beneficials early m the growth of the crop, with broader spectrum pesticides toward harvest. To mmimtze the effects of persistent residues, growers are urged to destroy the crop immediately after harvest to prevent pest population growth and exposure on regrowing plants. Mixtures and high dosesare to be avoided, with growers encouraged to use application rates at the low end of those on the pesticide label. 5. Implementation Identifying an appropriate resistance management strategy is relatively easy compared to gaining adoption of the strategy, and indeed, this has been the

Strategies for Resistance Management

589

most hmitmg factor to the success of resistance management efforts. Ultimately, any resistance management strategy wtll be most effective if it has the support of the private sector. The pesticide industry has established the Fungicide, Insecticide, and Herbicide Resistance Action Committees (FRAC, IRAC, and HRAC, respectively) to assist in this process. However, contrary to popular myth, it is the pesticide users, not the pestttide manufacturers, who have the most to lose when resistance evolves in those caseswhere it is such a problem that we must really be concerned with managing it. In these cases,generally only one or two pesticide options are generally available at any given time (e.g., 28). If the pesticide does not more than pay for its purchase cost in terms of improved price for the crop, the grower would be foolish to use it. On the other hand, this cost far exceeds the actual profit to a company. Thus, on a per unit basis, the cost of a lost pesticide to a grower, especially when there are few, if any, alternative pesticides, must considerably exceed the revenue lost to a pesticide company if the pesticide fails. In the case of the Colorado potato beetle, for example, resistance to all of the existmg pesticides can easily cost more than $750 per hectare (30) in the northeastern United Statesand Canada; a recently registered insecticide that provides effectively complete control is sold for less than $200 per hectare, of which probably much less than half is profit to the company. Because government scientists in some measure represent the interests of the general public and growers, the public sector thus has an obligation to address resistance issues, Both the pesticide companies and pesttcide users have strong economic incentives to manage resistance, yet resistance is still generally poorly managed. Ultimately, a stronger partnership is required between the public and prtvate sectors to assure that the promise of reststance management is fulfilled. 6. Conclusions Although resistance management is often perceived as a complex problem, the list of potential tactics 1sgenerally so short that choosing which would be useful is not difficult. Most of these tactics are also complementary and are most effective when adopted before selection commences. Thus, even though reststance management can always be improved with additional data, one should also aim to adopt a resistance management plan at the first introductton of the product. References 1. Georghiou, G. P. and Lagunes-Tejeda,A. (1991) The Occurrence of Remtance to Pesticides m Arthropods. FAO, Rome, Italy. 2. Jasienuik,M., Brule-Babel, A. L., and Morrison, I. N. (1996) The evolution and geneticsof resistancein weeds. Weed Science 44, 176-l 93.

590

Roush

3. McGaughey, W. H. (1985) Insect resistance to the biological Insecticide Baczllus thurzngzensu. Science 229, 193,194 4. Tabashnik, B. E., Cushmg, N. L , Fmson, N., and Johnson, M. W (1990) Field development of resistance to Bacillus thurzngzenszs in diamondback moth (Lepidoptera: Plutelhdae) Jr Econ Entomol 83, 1671-1676 5. Hama, H., Suzuki, K., and Tanaka, H. (1992) Inheritance and stability of resistance to Bactllus thunngzensts formulations in the diamondback moth, PlutelEa xylostella (Lmnaeus) (Leprdoptera Yponomeutidae). Appl Entomol Zoo1 27,355-362 6. Shelton, A. M., Robertson, J. L , Tang, J. D , Perez, C , Ergenbrode, S D , Preisler, H K., Wtlsey, W. T., and Cooley, R J (1993) Resistance of dramondback moth (Leprdoptera. Plutelhdae) to Bactllus thurzngzenszs subspecies in the field. J Econ Entomol 86,697-705 7. Tabashnik, B E. (1994). Evolution of resistance to Bactllus thurzngzenszs. Annu Rev Entomol 39,47-79 8. Perez, C. P. and Shelton, A. M (1997) Resistance of Plutella xylostella to Bactllus thurzngtenszs Berlmer in Central America. J Econ Entomol. 90, 87-93. 9 LI, H and Lerfert, C. (1994) Development of resistance in Bottyottnta fuckelzana (de Bary) Whetzel against the brological control agent Bacillus subtzlzs CL27. Ztetschnftfur Pflanzenkrankheiten und Pflanzenschutz 101,4 14-418 10. Roush, R. T. and Daly, J. C (1990) The role ofpopulatton genetics tn resistance research and management, m Pesttcrde Reststance m Arthropods (Roush, R. T and Tabashmk, B E , eds.), Chapman and Hall, New York, pp. 97-l 52 11 Natronal Research Council, ed (1986) Pesticide resistance. strategies and tacttcs for management Natronal Academy Press, Washington, DC 12. Van Rie, J., McGaughey, W H , Johnson, D E , Barnett, B. D , and van Malaert, H. (1990) Mechanism of insect resistance to the microbial msectrctde Baczllus thurtngtensu. Science 247, 72-74 13 Fe&, J , Real, M D., Van Rie, J., Jansens, S., and Peferoen, M (1991) Resistance to the Bactllus thurtngiensts biomsectictde m a field population ofPlutella xylostella IS due to a change m a midgut membrane receptor Proc Nat1 Acad Scz USA 88,5119-5123 14 Tang, J D , Shelton, A M , Van Rie, J., De Roeck, S , Moar, W J., Roush, R T , and Peferoen, M. (1996) Toxlcrty of Baczllus thurzngzenszs spore and crystal protein to the resistant diamondback moth (Plutella xylsotella). Appl Envtron. Mtcrobrol. 62,564-569. 15. Tabashnrk, B. E., Lm, Y.-B., Fmson, N., Masson, L , and Heckel, D G. (1997) One gene m diamondback moth confers resistance to four Bactllus thurtngtensrs toxins Proc Nat1 Acad Sci. USA 94, 1640-644. 16. Tang, J. D., Grlboa, S., Roush, R T., and Shelton, A. M (1997) Inheritance, stability, and fitness of resistance to Bactllus thurzngtensu in a field colony of Plutella xylostella (L.) (Lepidoptera. Plutelhdae) from Florida. J Econ Entomol 90,732-741 17 Briese, D T. (198 1) Resistance of insect species to mtcrobial pathogens, m Pathogenesis of Invertebrate Mtcrobtal Diseases (Davidson, E. W., ed.), AllanheldOsmun, Totowa, NJ, pp 5 1 l-545

Strategies for Resistance Management

591

18 Ignoffo, C M., Huettel, M. D., McIntosh, A. H., Garcia, C , and Wtlkenmg, P (1985) Genetics of resistance of Helzothzs subfexa (Lepidoptera. Noctuidae) to Baculovwus hellothis. Ann. Entomol Sot. Amer 78,468-473 19 Helle, W (1965) Resistance in the Acarina: mites. Adv Acarol. 2, 71-93. 20. Sanderson, J. P. and Roush, R. T (1995) Management of insecticide resistance m the greenhouse, m Proceedmgs of the I1 th Conference on Insect and Disease Management on Ornamentals (Bishop, A., Hausbeck, M , and Lmdqurst, R , eds.), Society of American Florists, Alexandria, VA, pp. 23-36. 21 Tabashmk, B E. and Croft, B. A. (1985) Evolutron of pesticide resistance in apple pests and their natural enemies. Entomophaga 30,37-49 22 Roush, R. T and Croft, B A. (1986) Experimental population genetics and ecological studres of pestictde resistance m msects and mites, in Pestrclde Reswtance Strategies and Tactics for Management (National Research Council, ed.), National Academy Press, Washington, DC, pp 257-270 23 McDonald, P T., Schmidt, C. D., Fisher, W. F , and Knuz, S. E. (1987) Survival of permethrin susceptible, resistant and Fl hybrid strains of Haematobia rrrltans (Diptera* Muscidae) on ear-tagged steers J Econ Entomol 80, 1218-1222. 24 Roush, R. T , Combs, R L , Randolph, T C., MacDonald, J , and Hawkms, J A (1986) Inheritance and effective dominance of pyrethroid resistance m the horn fly (Diptera: Muscidae) J Econ Entomol 79, 1178-l 182. 25. Rotteveel, T. J. W., de GoeiJ, J W F. M , and van Gemerden, A F. (1997) Towards the construction of a resistance risk evaluation scheme Pestzc Scz 51,407-411. 26 Roush, R. T. and McKenzie, J A (1987) Ecological genetics of insecticide and acaricide resistance Ann Rev Entomol 32,361-380. 27 Gould, F , Anderson, A., Jones, A, Sumerford, D., Heckel, D. G., Lopez, J , Micmski, S , Leonard, R , and Laster, M (1997) Initial frequency of alleles for resistance to Bacillus thuruzgzenszs toxins in field populations of Helzothzs vu-escens. Proc Natl Acad. Sci USA 94,3519-3523. 28. Brent, K. J. (1995) Fungicide Resistance in Crop Pathogens How Can It Be Managed7 Fungicide Resistance Action Commitee Monograph No 1, GIFAP (International Group of National Associations of Manufacturers of Agrochemical Products), Brussels 29. Tabashmk, B E (1997) Seeking the root of insect resistance to transgenic plants. Proc Nat1 Acad Sci USA 94,3488-3490. 30 Roush, R T. and Tingey, W. M. (1992) Evolution and management of resistance in the Colorado potato beetle, Leptmotarsa decemllneata, m Resistance ‘91 Achievements and Developments in Combating Pestzczde Resutance (Denholm, I., Devonshire, A. L , and Holloman, D. W , eds.), Elsevier Applied Science, Essex, UK, pp. 61-74 31 Roush, R T., Hoy, C W , Ferro, D. N., and Tingey, W. M (1990) Insectrcide resistance m the Colorado potato beetle (Coleoptera. Chrysomebdae). Influence of crop rotation and insecticide use. J Econ Entomol 83,3 15-3 19.

592

Roush

32 Brown, A W A. (1967) Insectrctde resrstance-genetrc implicattons and apphcatrans. World Rev Pest Control 6, 104-I 14. 33 Denholm, I , Farnham, A W , O’Dell, K , and Sawicki, R. M. (1983) Factors affecting resistance to msecttcrdes m house-fhes, Musca domestzca L (Dtptera Musctdae) I. Long-term control wrth bioresmethrm of fltes with strong pyrethrotd-resrstance potential. Bull. Entomol Res. 73,48 l-489. 34 Roush, R T. (1989) Destgnmg resistance management programs* How can you choose? Pesttc Scz 26,423441. 35 Schroeder, H. E , Gollasch, S., Moore, A., Tabe, L M , Cratg, S , Hardte, D C , Chrrspeels, M. J., Spencer, D., and Hrggms, T J V (1995) Bean alpha-amylase mhtbttor confers resistance to the pea weevil (Bruchus puorum) m transgenic peas (Ptsum sativum L ) Plant Physiology 107, 1233-1239 36 Gould, F (1998) Sustamabtlrty of transgemc msectictdal culttvars. mtegratmg pest genettcs and ecology. Annu Rev Entomol 43,701-726 37 Roush, R, T (1994) Managmg pests and their resistance to Baczllus thunngzensu Can transgemc crops be better than sprays? Bzocontrol Set Technol 4,501-5 16 38 Roush, R. T. (1997) Managing resistance to transgemc crops, m Advances zn Insect Control The Role of Transgentc Plants (Carozzi, N , and Koziel, M , eds ), Taylor and Francis, London, pp. 271-294. 39 Roush, R T (1997) Bt-transgemc crops: Just another pretty msectrctde or a chance for a new start in resistance management7 Pesttc Set 51, 328-334 40 Tabashmk, B E and Croft, B A. (1982) Managing pesttctde resistance in croparthropod complexes Interactions between brologtcal and operatronal factors Envrron Entomol 11, 1137-l 144. 41 Denholm, I. and Rowland, M W. (1992) Tactics for managing pesttcrde reststance m arthropods. theory and practice Ann Rev. Entomol 37,91-l 12 42 Daly, J., Fisk, J. H., and Forrester, N W. (1988) Selective mortality m field trials between strams of Helzothzs armzgera (Lepidoptera. Noctuidae) resistant and susceptible to pyrethroids functronal dominance of resistance and age class J Econ Entomol. 81, 1000-1007. 43 Metz, T. D , Roush, R T , Tang, J D , Shelton, A M., and Earle, E D (1995) Transgemc broccoli expressing a Bacillus thurzngtenszs msecticrdal crystal protein. tmplrcattons for pest reststance management strategies. Molecular Breeding 1,309-3 17. 44 Forrester, N W., Cahrll, M , Bird, L. J., and Layland, J. K. (1993) Management of pyrethroid and endosulfan resistance in Heltcoverpa armtgera (Leprdoptera. Noctutdae) m Austraha. Bull Entomol Res Suppl 1. 45 Roush, R T (1993) Occurrence, genetics and management of msectrcide reststance. Parasttology Today 9, 174-179. 46. Curtts, C. F (1987) Genetic aspects of selection for resistance, in Combatrng Resistance to Xenobzotzcs (Ford, M G , Holloman, D W., Khambay, B P S , and Sawtcki, R. M , eds.), Elhs Hot-wood, Chichester, UK, pp 15 1-16 I 47 Tabashmk, B. E (1989) Managmg resistance wrth multtple pestrctde tactrcs theory, evidence, and recommendations. J Econ Entomol 82, 1263-1269

Strategies

for Resistance

Management

593

48 Immaraju, J A., Morse, J G., and Hobza, R. F. (1990) Fteld evaluatton of msectrcide rotatton and mixtures as strategies for citrus thrlps (Thysanoptera Thrtptdae) resistance management in California. J Econ Entomol 83,306-3 14 49. Gould, F (1986) Srmulatton models for predtcting durabilny of Insect-resrstant germplasm a determnustrc drplotd, two locus model. Envu-on Entomol. 15, l-10 50. McGaughey, W. H. and Johnson, D. E. (1987) Toxtctty of different serotypes and toxins of Bacillus thurlngzensrs to resistant and susceptible Indtanmeal moth (Leptdoptera: Pyrahdae). J Econ Entomol 80, 1122-1126. 51 Kozrel, M G , Carozzr, N B., Currier, T. C , Warren, G. W , and Evola, S V (1993) The msectictdal crystal protein of Bacillus thunngzensx past, present, and future uses. Bzotechnol Genet Eng Rev 11, 171-228. 52. Tang, J D , Shelton, A M., Roush, R. T., and Moar, W. J. (1995) Consequences of shared toxins m strains of Bacillus thurzngiensu for resistance m dtamondback moth Pestlclde Resistance Management Newsletter 7(l), 5-7 (also on World Wide Web at http.//www msstate.edu/Entomology/vln Us95rpm html#art03) 53 Roush, R. T. and Miller, G L (1986) Constderattons for destgn of msecttctde resistance momtoring programs J Econ Entomol 79, 293-298.

31 Field Management Delivery of New Technologies to Growers Mark E. Whalon and Deborah L. Norris 1. Introduction Since ancient times, man has recognized and utilized biological agents for pest control (I). Modern agriculture has built on this foundation of indigenous knowledge to explore and advance the use of novel methods of biological control of insect pests. In the past 30 yr, biotechnological innovation, including natural product chemistry, fermentation, and genetic engineering, has led to the development of many revolutionary products that have fundamentally changed the way humans manage pest-control agents. The use of chemicals in arthropod control can now be divided into two categories: conventional chemical insecticides and bioinsecticides. Furthermore, the definition of bioinsecticides has expanded to include the use of genes and gene products, microbes or products derived from microbes, plants, and other biological entities. The development of recombinant DNA technology and other factors, e.g., concerns over the environmental and health risks of conventional chemicals, the development of resistance to existing chemicals, and a growing interest in IPM has accelerated research interest in biopesticides as chemical-pesticide alternatives. Growers today find ever-increasing bioinsecticide products available to them, from conventional spray deployment of Bacillus thuringiensis (Bt) to insect-specific viruses, sprayable pheromones, and transgenic plants with insecticidal genes incorporated into plant tissue. However, biopesticide development and utilization has many challenges today (2). The US Environmental Protection Agency (EPA) has officially defined a biopesticide as belonging to one of three groups: biochemical pesticides, microbial pesticides, or transgenic plant pesticides (3). Biochemical pesticides From: Methods in Biotechnology, vol. 5: Biopesticides: Use and Delivery Edited by: F. Ft. Hall and J. J. Menn 0 Humana Press Inc., Totowa, NJ

595

596

Whalon and Norris

include pheromones, hormones, natural Insect- and plant-growth regulators, repellents, and enzymes as active ingredients. Microbial pesticides include microorgamsms (bacteria, fungi, and viruses) and then products, usually proteins, as active ingredient (4). Becauseof growing consumer food-safety awareness and demands for less-toxic substitutes for conventional chemical pesticides, there has been a commitment on the part of the EPA to facilitate registration and bring these new products to market. By redefining biopesttctdes as a separate class of control agents, the EPA has been able to revise the registration requirements for these compounds, most of which now have fewer data requirements for registration than conventional broad-spectrum chemical pesticides. By promotmg a more rapid path to market, the EPA has helped encourage the agrochemical and biotechnology industries to spend more time and resources m the development of new btomsecticides and related technologies. Implementation of these new technologies will require an mterdtsciplmary approach to pest management and Increased grower access to information on use and deployment. The goal of this chapter is to discuss some of the challenges to the ecologically sound implementation and adoption of biopesticides, challenges that the authors believe can only be overcome through changes m the structural and phrlosophical aspects of agrrcultural production, mcludmg the consideration of the mtrmsic value of our natural and ecological resources, when making pest-management decisions. 2. Why Have Growers Been Slow to Adopt Bioinsecticides? Despite the benefits of many new mnovattons, a considerable time lag is generally required before an innovation receives widespread acceptance (5). Cultural and social values sometimes play a role m grower dectsion-making, and consumer fears of microbial agents can also be a barrier to the acceptance of new technologies, especially those related to food products (6) More importantly, economic feasibihty and product performance play a large role m the adoption of new technologies by growers. Btopesticides usually do not perform as effecttvely and/or as quickly as synthetic chemical pesttcides. In addtnon, many requtre a higher level of management skill and knowledge than do conventional chemicals (7,s). These and other factors, such as target-pest spectrum, crop value, and production and operational challenges, can affect the adoptton and field management of btomsecticide technologies. In addition, tt is not likely that biopesticides will replace conventional pesticides in the nearterm, but they are already being broadly accepted in IPM programs. 2.1. The Case of Transgenic Plants For the past decade, the development and commercialization of msect-resistant transgenic plants has focused almost exclusively on transferring toxin

he/d Management genes from Bt to crop plants (9). Bt 1s a common so11 bacteria that produces many different crystal toxins that are selective against specific pest species. Bt sprays, which have been used commercially in the United States since 1958 (IO), have several advantages over synthetic insecticides, including httle or no Impact on nontarget and beneficial organisms, low environmental persistence because of rapid degradation, and little or no known toxic effects against humans and other animals. From the growers’ perspective, however, some of these benefits actually make Bt less attractrve as an alternative control tactic. For example, insect specificity makes Bt less competrtive with many broad-spectrum synthetics, and rapid degradation increases grower input requirements m the form of repeated applications and greater management of timmg and location of sprays to achieve comparable results. These drawbacks are rllustrattve of the potential problems with most bropesticides, which exhibtt selective rather than broadspectrum activity. Transgenic Bt technology aims to overcome some of these delivery barriers by engineering crop plants to express high levels of Bt toxin(s) within plant tissues continuously throughout the growing season, thus eliminating the burden of spatial and temporal management of apphcatrons.

3. Feasibility of Biopesticide Acceptance The economic feasibility of bropesticide acceptance IS determined by many factors, all of which can contribute to the eventual adoption or rejection of new technologies by growers. As summanzed by Reichelderfer (II), these factors include. 1. Economic incentive: The net gain from the use of a bropesticide must equal or exceed the gain that the grower would recerve from conventronal pest-control tactrcs (chemicals) 2 Efficacy: Given equal costs, broinsecticldes must be as effective as the chemrcal they would replace. As efficacy of bioinsecticides increases, economrc feasibllrty wrll increase. 3. Pest spectrum: Pest specificity can be an advantage of bioinsecticrdes in smglespecres outbreaks, but, m many cases, growers are faced with the challenge of elrmmatmg several pests at once When a srmultaneous outbreak of several pests occurs, there IS little justrficatron m using a species-specific broinsectrcrde, If several broad-spectrum chemical alternatives are avarlable Therefore, the development of some broad-spectrum biomsectrctdes or mrxtures, to compete with then synthetic chemical predecessors, can increase the economic feaslbrlity of those products. 4. Crop price As crop value increases, there is greater interest m using rapid, rehable, and effective pest-control measures. This is achieved most frequently through the use of chemical treatments that are already familiar to the grower Growers are less hkely to use a more expensive and more specific bioinsecticide, unless its performance IS better than Its conventional chemical counterpart or

Whalon and Norris

598

other incentives of reduced residues, pest resistance, consumer acceptance, or loss of conventional-control chemicals 5 Price of blopestlclde. The lower the market price of a blopestlclde, the more economically feasible it IS to the user However, if a blopestlclde product offers other relative advantages, such as low human and environmental toxlclty, and consumer or regulatory pressures like the passage of the 1996 US Food Quahty Protection Act (FQPA) make its use more attractive, then a htgher price may be more readily accepted by growers 6 Varlabllity of performance. Users prefer consistency and rehablllty m the pestcontrol methods they choose, therefore, economic feaslblllty will increase as blomsecticldes become more consistent m their field performance 7. User costs Growers factor m costs other than the market price of a product when they are making pest-management choices The costs associated with time, labor, and management, when usmg a new technique or product, can increase the overall cost to the user and make the product less economically feasible Many growers ~111 continue to use conventional chemical tools if they beheve that the additional costs associated with using a bioinsectlcide, Including transgemc seed license fees, scouting, timing, multiple apphcatlons, and so on, will outwelgh the potential benefits.

4. Implementation 4.1. Biopesticides

Issues vs Conventional

Systems

Biopestlclde products function within a narrow host-target range, which, although advantageous for preserving the ecological balance of nontarget and beneficial insects, is nonetheless disadvantageous to the grower during multlple-pest outbreaks. Conventional systems usually provide greater broad-range control, but, as a result, have more negative effects on nontarget and beneficial populations. In addition, conventional systems often preclude an understandmg of other factors m pest management, espectally pest biology, insect behavior, agroecology, and mode of action, because of their broad-spectrum activity and multiple mechanisms of toxlclty (fumigant, contact, systemic, and per OS) With the exception of a few contact bloinsectlcldes like Beauveria basszana, the use of species-specific, noncontact, ingestible blomsecticldes has created a need for greater understandmg of the complete pest complex, includmg mul-

tiple insect mteractlons and behavior. To illustrate this point, consider the mtroductlon of tebufenozlde, an insect-growth regulator (IGR) for lepidoptera control in apples This product has not performed as well as expected in mldwest apple pest control, even though it utilizes a novel mode of actlon and

may cn-cumvent organophosphate msectlclde resistance m target leafrollers. In the case of organophosphate, carbamate, and synthetic pyrethrold products, failure in the field has often been associated with application

problems,

timing,

or resistance in the target population. In the case of this narrow-spectrum TGR,

Field Management

599

leafroller biology, ecology, and behavior are more tmportant contrtbutmg factors to field failure. For example, behavioral avoidance, whereby early-season larvae move their feedmg to newly emerging leaves, avoiding mgestion of the IGR, has led to recent product failures. Since the effectiveness of many IGRs require mgestion of a treated plant part, a behavioral characteristic, such as feedmg preference for new-growth plant tissues or product detectton, may result in an effective avotdance to exposure. Thus, the use of this product challenges pest managers to rethmk then approach to management by mcluding better momtormg, more knowledge of pest behavior and host-plant mteractions, preservation of beneticials, as well as a change m the acceptable damage thresholds and treatment frequency. Tables 1 and 2 compare some of the fielduse considerations for several types of conventional pesticides and biopestictdes.

4.2. Will an Era of Biopesticides Replace the Age of Synthetic Pesticides? Since the mtroduction of DDT and the advent of the chemical age m agrtculture in the 194Os,pest control has been primarily dependent on insect neurotoxins that are broad-spectrum, fast-acting, active by contact, and persistent (Table 1). These characteristics, along with low cost and relative ease of apphcation, have provided positive economic feedback through effective, simple control. Meanwhile, however, a second system of negative feedback has been moving m the opposite direction to counteract the positive effects of conventional insecticide use (Fig. 1). Negative feedback (externalities) has developed relatively unchecked, in part because of the necessary lag time between the mtroduction of an innovation (i e., chemical pesticides) and the development or evidence of these negative impacts. The fact that optimal use of insect neurotoxins has traditionally been assessedin purely economic (production) terms has further hindered the ability to detect long-term, noneconomtc, negative feedbacks. The cycle of negative feedbacks includes environmental impacts, ecological impacts, lack of sustainability, pest resistance, human health impacts, trade issues, regulatory burdens, and consumer concerns (Fig. 1). Many of these factors are often ignored m traditional cost-benefit analyses, because of difficulties in quanttfying or assessingtheir value (12). In the United States, under FQPA codificatton, many conventtonal pesticides, nerve toxms, and B-l and B-Z carcinogens will dramatically decrease, especially on minor crops (fruits, nuts, vegetables, and ornamentals). A pestmanagement system dependent on biopesticides is likely to evolve in the very near future, especially m these minor crops. These systems will have to be supported with better monitoring, as with apple leafroller control with tebufenozide. Newly developed pest- and natural enemy mteraction thresholds (I.?), similar to those developed for predaceous mites m deciduous fruits (14),

Low Moderate

Low

Chlormated hydrocarbons Synthettc pyrethrotds

011

Low

Low

(OPs)

Attentton to application

Precision and timing requnement

Chemical

Low-moderate

Low Low-moderate

Low-moderate

Low-moderate

for Conventional

Carbamates

Organophosphates

Compound

Table 1 Field Use Considerations

Usually not

No No

Usually not

Usually not

Thresholds necessary?

Insecticides

Vanable low-htgh

Little Little

Some

Some

Phytotoxictty

Wash off, volatrhty Wash off, volatihty Less than OPs Temperature inversion Wash off

Weather effects

High; short- to medium-term Hugh; long-term High; short- to long-term High

High, short-term

Environmental and ecologtcal tmpact

2

High Moderate-htgh Refugta

High Low-moderate Moderatehtgh High

Compound

Pheromone Bt sprays Transgemcs

Spinosad Nicotmyl compounds Avermectm Tebufenozide Fenoxycarb

High Low-moderate Moderate Htgh

Low-moderate High None

Prectsion and timing requirement

for Biopesticides

Attention to application

Table 2 Field Use Considerations

Sometimes Sometrmes Somettmes Yes

No Usually No

Thresholds necessary7

None Non+low None None-low

None Low None

Phytotoxicity

Moderate Moderate Low-moderate Htgh

Hrgh Hrgh Some

Weather effects

None+? Some Moderatehigh (gene transfer, resistance) Some Some Some Some

Envnonmental and ecological impact

602

Whalon and Norris

Envtronmental impacts EcologIcal impacts Non-renewable use of resources Resistance Human health Impacts Consumer concerns

Broad-range control Persistence Multiple modes of

Greater management requlrementslcosts Greater technical expertise needed lnitlally Reassessment of acceptable damage levels Resistance If mismanaged Loss of susceptible genes if mismanaged

resources Less Impact on beneficlals Less envlronmental

Renewable Potentially

Fig 1. Integrated production

system low cost

feedback mechanisms

will be very useful. In addition,

for conventional

and blomtenslve

better weather forecastmg, and even microcli-

mate monitormg, will significantly improve reapphcation decisions where washoff and heat-moisture weathering are problematic to blopestlclde performance. A pest-management approach dependent primarily on blopesticides 1s likely to involve

a new set of negative externahtles,

but these are likely

to be longer-

term and of lower intensity than conventional chemical pesticides The eco-

Field Management

603

logical, environmental, and human health advantages Inherent m biopesticide systems will eventually be challenged by these negative feedback factors, some of which we cannot yet fully predict or identify (Table 2). In the short term, the most important factors are likely to be the technical, educational, and cost barriers that could adversely affect a grower’s choice to implement brologically intensive IPM, but under FQPA most minor-use producers will not have a choice. These barriers mclude the challenge of balancing kill rates and damage levels with fluctuations in pest populations, the knowledge requirements for understanding the biological and ecological mteracttons of the insect-crop complex, and the technical and managerial expertise needed to run the system effectively, These factors affect economic feasibihty and net dollar returns that ultimately dictate the extent to which the grower IS willmg or economically able to adopt a new control paradigm, especially a more biologically intensive program. 4.3. Avoiding

the New Biopesticide

In conventional

chemical

Treadmill

systems, producers

catastrophically

eliminated

many species,greatly simplifying biological diversity and interactions (2627). These simplified systems often undergo crisis when a new pest is mtroduced, because resistance occurs or secondary pests rebound. Biopesticide pest-management systems may also be at risk of failure because of then- increased reltante on system mteractions, and ultimately on the systems’ total genetic resources (plant genes, pest genes, beneficial insect genes, bacterial and virus genes, and so on). Of great concern is the potential for resistance and loss of genes that confer susceptibility

in pest populations

Figure

1 illustrates

the

potential complexity of the biointensive pest-management feedback systems. A short-sighted approach to biopestrcide use could easily disrupt the finer balance necessary for this system’s ecological stability, and the ensumg economic consequences will be Just as significant as with conventional pestmanagement systems. The study reported by Trumble et al. (28) IS an example of a biopesticide pest-management system m a minor crop. It compared the economic and environmental

impact of a Bt-based IPM program with the current standard chemi-

cal program m California celery production. The program utilized scoutmg of both pests and natural enemies (parasites) to implement

a system of threshold-

driven msecticide applications, which reduced Inputs and provided protection of the natural parasites. The IPM materials (Bt and abamectin) were chosen based on thetr control potential of the target pests (mainly Spodoptera exigua), balanced with minimal disruption of natural parasites. Both the conventional and btopesticide program was evaluated for yield, crop value, and cost of control. Although

the economic

evaluation

did not take mto account hidden costs

604

Whalon and Norris

and externalities, such as ecological and environmental effects, the researchers did separately evaluate the potential for air pollution of both programs, to compare the environmental benefits of biopesticide vs the conventional chemical standard. Their results indicated that the biopesticide program achieved a net profit of $4 1O-l 485/ha greater than the standard conventional program m three trials over 3 yr. These researchers recognized that the benefits of the biopesticide program went far beyond net monetary profit, and included decreased negative human health effects, decreased environmental impacts, and the primary advantage of decreased pest resistance in the short-term. As a single strategy in a one-crop system, thts IPM approach adheres to many of the prmciples of a more durable production paradigm (28). The system utihzes multidisciplmary information to select which compounds to apply, and when to apply them, taking mto consideration both pest and nonpest toxtctties, population dynamics, envnonmental concerns, and newly developed damage thresholds. Thrs approach is proactive and preventive (e.g., scoutmg and allowing natural enemies to control the pest before resorting to treatment), rather than based on a standardized treatment regime. The described program is also likely to preserve genetic resources (susceptible genes, nontarget organisms), and reduce risks to humans and other mammals. It should be noted, however, that the use of a smgle bioinsecticide with scoutmg is insufficient for long-term prevention of pest outbreaks. Trumble et al. (17) suggest that alternative strategies could be equally viable. Crop rotation, ground-cover management (for orchards), and strip cropping are key strategies m this process. 5. Agricultural Research and Field Management: Role in IPM We have mentioned some of the drivmg forces m the transition to renewable production practices This transition will require a more information- and education-intensive commitment on behalf of industry, regulatory agencies, landgrant umversmes, and producers. The implementation of a truly integrated, durable system will demand an interdisciplinary approach, with more elastic and targeted field research that will probably require more producer input. Government agricultural and environmental agencies could help direct thts shift by priorittzing funding for appropriate science, i.e., problem-driven research that addresses basic agricultural and ecological problems associated with btomtenstve productron. Although prevention-oriented pest management may not be the marketing strategy of most biomsectictde companies, some companies do take an active role m educating growers and the public on the use of their products and new technologies. Amertcan Cyanamid (Parsippany, NJ) has created citizens’ advisory panels, composed of local cttizens, environmental groups, university extension personnel, growers, and other stakeholders, in areas where they are

Field Management running field trials on biopesttcides, e.g., on recombinant baculoviruses for insect control. These panels of approx 1540 people observe and assessthe trial programs, and then have opportunities to ask questions and discuss concerns with company representatives. The company also has a targeted technology public relations program that works with some environmental groups to educate, inform, and create public dialog on its recombinant baculovnus technologies. Even in the face of sometimes violent opposition to biotechnology products, their open, consumer-Informed education and labelmg programs have been successful. These types of industry-academic-citizen-group partnerships help foster a more open and dynamic envn-onment for creating change and tmplementmg new biotechnologies, especially technologies that may engender fear m the public. 5.1. Role of Industry, Government, and Academia in Bringing New Biotechnologies to the Field Because proper use of biopesticrdes often requires different kmds of mformation and actions that are new to the grower, performance expectations may not be met, and many growers will return to the use of quick-acting pesticides, if available. Manufacturers and/or registrants of biopesticides and the university extension service, therefore, must educate growers on the proper use of btopesticides in order to bring expectations in line with reality. Informatton and trainmg on timmg and placement is critical, since these factors are paramount m achieving high-efficacy biopesticide performance. In general, companies need to provide more technical support for bioinsecticrdes than conventional synthetics-a necessaryinvestment if they hope to maintain longterm grower interest m then products (18). Benbrook et al. (19) proposed a framework for accelerating progress toward implementation of a biomtenstve IPM programs, which requires a structural foundation in ecology. These control programs would exploit natural and augmentattve biological control and build on an emerging cadre of private sector biological control companies. The first step in achieving this 1sto develop rigorous methodologies for measuring pesticide use and risks for individual crops, both regionally and nationally. Without the establishment of an initial usage baseline, specific changes in pesticide use cannot be effectively measured. Additionally, comparative data is lacking on the performance and cost-effectiveness of biointenstve IPM systemsvs conventional systems.Documentation and monitoring of biointensive programs that currently exist could help in setting future goals for the transition to IPM and use of biologically based cropprotection tools. Quality control, standardization, and an industry-based enforcement mechanism for btological control supply houses are also critical in this transition.

606

Whalon and Norris

Changes in government mfrastructure are also needed, such as re-evaluation of agricultural policies and public funding priorities that promote reliance on conventional chemicals. Currently, the USDA spends approx I2 7% of its total pest-management funding support on btointenstve IPM research (20). A shift m fundmg priorities and an emphasis on researchmg plant-Insect mteractions, ecology, plant transduction and signaling, biopesticide/natural enemy thresholds, and btodiversity are necessary changes m the public-sector research agenda. Likewise, umversities must reprioritize their research agendas to promote more on-farm, producer-assisted research, and increase access to mformatton via databases of pest and pesticide-alternative mformation, field diagnosttc tools, and internet resources. This approach to date has been largely ineffectual, because deciston-support systems,databases,and electronic mformation networks have failed to reach the producers, or address theu needs A good example of the use of a database that helped define research priorities is the PesticidesAt Risk (PAR) program at Mtchigan State University. The database is designed to identify and assessthe risk of loss of various pesticides because of industry withdrawal, reregulation, environmental and consumer concerns, and the 1996 FQPA, and to identify the crops that will be affected if these pesticides are removed from the market. The loss risk ratmg assigned to the pesticides has been used to direct agricultural research funding priorities. Thus, m the United States,the loss of key conventional pesticides used for specialty crops is driving the research agenda in the directton of alternative pest-management techniques that can replace these pesticides when they are removed from the market. The PAR process is illustrative of how new regulations can create a force for transition toward more biointensive IPM research and practices. Other mformation-based delivery systemsare uttlizmg data transfer networks that producers can subscribe to, or real-time weather and biological interface software systemsthat provide on-site predictions of pest development and progress. 6. Conclusion In the past, products that exhibited less-than-optimal performance, poor product support packages, and lack of economic mcentives, compared to conventional chemical pesticides, have limited the adoption of new biopesticides by growers. Advances in biopesticide technology are now increasing the availability of more efficacious and reliable biologically based insecticides In the United States, the EPA has implemented policy changes for bioinsectmides that have facilitated and streamlined registration. With biopesticides, pest managers now have the emerging tools to allow current agricultural production tools to be more environmentally friendly and establish potentially durable pest-control systems. The challenge involved m transition from conventional synthetic pesticides is twofold. Changes are

Field Management

607

required m the structure and function of agricultural researchand Implementation at the local, private, and public levels, both nationally and mternatlonally; and subsequentadoption of a more biologlcally mtenslve understanding of agriculture production by producers is necessary.The structural changes should involve, m part, refocusing public institutional researchandimplementation priorities to target longterm integration of biologically intensive pest-suppreswonstrategtesand resistance management.The biologically Intensive changesinvolve a commitment to the prmclples of ecological management and more durable production systems,i.e., creating an agricultural production system that utilizes a comprehensive blologlcal systemsapproach to problem-solving in production, pest management, and marketing; emphasizesecologically sound and equitable infrastructures; and places an appropriate economic valuation on the agroecosystem’snatural capital, including predators, parasites,pest-damageinduced-nnmunity, diseasehypersensltlvlty, and so on. More research m the field of ecological economics would also help assess and define the intrinsic value of natural resourcesand dependenceon the underlymg genesthat confer susceptiblegenesm target-pest populations. This will lead to amore thorough and enhghtened economic analysesof agricultural production and better use of biopestlcldes. Somecomparative studiesof blologlcally intensive IPM programs vs standard chemical production have already demonstrated the transaction level economic advantagesin utilizing blopestlcides, yet the intrinsic ecologlcal advantages of these systemshave not yet been addressed. Ultimately, the transition to blologlcally intensive field management, including biopesticides, will be market-driven, with industry, academia, government, growers, environmentalists, and consumers sharing m the process. Growers and biotechnology-based companies will have more incentives to switch to biological products and approaches, as consumer demands and the regulatory environment dictate The US FQPA of 1996 has already been a catalyst for transition in the Umted States. Policymakers and government regulatory agencies can facilitate this process by relaxing market quality standards that currently dictate chemically intensive production. Several private sector companies, including many start-up biotechnology companies, find it more economically feasible to pursue the development of biopestlcides, rather than deal with the regulatory resistance and consumer burdens on registration of conventlonal chemicals Finally, the development of better timing tools (electronic, biologically tlmed, Information delivery systems), critical agroecosystem-based treatment thresholds, and on-farm evaluations are likely to accelerate and facilitate adoption of these tools. References 1. Jacobson,M. (1988) Botanical pesticidespast,present,andfuture, m Znsectmdes ofPlant Orzgzn (Amason,J. T., Philgkne,B J R., and Morand, P , eds.),Amencan Chemistry SocietySymposiumSeries387, pp. I-10

2 Gaugler, R. (1997) AlternatIve paradigms for commerclahzmg biopestlcldes Phytoparasztzca 25, 3. 3 Code of US Federal Register 40 parts 152, 174, and Plant Pesticides Supplemental Notlce 180 4 McClmtock, J. T , Kough, J L , and SJoblad, R D. (1994) Regulatory oversIght of biochemical pestlcldes by the U. S. EnvIronmental Protection Agency health effects conslderatlons. Regul Tox Pharmacol 19, 115-124 5 Rogers, E M (1962) Dzfiszon ofhznovatzons Colher-Macmillan, Toronto, Canada. 6 Remecke, P. (1990) Biological control products. demands of industry for successful development, m Pestzczdes and Alternatzves Innovatzve Chemzcal and Bzologzcal Approaches to Pest Control (Caslda, J. E , ed.), Elsevler, New York, pp 99-108 7 Carlson, G A (1988) Economics of biological control of pests. Am J Alternatzve Agrzc 3, 110-116 8 Hollander, A K., Wood, H A , and Evans, F (1991) Blopestlcldes (workshop report), m Bzotechnology and Sustaznable Agriculture Polzcy Alternatzves National Agriculture Biotechnology Center Report 1 (MacDonald, J F., ed ), National Agriculture Biotechnology Center, Ithaca, NY, pp. 14-20. 9 James, C and Krattiger, A. F. (1996) Global review of the field testing and commercializatlon of transgenic plants, 1986-1995. the first decade of crop blotechnology ISAAA Brzefi No 1, Ithaca, NY. 10 Andrews, R E , Faust, R M , Wablko, H , Raymond, K. C., and Bulla, L A. (1987) The biotechnology of Baczllus thurzngzenszs.CRC Crzt Rev. Bzotechnol 6, 163-232, 11 Relchelderfer, K H ( 1989) Economic aspects of biopestlcldes, m Bzotechnology and Sustaznable Agrzculture Polzcy Alternatives. National Agriculture Blotechnology Center Report 1 (MacDonald, J. F., ed.), Natlonal Agriculture Blotechnology Center, Ithaca, NY, pp. 82-89. 12 Bowles, R. G. and Webster, J. P. G. (1995) Some problems associated with the analysis of the costs and benefits of pesticides. Crop Protect. 14, 593-600 13. Brown, G. C (1997) Simple models of natural enemy action and economic thrcsholds Am Entomol. 43, 117-124. 14 Croft, B. A. (1975) Integrated control of apple mites Mlchlgan State Umverslty Cooperative Extension Services Extension Bulletin E-825 15. Stnckler, K., Cushmg, N , Whalon, M , and Croft, B A (1987) Mite (Acan) species composltlon m Mlchlgan apple orchards. Envzron Entomol 16, 30-36 16. Stickler, K and Whalon, M (1985) Microlepidoptera species composltlon In Mlchlgan apple orchards Environ Entomol 14,486-495 17 Trumble, J. T., Carson, W. G., and Kund, G. S (1997) Economics and envlronmental impact of a sustainable integrated pest management program m celery J Econ. Entomol. 90, 139-146 18 Dahlberg, K A (1993) Government polictes that encourage pestlclde use m the United States, in The Pesticzde Question Envzronment, Economzcs, and Ethics (Pimentel, D. and Lehman, H , eds.), Chapman and Hall, New York, pp 28 l-306 19 Knmsky, S and Wrubel, R P. (1996) Agrzcultural Bzotechnology and the Environment Sczence,Polzcy, and SoczaZIssues Unlverslty of Illmols Press, Urbana and Chlcago. 20 Benbrook, C M., Groth, E , Halloran, J M., Hansen, M K., and Marquardt, S (1996) Pest Management at the Crossroads. Consumers Umon, Yonkers, NY

Index A

Anticarsia gemmatalis, 5 Anticarsia gemmatalis NPV, 3 16 Antifeedant bioassays, 143, 144 Aphids, 32, 165 Apic, 159 Apochemia cineraius, 223 Application equipment biopesticide product development, 515-517 AQlO, 2, 3,97 disease incidence, 86-89 spray applications, 92-95 AQ 10 development, 82-95 field trials, 83-91 assessment, 85-91 Arizona Cotton Research and Protection Council pink bollworm, 39 1 Army worm, 164 Arthropod viruses, 34 1 Article 130 A Europe registration requirements biopesticides, 454,455 ASPIRE, 96 A-terthienyl, 147 Augmentative biocontrol, 37 1 Australia, 159, 160, 218 Autographa californica nucleopolyhedroviruses, 40, 3 16, 322-336 Avermectins, 3 Azadirachta indica, 156 Azadirachtin, 146, 155 adjuvants, 163-l 66 formulation effects, 161-163

A. quisqualis, 82, 83 Abamectin, 3 AcJHE.KK, 347 AcMNPV, 3 16 recombinant lethal times, 342, 342t AcNPV, 40,3 16,322-336 ACRPC pink bollworm, 391 Acyrthosiphon pissum, 162 Adjuvants, 163-l 66 Aedes aegypti, 150 Aedes vexans, 39 A83543 factors, 172 Africa, 52 Agaricus Bosporus, 284 AgNPV, 3 16 Agree, 198 Agriculture alternative production practices, 64 chemical regulations, 6 l-63 research IPM role, 604-606 scouting, 64 threshold use, 64 Agrobacterium radiobacter, 30, 104 Agrobacterium tumefaciens, 30, 2 13 Agrotis ipsilon, 17, 289 Align, 148 Annexes EU directive 91/414EEC, 457t, 458f, 4592 Antibiosis microbial antagonism, 122-123 609

610 mode of action, 160, 161 origin, 156, 157 regulation, 158 Azadirachtin-based insecticides development, 159, 160 Azadirachtin-based pesticides commercialization history, 157-l 59 Azatin, 148, 158, 162 B B. bassiana, 238-245 B. sphaetkus, 35 Bacillus papillae, 14 Bacillus subtilis, 3 1, 104 Bacillus thwit~gietuis, 1, 189-204. See also Bt agricultural applications, 57 analysis, 543-546 crystal proteins, 189, 190 development, 190, 19 1 discovery, 13 efficacy, 193 genetically modified strains, 198, 199 genetic manipulation, 193-198 improving, 19 l-l 93 market, 24 recombinant strains, 199-201 sales, 189 Bacterial biofungicide disease control and yield enhancement, 492 Bacterial biofungicide formulations, 492 Bacterial bioherbicides formulation, 491,492,493t target weed and biocontrol agent, 493t trade name and manufacturers, 493t Bacterial bioinsectjcide formulations, 4921194 Bacterial biopesticide formulations, 488494

Index development, 489-49 1 requirements, 488, 489 storage time, 488 Bacterial insecticides field performance, 35-39 Bactimos, 38, 19 1 Baculoviridae genera, 32 1 Baculoviruses, 5 augmentative control agents, 307-309 cost effectiveness, 302 developing countries, 3 16, 3 17 genetically modified interactions, 346349 insect control agents, 305, 306 example, 306 insect pest control, 30 l-3 18 integrated pest management system, 342 Lepidoptera examples, 32 1 new developments, 39-41 other control agents joint actions, 341-35 1 performance expectations, 303, 304 pest control advantages, 34 1 recombinant, 32 l-336 registered insect pest control, 3 1Ot synthetic jnsecticides interactions, 343-346 types and properties, 304, 305 viral insecticides, 309-3 16 efficacy, 312, 313 vs chemical insecticides cost comparison, 303, 304 Baculovirus field test US EPA approval, 547 Bassi, Aogostino, 13 BCA, 25,26 formulation and application, 5 12-5 14 inactivation, 26

Index

611

induced systemic resistance, 123, 124 microbial antagonism, 120-123 single strain, 129 ultraviolet (UV) radiation, 26 Beauveria bassiana, 4, 5, 27, 32-34

China, 147 discovery, 13 Beauveria brongniartii,

32-34

Beet army worm, 3 14 Beneficial insects Neemix, 166, 167 Beneficial organisms, 58, 59 Betel, 34 Billbugs, 288,289 Bioassay methods Bt preparations, 546 Biochemical active ingredients, 426t428t Biochemical agents analysis and instrumentation, 532 classes, 532 Biochemical pesticides description, 4 16 list, 595, 596 US EPA registration process active ingredient classification, 424,425 classification guidance, 425-429 data requirements, 424-434 FIFRA exemptions, 429,430 indentitylanalysis, 430,43 1 mammalian toxicology, 43 l-433 nontarget organism testing, 433, 434 Biocontrol agents classes, 53 1 Biocontrol products delivery methods plant treatment, 501 seed treatment, 499 soil treatment, 500, 501 formulation technology future needs, 502

Biofungicide development bioassay development, 60 current status and future prospects, 95-97 demonstration program, 8 1, 82 fermentation process selection, 60 field-testing, 8 1 formulation development, 80, 8 1 microorganism screening, 79, 80 protocol design, 82 registration package, 8 1 Biofungicides, 2, 3 biopesticide formulations, 487-502 commercial development, 77-97 criteria, 78, 79 Bioherbicides, 5,6, 359-378 biopesticide formulations, 487-502 hunting and gathering, 361, 362 Bioinsecticides, 3-5 biopesticide formulations, 487-502 dose acquisition cost benefit analysis, 566 delivery vs target, 562-569 tield dosage determination, 563568 future needs, 569 historical perspectives, 544, 545 host biology, 558-561 host susceptibility, 555-558 microbial agent, 561, 562 mortality relationships, 555-557 parameter and constraints, 568t principles, 553-569 relationship formula, 564 replication, 557, 558 target, 555-561 growers, 596, 597 insect target and pathotype, 494t trade name and manufacturers, 494t Biological control areas, 371 microbial associations, 124, 125 prospects and constraints, 126, 127

612 Biological control agents, 23 formulation and application, 5 12-5 14 inactivation, 26 induced systemic resistance, 123, 124 microbial antagonism, 12&l 23 single strain, 129 ultraviolet (UV) radiation, 26 Biological herbicides synergized, 377 Biological insecticides field performance, 32-35 IPM, 32-35 Biological macromolecules analysis, 535-537 Biological pesticides categories, 4 15,416 developmental incentives, 439 Biological weed control, 5, 6 plant pathogens formulation and application, 37 l-378 Bioneem, 158, 160 Biopesticide acceptance factors, 597, 598 Biopesticide analysis sequencing, 537 Biopesticide companies competitiveness, 70, 7 1 Biopesticide conversion, 14 Biopesticide delivery system field trials, 5 19-523 mortality guideline assessment, 523 prefield trials, 5 17-5 19 techniques, 5 12 ULV drift spraying, 520-522 variables, 5 19, 520 vs chemical delivery system, 5 12,5 13 Biopesticide development phases, 5 11f regulation, 65-67 successful conditions, 509, 5 10 Biopesticide efficacy, 546, 547 Biopesticide market future needs, 450,45 1

Index Biopesticide product commercialization industries view future needs, 482,483 Biopesticide product development application equipment, 5 15-5 17 spray parameters, 5 15-5 17 step by step approach, 5 14-523 Biopesticide production developing countries, 47 Biopesticide products advantages, 598 disadvantages, 598 implementation issues, 5988604 Biopesticide registration, 4 15-483 Biopesticide regulation, 67 Biopesticides, l-8 analysis, 529-548 categories, 4 15, 4 16 conventional systems, 598, 599 definition, 473, 5 12 development, 45,46 dose acquisition, 7 EPA definition, 63 EU directive 9 l/4 14EEC active ingredients, 46 1 harmonization, 469-47 1 principles, 46 1, 465 product requirements, 46 I European market, 24 Europe registration requirements, 453-471 field use considerations, 60 1t formulations, 487-502 deployment compatibility, 487,488 principles, 5 17, 5 18 government commodity and conservation programs, 60,6 1 growth rate, 1,2, 15 management protocols, 7 market, 53, 54, 70 market acceptance, 15 market potential, 23 monitoring, 529-548

Index opportumttes, 2 product development Industry’s progress, 476-479 production, 47 protocol and delivery systems, 509-524 registration requirements, 49, 50 regulation, 2 regulatory tmplicattons, 529-548 relatrve factor prtces, 60 shelf ltfe, 7 US EPA definitron, 595 groups, 595 regtstratlon process, 6, 4 16-440 use advantages and disadvantages, 477 Biopesticides Pollution and Prevention Dtvision EPA, 6,63 Btopestictde technology pestictde pohcy, 55-7 1 Blopestrcrde use policres, 63, 64 Bioratronals defimtron, 512 Bioratlonal technologies, 385-404 B~osys, 158, 159 BIO Systems, 32 Blotechnologies future needs, 605, 606 Black cutworms, 17,289 Black vine weevil, 34,284,286 Bluegrass btllbug, 288 Bollgard cotton, 2 17 Botamcal msecticides commercialization barriers, 141 Botamc pesttctdes, 45 BPPD, 6 EPA, 63 Breedmg programs, 59, 60 Brown rot, 30, 3 1 Bt, 35

673

Bt Bt Bt Bt

analysts, 543-546 exotoxm presence, 545, 546 prmciple toxms, 543 regulatory methods, 543-545 Berliner safety evaluations, 53 1 corn, 18, 59 cotton, 18 cost, 60 H- 14 productton guidelines World Health Orgamzatron, 52

Btl larvmde,

39

Btk, 37 Bt kurstakl,

51

Bt preparations quantttation methods, 546 Bt production China, 48,49 quahty control, 49 Bts, 4 Bt subspp rsraelensis (Bti), 37, 38 Bt toxins plant expresston, 16 transgenic plants, 2 1 l-225 transgemc technology, 2 1 l-2 14 Bt transgenic crops EPA registration and reststance management, 224,225 future, 225 Burkholderza cepacla, 104, 105 brological control, 105 ecology, 104 mode of actton, 105 productton and appltcatton, 107, 108 C Cages disadvantages, 5 19 Candtda sake, 30 Capillary zone electrophoresrs polypeptide analysrs, 536 Carbon competltlon microbial antagomsm, 12 1, 122 CASST, 495

index

614 Cat flea, 289 CDA techmques, 5 15, 5 16 Cell-u-wet, 164 Chemrcal mrcrobtal mteracttons statrstrcal methods, 349 Chemtcal apphcatton techmques, 5 13,5 14 Chemtcal delrvery system vs btopesttcrde dellvery system, 5 12,5 13 Chemical herbtctdes integration, 377 Chemrcal msectrcrdes field use constderatrons, 600t Chemical pestrcides, 45 lust, 415 Chemical regulatrons agrrcultural, 61-63 Chemical screenmg, 14 Chma, 238 Beauverla

basslana,

147

Bt productron, 48,49 fungal pathogens, 250 opportunrttes, 53 regtstratron requrrements, 49 Chrnaberry tree, 149 Chmch bug, 13 Chlamydospores, 366 Chorzstoneura

fumtferana,

19 1

Classical brocontrol, 37 I Codlmg moth, 3 15 case study, 393-401 Cold fogging equrpment, 5 17 Coleoptera, 17 Collego, 495 Colletotrtcum orblculare, 123 Colorado beetle, 532 Colorado potato beetle, 147, 22 1 COM (89) 34 plant protectron products Europe, 455,456 Commumty pohcy Europe regtstratron requirements, 455,456

Condor, I98 Conidla, 25, 237-245 Consumer acceptance nematode products, 279 Consumer concern, 59,60 Controlled droplet appbcatron techniques, 5 15, 5 16 Corn Insect control, 16 transgemc, 16 Cotton, 150 case study, 389-393 insect control, 16 Cotton bollworm, 532 Cotton leaves recombinant baculovnus, 326t Councrl directive plant protection products Europe, 455,456 CpGV, 3 15 Crop pests natural enemies, 58, 59 Crop rotatton, 64 Crops planted outlook, 68 Cry3A protem, 198, 199 Cry genes, 193, 194 CryIa Bt toxms, 16 CRYMAX, 15, 199,200 Cryphonectrza parasztlca, 12 1 Cry protems, 190 Cryptic pests entomopathogemc fungr, 245-247 Ctenocephalldes

fells fells, 289

Cucumber, 123 Culex pzplens, 39

Culture techmques microorgarusm recogmtron, 539, 540 Cutlass, 198 Cydla pomonella GV, 3 15 CZE polypeptrde analysrs, 536

lncfex D

DDT, 13, 17 Dehvery system bropesticides, 509-524 chemical vs bropestlclde, 5 12, 5 13 definition, 5 10 Deny, 109 Design, 198 Developmg countrtes, 45-53 baculovrruses, 3 16, 3 17 biopestrcrde productron, 47 opportunitres, 53 DeVme, 494,495 Drabrotica, 16 Dramondback moth example resrstance management, 588 Dipel ESNT, 25 Drptera, 532 Directrve 9 l/4 14EEC Europe regrstratron reqmrements basrcsldescriptron, 456-461 blopestrcrdes, 454, 455 specrfied procedures, 456461 Diseases bropestrcide formulatrons, 487-502 Dormant propagules, 489 Dose acquismon broinsectrctdes fundamental aspects, 553 prmcrples, 553-569 Dose-response analyses bropestrctde product development, 514,515 Doses resrstance management, 582-584 Douglas fir tussock moth, 302, 307-309 Dow Agrochemrcals, 532 DowElanco, 532 Droplet stze spectra measurmg and mterpretmg, 5 16, S17 E

Ecogen, 15 Ecologrcally based management systems, 68, 69

67.5 Ecologrcal restramts weed control, 363, 364 E 1-D Parries, 159 Elcar, 5,3 13,3 14 ELISA, Bt preparations and, 546 Emamectm benzoate, 3,4 Emergency use permit, 6 Emulsifiable concentrate formulatrons, I46 Entomopathogemc fungr, 4, 5 cryptic pests, 245-247 development, 234-237 product stabrlizatron, 237-240 submerged culture, 235 Entomopathogenic nematodes, 5,27 1-29 1 Enzymes, 532 EPA, U S See U.S. EPA EPA approved CellCap products, 478 EPA Pestrcrde Testing Gmdelmes, 109 Eptcoccum mgrum, 30 Eplzootrcs, 544, 545 EU directrve 9 l/4 14EEC bropestrcrdes, 454, 455 US EPA registration process comparison, 470t EUP, 6 Europe formulatron and dehvery, 24-27 future prospects, 41,42 insectrcides, 32-35 mrcrobtal biopestrcides, 23-42 new developments, 29-4 1 plant pathogen control, 27-32 European cockchafers, 252 European corn borer, 34,219,246 European pme sawfly, 3 15 European spruce sawfly nuclear polyhldrosrs vn-us (NPV), 302,306 Europe registration reqmrements bropestrcrdes, 453-471 dtrectrve 91/414EEC, 454,455 US EPA registration process comparison, 470t commumty and pestrcrde pohcy, 455,456

616 F FAIR Act, 68 Fax Act and Food Qualrty Protectron Act, 2 bropestrcrdes, 529, 530 Fall army worm, 201 Federal Food, Drug and Cosmetic Act, 61 biopesttcide registration, 4 15-440 Federal Insectnxde, Fungrcrde, and Rodentictde Act, 61, 65 bropesticide regrstratron, 415-440 Fermentation-derived insect control agents, 17 I-1 85 dtscovery, 17 1 FFDCA, 61 bropestrcrde regrstratton, 4 15-440 Freld crop pests spray, 20-257 Field efficacy recombinant baculovrrus, 329, 330 Field management history, 595 IPM role, 604-606 new technologres, 595-607 Freld trials bropestrcrde delivery system, 5 19523 FIFRA bropesttcrde regrstration, 415-440 bropestrcrdes, 529-530 Fmal Rule for Mrcrobral Pestrctdes, 480 Flea Halt, 289 Floral lures/attractants/repellents target pest, 427t, 428t Florida IPM, 18 Foil, 198 Follar application weed control, 375-377 Fohar fungal brocontrol agents invert emulsions, 495 Fomes annosus, 3 1

Index Food Quality Protection Act, 6 1, 62 Food Quahty Protection Act 1996 pesticrde minor use definitron, 445 Foray 48B, 36 Forest tent caterpillars, 162, 223 Formulatron biologtcal agents, 5 12-5 14 weed control, 373-375 plant pathogens, 37 l-378 FQPA, 2 bropestrcrdes, 529, 530 France mosquito control, 38 Frankluuella occrdentalrs, 34 Frmt, pome case study, 393-401 Fungal btofungrcrdes brocontrol agent, 497t delivery, 497t formulatrons, 496,497t trade name and manufacturers, 497t Fungal broherbrcides formulation, 493t target weed and brocontrol agent, 493t trade name and manufacturers, 493t Fungal bromsecttcide formulatrons, 496-498 Fungal bropesticrde formulatron, 494498 Fungal pathogens granular formulations, 245-247 inoculative augmentation, 252 oil formulations, 240-245 Fungi entomopathogemc, 4, 5 Fungrcrdes target pest, 428t Fungus formulatron, 240-245 Fungus gnats, 286 Fusarwm oxysporum, 30, 3 1, 124, 125 Fusarlum prollforatum,

96

index Fusarmm welts suppressive soils, 118-l 20 G Gas chromatography biopesticide analysis, 53 1 Genetically modtfied baculoviruses interactions, 346-349 Genetic engineering, 15 Genetic techniques microorganism recognition, 54 l-543 Germany mosquito control, 38, 39 Gmkgohdes, 147 Ghocladium, 30 Ghocladmm virens, 105-107 apphcation, 107-l 09 biological control, 107 ecology, 105, 106 formulation development, 108 mode of action, 106 production, 107, 108 registratton, 109, 110 GhoMix, 30 Ghotoxm, 106 Globus etunicatum, 127 Government commodity and conservation programs btopesttcides, 60, 6 1 Granular formulattons fungal pathogens, 245-247 Granuloviruses, 32 1 Grapeleaf skeletonizer, 302, 307-309 Grasshoppers, 247-250 Gray mold, 96 Greenhouse assays recombinant baculovnus, 328, 329 Greenhouse whitefly, 164 Green products consumer demand, 69-7 1 Growers biomsecticldes, 596, 597 transgemc plants, 596, 597

617 Gusano, 3 16 Gypcheck, 3 14,3 15 Gypsy moth, 37, 191, 223, 3 14, 3 15 H H vzrescens, 17, 178, 182 Hellcoverpa armlgera, 39 Helxoverpa zea, 5 Hellothls armrgera, 5 1, 2 18

Heliothts NPV, 3 13, 3 14 Hekothzs vwescens,

4

Herbicides synergy mycoherbicides, 365 Heterorhabditis, 27 l-276 High doses resistance management, 582-584 High pressure liquid chromatography biopesticide analysis, 53 1 Honduras, IPM, 18 Hoplochelus

marglnalw,

34

Hormones, 532 Host-plant resistance, 5, 60 Host range genetic manipulatton mycoherbmtdes, 365, 366 HPLC biopesticlde analysts, 53 1 Hydraultc sprayers vs rotary atomizers, 5 16 Hydraulic spraying disadvantages, 5 15 Hydroponics, 129 Hyphomycetes solid substrate culture, 235 submerged culture, 236 HzNPV, 313,314 I IACR-Rothamsted, ICAMA, 49,50 ICIPE, 52 IGR, 160, 161

35

618 Immunological methods microorgamsm recogmtion, 540, 541 India, 50, 5 1, 156 azadirachtm-based Insectlades, 1.59, 160 Indonesia, 159, 160 Induced systemic resistance, 123, 124 Insect control pheromones, 385-404 case studies, 389-401 future needs, 401-404 Insectlclde production bloactlvlty screening, 143, 144 choice of plants, 140, 141 collection sites, 142 extraction, 144, 145 formulation, 146, 147 standardization, 145, 146 tissue harvested, 141, 142 Insecticides compatlblllty with IPM, 148 development, 140 Insect pest control baculovlruses, 30 l-3 18 reglstered baculovlrus, 3 1Ot Insect pests blopesticide formulations, 487-502 Insects bacterial pathogens, 13 Institute for the Control of Agrochemlcals, M1nW-y of Agriculture, 49, 50 Integrated pest management, 17,438, 439,578 See also IPM baculovirus, 342 International Center of Insect Physiology and Ecology, 52 International orgamzatlons, 52, 53 Interrupt, 289 Invert emulsions follar fungal blocontrol agents, 495 IPM, 17,438,439, 578 baculovlrus, 342 blologlcal insecticides, 32-35

Index Florida, 18 Honduras, 18 Mexico, 18 Nicaragua, 18 North America, 17, 18 tomatoes, 18 IPM approach one crop system, 604 IR-4 Blopestlclde Grants Program funded proposals, 44&448 IR-4 blopestlclde program admimstratlon, 443,444 food crop successes, 445t minor crops, 443-45 1 Iron competition, 122 J

Japanese beetle, 14 K

Koppert, 32 L

Laboratory assays blopestlclde product development, 514,515 recombinant baculovlrus, 325-328 Lawns, 158 LdMNPV Neem extract, 344 LdNPV, 314,315 Lepldoptera, 16, 147, 177, 217, 532 baculovlrus examples, 32 1 Lepldopteran pheromones isolation and ldentlficatlon, 535 regulations, 534 volatihzatlon, 534 Leptmotarsa decemllneata, 22 1 Life stage targeting resistance management, 584 Lilly Research Laboratories, 172 Llmmlods, 162, 163 Llmonin, 147, 150

Index Liquid fermentation, 108 Locusts Low doses resistance management, 584 Low persistence formulations resistance management, 58 1 LUBILOSA project, 5 12 Lupins, 150 Lymantrla dzspar, 5, 37, 191 Lyman tria dzspar NPV, 3 14-3 15 Lymantrla monacha, 36, 192 M

Macromolecules characterization, 536 Malacosoma dzsstrza, 162, 223 MALDI-TOF technique btologtcal macromolecule analysts, 536 Mamestra brasslcae NPV, 39 Mamestrin, 39 Mammalian toxicology data US EPA registration process btochemmal pestictdes, 43 It mtcrobtal pesticides, 4 19t Management protocols biopesticides, 487-607 MargoBiocontrols, 159 Margosan-0, 148, 157, 158 Marigolds, 150 Mass spectrometry biopesticide analysis, 53 1 Mating dtsruption products, 386t Matrix assisted, laser desorptton ionization time of flight btological macromolecule analysis, 536 Me&a volkensl, 150 Melolontha melolontha, 34 Metabolism microorganism recognitton, 540 Metarhzium anlsopllae, 13, 32, 34 Metarhlzrumflavovtrzde, 25, 34

619 Methomyl efficacy, 35Of Mexico IPM, 18 MFP, 15 Microbial chemical mteracttons statistical methods, 349 Mtcrobial antagomsm, 120-I 23 anttbtosis, 12, 123 btological control agents, 120-123 nutrtent competition, 12 1, 122 parasitism, 120, 12 1 Microbial assoctattons biological control, 124, 125 mtcrobtal product development, 128, 129 Microbial biomsecttctdes prmctpal features, 553 Microbial biopesttcides, 23-42 formulation and delivery, 2427 future prospects, 4 1, 42 msecttcides, 32-35 new developments, 294 1 plant pathogen control, 27-32 Mtcrobtal fermentation products, 533 Microbial msecttcides distribution, 27 Microbial Joint action, 117-l 30 btological control agents mode of action, 120-l 24 induced systemic resistance, 123, 124 mmrobtal antagomsm, 120-l 23 microbial associations, 124-l 26 biological control, 124, 125 plant growth promotion, 125, 126 microbtal product development, 128-130 microorganism compattbiltty, 127, 128 prospects and constraints, 126, 127 suppressive ~011s 118-l 20

620 Mtcrobtal pest control agents tier 1 toxicology requirements, 449t Microbtal pesticides descrtptlon, 4 16 ltst, 596 pheromones, 56, 57 US EPA registration process data requirements, 4 16-424 tdenttty/analysis, 4 17 manufacturmg process descriptton, 418 nontarget organism data requirements, 42 l-424 toxtctty testmg, 4 18-42 1 Microbial weed control htstory, 359-361 status, 359 Mtcrobtological pest control agents analysts, 53 1 MicroGermm, 32 Microorganism compattbility, 127, 128 Mtcroorgamsms, 538-543 detection methods, 539 genetically engmeered, 538 monitoring methodologies, 538, 539 Minor crops IR-4 biopesttctde program, 44345 1 Mint flea beetle, 287 Mint root borer, 287 Mole crickets, 288 Monhua fructicola, 3 1 Mondinia laxa, 30, 3 1 Momtormg btopesttcides, 529-548 Monsanto transgemc plant regulations, 530, 53 1 Mosquito control Europe, 38, 39 MPCAs analysis, 53 1 M-PEde, 165 M-Trak, 15 Mycogen, 15

index Mycoherbtcides, 359-367 barriers, 363, 364 containment, 366 ecologmal restraints, 363, 364 efficacy improvement, 365,366 enhancement, 364-366 formulations, 494-496 genettc mampulation virulence and host range, 365, 366 registered, 362, 363 synergy herbicides, 365 Mycomsectictdes, 4, 5, 223-259 agrochemlcal mtegration, 256, 257 commerctahzatton, 257 development, 234-240 field performance, 32-35 fungus formulation, 240-245 future, 258 inundattve applmations, 252, 253 product stability, 253 spray apphcatton, 253-256 use and delivery, 245-247 UVL spray applications, 247-250 Mycorrhizas, 125, 126 Mycorrhization helper bacteria, 126 Mycostop, 492 Mycotrol, 252-257 agrochemtcal integration, 256, 257 commerctahzation, 257 product stabtlity, 253 spray application, 253-256 N N tabacum, 214

National Agricultural Science and Technology Institute (NASTI), 5 1 Nattonal Organic Standards Board, 70 National Research Council, 69 Natural baculovtruses insect pest control, 301-3 18 Natural insect regulators target pest, 428t

Index Natural plant protection, 34 Natural plant regulators, 532 Naturalyte, 532 Neem extracts, 533 LdMNPV, 344 Neem, 139-151 applications, 147 bioactivlty screening, 143, 144 collection sues, 142 commercral experience, 155-I 68 compatibility with IPM, 148 extraction, 144, 145 formulation, 146, 147 future trends, 150, 15 1 materials, 140-142 methods, 143-148 productton, 156 recent products, 140-148 seeds, 141, 156 standardtzatton, 145, 146 trees, 156 Neemazad, 158, 160 Neemazal, 148, 159 Neem-based insectrcides regulatron, 158 Neemix, 160, 162, 164 beneficial insects, 166, 167 Neemix 4.5, 148, 158 Nematicides target pest, 428t Nematode products consumer acceptance, 279 Nematodes applicatron, 28 l-284 biology, 27 l-276 bulk storage, 257 emus, 287 entomopathogenic, 5 field efficacy, 284-289 foliar application, 283 formulation, 257-259 future, 290, 291

621 glasshouse crops, 286 host range, 274 Insect traps, 284 inundate biological control, 28 I, 282

mass production, 276, 277 mmt and berries, 287, 288 mushrooms, 284-286 parasitic cycle, 27 1, 272 pet/vet, 289 plan propagation application, 283 product quality, 279-28 1 soil application, 282, 283 strain discovery, 274 trap crop applicatron, 283 turf, 288-289 Neodiprlon sertrfer NPV, 3 15 Newleaf, 22 1 Nicaragua IPM, 18 Nicottana gosset, 149 Nitrogen-9nO-fixing bacteria, 126 Noctuid moths, 3 13, 3 14 Nonproteinaceous pesticides, 436 Nontarget organism data requirements US EPA registration process biochemmal pesticides, 433,434 microbial pesticides, 422t transgemc plant pesticides, 435t North America biopestrcrde converston, 14, 15 future, 19-20 genetic engineering, 15 historical trends, 13, 14 IPM implementation, 17, 18 toxin expression, 16, 17 NsNPV, 3 15 Nuclear polyhedrosls virus, 39, 321 See also NPV European spruce sawfly, 302, 306 Nun moth, 36, 37, 191 Nutrient competition microbial antagonism, 12 1, 122

622

Index

0 Office of Pesticide Program, 4 16 011 drlutents, 242, 243 Oil formulattons fungal pathogens, 240-245 OPP, 416 Orgamc Foods Productton Act, 70 Oryza satwa, 2 13 Ostrlnla furnacalrs, 219 Ostnnla nubdulls, 34, 2 19 Otlorhynchus sulcatus, 34

P Paecdomycesfumosooseus,

32

Pakistan, 5 1 Parasittsm, 120, 12 1 Parker Valley Program pmk bollworm, 39 l-393 Pathogen biology weed control, 372, 373 Pathogen marking mycoherbtctdes, 364 Pawpaw tree, 149 Pea aphid, 162 Penicillium

oxalicum, 30

Pepper weevtl, 164 Pesttctdal active ingredients categortes, 436 Pesttcrde analysis objectives, 529 classes, 4 15 defimtton, 4 15 minor use defimtton, 445 mixtures resistance management, 586, 587 rotation resistance management, 585, 586 safety, 19 Pesticide pohcy btopestictde technology, 55-7 1 Europe regtstratton requirements, 455,456

Pheromones, 6 analysts, 533-535 history, 385-389 msect control, 385404 case studtes, 389-40 1 future needs, 401-404 mlcroblal pesttcldes, 56, 57 purposes, 533,534 target pest, 426t428t USEPA regtstered, 386t Phlebtopsts glgantea, 3 1 Phthormmaea operculella Zeller, 22 1 Pmk bollworm Arizona Cotton Research and Protectton Counctl, 39 1 case study, 389-393 Parker Valley Program, 39 l-393 Plant btology weed control, 372 Plant expresston Bt toxm, 16 Plant growth promotton mtcrobtal assocratlons, 125, 126 Plant growth regulators target pest, 426t-128t Planthoppers, 147 Plant-parasmc nematodes suppresston, 289,290 Plant pathogemc fungt blologtcal control, 27-32 Plant pathogens blologtcal weed control formulation and apphcatton, 37 l378

weed control research needs, 377, 378 Plant protectton products Europe proposal, 455,456 Plutella xylostella,

PMD, 82-95 Poland softwood, 36

39, 52

Index Polypepttde analysts CZE, 536 Pome frurt case study, 393-40 1 Population growth, 19 Potato tuber moth, 22 1 Prefield trials btopesttcide delivery system, 5 17519 Proactrve plans resistance management, 578 Proteinaceous pesttctdes, 436 Protein method Bt preparatrons, 546 Pseudomonasfluorescens, 30, 127, 129 Pyrethrum, 147

Q Qualny control Bt production China, 49

R Rangeland pests UVL spray, 247-250 Raven Biomsecticide, 199 Reactive plans resistance management, 578 Recombinant AcMNPV lethal times, 342, 342t Recombinant baculovnus, 32 l-336 biological selectivity, 330-332 deployment strategtes, 334-336 envnonmental fate, 332-334 gene deletron, 322 gene insertron, 323-325 genetic fitness, 332 msectrcldal actrvny, 325-330 molecular desrgn, 334336 safety, 330-332 vs wild-type, 332 Reduced-Risk Pesticide Imtrattve, 62 Reduced-risk pestlctdes US EPA, 438

623 Reduced-risk program purpose US EPA, 438,439 Refuges resistance management, 579-58 1 Regtonal Network on Pesticides for Asia and the Pacific, 50 Registered baculovuus insect pest control, 3 1Ot Regrstratron reqmrements, 49, 50 resistance management, 577, 578 Regulations agrmultural chemical, 6 l-63 biopesticide, 67 Regulatory rmplrcatrons biopesticides, 529-548 Relatrve factor prices biopesticides, 60 RENPAP, 50 Research needs plant pathogens weed control, 377, 378 Reststance management diamondback moth example, 588 tmplementatton, 588, 589 mode of actron, 567-576 regrstration requnements, 577, 578 resistance momtormg, 587, 588 strategies, 575-589 tactics, 578-587 targeted pests, 576, 577 Resrstance morutormg reststance management, 587, 588 RH-9999, 162 Rice, 213 Rice leaffolder, 222 Rotary atomrzers bropestrcrde product development, 515,516 advantages, 5 16 vs hydraulic sprayers, 5 16 Rotenone, 147

624

Index

S S carpocapsae, 289 S glaserz, 276 Saccharopolyspora spznosa, 172 Scaptertscus borelltt, 288 Scaptertscus rtobravts, 288 Scapteriscus vzctnus, 288

Sclartd flies, 284 Seedling disease brocontrol agents application, 109 compattbtltty with chemical pesticides, 110 current status and future prospects, 111 formulatron development, 108, 109 hqutd fermentation, 108 registration, 109, 110 brologtcal control, 103-l 11 Burkholderta cepacta, 104, 105 Gtlocladtum wrens, 105-107 Semtochemtcals, 532 analysis, 533-535 defimtion, 533 SeMNPV efficacy, 350f SeNPV, 3 14 Stlkworm infectious disease, 13 Site-specific recombmatton system, 194, 195 Skeetal, 19 1 Slow acting pesttctdes mortality gutdelme assessment, 523 Softwood, m Poland, 36 SotlGard, 109 Soursop frmt, 149 South Korea, 5 1, 52 Sphenophorus

parvulus,

Spmosads, 4, 532 Spinosyns, 4 chemistry, 173-177 discovery, 172

288

envtronmental and toxicological profile, 181, 182 insect spectrum, 177-180 mode of action, 180, 181 resistance, 182-l 84 Spinosyn structure-activity relattonshtps, 178-l 80 Spodex, 3 14 Spodoptera extgua NPV, 3 14 Spodoptera frugtperda, 20 1 Spodoptera ltttoralu, 2 18 Spodoptera lttura, 143 Sportdesmrum sclerottvorum,

12 1

Spray field crop pests, 250-257 parameters btopesttctde product development, 515-517 Spray coverage improvement reststance management, 584, 585 Spray Drift Task Force, 5 16 Spruce budworm, 19 1 Steinernema, 27 l-276 Step by step approach biopestictde product development, 514-523 Streptomyces grrseovtrtdts, 104 Striped stem borer, 222 Stylet 011, 165 Sugar cane white grub, 34 Suppressive soils mtcrobtal Joint action, 118-l 20 Surfactants, 495 Synthettc msecttctdes weld-type baculovirus interactions, 343-346 Synthetic pesticides history, 599 T

TalaromycesJavus, 110 Target pest pheromones and plant growth regulators, 426t-428t

Index Teknar, 19 1 Termmites, 532 Tetranychus urtlcae, 180 Thailand, 50, 5 I Thermal fogging equipment, 5 17 ThermoTrilogy, 159 Tier 1 toxtcology reqmrements microbial pest control agents, 449t Tobacco budworm, 177, 178,215,532 Tobacco cutworm, 143 Tobacco hornworm, 2 14 Tomatoes, IPM and, 18 Tomato wilt, 3 1 Totipotency, 2 11 Toxicology requirements, tier 1 microbial pest control agents, 449t Transgenic corn, 2 19-22 1 future, 220, 22 1 history, 2 19 registration and commercialization, 219,220 Transgenic cotton, 2 14-2 19 field tests, 2 15, 2 16 future, 218, 219 history, 2 14, 2 15 registration and commercialrzatton, 216218 Transgemc eggplant, 22 1, 222 Transgenic plant pesticides definition, 434 description, 4 16 US EPA registration process, 434438 active ingredients, 436 characterization data/mformatton, 436,437 nontarget organism data requirements, 435t Transgemc plants Baczllus thurwg?enszs (Bt) toxins, 2 1 l-225 dose acquisition, 569 growers, 596,597 regulations, 530, 53 1

625 Transgenic Transgemc Transgemc Transgenic Transgenic

potato, 22 1, 222 rice, 222 soybeans, 222,223 tomato, 222 trees, 223 Trlaleurodes vaporarlorum, 164 Trichoderma harzianum, 30, 127 Twhoderma

vwzde, 30

Trtchodex, 96, 97 Turfplex, 158 Two-spotted spider mites, 180

U ULV application hydraulic sprayers, 5 15, 5 16 UNIDO, 52-54 US EPA biopesttcrde reglstratton, 4 15-440 blopesttctdes, 529, 530 US EPA registration process biopesticides, 4 15-440 Europe registration requirements comparison, 470t mdustries view on EPA role, 479, 480 industry view and approach, 473-

483 EU directive 91/414EEC comparison, 470t USDA industries view on role biopesttcides, 48 1,482 USDA Forest Service, 69 UVL spray rangeland pests, 247-250

V VA-mycorrhiza, 126 VA mycorrhizal fungi, 127 Vectobac, 38, 191 Vector MC, 288 Velvetbean caterptllar, 5, 3 16 Verticilltum

Vietnam, 52

lecanu, 32

626 Vtp msecttctdal proteins, 225 Vu-al msectlctdes baculovuus, 309-3 16 drsadvantages, 349-35 1 Vtrulence btopestnxdes defimtion, 5 12 genetic mampulatton mycoherbtctdes, 365, 366 Virulence range recombmant baculovirus, 33 1,332 Vnus chemical mteracttons commercial potential, 349-351 Virus formulation, 498, 499 development, 498,499 requirements, 498 W Weed control barriers, 363, 364 btologtcal, 5, 6

Index btologtcal vs chemtcal, 363 history, 359-361 plant pathogens formulanon and apphcatron, 371-378 research needs, 377,378 Weeds btopesttctde formulatrons, 487-502 Weevrls, 287, 288 Western flower thrtps, 34 Whnefbes, 32 Whole cell analysts mtcroorgamsm recognition, 540 Wild-type baculovnuses dtsadvantages, 40 synthetic msecttctdes mteracttons, 343-346 World Health Orgamzatton Bt H- 14 productron gutdelmes, 52 Y Yellow fever mosquito, 150