Miguel Angel Altieri, PhD Clara Ines Nicholls, PhD
Biodiversity and an d Pes estt Ma Mana nage geme ment nt in Agr Agroec oecosy osyste stems ms Second Sec ond Edi Editio tion n
Pre-publication REVIEWS, COMMENTARIES, EVALUATIONS . . .
“T
his book is an important contribution to the literature on ecologically based pest management. In addition tio n to re revie viewin wing g re relev levant ant asp aspect ectss of ecological theory and the broader agroecolog eco logical ical con contex textt of pes pestt man manage agemen ment, t, it presents numerous data sets and case studie stu diess fro from m bot both h tem temper perate ate and tr tropi opical cal farming farmi ng syste systems. ms. Biodi Biodiversit versityy and Pest Management in Agroecosystems reflects the authors’ many years of experience, particularl ticu larly y in the thecha chapte pters rs on inse insect ct man manage age-mentin multip multiple le cropp cropping ing syste systems, ms, ins insec ectt ecolog eco logy y in or orcha chard rdss con contai tainin ning g cov cover er crops, and non-crop vegetation effects on insect popul po pulat ation ionss in cr crop op fie fields lds.. Many new references have been added to this edition. This second edition also includes a new chapte cha pterr on ins insect ect dyn dynami amics cs in agr agrofo ofore reststry systems. Anyone interested in rela-
tionships between insect pest management,diversifie agement, diversified d cropping systems systems,, and field boundary vegetation should read this book.” Matt Liebman, PhD Professor of Agronomy, Iowa State University
“T
his boo book k is ess essent ential ial re readi ading ng for anyone interested in developing tru truly ly logi logical cal pes pestt man manage agemen mentt practices in any country. country. The authors go wa way y be beyo yond nd IP IPM M an and d in inpu putt su subs bsti ti-tution approaches to show how a well-designed agroecosystem can actually sponsor its own pest management.” Peter Rosset, PhD Co-Director,, Food First/Institute for Food Co-Director and Development Policy
More pre-publication REVIEWS, COMMENTARIES, EVALUATIONS . . .
“A
ltieri and Nicholls, ltieri Nicholls, in thi thiss completely ple telyre rewor worked ked and upda updated ted edition,make ti on,make oneof th thee be best st at atte tempt mptss to da date te to pr prese esent nt pes pestt man manage agemen mentt with the whole-systems approach—biodiversity and habitat management in an agroecosystems framework. This is a book about how to use eco ecolog logica icall kno knowle wledge dge to des design ign agr agrooecosystems with built-in pest managementt pr men proce ocesse ssess and and,, as a re resul sult, t, re reduc ducee or eliminate the use of synthetic chemical pesticid pest icides. es. The Theaut authorsbuild horsbuild on the thegro growwing body of knowledge about the vital importance of biodiversity for maintaining balanced, long-term ecosystem function, tio n, and its imp import ortant ant ro role le in agr agricul icultur ture. e. This book will appeal to researchers developi vel oping ng agr agroec oecolog ological ically ly bas based ed pes pestt management strategies and to practitioners looking for possible options to apply on their own farms. The ver very y ext extens ensive ive and curr current ent re revie view w of th thee lit liter erat atur uree ma make kess th this is boo book k a re remar markkably use useful ful cont contrib ributio ution n to agr agroec oecologic ological al knowledge. Most chapters complement this review with several very in-depth
case studies. The first three chapters provide a firm grounding for pest managemen manag ementt in ecolog ecological ical theor theory y. The nex nextt fiv fivee cha chapte pters rs ver very y ef effec fec-tively tivel y pre present sent examp examples les that apply the the theory ory in agr agroec oecosys osystem temss fro from m aroun ar ound d th thee wo world rld,, an and d in sy syst stem emss th that at range from multiple cropping, to orchard cha rdss wit with h cov cover er cr crops ops,, to agr agrooforestry systems with the integration of trees. The review of the influence of surrounding habitats on pest management very appropriately places the process off the farm into the landscape level. The concluding two chapterss of the ter theboo book k exp explor loree howbiodiv howbiodivers ersity ity and habitat management must be part of the transition to sustainability in agriculture, presenting exciting and challenging goals for the future.” Stephen R. Gliessman, PhD Alfred Heller Professor of Agroecology, University of California, Santa Cruz
Food Products Press® An Imprint of The Haworth Press, Inc. New York • London • Oxford
NOTES FOR PROFESSIONAL LIBRARIANS AND LIBRARY USERS
This is an or This origi igina nall bo book ok ti titl tlee pub publi lish shed ed by Fo Food od Pr Prod oduc ucts ts Pr Pres ess, s, an im im-print of The Haworth Press, Inc. Unless otherwise noted in specific chapters with attribution, materials in this book have not been previously published elsewhere in any format or language. CONSERVATION CONSERV ATION AND PRESERVA PRESERVATION TION NOTES
All books published by The Haworth Press, Inc. and its imprints are prin pr inte ted d on ce cert rtif ifie ied d pH ne neutr utral al,, ac acid id fr free ee bo book ok gr grad adee pa pape perr. Th This is pa pape perr meets the minimum requirements of American National Standard for Informati Infor mation on Scien Sciencesces-Perm Permanen anence ce of Pap Paper er for Print Printed ed Mate Material, rial, ANSI Z39.48-1984.
Biodiversity and Pest Management in Agroecosystem Agroecosystemss Second Edition
®
FOOD PRODUCTS PRESS Crop Science Amarjit S. Basra, PhD Senior Editor
Mineral Nutrition of Crops: Fundamental Mechanisms and Implications by Zdenko Rengel Conservation Tillage in U.S. Agriculture: Environmental, Economic, and Policy Issues by Noel D. Uri Cotton Fibers: Developmental Biology Biology,, Quality Improvement, and Textile Processing edited by Amarjit S. Basra Heterosis and Hybrid Seed Production in Agronomic Crops edited by Amarjit S. Basra P. R. Intensive Cropping: Efficient Use of Water, Nutrients, and Tillage by S. S. Prihar, P. Gajri, D. K. Benbi, and V. K. Arora Physiological Bases for Maize Improvement edited by María E. Otegui and Gustavo A. Slafer Plant Growth Regulators in Agriculture and Horticulture: Their Role and Commercial Uses edited by Amarjit S. Basra Crop Responses and Adaptations to Temperature Temperature Stress edited by Amarjit S. Basra Plant Viruses As Molecular Pathogens by Jawaid A. Khan and Jeanne Dijkstra In Vitro Plant Breeding by Acram Taji, Prakash P. Kumar, and Prakash Lakshmanan Crop Improvement: Challenges in the Twenty-First Twenty-First Century edited by Manjit S. Kang Barley Science: Recent Advances from Molecular Biology to Agronomy of Yield Yield and Quality edited by Gustavo A. Slafer, José Luis Molina-Cano, Roxana Savin, José Luis Araus, and Ignacio Romagosa
P. R. Gajri, V. K. Arora, and S. S. Prihar Tillage Till age for Sustainable Cropping by P. Bacterial Disease Resistance in Plants: Molecular Biology and Biotechnological Applications by P. Vidhyasekaran Handbook of Formulas and Software for Plant Geneticists and Breeders edited by Manjit S. Kang Postharvest Po stharvest Oxidative Stress in Horticultural Crops edited by D. M. Hodges Schlegel el Encyclopedic Dictionary of Plant Breeding and Related Subjects by Rolf H. G. Schleg Handbook of Processes and Modeling in the Soil-Plant System edited by D. K. Benbi and R. Nieder The Lowland Maya Area: Three Millennia at the Human-Wildland Human-Wildland Interface edited by A. Gómez-Pompa, M. F. Allen, S. Fedick, and J. J. Jiménez-Osornio Biodiversity and Pest Management in Agroecosystems, Second Edition by Miguel A. Altieri and Clara I. Nicholls Plant-Derived Antimycotics: Current Trends and Future Prospects edited by Mahendra Rai and Donatella Mares Concise Encyclopedia of Temperate Tree Fruit edited by Tara Auxt Baugher and Suman Singha
Wojtkowski Landscape Agroecology by Paul A. Wojtkowski Concise Encyclopedia of Plant Pathology Pathology by P. Vidhyasekaran
Biodiversity and Pest Management in Agroecosys Agroecosystems tems Second Edition
Miguel Angel Altieri, PhD Clara Ines Nicholls, PhD
Food Products Press® An Imprint of The Haworth Press, Inc. New York • London • Oxford
Published by Food Products Press®, an imprint of The Haworth Press, Inc., 10 Alice Street, Binghamton, NY 13904-1580. © 2004 by The Haworth Press, Inc. All rights reserved. No part of this work may be reproduced or utilized util ized in any form or by any mean means, s, ele electron ctronic ic or mec mechanic hanical, al, incl including uding photo photocopyi copying, ng, micr microfilm ofilm,, and recording, or by any information storage and retrieval system, without permission in writing from the publisher. Printed in the United States of America. Cover design by Brooke R. Stiles. Cover photos of vineyards by Miguel Altieri. Library of Congress Cataloging-in-Publication Cataloging-in-Publication Data
Altieri, Miguel A. Biodiv Bio divers ersityand ityand pes pestt man manage agemen mentt in agr agroec oecosy osyste stems ms / Mig MiguelAngel uelAngel Alt Altier ieri, i, Cla Clara ra I. Nic Nichol holls. ls.— — 2nd ed. p. cm. Includes bibliographica bibliographicall references (p. ) and index. ISBN 1-56022-922-5 (hardcover : alk. paper)—ISBN 1-56022-923-3 (softcover : alk. paper) 1. Insect pests—Biological control. 2. Insects—Ecology. 3. Biological diversity. 4. Agricultural ecology. I. Nicholls, Clara I. II. Title. SB933.3 .A38 2003 632'.7—dc21 2002014860
CONTENTS Preface to the Second Edition
ix
Acknowledgments
xi
Introduction
1
Chapter 1. The Ecological Role of Biodiversity in Agriculture
3
Traditional Traditio nal Agro Agroecos ecosystem ystemss As As Models Models of Biodi Biodiver verse se Farm Farmss 6 The Ecological Role of Biodiversity 9 The Nature of Biodiversity in Agroecosystems 10 Chapter 2. Agroecology and Pest Management
The Nature of Agricultural Habitats and Its Relation to Pest Buildup Crop Diversification and Biological Control Chapter 3. Plant Diversity and Insect Stability in Agroecosystems
Ecological Theory Theory Dilemmas Chapter 4. Insect Manipulation Through Weed Management
17
17 23 29
29 39 47
Wee eeds ds As So Sour urce cess of In Inse sect ct Pe Pest stss in Ag Agro roec ecos osys ystem temss Thee Ro Th Role le of Wee eeds ds in th thee Ec Ecol olog ogy y of Na Natu tura rall En Enem emie iess Inse In sect ct Dy Dyna nami mics cs in Wee eedd-Di Div ver ersi siffie ied d Cr Crop op Sy Syst stem emss Isol Is olat atin ing g th thee Ec Ecol olog ogic ical al Ef Effe fect ctss of Wee eed d Di Div ver ersi sity ty Crop-Weed Management Considerations
48 49 55 67 68
Chapter 5. Insect Management in Multiple-Cropping Systems
77
Patterns of Insect Abundance in Polycultures Herbivore Trends in Polycultures
78 85
Case Study 1: Maize Inte terrcrops and Pest Att ttaack Case Ca se St Stud udy y 2: Ca Cass ssaava In Inte terc rcro rops ps an and d Pe Pest st In Inci cide denc ncee Case Study 3: Reducing Stemborers in Africa Living Mulches: A Special Type of Intercrop Metho Met hodo dolog logies ies to Stu Study dy Ins Insect ect Dyn Dynami amics cs in Pol Polycu ycultu ltures res Management Considerations Chapter 6. Insect Ecology in Orchards Under Cover-Crop Management
Select Sele ctin ing g an and d Ma Mana nagi ging ng Co Cov ver Cr Crop opss in Or Orch char ards ds Case Study 1: Apple Orchards in California Case Study 2: Pecan Orchards in Georgia Case Ca se Stu tud dy 3: Sum umme merr Co Cov ver Cr Cro ops in Vin ineeyar ard ds Chapter 7. The Influence of Adjacent Habitats on Insect Populations in Crop Fields
91 93 95 96 99 107 111
116 116 118 121 123 127
Crop Edges and Insect Pests Field Boundaries and Natural Enemies Designing and Managing Borders Case Study 1: Exchange of Arthropods at the Interface of Apple Orchards and Adjacent Woodlands Manipulating Crop-Field Border Vegetation Case Ca se Stu tud dy 2: Bi Bio olo log gic ical al Co Corr rrid ido ors in Vin ineeyar ard ds Casee Stu Cas Study dy 3: Str Strip ip Man Manage agemen mentt to Aug Augmen mentt Pre Predat dator orss
143 148 151 154
Chapter 8. The Dynamics of Insect Pests in Agroforestry Systems
161
The Ef The Effe fect ctss of Tre rees es in Ag Agro rofo fore rest stry ry on In Inse sect ct Pe Pest stss Designing Natural Successional Analog Agroforestry Systems The Need for Further Research Chapter 9. Designing Pest-Stable Vegetationally Vegetationally Diverse Agroecosystems
Monocultures and the Failure of Conventional Pest-Control Approaches Toward Sustainable Agriculture Requirements of Sustainable Agroecosystems Designing Healthy Agroecosystems
129 131 138
162 16 2 166 170 171
171 174 175 176
Healthy Soils—Healthy Plants Restoring Diversity in Agricultural Systems Enhancing Surrounding Biodiversity Case Study 1: Diver Diversification sification of an Onion Agroecosystem in Michigan Case Study 2: A Div Diversified ersified Small Farming System in Chile
179 183 187 188 191
Conclusion
197
References
203
Index
225
ABOUT THE AUTHORS
M. A. Altieri, PhD, has been a Professor of Agroecology at UC
Berkeley for more than two decades. He is recognized as a pioneer in developing dev eloping ecologically based pest management systems, especially through the diversification of agroecosystems. He has published more than 10 books, including Agroecology: The Science of Sustaini n various journals. able Agriculture, and more than 200 articles in C. I. Ni Nich chol olls ls,, Ph PhD, D, is a Research Fellow in the Division of Insect
Biology at UC Berkeley. She is an expert in biological control and has conducted considerable research on habitat management strategies to enhance beneficial insects in agricultural systems. She is the author of numerous articles on biological control of insect pests in many international journals.
Preface to tothe Edition Preface the Second Second Edition Ten years ago, the writing of this t his book was a long overdue task. At that time, the pioneering work of researchers such as Helmut van Emden, Robert van den Bosch, David Pimentel, and Richard Root prov pr ovide ided d the ini initial tial enl enligh ighten tenmen mentt and ins inspir pirati ation. on. Mu Much ch of thi thiss wo work rk was continued by various colleagues, many of them former UniverUniver sity of California at Berkeley graduate students (Deborah K. Letourneau, Matt Liebman, Cliff Gold, David Nestel, Michael Costello) who, with much scientific elegance, have contributed substantially to the advancement of this field. After Aft er con consid sidera erable ble deb debate ate on the ent entom omolo ologic gical al mer merits its of di dive versi rsifi fi-cation, involving involving colleagues such as the late Steve Risch, John VanVandermeer, David Andow, and Peter Kareiva who cautioned about the universality univ ersality of the effects of diver diversity sity on pest populations and called call ed for more research, there has been a renewed interest in the effects of habitat diversification on insect ecology. This contemporary reevaluation of the field has been championed by a new generation of scientists such as Robert L. Bugg, G. M. Gurr, J. A. Lys, and Doug A. Landis. Despite all of the scientific evidence, years ago we became convinced that diversity in agriculture is not only essential for pest suppression but also crucial to ensure the biological basis for the sustainability of production. We have witnessed the benefits of mimicking nature in our own experimental experimental plots and, more important, in hundreds of farmers’fields farmers’ fields in developing countries countries and in California. Projects led by many nongovernmental organizations (NGOs) in Latin America have used biodiversity as the basis of an agroecoagroeco logical approach tailored to meet the needs of resource-poor farmers throu thr ough gh mor moree sta stable ble yie yields lds and con conser serv vatio ation n of loc local al res resour ources ces.. The These se actions have resulted in enhanced food security, reduced use of toxic pesticides, and consequently healthier and better-fed rural families. These testimonies have been significant enough, often more so than what wh at st stat atis isti tica call an anal alys ysis is ma may y su sugg gges est, t, to co con nvi vinc ncee us of th thee be bene neffits of biodiversity biodiv ersity in agriculture.
We write the second edition of this book perhaps not so much as entomo ent omolog logist istss bu butt as agr agroec oecolo ologis gists ts or or,, bet better ter,, as sel self-m f-made ade soc social ial sci sci-entists. This evolution was natural since our research on intercropping placed us face to face with the challenges of rural development development in Latin America and the powerful agrarian structure of California. It is a healthy evolution, evolution, as env environmental ironmental problems in agriculture are not only ecological but also part of a social, economic, and political process. Therefore, the root causes of most pest problems affecting agriculture are inherent in the structural features of the prevailing agroeconomic system, which encourages energy-intensive, largescale, and specialized monocultures. It is our expectation that pest mana ma nage gers rs wi will ll be beco come me mo more re se sens nsiti itive ve to so socia cial, l, ec econ onom omic ic,, an and d cu cultu ltura rall issues, as pest problems cannot be understood by disentangling ecological fro from m soc socioe ioecon conomi omicc fac factor tors. s. On the con contra trary ry,, the agr agroec oecolo ologic gical al paradigm maintains that both must be examined holistically. Only a broader understanding will ensure that the benefits of a diversified agriculture can expand beyond the obvious entomological advantages to include concerns for social equity, economic viability, and cultural compatibility compatibility..
Acknowledgments Acknowledgments We ar aree in inde debt bted ed to th thee hu hund ndre reds ds of fa farm rmer erss in La Latin tin Am Amer erica ica,, Ca Cali li-fornia, and many other parts of the t he world who have showed us ecologi lo gica call lly y so soun und d wo work rkin ing g fa farm rms, s, wh whic ich h he help lped ed us be beco come me aw awar aree of th thee importance of diversification in agriculture. This book was initially written while the then-existing Division of Biological Control, UniUni versity of California, Berkeley, was a world-leading center of research on alternatives to pesticides. During that time, many colleagues and faculty provided encouragement encouragement and support, including Linda Schmidt, Javier Trujillo, Jeff Dlott, Andres Yurjevic, Deborah Letourneau, Steve Steve Gliessman, Matt Liebman, Leo Caltagirone, Donald Dahlsten, M. Alice Garcia, Marta Astier, and many others. The Jesse Smith Noyes Foundation of New York provided generous financial support for many years. Almost ten years after the first edition of this book, it is necessary to th than ank k ma many ny ad addi ditio tiona nall pe peop ople le fo forr th thei eirr en enth thus usias iasm, m, fa faith ith,, an and d es espe pe-cially dedication for translating many of the ideas in this book into practice: Christos Vasilikiotis, Vasilikiotis, Marco Barzman, Fabian Banga, Peter Rosset, Patrick Archie, Josh Mimer, Raul and Carlos Venegas, Santiago Sarandon, Eddy Peralta, Alfredo Jimenez, the extensionists of EMATER (Instuto de Assistencia e Extensão Rural) and EPAGRI (Emp (E mpre resa sa de Pe Pesq squi uisa sa e Ex Exte tens nsão ão Ru Rura rall de Sa Sant ntaa Ca Catar tarim ims) s) in Br Braz azil il,, the research team of CATIE (Tropical Agricultural Research and Higher Education Center) in Nicaragua, the graduate students of the agroecology master’s program at the International University of Andalu da lucia cia an and d th thee Me Medi dite terr rran anea ean n Ag Agro rono nomi micc In Inst stit itut utee of Ba Bari ri,, as we well ll as the hundreds of NGO personnel and farmers that constantly strive to enhance biodiversity biodiversity in the rural landscapes of Latin America. We are solely responsible for the views expressed in this book and hope that the assembled information provides a useful tool for students and field practitioners to understand and obtain criteria on the diversification diver sification of agroecosystems for enhanced pest management. We de dedi dica cate te th this is bo book ok to ou ourr mo moth ther ers, s, fa fami mili lies es,, an and d fr frie iend nds, s, as we well ll as all of the current and future farmers of the world, so that they can
use the principles of agroecology to better nurture the land and attain a truly sustainable agriculture. New donors have supported our efforts, so our special gratitude goes to the Foundation for Deep Ecology, Ecology, the Clarence Heller Charitable tab le Fo Foun undat dation ion,, the Or Organ ganic ic Fa Farmi rming ng Res Resear earch ch Fo Found undatio ation, n, and the California Department of Food and Agriculture (CDFA) Department of Pesticide Regulation. We also thank the Rockefeller Foundation for hosting us at their Bellagio Center where, while contemplating enchanting Lake Como, we wrote much of this second edition.
Introduction Introduction
Agriculture implies the simplification of nature’s biodiversity and reac re ache hess an ext xtre reme me fo form rm in cr crop op mo mono nocu cult ltur ures es.. Th Thee en end d re resu sult lt is th thee pr prooduction of an artificial ecosystem requiring constant human intervention. In most cases, this intervention is in the form of agrochemical inputs pu ts wh whic ich, h, in ad addi ditio tion n to te temp mpora orari rily ly boo boost sting ing yi yiel elds, ds, re resu sult lt in a num numbe berr of undesirable environmental and social costs (Altieri, 1987). As agricul agricultural tural modernization modernization progres progresses, ses, ecologi ecological cal princi principles ples are continuously ignored or overridden. As a consequence, modern agroecosystems are unstable. Breakdowns manifest themselves as recurrent pest outbreaks in many cropping systems and in the forms of salinization, soil erosion, pollution of water systems, etc. The worsening of most pest problems has been experimentally linked to the expansion of crop monocultures at the th e expense of vegetation diversity, diversity, which is an eses sential sen tial lan landsca dscape pe com compone ponent nt pro providi viding ng ke key y ecol ecologi ogical cal ser servic vices es to ens ensure ure crop protection (Altieri and Letourneau, 1982). Ninety-one percent of the world’s 1.5 billion hectares of cropland are under annual crops, mostl mo stly y mo mono nocu cult lture uress of wh whea eat, t, ri rice, ce, ma maiz ize, e, co cott tton, on, an and d so soyb ybea eans. ns. On Onee of the main problems arising from the homogenization of agricultural systems is an increased vulnerability of crops to insect pests and diseases, whic wh ich h ca can n be de dev vast astat atin ing g if th they ey inf infes estt a un unif ifor orm m cr crop op,, es espec pecia iall lly y in la larg rgee plantations. To To protect these crops worldwide, about 4.7 million pounds of pesticides were applied in 1995 (1.2 billion pounds in the United Stat St ates) es);; suc such h pe pesti stici cide de in inje jecti ction on has in incre creas ased ed in the pa past st te ten n ye year ars. s. In th thee United Uni ted Sta States, tes, en enviro vironme nmental ntal and soci social al cost costss asso associa ciated ted wit with h such pes pestiticide levels levels have been estima estimated ted at $8 billion per year (Pimentel (Pimentel et al., 1980). Such costs are still valid today. Crop losses due to pests remain at 30 pe perc rcent ent,, no di difffer feren entt fr from om th thir irty ty to for forty ty ye year arss ago ago.. In Ca Cali lifo forn rnia, ia, pe pesstici ti cide de us usee in incr crea eased sed fr from om 16 161 1 to 212 mi mill llio ion n po poun unds ds of ac acti tive ve in ingr gredi edien ent, t, despite the fact that crop acreage remained constant and that re search in integrated pest management (IPM) is quite advanced (Liebman, 1997). These are clear signs that the pesticide-based pesticide-based approach to pest control has reached its limits. An alternative approach is needed—one based on
the us the usee of ec ecol olog ogic ical al pr prin inci cipl ples es in or orde derr to ta take ke fu full ll ad adv van anta tage ge of th thee be bennefits of biodiversity in agriculture. This book analyzes the ecological basis for the maintenance of biodive di versi rsity ty in ag agric ricul ultu ture re an and d th thee rol rolee it ca can n pla play y in re rest stori oring ng th thee eco ecolo logi gical cal balan ba lance ce of agr agroe oecos cosys ystem temss so tha thatt sus sustai tainab nable le pr produ oducti ction on ma may y be ac achie hieve ved. d. Biodiversity performs a variety of renewal processes and ecological services in agroecosystems; when they are lost, the costs can be significant (Altieri, 1991b). The book focuses particularly on the ways in which biodiversity can contribute to the design of pest-stable agroecosystems. The effects of intercr int ercropp opping, ing, cov cover er crop croppin ping, g, wee weed d man manage agemen ment, t, and cro crop-f p-fiel ield d bor border der vegetation manipulation are discussed. A considerable amount of attention is paid to understanding the effects of these vegetationally diverse systems on pest population density and the mechanisms underlying pest reduction in polycultures. This is essential if vegetation management is effectively as the basis of ecologically based pest manageto be used effectively ment (EBPM) tactics in sustainable agriculture. Although insect communities in agroecosystems can be stabilized b y constructing vegetational architectures that support natural enemies and/ an d/or or di dire rect ctly ly in inhi hibit bit pe pest st att attac ack, k, th this is boo book k st stres resse sess th thee fa fact ct th that at ea each ch si sittuation must be assessed separately, given that long-term vegetationmanagement strategies are site specific and need to be developed with regard to local and regional environmental, socioeconomic, and cultural factors. In this way, crop mixtures may serve to meet the broader needs and an d pre prefe fere renc nces es of lo local cal fa farm rmers ers and, and, at the sa same me ti time me,, en enhan hance ce en envi viro ronnmental quality qualit y. This book builds on information emerging from renewed interest among am ong sc scie ienti ntist stss in th thee fi fiel eld d of ha habit bitat at ma manag nagem ement ent in en enhan hanci cing ng bi biol ologogical control of insects (Barbosa, 1998; Pickett and Bugg, 1998; Landis, Wratten, and Gurr, 2000; Smith and McSorley McSorley,, 2000).
Chapter 1
The Ecological Role of Biodiversity The Ecological Role of Biodiversity in Agriculture in Agriculture
Biodiversity refers to all species of plants, animals, and microorganisms existing and interacting within an ecosystem (McNeely et al., 1990). Global threats to biodiversity should not be foreign to agriculturalists, since agriculture, which covers about 25 to 30 percent ce nt of wo worl rld d lan land d ar area ea,, is pe perh rhap apss on onee of th thee ma main in ac acti tivi viti ties es af affe fect ctin ing g biological diversity. diversity. It is estimated that the global extent of cropland increased from around 265 million hectares (ha) in 1700 to around 1.5 bil billio lion n hec hectar tares es tod today ay,, pre predom domina inantl ntly y at the ex expen pense se of for forest est hab hab-itats (Thrupp, 1997). Very limited areas remain totally unaffected by agriculture-induced agriculture-indu ced land-use changes. Clearly, agriculture implies the simplification of the structure of the environment over vast areas, replacing nature’s diversity with a small sma ll num number ber of cul culti tiva vated ted pla plants nts and dom domest estica icated ted ani animal malss (An (Ando dow w, 1983a). In fact, the world’s agricultural landscapes are planted with only on ly som somee twe twelve lve spe specie ciess of gra grain in cro crops ps,, twe twenty nty-th -three ree ve veget getabl able-c e-crop rop species, and about thirty-five fruit- and nut-crop species (Fowler and Mooney, 1990); that is no more than seventy plant species which spread over approximately 1,440 million ha of presently cultivated land in the world (Brown (B rown and Young, Young, 1990), a sharp contrast cont rast with wit h the diversity diver sity of plant species found within one hectare of a tropical rainforest which typically contains over 100 species of trees (Myers, 1984 19 84). ). Of th thee 7, 7,00 000 0 cr crop op sp spec ecie iess us used ed in ag agri ricu cult ltur ure, e, on only ly 12 120 0 ar aree im im-port po rtan antt at a na natio tiona nall le leve vel. l. An es esti tima mate ted d 90 pe perc rcen entt of th thee wo worl rld’ d’ss ca callorie intake comes from just thirty crops, a small sample of the vast crop diversity available. Monocultures may have temporary economic advantages advantag es for farmers, but in the long term they do not represent an ecological optimum (USDA, 1973). Rather, the drastic narrowing of cultivated plant diversity has put the world’s food production in
greater peril (National Academy of Sciences [NAS], 1972; Robinson, 1996). The process of environmental simplification reaches an extreme form in agricultural monocultures, which affects biodiversity biodiversity in various ways: • Expansion of agricultural land with loss of natural habitats • Conversion into homogeneous agricultural landscapes with low
habitat value for wildlife Loss ss of wi wild ld sp spec ecie iess an and d be bene neffic icial ial ag agro rodi dive vers rsit ity y as a di dire rect ct co conn• Lo sequence of agrochemical inputs and other practices • Erosion of valuable genetic resources through increased use of uniform high-yielding varieties (HYV) In the deve developing loping world, agricultural diversity has been eroded as monocultures dominate. For example, in Bangladesh the promotion of Green Revolution rice led to the loss of diversity, including nearly 7,000 7,0 00 tra tradit dition ional al ric ricee va varie rieties ties and man many y fi fish sh spe specie cies. s. Sim Simila ilarly rly,, in the Philippines, the introduction of HYV rice displaced more than 300 traditional rice varieties. In North America similar losses in crop diversity are occurring. Eighty-six percent of the t he 7,000 apple varieties used us ed in th thee Un Unit ited ed St Stat ates es be betw twee een n 18 1804 04 an and d 19 1904 04 ar aree no lo long nger er in cu culltiv ti vat atio ion; n; of 2, 2,68 683 3 pe pear ar var arie ieti ties es,, 88 pe perc rcen entt ar aree no lo long nger er av avail ailab able le.. In Europe thousands of varieties of flax and wheat vanished following following the tak take-o e-ove verr by mod modern ern va varie rieties ties (Th (Thrup rupp, p, 19 1997) 97).. In fac fact, t, mo moder dern n agriculture is shockingly dependent on a handful of varieties for its ma jor crops. For example, in the United States, three decades ago 60 to 70 percent of the total bean acreage was planted with two to three bean varieties, 72 percent of the potato acreage with four varieties, and 53 percent of cotton with only three varieties (NAS, 1972). Resear Res earche chers rs ha have ve rep repeat eatedl edly y wa warne rned d abo about ut the ex extre treme me vul vulner nerabi abillity associated with this genetic uniformity. uniformity. Perhaps the most striking example of vulnerability associated with homogenou homogenouss uniform agricult cu ltur uree wa wass th thee co coll llap apse se of Ir Iris ish h po pota tato to pr prod oduc ucti tion on in 18 1845 45,, wh wher eree th thee uniform stock of potatoes was highly susceptible to the blight (Phytophthora infestans). During the nineteenth century in France, wine grape production was wiped out by a virulent pest (Phylloxera vitiviti foliae) which eliminated 4 million hectares of uniform grape varieties. Banana monocultural plantations in Costa Rica have been repeatedly
seriously jeopardized by diseases such as Fusarium oxysporum and yellow Sigatoka. In the United States, in the early 1970s, uniform high-yielding maize hybrids constituted about 70 percent of all the maize varieties; a 15 percent loss of the entire crop by leaf blight occurred in that decade (Adams, Ellingbae, and Rossineau, 1971). Recent expansion of transgenic maize and soybean monocultures, reaching about 45 million hectares mostly grown in the United States in less than six years, represents a worrisome landscape-simplification trend of homogenization (Marvier, (Marvier, 2001). The net result of biodiversity simplification for agricultural purposes is an artificial ecosystem that requires constant human intervention. Commercial seedbed preparation and mechanized planting replace natural methods of seed dispersal; chemical pesticides replace natural controls on populations of weeds, insects, and pathogens; and genetic manipulation replaces natural processes of plant evolution and selection. Even decomposition is altered since plant growth grow th is harvested and soil fertility maintained not through biologically mediated nutrient recycling but with fertilizers (Altieri, 1995). Another important way in which agriculture affects biodiversity biodiversity is through the externalities associated with the intensiv intensivee agrochemical and mechanical technology used to boost crop production. In the United States about 17.8 million tons of fertilizers are used in grain production systems, and about 500 million pounds of pesticides are applied annually to farmlands. Although these inputs have boosted crop yields, their undesirable environmental environmental effects are undermining the sus sustai tainab nabili ility ty of agr agricu icultu lture. re. En Envir viron onmen mental tal (in (inclu cludin ding g the los losss of key biodiversity elements such as pollinators and natural enemies) and social costs associated with pesticide use today are crudely estimated to reach more than $850 million per year in the U.S. (Pimentel et al al., ., 19 1980 80). ). Ab Abou outt 1. 1.5 5 mi mill llio ion n he hect ctar ares es wor world ldwi wide de an and d ab abou outt 27 pe perr cent of the irrigated land in the United States is damaged by salinizat iz atio ion n du duee to ex exce cess ssiv ivee or im impr prop oper er ir irri riga gatio tion. n. Be Beca caus usee of la lack ck of ro ro-tations and insuff insufficient icient vegetation cover, soil erosion levels average 185 tons/ha per year in U.S. croplands, well above the acceptable threshold. It is estimated that soil degradation has reduced crop productivity by around 13 percent (Brown and Young, 1990). Nowhere are the consequences of biodiversity reduction more evident than in the realm of agricultural pest management. The instabilinstability it y of ag agro roec ecos osys yste tems ms be beco come mess ma mani nife fest st as th thee wo wors rsen enin ing g of mo most st in in--
sect-pest problems is incre sect-pest increasing asingly ly link linked ed to the exp expansio ansion n of crop monocultures at the expense of the natural vegetation, thereby decreasing local habitat diversity (Altieri and Letourneau, 1982; Flint and Roberts, 1988). Plant communities that are modified to meet the special needs of humans become subject to heavy pest damage, and generally the more intensely such communities are modified, the more abundant and serious the pests. The inherent self-regulation characteristics of natural communities are lost when humans modify such suc h com commun munitie itiess by bre breaki aking ng the fra fragil gilee thr thread ead of com commu munit nity y int intereractions (Turnbull, (Turnbull, 1969). This breakdo breakdown wn can be repaired by restoring the elements of community homeostasis through the addition or enhancement of biodiversity. In this book, we explore practical steps to break the tendency for monoculture crop production and thus reduce ecological vulnerability by restoring agricultural biodiversity at the field and landscape level. The most obvious advantage of diversification diversificati on is a reduced risk of total crop failure due to invasions by unwanted species and subsequent pest infestations.
TRADITIONAL AGROECOSYSTEMS AS MODELS OF BIODIVERSE FARMS
Not all forms of agriculture lead to the extreme simplification of biodiversity. A salient feature of traditional farming systems managed ag ed by sm smal alll far arme mers rs in th thee th thir ird d wor orld ld is th thei eirr de degr gree ee of pl plan antt di div ver er-sity in the form of polycultures and/or agroforestry patterns. patterns. In fact, the species-rich quality of all biotic components of traditional agroecosystems is comparable with that of many natural ecosystems. These systems offer a means of promoting diversity of diet and income, stability of production, minimization of risk, reduced insect and disease incidence, efficient use of labor, intensification of production with limited resources, and maximization of returns under low levels of technology. Traditional, multiple-cropping systems are estimated to still provide as much as 15 to 20 percent of the world’s food supplies. In Latin America, farmers grow 70 to 90 percent of their beans in combination with maize, potatoes, and other crops. Maize is intercropped in 60 percent of the region’s maize-growing area (Francis, 1986).
Traditional cropping systems are also genetically diverse, containTraditional ing numerous varieties of domesticated crop species as well as their wild relatives. In the Andes, farmers cultivate as many as 50 potato var arie ietie tiess in th thei eirr fiel ields ds.. Ma Main inta tain inin ing g ge gene neti ticc di dive vers rsit ity y ap appe pear arss to be of even greater importance as land becomes more marginal and hence farming more risky. In Peru, for example, the number of potato varietie et iess cu culti ltiv vate ated d in incr crea ease sess wi with th th thee al alti titu tude de of th thee la land nd fa farm rmed ed.. Ge Gene neti ticc dive di versi rsity ty con confer ferss at lea least st par partia tiall res resist istanc ancee to dis diseas eases es tha thatt are spe specif cific ic to particular strains of crops and allows farmers to exploit different soil so il ty type pess an and d mi micr croc oclim limate atess fo forr a var ariet iety y of nu nutr triti ition onal al an and d ot othe herr us uses es (Brush, 1982). On the other hand, traditional agroforestry systems throughout throughout the tropics commonly contain well over 100 annual and perennial plant species per field, species used for construction materials, firewood, tools, medicine, livestock feed, and human food. In these systems, besides providing useful products, trees minimize nutrient leaching and soil erosion and restore key nutrients by pumping them from the lower soil strata (Nair, 1993). Examples include the home gardens of the Huastec Indians in Mexico and the agroforestry systems of the Amazonian Kayapo and Bora Indians (Toledo et al., 1985). Intercropping, agroforestry, shifting cultivation, and other traditional tio nal far farmin ming g met method hodss mim mimic ic nat natura urall eco ecolog logica icall pr proce ocesse sses, s, and the their ir sustainability lies in the ecological models they follow. This use of natural analogies suggests principles for the design of agricultural systems that make effective use of sunlight, soil nutrients, rainfall, and biological resources. Several scientists now recognize how tradition ti onal al fa farm rmin ing g sy syst stem emss ca can n se serv rvee as mo mode dels ls of ef efffic icie ienc ncy y as th thes esee sy sysstemss inc tem incorp orpora orate te car carefu efull man manage agemen mentt of soi soil, l, wa water ter,, nu nutri trient ents, s, and biological resources. In developing countries, biodiversity biodiversity can be used to help the great masss of res mas resou ource rce-po -poor or far farmer mers, s, mos mostly tly con confi fined ned to mar margin ginal al soi soils, ls, hil hilllsides, and rain-fed areas, to achieve year-round food self-sufficiency, reduce their reliance on scarce and expensive agricultural chemical inputs, and develop production systems that rebuild the productive capacities of their small holdings (Altieri, 1987). The objective objective is to assist farmers in developing sustainable farming systems that satisfy food self-sufficiency, as well as stabilize production by avoiding soil erosion (Beets, 1990). Technically, the approach consists of devising multiple-use farming systems emphasizing soil and crop protection
and achieving soil-fertility improvement and crop protection through the integration of trees, animals, and crops (Figure 1.1). Exampl Exa mples es of gra grassr ssroot ootss rur rural al de deve velop lopmen mentt pro progra grams ms in Lat Latin in Ame Amerrica analyzed by Altieri (1991d, 1999) suggest that the maintenance and/or enhancement of biodiversity in traditional agroecosystems represents a strategy that ensures diverse diets and income sources, stable production, minimum risk, intensive production with limited resources, and maximum returns under low levels of technology technology.. In these systems, the complementarity of agricultural enterprises reduces the need for outside input. The correct spatial and temporal assemblage of crops, trees, animals, soil, and so forth enhances the interactions that sponsor yields dependent on internal sources and recycl rec ycling ing of nu nutri trient entss and or orga ganic nic mat matter ter and on tro troph phic ic rel relati ations onship hipss among plants, insects, or pathogens, which enhance biological pest control (Altieri and Nicholls, 2000). Since traditional farmers generally have a profound knowledge of biodiversity, their knowledge and environmental perceptions should Fruits, timber, firewood, forage, windbreaks
Biomass Mulch Green manure
Tree integration
Soil and water conservation practices
Alley cropping mulching Windbreaks Minimum tillage Contour planting
Soil protection
Soil fertility improvement
Crop protection
Rotations Cover crops Multiple cropping
Food and industrial crops
Manure compost
FORAGE
Animal integration
Meat, eggs, milk
FIGURE 1.1. Assemblage of a diversified agroecosystem resulting in enhanced soil protection, soil fertility, and biological crop protection (after Altieri, 1987).
be integrated into schemes of agricultural innovation that attempt to link resource conservation and rural development (Altieri and Hecht, 1991). For a resource conservation strategy compatible with a diversified production strategy to succeed among small farmers, the process must be linked to rural development efforts efforts that give equal importance to local resource conservation and food self-sufficiency and/or participation in local markets. Any attempt at soil, forest, or crop cro p gen genetic etic con conser serv vatio ation n mu must st str strugg uggle le bo both th to pr prese eserv rvee the di dive versi rsity ty of the agroecosystems in which these resources occur and to protect the local cultures that nurture them. Cultural diversity is as crucial as biological diversity.
THE ECOLOGICAL ROLE OF BIODIVERSITY
In addition to producing valuable valuable plants and animals, biodiversity performs many ecological services. In natural ecosystems, the vegetative cover of a forest or grassland prevents soil erosion, replenishes groundwater, and controls flooding by enhancing infiltration and reducin du cing g wa water ter run runof off. f. Nat Natura urall hab habitat itatss als also o con contai tain n wil wild d pop popula ulatio tions ns of domesticated plants and animals, and these populations contain useful genes that are often absent in the domesticated gene pool. The entire ti re ra rang ngee of ou ourr do dome mest stic ic cr crop opss is de deri rive ved d fr from om wi wild ld sp spec ecie iess th that at ha have ve been modified through domestication, selective breeding, and hybridization. Most remaining world centers of diversity contain populatio la tions ns of var ariab iable le an and d ad adap aptab table le la land ndra race ces, s, as we well ll as wi wild ld an and d we weed edy y relatives of crops (Harlan, 1975). Many traditionally managed farming in g sy syst stem emss in th thee Th Thir ird d Wor orld ld co cons nsti titu tute te in si situ tu re repo posi sito tori ries es of na nati tive ve crop cr op di dive vers rsity ity (A (Alt ltier ierii an and d He Hech cht, t, 19 1991 91). ). Th Ther eree is gr grea eatt co conc ncer ern n to toda day y about genetic erosion of crops in areas where small farmers are pushed by agricultural modernization to adopt new or modified varieties at the expense of traditional ones. In agricultural systems, biodiv biodiversity ersity performs ecosystem services beyon be yond d pro produc ductio tion n of fo food, od, fi fiber ber,, fue fuel, l, and inc income ome.. Exa Exampl mples es inc includ ludee recycl rec ycling ing of nu nutri trient ents, s, con contro troll of loc local al mic microc roclim limate ate,, re regul gulati ation on of local hydrological processes, regulation of the abundan abundance ce of undesirable organisms, and detoxification of noxious chemicals. These renewal processes and ecosystem services are largely biological; therefore, their persistence depends upon maintenance of biological diversity.
When these natural services are lost due to biological simplification, the economic and environmental costs can be quite significant. Economic no micall ally y, agr agricu icultu ltural ral bu burde rdens ns include include the nee need d to su suppl pply y cro crops ps with costly external inputs, since agroecosystems deprived deprived of basic regulatory functional components lack the capacity to sponsor their own soil fertility and pest regulation. Often the costs also involve involve a reduction ti on in th thee qu quali ality ty of lif lifee du duee to de decr crea ease sed d so soil, il, wa wate terr, an and d fo food od qu qual ality ity when pesticide, nitrate, or other type of contamination occurs. Clearly,, the fates of agriculture and biodiversity are intertwined. It Clearly is possible to intensify agriculture in a sustainable manner in order to secure some of the remaining natural habitats, thus ensuring the provision of environmental services to agriculture. Agroecological forms of intensification can also enhance the conservation and use of agrobiodiversity, which can lead to better use of natural resources and agroecosystem stability (Gliessman, 1999). THE NATURE OF BIODIVERSITY IN AGROECOSYSTEMS
Biodive Biodi vers rsity ity in ag agro roec ecos osys yste tems ms ca can n be as var arie ied d as th thee cr crop ops, s, we weed eds, s, arthropods, or microorganisms involved or the geographical location and climatic, edaphic, human, and socioeconomic factors. In general, the degree of biodiversity in agroecosystems depends on four main characteristics of the agroecosystem (Southwoo (Southwood d and Way ay,, 1970): • • • •
The di dive versi rsity ty of ve veget getatio ation n wit within hin and aro aroun und d the agr agroec oecosy osyste stem m The per perman manenc encee of the va vario rious us cro crops ps wit within hin the agr agroec oecosy osyste stem m The intensity of management The extent of the isolation of the agroecosystem from natural vegetation
The biodiversity biodiversity components of agroecosystems can be classified cl assified in relation to the roles they play in the functioning of cropping systems. Accord Acc ording ing to thi this, s, agr agricu icultu ltural ral bio biodi dive versi rsity ty can be gro groupe uped d as fol follo lows ws (Swift and Anderson, 1993): • Productive biota: crops, trees, and animals chosen by farmers
that play a determining role in the diversity and complexity of the agroecosystem
• Resour organ ganism ismss tha thatt con contrib tribute ute to pro produc ducti tivit vity y thr throu ough gh Resource ce biota: or
pollination, biological control, decomposition, etc. • Destructive biota: weeds, insect pests, microbial pathogens, etc., etc ., tha thatt far farmer merss aim at red reduci ucing ng thr throug ough h cul cultur tural al man manage agemen mentt Two distinct components of biodiversity can be recognized in agroecosystems (Vandermeer (Vandermeer and Perfecto, 1995). The first component, planne includ ludes es the cro crops ps and li live vesto stock ck pur purpo posel sely y planned d biodive biodiversity rsity,, inc included in the agroecosystem by the farmer, which will vary depending on the management inputs and the spatial and temporal arrangements of crops. The second component, associated biodiversity bio diversity,, inclu in clude dess all soi soill flo flora ra and fa fauna una,, her herbi bivo vores res,, car carni nivo vores res,, dec decom ompos posers ers,, etc., etc ., tha thatt col colon onize ize the agr agroec oecosy osyste stem m fro from m sur surrou roundi nding ng en envir viron onmen ments ts and that will thrive in the agroecosystem depending on its management and structure. The relationship of both types of biodiversity comp co mpon onen ents ts is ill illus ustr trat ated ed in Fi Figu gure re 1. 1.2. 2. Pl Plan anne ned d bi biod odiv iver ersi sity ty ha hass a di di-rect function, as illustrated by the bold arrow connecting the planned Agroecosystem Management Planned Biodiversity
Creates conditions that promote
Associated Biodiversity
Biodiversity of Surrounding Environment
Promotes
Ecosystem Function e.g., pest regulation, nutrient cycling, etc. Promotes
Key: Direct function Indirect function
FIGURE 1.2. The relationship between planned biodiversity (that which the farmer determines based on management of the agroecosystems) and associassoci ated biodiversity and how the two promote ecosystem function (modified from Vandermeer and Perf Perfecto, ecto, 1995).
biodiversity box with the ecosystem function box. Associated biodiversity also has a function, but it is mediated through planned biodi bio dive versi rsity ty.. Thu Thus, s, pla planne nned d bio biodi dive versi rsity ty als also o has an ind indire irect ct fun functio ction, n, illu il lust stra rate ted d by th thee do dotte tted d ar arro row w in th thee fig igur ure, e, wh whic ich h is re real aliz ized ed th thro roug ugh h its influence on the associated biodiversity. biodiversity. For example, the trees t rees in an agroforestry system create shade, which makes it possible to grow sun-intolerant crops. So, the direct function of this second species (the trees) is to create shade. Yet along with the trees might come wasps that seek out the nectar in the tree’s flowers. These wasps may in turn be the natural parasitoids of pests that normally attack crops. Thee wa Th wasp spss ar aree pa part rt of th thee as asso socia ciate ted d bi biod odiv iver ersi sity ty.. Th Thee tr tree ees, s, th then en,, cr creeate shade (direct function) and attract at tract wasps (indirect function) (Van(Vandermeer and Perfecto, 1995). Complementary interactions between the various biotic components can also be of a multiple nature. Some of these interactions can be used to induce positiv positivee and direct effects on the biological control of spe specif cific ic cro crop p pes pests, ts, soi soil-f l-fert ertilit ility y re regen genera eratio tion n and and/or /or enh enhanc anceme ement, nt, and soil conservation. The exploitation of these interactions in real situations involves agroecosystem design and management and requires an understanding of the numerous relationships among soils, microorganisms, microorg anisms, plants, insect herbivores, and natural enemies. According to agroecological theory, the optimal behavior of agroecosystems depends on the level of interactions between the various biotic and abiotic components. By assembling a functional biodiversity (Figure 1.3), it is possible to initiate synergisms that subsidize agroecosystem processes by providing ecological services such as the activa ti vatio tion n of soi soill bio biolog logy y, the rec recycl ycling ing of nut nutrie rients nts,, the enh enhanc anceme ement nt of beneficial arthropods and antagonists, and so on (Altieri, 1995; Gliessman, 1999), all important in determining the sustainability of agroecosystems. In modern agroecosystems, the experimental evidence suggests thatt bio tha biodi dive versi rsity ty can be use used d fo forr imp impro rove ved d pes pestt man manage agemen mentt (An (Ando dow w, 1991a). Several Several studies have shown that it is possible to stabilize the t he insect ins ect com commu munit nities ies of agr agroec oecosy osyste stems ms by des design igning ing and con constr struct ucting ing vegetational architectures that support populations of natural enemies or have direct deterrent effects on pest herbivores. The key is to identify the type of biodiversity that is desirable to main ma inta tain in an and/ d/or or en enha hanc ncee in or orde derr to ca carr rry y ou outt ec ecol olog ogic ical al se serv rvic ices es,, an and d
FIGURE 1.3. System dynamics in diverse agroecosystems (after Gliessman, 1999).
then to determine the best practices that will encourage the desired biodiversity components (Figure 1.4). There are many agricultural practices and designs that have the potential to enhance functional biodiversity biodiv ersity and others that negati negatively vely affect it. The idea is to apply the best management practices in order to enhance or regenerate the kind of biodiversity that can subsidize the sustainability of agroecosystems by providing ecological ecological services such as biological pest control, nutrient cycling, water and soil conservation, etc. The role of agroecologists should be to encourage those agricultural practices that increase increa se the abundance and diversity of above- and belowground organisms, which in turn provide key ecological services to agro ecosystems (Figure 1.5). Thus, a key strategy of agroecology is to exploit the comple mentarity and synergy that result from the various combinations of crops, trees, and animals in agroecosystems that feature spatial and temporal arrangements such as polycultures, agroforestry systems, and crop-livestock mixtures. In real situations, the exploitation of
COMPONENTS Pollinators
Predators and parasites
Herbivores
Noncrop Earthworms vegetation
Soil mesofauna
Soil microfauna
AGROECOSYSTEM BIODIVERSITY
Pollination FUNCTIONS
Population regulation
Genetic Biological introgression control
Competition Biomass consumption Allelopathy Sources of Nutrient natural enemies cycling Crop wild relatives
ENHANCEMENT Intercropping Agroforestry Rotations
Decomposition Nutrient Soil structure Predation cycling Nutrient cycling
Nutrient cycling
Disease suppression
Organic Cover Green No-tillage Composting Windbreaks crops manuring matter addition
FIGURE 1.4. The components, functions, and enhancement strategies of biodiversity in agroecosystems (after Altieri, 1991a).
these interactions involves agroecosystem design and management and requires an understanding of the numerous relationships among soils, soi ls, mic micro roor organ ganism isms, s, pla plants nts,, ins insect ect her herbi bivo vores res,, and nat natur ural al ene enemie mies. s. This book analyzes options for agroecosystem design in detail. INCREASE IN NATURAL ENEMIES SPECIES DIVERSITY LOWER PEST POPULATION DENSITIES
Hedgerows Shelterbelts Windbreaks
Pol olyc ycul ultu turres
Rota Ro tati tion ons s Cover crops
Habitat diversification
Low soil Organic soil management disturbance tillage practices
AGROECOSYSTEM MANAGEMENT
Cultural practices Conventional Total weed tillage removal
Monoculture
Pesticides Chemical fertilization
Decrease in Natural Enemies Species Diversity Population Increases of Pestiferous Species FIGURE 1.5. The effects of agroecosystem management and associated cultural practices on the biodiversity of natural enemies and the abundance of insect pests (after Altieri and Nicholls, 2000).
Chapter 2
Agroecology and Management Agroecology and Pest Pest Management
THE NATURE OF AGRICULTURAL HABITATS AND ITS RELATION TO PEST BUILDUP
Each re Each regi gion on ha hass a un uniq ique ue se sett of ag agro roec ecos osys yste tems ms th that at re resu sult lt fr from om lo lo-cal climate, topography, soil, economic relations, social structure, and history. history. Each region contains a hierarchy of systems (Figure 2.1) in which the regional system is a complex of land-utilization units with farming subsystems and cropping subsystems which produce and transform primary products, involving involving a large service sector including urban centers (Hart, 1980). The agroecosystems of a region often include both commercial and local-use agricultural systems that rely on technology to different extents depending on the availability of land, capital, and labor. labor. Some technologies in modern systems te ms ai aim m at ef effi fici cien entt la land nd us usee (r (rel elian iance ce on bi bioc oche hemi mica call in inpu puts ts); ); ot othe hers rs reduce labor (mechanical inputs). i nputs). In contrast, resource-poor farmers usuall usu ally y ado adopt pt lo low-i w-inpu nputt tec techno hnolog logy y and lab labor or-in -inten tensi sive ve pra practi ctices ces tha thatt optimize production efficiency and recycle scarce resources (Matteson, Altieri, and Gagne, 1984). Although each farm is unique, farms can be classified by type of agricu agr icultu lture re or agr agroec oecosy osyste stem. m. Fu Funct nction ional al gro groupi uping ng is ess essent ential ial for de de-vising appropriate management strategies. Five criteria can be used to classify agroecosystems in a region: 1. The types of crop and livestock 2. The methods used to grow the crops and produce the stock 3. The relative intensity of use of labor, capital, and organization, and the resulting output of product
4. The disposal of the products for consumption (whether (whether used for subsistence or supplement on the farm or sold for cash or other goods) 5. The structures used to facilitate farming operations (Norman, 1979)
FIGURE 2.1. Agriculture as a hierarchy of systems (after Hart, 1980).
Based Base d on th thes esee cr crit iter eria ia,, it is po poss ssib ible le to re reco cogn gniz izee se sev ven ma main in ty type pess of agricultural systems in the world (Grigg, 1974): 1. Shifting cultivation systems 2. Semipermanent rain-fed cropping systems 3. Permanent rain-fed cropping systems 4. Arable irrigation systems 5. Perennial crop systems 6. Grazing systems 7. Sys System temss wit with h re regul gulate ated d ley fa farmi rming ng (al (alter ternat nating ing ara arable ble cro croppi pping ng and sown pasture) Systems 4 and 5 have evolved evolved into habitats that are much simpler in form and poorer in species than the others, which can be considered more diversified, permanent, and less disturbed. Within the range of world agricultural systems, traditional polycultures require less energy and external inputs than modern orchards, field crops, and vegetable cropping systems to achieve a similar level of desired stability (Figure 2.2). This greater stability apparently results from certain ecological and management attributes inherent to polycultural Cropping systems
Energy inputs required to achieve desired stability
Field crops
Fruit orchards
Cereals
Mixed farming Traditional polycultures STABILITY Desired stability level
FIGURE FIGUR E 2.2 2.2.. Ener Energy gy req requir uireme ements nts to sust sustain ain a desi desired red le leve vell of pro produc ductio tion n sta stabilbility in a range of farming systems (after Altieri, 1987).
cropping systems. Modern systems require more radical modifications of their structure to approach a more divers diversified, ified, less disturbed state. Across the world, agroecosystems differ in age, diversity, structure, and management. In fact, there is great variability in basic ecological and agronomic patterns among the various dominant agroecosystems (Figure 2.3). In general, agroecosystems that are more dive di verse rse,, mor moree per perman manent ent,, iso isolate lated, d, and man manage aged d wit with h lo low-i w-inpu nputt tec techhnology (i.e., agroforestry systems, traditional polycultures) take fuller advantage of work usually done by ecological processes associated with wit h hig higher her bio biodi dive versi rsity ty tha than n hig highly hly sim simpli plifi fied, ed, inp inputut-dri drive ven, n, and dis dis-turbed systems (i.e., modern vegetable monocultures and orchards). All agroecosystems are dynamic and subjected to different levels levels of management so that the crop arrangements in time and space are continually changing in the face of biological, cultural, socioeconomic, and environmental factors. Such landscape variations determine the degree of spatial and temporal heterogeneity characteristic of ag agri ricu cult ltur ural al re regi gion ons, s, wh whic ich h ma may y or ma may y no nott be bene nefi fitt th thee pe pest st pr prot otec ec-tion of particular agroecosystems. Thus, one of the main challenges
FIGURE 2.3. Ecological patterns of contrasting agroecosystems (after Altieri, 1987). Longer bars indicate a higher degree of the characteristic.
facing agoecologists today is identifying the types of heterogeneity (either at the field or regional level) that will yield desirable agricultural results (i.e., pest regulation), given the unique environment environment and entomofaunaa of each area. This challenge can be met only by further entomofaun analyzing the relationship between vegetation diversification diversification and the popul po pulati ation on dy dynam namics ics of her herbi bivo vore re spe specie ciess in in lig light ht of the di dive versi rsity ty and complexity of site-specific agricultural systems. A hypothetical pattern in pest regulation according to agroecosystem temporal and spatial diversity is depicted in Figure 2.4. According to this “increasing probability for pest buildup” gradient, agroecosystems on the left side of the gradient are more biodiv biodiverse erse and tend to be more amenable to manipulation since polycultures already contain many of the key factors required by natural enemies. There are, however, habitat manipulations that can introduce appropriate diversity into the important (but biodiversity impoverished) grain, vegetable, and row crop systems lying in the right half of Figure 2.4. Althou Alt hough gh her herbi bivo vores res va vary ry wid widely ely in the their ir res respo ponse nse to cro crop p dis distri tribu bu-tion, abundance, and dispersion, the majority of agroecological studies show that structural (i.e., spatial and temporal crop arrangement) and management (i.e., crop diversity, diversity, input levels, etc.) attributes of agroecosystems influence herbivore dynamics. Several of these attrib tr ibut utes es ar aree re rela late ted d to bi biod odiv iver ersi sity ty,, an and d mo most st ar aree am amen enab able le to ma mana nage ge-ment (i.e., crop sequences and associations, weed diversity, genetic diversity, etc.).
FIGURE 2.4. A classification of dominant agricultural agroecosystems agroecosystems on a gradient of diversity and vulnerability to pest outbreak (after Altieri, 1991c).
Crop temporal and spatial patterns in agroecosystems throughout the world vary tremendously. tremendously. There has been little analysis and discussion of whether temporal and spatial patterns that characterize crop phenology in agroecosystems influence the potential success of biological control through conservation. Crop spatial and temporal patt pa tter erns ns de dete term rmin inee th thee ex exte tent nt an and d pe pers rsis iste tenc ncee of cr crop op pl plan ants ts an and, d, th thus us,, the availability of key resources associated with the crop (Kareiva, 1983; Van Emden, 1990). The spatial and temporal availability of those tho se cro cropp-ass associ ociate ated d res resour ources ces (an (and d tho those se in sur surro round unding ing or adj adjace acent nt unmanaged habitats or in managed refuges) also may be critical determinants of whether, when, or how they respond to herbivores or other resources provided by crops (or other plants). Indeed, in simulation studies, Corbett and Plant (1993) have noted that the timing of the availability of interplanted (refuge) vegetation relative to t he germination of crop plants may determine if the refuge is likely to act as a source of natural enemies or as a sink (i.e., taking natural enemies away from crops). Agroecosystems are managed habitats with concentrations of perennial crops, annual crops, or both. The crop plant’ plant’ss life cycle, to a great extent, dictates the nature of the habitat (i.e., its structure and texture, longevity, and the composition and complexity of its fauna and flora). The permanence of a crop determines the intensity and complexity of the interactions that unfold in a given agroecosystem. In both annual and perennial agroecosystems, crop phenology may cause asynchrony between resource availability and the natural enemy (predator or parasite) stage requiring that resource. Effective conservation conserv ation of natural enemies must ameliorate and/or compensate for the elimination, reduction, or disruption of needed resources and conditions that result from patterns of crop phenology in agroecosystems. In annual agroecosystems, the availability of the crop varies in time and space depending on agroeconomic as well as biological constraints (Barbosa, 1998). At one end of a gradient, an agroecosystem may consist of a sequence of a single crop cultivated throughout most or all of the growing season (see Figure 2.5, A). At thee ot th othe herr en end d of th thee gr grad adie ient nt,, an ag agro roec ecos osys yste tem m ma may y be ch char arac acte teri rize zed d by a sequence of plantings and harvests of different different crops (see C). A third point in this hypothetical gradient is represented by agroeco-
FIGURE 2.5. Hypothetical sequence of crop plants. Bars with different patterns represent different crops or crop cultivars (after Barbosa, 1998).
systems in which a giv given en crop may occur discontinuously at two different times during a season (see B). Implementation of biological control tactics will vary among crop sequences. In sequence 1A, resource diversity for natural enemies is low,, and the monoculture ensures completion of pest life cycles. Selow quence 1C provides more diversity of resources for natural enemies and breaks the pest life cycles more effectively. In a dis discon contin tinuou uouss cro cropp pping ing sys system tem imp implem lement entatio ation n of con conser serva vatio tion n biol bi olog ogic ical al co cont ntro roll du duri ring ng on onee or bo both th cr crop op ph phas ases es is li like kely ly to ha have ve lit little tle impa im pact ct wi with thou outt a th thor orou ough gh pl plan an fo forr th thee co cons nser erv vati ation on of na natu tura rall en enem emies ies during the interval between the first and the second crop plantings. Natur Nat ural al ene enemie mies, s, par partic ticula ularly rly mon monoph ophago agous us spe species cies,, mus mustt sur survi vive ve wh when en the crop and hosts are not present. Alternatively, the conservation of natur nat ural al ene enemie miess may in invo volve lve rel relian iance ce on or man manipu ipulati lation onss of unm unmana anaged ged habi ha bita tats ts (o (orr ma mana nage ged d re refu fuge ges) s) in th thee la land ndsc scap apee of th thee ag agro roec ecos osys yste tem m to compensate for the discontinuous crop pattern.
CROP DIVERSIFICATION AND BIOLOGICAL CONTROL
Crop monocultures are difficult environments environments in which to induce efficient eff icient biological pest control because these systems lack adequate resources for effective effective performance of natural enemies and because
disturbing cultural practices are often utilized in such systems. More diversified cropping systems already contain certain specific resources for natural enemies provided by plant diversity and are usually not disturbed with pesticides (especially when managed by resource-poor farmers who cannot afford high-input technology). They are also more amenable to t o manipulation. Thus, by replacing or adding diversity to existing systems, it may be possible to exert changes in habitat diversity that enhance natural enemy abundance and effectiveness by 1. providing alternative hosts/prey hosts/prey at times of pest-host scarcity; 2. providing food (pollen and nectar) for adult parasitoids and predators; 3. providing refuges for overwintering, nesting, and so on; and 4. maintaining acceptable populations of the pest over extended periods to ensure continued survival of beneficial insects (van den Bosch and Telford, 1964; Altieri and Letourneau, 1982; Powell, 1986). The specific resulting effect or the strategy to use will depend on the species of herbivores and associated natural enemies, as well as on prop properties erties of the ve vegetati getation, on, the phy physiolo siological gical condition condition of the crop, or the nature of the direct effects of particular plant species (Letourneau, 1987). In addition, the success of enhancement measures can be influenced by the scale upon which they are implemented (i.e., field scale, farming unit, or region) since field size, within-field and surrounding vegetation composition, and level of field isolation (i.e., distance from source of colonizers) will all affect immi im migr grat atio ion n ra rate tes, s, em emig igra rati tion on ra rate tes, s, an and d th thee ef effe fect ctiv ivee te tenu nure re ti time me of a particular natural enemy in a crop field. Whatever diversity enhancement strategy is used, it must be based on a thorough knowledge of the natural enemies’ ecological requirements. requirements. Perh Pe rhap apss on onee of th thee be best st st stra rate tegi gies es to in incr crea ease se ef effe fecti ctive vene ness ss of pr pred ed-ators and parasitoids is the manipulation of nontarget food resources (i.e., alternate hosts/prey and pollen/nectar) (Rabb, Stinner, and van den Bosch, 1976). Here it is important that not only the density of the nontarget resource be sufficiently high to influence enemy populations but also that the spatial distribution and temporal dispersion of the resource be adequate. Proper manipulation of the nontarget re-
source should result in the enemies colonizing the habitat earlier in the season than the pest and frequently encountering an evenly distributed resource in the field, thus increasing the probability that the enemy will remain in the habitat and reproduce (Andow and Risch, 1985). Certain polycultural arrangements increase and others reduce the spatial heterogeneity of specific food resources; thus, particular spec sp ecie iess of na natu tura rall en enem emie iess may be mo more re or le less ss abun abunda dant nt in a sp spec ecif ific ic polyculture. These effects and responses can be determined only experimentally across a whole range of agroecosystems. The task is indeed overwhelming overwhelming since enhancement techniques t echniques must necessarily be site specific. The literature is full of examples of experiments documenting that diversification diver sification of cropping systems often leads to reduced herbivore populations. The studies suggest that the more diverse the agroecosystem and the longer this diversity remains undisturbed, the more internal links develop to promote greater insect stability (Way, 1977). It is clear, however, however, that the stability of the insect community depends not only on its trophic diversity but also on the actual density-dependence nature of the trophic levels (Southwood and Way, 1970 19 70). ). In ot othe herr wo word rds, s, st stab abili ility ty wi will ll de depe pend nd on th thee pr prec ecis isio ion n of th thee re re-sponse of any particular trophic link to an increase in the population at a lower level. Although most experiments have moved forward on documenting inse in sect ct po popu pulat latio ion n tr tren ends ds in si sing ngle le ve vers rsus us co comp mple lex x cr crop op ha habi bita tats ts,, a fe few w have concentrated on elucidating the nature and dynamics of the trophic relationships between plants and herbivores and herbivores and the their ir nat natur ural al ene enemie miess in di dive versi rsifi fied ed agr agroec oecosy osyste stems. ms. Se Seve veral ral lin lines es of study have developed. • Crop-weed-insect interaction studies: Evidence indicates that
weeds influence the diversity and abundance of insect herbivores and associated natural enemies in crop systems. Certain flowerin flo wering g weed weedss (mos (mostly tly Umbe Umbellifer lliferae, ae, Legu Legumino minosae, sae, and Compo Com posita sitae) e) pla play y an imp impor ortan tantt eco ecolog logical ical rol rolee by har harbo borin ring g and supporting a complex of beneficial arthropods that aid in suppressing pest populations (Altieri, Schoonhoven, Schoonhoven, and Doll, 1977; Altieri and Whitcomb, 1979b, 1980). • Insect dynamics in annual polycultures: Overwhelming evidence suggests that polycultures support a lower herbivore load
than monocultures do. One factor explaining this trend is that relatively more stable natural-enemy populations can persist in polycultures due to the more continuous availability of food sources and microhabitats (Letourneau and Altieri, 1983; Helenius, 1989). The other possibility is that specialized herbivores are more likely to find and remain on pure crop stands that provide concentrated resources and monotonous physical conditions (Tahvanainen and Root, 1972). • Herbivores in complex perennial crop systems: Most of these studies have explored the effects of the manipulation of ground ground-cover vegetation on insect pests and associated enemies. The dataa ind dat indica icate te tha thatt orc orchar hards ds wit with h ric rich h flo floral ral un under dergro growth wth ex exhib hibit it a lower incidence of insect pests than clean-cultivated orchards, mainly because of an increased abundance and efficiency of predators and parasitoids (Altieri and Schmidt, 1985). In some cases, groundcover directly affects herbivore species that discriminate among trees with and without cover beneath. • The effects of adjacent vegetation: These studies have documented the dynamics of colonizing insect pests that invade crop fields from edge vegetation, especially when the vegetation is botanically related to the crop. A number of studies document the importance of adjoining wild vegetation in providing alternate food and habitat to natural enemies that move into nearby crops (Van Emden, 1965b; Wainhouse and Coaker, 1981; Altieri and Schmidt, 1986a). The available literature suggests that the design of vegetationmanage man agemen mentt str strate ategie giess mus mustt inc includ ludee kn know owled ledge ge and con consid sidera eratio tion n of (1) crop arrangement in time and space, (2) the composition and abundance of noncrop vegetation within and around fields, (3) the soil so il ty type pe,, (4 (4)) th thee su surr rrou ound ndin ing g en envi viro ronm nmen ent, t, an and d (5 (5)) th thee ty type pe an and d in inte tennsity of management. The response of insect populations to env environironmental manipulations depends upon their degree of association with one or more of the vegetational components of the system. Extension of the cropping period or planning temporal or spatial cropping sequences may allow naturally occurring biological control agents to attain higher population levels on alternate hosts or prey and to persist in the agricultural environment throughout the year.
Since farming systems in a region are managed over a range of energy inputs, levels of crop diversity, and successional stages, variations in insect dynamics are likely to occur and may be difficult to predict. However, based on current ecological and agronomic theory, low pest potentials may be expected in agroecosystems that exhibit the following characteristics: 1. High crop diversity through mixtures in time and space (Cromartie, 1981; Altieri and Letourneau, 1982; Risch, Andow, and Altieri, 1983; but see also Andow and Risch, 1985; Nafus and Schreiner, 1986) 2. Discontinuity of monoculture in time through t hrough rotations, use of short-maturing varieties, varieties, use of crop-free or preferred host-free periods, etc. (Stern, 1981; Lashomb and Ng, 1984) 3. Small, scattered fields creating a structural mosaic of adjoining crops cro ps and unc uncult ultiv ivated ated lan land d whi which ch po poten tential tially ly pro provid videe she shelte lterr and alternative food for natural enemies (V (Van an Emden, 1965a; Altieri and Letourneau, 1982) (Pests also may proliferate in these environments depending on plant species composition [Altieri and Letourneau, 1982; Slosser et al., 1984; Collins and Johnson, 1985; Levine, 1985; Lasack and Pedigo, 1986]. However, the presence of low levels of pest populations and/or alternate hosts may be necessary to maintain natural enemies in the area.) 4. Farms with a dominant perennial crop component. (Orchards are considered to be semipermanent ecosystems and more stable than annual crop systems. Since orchards suffer less disturbance and are characterized by greater structural diversity, diversity, possibilities for the establishment of biological control agents are generally higher, especially if floral undergrowth diversity is encour enc ourage aged d [H [Huf uffa faker ker and Mes Messen senger ger,, 197 1976; 6; Alt Altier ierii and Sch Schmid midt, t, 1985].) 5. High crop densities or presence of tolerable levels of weed background (Shahjahan and Streams, 1973; Altieri, Schoonhoven, and Doll, 1977; Andow, 1983b; Mayse, 1983; Buschman, Pitre, and Hodges, 1984; Ali and Reagan, 1985) 6. Hig High h gen geneti eticc di dive versi rsity ty res result ulting ing fro from m the use of va varie riety ty mix mixtur tures es or several lines of the same crop (Perrin, 1977; Gould, 1986; Altieri and Schmidt, 1987) By considering various spatial, temporal, and varietal features of cropping systems, Litsinger and Moody (1976) suggested the impli-
cations for pest suppression of various crop-management schemes (Fig (F igur uree 2. 2.6) 6).. Th Thes esee ge gene nera rali liza zatio tions ns ca can n se serv rvee in th thee pl plan anni ning ng of a ve veggetation-management strategy in agroecosystems; however, they must take into account local variations in climate, geography, geography, crops, local vegetation,, inputs, pest complexes, etc., which might increase vegetation i ncrease or decrease the potential for pest development under some vegetationmanagement conditions. The selection of component plant species can ca n al also so be cr crit itic ical. al. Sy Syst stem emati aticc st stud udies ies on th thee “q “qua uali lity ty”” of pl plan antt di dive verrsiffic si icati ation on wi with th re resp spec ectt to th thee ab abun unda danc ncee an and d ef effi fici cien ency cy of na natu tura rall en eneemies mi es ar aree ne need eded ed.. As po poin inted ted ou outt by So Sout uthw hwoo ood d an and d Way (1 (197 970) 0),, wh what at seem se emss to ma matte tterr is “f “fun unct ctio iona nal” l” di dive vers rsit ity y an and d no nott di div ver ersi sity ty pe perr se se.. Th This is highli hig hlight ghtss the imp import ortanc ancee of rec recog ogniz nizing ing tha thatt agr agreco ecosys system temss may not benefit from a “hit-and-miss” approach to diversification but require certain cer tain elem element entss of di dive versi rsity ty wh which ich,, on once ce ide identi ntifi fied, ed, cou could ld be ret retain ained ed or rei reintr ntrodu oduced ced.. Mec Mechan hanisti isticc stu studie diess to det determ ermine ine the un under derlyi lying ng ele ele-ments of plant mixtures which disrupt pest invasion and which favor colonization and population growth of natural enemies will allow more precise planning of cropping schemes and increase the chances of a beneficial effect beyond the current levels. It is important that changes in habitat diversity are purposely designed to obtain specific effects within the socioeconomic constraints of the agroecosystem.
PEST POTENTIAL RELATED RELATED TO CROP MANAGEMENT MANAGEMENT High Pest Potential
Low Pest Potential
CROP ARRANGEMENT IN TIME Monoculture Annual crop Long-maturing crop Continuous planting Asynchronous planting Season favorable to pest
Crop species rotation Perennial crop Short-maturing crop Discontinuous Discontinuou s planting Synchronous planting Synchronous Season unfavorable to pest
CROP ARRANGEMENT IN SPACE Row Ro w or stri strip p cro cropp pping ing Sole cropping Low planting density Large field Fields aggregated
Mixed int Mixed interc ercrop roppin ping g High planting density Small field Fields scattered
FIGURE 2.6. Hypotheti FIGURE Hypothetical cal trends of increased or decrea decreased sed pest potent potential ial in agroecosyst agroe cosystems ems depending on crop arrangement arrangement in time and/or space (afte (afterr Litsinger and Moody Moody,, 1976).
Chapter 3
Plant Diversity Insect Stability Plant Diversity and Insectand Stability in Agroecosystems in Agroecosystems ECOLOGICAL THEORY
Monocultures are dominated by a single plant species and, therefore, represent an extreme example of agroecosystems with low diversity.. Such systems are more susceptible to weather disasters, pest versity or disease outbreaks, and other catastrophes. A high degree of management and external inputs is required to maintain these types of agroecosystems. In contrast, many natural ecosystems appear to be more stable and less subject to fluctuations in populations of their component organisms. Ecosystems with higher diversity are more stable because they exhibit higher • resistance, or an ability to avoid or withstand disturbance; and • resilience, or an ability to recover following following disturbance.
Diversi Dive rsity ty is on only ly on onee mea measur suree of eco ecosys system tem com comple plexit xity y. The com commu mu-nity of organisms becomes more complex when a larger number of differ dif ferent ent kin kinds ds of or orga ganis nisms ms is inc includ luded, ed, wh when en the there re are mor moree int intera eracctions among organisms, and when the strength of these interactions increases. As diversity increases, so do opportunities for coexistence and beneficial interference between species that can enhance agroecosystem sustainability. Diverse systems encourage complex food webs that entail more potential connections and interactions among members, and many alternative paths of energy and material flow through it. Thus, a more complex community is more stable, and much data supports this idea. Nevertheless, ecologists have for years debated the assumption that increased diversity fosters stability. Critical theoretical reviews
on this subject are available (Watt, 1973; Van Emden and Williams, 1974; Goodman, 1975; Murdoch, 1975), as are reviews that use agricultural examples to bolster the theory (Pimentel, 1961; Root, 1973; Dempster and Coaker, 1974; Litsinger and Moody, 1976; Perrin, 1977). Regardless of those discussions, research has shown that mixing certain plant species with the primary host of a specialized herbivore gives a fairly consistent result (Figure 3.1): specialized species usually exhibit higher abundance in monocultures than in polycultures. In a review of 150 published investigations, Risch and colleagues (1983) found evidence to support the notion that specialized insect herbivores were less numerous in diverse systems (53 percent of 198 cases). Another comprehensive review by Andow (1991a) identified 209 20 9 pub publis lished hed stu studie diess tha thatt dea deall wit with h the ef effec fects ts of ve veget getati ation on di dive versi rsity ty in agroecosystems on herbivorous arthropod species. Fifty-two percent ce nt of th thee 28 287 7 to tota tall he herb rbiv ivor oree sp spec ecie iess ex exam amin ined ed in th thes esee st stud udie iess we were re found to be less abundant in diversified systems than in monocultures, while only 15.3 percent (44 species) exhibited higher densities in polycultures (Table 3.1). In a more recent review of 287 cases,
FIGURE 3.1. Consistent population trends of specialized herbivores in monocultures and polycultures where host plants are mixed with nonhost plants (after Strong, Lawton, and Southwood, 1984).
TABLE 3.1. Numbers of Arthropod Species with Particular Responses to Additive and Substitutive Polycultures a Pop opul ulat atiion de dens nsiity of ar arth thro ropo pod d sp spec ecie ies s in po pollyc ycul ultu ture re co comp mpar ared ed to monoculture: Variableb
Higher
No Change
Lower
58 (20.2)
44 (15.3)
36 (12.5)
149 (51.9)
Monophagous
42 (19.1)
17 (7.7)
31 (14.1)
130 (59.1)
Polyphagous
16 (23.9)
27 (40.3)
5 (7.5)
19 (28.4)
Natural enemies
33 (25.6)
68 (52.7)
17 (13.2)
12 (9.3)
Predators
27 (30.3)
38 (42.7)
14 (15.7)
11 (12.4)
Parasitoids
6 (15.0)
30 (75.0)
3 (7.5)
1 (2.5)
Herbivores
Source: After Source: After Andow, 1991a. aPercentage of total number of species is in parentheses. bA va vari riab able le re resp spon onse se me mean ans s th that at an ar arth thro ropo pod d sp spec ecies ies did no nott co cons nsist isten entl tly y hav have e
a higher or lower population density in polycultures compared to monocultures when the species response was studied several times.
Helenius (1998) found that reduction of monophagous pests was greater in perennial systems and that the reduction of polyphagous pest numbers was less in perennial than in annual systems (Table 3.2). Four main ecological hypotheses have been offered to explain lower pest-population loads in multispecies plant associations. Associational Resistance
Ecosystems in which plant species are intermingled possess an associational resistance to herbivores herbivores in addition to the resistance of individual plant species (Root, 1975). Tahvanainen and Root (1972) suggest that, in addition to their taxonomic diversity, polycultures have a relatively complex structure, structure, chemical environment, environment, and and associated patterns of microclimates. These factors of mixed vegetation
TABLE 3.2. Percentage of Cases with Lowered Numbers of Arthropod Herbivores in Crops of Increased Vegetational Diversity Than in Monocrops Monophagous Species
Polyphagous Species
All
Annual systems
53.5
33.3
48.5
Perennial systems
72.3
12.5
60.5
All
59.1
28.4
51.9
Source: From Source: From Andow, 1991a.
work synergistically synergistically to produce an “associational resistance” to pest attack. In stratified vegetation, insects may experience difficulty in locating and remaining in small, favorable spots if microclimatic conditions are highly fractionated. Thus, diversity ameliorates the herbivore herbiv ore pressure on the crop system as a whole. Host-plant finding by insect pests often involves olfactory mechanisms, and host plants grown in association with unrelated plants may be an important component in the defense against herbivores, the nonhost-plant odors leading to disruption of the host-finding behavior of the insect based on odor cues. This type of protection derives from the masking effect of the nonhost-plan nonhost-plantt odors on the odors emitted by the host plants. This effect has been demonstrated with collards interplanted with tomato or tobacco on the flea beetle, Phyllotreta cruciferae (Root, 1973), and the diamondback moth, (Litsinger nger and Moody, 1976), and the carrot fly on Plutella xylostella (Litsi carrots interplanted with onions (Uvah and Coaker, 1984). In the latter example, reduction in infestation occurred only when the onion leaves were expanding and not when the plants had started to bulb, suggesting that the masking odor emanated from young leaves only. Other aromatic herbs have been claimed, mostly by organic gardeners, to be repellent to insect pests of vegetable crops, but but little experimental work has been done to substantiate these claims. Natural Enemy Hypothesis
This proposition predicts that there will be a greater abundance and an d di dive vers rsity ity of na natu tura rall en enem emie iess of pe pest st in inse sect ctss in po poly lycu cult ltur ures es th than an in monocultures (Root, 1973). Predators tend to be polyphagous and
have broad habitat requirements, so they would be expected to encounter a greater array of alternative prey and microhabitats in a heterogen ero geneou eouss en envir vironm onment ent (Ro (Root, ot, 197 1975) 5).. An Annua nuall cro crop p mo monoc nocult ultur ures es do not provide adequate alternative sources of food (pollen, nectar, prey), shelter, and breeding and nesting sites for the effective performance man ce of nat natur ural al ene enemie miess (Ra (Rabb bb,, Sti Stinne nnerr, and va van n den Bos Bosch, ch, 19 1976) 76).. The natural-enemy hypothesis has been stated in the following way: 1. A greater diversity of prey and microhabitats is available within comple com plex x en envir viron onmen ments. ts. As a res result ult,, rel relati ative vely ly sta stable ble po popu pulati lations ons of generalized predators can persist in these habitats because they can exploit the wide variety of herbivores which become available availab le at different times or in different microhabitats (Root, 1973). 2. Specialized predators are less likely to fluctuate widely because the refuge provided by a complex environment enables their prey to escape widespread annihilation (Risch, 1981). 3. Diver Diverse se habitats offer many important requisites for adult predator at orss an and d pa para rasi site tes, s, su such ch as ne necta ctarr an and d po polle llen n so sour urce ces, s, wh whic ich h ar aree not available in a monoculture, reducing the probability that they will leave or become locally extinct (Risch, 1981). According to Root’s enemies hypothesis, generalist and specialist natural enemies are expected to be more abundant in polycultures and, therefore, more effectively suppress herbivore population densities in polycultures than in monocultures. Generalist predators and parasitoids should be more abundant in polycultures than monocultures because (1) they switch and feed on the greater variety of herbivores that become available in polycultures at different times during the growing season; (2) they maintain reproducing populations in polycultures although in monocultures only males of some parasitoids are produced; (3) they can utilize hosts in polycultures that they would normally not encounter and use in monocultures; (4) they can exploit the greater variety of herbivores available in different microhabitats in the polycultures; and (5) prey or hosts are more abundant or more available in polycultures (Smith and McSorley, 2000).
Specialist predator and parasitoid populations are expected to be more abundant and effective in polycultures than monocultures becaus ca usee pr prey ey or ho host st re refu fuge gess in po poly lycu cultu lture ress en enab able le th thee pr prey ey or ho host st po poppulations to persist, which stabilizes predator-prey predator-prey and parasitoid-host interactions. In monocultures, predators and parasitoids drive their prey or host populations to extinction and become extinct themselves t hemselves shortly thereafter. Prey or host populations will recolonize these monocultures and rapidly increase (Andow, 1991a). Finally, both generalist and specialist natural enemies should be more abundant in polycultures than monocultures because more pollen le n an and d ne nect ctar ar re reso sour urce cess ar aree av avai aila labl blee at mo more re ti time mess du duri ring ng th thee se seas ason on in the complex systems (Altieri and Letourneau, 1982). Resource Concentration
Insect populations can be influenced directly by the concentration or spatial dispersion of their food plants. There can be a direct effect of associated plant species on the ability of the insect herbivore to find and utilize its host plants. Many herbivores, particularly those with narrow host ranges, are more likely to find and remain on hosts that are growing in dense or nearly pure stands (Root, 1973), which are thus providing concentrated resources and monotonous physical conditions. For any pest species, it is the total strength of the attractive stimuli thatt det tha determ ermine iness the res resour ource ce con concen centra tratio tion, n, and thi thiss va varie riess wit with h int intereracti ac ting ng fa fact ctor orss su such ch as th thee de dens nsity ity an and d sp spat atia iall ar arra rang ngem emen ents ts of th thee ho host st plant and the interfering effects of nonhost plants. Consequently, Consequently, the lower the resource (host plant) concentration, the more difficult it will wi ll be fo forr th thee in inse sect ct pe pest st to lo loca cate te a ho host st pl plan ant. t. Re Rela lati tive ve re reso sour urce ce co conncentration also increases the probability of the pest species leaving the habitat once it has arrived; for instance, the pest may tend to fly sooner and further after landing on a nonhost plant, which can result in a hig higher her emi emigra gratio tion n rat ratee fro from m pol polycu ycultu ltures res tha than n fro from m mon monocu ocultu ltures res (Andow, 1991a). Such an “emigration effect” should be evident in polycultures when the pest’s trivial movement involves (1) mistakenly alighting on nonhost plants, (2) moving off nonhost plants more frequently than moving off off host plants, and (3) being at risk of leaving the crop area during movement.
Using diffusion models (calculating movement coefficients and disappearance rates), Power (1987) compared movement rates of the t he leafhopper Dalbulus maidis in maize monocultures and maize intercropped with beans. The rate of movement along rows and the disappearance rate were twice as rapid in the intercrop as in the monoculture, but the rate of movement across rows was dramatically reduced in the intercrop. As would be expected from the disruptive crop, movement along the rows was more rapid when an effective block of beans was between the rows, suggesting a more rapid rate of disappearance in the intercrop. This hypothesis has also been experimentally tested by Bach (1980b) and Risch (1981). Plant “Apparency”
Most crops are derived from early successional herbs that escaped from fro m her herbi bivo vores res in spa space ce and tim timee (Fe (Feen eny y, 197 1976) 6).. The ef effec fecti tive venes nesss of natural crop-plant defenses is reduced by present agricultural methods: monocultures make crop plants “more apparent” to herbivores than were their ancestors. In agriculture, the “apparency” of a crop plant is increased by close association with related species (Feeny, 1977). Therefore, crop plants grown in monoculture are subjected to artificial conditions for which their qualitative chemical and physical defenses are inadequate. The theory of plant chemical defense developed by Feeny (1976) and Rhoades and Cates (1976) discusses the classification of plants as “apparent or predictable” and “unapparent or unpredictable” and the implications of such divisions for agricultural crops in relation to pest susceptibility. susceptibility. Crop apparency can be increased or decreased either ei ther by intracrop diversity div ersity or by high high-den -density sity crop cropping ping.. Veget egetation ational al back backgrou ground nd amon among g crops can have different effects on the associated insect fauna depending on the situation the pest is adapted to exploit. Pieris rapae and Bre Brevicoryne vicoryne brassicae favor mainly open successional habitats and are more attracted to host plants that stand out against a bare soil; in contrast, the fruit fly dwells in dense stands and would be less attracted to open plantings of grasses and cereals (Burn, Coaker, and Jepson, 1987).
Critical Assessments of the Hypotheses
Of the four hypotheses, insect ecologists have tested more systematically the “enemy hypothesis” (predators and parasites are more effective in complex systems) and the “resource concentration hypothesis es is”” (s (spe peci cial alis istt he herb rbiv ivor ores es mo more re ea easi sily ly fi find nd,, st stay ay in in,, an and d re repr prod oduc ucee in simple sim ple sys system tems). s). Tabl ablee 3.3 sum summar marize izess the mai main n fi findi nding ngss and con conclu clu-sions of major reviews that explore the ecological implications of both hypotheses. Finch and Collier (2000) argue that both hypotheses fail to produce du ce a ge gene nera rall th theo eory ry of ho host st-p -plan lantt se sele lecti ction on,, an and d th thus us,, th they ey pr prop opos osee a new theory based on “appropriate/inappr “appropriate/inappropriate opriate landings.” landings.” This theory is based on the fact that during the host-plant-finding host-plant-finding phase, the searching insects land indiscriminately on leaves of host plants (appropriate landings) and nonhost plants (inappropriate landings) but avoid avo id landing on brow brown n surfaces, such as soil. This process seems to be governed by visual stimuli, as opposed to chemical cues, a central link within the three-link chain of events that governs host-plant selection. In a study of eight different phytophagous species, Finch and Kienegger (1997) found that the ability of each species to find cabbage was affected adversely, though to differing degrees, when their host plants were surrounded surrou nded by clover. clover. From a crop-protection crop-protecti on standpoint, the more host plants are concentrated in a given crop area, the greater the chance an insect has of finding a host plant. Current agricultural methods are exacerbating pest control problems, as “baresoil so il”” cu cult ltiv ivat atio ion n en ensu sure ress th that at cr crop op pl plan ants ts ar aree ex expo pose sed d to th thee ma maxi ximu mum m pest-insect attack possible in any given locality. One mai main n cri critic ticism ism of the de deve velop loping ing the theori ories es of ins insect ect-pl -plant ant int intereractions in diverse agroecosystems has been advanced by Price and colleagues (1980). They contend that this theory has involved only the plant (first trophic level) and the herbivore (the second level) but hass no ha nott se seri riou ousl sly y co cons nsid ider ered ed th thee na natu tura rall en enem emie iess th that at pr prey ey on on,, or pa para ra-sitize, the herbiv herbivores ores (third trophic level); this level should be viewed as part of a plant’s battery of defenses against herbivores. This is especially relevant in multispecies crop associations because (1) the herbivore-enemy interaction of one plant species can be influenced by the presence of associated plants, and (2) the herbivore-enemy herbivore-enemy interaction of one plant species can be influenced by the presence of herbi her bivo vores res on ass associ ociate ated d pla plant nt spe species cies.. In man many y cas cases, es, ent entomo omopha phagou gouss
TABLE 3.3. Main Findings and Conclusions of Reviews of Enemy Hypothesis (NE) and Resource Concentration (RC) Hypothesis Review
Main Findings
Main Conclusions
Risch Risc h et al. (19 (1983) 83)
150 stu studie dies, s, 198 her herbiv bivore ore spe specie cies. s. “Div “Divers erse” e” systems: herbiv herbivore ore populations populations lower in 58 percent of cases, higher in 18 percent, no change in 9 percen percent, t, and variab variable le in 20 percent.
Resource concentration hypothesis (Root, 1973) most likely explanation but mechanisms mechanisms rarely rarely studied, e.g., host-plant finding behavior, predation rate, etc.
Risch (1981)
Review of mechanisms.
Resource concentration hypothesis in annual annual systems, systems, natural enemies in perennial perennial systems. systems.
Karei Ka reiva va (1 (198 983) 3)
Modeliling Mode ng an and d fie field ld stu studi dies es sh show ow th that at si simp mple le systems can be stable. Specie Species s number and “connectance” “conne ctance” are important in stability. stability.
Resource concentration (RC) hypothesis supported rather than an effect via the community’s trophic structure.
Russell Russ ell (19 (1989) 89)
Reviewed Revie wed nin ninete eteen en stu studie dies s whi which ch ex expli plicitl citly y tested the natural enemies hypothesis (NE).
Of the 19 studies, mortality rates from predators and parasitoids parasi toids in “diverse” “diverse” systems were higher in 9, lower in 2, unchanged unchan ged in 3, and variable variable in 5. When mechanisms mechanisms are evaluated, natural enemy hypothesis more likely to be confirmed. This hypothesis hypothesis and resource concentration concentration hypothesis “complementary. “complementary.””
Baliddawa Balidd awa (1985)
Reviewed Revie wed crop/we crop/weed ed systems and intercr intercrops. ops.
Crop/weed studie Crop/weed studies: s: 56 percen percentt of pest decrea decreases ses caused by naturall enemies. natura enemies. Intercro Intercrops: ps: 25 percent. Intercro Intercrops ps probably probably slowed slowe d colon colonization ization by pests. (Plant (Plants s may limit predator movement also—see text.)
Ando An dow w (1 (199 991a 1a))
Classifi Class ified ed pe pests sts as mo mono noph phag agou ous s or polyphagous. 254 herbivore species reviewed.
56 percent of herbi herbivore vore species’ species’ popul populations ations were lower in diverse systems, compared with 66 percent for monophagous herbivores herbiv ores and 27 percent for polyph polyphagous agous ones. Predat Predator or numbers numbe rs were highe higherr in 48 percen percentt of diverse habitats studied, as were 81 percen percentt of parasi parasitoids’ toids’.. Conclud Concluded ed that both hypotheses applied but RC hypothesis most important.
Cromartie Crom artie (19 (1991) 91)
Reviewed Revie wed data from ann annual ual row crop crops s and from perennial crops.
RC hypothesis most likely for annual crops, with NE hypothesis most likely for perennial (and some annual) crops.
insects are directly attracted to particular plants, even in the absence of host or prey, or by chemicals released rel eased by the herbivore’ her bivore’ss host plant or ot othe herr as asso soci ciat ated ed pl plan ants ts.. In Inter terac acti tion onss ca can n be ve very ry co comp mple lex x an and d ty typi pi-cally involve different trophic levels and several species of plants, herbivores, and natural enemies, as illustrated by Price and colleagues (1980) (Figure 3.2). An additional criticism raised by Vandermeer Vandermeer (1981) is that none of the hypotheses includes what Vandermeer calls the trap-cropping effect. The idea is that the t he presence of a second crop in the t he vicinity of a principal crop attracts a pest which would otherwise normally attack the principle crop. Corn, when planted in strips in cotton fields, reportedly may attract the cotton bollworm away from the cotton
FIGURE FIGUR E 3.2 3.2.. Int Intera eracti ctions ons in a com commun munity ity of fo four ur tro trophic phic le leve vels ls inv involv olving ing semiochemicals. Arrows are placed against the responding organism. Thick solid lines and solid arrows illustrate attraction to a stimulus (e.g., 1, 4, 11, 24). Thin solid lines and open arrows illustrate repulsion (e.g., 3, 13, 17, 26). Thin dashed lines show indirect effects such as interference with another response (e.g., 2, 12, 19) (after Price et al., 1980).
(Lincoln and Isley, 1947). 1947). Sorghum may act as an effecti effective ve trap for the stem borer Chilo partellus in India. Trap crops have been used to control jassids in cotton (Ali and Karim, 1989). In Central America, Rosset and colleagues (1985) found that the attack by the armyworm (Spodoptera sunia) totally destroyed a monoculture of tomatoes, although an intercrop of tomatoes and beans was effective effective in reducing the attack to virtually zero. It was clear that the caterpillars were being attracted to the beans in the intercrop that acted as a trap crop. More specific examples exampl es of trap crops that have proven effective to attract insect pests away from target plants in a variety of cropping systems are presented in Table 3.4. A remarkable example of trap cropping from Canada concerns the use of sterile bromegrass as a trap plant for the wheat stem sawfly Cephus cinctus. The bromegrass traps a large proportion of the incoming sawflies. They lay their eggs there, and the larvae bore into the stems. The elegant feature of this system is that the larvae die in the brome stems not before pupating but after larval parasitoids have emerged. emerged. No control of the insect on thee tr th trap ap pl plan ants ts is th ther eref efor oree ne nece cess ssar ary y. Mo More reov over er,, th thee gr gras asss ac acts ts as a fi fillter which converts pests into beneficial biomass (Van Emden and Dabrowski, 1997). According to Vandermeer (1989), trap crops act preferentially to attract generalist herbivores in such a way that the plant to be protected is not as likely to be directly attacked. He proposes some general er al mo mode dels ls to op optim timize ize th thee le leve vell of tr trap ap-c -cro rop p co conc ncen entr trat atio ion n to in incr crea ease se the trap’s ability to draw pests away from the crop (local-attraction probability) as opposed to attracting pests from afar (regional-attraction probability) (Figure 3.3).
THEORY DILEMMAS
Sheehan (1986) and Russell (1989) have questioned the universal val alid idity ity of th thes esee th theo eori ries es th that at ex expl plai ain n th thee ef effe fect ctss of ag agro roec ecos osys yste tem m di di-versification versif ication on searching behavior and success of arthropod natural enemies and claim that such interactions are still poorly understood. None of the proposed hypotheses really include all of the mechanisms that are known to operate in the general area of diversity and pest attack (Vandermeer, 1989). Fifty percent of the eighteen studies reviewed by Russell (1989) found higher herbivore mortality rates
TABLE 3.4. Examples of Trap Trap Crop Systems Successfully Applied in Agricultural Practice Controlled Pests
Main Crop
Trap Crop
Location
Lygus bugs cotton (Lygus hesperus, Lygus elisus)
alfalfa
California
Cotton boll weevil (Anthonomus grandis)
cotton cotton
cotton cotton
United States Nicaragua
Stinkbugs (Nezara viridula Euschistus spp., Euschistus spp., Acrosternum hilare, Piezodorus guildinii ) guildinii )
soybeans soybeans soybeans
soybeans soybeans soybeans cowpea
United States Brazil Nigeria
Mexican bean beetle (Epilachna varivestis)
soyb so ybea eans ns
snap sn ap be bean ans s
Uni nitted St Stat ates es
Bean leaf beetle (Cerotoma trifurcata)
soybeans
soybeans
United States
Colorado potato beetle (Leptinotarsa decemlineata)
potato potato
potato potato
USSR Bulgaria
Blossom beetle (Meligethes aeneus)
rape cauliflower
rape marigold
Finland
Pine shoot beetle (Tomicus (T omicus piniperda)
pine trees
pine logs
Great Britain
Spruce bark beetle (Ips typographus)
spruce
spruce trees logs
Europe
Source: After Source: After Hokkanen, 1991.
from predation or parasitism in diverse systems, but according to Russell, a lack of adequate control in all but one study prevented prevented researchers from concluding that the difference in mortality was what really reduced the number of herbivores in complex systems. He further argues that the enemies hypothesis and the resource concentration hypothesis act as complementary mechanisms in reducing numbers of herbivores in polycultures, and both should be enhanced simultaneously to achieve maximum control. Accord Acc ording ing to She Sheeha ehan n (1 (1986 986), ), the ene enemie miess hy hypot pothes hesis is is sim simpli plistic stic in se seve veral ral res respec pects. ts. Victi ictim m loc locatio ation n by gen genera eralis listt ene enemie miess may be hin hin--
Trap crop not feasible Trap crop feasible Region in which trap crop will be successful
Trap concentration Optimum trap crop
Nonfunctional trap crop
FIGURE FIGU RE 3. 3.3. 3. Ra Rati tio o of ge gene nera rall to loc local al at attr trac acti tion on pr prob obab abili iliti ties es as a fu funct nction ion of tr trap ap concentration, illustrating the region of trap-crop feasibility (after Vandermeer, 1989).
dered by increased plant density or patchiness in diverse agricultural system sys tems. s. In fac fact, t, cro crop p di dive versi rsifi ficat cation ion may red reduce uce ene enemy my sea search rching ing efficiency fici ency and destab destabilize ilize pred predator ator-pre -prey y intera interaction ctions. s. Speci Specialist alist enemi enemies, es, often important in biological control programs, may be particularly sensit sen sitiv ivee to ve veget getati ation on tex textur ture. e. Pes Pestt con contro troll by spe specia cialis listt ene enemie miess may be more effective in less diverse agroecosystems if concentration of host plants increases attraction or retention of these enemies. Thus, it is possible that certain specialist enemies may not necessarily respond to habitat diversification in the same way as generalists. Sheehan (1986) suggested that specialist parasitoids might be less abundant in polycultures than monocultures because (1) chemical cues used in host finding will be disrupted and the parasitoids will be less able to find hosts to parasitize parasitize and feed upon in polycultures, polycultures, and (2)) th (2 thee in indi disti stinc nctt bo boun unda dary ry at th thee ed edge gess of po poly lycu cultu lture ress wi will ll be ha hard rd to recognize and they will be more likely to leave polycultural habitats
than monocultures. In addition, Andow and Prokrym (1990) showed that structural complexity, or the connectedness of the surface on which the parasitoid searches, can strongly influence parasitoid hostfinding rates. An implication of their study is that structurally complex polycultures would have have less parasitism than structurally simple monocultures. Facto Fa ctors rs tha thatt inc increa rease se imm immigr igrati ation on to and dec decrea rease se emi emigra gratio tion n fr from om host-plant areas by specialist enemies (e.g., large patch size, close plan pl antt sp spac acin ing, g, th thee pr pres esen ence ce of sp spec ecif ific ic ch chem emica icall or vi visu sual al st stim imul uli, i, an and d lower chemical or structural diversity of associated vegetation) may cause those enemies to remain longer and hunt more effectively in simple than in diverse agroecosystems, at least in those that are not too extensive. Using a mathematical model analyzing the role of movement in the response of natural enemies to diversification, Corbett and Plant (1993) predicted that interplanted strip vegetation can act as either a sink or a source of natural enemies, depending on the mobility of the natu na tura rall en enem emy y an and d th thee sp spec ecif ific icss of sy syst stem em de desi sign gn.. Fo Forr a hi high ghly ly mo mobi bile le predator, interplanted vegetation might be converted from a severe sink into a valuable resource simply by having interplanted vegetation germinate before the crop germinates, allowing early colonization by the predator. When the crop and interplanted vegetation germinate simultaneously, the model predicts that interplants act as a sink, but the magnitude of this effect will vary with natural enemy mobility, which determines spatial distributional patterns of predators and parasites. Another shortcoming of the theory is that, so far, enemy and resource concentration hypotheses provide the basis to predict herbivore response in polycultures if 1. th thee he herb rbiv ivor oree is a sp spec ecia ialis listt or ex expl ploi oits ts a na narr rro ow ra rang ngee of pl plan ants ts;; 2. the polyculture is composed of a preferred host plant and one or more nonhost plants; or 3. host host and non nonhost host plants overlap overlap in time and space within the mixture. Given these limitations, not all results from studies of agricultural diversification diver sification leading to reduced pest populations can be adequately explained by the two hypotheses as they stand. One such case is a
study by Gold (1987) on the effects of intercropping cassava with cowp co wpeas eas on the pop popula ulatio tion n dyn dynami amics cs of the cas cassa sava va whi whitef teflie liess ( Aleur Aleurotrachelus socialis and Trialeurodes variabilis variabi lis) in Colombia. As expected, intercropping reduced egg populations of both species of whitefly on cassava (a twelve-month crop), but the reductions were residual, persisting up to six months after the t he harvest of the cowpea, which lasts in the field only three to four months (Figure 3.4). The natural enemy hypothesis was rejected as a mechanism explaining reduced herbivore load, because predators were more abundant in monocultures than in polycultures, showing a numerical response. The resource concentration hypothesis could explain lower whitefly densities during the period that cowpea was present but cannot explain the reductions in whitefly populations observed long after the removal of the cowpea. Instead, it appeared that intercrop competition caused a reduction in cassava size or vigor that persisted through the remainder of the season. Thus, whitefly populations in polycultures were a function of host-plant selection and/or tenure time which, in turn, was related to host condition. Another case is a study by Altieri and Schmidt (1986b), which found lower densities of the specialist herbivore Phyllotreta cruci ferae on broccoli mixed with another crucifer host plant, the wild mustard Brassica kaber. Densities of P. cruciferae were greater on
FIGURE 3.4. Expected population response (egg densities) of the whiteflies Aleurotrachelus socialis and socialis and Trialeurodes variabilis variabilis in in cassava and cowpea mixtures in Colombia (after Altieri, 1991a).
brocco broc coli li pl plan ants ts gr gro own in mo mono nocu cultu lture ress th than an in po poly lycu cult ltur ures es bu butt no nott so on a per-plot basis. The differences in beetle abundance on a perplan pl antt ba basi siss we were re ba basi sica call lly y du duee to th thee fa fact ct th that at th thee be beet etle less co conc ncen entr trate ated d more on wild mustard than on broccoli in the mixture. This Th is pr pref efer eren ence ce ha hass a ch chem emic ical al ba basi siss si sinc ncee wi wild ld mu must star ard d ha hass hi high gher er concentrations of glucosinolates than does broccoli, a strong beetle attractant. Thus, in this case, the differences in beetle abundance caused by diversity resulted from differenc differences es in beetle-feeding preferences rather than from differences in colonization, reproduction, or preda pr edatio tion. n. Ob Obvio viousl usly y, the these se tre trends nds do no nott con confor form m to the ass assump umptio tions ns of either of the two hypotheses. These two case studies do not necessarily contradict the two hypothe po theses ses bu butt ins instea tead d pro provid videe add additio itional nal ex expla planat nation ionss to cur curren rentt the theor ory y and call for more caution and flexibility since the responses of herbivores to vegetational diversity are not uniform and cannot always be explained by diversity per se. In fact, differences in tenure time or movement patterns between monocultures and polycultures are not always evident, although interplot differences in abundan abundance ce persist. As suggested by the studies mentioned, other factors such as visual cues, cue s, mic microc roclim limatic atic cha chang nges, es, fee feedin ding g pr prefe eferen rences ces,, or dir direct ect ef effec fects ts on hostho st-pla plant nt vig vigor or cou could ld inf influe luence nce hab habita itatt loc locati ation on and and/or /or sea search rching ing behavior of both herbivores and natural enemies. Despite all the experimental studies described here, we still have not been able to deve develop lop a predictive theory that enables us to t o determine what specific elements of biodiversity should be retained, added, or eliminated to enhance natural pest control. In a few cases, the simple addition of one element of diversity is all that is needed to ensure biological control of a pest species (e.g., incorporating blackberries in vineyards to control Erythroneura leafhoppers in California) (Doutt and Nakata, 1973). At times, all that is needed is to halt insecticide treatments to restore ecosystem regulating functions, as wass th wa thee ca case se in Co Cost staa Ri Rica can n ba bana nana na pl plan anta tati tion ons, s, wh wher eree af afte terr tw two o ye year arss of nontreatment, major insect pests decreased and many former insect pests nearly disappeared. After ten years, most pests were under total biological control (Stephens, 1984). Recent studies comparing arthropod fauna in organic and conventional farming systems confir irm m th thee be bene neffits of pe pest stic icid idee re remo mov val on th thee di dive vers rsit ity y of be bene nefi fici cial al ar ar-thropods on the foliage and soils (Paoletti, Stinner, and Lorenzoni, 1989).
We also know little to suggest whether the ecological mechanisms proposed by the hypotheses work at a more regional level (i.e., at the level of agroecosystem mosaics interspersed among natural vegetation ti on)) an and d th that at pe pest st pr prob oble lems ms wi will ll di dimi mini nish sh du duee to sp spat atial ial an and d te temp mpor oral al heterogeneity of agricultural landscapes. Some studies suggest that the vegetational settings associated with particular crop fields influence the kind, abundance, and time of arrival of herbivores and their natural enemies (Gut et al., 1982; Altieri and Schmidt, 1986a). In perennial orchards (e.g., pear and apple) at temperate latitudes, a di dive vers rsee co comp mple lex x of pr pred edat ator orss us usua ually lly de deve velo lops ps ea earl rly y in th thee se seas ason on in orchards surrounded by woodlands. In these systems, main pests (e.g., Psy etc. c.)) ar aree qu quic ickl kly y re redu duce ced d an and d Psylla lla pyr pyrico icola, la, Cyd Cydia ia pom pomone onella, lla, et main ma inta tain ined ed at lo low w le lev vels th thro roug ugho hout ut th thee se seas ason on.. In co cont ntra rast st,, ea earl rly y se seaason predators are absent in more extensive extensive commercial orchards, and therefore, pest pressure is more intense (Croft and Hoyt, 1983). Assuming that these trends also occur in the tropics, t ropics, one would expect that in certain tropical agroecosystems (e.g., shifting cultivation in the lowland tropics), forest and bush fallows have potential value in cont co ntro rolli lling ng pe pest sts. s. Cl Clea eari ring ng sm smal alll pl plot otss in a ma matr trix ix of se seco cond ndar ary y fo fore rest st vegetation may permit easy migration of natural enemies from the surrounding surroundin g jungle (Matteson, Altieri, and Gagne, 1984). This positive potential role of natural vegetation on biological control is expected to change in view of current deforestation rates and modernization trends toward commercial monocultures.
Chapter 4
Insect Manipulation Insect Manipulation Through Weed Management Through Weed Management
The presence of weeds within or around crop fields influences the dynam dy namics ics of the cro crop p and ass associ ociate ated d bio biotic tic com commun munitie ities. s. Stu Studie diess ov over er the past thirty years have produced a great deal of evidence that the manipulation of a specific weed species, a particular weed-control practice, or a cropping system can affect the ecology of insect pests and associated natural enemies (Van Emden, 1965b; Altieri, Schoonhoven, and Doll, 1977; Altieri and Whitcomb, 1979a,b; Thresh, 1981; 19 81; Willi illiam, am, 19 1981; 81; Nor Norris ris,, 19 1982; 82; An Ando dow w, 19 1983a 83a). ). Weed eedss ex exert ert a direct biotic stress on crops by competing for sunlight, moisture, and some nutrients, thus reducing crop yields. Weeds indirectly affect crop plants through positive and/or negative effects on insect herbivores and also on the natural enemies of herbivores (Price et al., 1980). Herbivore-natural enemy interactions occurring in a crop system can be influenced by the presence of associated weeds or by t he presence of herbivores on associated weed plants (Altieri and Letourneau, 1982). On the other hand, herbivores can mediate the interacti ac tion on be betw twee een n cr crop opss an and d we weed eds, s, as in a na natu tura rall co comm mmun unity ity wh wher eree th thee competitiveness competitiv eness of two plant species was altered substantially by the selective feeding of a foliage-consuming beetle (Bentley and Whittaker,, 1979). Such relationships have been little taker li ttle explored in agricultural systems. In this chapter, the multiple interactions among crops, weeds, herher bivores, and natural enemies and, in particular, weed ecology and management that affect the dynamics of insect populations and thus crop health are discussed. The stress imposed on crops by weeds is viewed beyond the mere competitive interaction, incorporating an analysis of three trophic-le trophic-level vel system interactions.
WEEDS AS SOURCES OF INSECT PESTS IN AGROECOSYSTEMS
Weeds are important hosts of insect pests and pathogens in agroecosystems. Van Emden (1965b) cites 442 references relating to weeds as reservoirs of pests. One hundred such references concern cereals. A series of publications concerning weeds as reservoirs for organisms affecting crops has been published by the Ohio Agricultural Research and Development Center. More than seventy families of arthropods affecting crops were reported as being primarily weed associated (Bendixen and Horn, 1981). Many pest outbreaks can be traced to locally abundant weeds belonging to the same family as the affected crop plants. Many pest insects are sufficiently polyphagous polyphagous that weeds unrelated to the crop may also be pest reservoir reservoirs. s. For exAphis is gos gossyp sypii ii fe ample, Aph feed edss on ov over er tw twen enty ty un unre rela lated ted we weed ed sp spec ecies ies in and an d ar arou ound nd co cott tton on fie ield lds. s. De Deta tail iled ed exa xamp mple less of th thee ro role le of we weed edss in th thee epidemiology of insect pests and plant diseases can be found in Thresh (1981), especially for crop diseases transmitted from weeds to adjacent crop plants by insect i nsect vectors. A major example provided is the role of the leafhopper Circulifer tenellus in transmitting the beet curly top virus from the western states of Utah and Colorado, where the vector itself is unable to survive the winter. winter. Weedy plants near crop fields can provide the requisites for pest outbreaks. The presence of Urtica dioica in the host layer of noncrop habitats surrounding carrot fields was the most important factor determining high levels of carrot fly larval damage to t o adjacent carrots (Wainhouse and Coaker, 1981). Adult leafhoppers invade peach orchards from edge vegetation and subsequently colonize trees under which groundcover is composed of preferred wild hosts (McClure, 1982). Plantains ( Plantago spp.) provide alternative altern ative food for the rosy apple aphid Dysaphis plantaginea, an important pest of apple orchards in England. The rosy apple aphid spends most of the summer on plantains, returning to apples in late summer. The dock sawfly Ametrastegia glabrata normally feeds on docks ( Rumex spp.) and knotgrass (Polygonum spp.), and the larvae of the last generation can move on to adjacent apple trees and bore into fruits or shoot tips (Altieri and Letourneau, 1982). Certain grasses can act as hosts for cereal pests, and these species should be excluded from undersowings. Bromus spp., Festuca spp.,
and Lolium multiflorum (italicum) are among the hosts for Sitobium avenae and Rhopalo Rhopalosiphum siphum padi which transmit the bean yellow dwarf virus (BYDV). Such species left as undersowings after harvest coul co uld d ac actt as a “g “gre reen en br brid idge ge”” an and d en enco cour urag agee th thee tr tran ansf sfer er of th thee pe pest st to the following crop. Less susceptible species are Agropyr Agropyron on repens and Arrhenatherum elatius; however, A. repens favors the buildup of saddle gall midge on Lolium multiflorum and is also highly susceptible to fruit fly fl y, as are various Festuca and Poa spp., the latter also acting as hosts for wheat bulk fly (Burn, 1987). Puvuk and Stinner (1992) report that the presence of grass weeds in co corn rnffie ield ldss is a fa fact ctor or th that at in incr crea ease sess th thee at attr trac acti tive vene ness ss of th thos osee fie ield ldss to the second flight of the European corn borer (Ostrinia nubilalis). Grassy vegetation appears to be preferred mating habitats for O. nubilalis.
Often Ofte n th thee do dome mest stic icat ated ed ver ersi sion on of a pl plan antt sp spec ecie iess is th thee re resu sult lt of an intensivee breeding program that results in, among other factors, a reintensiv duction in the concentration of secondary substances in various plant parts, producing plants with simpler and less stable defenses against herbi her bivo vores res (Ha (Harla rlan, n, 19 1975) 75).. The pr prese esence nce of wil wild d rel relati ative vess in cro crop p bor bor-ders may have local effects on the population genetics of a crop pest. Presumably,, the wild plants can contribute to the genetic diver Presumably diversity sity of pest populations, at least if migration is limited. Thus, maintenance of a variety of plants that possess different complements of plant defenses in crop-border flora could result in the preservation of genetic diversity in the local pest population and decrease the rate of selection for new biotypes that have the ability to overcome host-plant resistance or withstand the application of pesticides (Thresh, 1981).
THE ROLE OF WEEDS IN THE ECOLOGY OF NATURAL ENEMIES
Certain weeds are important components of agroecosystems because they positively affect the biology and dynamics of beneficial insects. Weeds offer many important requisites for natural enemies such as alternative prey/hosts, pollen, or nectar, as well as microhabitats that are not available in weed-free monocultures (Van (Van Emden, 1965b). Many insect pests are not continuously present in annual crops, and their predators and parasitoids must survive elsewhere
during their absence. Weeds usually provide such resources (alternate host or pollen/nectar), aiding in the survival of viable natural enemy populations. The beneficial entomofauna associated with weeds has been surveyed for many species, including the perennial stinging (Urtica ica dioica) dioica),, Mexican tea (Chenopodium ambrosioides), nettle (Urt camphorweed (Heter (Heterotheca otheca subaxillaris), sub axillaris), and a number of ragweed species (Altieri and Whitcomb, 1979a). Perhaps the most exhaustive study of the fauna associated with various weeds is the work of Nentwi Nen twig g and ass associ ociate atess in Ber Berne, ne, Swi Switze tzerla rland nd,, wh where ere the they y mon monito itored red the ins insect ectss ass associ ociate ated d wit with h eig eighty hty pla plant nt spe specie ciess so sown wn as mon monocu ocultu ltures res in a total of 360 plots (Nentwig, 1998). According to this survey, weed species are insect habitats of of widely differing differing quality. quality. Plants such as chervil of France (Anthriscus cerefolium), comfrey (Sym phytum officinale), and gallant soldier (Galinsoga ciliata) have extremely low arthropod populations of less than 15 individuals/m 2, whereas most plants have 100 to 300 arthropods/m 2 according to the D-vac sampling method used by these researchers. Extremely high lev le vel elss we were re fo foun und d on po popp ppy y (P (Papaver apaver rhoeas rhoeas), ), rape (Br (Brassica assica napus) napus),, buckwheat (F (Fagopyrum agopyrum esculentum), and tansy (Tanacetum vulgare), where up to 500 or more arthropods were found per square meter. Considering the trophic structure of the arthropod communities, results were even more striking. Of all al l arthropods, phytophagous phytophagous insects constituted about 65 percent of the species (most values between 45 and 80 per percen cent), t), bu butt the com compo posit sition ion of the rem remain aining ing art arthr hropo opods ds va varries greatly among predators and parasitoids and phytophagous arthropods or between aphidophagous predators and aphids according to the plant species. Most parasitoids were Hymenoptera of the families Aphidiidae, Braconidae, and Ichneumonidae, and also Proctotrupidae and Chalcidoidea, reaching about five to thirty individuals per square meter of vegetation, especially on Asteraceae and Brassicaceae weeds. Dominant predators included Empididae flies, Coleoptera (Cocinellidae, Carabidae, Staphilinidae, and Cantharidae), and chrysopid green lacewings. High densities (seventy predators/m 2) weree ob wer obser serve ved d on bor borage age,, blu bluee kn knapw apweed eed,, and Papaver rhoeas. Ap Aphid hid-ophagous opha gous syrphids syrphids require a succession of plants, early- to late-flowering species, including highly attractive Bras Brassica, sica, Sinapsis Sinapsis,, and Raphanus species. Preferred oviposition sites for the common lacewing (Chrysoperla carnea) included about sixteen plants such as
Agrostemma githago, Trifolium arvense, Echium vulgare, Oenothera Agrostemma others ers (Za (Zands ndstra tra and Mot Motoo ooka, ka, 19 1978) 78).. biennis,, Centaur biennis Centaurea ea jacea, and oth
The Importance of Flowering Weeds
Most adult hymenopteran parasitoids require food in the form of pollen and nectar nec tar to ensure effective e ffective reproduction reproducti on and longevity. Van Van Emden (1965b) demonstrated that certain Ichneumonidae, such as Mesochorus spp., must feed on nectar for egg maturation, and Leius (1967) reported that carbohydrates carbohydrates from the nectar of certain Unbelliferae are essential in normal fecundity and longevity in three Ichneumonid species. In studies of the parasitoids of the European pine shoot moth, Rhyacionia buoliana, Syme (1975) showed that fecundity and longevity of the wasps Exeristes comstockii and Hyssopus thymus were significantly increased with the presence of several flowering weeds. In Hawaii, Euphorbia hirta was reported as an important nectar source for Lixophaga sphenophori, a parasite of the sugarc sug arcane ane wee weevil vil (T (Top opham ham and Bea Beards rdsley ley,, 197 1975). 5). In San Joa Joaqu quin in Valley, California, Apanteles medicaginis wasps, parasites of the alfalfa al falfa caterpillar (Colias eurytheme), were often observed feeding on several er al we weed edss sp spec ecie iess (Con livin ving g Convolvulu volvulus, s, Helian Helianthus, thus, and Polygonum), li longer and exhibiting a higher fecundity. Similar dependence on flowers have have been reported for Orgilus obscurator, a parasite of the Euro Eu rope pean an pi pine ne sh shoo oott mo moth th,, an and d Larr aras asit itee of th thee mo mole le Larra a americ americana, ana, a par cricket (Zandstra and Motooka, 1978). Wildflowers Wildflo wers such as Brassica kaber, kaber, Barbarea vulgaris, vu lgaris, and Daucus carota provided nectar flowers to female parasitoids of Diadegma insulare, an ichneumonid parasitoid of the diamondback moth (Idris and an d Gr Graf afiu ius, s, 19 1995 95). ). An in incr crea ease sed d fe fecu cund ndit ity y an and d lo long ngeevi vity ty of th thee wa wasp sp wass cor wa correl relate ated d wit with h flo flower wer cor coroll ollaa ope openin ning g dia diamet meter er and flo flower wer sha shadding provided to the parasitoid by the plants. Because of its long flowerin er ing g pe peri riod od ove verr th thee su summ mmer er Phaceli hass be been en us used ed as Phacelia a tanacet tanacetifolia ifolia ha a pollen source to enhance syrphid fly populations in cereal fields in the United Kingdom (Wratten and Van Emden, 1995). Spectacular parasitism increase has been observed in annual crops and orc orchar hards ds wit with h ric rich h und under ergro growth wthss of wil wild d flo flower wers. s. In app apple le orc orchar hards, ds, parasitism of tent caterpillar eggs and larvae and codling moth larvae was eighteen times greater in those orchards with floral undergrowths than in orchards with sparse floral undergrowth (Leius, 1967).
Soviet researchers at the Tashkent Laboratory (Telenga, 1958) cited lack of adult food supply as a reason for the inability of Aphytis proclia to control its host, the San Jose scale (Quadraspidiotus perniciosus). The effectiveness of the parasitoid improved as a result of planting a Phacelia sp. cover crop in the orchards. Three successive plantings of Phacelia increased scale parasitization from 5 percent in clean-cultivated orchards to 75 percent where these nectarproducing plants were grown. These Soviet researchers also noted that Apanteles glomeratus, a parasite of two cabbageworm species (Pieris spp.) on crucifer crops, obtained nectar from wild mustard flowers. The parasites lived longer and laid more eggs when these weeds were present. When quick-flowering mustards were actually planted in the fields with cole crops, parasitization of the host increased from 10 to 60 percent (Telenga, 1958). Weed flowers are also important food sources for various insect predators (Van (Van Emden, 1965b). Pollen appears to be instrumental in egg production of many syrphid flies and is reported to be a significant food source for many predaceous Coccinellidae. Lacewings seem to prefer sever several al composite flowers that supply nectar, thus satisfying their sugar requirements (Hagen, 1986). Weed eedss may inc increa rease se po popul pulati ations ons of non nonpes pestif tifero erous us her herbi bivo vorou rouss insects (neutral insects) in crop fields. Such insects serve as alternative hosts or prey to entomophagous insects, thus improving the survival and reproduction of these beneficial insects in the agroecosystem. For example, the effectiveness of the tachinid Lydella grisescens, a parasite of the European corn borer, Ostrinia nubilalis, can be increased in the presence of an alternate host Papaipema nebris, a stalk borer on giant ragweed ( Ambrosia spp.) (Syme, 1975). Several other authors have reported that the presence of alternate hosts on ragweeds near crop fields increased parasitism of specific crop pests. Examples include Eurytoma tylodermatis against the boll weevil, Anthonomus grandis, and Macrocentrus ancylivorus against the oriental fruit moth, Grapholita molesta, in peach orchards. The parasite Herogenes spp. uses the caterpillar of Swammerdamia lutarea on the weed Crataegus sp sp.. to ov over erwi wint nter er ea each ch ye year ar af afte terr em emer erge genc ncee fr from om th thee diamondback moth, Plutella maculipennis (Van Emden, 1965a). A similar situation occurs with the egg parasitoid Anagrus epos, whose effectiveness in regulating the grape leafhopper, Erythroneur Erythroneura a elegantula, is greatly increased in vineyards near areas invaded by wild
blackberry ( Rubus Rubus sp.). This plant hosts an alternate leafhopper Dikrella cruentata that breeds in its leaves during winter (Doutt and Nakata, 1973). Stinging nettle (Urtica dioica) is a host of the aphid Microlophium carnosum. A large complex of predators and parasites attacks the aphid, the numbers of which increase rapidly in April and May in England before pest aphids appear on the crop plants. These natu na tura rall en enem emies ies bui uild ld up in nu numb mber erss on th this is pl plan antt an and d th then en mo mov ve on to adjacent crops once the nettles are cut in mid-June (Perrin, 1975). Some entomophagous insects are attracted to particular weeds, even in the absence of host or prey, by chemicals released by the herbiv bi vor ore’ e’ss ho host st pl plan antt or ot othe herr as asso soci ciate ated d pl plan ants ts (A (Alti ltier erii et al. al.,, 19 1981 81). ). Fo Forr example, the parasitic fly Eucelatoria sp. prefers okra to cotton, and the wa wasp sp Peristenus pseudopallipes, wh which ich att attack ackss the tar tarnis nished hed pla plant nt bug, bu g, pref prefers ers Erigeron to oth other er wee weed d spe specie ciess (M (Mont onteit eith, h, 19 1960; 60; Net Nettle tles, s, 1979). Parasitism by Diaeretiella rapae was much higher when the aphid Myzu wass on co coll llar ard d th than an wh when en it was on be beet et,, a pl plan antt Myzuss persi persicae cae wa lacking attractive mustard oil (Read, Feeny, and Root, 1970). Of significant practical interest are the findings of Altieri and colleagues (1981), which showed that parasitization rates of Heliothis zea eggs by Trichogramma sp. were greater when the eggs were placed on soybeans next to corn and the weeds Desmodium sp., Cassia sp., and Croton sp. than on soybeans grown alone (Figure 4.1). Although the same number of eggs were placed on soybean and on the associated plants, few of the eggs placed on the weeds were parasitized, sitize d, sugg suggestin esting g that these plan plants ts were not acti actively vely searched searched by efficiency cy of paraTrichogramma sp. but nevertheless enhanced the efficien sitization sitiza tion on the associated soybean soybean plants. It is poss possible ible that they emitted volatiles with kairomonal action. Further tests showed that appl ap plica icati tion on of wa wate terr ex extr trac acts ts of so some me of th thes esee as asso soci ciate ated d pl plan ants ts (e (esp speecially Amaranthus sp.) to soybean and other crops enhanced parasitization of H. zea eggs by Trichogramma spp. wasps (Table 4.1). The authors stated that a stronger attraction and retention of wasps in the extract-treated plots may be responsible for the higher parasitiza sit izatio tion n le leve vels. ls. The po possi ssibil bility ity tha thatt ve veget getatio ational nally ly com comple plex x plo plots ts are more chemically diverse than monocultures, and therefore more acceptable and attractive to parasitic wasps, opens new dimensions for biolog bio logica icall con contro troll thr throug ough h wee weed d man manage agemen mentt and beh behav avior ior mod modif ifica ica-tion.
FIGURE 4.1. The effect of plant assemblages on parasitization of corn earworm (Heliothis zea) eggs zea) eggs by Trichogramma Trichogramma parasitic parasitic wasps in soybean fields (after Altieri and Letourneau, 1984).
In general, most beneficial insects present on weeds tend to dispers pe rsee to cr crop ops, s, but in a fe few w in inst stan ance cess th thee pr prey ey fo foun und d on we weed edss pr preeven entt or delay this dispersal. In such cases, allowing all owing weeds to grow grow to ensuree con sur concen centra tratio tions ns of ins insect ectss and the then n cut cuttin ting g the them m re regul gularl arly y to fo force rce movement could be an effective strategy. For example, by cutting patches of stinging nettle (U. dioica) in May or June, predators (mainly Coccinellidae) were forced to move into crop fields (Perrin, 1975). Similarly, cutting the grass weed cover drove Coccinellidae into orchard trees in southeastern Czechoslovakia Czechoslovakia (Hodek, 1973). By cutting hedges of Ambrosia trifida infested with the weevil Lixus scrobicollis, a 10 percent increase of boll weevil parasitization by E. tylodermatis was obtained in two test plots of cotton adjacent to the hedgerow (Pierce, Cushman, and Hood, 1912). These practices must be carefully timed and based on the biology of beneficial inin sects. For example, in California the annual cleanup cl eanup of weeds along the edges of alfalfa fields should be delayed until after mid-March, when aggregations of dormant Coccinellidae have largely dispersed (van den Bosch and Telford, 1964).
TABLE 4.1. Percent Parasitization of Heliothis zea (Boddie) zea (Boddie) Eggs by Naturally Occurring Trichogramma sp sp.. in Cro Crop p Syst Systems ems Spr Spray ayed ed wit with h Vario arious us Plan Plantt Extracts Treatment
Soybean
Cowpeas
Tomato
Cotton
Water extract of Amaranthus
21.4a
a
45.4a
24.3a
13.6a
Water extract of corn
17.4b
45.8a
21.1a
—
Water
12.6c
31.6b
17.6b
4.2b
Source: After Source: After Altieri et al., 1981. aMeans followed by the same letter in each column col umn are not significantly different
according to Duncan’s multiple range test (P ( P < < 0.05).
INSECT DYNAMICS IN WEED-DIVERSIFIED CROP SYSTEMS
In the past thirty-five years, research has shown that outbreaks of certain types of crop pests are less likely to occur in weed-diversified crop systems than in weed-free fields, mainly due to increased mortality imposed by natural enemies (Pimentel, 1961; Adams and Drew, 1965; Dempster, 1969; Flaherty, 1969; Smith, 1969; Root, 1973; Altieri, Schoonhoven, Schoonhoven, and Doll, 1977). Crop fields with a dense weed cove co verr and hig high h di dive versi rsity ty usu usually ally ha have ve mor moree pre predac daceou eouss art arthr hropo opods ds tha than n do weed-free fields (Pimentel, 1961; Dempster, 1969; Flaherty Flaherty,, 1969; Smith, 1969; Root, 1973; Perrin, 1975; Speight and Lawton, 1976). The successful establishment of several parasitoids has depended on the presence of weeds that provided nectar for the adult female wasps. Relevantt examples of cropping systems in which the Relevan t he presence of specific weeds has enhanced the biological control of particular pests are given in Table Table 4.2 (Altieri and Letourneau, 1982). A literature survey by Bali Balidda ddawa wa (19 (1985) 85) sho showed wed tha thatt pop popula ulatio tion n den densit sities ies of twe twenty nty-se -seve ven n insect species were reduced in weedy crops compared to weed-free crop cr ops. s. In ad addi diti tion on to th thee in inse sect ct st stud udie ies, s, po popu pula latio tion n de dens nsiti ities es of on onee mi mite te Eotetranych anychus us willam willamette, ette, we species, Eotetr were re fo foun und d to be re relat lativ ivel ely y lo lowe werr in grap gr apee vi vine ness wh wher eree we weed edss we were re al allo lowe wed d th than an in we weed eded ed pl plot otss (F (Fla lahe hert rty y, 1969). Table 4.3 lists the insect species under their respective orders and scores the number of occurrences that they appeared in the literature. The cabbage aphid and the cabbage white butterfly, Pieris rapae, seem to have received most attention.
TABLE 4.2. Selected Examples of Cropping Systems in Which the Presence of Weeds Enhanced the Biological Control of Specific Crop Pests
Cropping Systems
Factor(s) Involved
Weed Species
Pest(s) Regulated
Alf lfa alfa
Natural bloo oom min ing g weed complex
Alfalfa caterpillar (Colias eurytheme)
Increased activity of the parasitic wasp Apanteles medicaginis
Alfalfa
Grass weeds
Empoasca fabae
Unknown
Apple
Phacelia sp. Phacelia sp. and Eryngium Eryngium sp. sp.
San Jose scale (Quadraspidiotus perniciosus) and perniciosus) and aphids
Increased activity and abundance of parasitic wasps (Aphelinus mali and Aphytis proclia )
Apple
Natural weed complex
Tent caterpillar cater pillar Increased activity (Malacosoma and abundance of americanum) and americanum) and parasitic wasps codling moth (Cydia pomonella)
Beans
Goosegrass Leafhoppers (Eleusine indica) (Empoasca and red sprangletop kraemeri) (Leptochloa filiformis)
Broccoli
Wild mustard
Phyllotreta cruciferae Trap cropping
Brussels sprouts
Natural weed complex
Imported cabbage butterfly (Pieris rapae) and rapae) and aphids (Brevicoryne brassicae)
Alteration of colonization background and increase of predators
Brussels sprouts
Sper Sp ergul gula a arv arvens ensis is
Delia De lia br bras assic sicae ae
Unknown
Brussels sprouts
Spergul Spe rgula a arv arvensi ensis s
Mamestra Mamest ra bra brassi ssicae cae,, Evergestis forficalis, Brevicoryne brassicae
Increase of predators and interference with colonization
Cabbage
Crataegus s Crataegus sp.
Diamond moth (Plutella maculipennis)
Provision of alternate hosts for parasitic wasps (Horogenes Horogenes sp.) sp.)
Citrus
Hedera helix
Lachnosterna a sp spp. p.
Enhance Enhan ceme ment nt of Aphytis lingnanensis
Chemical repellency or masking
Cropping Systems
Factor(s) Involved
Weed Species
Pest(s) Regulated
Citrus
Natural weed complex
Mites (Eotetranychus Eotetranychus sp., sp., Panonychus citri, Metatetranychus citri )
Unknown
Citrus
Natural weed complex
Dias Di aspi pidi did d sc scal ales es
Unkn Un know own n
Coffee
Natural weed complex
Pentatomid Unknown Antestiopsis intricata
Collards
Ragweed (Ambrosia artemisiifolia)
Flea beetle (Phyllotreta cruciferae)
Collards
Amaranthus Green peach aphid retroflexus, (Myzus persicae) Chenopodium album, Xanthium strumarium
Corn
Giant ragweed
European corn borer Provision of alter(Ostrinia nubilalis) nate hosts for the tachinid parasite Lydella grisescens
Corn
Natural weed complex
Heliothis zea, Spodoptera frugiperda
Enhancement of predators
Corn
Setaria viridis and S. fa faberi beri
Diabrotica virgifera and D. barberi
Unknown
Cotton
Ragweed
Boll weevil (Anthonomus grandis)
Provision of alternate hosts for the parasite Eurytoma tylodermatis
Cotton
Ragweed and Rumex Heliothis Heliothis s spp. crispus
Increased populations of predators
Cotton
Salv Sa lvia ia co cocc ccin inea ea
Lyg ygus us vos osse sele leri ri
Unknown
Cruciferous crops
Quick-flowering mustards
Cabbageworms (Pieris Pieris spp.) spp.)
Increased activity of parasitic wasps (Apanteles glomeratus)
Mungbea Mun gbeans ns
Naturall we Natura weed ed complex
Beanfly (Ophiomyia phaseoli)
Alteration of colonization background
Chemical repellency or masking Increased abundance of predators (Chrysoperla carnea, Coccinellidae Syrphidae)
TABLE 4.2 (continued) Cropping Systems
Factor(s) Involved
Weed Species
Pest(s) Regulated
Oil palm
Pueraria sp., Pueraria sp., Flemingia sp., Flemingia sp., ferns, grasses, and creepers
Scarab beetles Oryctes rhinoceros and Chalcosoma atlas
Unknown
Peach
Ragweed
Oriental fruit moth
Provision of alternate hosts for the parasite Macrocentrus ancyclivorus
Peach
Rosaceous weeds and Dactylis glomerata
Leafhoppers Paraphelepsius sp. Paraphelepsius sp. and Scaphytopius actus
Unknown
Sorghum
Helianthus spp. Helianthus spp.
Schizaphis graminum
Enhancement of Aphelinus Aphelinus spp. spp. parasitoids
Soyb So ybea ean n
Broadl Broa dlea eaff wee eeds ds an and d grasses
Epilachana varivestis
Enhancement of predators
Soybean
Cass Ca ssia ia ob obtu tusif sifol olia ia
Nezara Neza ra vi virid ridula ula,, Anticarsia gemmatalis
Increased abundance of predators
Soybean
Crotalaria usaramoensis
Nezara viridula
Enhancement of tachnid Trichopoda sp. Trichopoda sp.
Sugar cane
Euphorbia spp. Euphorbia spp. weeds
Sugar-cane weevil (Rhabdoscelus obscurus)
Provision of nectar and pollen for the parasite Lixophaga sphenophori
Sugar cane
Grassy weeds
Aphid (Rhopalosiphum maidis)
Destruction of alternate host plants
Sugar cane
Borreria verticillata Cricket and Hyptis atrorubens (Scapteriscus vicinus)
Sweet potatoes
Morning glory (Ipomoea asarifolia)
Provision of nectar for the parasite Larra americana
Argus tortoise beetle Provision of alter(Chelymorpha nate hosts for the cassidea) parasite Emersonella sp. Emersonella sp.
Cropping Systems
Factor(s) Involved
Weed Species
Pest(s) Regulated
Vegetable crops
Wild carrot (Daucus carota)
Japanese beetle (Popillia japonica)
Increased activity of the parasitic wasp Tiphia popilliavora
Viney Vin eyar ards ds
Wild bl Wild blac ackbe kberry rry (Rubus Rubus sp sp.. )
Grape leafhopper (Erythroneura elegantula)
Increase of alternate hosts for the parasitic wasp Anagrus epos
Viney Vin eyar ards ds
Johnson John son gr grass ass (Sor- ghum halepense)
Pacific mite (Eotetranychus willamette)
Buildup of predaceous mites (Metaseiulus occidentalis)
Source: Based Source: Based on Altieri and Letourneau, 1982; Andow, 1991a.
Table 4.4 lists the main pest population regulating factors in the weed-diversified crops. These factors include parasites and predators, camouflage and masking, and reduced colonization. Natural enemies alone account for more than half (56 percent) of the cases where the pest population was claimed to be controlled. Researchers have found at least two underlying mechanisms explaining how careful diversification diversification of the weedy component of agricultural systems often lowers pest populations significantly (Altieri, Schoonhoven, and Doll, 1977; Risch, Andow, and Altieri, 1983). In somee cas som cases, es, pla plant nt dis disper persio sion n and di dive versi rsity ty app appear earss to inf influe luence nce her herbi bi-vore density, primarily by altering herbivore movement or searching behavior (Risch, 1981; Bach, 1980b; Kareiva, 1983). In other cases, predators and parasites encounter a greater array of alternative resources and microhabitats in weedy crops, reach greater abundance and diversity levels, and impose greater mortality on pests (Root, 1973; Letourneau and Altieri, 1983). Many Man y stu studie diess ind indica icate te tha thatt ins insect ect-pe -pest st dy dynam namics ics are af affec fected ted by the lower concentration and/or greater dispersion of crops intermingled with weeds. For example, adult and nymph densities of Empoasca kraemeri, the main bean pest of the Latin American tropics, were reduced significantly as weed density increased in bean plots. Conversely, the chrysomelid Diabrotica balteata was more abundant in weedy bean habitats than in bean monocultures; bean yields were not affected because feeding on weeds diluted the injury to beans.
TABLE 4.3. Occurrence of Individual Pests in Weed-Crop Diversity Diversity Studies Crop Pest
Occurrence
Order: Coleopt Coleoptera era Anthonomus grandis Boheman grandis Boheman Popillia japonica Newm. japonica Newm. Phyllotreta striolata (Fabricius) striolata (Fabricius) P. cruciferae cruciferae (Goeze) (Goeze) Rhaboscelus obscurus (Boisduval) obscurus (Boisduval)
2 1 2 3 1
Order: Lepido Lepidoptera ptera Anticarsia gemmatalis Hubner gemmatalis Hubner Cydia pomonella (Linnaeus) pomonella (Linnaeus) Colias eurytheme Boisduval eurytheme Boisduval Evergestis forficalis Linnaeus forficalis Linnaeus Grapholitha molesta (Busck) molesta (Busck) Heliothis sp. Heliothis sp. Malacosoma americanum (Fabricius) americanum (Fabricius) Mamestra brassicae (Linnaeus) brassicae (Linnaeus)
1 1 1 1 1 1 1 1
Order: Lepido Lepidoptera ptera Ostrinia nubilalis (Hubner) nubilalis (Hubner) Pieris rapae (Linnaeus) rapae (Linnaeus)
1 8
Order: Diptera Ophiomyia (Melanogromyza) phaseoli (Tryon)
2
Order: Homopte Homoptera ra Aleyrodes brassicae (Walker) brassicae (Walker) Brevicoryne brassicae (Linnaeus) brassicae (Linnaeus) Erythroneura elegantula (Osborn) elegantula (Osborn) Erioischia brassicae (Linnaeus) brassicae (Linnaeus) Empoasca kraemeri (Ross kraemeri (Ross & Moore) Lygus hesperus (Knight) hesperus (Knight) L. eli elisus sus (Van (Van D.) Myzus persicae (Sulzar) persicae (Sulzar) Quadraspidiotus pernicious (Comstock) Order: Heterop Heteroptera tera Nezara viridula Scudder viridula Scudder Order: Orthopter Orthopteraa Scapteriscus vicinus Scudder vicinus Scudder Order: Acarina Eotetranychus willamette Ewing willamette Ewing Total
Source: After Source: After Baliddawa, 1985.
1 7 1 2 1 1 1 4 1 1 1 1 50
TABLE 4.4. Examples of Crop Pest Population Management Through WeedCrop Diversity
Agroecosystem
Pest
Factor (Suggested or Proved)
Reference
1 Cotton and cowpe cowpea a Bollweevil, strip planted with Anthonomus weeds grandis
Greater parasitic Pierce (1912) wasp, Eurytoma Eurytoma sp. sp. quoted in population Marcovitch (1935)
2 Vegetab egetables les grown among wild carrot (D. carota)
Greater activity of the parasitic wasp, Tiphia popilliavora
King and Holloway quoted by Altieri and Letourneau (1982)
3 Peach Peach and ragwe ragweed ed Oriental fruit moth, (Ambrosia Ambrosia sp.), sp.), Grapholitha smart weed molesta (Polygonum sp.) and lambs quarter (Chenopodium al- bum), golden rod (Solidago Solidago sp.) sp.)
Alternate hosts for the parasite, Macrocentrus ancylivorus
Bobb (1939)
4 Sugar-cane Sugar-cane with Borreria verticillata and Hyptis atrorubens
Cricket, Scapteriscus vicinus
Nectar source for the parasite, Larra americana
Wolcott (1942) quoted by Altieri and Letourneau (1982)
5 Apple trees grown with Phacelia Phacelia sp. sp. and Bryngium Bryngium sp. sp.
San Jose scale, Quadraspidiotus perniciosus and perniciosus and various species of aphids
Greater abundance and activity of the parasites, Aphytis proclia
Telenga (1958) quoted by Altieri and Letourneau (1982)
6 Collard Collards, s, Brassica oleracea, and other brassicas grown among weeds
Pyllotreta striolata, Myzus persiscae, Brevicoryne brassicae, Pieris rapae
Camo Ca mouf ufla lage ge
Pime Pi ment ntel el (1 (196 960) 0)
7 Cotton grown with Bollworm, ragweed Ambrosia Anthonomus sp. grandis
Alternate hosts for the parasite, Eurytoma tylodermis
van den Bosch and Telford (1964)
8 Alfalfa Alfalfa with natura naturall blooming weed complex
Alfalfa caterpillar, Colias eurytheme
Increased activity of Apenteles medicaginis
van den Bosch and Telford (1964)
9 Apple plants with weeds
Tent caterpillar, Malacasoma americanum and americanum and Carpocapsa pomonella
Increased activity and abundance of parasitic wasps
Lewis (1965)
Plant bugs, Lygus herperus and L. eli elisus sus
Retention of natural van den Bosch and enemy and synchro- Stern (1969) nizing intense natural enemy activity with pest populati population on
10 Strip croppi cropping ng cotton and alfalfa
Japanese beetle, Popillia japonica
TABLE 4.4 (continued)
Agroecosystem
Pest
11 Grape vines with weeds
Pacific mite, Eotetranychus williamettei
Factor (Suggested or Proved)
Reference Flaherty (1969)
12 Brussels sprout sprouts s Myzus persicae, with weeds (hoed or Brevicoryne cut back to 15 cm) brassicae, Aleyrodes brassicae, Pieris rapae
Camo Ca mouf ufla lage ge
Smith Smit h (1 (196 969) 9) an and d (1976a)
13 Cruciferous Cruciferous crops Cabbage worm, with quick flowering Pieris Pieris sp. sp. mustards
Increased activity of the parasite, Apanteles glomeratus
National Academy of Sciences (1969)
14 Beans with goose Leafhopper, grass, Eleusine Empoasca indica and indica and red kraemeri spragletop, Leptochloa filiformis
Chemical repellency or masking
Tahvanainen and Root (1972)
15 Collards Collards and and ragweed, Ambrosia artemislifolia
Flea beetle, Phyllotreta cruciferae
Chemical repellency and masking
Tahvanainen and Root (1972)
16 Cotton and ragweed plus Rumex crispus
Heliothis s Heliothis sp.
Greater predator population
Smith and Reynolds (1972)
17 Vineyards Vineyards with wild blackberry, Rubus sp.
Grape leafhopper, Erythroneura elegantula
More alternate hosts for Anagrus epos
Doutt and Nakata (1973)
18 Collards Collards in “dive “diverse” rse” P. cruciferae, row in an old field M. periscae, periscae, B. brassi brassicae, cae, P. rapae
Resource concentration
Root (1973)
19 Cabbage Cabbage with with white Erioischia and red clover brassicae, B. brassi brassicae, cae, P. rapae
Less colonization and greater predator population of Harpalus rufipes, Phalangium sp. Phalangium sp.
Dempster and Coaker (1974)
20 1, 10, 10, 100 100 plant plant monocultures, 10 plant plots in old mowed field
P. cruciferae, P. striolata, aphids, and P. rapae
Camo Ca mouf ufla lage ge
Crom Cr oma art rtie ie (1 (198 981) 1)
21 White clov clover er undersown in brussels sprouts
E. brassi brassicae, cae, B. brassi brassicae, cae, P. rapae
Camouflage and greater predation
Dempster and Coaker (1974)
22 Mungbeans Mungbeans grown among weeds
Bean fly, O. phaseo phaseoli li
Camouflage
Altieri and Whitcomb (1979b)
Factor (Suggested or Proved)
Agroecosystem
Pest
Reference
23 Corn gro grown wn with giant ragweed
European cornborer, O. nubil nubilalis alis
Alternate hosts for Lydella grisesens
Syme (1975)
24 Sugar-cane Sugar-cane with Euphorbia sp. Euphorbia sp.
Sugar-cane weevil, R. obs obscuru curus s
Nectar and pollen sources for Lixophaga sphenophori
Topham and Beardsley (1975)
25 Brussels Brussels sprouts P. rapae, grown among natu- B. brassicae brassicae ral weeds complex
Camouflage and more predation
Smith (1976a)
26 Mungbean Mungbean grown among natural weed complex
Bean fly, O. phaseo phaseoli li
Less Les s col coloni onizati zation on
Litsinger Litsin ger and Moody (1976)
27 Beans growin growing g among weeds or surrounded by weedy borders
E. kra kraeme emeri, ri, D. baltea balteata ta
28 Brussels Brussels sprouts grown with Spergula arvensis weeds
M. brass brassicae, icae, E. forfi forficalis calis,, B. brassi brassicae cae
Altieri et et al. (1977 (1977))
Lower colonization and greater predator population
Theunissen and den Ouden (1980)
29 Collards Collards grown Green peach aphid, Greater abundance Horn (1981) among weeds, M. per persica sicae e of predators, mainly Amaranthus Chrysopa camea retroflexus, Chenopodium al- bum, Xanthium stramonium 30 Soybean Soybean grown with The green stinkbug Cassia obtusifolia N. viri viridul dula, a, and velvet bean caterpillar Anticarsia gemmatalis
Greater abundance Altieri and Todd of predators (1981)
Source: After Source: After Baliddawa, 1985.
In other experiments, E. kraemeri populations were reduced significantly in weedy habitats, especially in bean plots with grass weeds ( Eleusine Eleusine indica and Leptochloa filiformis). D. balteata densities fell by 14 percent in these systems. When grass-weed borders one meter wide surrounded bean monocultures, populations of adults and nymphs of E. kraemeri fell drastically (Figure 4.2). Also, pure stands of L. filiformis reduced adult leafhopper populations significantly more than E. indi (Sch choo oonh nho ove ven n et al al., ., 19 1981 81); ); th this is ef effe fect ct ce ceas ased ed wh when en th thee indica ca (S weed was killed with paraquat. If bean plots were sprayed with a wa-
300 st n al p n a e b 0
200 8 n o n oi t al u p o p
a c s a 100 o p m E
Beans alone
tl u
Broadleaved weeds d A
Grass weeds 0
15
25 Days after planting
35
45
FIGURE FIGU RE 4. 4.2.Eff 2.Effec ectt of gr gras ass s we weed ed bo bord rder ers s ar aroun ound d 16 m2 be bean an plo plots ts on th the e po popu pu-lation lat ion of adu adult lt Empoas Empoasca ca kraem kraemeri eri (af (after ter Alt Altieri ieri,, Scho Schoonh onhov oven, en, and Dol Doll, l, 197 1977). 7).
ter homogenate of fresh grass-weed leaves, adult leafhoppers were repelled. Continuous applications affected the reproduction of leafhoppers, as evinced by a reduction in the number of nymphs (Altieri et al., 1977). Their regulatory effect was greater than that exhibited by extracts of broadleaf weeds such as Amaranthus dubius. Weeds within a crop system can reduce pest incidence by enticing pest insects away from the crop. For example, flea beetles, Phyllotreta cruciferae, concentrate their feeding more on the intermingled Brassica campestris plants than on collards (Altieri and Gliessman, 1983). The weed species has significantly higher concentrations of allylisothiocyanate (a powerful powerful attractant of flea beetle adults) than collards, thuss di thu dive verti rting ng the bee beetle tless fr from om the cro crops. ps. Sim Similar ilarly ly,, in Tla Tlaxca xcala, la, Me Mexxico, the presence of flowering Lupinus spp. in tasseling cornfields often diverts diverts the attack of the scarab beetle, Macrodactylus sp., from female corn flowers to lupine flowers (Trujillo-Arriaga and Altieri, 1990).
Several studies have documented Several documented pest reduction due to an increase i ncrease of natural enemies in weedy crop fields. Fall armyworm, Spodoptera incide idence nce wa wass con consis sisten tently tly hig higher her in wee weed-f d-free ree cor corn n plo plots ts frugiperda, inc than in corn plots containing natural weed complexes or selected weed associations. Corn earworm (Heliothis zea) damage was similar in all weed-free and weedy treatments, suggesting that this i nsect is not affected greatly by weed diversity. Experimental design was a crucial factor in demonstrating the effect of weeds on predator population lat ions. s. In an ex exper perime iment nt con conduc ducted ted by Alt Altier ierii and Whi Whitco tcomb mb (19 (1980 80), ), field plots were close together (8 m apart), and predators moved freely between habitats, confoundin confounding g results. Therefore, it was diff diffiicult to identify between-treatment differences (i.e., weedy versus weed-free plots) in the composition of predator communities. In another experiment, increased distances (50 m) between plots minimized such migrations, resulting in greater population densities and diversity of common foliage insect predators in the weed-manipulated corn systems than in the weed-free plots. Trophic relationships in the weedy habitats were more complex than food webs present in monocultures. When comparing parasitism of second generation O. nubilalis larvae by the ichneumonid Eriborus terebrans, Puvuk and Stinner (1992) found that parasitism was not significantly influenced by the presence of weeds, although there was a trend for greater parasitism in treatments with weeds than in weedless plantings. In England, winter barley plots with grass weeds had fewer aphids and more than ten times the number of staphylinid beetles than plots without weeds (Burn, 1987). Similarly, spring-planted alfalfa plots infest inf ested ed wit with h wee weeds ds had a les lesss di dive verse rse sub substr strate ate-pr -preda edator tor com comple plex x bu butt a greater foliage-predator complex than did weed-free plots (Barney et al., 1984). The carabid Harpalus pennsylvanicus and foliage predators (i.e., Orius insidiosus and Nabidae) were more abundant in alfalfa fields where grass weeds were dominant. Smith (1969) concluded that weeds within brussels sprouts enhanced natural enemy action against aphids by providing predator oviposition sites. This partly explained the lower aphid populations recorded in weedy plots. By selectively allowing a cover of Spergula arvensis within brussels sprouts plots, populations of Mamestra brassicae, Evergestis forficalis, cabbage root fly, and Brevicoryne
brassicae were drastically reduced (Theunissen and den Ouden,
1980). Schellhorn and Sork (1997) compared population densities of herbivores, predators, and parasitoids on collards in monocultures and on collards interplanted with two t wo different groups of weeds, one with weed wee d sp spec ecies ies fr from om th thee sa same me pl plan antt fa fami mily ly as th thee co colla llard rdss (B (Bra rass ssic icac acea eae) e) and one with weed species from unrelated plant families (nonBrassicaceae). The collards in the Brassicaceae-weed polyculture had higher densities (number of herbivores/mean leaf area [cm 2] per plant) of specialist herbiv herbivores ores than collards in i n the non-Brassicaceaeweed polyculture and in collard monoculture. The resource concentration hypothesis is supported by the observation of higher populations tio ns of Phyllotreta spp spp., ., act acting ing as fa facul cultati tative ve po polyp lyphag hages, es, in the Brassicaceae-weed polyculture than in the non-Brassicaceae-weed polyculture where Phyllotreta spp. are facultative monophages. Population densities of natural enemies (mostly coccinelids, carabids, and a nd staphylinids) were higher in the polycultures than in the monoculture tu re:: ca cara rabi bid d an and d st stap aphy hylin linid id pr pred edat ator orss ma may y be re resp spon onsi sibl blee fo forr th thee ob ob-served larval mortality in the imported cabbage worm, Pieris rapae, and in the diamondback larvae, larvae, Plutella xylostella. In spite of differences in densities of specialist herbivores across treatments, crop yield, leaf area (cm2), the proportion of leaf area damaged, and the number of leaves undamaged did not differ. The authors concluded that the use of weedy cultures can provide effective effective means of reducing herbivores herbivores if the crop and weed species are not related and plant competition is prev prevented. ented. Based on his extensive studies of weeds as sources of habitat for natural enemies, Nentwig (1998) prepared seed mixtures consisting of twenty-four species of wildflowers sown as 3- to 8-m-wide weed strips planted every 50 to 100 m within fields. Those strips are referred to as ecological compensation areas, which serve as a refuge area and/or dispersal center of natural enemies, compensating, at least partially, for the negative effects of monoculture (Frank and Nentwing, 1995). These studies have demonstrated enhanced biodiversity diver sity of beneficial insects and lower insect-pest incidence in crop fields enriched with weed strips.
ISOLATING THE ECOLOGICAL EFFECTS OF WEED DIVERSITY
Although all hypotheses explaining herbivore herbivore population dynamics in weed-diversified cropping cropping systems attribute a role to the physical structure of the plant canopies in achieving decreased herbivore abundance and/or increased natural enemy densities, few experiments men ts ha have ve rem remov oved ed the con confou foundi nding ng inf influe luence nce of pla plantnt-spe species cies ric richhness to assess the effects of plant architecture and density per se (Altieri and Letourneau, 1984). One problem with these experiments is that they do not isolate crop-weed diversity diversity as an independent variable. In most cases, weed diversity could possibly have reduced herbiv bi vor oree ab abun unda danc ncee be beca caus usee it re redu duce ced d th thee si size ze or qu qual ality ity of cr crop op pl plan ants ts (Kareiva, (Kareiv a, 1983). Weed density, diversity diversity,, plot, or patch sizes are all intera int eracti cting ng fa facto ctors rs tha thatt may inf influe luence nce cro crop p qua qualit lity y and her herbi bivo vore re den den-sities. Thee pr Th pres esen ence ce of we weed edss in cr crop opss af affe fects cts bo both th pl plan antt de dens nsit ity y an and d sp spac ac-ing patterns, factors known to significantly influence insect populations (Mayse, 1983). In fact, many herbivores respond specifically to plant density; some proliferate in close plantings, whereas others reach rea ch hig high h num number berss in ope openn-can canop opy y cro crops ps.. Pre Predat dator or and par parasi asite te po poppulations tend to be greater in high-density plantings. Mayse suggests that the microclimate associated with canopy closure, which occurs earlier in dense plantings, may increase development rates of some predators and possibly facilitate prey capture. Careful consideration of the units to be used in i n expressing population numbers is crucial for meaningful interpretations of results and for determining general patterns among various research findings. For example, Mayse and Price (1978) found that numbers of certain arthropod species sampled in different soybean row-spacing treatments were significantly different on a per-plant basis, but those same population values converted converted to a square meter of soil-area basis were not significantly different. Based on the preceding considerations, in addressing the effects of crop-weed density/spacing on insect populations, one should fully consider (1) the effects on the growth, development, and nutritional status of the crop plants and weeds, (2) the effects on the microclimate and microhabitats available for the life processes of herbivores and
their natural enemies, and (3) the effects of potentially different levels of herbivores herbivores in the population dynamics of predators and parasites. Andow (1991b) purposely designed an experiment to separate the effects of decreased pest attack (from Epilachna varivestis and Em poasca fabae) and increased plant competition that often occur simultaneously in weed-diversified weed-diversified bean systems. In this system, weeds directly affected bean herbivores herbivores by reducing their population densities and intensity of attack, indirectly benefiting the bean plants. At thee sa th same me ti time me,, th thee we weed edss di dire rect ctly ly co comp mpet eted ed wi with th th thee be bean ans. s. Th Ther eree is a negative correlation between the intensity of insect attack on beans and the intensity of competition. At one extreme, in the monoculture, there was very high pest attack but no interspecific competition, wher wh erea eass at th thee ot othe herr ex extr trem eme, e, th thee un unwe weed eded ed tr trea eatm tmen ent, t, th ther eree wa wass ver ery y intense competition but very little insect attack. Without insects, competition reduced the yield of beans in these treatments. However, However, when insects were present, there was no difference in yield among the three treatments. Thus, at low levels of crop-weed competition, the effects of reduced insect-pest damage were large l arge enough to balance out the effects of increased plant competition (Figure 4.3). In con contra trast, st, the there re wa wass no sig signif nifica icant nt int intera eractio ction n bet betwee ween n yie yields lds and insect damage in the unweeded treatment. This was the treatment with the greatest reduction in insect-pest populations and herbivore damage. Despite this great decrease in the intensity of herbivory, there was no detectable yield response. Apparently, Apparently, the reduction in yiel yi eld d fr from om th thee in inten tense se we weed ed co comp mpet etit itio ion n wa wass so la larg rgee th that at th thee po posi siti tive ve effect of reduced herbivory was swamped out. Andow (1991b) concluded that three-way interactions among beans, weeds, and bean herbivores herbiv ores were important when bean-weed competition was not very intense.
CROP-WEED MANAGEMENT CONSIDERATIONS
As indicated by the studies discussed in this chapter, much evidence suggests that encouragement of specific weeds in crop fields may improve the regulation of certain insect pests (Altieri and Whitcomb, 1979a). Naturally, Naturally, careful manipulation strategies need to be defined in order to avo avoid id weed competition with crops and interferi nterference with certain cultural practices. Moreover, Moreover, economic thresholds t hresholds
FIGURE 4.3. Yields of beans (as seed weight per plant) with and without herbivores and with and without weeds. Insects were eliminated with insecticides. The four levels of weediness were (1) no weeds (monoculture), (2) interplanted with 8,200 clumps of wild mustards (Brassica kaber) /ha midway between alternate nat e be bean an ro rows ws,, (3 (3)) na natu tura rall pop popul ulat atio ions ns of we weed eds s for th the e fi firs rstt th thirt irtyy-fi five ve da days ys af afte terr planting and weeded from then on (July weeded), and (4) natural populations of weeds for the entire growing season (unweeded) (after Andow An dow,, 1991b).
of weed populations, as well as factors affecting crop-weed crop-weed balance within a crop season, need to be defined (Bantilan, Palada, and Harwood, 1974). Shifting the crop-weed balance so that insect regulation is achieved and crop yields are not economically reduced may be accomplished by car carefu efully lly usi using ng her herbic bicide idess or sel select ecting ing cul cultur tural al pra practi ctices ces that that fa favo vorr the crop cover over weeds. Suitable levels of desirable weeds that suppo sup port rt po popul pulatio ations ns of ben benef efici icial al ins insects ects can be att attain ained ed wit within hin fi field eldss by (1) designing competitive crop mixtures, (2) allowing weed growth as alternate rows or in field margins only, (3) using cover crops, (4)) ado (4 adopti pting ng clo closese-ro row w spa spacin cing, g, (5 (5)) pro provid viding ing wee weed-f d-free ree per period iodss (i. (i.e., e., keeping crops free of weeds during the first third of their growth cycle), (6) mulching, (7) practicing soil-fertility management, and (8) practicing cultivation regimes.
In addition to minimizing the competitive interference of weeds, changes in the species composition of weed communities are desirable to ensure the presence of plants that attract beneficial insects. Manipulation of weed species can be achieved by several means, such as changing levels of key chemical constituents in the soil, use of herbicides that suppress certain weeds while encouraging others, dire di rect ct so sowi wing ng of we weed ed se seed eds, s, an and d ti timi ming ng so soil il di dist stur urba banc nces es (A (Alt ltie ieri ri an and d Whitcomb, 1979b; Altieri and Letourneau, 1982). Changes of the Levels of Key Chemical Constituents in the Soil
The local weed complex can be affected indirectly by the manipulation of soil fertility. Fields in Alabama, with low soil potassium, were dominated by buckhorn plantain (Plant (Plantago ago lanceol lanceolata) ata) and curly dock (Rumex crispus), whereas fields with low soil phosphorus were dominated by showy crotalaria (Crotalaria spectabilis), morning glory (Ipomoea purpurea), sicklepod (Cassia obtusifolia), Geranium carolinianum, and coffee senna (Cassia occidentalis) (Hoveland, Buchanan, and Harris, 1976). Soil pH can influence the growth of certain weeds. For example, weeds of the genus Pteridium occur on acid soils while Cressa sp. inhabits only alkaline soils. Other species (many Compositae and Polygonaceae) can grow in saline soils (National Academy of Sciences, 1969). Other major soil-management practices that affect soil processes related to soil-weed dynamics are tillage, crop rotation, and use of cover crops and green manures. Combined in a cropping system these practices (1) can reduce the persistence of weed seeds in the soil so il,, (2 (2)) re redu duce ce th thee ab abun unda danc ncee of sa safe fe-s -site itess an and d th thee fill illin ing g of av avai aila labl blee sites, and (3) reduce individual crop-yield loss per individual weed (Liebman and Gallandt, 1997). Use of Herbicides
Repeated herbicide treatments can cause a shift in weed populations or select for the development of resistant weed biotypes at the expen ex pense se of sus suscep ceptib tible le com commun munity ity mem member berss (H (Horo orowit witzz et al. al.,, 19 1962) 62).. Buchanan (1977) has published a list of herbicides that suppress certain weeds while encouraging others. When a maximum rate of 0.6 kg/ha of trifluralin (a,aa-trifluoro-2, 6-dinitro-N,N-dipropylp-
toluid tol uidine ine)) is app applie lied d bef before ore so sowin wing, g, pop popula ulatio tions ns of ve velv lvetle etleaf af (Abutilon theophrasti), jimson weed (Datura stramonium), venice mallow (Hiand d pr pric ickl kly y si sida da (Si can n be gr gro own am amon ong g biscus trionum trionum), ), an (Sida da spi spinosa nosa)) ca cotton and soybeans without the presence of other unwanted weed species. Although most examples cited by Buchanan (1977) relate to weed-control studies, similar methods may be developed to favor particular beneficial weeds in order to achieve early increases of natural enemy populations. Of course such an approach has no applicability in organic farming systems. Direct Sowing
Direct sowing of the grasses Eleusi Eleusine ne indica and Leptochloa form rm a on onee-me mete terr bo bord rder er ar arou ound nd be bean an fiel ields ds in Co Colo lomb mbia ia,, filiformis to fo decreased colonization and reproducti reproductive ve efficiency in the leafhopper Empoasca kraemeri (Altieri and Whitcomb, 1979a). The application of this method, however, demands careful investigation of certain weed-seed germination requirements. Some seeds remain in enforced dormancy and germinate only under certain environmental conditions. Most weed seeds have specialized requirements for germination, making it diff difficult icult to sow weeds for experimental purposes (Anderson, 1968). 1968). Nevertheless, Nevertheless, today in the market it is possible to to find fi nd man many y wee weed-s d-seed eed mix mixtur tures es (mo (mostly stly flo flower wering ing pla plants nts)) tha thatt are rec rec-ommended for planting in and around crop fields to create habitats for beneficial insects. Soil Disturbance
The weed-species composition of recently cropped fields can be mani ma nipu pula lated ted by ch chan angi ging ng th thee se seas ason on of di dist stur urba banc nce. e. In no nort rthe hern rn Fl Flor orid ida, a, field plots plowed at different times of year exhibited different weedspecies composition. Within these plots, populations of herbiv herbivorous orous insects fluctuated according to composition and abundance of weed hosts. Large numbers of chrysomelids and leafhoppers were collected in treatment plots where preferred weed hosts reached high cove co verr va value lues. s. As the these se her herbi bivo vores res ser serve ved d as alt altern ernati ative ve pre prey y, the num num-ber of pr preda edaceo ceous us art arthro hropo pods ds fee feedin ding g on the them m va varie ried d in dir direct ect pro propor por-tion to the size of populations of their t heir preferred herbivorous herbivorous prey as determined by the presence of weed hosts and the season of plowing
(Altieri and Whitcomb, 1979a). The authors proposed plowing strips of la land nd wi with thin in a cr crop op in di difffe fere rent nt se seas ason onss to en enco cour urag agee sp spec ecif ific ic we weed edss that, in turn, provide an alternative food and habitat to specific predators to rs (T (Tab able le 4. 4.5) 5).. If th this is is do done ne ea earl rly y in th thee se seas ason on,, a ba bala lanc ncee of na natu tura rall enemies can be maintained in the field, before outbreaks of pest species occur. Modifying Weed Spatial Patterns
It may be possible to influence weed spatial distributions and promote weeds to occur in clumps within fields rather than being uniformly distributed. For a given average density over a broad area, clumped weeds are expected to be less damaging to crop yield than are randomly or evenly distributed weeds (Liebman and Gallandt, 1997). Clumped weeds in a field spot may reduce yields there but prov pr ovid idee a so sour urce ce of be bene nefi fici cial alss th that at co colo loni nize ze th thee re rest st of th thee fi fiel elds ds fr from om the clump. Manipulating the Weed’s Critical Competition Period
Perhap Perh apss on onee of th thee mo most st us usef eful ul co conc ncep epts ts in ma mana nagi ging ng we weed edss fo forr in in-sect regulation within fields is the “critical period.” period.” This is the maximum period that weeds can be tolerated without affecting final final crop yields or the point after which weed growth does not affect final yield. In general, weeds that emerge earlier in the grow growing ing season are more damaging to crop yields than are populations that emerge later. later. On the other hand, crops differ in their sensitivity to different durations of weed competition, but most are most susceptible during the first third of their life cycle. The guiding principle here is to delay weed emergence relative to crop emergence (Liebman and Gallandt, 1997) Duration of weed-competition data for particular crops has been compiled by Zimdahl (1980) and critical weed-free maintenance periods have been identified for various crop-weed associations. The important question becomes how long exclusion efforts must be maintained before they can be relaxed, so that weeds emerge and provide the desired entomological benefits. As might be expected, the critical weed-free period for a given crop varies considerably ab ly am amon ong g si site tess an and d ye year ars, s, du duee to cl clim imat atee an and d ed edap aphi hicc co cond ndit itio ions ns af af-fecting crop and weed emergence and growth rates, weed-species composition, and weed density.
TABLE 4.5 4.5.. Sele Selecte cted d Exam Examples ples of Asso Associat ciations ions Betw Between een Her Herbiv bivorou orous s Ins Insects ects and Predaceous Arthropods Occurring on Specific Weed Species Which Respond to Particular Dates of Soil Disturbance in North Florida
Weed Species
Date(s) of Soil Disturbance That Enhances the Weed Population
Herbivore(s) Associated with the Weed That Serve As Alternate Prey to Various Predators
Predaceous Arthropods That Feed on the Herbivore(s) Associated with the Weed
1
Oenothera laciniata August (early evening primrose)
Altica sp. Altica sp. (leaf beetle)
Lebia viridus (ground beetle)
2
O. bie bienni nnis s (evening primrose)
December
Altica sp. Altica sp.
L. viri viridus dus
3
Amaranthus sp. Amaranthus sp. (pigweed)
April
Disonycha glabrata Lebia analis (leaf beetle) (ground beetle)
4
Heterotheca subaxillaris (camphorweed)
October
Zygogramma heterothecae (leaf beetle) and 30 other herbivorous insect species
30 predaceous arthropod species
5
Chenopodium ambrosiodes (Mexican tea)
December
Z. sutu sutural ralis is and and 31 other herbivorous insect species
34 predaceous arthropod species
6
Solidago altissima (goldenrod)
December
Uroleucon spp. Uroleucon spp. 58 predaceous (11 aphid species) arthropod species and 28 other herbivorous species
7
Ambrosia artemisiifolia (ragartemisiifolia (ragweed)
December October
Z. sutura suturalis, lis, Nodonota sp. Nodonota sp. (leaf beetles) Reuteroscopus ornatus (plant ornatus (plant bug) Uroleucon ambrosiae (aphid) ambrosiae (aphid) Epiblema sp., Epiblema sp., Tarachidia sp. Tarachidia sp. (caterpillars) and 17 other herbivorous insect species
Cycloneda sanguinea (lady beetle) Zelus cervicalis and Sinea Sinea spp. spp. (assassin bugs) Peucetia viridan s (lynx spider) and 4 other predaceous arthropod species
8
Lactuca canadensis Control (wild lettuce) (not disked)
Uroleucon sp. Uroleucon sp. (aphids)
Podabrus sp. Podabrus sp. (soldier beetle) Cycloneda sanguinea Chrysopa sp. Chrysopa sp. (lacewing) Doru taeniatum (earwig) Syrphids and spiders
Source: After Source: After Altieri and Letourneau, 1982.
In studies in southern Georgia, Altieri et al. (1981) observed that popul po pulati ations ons of the ve velv lvetb etbean ean cat caterp erpill illar ar (Antic (Anticarsia arsia gemm gemmatalis) atalis) and of the southern green stink bug (Nezara viridula) were greater in weed we ed-f -fre reee so soyb ybea eans ns th than an in so soyb ybea eans ns le left ft we weed edy y fo forr ei eith ther er tw two o or fo four ur weeks after crop emergence or for the whole season. Soybeans maintained weed free for two or four weeks after emergence required no furth fu rther er wee weed d con contro troll to pro produc ducee opt optimu imum m yie yield ld (W (Walk alker er et al. al.,, 19 1984) 84).. In another experiment conducted in California, allowing weed growth grow th during selected periods of the collard crop cycle (two or four weeks weed free or weedy all season) resulted in lower flea beetle dens nsit itie iess in th thee we weed edy y mo mono nocu cult ltur ures es th than an in th thee we weed ed-(P.. cruciferae) de (P free monocultures. Lowest densities occurred in systems allowed to remain weedy all season. No differences in the abundance of beetles were we re ob obse serv rved ed be betw twee een n co colla llard rdss ke kept pt we weed ed fr free ee fo forr tw two o or fo four ur we week ekss after transplanting (Altieri and Gliessman, 1983). Beetle densities were lower in these systems than in the weed-free system. In collard systems with “relaxed” weeding regimes, flea beetle densities were at least l east five times greater on a per-plan per-plantt basis on Brassica campestris (the dominant plant of the weed community) than on collards. B. campestris germinated quickly and flowered early, each plan pl antt av aver erag agin ing g a he heig ight ht of 39 cm wi with th tw twel elve ve le leav aves es an and d si sixt xtee een n op open en flowers, sixty days after germination. This apparent preference of Phyllotreta cruciferae for B. campestris over collards resulted in a higher concentration of flea beetles on the wild crucifer, diverting flea beetles from collards and consequently diluting their feeding on the collards colla rds (Table (Table 4.6). Collards Col lards grown under various levels of weed diversity exhibited significantly less leaf damage than monoculture collar col lards ds gr grow own n in wee weed-f d-free ree sit situat uation ionss (A (Altie ltieri ri and Gli Gliess essman man,, 198 1983). 3). In a similar experiment, wild mustard sowed one week after broccoli transplanting showed no reduction of broccoli yield and reduced aphid aph id nu numbe mbers rs whi while le inc increa reasin sing g ef effec fecti tive ve pre predat dation ion by syr syrph phid id lar larva vaee (Kloen and Altieri, 1990). Defining periods of weed-free maintenance in crops so that numbers be rs of pe pest stss do no nott su surp rpas asss to toler lerab able le le leve vels ls mi migh ghtt pr prov ovee to be a si sign gnif if-icant compromise between weed science and entomology, a necessary sar y ste step p to fur furthe therr ex explo plore re the int intera eracti ction onss des descri cribed bed in thi thiss cha chapte pterr. Unquestionably,, weeds stress crop plants through interference proUnquestionably cesses. How However ever,, substantial evidence evidence suggests that weed presence presence in crop fields cannot be automatically judged damaging and in need of
TABLE 4.6. Mean Flea Beetle (Phyllotreta cruciferae) Densities, cruciferae) Densities, Weed and Crop Biomass in Various Collard Cropping Systems in Santa Cruz, California
No. of flea beetles
Cropping System
Per 10 Collard Plants
Leaves in Each Collard Plant with Beetle Damage (45 Per 5 Brassica Days After Transplanting) Weed Biomass campestris Plants (%) (g/m2)
Weed free all season
34.0
2.6a
—
Weed free for 4 weeks after collard transplanting
29.3
1.7b
78.1
Weed free for 2 weeks after collard transplanting
29.0
1.7b
Weedy all season
6.6
3.8c
Whole-Plant Collard Dry Weight (g/m2)
54.4a
0
213.6
16.3a
44.6b
52.3
361.3
73.7
20.1a
44.5b
55.2
243.0
25.0
11.5b
29.9c
483.2
226.1
Source: After Source: After Altieri and Gliessman, 1983. Note: Means followed by the same letter in each column are not significantly different (P ( P = 0.05). All means are averages of the three sampling dates (15, 30, and 45 days after transplanting)
immediate control. In fact, crop-weed interactions are overwhelmingly site specific and vary according to plant species involved, environmental factors, and management practices. Thus, in many agroecosystems, weeds are ever-present components adding to the complexity of interacting trophic levels mediating a number of cropinsect interactions with major effects on final yields. It is here argued that in weed-diversified systems we cannot understand plant-herbivore interactions without understanding the effects of plant diversity on natural enemies, nor can we understand predator-prey and parasite-host interactions without understanding the role of the plants involved volv ed in the system. An increasing awareness of these ecological relationships should plac pl acee em emph phas asis is on we weed ed ma mana nage geme ment nt,, as op oppo pose sed d to we weed ed co cont ntro rol, l, so that herbicides may be considered merely a component part of a total system for managing weeds, where season-long, weed-free monocultures are not always assumed to be the best crop-production strategy (Aldrich, 1984).
Chapter 5
Insect inManagement Insect Management Multiple-Cropping Multiple-Cro pping Systems in Multiple-Cropping Systems The single-species nature of crop systems can be broken by growing crops in polycultural patterns. Polycultures are systems in which two tw o or mor moree cro crops ps are usu usually ally pla plante nted d sim simult ultane aneou ously sly and suf suffi ficie cientl ntly y close together to result in interspecific competition and/or complementation. These interactions may have inhibiting or stimulatory effects on yields (Hart, 1974). In the design and management of these syst sy stem ems, s, on onee st stra rate tegy gy is to mi mini nimi mize ze co comp mpet etit itio ion n an and d ma maxi ximi mize ze co commplementation among species in the mixture (Francis, Flor, and TemTemple, 1976). Among the potential advantages that can emerge from the intelligent design of polycultures are population reduction of insect pests, suppression of weeds through shading by complex canopies or allelopathy (Gliessman and Amador, 1980), better use of soil nutrients (Igzoburkie, 1971), and improved productivity productivity per unit of land (Harwood, 1974). In th thee tr trop opic ics, s, po poly lycu cult ltur ures es ar aree an im impo port rtan antt co comp mpon onen entt of tr trop opica icall small-farm agriculture and, in addition to lowering risks, one of the reasons for the evolution and adoption of such cropping patterns may be the reduced incidence of insect pests (Altieri and Liebman, 1988; Alghail, 1993). Polyculture systems may also provide a potential for impro imp rove ved d cro crop p pr produ oducti ctivit vity y ev even en in tem temper perate ate agr agricu icultu lture. re. In fac fact, t, the systems are so prevalent that quantitative estimates suggest that 98 perc pe rcen entt of th thee co cowp wpea eass gr gro own in Af Afri rica ca an and d 90 pe perc rcen entt of th thee be bean anss in Colombia are intercropped. The percentage of cropped land in the tropics actually devoted to intercropping varies from a low of 17 percent in India to a high of 94 percent in Malawi. Apparently, Apparently, in El Salvador, it used to be impossible to find sorghum that was not intercropped with maize. Even in temperate North America, before the widespread use of modern varieties and mechanization, intercropping was apparently common (e.g., 57 percent of the soybean acreage in
Ohio was grown in combination with maize in 1923) (Vandermeer (Vandermeer,, 1989). In the Midwest of the United States, the combination of soybean be an an and d co corn rn in st stri rip p in inte terc rcro ropp ppin ing g ha hass be been en tr trie ied d as an ec econ onom omic ic al al-ternativee to monocultures (Francis, 1990). ternativ Polycu Pol ycultu lture re man manage agemen mentt bas basica ically lly con consis sists ts of the des design ign of spa spatia tiall and temporal combinations of crops in an area. The arrangement of crops in space can be in the form of such systems as strip cropping, intercropping, mixed-row cropping, and cover cropping (Andrews and an d Ka Kass ssam am,, 19 1976 76). ). Th Thee cr crop op ar arra rang ngem emen entt in tim timee ca can n var ary y ac acco cord rdin ing g to whether mixed crops are planted simultaneously or in sequence as rotational crops, relay crops, or ratoon crops, or whether crops are combined in a synchronous or an asynchrono asynchronous us fashion or in a continuous or discontinuous planting pattern (Litsinger and Moody, 1976). According to Francis, Flor, and Temple (1976), desirable features of crops to be considered for intercropped systems include photo ph otoper period iod ins insens ensiti itivit vity y, ear early ly and un unifo iform rm mat maturi urity ty,, lo low w sta statur turee and nonlodging effects, effects, good population response, insect and disease resistance, efficient soil-fertility response, and high yield potential. Multiple-c Multi ple-cropp ropping ing syste systems ms cons constitute titute agric agricultur ultural al syste systems ms div diversiersifie ied d in ti time me an and d sp spac ace. e. As me ment ntio ione ned d ea earl rlie ierr, mu much ch evi vide denc ncee su sugg gges ests ts that this vegetational diversity often results in a significant reduction of insect-pest problems (Altieri and Letourneau, 1982). A large body of literature cites specific crop mixtures that affect particular insect pests (Litsinger and Moody Moody,, 1976; Perrin, 1977; Perrin and Phillips, 1978; Andow, Andow, 1983a), and other papers explore explor e the ecological ecologic al mechanisms involved in pest regulation (Root, 1973; Bach, 1980a,b; Risch, 1981). In polycultures, apart from the evident increase in plant-species diversity, there are changes in patch size, plant density, and plant quality. All of these changes affect density of pests and other oth er or organ ganism isms. s. Cle Clearl arly y, muc much h kn know owled ledge ge has acc accumu umulat lated, ed, and thi thiss acquired information is slowly providing a basis for designing complex crop systems so that pest problems and the need for active control measures are minimized (Murdoch, 1975). PATTERNS OF INSECT ABUNDANCE IN POLYCULTURES
In recent years, agroecologists have conducted experiments in multiple-cropping systems to test the theory that increased plant di-
versity fosters stability of insect populations (Pimentel, 1961; Root, 1973; Van Van Emden and Williams, 1974). A recent examination of all available ava ilable studies on the effects of these patterns on insect-pest populations tends to support the theory, theory, although confusion may arise depending on how diversity and stability are defined (Risch, Andow, and Altieri, 1983). In multiple cropping, structural and species vegetational diversity (a measure of the biotic, structural, and microclimatic complexity arising from the mixing of different plants) results from the addition of crop species in time and space. Stability herein refers to low pest-population densities over time. Exam Ex ampl ples es of sp spec ecif ific ic cr crop op mi mixt xtur ures es th that at re resu sult lt in re redu duce ced d pe pest st in inci ci-dence can be found in Litsinger and Moody (1976), Altieri and Letourneau (1982), Andow (1983b), and Altieri and Liebman (1988) and are summarized in Table 5.1. Thirty-five insect species were investigated in fifty insect studies. The majority of the insects were in the orders Lepidoptera, Coleoptera, and Homoptera, accounting for 42, 32, and 18 percent, respectively, of the total crop pests. A combination of lowered resource concentration, trap cropping, various various diversi ve rsion onary ary mec mechan hanism isms, s, pla planti nting ng den densit sity y, and pla plant nt ph physi ysical cal obs obstru trucction account for 22.5 percent of the factors explaining pest reduction. Predators and parasites account for only 15 and 10 percent, respective ti vely ly,, wh where ereas as mas maskin king g and and/or /or cam camou oufla flagin ging g and rep repelle ellenc ncy y acc accoun ountt for 12.5 percent each. Overall natural enemy action was responsible forr up to 30 pe fo perc rcen entt of th thee co cont ntro roll of th thee st stud udie ied d cr crop op pe pest sts, s, an and d th thee re re-maining known cases were controlled by other factors. By analyzing a series of case studies, Helenius (1991) showed that monophagous insects are more susceptible to crop diversity than are polyphagous insects and cautioned the increased risk of pest attack if the dominant herbivore fauna in a given agroecosystem is polyphagous. The reduction in pest numbers for monophagous insects was almost twice (53.5 percent of the case studies showed lowered numbers in polycultures) that for polyphagous insects (33.3 percent). Coll (1998) compared parasitoid density and parasitism rates in forty fo rty-tw -two o dif differ ferent ent mon monocu ocultu lturere-int interc ercro rop p sys system temss rep report orted ed in the lit lit-erature. In two-thirds of the comparisons, the parasitoids were more abundant or attacked more hosts in the intercropped than monocultur cul tured ed hab habita itats. ts. Ho Howe weve verr, in abo about ut on one-t e-thir hird d of the com compar pariso isons, ns, no consistent differences were recorded in parasitoid density or parasitism rate among habitats. A lower rate of parasitoid enhancement was
TABLE 5.1. Selected Examples of Multiple-Cropping Systems That Effectively Prevent Insect-Pest Outbreaks Multiple-Cropping System
Pest(s) Regulated
Factor(s) Involved
Beans grown in relay intercropping with winter wheat
Empoasca fabae and Aphis fabae
Impairment of visual searching behavior of dispersing aphids
Brassica crops and beans
Brevicoryne brassicae and Delia brassicae
Higher predation and disruption of oviposition behavior
Brussels sprouts intercropped with fav fava a beans and/or mustard
Flea beetle Phyllotreta crucifecae and crucifecae and cabbage aphid Brevicoryne brassicae
Reduced plant apparency trap cropping, enhanced biological control
Cabbage intercropped Erioischia brassicae, with white and red clover cabbage aphids, and imported cabbage butterfly (Pieris rapae)
Interference with colonization and increase of ground beetles
Intercropping of Cajanus Pod borers, jassids, cajan with cajan with red, black, and and membracids green gram
Delayed colonization of herbivores
Cassava intercropped with cowpeas
Changes in plant vigor and increased abundance of natural enemies
Whiteflies Aleurotrachelus socialis and Trialeurodes variabilis
Cauliflower strip cropped Blossom beetle with rape and/or marigold Meligethes aeneus
Trap cropping
Corn intercropped with beans
Leafhoppers (Empoasca kraemeri) leaf beetle (Diabrotica balteata) and balteata) and fall armyworm (Spodoptera frugiperda)
Increase in beneficial insects and interference with colonization
Corn intercropped with fava fav a beans and squash
Aphids, Tetranychus urticae and Macrodactylus sp.
Enhanced abundance of predators
Corn intercropped with clover
Ostrinia nubilalis
Unknown
Corn intercropped with soybean
European corn borer Ostrinia nubilalis
Differences in corn varietal resistance
Corn intercropped with sweet potatoes
Leaf beetles (Diabrotica Diabrotica spp.) spp.) and leafhoppers (Agallia lingula)
Increase in parasitic wasps
Multiple-Cropping System
Pest(s) Regulated
Factor(s) Involved
Intercropping corn and beans
Dalbulus maidis
Interference with leafhopper movement
Cotton intercropped with forage cowpea
Boll weevil (Anthonomus grandis)
Population increase of parasitic wasps (Eurytoma Eurytoma sp.) sp.)
Intercropping cotton with sorghum or maize
Corn earworm (Heliothis zea)
Increased abundance of predators
Cotton intercropped with okra
Podagrica s Podagrica sp.
Trap cropping
Strip cropping of cotton and alfalfa
Plant bugs (Lygus hesperus and L. elis elisus) us)
Prevention of emigration and sychrony in the relationship between pests and natural enemies
Strip cropping of cotton and alfalfa on one side and maize and soybean on the other
Corn earworm (Heliothis zea) and cabbage looper (Trichoplusia (T richoplusia ni)
Increased abundance of predators
Intercropping cowpea and sorghum
Leaf beetle (Oetheca benningseni)
Interference of air currents
Cucumbers intercropped Acalymma vittata with maize and broccoli
Interference with movement and tenure time on host plants
Groundnuts intercropped Aphis craccivora with field beans
Aphids trapped on epidermal hairs of beans
Maize intercropped with canavalia
Not reported
Prorachia daria and daria and fall armyworm (Spodoptera frugiperda)
Maize-bean intercropping Spodoptera frugiperda and Diatraea lineolata
Lower oviposition rates, trap cropping
Strip cropping of muskmelons with wheat
Myzus persicae
Interference with aphid dispersal
Oats intercropped with field beans
Rhopalosiphum sp Rhopalosiphum sp..
Interf Inte rfer eren ence ce wit ith h ap aphi hid d dispersal
Peaches intercropped with strawberries
Strawberry leafroller (Ancylis comptana) Oriental fruit moth (Grapholita molesta)
Population increase of parasites (Macrocentrus ancyclivorus, Microbracon gelechise, and Lixophaga variabilis)
Peanut intercropped with maize
Corn borer (Ostrinia furnacalis)
Abundance of spiders (Lycosa Lycosa sp.) sp.)
TABLE 5.1 (continued) Multiple-Cropping System
Pest(s) Regulated
Factor(s) Involved
Sesame intercropped with corn or sorghum
Webworms (Antigostra Antigostra sp.) sp.)
Shading by the taller companion crop
Sesame intercropped with cotton
Heliothis s Heliothis spp.
Increase of beneficial insects and trap cropping
Soybean strip cropped with snap beans
Epilachna varivestis
Trap cropping
Squash intercropped with maize
Acalymma thiemei, Diabrotica balteata
Increased dispersion due to avoidance of host plants shaded by maize and interference with flight movements by maize stalks
Tomato and tobacco intercropped with cabbage
Flea beetles (Phyllotreta Feeding inhibition cruciferae) by odors from nonhost plants
Tomato intercropped with cabbage
Diamondback moth (Plutella xylostella)
Chemical repellency or masking
Source: Based Source: Bas ed on Al Alti tieri eri et al al.. (1 (1978 978), ), Alt Altier ierii an and d Le Leto tourn urnea eau u (1 (198 982) 2),, an and d And Andow ow (1991a).
found foun d wh when en da data ta we were re an analy alyze zed d by pa para rasi sito toid id sp spec ecie iess or gr grou oup p of sp speecies. Only 54 percent of the thirty-one studied species had a higher parasitism rate or density in intercrops compared to monocultures (39 percent showed similar or variable activity levels in simple and diverse diver se habitats). These data suggest that the response of some species to intercropping differs with different crop combinations, geographic location, and experimental procedures. Experiments reporting results in which no differences were observed or in which higher pest incidence occurred in multicrops are quite uncommon. A particular crop mix might be of value in controlling one pest in one area (i.e., Heliothis virescens in corn [Zea mays] and cotton [ Gossypium sp.] strip cropping in Peru), while increasing the same pest in other areas (i.e., H. virescens in Tanzania) (Smith and Rey Reyno nolds lds,, 19 1972) 72).. In Nig Nigeri eria, a, pop popula ulatio tions ns of flo flower wer thr thrips ips (Megalurothrips sjostedti) were reduced by 42 percent on cowpea (Vigna However, cropping pattern had no unguiculata)-maize polycultures. However,
effect on infestations of Maruca testulalis, pod-sucking bugs, and meloid beetles (Matteson, Altieri, and Gagne, 1984). Early infestations of Maruca were no different in monocrops and polycultures of maize mai ze and co cowpe wpea, a, bu butt twe twelv lvee wee weeks ks aft after er pla planti nting ng,, inf infest estatio ations ns wer weree significantly higher in the monocrops. Similar shifts were observed with Laspeyresia and thrips (Matteson, Altieri, and Gagne, 1984). In India, larval populations of Heliothis armigera were higher in sorghum (Sorghum bicolor)-pigeon pea (Cajanus cajan) intercropping systems than in sole pigeon pea plots, which led to higher grain losses in polycrops (Bhatnagar and Davies, 1981). In home-garden plots of beans (Phaseolus vulgaris) bordered by marigolds ( Tagetes spp.), Latheef and Irwin (1980) reported that their designs did not favor control of Heliothis zea and Epilachna varivestis. In Georgia, Nordlund, Chalfant, and Lewis (1984) did not find significant reductions of Heliothis zea damag damagee in maize ears, bean pods, or tomato fruits in polycultures of maize, bean, and tomato. In the Philippines, Hasse and Litsinger (1981) found that intercropping maize with legumes did not reduce the numbers of egg masses laid by corn borers (Ostrinia furnacalis).
Certain associated plants can have an adverse effect on parasitoids by acting as traps for searching adults. For example, Lysiphlebus testaceipes (Hymenoptera: Aphiidae), a parasitoid of cotton aphid (Aphis gossypii) (Homoptera: Aphidae), was entrapped by glandular exudates of petunia, and thereby prevented from protecting nearby okra (Abelmoschus esculentus) plants from aphid attacks (Marcovitch, 1935); Trichogramma minutum (Hymenoptera: Tricho Trichogramgrammatidae) adults could not parasitize hornwor hornworms ms Manduca sp. (Lepidoptera: Sphingidae) on tomato (Lycopersicon esculentum) plants because of entrapment by the sticky glandular trichromes on the leaves of the associated tobacco (Nicotiana tobacum) plants (Marcovitch, vit ch, 19 1935) 35).. Alt Altho hough ugh hig high h par parasi asitis tism m of Heliothis eggs gs by Heliothis armig armigera era eg Trichogramma sp. occurred in sorghum (Sorghum halepense), a low parasitism rate was recorded in sorghum associated with chickpea (Cicer arietinum) because adult wasps were trapped in its sticky trichromes (Van Emden, 1990). Despit Des pitee the these se rep report orts, s, use of int interc ercro roppi pping ng has bee been n wid widely ely rec recom om-mended as a management strategy to reduce insect damage. A reduced insect-pest incidence in multicrops may be the result of increased parasitoid and predator populations, higher availability of alternate food for natural enemies, decreased colonization and repro -
duction ductio n of pes pests, ts, che chemic mical al rep repell ellenc ency y, mas maskin king g and and/or /or fee feedin ding g inh inhibi ibi-tion from nonhost plants, prevention of pest movement and/or emi gration, and optimum synchrony between pests and natural enemies (Matteson, Altieri, and Gagne, 1984). Perrin and Phillips (1978) described the stages in pest population development development and dynamics that may be affected by mixed cropping. At the crop colonization stage, they the y po postu stulate late tha thatt dis disrup ruptio tion n of olf olfact actory ory and vis visual ual res respo ponse nses, s, ph physysical barriers, and diversion to other hosts are important mechanisms regulating herbivores in multiple-cropping systems. Once the pests become established in the field, their populations may be regulated by limitation of dispersal, feeding disruption, reproduction inhibition, and mortality imposed by biotic agents (Figure 5.1).
SURVIVAL + BUILDUP AT SOURCE SOUR CE
DISRUPTION OF OLFACT OLFACTORY/ ORY/ VISUAL RESPONSES INVASION AND SETTLING ON CROP
PHYSICAL BARRIERS TO COLONISTS DIVERSION TO TOLERANT/ LESS IMPORTANT HOSTS
LIMITED DISPERSAL BETWEEN PLANTS POPULATION DEVELOPMENT AND SURVIVAL
DISRUPTION OF FEEDING AND REPRODUCTION CHANGES IN MORTALITY DUE TO NATURAL ENEMIES
EMIGRATION TO OTHER CROPS AND/OR OFF-SEASON HABITATS
FIGURE 5.1. Stages in pest-population dynamics which may be affected by mixed cropping (after Perrin and Phillips, 1978).
Hasse and Litsinger (1981) summarized several several mechanisms that supposedly explain pest reduction in intercropping systems. A list of the proposed mechanisms mecha nisms is given in Table Table 5.2. Most of the proposed propose d mechan mec hanism ismss are acc accoun ounted ted for by the res resou ource rce con concen centra tratio tion n and ene ene-mies hypotheses discussed in Chapter 3.
HERBIVORE TRENDS IN POLYCULTURES
Andow (1991a) contends that herbivore movement patterns are more important than the activities of natural enemies in explaining the reduction of monophagous pest populations in diverse annual crop systems. Two classic studies support this view. The first study (Risch, 1981) looked at the population dynamics of six chrysomelid beetles in monocultures and polycultures of maize-bean-squash (Cucurbita pepo). In polycultures containing at least one nonhost plant (maize), the number of beetles per unit was significantly lower relative to the numbers of beetles on host plants in monocultures. Measure su reme ment nt of be beetl etlee mo move veme ment ntss in th thee fie ield ld sh sho owe wed d th that at be beetl etles es te tend nded ed to emigrate more often from polycultures than from host monocultures. Apparently, this was due to several factors: (1) beetles avoided avo ided host plants shaded by maize; (2) maize stalks interfered with flight movements of beetles; and (3) as beetles moved through polycultures, they remained on nonhost plants for a significantly shorter time than spent on host plants. There were no differences in rates of parasitism or predation of beetles between systems. Thee se Th seco cond nd st stud udy y ex exam amin ined ed th thee ef effe fect ctss of pl plan antt di dive vers rsit ity y on th thee cu cu-cumber cumb er beetle beetle,, Acalym (Bach,, 1980 1980a). a). Popu Population lation dens densities ities Acalymma ma vittata (Bach were significantly greater in cucumber (Cucumis sativus) monocultures than in polycultures containing cucumber and two nonhost species. Bach also found greater tenure time of beetles in monocultures than in polycultures. She determined that these differences were caused by plant diversity per se and not by differences in hostplant density or size. Thus, they do not reveal if differences differences in numbers of herbivores herbivores between monocultures and polycultures are due to diversity diver sity or rather to the interrelated and confounding effects of plant diversity, plant density, and host-plant patch size. In northern California, densities of cabbage aphids (Brevicoryne brassicae) and flea beetles (Phyllotreta cruciferae) were signifi-
TABLE 5.2. Possible Effects Effects of Intercropping on Insect-Pest Populations Factor
Explanation
Example
Interference with host-seeking behavior Camo Ca mouf ufla lage ge
A ho host st pl plan antt ma may y be pr prootected from insect pests by the physical presence of other overlapping plants.
Camouflage of bean seedlings by standing rice stubble for beanfly
Crop backg background round
Certain pests pref prefer er a crop background of a particular color and/or texture.
Aphids, flea beetles, and Pieris rapae are rapae are more attracted to cole crops with a background of bare soil than to ones with a weedy background
Masking or dilution of attractant stimuli
Presence of nonhost plants Phyllotreta cruciferae in cruciferae in colcan mask or dilute the attrac- lards tant stimuli of host plants leading to a breakdown of orientation, feeding, and reproduction processes.
Repellent chemical stimuli
Aromatic odors of certain Grass borders repel plants can disrupt host-find- leafhoppers in beans; ing behavior. populations of Plutella xylostella are xylostella are repelled from cabbage-tomato intercrops Interference with population development and survival
Mechanical barriers
All companion crops may block the dispersal of herbivores across the polyculture. Restricted dispersal may also result from mixing resistant and susceptible cultivars of one crop by settling on nonhost components.
Lack of arrestant stimuli
The presence of different host and nonhost plants in a field may affect affect colonization of herbivores. If a herbivore descends on a nonhost, it may leave the plot more quickly than if it descends on a host plant.
Microclimatic influences
In an intercropping system, favorable favorable aspects of mircroclimate conditions are highly fractioned; fractioned; therefore, insects may experience difficulty in locating and remaining in suitable microhabitats. Shade derived from denser canopies may affect feeding of certain insects and/or increase relative humidity which may favor favor entomophagous fungi.
Biotic influences
Crop mixtures may enhance natural-enemy complexes co mplexes (see natural-enemy hypothesis in text).
Source: After Source: After Hasse and Litsinger, 1981; and Litsinger et al., 1991.
cantly lower on cauliflower plants grown in association with vetch (Vicia sp.) than in clean-cultiv clean-cultivated ated monocultures (Altieri, 1984). The depression of crop growth and biomass in the diverse plots added a confounding confoundin g effect in that it was not clear cl ear if herbivore reduction resulted sul ted fro from m po poore orerr pla plant nt qu qualit ality y, wh which ich mad madee cau caulif liflo lower werss les lesss attr attracactive to the herbivores. In another study, flea beetle numbers were significantly lower in collards associated with wild mustard (Brassica campestris) than in monocultures (Altieri and Gliessman, 1983). Flea beetles preferred this plant over collards, thus flea beetles were diverted from collards resulting in diluted feeding on the collards. The authors argue that wild wil d mus mustar tards ds ha have ve hig higher her con concen centra tratio tions ns of ally allylis lisoth othioc iocyan yanate ate (a po powwerful attractant to flea beetle adults) than do collards, and therefore the preference of flea beetles for wild mustard simply reflected different degrees of attraction to the foliage levels of this particular glucosinolate in the weeds and collards. Figure 5.2 illustrates this t
14 n al p
12 dr al l o c/
10 s el t e
8 e b a el
6 f r e b
4 m u n e
2 g ar e v A
0
19
27
34
41
48
55
62
66
Days after planting collard monoculture-normal density collard monoculture-double density wild mustard intercrop barley intercrop
FIGURE 5.2. Population FIGURE Populations s trend trends s of Phyllotreta cruciferae in collard monocultures and in collard polycultures mixed with a host plant (wild mustard) and a nonhost plant (barley) (after Altieri and Schmidt, 1986b).
preference in the field by showing that population densities of flea beetles on collard plants grown as monocultures are greater than on collards intercropped with wild mustards and with nonhost barley (Hordeum vulgare) (Altieri and Schmidt, 1986b). Although the barley effect might support the resource concentration hypothesis, the trap-cropping effect effect of wild mustards exerts a stronger influence on beetle abundance in this case. Thee sa Th same me st stud udy y al also so sh sho owe wed d th that at re remo mov val of fl flo owe wers rs of wi wild ld mu musta stard rdss results in a substantial reduction of the attractant effect (Table 5.3). Consequently, collard plants in flowerless intercrops experienced greater flea beetle loads than collards within the intercrop with flowers and even monocultures. Risch, Andow, Andow, and Altieri (1983) collected additional data that do not support the enemies hypothesis. They found that predation rates on egg masses of the European corn borer (Ostrinia nubilalis) by a predaceou pred aceouss beetl beetlee (Coleo signific ificantly antly high higher er (Coleome megilla gilla macula maculata) ta) were sign in maize monocultures than in the more densely planted maize-beansquash polyculture. They argue that in polycultures, the beetles apparently spent more time foraging on plants (beans and squash) that contain con tained ed no bo borer rer eg eggs, gs, thu thuss dec decrea reasin sing g the their ir fo forag raging ing ef effec fecti tive venes ness. s. Even if prey densities per maize plant were the same in the two culture types, beetles might forage less efficiently in the polyculture due to unrewarded time spent foraging on bean and squash plants. This lower reward rate leads to faster emigration of beetles from polycultures (Wetzler and Risch, 1984). TABLE 5.3 5.3.. Flea Bee Beetle tle Num Number bers s per Col Collard lard Plant in Mon Monocu ocultu ltures res and Polycultures with and Without Wild Mustard (Brassica kaber) Flowers kaber) Flowers Days after planting* 30
44
57
Monoculture:
norm rma al density
30.0
6.9
49.6
19.7
10.1
6.6
Monoculture:
double density
40.0
24.0
79.6
43.6
6.5
2.8
Polyculture:
with flowers
5.6
2.5
10.6
5.5
1.6
0.6
Polyculture:
without flowers
**7.6
3.5
81.0
38.3
5.8
2.4
Source: After Source: After Altieri and Schmidt, 1986b. *Means derived from different counts on 10 random plants per plot **Flowers not yet removed
Wrubel (1984) contends that visual camouflage from nonhost plants may have resulted in more Mexican bean beetles colonizing soybean (Gl monocultu culture re than maize maize-soy -soybean bean inter intercrop cropped ped (Glycin ycinee max max)) mono plots. Conversely, the higher concentration of food resources in clover-soybean clover -soybean (two legumes) than soybean monoculture plots may explain the slightly higher abundance of polyphagous acridids that Wrubel found in the t he clover-soybean clover-soybean plots. Differences in the structure of the crop canopy in tall maize-soybean and short maize-soybean plots appeared to affect the behavior of several groups of herbivores, with lower abundance abundance of Japanese beetles due to shading of the soy soy-bean canopy by the taller maize plants. There The re are are,, ho howe weve verr, stu studie diess tha thatt sup suppor portt the ene enemie miess hy hypot pothes hesis. is. In tropical corn-bean-squash systems, Letourneau (1983) studied the importance of parasitic wasps in mediating the differences in pest abund ab undanc ancee bet betwee ween n sim simple ple and com comple plex x cro crop p arr arrang angeme ements nts.. A squ squash ash-Diaphania nia hyalina hyalinata ta (Lepidoptera: Pyralidae), feeding caterpillar, Diapha occurred at low densities on intercropped squash in tropical t ropical Mexico. Part of the effect of the associated maize and bean plants may have been to render the squash plants less apparent to ovipositing moths. Polyculture fields also harbored greater numbers of parasitic wasps than did squash monocultures. Malaise trap captures of parasitic wasps in monoculture consisted of one-half the number of individu individu-als caught in mixed culture. The parasitoids of the t he target caterpillars were also represented by higher numbers in polycultures throughout the season. Not only were parasitoids more common in the vegetationally diverse, traditional system, but but also the parasitization rates of D. hyalinata eggs and larvae on squash were higher in polycultures. Approximately 33 percent of the eggs in polyculture samples pl es ov over er th thee se seas ason on we were re pa para rasi siti tize zed d wh whil ilee on only ly 11 pe perc rcen entt of eg eggs gs in monocultures were. Larval samples from polycultures showed an incidence of 59 percent parasitization for D. hyalinata larvae, whereas samples from monoculture larval specimens were 29 percent parasitized. Another study conducted in Davis, California, tested whether predator colonization rates could be manipulated through vegetational diversity (Letourneau and Altieri, 1983). The densities of Orius tristicolor and its preferred prey, Frankliniella occidentalis, were compared between squash monocultures and polycultures of squash squ ash,, cor corn, n, and co cowp wpea. ea. The pat patter terns ns of pr preda edator tor col coloni onizat zation ion rat rates es
and pes pestt den densit sities ies in the these se two cro croppi pping ng sys system temss par parall allele eled d tho those se do doccumente ume nted d fo forr pre predat dator or-pr -prey ey int intera eracti ction onss in the mit mite-g e-grap rapev evine ine sys system temss of Flaherty (1969). In both studies, the colonization rate of predators was increased in diverse habitats, and the prey (pest) populations in each case reached lower maximum levels. In Flaherty’s study, the causativee factor was the close proximity of the source of colonizing causativ predators. The great variation between levels of Williamette mite (Eotetranychus willamette) infestation on individual grape vines was caused by their variable proximity to clumps of johnsongrass (Sorghum halepense). The grass supported an alternate host for a predatory mite, Metaseiulus occidentalis. The predators then colonized contiguous vines sufficiently sufficiently early to suppress the pest-mite populations. In the Davis study, the sources of colonizers were presumably at similar distances to randomly assigned plots of monoculture and polyculture. The authors suggest that the determining factor for differential colonization by Orius sp. in monocultures and polycultures of squash was attraction to the early-season complex-crop habitat during the host-location process. Results showed that mean density of thrips on squash leaves was initially much greater in monoculture than in polyculture and remained at significantly higher levels until sixtysix ty-fi five ve day dayss aft after er so sowin wing. g. The Orius sp. den densit sity y, ho howe weve verr, wa wass sig sig-nificantly higher on squash early in the season (days 30 and 42) in polyculture. A decrease in prey density accompanied an increase in adult Orius sp. colonization in both treatments until thrips reached low densities (Figure 5.3). Predator manipulation experiments conducted in field cages, in which Orius sp. pop popula ulatio tions ns wer weree eit either her inc includ luded ed or ex exclu cluded ded,, sho showed wed that the density of thrips was influenced by predation by Orius sp. (Letourneau and Altieri, 1983). On uncaged control plants, the mean density of thrips per leaf declined steadily from day 50, as it had in the general field samples. Inside the exclusion/inclu exclusion/inclusion sion cages, thrip density more than tripled the first week after Orius and Erigone spp. spiders were eliminated. When predators, equal in number to those that were eliminated, were added to cage 1, the thrip density fell in this cage. Altieri (1984) found that brussels sprouts grown in polycultures with fava beans or wild mustard supported more species of natural enemies (six species of predators and eight species of parasites) than monocultures (three species of predators and three species of parasites). Apparently, Apparently, the presence of flowers, extrafloral nectaries, and
250
200
150
100
50
0 4 3 2 1 0 30
35
40
45
50
55
60
65
70
75
80
FIGURE 5.3. Predator-prey relationships in monoculture and polyculture plots from samples taken on squash leaves (after Letourneau and Altieri, 1983).
alternate prey/hosts associated with the companion plants allowed this enhancement. Aphid densities were also lower in such systems, apparently due to increased mortality imposed by the more diverse species complex of natural enemies. CASE STUDY 1: MAIZE INTERCROPS AND PEST ATTACK
Corn-bean polycultures are common diversified systems used by small sma ll fa farme rmers rs in Lat Latin in Am Ameri erica. ca. The These se int interc ercro roppi pping ng sys system temss usu usually ally result in higher yields for several reasons, including reduced weed
compet comp etit itio ion n du duee to de dens nsee cr crop op co cove verr, so soil il co cons nser erv vat atio ion, n, an and d be bett tter er us usee of incident radiation, water, and soil nutrients. Another advantage is that insect-pest attack is often less than that on sole crops (van Huis, 1981). In order to test the built-in pest-control pest-control features of those systems, several experiments were conducted in tropical Colombia (Altieri and Dol Doll, l, 197 1978). 8). Sim Simult ultane aneous ously ly pla plante nted d cor cornn-bea bean n pol polycu ycultu ltures res sho showed wed significantly fewer adult leafhoppers (Empoasca kraemeri) on beans in the maize-bean polyculture compared to monoculture beans, until the final sampling date seven seventy ty days after planting. Nymphal populations were not affected by diversity in cropping systems. Anagrus sp. (Hymenoptera: (Hymenopter a: Mymaridae), the main egg parasitoid of E. kraemeri, showed 20 percent higher activity in polycultures with 48.5 percent parasitism in monoculture versus 60.7 percent parasitism in the crop association. The occurrence of natural predators was significantly higher in polyculture at forty days from planting. Principal predators were Condylostylus sp. (Diptera: Dolichopodidae) and some Hemiptera (Reduviidae and Nabidae). These insects showed higher densities ti es in po poly lycu cultu lture ress th than an in mo mono nocu cultu lture res, s, but fi fift fty y da days ys af afte terr pl plan anti ting ng they showed an opposite pattern suggesting a migratory trend toward monocultures where prey concentrated. Diabrotica balteata, a polyphago ph agous us lea leaff bee beetle tle,, sho showed wed 45 per percen centt lo lower wer adu adult lt pop popula ulatio tion n den densisitiess in pol tie polycu ycultu ltures res.. The inc increa reased sed po popul pulatio ation n of the their ir red reduv uviid iid pre predadators in polycultures probably was an important regulatory factor. factor. It is possible that the presence of other chrysomelids, which was 30 percent higher in polyculture, exerted a competitive displacement of D. balteata, decreasing its feeding and colonization efficiency. The number of bean plants with damaged leaves was similar in polycultures and monocultures. In general, the percentage of corn seedlings damaged by cutworm atta at tack ck wa wass lo low w in al alll tr trea eatm tmen entt pl plot ots; s; ho howe wev ver er,, it ten tende ded d to be lo lowe werr in plots with high vegetational diversity. Corn-bean polycultures had the lo lowes westt nu numbe mberr of dam damage aged d see seedli dlings ngs.. Twen wenty ty day dayss aft after er pla planti nting, ng, fall armyworm (Spodoptera frugiperda) (F (FA AW) larval densitie densitiess were significantly lower in polycultures than in monocultures at all sampling dates. Parasitism of FAW larvae by the braconid Meteorus sp. forty days after planting was higher in the polycultures than in the monocultures. Interc Int ercro roppi pping ng cor corn n wit with h dif differ ferent ent bea bean n va varie rieties ties had di dive verse rse ef effec fects ts upon FAW FAW. Corn interplanted with the bush bean variety ICA Pijao
presented 14 percent less whorl damage than corn interplanted with the climber bean variety P-589. Systems with sequential planting of corn in relation to beans resulted in reduced population densities of adult E. kraemeri when maize was planted twenty to forty days before the beans. Significantly lower infestation levels of FAW in corn were observed with beans planted twenty to thirty days before, and higher and uniform populations in treatments where beans were planted ten days before to twenty days after corn. The FAW populations in corn were reduced by 88 percent by the early plantings of beans. Corn yields did not vary significantly among these planting dates. In sur surve veys ys of mai maizeze-bea bean n int interc ercrop ropss in Nic Nicara aragu gua, a, va van n Hu Huis is (19 (1981) 81) found trends of FAW similar to those found in Colombia. In addition to lower incidence of FAW, he noted population reductions of the stalk borer Diatraea lineolata. In Mexico, Trujillo-Arriaga and Altieri (1990) found that maize associated with fava bean and squash exhibited lower damage by Tetranychus urticae than monocultures, apparently because maize in monocultures was more affected by water stress, a condition that makes these plants more susceptible to mite attack att ack.. In po polyc lycult ultur ures, es, the aph aphid id Rhopa Rhopalosiphu losiphum m maidis experienced higher attack by several species of lady beetles, Hippodamia convergens and H. koebelei. A corn-diversification corn-diversification strategy amenable to U.S. midwestern farmers is strip cropping of corn and soybeans. Tonhasca (1993) reported a gen genera erall po posit sitiv ivee res respon ponse se to str stripip-cro croppi pping ng sys system temss by se seve veral ral pre preddator species (including lady beetles, spiders, minute pirate bugs, and nabids nab ids)) and by par parasi asitic tic wa wasps sps.. Cor Corn n in str stripip-cro cropp pping ing plo plots ts pro provid vided ed shade, sha de, red reduce uced d win wind d spe speed, ed, hig higher her hum humidi idity ty,, lo lower wer tem temper peratu atures res,, and altern alt ernate ate fo food, od, wh which ich fa favo vored red all fac factor torss for so soyb ybean ean nat natura urall ene enemie mies. s.
CASE STUDY 2: CASSAVA INTERCROPS AND PEST INCIDENCE
Traditional farmers farmers in Latin America and Africa commonly intercrop cassava, partly as a defense against herbivore herbivore attack. For example, in Nigeria, farmers intercropping cassava with maize and sorghum reported lower populations of Zonocerus variegatus than in monocultures. In Colombia, cassava intercropped with beans exhib-
its reduced populations of various pests such as the hornworm, shoot Aleurotrachelus achelus socialis socia lis fly,, and lacebug. Cassava fly Cassava whitefly species ( Aleurotr and Trialeurodes variabilis) exhibited lower per-leaf densities in cassava-cowp cassav a-cowpea ea intercrops than in i n monoculture. Intercropping with maize had no significant effect on whitefly populations per leaf (Gold, Altieri, and Bellotti, 1989a). Rates of parasitism and overall immature mortality of cassava whiteflies were similar among monoand polyculture. In or orde derr to de dete term rmin inee yi yiel eld d lo loss sses es un unde derr th thee di difffe fere rent nt cr crop oppi ping ng sy sysstemss and to sep tem separa arate te the ef effec fects ts of int interc ercrop rop com compet petitio ition n and dif differ ferenential herbivore numbers, applications of the insecticide monocrotophos ph os wer weree mad madee to pro protect tected ed plo plots. ts. Int Interc ercrop roppin ping g wit with h co cowp wpea ea lo lower wered ed yields of a regional cultivar cultivar of cassava (‘MCOL 2257’) in protected plots. However, in nonprotected plots, the regional cassava variety intercropped with cowpea had higher yields and sustained lower yield losses than other systems. Yields of regional cassava intercropped with maize, grown in monoculture, or mixed with cassava cultivar CMC 40 were equivalent in both protected and nonprotected envir en viron onmen ments. ts. Yiel ield d los losses ses clo closel sely y fol follo lowed wed po popu pulati lation on tre trends nds of cas cas-sava whiteflies. Whiteflies were attracted to more vigorous plant assemblages, as in monocultures, with lowest numbers in cassavacowpea systems. However However,, the data indicate that under stress, cassava favors top growth over roots, and large plant size did not ensure high yield. Land equivalent ratios exceeded 1.5 for intercropped systems (Gold, Altieri, and Bellotti, 1989b). One of the intriguing aspects of the cassava-cowpea cassava-cowpea association is that reductions in whitefly densities persist even after the harvest of cowp co wpea ea (a (att th thre reee mo mont nths hs,, wh when en ca cass ssav avaa st stay ayss in th thee fi fiel eld d an ad addi ditio tiona nall six to nine months). Both species of whitefly ( A. socialis and T. variabilis) had lower egg densities on cassava-cowpea mixtures than on cassava monoculture, with lower levels remaining for six months. These residual effects apparently resulted from two possible sources. The existent populations of herbivores and their natural enemies at thee ti th time me of in inte terc rcro rop p ha harv rves estt ma may y ha have ve in infl flue uenc nced ed pe pest st po popu pula latio tion n dy dy-namics during the postintercrop period. The relative mobility of t he insects and the determination of immigration and emigration rates may ma y in indi dica cate te ho how w lo long ng su such ch a re resi sidu dual al af affe fect ct ma may y pe pers rsis ist. t. In ad addi ditio tion, n, differences differen ces in plant quality may extend far into the postintercrop pe-
riod. Apparently, intercrop competition retarded cassava growth, causing reductions in host-plant size well beyond the intercrop period. Greater whitefly numbers were associated with more vigorous plants and, hence, were higher in monoculture (Gold, Altieri, and Bellotti, 1990).
CASE STUDY 3: REDUCING STEMBORERS IN AFRICA
Many intercropping studies have transcended the research phase and have found applicability to control specific pests such as the stemborers in Africa. Scientists at ICIPE developed a habitat management system that uses two kinds of crops planted together with maize: a plant that repels these borers (the push) and another that attracts (pulls) its natural enemies (Khan et al., 2000). The push-pull system has been tested on over 450 farms in two districts of Kenya and has now been released for uptake by the national extension systems in East Africa. Participating farmers in the breadbasket of Trans Nzoia have reported reported a 15 to 20 percent increase in maize yield. In the semiarid Suba district, plagued by both stemborers and striga, a substantial increase in milk yield has occurred in the past years, with farmers now being able to support cows on the fodder produced. When farmers plant maize, napier grass, and desmodium together, together, a return of US$2.30 for every dollar invested is made, as compared to $1.40 obtained by planting maize as a monocrop. Two of the most useful trap crops that pull in the borers’ natural enemies are napier grass (P (Pennisetum ennisetum purpureum) and Sudan grass (Sorghum bicolor both th im impo port rtan antt fo fodd dder er pl plan ants ts;; th thes esee ar aree pl plan ante ted d in a bo bord rder er Sudanese), bo around the maize. Two excellent borer-repelling borer-repelling crops that are planted between the rows of maize are molasses grass (Mellinis minutifolia), which also repels ticks, and the leguminous silverleaf maize (Desmodium), which can suppress the parasitic weed Striga by a factor of forty fo rty com compar pared ed to mai maize ze mon monocr ocrop op;; its NN-fi fixin xing g abi ability lity inc increa reases ses soi soill fertility, and it is an excellent forage. As an added bonus, sale of desmodium seed is proving to be a new income-generating opportunity for women in the project areas.
LIVING MULCHES: A SPECIAL TYPE OF INTERCROP
The use of legume cover crops as sod-strip intercropping and/or living mulches in year-round cropping systems and rotations has been bee n pro propo posed sed as ho holdi lding ng pot potent ential ial for sus sustai tained ned cro crop p pro produc ductio tion n and self-sufficiency self-suff iciency in soil nutrients (Vrabel, Minnotti, and Sweet, 1980; Palada et al., 1983). Researchers at Cornell University and at the Rodale Rod ale Res Resear earch ch Cen Center ter ha have ve fou found nd tha thatt ov overs erseed eeding ing le legum gumes es in cor corn n and other annual crops maintains yields while providing increased soil protection on highly erosive soils. Moreover, when the legume sod is pro proper perly ly man manage aged, d, wee weed d sup suppre pressi ssion on is enh enhanc anced ed sig signif nifica icantl ntly y, reducing the need for chemical herbicides. Although the entomological advantages of these systems are still poorly understood, experimental work suggests that many livingmulch systems have built-in biological control advantages. Most research has focused on Brassica crops. For example, Dempster and Coaker (1974) found that the maintenance of a clover cover cover aided in the reduction of three insect pests ( Brevicoryne brassicae, Pieris rapae, and Erioischia brassicae). In the case of P. rapae, the reduction was attributable to increased numbers of the predacious ground beetle Harpalus rufipes in the clover clover-seeded -seeded plots. Similar enhancements were observed when planting clover between rows of cabbages, which resulted in a 34 percent increased predation of eggs of the cabbage root fly, Delia brassicae (Cromartie, 1981). Hooks, Valenzuela, and Defrank (1998) report decreased pest incidence due to enhanced natural enemy activity in zucchini grown with living mulches. In New York State, an experiment was conducted using cabbage interplanted with several living mulches and in bare-ground monocultur cul tures es (An (Ando dow w et al. al.,, 19 1986) 86).. Li Livin ving g mul mulche chess inc includ luded ed cre creepi eping ng ben bentt grass, red fescue, Kentucky bluegrass, and two white clovers. Populations of Phyllotreta cruciferae and Brevicoryne brassicae were lower on cabbage grown with any living mulch than on cabbage in bare-ground monocultures. First-generation larvae of Pieris rapae were more common on cabbage with clover living mulches, but second-generation eggs and larvae were less common on cabbage with clover living mulches. These differences in population density were probably determined by variation in herbivore colonization
rates, not by variation in herbivore herbivore mortality. The authors suggested that th at ea earl rlyy- se seas ason on ch chem emic ical al tr trea eatm tmen ents ts fo forr fl flea ea be beet etles les mi migh ghtt be el elim imiinated when living mulches are used. However, this potential gain may be offset by yield reduction from competition between cabbage and living mulches. Helenius (1998) reports a study of cabbage with or without clover in the interrow spaces in which cabbage root fly eggs were consistently reduced by 25 to 64 percent in the intersown plots. Laboratory experiments suggested that intersowing had not deterred oviposition. Similarly, pitfall traps did not provide evidence of higher numbers of epigeal predators. However, However, predator-exclusion experiments revealed that th at ov ovip ipos osit itio ion n wa wass in inde deed ed re redu duce ced d (b (by y 18 pe perc rcen ent) t) in ca cabb bbag ages es wi with th clover and that predators further reduced egg numbers to 41 percent of control values. In tw two o loc locati ations ons in Cal Califo iforni rnia, a, Alt Altier ieri, i, Wils ilson, on, and Sch Schmid midtt (1 (1985 985)) further tested the effects of vegetation background in the form of living mulches and natural weed cover on the population dynamics of foliage and soil arthropods in corn, tomato, and cauliflower crop systems. In Davis, California (Central Valley site), herbivores (especially aphids and lygaeids) were more abundant in the plots with weed cover than in the clover clover-mulch -mulch plots, whereas leafhoppers were most common in the clover mulch. Higher numbers of natural enemies were observed in the clover plots. Significantly more ground predators (Carabidae, Staphylinidae, spiders) were caught in pitfalls placed in the weedy and clover plots than were caught in the cleancultivated cultiv ated plots. In Albany, Albany, California (coastal area), specialized herbivo bi vore re (ca (cabba bbage ge aph aphid id and fle fleaa bee beetle) tle) den densit sities ies wer weree sig signif nifican icantly tly reduced in plots with living-mulch cover cover.. It is i s not clear if this reduction wass du wa duee to pl plan antt di dive vers rsity ity or de dens nsit ity y ef effe fect cts, s, th thee ef effe fect ctss of na natu tura rall en eneemies, or the lower quality of plants in the weedy and mulched plots, as crop growth and yields were drastically reduced in these plots at both sites. In Salinas, California, Costello and Altieri (1994) designed an experiment to test whether a living mulch of white clover clover,, strawberry clover, and bird’s-foot trefoil, could be useful in place of insecticides to protect broccoli from cabbage aphid infestation. The stud st udy y sh sho owe wed d th that at li livi ving ng mu mulch lches es ca can n re redu duce ce ca cabb bbag agee ap aphi hid d de dens nsiti ities es in har harve veste sted d br brocc occoli oli hea heads ds rel relati ative ve to bro brocco ccoli li gro grown wn wit witho hout ut a co cove verr crop. Differences in infestation levels between living mulches and
clean cultivation are best explained by differences in aphid colonization rates. The correlation between light intensity and numbers of alates suggests that the lower-intensity light reflected from broccoli grown grow n with living mulches is less attractive to incoming aphids than the higher-intensity light reflected from clean-cultivated broccoli. Thee st Th stro rong ng as asso socia ciatio tion n be betw twee een n nu numb mber erss of al alat ates es an and d in inte tens nsit ity y in th thee yellow wavelength is interesting given that the cabbage aphid is kno kn own to be at attr trac acte ted d to th thee co colo lorr ye yello llow w (C (Cos oste tell llo o an and d Al Alti tier eri, i, 19 1995 95). ). The authors suggested that further agronomic work is warranted to minimize the competitive effects of legume covers on crops, so that the obs observ erved ed ent entomo omolog logica icall adv advant antage agess can be use used d in a pra practi ctical cal wa way y. Planting lucerne (Medicago littoralis) in the same row as carrot significantly reduced the damage levels of the carrot rust fly. Although the abundance of generalist predators was higher in the lucernecarrot plots than in monocultures, this enhancement did not explain the observed decreased carrot fly rust damage. Emigration of carrot rust flies from the intercropping system was higher than from monocultures, thus flies exhibited longer tenure times in single habitats (Cromartie, 1981). In England, undersowing undersowing cereals with grass species (i.e., ryegrass) increases the activity and abundance of natural enemies. Potatoes undersown with perennial ryegrass have been found to have aphid popu po pula lati tion onss re redu duce ced d by up to 66 pe perc rcen entt co comp mpar ared ed wi with th pl plan ants ts in ba bare re ground. Since the number of colonizing aphids did not appear to t o be affected by undersowing, increased mortality from natural enemies was suspected. The practice of undersowing appears to be one of the mostt ef mos effec fecti tive ve mea means ns of enh enhanc ancing ing aph aphid id par parasi asitis tism m by Aphidius spp. in cereals (Burn, 1987). A similar effect was shown in Germany, Metopolophium ophium dirhodum by two parasitoids where parasitism of Metopol was higher in wheat underso undersown wn with clover than in wheat monoculturee (El Titi, 198 tur 1986). 6). In an ex exper perime iment nt in Eng Englan land, d, rye ryegra grass ss un under derso sown wn in wheat was purposefully infested with Myzus festucae followed by a release of the parasitoid Aphidius rhopalosiphi. Thus, a parasitoid popul po pulati ation on wa wass est establ ablish ished ed on the rye ryegra grass ss bef befor oree Sitobion avenae invaded the wheat in the spring (Powell, 1986). Populations of the pest aphi ap hid d on wh whea eatt we were re sm smal alles lestt on th thos osee pl plot otss wh whic ich h ha had d de deve velo lope ped d th thee largest M. festucae populations in the spring.
METHODOLOGIES TO STUDY INSECT DYNAMICS IN POLYCULTURES
Many methodological approaches have been tested in studies of monocu mon ocultu ltures res and po polyc lycult ultur ures es to ex expla plain in the eco ecolog logical ical mec mechan hanism ismss underlying the entomological effects of diversity. Risch (1981) studied beetle-movement behavior behavior to see if this could account for lower numbers of beetles in maize-bean intercrops. He placed directional mala ma lais isee in inse sect ct tr trap apss on ea each ch si side de of eve very ry pl plot ot.. Wh When en be beet etle less fl flew ew ou outt of the plot, some of them landed on the vertical trap walls and were caught in the collecting jars. By counting these trapped beetles and estimating the total number of beetles in the t he plot at that time t ime from direct re ct co coun unts ts,, he ca calc lcul ulat ated ed a ra rati tio o of th thee tw two o gr grou oups ps an and d ca call lled ed th thee ra rati tio o “tendency to emigrate,” which measures the beetle’s relative tendenc de ncy y to le leav avee a pl plot ot on once ce it ha hass ar arri rive ved. d. Af Afte terr si sixt xty y to si sixt xtyy-fi five ve da days ys,, there was a much greater tendency to emigrate from the bean monocult cu ltur uree th than an fr from om th thee be bean an po poly lycu cultu lture re.. Th This is co corr rres espo pond ndss wi with th th thee ob ob-servation that there were far fewer beetles on beans planted with maize in the polycultures than in the bean monocultures and that this large difference became apparent approximately sixty-five sixty-five days after planting. Maize has an inhibitory effect on the presence of this i nsect species. How does maize exert its inhibitory effect? Beans grown with maiz ma izee ar aree sh shad aded ed mo more re th than an be bean anss in mo mono nocu cult ltur ures es.. On Onee po poss ssib ibil ility ity is that the beetles avoid feeding in shaded areas, preferring to feed on plants that are not shaded. This was tested directly by constructing two large shade screens and suspending them 80 cm above the ground. One screen provided little shade, allowing 65 percent light transmission, and the other provided much more shade, allowing only 25 percent light transmission. Squash and bean plants were grown in the greenhouse and placed under these screens. Then the numbers of colonizing beetles on the plants were counted over a series of days. The results showed that there were always significantly more beetles under the light-shade screen than the dark. Yet shade might not be the only way that the presence of maize interferes with beetle-flight behavior. To determine if a vertical obstruction, such as a maize stalk, could discourage beetle colonization in other ways, dry maize stalks were staked among potted bean plants, and a light screen was erected over the plants. Potted beans
without maize stalks were also placed in a nearby area with a darker screen over them, so that the total amount of light reaching the plants in both areas was identical. Consistently, many more beetles were found in the beans without maize stalks, indicating that maize physically inhibited beetle colonization in ways other than by just increasing the overall shade of a microhabitat. Although these experiments provided an indication of the underlying causes of the reduction in beetle numbers in maize-bean polycultures, they did not help in predicting numbers of beetles in differ differ-ent variations of the entire maize-bean-squash system. Risch (1980) also studied the influence of size of the plot and relative proportions of maize, beans, and squash on the number of beetles in the field. He observed and modeled the movement of one beetle, Acalymma vittata, a squash specialist that is much more abundant in monocultures of squash than in maize-bean-squash polycultures. polycultures. The variables Risch thought might be important in ultimately determining the rate at which wh ich a bee beetle tle lea leave vess a mai maizeze-bea bean-s n-squa quash sh pol polycu ycultu lture re ve versu rsuss a squ squash ash monoculture are the following: following: the time a beetle spends on a maize, bean, or squash plant; the probability of moving to a maize, bean, or squash plant; the distance a beetle travels when it leaves a maize, bean, or squash plant and flies over an intercrop or monoculture; and its orientation behavior at the edge of a plot. Kareiva (1983) proposed that emigration models can be mathematically formalized as continuous-time, finite-state Markov processes in which insects move among three states (on host plant, on nonhost plant, outside plot) in polycultures but only between two states (on host plant, outside plot) in monocultures. Because computatio ta tions ns an and d pa para rame mete terr es estim timati ation on ar aree si simp mple le fo forr th thes esee mo mode dels ls,, th this is ap ap-proach is ideal for exploring the connection between herbivory and trivial movement. movement. Instantaneous transition rates between these states can be easily obtained by releasing and recapturing marked insects. Equilibrium herbivore herbivore densities for polycultures versus monocultures can ca n th then en be ca calc lcul ulat ated ed fr from om th thee mo mode del; l; wi with th th this is ap appr proa oach ch,, it mi migh ghtt be possible to explicitly attribute reduced pest pressure in polycultures to high rates of movement, either from host to nonhost plants or from nonhost plants to areas outside the crop. One of the suggestions of Kareiva (1983) is that nonrandom pest movement is the process often responsible for the different damage levels observed in randomized-block field experiments. Individual
insects of these mobile species will move frequently frequently among the variouss cro ou croppi pping ng tre treatm atment entss inc includ luded ed in a ran rando domiz mized ed blo block ck des design ign;; typ typiically they will spend disproportionate amounts of time in the treatments that represent a preferred food or preferred habitat. To evaluate the dependence of apparent crop protection on local pest movement, Kareiva established one set of treatment plots in close proximity to one another with free access between them, and a seco se cond nd se sett of tr trea eatm tmen entt pl plot otss (a (als lso o in clo close se pr prox oxim imity ity)) se sepa para rate ted d fr from om one another by tall (1.5 m) curtain barriers, which interfere greatly with wi th fl flea ea be beet etle le mo move veme ment nt (F (Fig igur uree 5. 5.4) 4).. Th Thee cr crop oppi ping ng tr trea eatm tmen ents ts th that at he contrasted were (1) pure stands of collards versus collards inter-
FIGURE 5.4. Effects of between-treatment movements on the response of flea beetles to cropping systems (polyculture [P] versus monoculture [M] {a}, or low density [L] versus high density [H] {b}). Movement between the the top two subplots in each block was restricted by a curtain barrier. These cultivated blocks were surrounded surrou nded by old-fi old-field eld vege vegetation tation constituted constituted mainly of goldenr goldenrod od (Solidago spp.) (after Kareiva, 1986).
cropped with potatoes and (2) collards planted at a high density of 6.7/m2 versus collards planted at a lower density of 3.3/m 2. Where there were no barriers between intercropped i ntercropped and monoculture treatments, flea beetles were less abundant in the intercropped plots than in adjacent monocultures; where there were barriers between the treatments, intercropping yielded no reduction in beetle infestation (Kareiva, 1986). By marking beetles in California, Garcia and Altieri (1992) examined in ed th thee mo move veme ment nt be beha havi vior or an and d th thee ra rate te at wh whic ich h ma mark rked ed fl flea ea be beet etles les released in broccoli monocultures and broccoli- Vicia polycultures tended to either leave or stay in the system, or even migrate from one system to the other. After vacuuming of all naturally occurring flea beetles, three groups of 350 flea beetles marked with fluorescent blue, orange, or pink were released in each plot. The number of marked beetles remaining in each plot was estimated visually by inspecting all plants and the surrounding soil at six and twenty-four hours after release. More beetles tended to fly out and leave mixed cultures compared to monocultures. Apparently, the broccoli- Vicia system had a deterrent effect that resulted in massive emigration of the released beetles. Figure 5.5 depicts the t he movement patterns of flea beet be etle less ou outt of th thee pl plot ot,, be betw twee een n pl plot ots, s, an and d mo movi ving ng in into to pl plot otss fr from om su surrrounding habitats during a twenty-four-hour period after release. Bach (1980a) focused on the response of one specialist herbivore, the striped cucumber beetle (Acalymma vittata), to cucumber monocultures versus cucumber-broccoli-maize polycultures. By controlling total plant density, host-plant density, density, and plant diversity, diversity, Bach was able to distinguish the effects of these three confounding variables. Applying a three-way analysis of variance to censuses of beetles per cucumber plant, Bach reported a significant effect of both plant density and diversity on Acalymma abundan abundance, ce, but the results only partially support the resource concentration hypothesis. Although an increase in stand purity yielded the expected increase in beetles per cucumber plant, an increase in cucumber density reduced the number of beetles per plant (Bach, 1980a). Bach also found that cucumber plants were, on average, smaller in polycultures than in monocultures and that beetle density was positively correlated with plant size. Two polyculture plots with cucumbers were equal in size to monoculture cucumbers; in these two thriving polycultures, beetles were still significantly fewer per plant than in monocultures.
Broccoli-Vicia faba
Broccoli
Broccoli-Vicia sativa
FIGURE FIGU RE 5. 5.5.Flux 5.Flux of ma mark rked ed fl flea ea bee beetl tles es in th thre ree e dif diffe fere rent nt cr crop oppin ping g sy syst stem ems s dur dur-ing a twenty-four-hour period after release. A = beetles emigrating out of crop habitat; B = beetles moving moving from dicultures to monoculture; C = beetles colonizing plots from surrounding habitats; D = beetles staying in the plots (after Garcia and Altieri, 1992).
Thus, it is clear that the reduced beetle numbers in polycultures cannot simply be attributed to smaller cucumber plants. In a later study with Acalymma vittata, Bach (1980b) provided evidence for a surprising reduction in foliage palatability associated with cucumbertomato polycultures. Beetles given a laboratory choice between cucumber leaves grown grown in monoculture and cucumber leaves grown in a tomato-cucumber tomatocucumber mixture significantly preferred monoculture leaves. This illustrates the subtle links that are possible between plant diver-
sity and plant quality, quite apart from the conventional ideas concerning the influence of resource concentration on herbivores (Kareiva, 1983). Some studies have focused on the behavior of natural enemies in polycultures. Wetzler and Risch (1984) examined the behavior of a coccinellid beetle in the field in four diffusion experiments, experiments, Each involved volv ed the release of beetle populations in a matched pair of agricultural plots (10 m 10 m each) each) planted planted with various combinations of maize, beans, and squash. The day preceding each release, all Coleo Coleo-individuals ls were aspirated from every plant, ensurmegilla maculata individua ing “clean” fields for each experiment. One of the experiments focused on determining whether differences enc es in dif diffus fusion ion rat rates es fro from m mon monocu ocultu ltures res and po polyc lycult ulture uress mig might ht be caused in part by differences in the average time a beetle spent on maize, bean, and squash plants (i.e., tenure time per plant). Maize, squash, and bean plants were first grown in pots until all the plants were in flower. Approximately half of the maize plants had large numbers of the corn aphid Rhopalosiphum maidis. In the first trial, fifty Coleomegilla were placed on five aphid-infested maize plants, fifty beetles were placed on five bean plants, and fifty beetles were placed on five squash plants (ten beetles per plant). The beetles were cooled to approximately 6°C before being placed on plants. The number of beetles remaining on the plants was counted approximately every ten minutes for a period of 100 minutes. Sight counting proved to be an effective means of population census cen susing ing sin since ce the bee beetle tless are hig highly hly vis visibl ible, e, thu thuss av avoid oiding ing pr probl oblems ems associ ass ociate ated d wit with h tra trappi pping. ng. Car Carefu efull col collec lectio tion n of ind indiv ividu iduals als for rel releas easee enabled uniform, almost equivalent equivalent releases of adults. Since each experiment was run for only twenty-four hours and was preceded by a minimum of handling of beetles, mortality was extremely low (0.5 percent) and complications due to beetle reproduction were nonexisnonexistent. The timing of the one-hour, three-hour, and six-hour censuses was arranged to correspond with maximum periods of diurnal activity to ensure that the most conservative diffusion estimates would arise ari se du durin ring g the fi final nal cen census suses. es. Sin Since ce all ex exper perime iments nts wer weree con condu ducted cted within a fiv five-week e-week interval, seasonal variability (i.e., migratory movements) of Coleomegilla activity was restricted.
In their studies of corn-cowpea-squash polycultures, Letourneau and Altieri (1983) found visual-inspection sampling of thrips and Orius to produce a more representative measure of density than did sticky traps, pan traps, or malaise traps, each of which showed very few catches. Ten hills of squash (each hill consisting of two plants) were randomly selected, and the plant most southwest in the hill was sampled by gently turning each leaf and recording the numbers of Orius adults and nymphs (as well as any other common arthropods). Thrips were counted on one medium-sized leaf of each plant. During this season, Orius densities increased, and plants grew so large that thee nu th numb mber er of le leav aves es sa samp mple led d wa wass re redu duce ced d to fi fiv ve pe perr pl plan ant: t: th thee gr gro owing shoot, two young, and two old leaves. Biomass estimates were made ma de at tw twoo-we week ek in inter terv val alss by me meas asur urin ing g lea leaff wi widt dths hs on al alll th thee le leav aves es of ten plants per plot. The leaf width of squash plants was highly correlated with leaf biomass, determined as dry weight of the leaf blade (r = 0.93). To standardize for possible leaf size differences between treatments (and thus searching-area differences), predator numbers per plant were converted to numbers per 5 g of leaf biomass. Individual lea leave vess sampled sampled for thr thrips ips were were als also o mea measur sured ed to allo allow w for con conve verrsion of thrips per leaf to thrips per 5 g of leaf biomass. To determine whether predators were concentrated within treatments on plants with higher thrip densities, an index of aggregation was ca calc lcul ulat ated ed on da day y 30 30.. Ra Rati tios os of me mean an th thri rip p de dens nsit ity y on pl plan ants ts wi with th Orius to those without Orius present would be significantly greater than if Orius were showing such a preference within a treatment. Coll (1998) has warned about the limitations of using trapping or recording host-parasitism rates in simple and div diverse erse habitats. These methods are appropriate only when habitat type does not affect the precision of the sampling. However, sampling may not be equally effecti fec tive ve in dif differ ferent ent hab habitat itatss bec becaus ausee dif differ ferenc ences es in pla plant nt hei height ght and architecture may alter sampling effort and/or effectiveness. effectiveness. For example, Perfect (1991) used traps to compare parasitoid density when cowpea and maize were grown alone or intercropped. They found that when traps were placed 0.5 m above the ground, almost twice as many man y cha chalci lcidoi doids ds wer weree tra trappe pped d in the int interc ercrop rop tha than n in mon monocu ocultu ltures res.. However,, when the traps were situated However sit uated 2 m above the ground, similar simila r numbers of chalcidoids were trapped in the two monocultures and in the int interc ercrop rop.. Sim Similar ilarly ly,, mai maize ze hei heigh ghtt af affec fected ted the nu numbe mberr of tach tachini inid d flies caught in malaise traps. Thus, in both cases the traps are in-
appropriate for comparing parasitoid abundance in monocultural and intercropped habitats. Colored traps may be more apparent in one habitat, resulting in greater attractiveness and capture of parasitoids. Estimating parasitism rate may also depend on habitat type if, for example, host spatial distribution (i.e., clumpedness) differs among habitat hab itatss bu butt the sam samee sam sampli pling ng pro protoc tocol ol is use used d in dif differ ferent ent hab habita itats. ts. Given these limitations, Coll and Botrell (1996) used releaserecapture experiments to determine how the presence of maize (nonhost plant) influenced the movement of the parasitoid (Pediobius foveolatus) in bean plots in the absence of hosts. They tested three predictions that are based on the response of monophagous herbivores to plant diversification. In the absence of hosts they tested whether (1) the parasitoid immigrates more readily to taxonomically simple than diverse habitats, (2) it is more likely to remain in simple than in diverse plant stands, and (3) its movement in the habitat is hampered by the presence of tall nonhost plants. Then they assessed how vegetation diversity diversity affects wasp reproduction (parasitism) and subsequent density in the presence of its hosts, Mexican bean beetle larvae. Fewer female wasps immigrated into and more emigrated out of a bean and tall maize intercrop intercrop than bean mono monocultu cultures. res. Bean plan plantt density and the presence of maize per se did not significantly affect parasitoid immigration. Instead, maize height was the primary factor contributing contributin g to lower female immigration into the intercropped bean and tall maize plots. However, tall maize plants did not impede the wasps’ within-habitat movem movement. ent. When Whe n wa wasps sps wer weree rel releas eased ed out outsid sidee the plo plots, ts, hig higher her par parasi asitism tism wa wass recorded in monocultures, irrespective of host density. In contrast, when wasps were released within the plots, significantly higher parasitism rates were found in the bean and tall maize habitat. Results suggested that female wasps accumulate in the bean and tall maize habitat hab itat in res respo ponse nse to res resou ource rcess oth other er tha than n ho hosts sts and and,, ult ultima imatel tely y, wa wasp sp density may be determined primarily by differential emigration rather than by immigration rates. These The se res resear earch ch ex examp amples les ill illust ustrat ratee the ran range ge of met metho hodol dologi ogies es tha thatt have been employed to explain insect movement and differences in densities in simple and complex habitats. These studies have been crucial in advancin advancing g our understanding of how plant diversity influences insect ecology in polycultures.
MANAGEMENT CONSIDERATIONS
Multiple-crop management is basically the design of spatial and temporal combinations of crops in an area (Harwood, 1979). There are many possible crop combinations and arrangements, and each can have different effects on insect populations. The attractiveness of crop cr op ha habi bita tats ts to in inse sect ctss in te term rmss of si size ze of fi fiel eld, d, na natu ture re of su surr rrou ound ndin ing g vegetation, plant densities, height, background color and texture, crop diversity, diversity, and weediness is subject to manipulation. An important goal of intercropping i ntercropping research should be to fully understand the mechanisms involved in pest reduction in polycultures, so that informed manipulations can be done to improve the entomological advantages of intercropping systems. In intercrop systems, the choice of a tall or short, early or late maturing, flowering or nonflowering companion crop can magnify or decrease the effects on particular pests (Altieri and Letourneau, 1982). The inclusion of a crop that bears flowers during most of the growing season can condition the buildup of parasitoids, thus improving biological control. Similarly, the inclusion of legumes or other plants supporting populations of aphids and other soft-bodied insects that serve as alternate prey/hosts can improve survival survival and reproduction of beneficial insects in agroecosystems. Early planting of an aphid-supporting legume in patches within the field can initiate the buildup of parasitoids before the rest of the crop is planted and more aphids colonize the field. The presence of a tall associated crop such as maize or sorghum may serve as a physical barrier or trap to pests invading from outside the field. Tal alll pl plan ants ts ca can n af affe fect ct th thee vi visu sual al st stim imul ulii by wh whic ich h in inse sect ct pe pest stss or orie ient nt themselves to their suitable host plants or may interfere with the herbivore’s movement and dispersal within the system (Perrin, 1977). The inclusion of strongly aromatic plants such as onion (Allium cepa), garlic (Allium sativum), or tomato (Lycopersicon esculentum) can disturb mechanisms of orientation to host plants by several pests. Including onions as intercrops in carrot fields in England reduced the attack of carrot fly (Psila rosae) and carrot willow aphid (Cavanella aegopodii) (Uvah and Coaker, 1984). The date of planting of component crops in relation to one another can also affect insect interactions in these systems. An associated crop can be planted so that it is at its most attractive growth stage at
the time of pest immigration or dispersal, diverting diverting pests from other more susceptible or valuable crops in the mixture. Planting of okra (Hibiscus esculentus) to divert flea beetles (Podagria spp.) from cotton in Nigeria is a good example (Perrin, 1980). Maize planted thirty and twenty days earlier than beans reduced leafhopper population on beans by 66 percent compared to populations in plots under simultaneous planting. Fall armyworm damage on maize was reduced by 88 percent when beans were planted twenty to forty days earlier than maize when compared to simultaneous planting (Altieri, Schoonhoven, and Doll, 1977). We stil stilll un under dersta stand nd litt little le ho how w spa spatia tiall arr arrang angeme ements nts (e. (e.g., g., ro row w spa spaccings) of crops affect pest abundanc abundancee in intercrops. For example, there is greater reduction in damage to cowpea flowers by Maruca testulalis in intrarow rather than interrow mixtures of maize and cowpea. Selection of proper crop varieties can also magnify insect suppression effects. In Colombia, lower whorl damage by Spodoptera frugiwass ob obse serv rved ed in ma maiz izee as asso socia ciate ted d wi with th bus ush h be bean anss th than an in ma maize ize perda wa mixed with climbing beans. In the same trials, maize hybrid H-207 seemed see med to ex exhib hibit it lo lower wer Spodoptera dam damage age tha than n hy hybr brid id H-2 H-210 10 wh when en intercropped with beans (Altieri et al., 1978). Strip-cropping systems can preferentially act as trap crops or as sources of natural enemies that move from one strip to another as in the case of alfalfa strips within cotton fields (Stern, 1979; Robinson, Young, and Morrison, 1972). In intercropping systems where crops are more closely intermingled, other mechanisms (i.e., repellency repellency,, masking, natural enemy enhance enha ncemen ment, t, phy physic sical al bar barrie riers) rs) may af affec fectt ins insect ect pes pests. ts. Cle Clearl arly y, muc much h more work is needed to determine effective row spacings within crop mixtures that enhance pest suppression. The manipulation of weed abundance and composition in intercrops can also have major implications on insect dynamics (Altieri, Schoonhoven, and Doll, 1977). When weed and crop species grow together, each plant species hosts an assemblage of herbivores and their natural enemies; thus, trophic interactions become very complex. As discussed in Chapter 4, many weeds offer important requisites for natural enemies such as alternate prey/hosts, pollen, or nectar as well as microhabitats that are not available in weed-free cropping systems (Van Emden, 1965b). Weed species that support rich natural enemy faunas include the perennial stinging nettle (Urtica dioica), Mexican tea (Chenopodium ambrosioides), camphorweed
(Heterotheca subaxillaris), and goldenrod (Solidago altissima), as
well as many weeds of the Compositae and Umbelliferae families. However Howe ver,, weed background may be important in other ways during thee ea th earl rly y st stag ages es of th thee in inte terc rcro rop p wh when en th thee pe pest st ar arri rive ves. s. So Some me ev evid iden ence ce suggests that pests are deterred from ovipositing or remaining in crops with weedy background as opposed to bare soil. Also, many weeds wee ds may pro produc ducee dif differ ferent ent che chemic mical al sub substa stance nces, s, thu thuss con confus fusing ing insects that localize their crops through chemical-feeding cues. Selective ti ve man manage agemen mentt of the these se wee weed d spe specie ciess wit within hin int interc ercro rops ps may cha chang ngee the mortality of insect pests caused by natural enemies. The ecological basis for obtaining crop-weed mixtures that enhance insect biological suppression needs further development. development. The final choice of cropping design must be dictated by the t he local nutritional needs and preferences, economic feasibility, and yield advanta va ntages ges of the mix mixtur ture. e. Com Combin binati ations ons of cor corn-l n-leg egum umes es usu usuall ally y ov overeryield corn monocultures; in other words, more area is needed under corn monocultures to produce the same yield as one hectare of polyculture (Vandermeer (Vandermeer,, 1981). This overyielding capability is expressed as a land equivalent ratio (LER). If higher than one, the ratio implies that th at th thee in inte terc rcro rop p gi giv ves a be bett tter er yi yiel eld d th than an th thee mo mono nocu cult ltur uree (T (Tre renb nbath ath,, 1976). The LER is defined as the relative land area required for sole crops to produce the same yields as intercropping. Using a simpler notation for competition studies, LER can be expressed as: LER =
Y A
+
Y B
S A
+
SB
Where Y A and Y B are the individual yields from crops in intercropping, and S A and S B are those for the same species as sole crops. Vandermeer (1981) discusses methods for interpreting data from intercropping and believes believes the LER to be a useful and practical indicator.
Chapter 6
Insect Ecology Orchards Insect Ecology in Orchards Underin Cover-Crop Management Under Cover-Crop Management Most bio Most biolog logica icall con contro troll pr progr ograms ams ha have ve bee been n con condu ducted cted in orc orchar hards ds and in protected environments because these are considered to be more stable and permanent ecosystems than annual-crop agroecosystems (Huffaker and Messenger, 1976). Several authors have claimed clai med thatt ins tha insect ect po popul pulati ations ons are mo more re sta stable ble in com comple plex x orc orchar hard d com commun muniities because a diverse and more permanent habitat can maintain an adequate population of the pest and its enemies at critical times (van den Bosch and Telford, 1964). Orchards are semipermanent, relative ti vely ly un undis distur turbed bed sys system tems, s, wit with h no fa fallo llow; w; and cro crop p rot rotati ation on do does es no nott apply in the short term, so particular biological situations affecting insects occur in these systems. For most entomologists, the relative permanency of orchards affords the opportunity of manipulating the components of an orchard habitat to the benefits of ecologically sound orchard management practices (Prokopy, (Prokopy, 1994). One of these practices is the manipulation of ground-cover ground-cover vegetation to enhance biological control of orchard arthropod pests. In California, cover-crop management in orchards has been recommended because ground cultivation exacerbates soil erosion, reduces water penetration, and modifies the summer microclimate unfavorably (Finch and Sharp, 1976). Several legumes, such as lana vetch, clovers, and Medicago spp., and grasses such as brome, rye, and barley, have been recommended to be sown annually in orchards in the autumn or early spring, or at times farmers use cover crops that are self-regenerating and are thus sowed once. Cover crops are tilled or mowed yearly. Cover crops can add or retain soil nitrogen (N), facilitate the av availability ailability of their nutrients, produce organic matter, matter, reduce soil compaction, improve water infiltration, and in some cases enhance moisture retention. In fact, cover crops can act as an “eco-
logical turntable” influencing various agroecological processes simultaneously (Figure 6.1). The manipulation of ground-cover vegetation in orchards and vineyards can significantly affect tree growth by altering nutrient availability, ava ilability, soil physics and moisture, and the prev prevalence alence of weeds, plant pathogens, and insect pests and associated natural enemies (Haynes, 1980). A great number of entomological studies conducted in these systems indicate that orchards with rich floral undergrowth exhibit a significantly lower incidence of insect pests than cleancultivated orchards, mainly because of an increased abundance and efficiency eff iciency of predators and parasitoids (Smith et al., 1996). Early in the twentieth century, Peterson (1926) observed that uncultivated orchards were attacked less severely by codling moth (Cydia pomonella) than were continuously cultivated orchards. Peppers and Driggers (1934) and Allen and Smith (1958) showed that the percentage of fruit moth larval parasitism was always greater in orchards with weeds than in clean-cultivated orchards. In Ne New w Je Jers rsey ey pe peac ach h or orch char ards ds,, co cont ntro roll of th thee or orie ient ntal al fr frui uitt mo moth th in in-creased in the presence of ragweed ( Ambrosia sp.), smartweed ( Poly-
FIGURE 6.1. Multiple and interactive effects effects of cover crops on farming farming systems (after Michigan State University Bulletin E-2704, 2000).
gonum sp.), lamb’s-quarter (Chenopodium album), and goldenrod (Solidago sp.). These weeds provided alternate hosts for the parasite Macrocentrus ancylivorus (Bobb, 1939). Similarly, Leius (1967)
found that the presence of wildflowers in apple orchards resulted in an eighteenfold increase in parasitism of tent caterpillar pupae over nonweedy orchards; parasitism of tent caterpillar eggs increased fourfold, and parasitism of codling moth larvae increased fivefold. Considerable work was conducted by researchers in the Soviet Union on the role of nectar plants in increasing the effectiveness of biological control agents in orchards. Telenga (1958) reported that the parasitoid Scolie dejeani was attracted to its grub hosts when the honey plants Phacelia and Eryngium were sown. These same plants were we re sh sho own to in incr crea ease se th thee ab abun unda danc ncee of th thee wa wasp sp Aphel Aphelinus inus mali for the control of apple aphids and to improve the activity of Trichospp. p. wa wasp spss in ap appl plee or orch char ards ds.. So Sovi viet et re rese sear arch cher erss at th thee Tas ashhgramma sp kent Laboratory cited lack of adult food supply as a reason for the inability inab ility of Aphytis proclia to control its host, the San Jose scale (Quadraspidiotus perniciosus). The effectiveness of the parasitoid impr im prov oved ed as a re resu sult lt of pl plan antin ting g a Phacel cove verr cr crop op in Phacelia ia tanacet tanacetifolia ifolia co the orchards. Three successive plantings of the Phacelia cover crop increased parasitization of scales from 5 percent in clean-cultivated plots to 75 percent in the Phacelia plots (Churnakova, 1960). In the Solomon Islands, O’Connor (1950) recommended the use of a cover crop in coconut groves to improve the biological control of coreid pests by the ant Oecophylla smaragdina subnitida. In Ghana, coconut gave light shade to cocoa and supported, without apparent crop loss, high populations of Oecophylla longinoda, keeping the cocoa crop free from cocoa capsids (Leston, 1973). Wood (1971) reported that in Malaysian oil palm (Elaeis guineensis) plantations, heavy ground cover, irrespective of type, reduced damage to young trees caused by the rhinoceros beetle (Oryctes rhinoceros). The mode of action is not certain, but it appears that the ground cover impedes flight of the adult beetles or restricts their movement movem ent on the ground. Economic control of this pest was possible by simply encouraging the growth of weeds between the trees. Sluss (1967) reported another example of advantageous use of ground cover cover under trees. In California, the beetle Hippodamia convergens is the most important predator of walnut aphid (Chromaphis juglandicola) during the early season. This beetle moves from its
overwintering area in the mountains to the walnut orchards in Februoverwintering ary and early March, when there are no leaves on the trees and therefore fo re no ap aphi hids ds.. Ho Howe weve verr, so some me ap aphi hids ds ar aree pr pres esen entt in th thee gr grou ound nd co cov ver unde un derr th thee tr tree eess an and d se serv rvee as a tem tempo pora rary ry fo food od so sour urce ce fo forr th thee pr pred edat ator orss which would otherwise move on or die of starvation. The ground cover under the trees should be chopped or disked in late April or early May to force the beetles onto the walnut trees. If it is chopped too to o ea earl rly y, ho howe weve verr, th thee be beet etles les wi will ll em emig igra rate te be befo fore re th thee wa waln lnut ut ap aphi hids ds have appeared appeared on the trees; if it is chopped too late, the t he large number of beetles will decimate the aphid population on the trees without ovip ov ipos ositi iting ng,, re resu sulti lting ng in fe fewe werr be beet etle less la late terr. Th Thus us,, ti timi ming ng of th thee ch chop op-ping of the ground cover is critical to maintain ample beetle population for sufficient control of the aphids. Fye (1983) proposed manipulation of orchard ground vegetation for building predator populations. In pear orchards of the Yakima Valley, he established various small grain and crucifer cover crops and an d fo foun und d th that at se sev ver eral al sp spec ecie iess of ge gene nera rall pr pred edat ator orss we were re su supp ppor orte ted d by aphids and Lygus bugs harbored by the cover crops. In Michigan, ground plants can be allowed to grow up to the apple trees, since rainfall is sufficient for the trees not to suffer from competition tio n fo forr wa water ter by th thee gr gras ass. s. Th Thee ph phyt ytop opha hago gous us mi mite tess pr pres esen entt on th thee co cove verr consti con stitut tutee an ear earlyly-sea season son foo food d sou source rce fo forr the pre predat datory ory mit mitee Amblyseius fallacis, which later moves up into the trees and regulates the spider mites Panonychus ulmi and Tetranychus urticae (Croft, 1975). Bugg Bu gg an and d Wad addi ding ngto ton n (1 (199 994) 4) pr prov ovid idee a lis listt of un unde ders rsto tory ry we weed edss or “resident vegetation” that can become an asset when managed as cover crops as they harbor beneficial arthropods. Among the main species included are common knotweed, chickweed, toothpick ammi, sweet fennel, and sow thistle. In China, Liang and Huang (1994) report that Ageratum conyzoides as well as other plants ( Erigeron Erigeron annuus, Aster tataricus, etc.) which encourage natural enemies, especially Amblyseius spp., of the citrus red mite (P (Panonychus anonychus citri) have been planted or conserved as ground cov co ver in an ar area ea of 13 135, 5,00 000 0 ha of ci citr trus us or orch char ards ds wi with th ex exce celle llent nt re resu sult lts. s. Also in China, Yan and colleagues (1997) developed a cover-crop system sys tem in app apple le orc orchar hards ds con consis sistin ting g of Lago (Labiatae) atae) in Lagopsis psis supina (Labi stead of the traditional cover of Chinese rape (Brassica campestris) and/or and/ or alfal alfalfa. fa. Lago had d a su subs bsta tant ntia iall lly y gr grea eater ter ef effe fect ct in en en- Lagopis pis supina ha hancing natural enemy populations than the other two plants.
The cover-crop work of McClure (1982), although not directed toward towar d natural enemy enhancement, is proving useful for the manipulation of leafhopper pests in peaches. McClure’ McClure’ss experiments demonstrated that ground cover significantly impacted the number of leafhoppers, Scaphytopius acutus (vectors of x-disease), colonizing peac pe ach h tr tree ees. s. Th Thee gr grea eatt ma majo jori rity ty of ad adul ultt le leaf afho hopp pper erss oc occu curr rred ed on tr tree eess with undergrowth of red clover and rosaceous weeds. Relatively few adults inhabited trees in plots with orchard grass, an unsuitable host. Thes Th esee da data ta in indi dica cate te th that at in inv vas asio ion n of th thee or orch char ard d by le leaf afho hopp pper erss ca can n be discouraged by rendering the orchard floor free of naturally occurring wild host plants. In Califo California’ rnia’ss Centr Central al Valley vine vineyard yards, s, var variega iegated ted leafh leafhoppe opperr population differences between cover and noncover plots were clear-cut forr all th fo thre reee br broo oods ds,, but th thee re reas ason onss be behi hind nd th thes esee di difffe fere renc nces es we were re no nott so clear. Anecdotal reports from growers in the area suggest that weedy cover crops in early to midseason may result in smaller populations of leafhoppers. An increase in the abundance of generalist predators, especially spiders, may help reduce leafhopper populations in the weed-cover plots (Settle et al., 1986). In the same area, leaving a managed ground cover of johnsongrass or Sudan grass, a minor cultural practice modification in vineyards, resulted in a habitat modification which greatly enhanced the activity of predators again ag ainst st ph phyt ytop opha hago gous us mi mites tes su such ch as th thee Wil illam lamet ette te mi mite te.. Wh When en jo john hn-songrass (Sorghum halepense) was allowed to grow in grape vineyard ya rdss in Ca Calif lifor orni nia, a, th ther eree wa wass a bui uild ldup up of al alte tern rnat atee pr prey ey mi mites tes,, wh whic ich h supported populations of the predatory mite Metaseiulus occidenwhich ich,, in tu turn rn,, re rest stra rain ined ed th thee Pa Paci cifi ficc mi mite te,, Eotetrany talis, wh Eotetranychus chus willame willamette, tte, to noneconomic numbers (Figure 6.2) (Flaherty, 1969). Also in the San Joaquin Valley, the emergence of navel orangeworm adults (Amyelois transitella) was significantly higher in the complete, residual herbicide-treated almond orchards than in i n orchards with a vegetation cover. These results show that fewer navel orangeworms survive the winter on the ground if cover crops are present. The differences might be greater when nuts in cover crops are fully subjected to regular, early-spring mowing. mowing. Nuts in the t he residual herbicide treatments, which do not need mowing, would not be disturbed. Mowing, especially flail mowing, could reduce the navel orangeworm population further by physically destroying the overwintering nuts and larvae (Bugg and Waddington, 1994).
F A E L R E P S E TI M E T T E M A L LI W R E B M U N E G A R E V A CLEAN CU CULTIVATED
SUDAN GR GRASS
FIGURE 6.2. Effect of ground cover cover on Willamette mite populations in a California vineyard (after Flaherty Flahe rty,, 1969).
SELECTING AND MANAGING COVER CROPS IN ORCHARDS
In cases where ground cover of any type is desirable, easily managed plants should be encouraged in preference to more aggressive species. In general, low-growing, nonclimbing legumes would be preferable to climbers or tall grasses, since climbers could cover the trees and tall grasses could limit movem movement ent between the trees. In addition, legumes such as vetch and fava beans fix more than 150 kg of N/ha and produce considerable amounts of biomass, an important in-
put of organ put organic ic mat matter ter into into the or orcha chard rd soil. soil. If the gro ground und cove coverr will be regularly cut, it would be advantageous to use resident weeds that grow early in the spring and can regrow after repeated mowing or disking. These would probably be low-gr low-growing owing perennial grasses or perennial broad-leaved broad-leaved species, or a combination of the two. The advantage of weeds in this situation is that they can often take much more abuse than cultivated species, and are thus easier to manage. There are, however, some cultivated legumes that might also be suitable for continual mowing. The understory vegetation in an orchard need not be managed uniforml fo rmly y. Dif Differ ferent ent zon zones es may be tre treate ated d dif differ ferent ently; ly; thi thiss is ter termed med strip management, bec becaus ausee the dif differ ferent ent tre treatm atment entss are usu usuall ally y app applie lied d lin lin-early, and the different understory zones appear as bands or strips running through an orchard. Strip management of cover crops may entail (1) sowing cover crops of different floristic composition in different strips; (2) mowing strips at different times; (3) tilling strips at differentt times; or (4) combinations of these three processes. Sowing differen of different mixes leads to stands with differing statures and phenologie olo gies, s, thu thuss af affor fordin ding g di dive verse rse res resour ources ces to pes pestt and ben benef eficia iciall art arthro hro-pods (Bugg and Waddington, 1994). Wyss, Niggli, and Nentwig (1995) planted strips of weed mixtures to enhance populations of aphidophagous insects and spiders in Swiss apple orchards. During flowering of weeds more aphidophagous predators were observed on the apple trees within the stripsown so wn ar area ea th than an in th thee co cont ntro roll ar area ea.. Th Thee mo most st ab abun unda dant nt an and d pe perm rman anen entt aphidophagous aphidophago us predators were spiders, predaceous Heteroptera, Coccinellidae, and Chrysopidae. Both species of aphids were significantly less abundant in the area with weed strips than in the control areaa dur are during ing the ve veget getati ation on per period iod.. The Their ir res result ultss sup suppo port rt str strip ip man manage age-ment as a viable option to manage aphids. Ideally,, cover crops should be selected or managed so as to (1) not Ideally harbor important pests; (2) divert generalist pests; (3) confuse specialis cia listt pes pests ts vis visual ually ly or olf olfacto actory ry and thu thuss red reduce uce the their ir col colon onizat ization ion of orchard trees; (4) alter host-plant nutrition and thereby reduce pest success; (5) reduce dust and drought stress and thereby reduce spider mite outbreaks; (6) change the microclimate and thereby reduce pest success; and (7) increase natural enemy abundance or efficiency, thereby increasing biological control of arthropod pests (Bugg and Waddington, 1994).
CASE STUDY 1: APPLE ORCHARDS IN CALIFORNIA
During 1982 and 1983 a study was conducted on the effects of cover-crop cover -crop manipulation on arthropod communities in three northern California apple orchards. The objectives were to (1) compare population levels and fruit damage by insects such as codling moths, aphids, and leafhoppers in orchards grown under clean cultivation or with wi th co cov ver cr crop ops, s, (2 (2)) fi find nd ou outt wh whet ethe herr un unde ders rso owi wing ng co cove verr cr crop opss in or or-chards would enhance populations of resident beneficial insects, and (3) evaluate the effects of cover-crop manipulation on tree growth and productivity. The study comprised compris ed one disked orchard kept free of ground vegetation by one spring and one late summer disking. The other covercropped orchard was undersown undersown in the fall with approximately twenty pounds of bell bean (Vicia faba) seeds per acre. By early June, the cover was mowed and the residues allowed to remain on the soil as straw mulch for the rest of the season. The relative abundance of plant-feeding insects and associated natural enemies were monitored on five randomly selected trees per orch or char ard, d, on co cove verr cr crop ops, s, an and d on th thee or orch char ard d fl floo oorr. In ea each ch or orch char ard, d, th thee lower canopy of each tree was sampled for one minute with a D-Vac insectt suctio insec suction n machine. machine. A pitfall pitfall trap filled filled with 75 percen percentt wate waterr and 25 pe perc rcen entt an antif tifre reez ezee pl plac aced ed at th thee ba base se of ea each ch tr tree ee ca capt ptur ured ed gr grou ound nd-dwelling arthropods. Two Two Zoecon codling moth (Cydia pomonella) pheromone traps were placed in each orchard, indicating peak flights of male moths. To assess codling moth damage at the end of each season, larval entries in 100 fruits collected from each of the five sample trees in each orchard were examined. All fruits from each tree were counted and weighed to determine yields per tree and percentage of total fruit damaged. Weekly evaluations of the proportion of infested twigs per tree indicated aphid and leafhopper levels. Predation on tree foliage was assessed with twenty-five paper card ca rdss (3 in inch ches es 4 in inch ches es), ), ea each ch co cont ntai aini ning ng fi fift fty y Me Medi dite terr rran anea ean n fl flou ourr moth (Anagasta kuehniella) eggs per card, hung in five trees per orchard. Gr Grou ound nd pr pred edati ation on wa wass es estim timat ated ed by 50 ca card rdbo boar ard d sh shee eets ts (8 (8.5 .5 in inch ches es square), each with twenty glued potato tuberworm (Phthorimaea operculella) larvae, randomly placed on the floor of each orchard. The
cards and sheets were removed after twenty-four hours and remaining eggs and larvae counted. In 19 1982 82,, su subs bstan tanti tiall ally y mo more re ma male le co codl dlin ing g mo moth thss we were re ca caug ught ht in th thee disked orchard (a total of 275 caught on nine sampling dates) than i n the co cove verr-cro cropp pped ed orc orchar hard d (16 (164 4 mot moths) hs).. The These se dif differ ferenc ences es did no nott occurr du cu duri ring ng 19 1983 83.. De Dens nsiti ities es of th thee ro rosy sy ap appl plee ap aphi hid d (Anur (Anuraphis aphis roseus) were slightly higher during May and June 1982 in the disked than in the cover-cropped cover-cropped orchard. In 1983, rosy apple aphids were detected only in the disked orchard, where they reached high numbers in early June.. Leafh June Leafhoppe oppers rs (Hom (Homopte optera: ra: Cicade Cicadellidae llidae)) colon colonized ized the orch orchards ards late in the 1982 season, reaching substantially higher densities in the disked than in the cover-cropped orchard (Figure 6.3). In bo both th ye year ars, s, po popu pulat latio ions ns of na natu tura rall en enem emies ies on th thee tr tree eess re rema main ined ed low,, and no differences between orchards were apparent in seasonal low abundance of common predators, such as Coccinellidae, Chrysopidae, and Cantharidae. Only spiders reached higher numbers on the trees with a cover crop early in the 1983 season. Despite these undetectable general differences in predator abundance, the number of Anagasta eggs removed from the trees was substantially higher in the coverDisked (no cover crop) Cover-cropped s gi wt 0 2/ sr e p p o hf a el . o n e g ar e v A
Jul
Aug Sampling dates in 1982
Sep
FIGURE 6.3. Leafhopper densities on apple trees with and without cover crop in California (after Altieri and Schmidt, 1986a).
cropped orchard than in the disked orchard, especially during July and August in both 1982 and 1983. As expected, disking primarily affected ground predators. In both year ye ars, s, an ants ts an and d sp spid ider erss we were re ca caug ught ht in pi pitf tfal alls ls mo more re co cons nsis iste tent ntly ly in th thee cover-cropped cover -cropped orchard than in the t he disked orchard, and carabid ground beetles appeared to be more prevalent in the disked orchard, especially from July on. Predators (especially ants) appeared to be more effecti eff ective ve in remo removing ving Phtorimaea lar larv vae in th thee co cove verr-c -cro ropp pped ed th than an in the disked orchard. A variety of general predators and parasitic Hymenoptera (mainly larger Braconidae and Ichneumonidae) were present on the covercrop vegetation. Most were supported by the high numbers of alternativ nat ivee pre prey y (es (espec pecial ially ly aph aphids ids)) har harbo bored red by the co cove verr-cro crop p ve veget getatio ation n from early April through mid-June in both years. Leafhoppers were particularly prevalent on the cover crops in 1983. In 1982, there were no apparent differences in fruit yields between the two orchards, but codling moth damage seemed slightly lower in the cover-cropped orchard. In 1983, however, cover-cropped trees produced considerably more fruit, although they were smaller, than did trees in the disked orchard. Codling moth incidence was substantially lower in the cover-cropped orchard (Table 6.1). The results suggest the following tentative description of difference en cess be betw twee een n ap appl plee sy syst stem emss wi with th an and d wi with thou outt co cove verr cr crop ops: s: Ap Appl plee or or-TABLE 6.1. Apple Production and Codling Moth (Cydia pomonella) Damage pomonella) Damage in Organic Orchards with and Without Cover Crop in Northern California
Orchard and Year*
Total No. of Fruits/Tree
Total Fruit Weight/Tree (kg)
Fruit>2.5" Diameter (%)
Fruit with Codling Moth Damage (%)
Cover 1982
241
35.6**
29.1
3.6
44.8
9.8
68.0
9.7
1983
334
56.8
58.3
9.7
87.6
14.6
4.2
0.7
1982
260
37.1
26.9
4.5
39.0
6.5
78.0
9.7
1983
94
11.7
15.5
2.6
54.8
11.0
38.9
7.8
Disked
Source: Altieri and Schmidt, 1986b. *Total rainfall during the growing season (April-October) was 341 mm in 1982 and 367 mm in 1983. **Means ± SD
chards with cover crops generally had (1) lower infestation levels of aphids, leafhoppers, and codling moths, (2) more species and more individuals of soil-dwelling predaceous arthropods, and (3) higher removal remov al rates of artificially placed prey prey.. In contrast, disked systems were generally characterized by greater numbers of plant feeders on the trees and by relatively low low population levels of natural enemies. The cover crops generally general ly harbored large numbers number s of prey, prey, such as aphids and leafhoppers, which attracted varying numbers of predators. High numbers of predators on the cover crops, however, did not necess nec essari arily ly tra transl nslate ate int into o hig higher her num number berss on the tre trees. es. Exp Experi erimen ments ts to test whether the common practice of mowing the cover crop forces natural enemies to move up to the trees could be useful in designing manage man agemen mentt pla plans ns for enc encou ourag raging ing ef effi ficie cienc ncy y of nat natur ural al ene enemie mies. s. Although cover cropping significantly affected ground-predator ground-predator population lat ions, s, the these se stu studie diess cou could ld not det determ ermine ine ho how w the these se cha change ngess af affec fected ted pest species on the trees. The data also al so did not indicate how realistically predation on artificial baits related to t o reduction of apple pests, such as codling moths, aphids, and leafhoppers. Depending on the orchard system, cover-crop complex, and associated arthropod species, it seems that manipulation of the ground cover can have a significant effect on the number of arthropods that inhabit the orchard by (1) directly affecting plant-feeding populations which discriminate between trees with and without cover underneath or (2) attracting and retaining soil- and foliage-inhabiting natural enemies by providing alternative food and habitats. Critical testing of these effects in a range of orchard systems may lead to improvee biological control of certain orchard pests. prov
CASE STUDY 2: PECAN ORCHARDS IN GEORGIA
In so sout uthe hern rn Ge Geor orgi gia, a, pe peca can n tr tree eess ar aree att attac acke ked d by se seve vera rall ap aphi hid d sp speecies, including yellow pecan aphid (Monelliopsis pecanis), blackmargined aphid (Monellia caryella), and black pecan aphid (Melan (Melan-ocallis caryaefoliae). Through phloem feeding, these can reduce tree vigorr and prod vigo producti uctivity vity.. More Moreov over er,, late-s late-season eason outb outbreaks reaks may be prompted by insecticides used against other pests of pecan, and there is growing evidence of aphid resistance to t o available insecticides.
Winter legumes such as hairy vetch (Vicia villosa) and crimson clover (T (Trifolium rifolium incarnatum) are now being used in attempts to enhance early-season biological control of pecan aphids. This practice is an alte alterna rnativ tivee to the pr prev evale alent nt un under dersto storyry-man manage agemen mentt app appro roach ach in pecan that inv involves olves herbicide tree-row strips with intervening alleys of mown grass, to facilitate harvest by mechanical shakers and sweepers. These warm-season grasses employed (e.g., Bermuda grass [Cynodon dactylon] and centipede grass [Eremoc [Eremochloa hloa ophiuroides]) harbor few beneficial insects and add little to soil nitrogen. Therefore, there is a need to dev develop elop a low-input, minimum-tillage scheme for year-round management of cover crops in the pecan agroecosystem. syste m. In orde orderr to meet this goal, Bugg and Dutc Dutcher her (1989) conducted trials of several prospective warm-season cover crops that would serve as potential “insectary crops.” Among the various evaluated species Bugg and Dutcher (1989) found fo und tha thatt Sesbani wass th thee be best st so sour urce ce of co cowp wpea ea ap aphi hid d an and d Sesbania a exal exaltata tata wa the best reservoir for various hover flies and coccinellid beetles. Cowpea aphids on Indigofera hirsuta attracted various aphidophagous predators in November after pecan leaves were shed. They further found that during the summer, summe r, the flowers of buckwheat buckwheat (Fagopyrum esculentum) attracted many entomophagous wasps. The understory of sesbania cover crops supported high densities of banded-winged whitefly and cowpea aphid, which were colonized by cocinellids (Olla v-nigrum and Hippodamia conver convergens gens). In another study, Bugg and Waddington (1994) found that in mature pecan orchards under minimal or commercial management, cool-season understory cover cover crops of hairy vetch and rye sustained significantly higher densities of aphidophagous aphidophagous lady beetles than did unmown resident vegetation or mown grasses and weeds. In covercropped understories, mean densities of aphidophagous coccinellids were we re ne near arly ly si six x tim times es gr grea eate terr th than an in un unmo mown wn re resi side dent nt ve vege geta tatio tion n an and d approximately eighty-seven eighty-seven times greater than in mown grasses and weeds. Dutche Dut cherr (19 (1998) 98) con consid siders ers co cove verr cro crops ps ke key y com compo ponen nents ts of pec pecan an integrated pest management (IPM) combined with reduced frequency of pesticide sprays, planting legumes as intercrops in the orchard to produce alternate prey aphids for aphidophaga, and partitioning of thee fo th fora ragi ging ng be beha havi vior or of th thee re red d im impo port rted ed fir iree an antt wi with th tr trun unk k sp spra rays ys of insecticide that prevent ants from reaching aphids and mealybugs in
the tree, yet allowing ants to remain on the orchard floor as predators of pecan weevil larvae. Years of experimentation show that intercropping with sesbania alone or the combination of intercropping with hairy indigo and ant exclusion reduce pecan aphid populations in Georgia. CASE STUDY 3: SUMMER COVER CROPS IN VINEYARDS
In Cal Califo iforni rnia, a, som somee res resear earche chers rs ha have ve tes tested ted pla planti nting ng co cove verr cro crops ps as a ha habi bitat tat ma mana nage geme ment nt ta tact ctic ic in vi vine neya yard rdss to en enha hanc ncee na natu tura rall en enem emie ies, s, including spiders (Costello and Daane, 1998). Reductions in mite (Flaherty, 1969) and grape leafhopper (Daane et al., 1998) populations have been observed, but such biological suppression has not been sufficient from an economic point of view (Daane and Costello, 1998). Perhaps the problem lies in the fact that most of these studies were conducted in vineyards with winter cover crops and/or with weedy resident vegetation which dried early in the season or which was mowed or plowed under at the beginning of the growing season. Therefore, in early summer these vineyards are virtual monocultures without floral diversity. diversity. For this reason Nicholls, Parrella, and Altieri (2000) tested the idea to maintain a green cover during the entire growing grow ing season in order to provide habitat and an alternate food for natural enemies. They sowed summer cover crops (buckwheat and sunflower) sunflow er) that bloom early and throughou t hroughoutt the season, thus providing a hig highly hly con consis sisten tent, t, ab abun undan dant, t, and wel well-d l-disp ispers ersed ed alt altern ernati ative ve foo food d source, as well as microhabitats, for a diverse community of natural enemies. This study was conducted in two identical adjacent Chardonnay organic vineyard blocks from April to September in 1996 and 1997. Vineyards were located in Hopland, 200 km north of San Francisco, California. One block was kept free of ground vegetation by one spring and one late-summer disking (monoculture vineyard). vineyard). In April, the other block bloc k (cover-cropped vineyard) was undersown in every alternate row with a 30/70 mixture of sunflower and buckwheat. Buckwheat flowered from late May to July, and as the buckwh buckwheat eat senesced, sunflower sunflow er bloomed from July to the end of the season. These The se res resear earche chers rs fou found nd tha thatt mai mainte ntenan nance ce of flo floral ral di dive versi rsity ty throughout the growing season in vineyards in the form of summer
cover crops had a substantial impact on the abundance of western grape leafhoppers, Erythroneura elegantula (Homoptera: Cicadellidae), western flower thrips, Frankliniella occidentalis (Thysanoptera: Thripidae), and associated natural enemies. During two consecutive years, vineyard systems with flowering cove co verr cro crops ps wer weree cha charac racter terize ized d by lo lower wer den densit sities ies of adu adult lt and ny nymp mph h leafhoppers (Figure 6.4) and thrips, and larger populations and more species of general predators, including spiders. Although Anagrus epos (Hymenoptera: Mymaridae), the most important leafhopper parasitoid, achieved high numbers numbers and inflicted noticeable mortality of grape leafhopper eggs, no differences in egg parasitism rates were observed between cover-cropped and monoculture systems. Mowing of co cove verr cro crops ps for forced ced mo move vemen mentt of Anagrus preda edator torss to adj adjace acent nt Anagrus and pr vines resulting in the lowerin l owering g of leafhopper densities in such vines (Figure 6.5).
p ar t/ sr e p p o
No cover
hf a e
Cover
L . o N n a e M
Dates in 1996
FIGURE 6.4. Densities of adult leafhoppers E. elegant elegantula ula in in cover-cropped and monoculture vineyards in Hopland, California, during the 1996 growing season. Mean densities (number of adults per yellow sticky trap) and standard errors of two replicate means are indicated. In some cases error bars were too small to appear in the figure (after Nicholls, Parrella, and Altieri, 2000).
(a)
f a el / s h p m y N
No mow o
Mow
. N n a e M
Before mowing
1 week after
2 weeks after
(b)
T S Y/
s u r g a n A
No mow
.
Mow
o N n a e M
Before mowing
1 week after
2 weeks after
FIGURE 6.5. Effects of cover-crop mowing in vineyards on densities of leafhopper nymphs and Anagrus epos during epos during the 1997 growing season in Hopland, California (after Nicholls, Parrella, and Altieri, 2000).
These studies showed that cover crops harbored a large number of Orius, coccinellid, thomisid spiders, and a few other predator species. Comparisons of predator abundan abundance ce in both blocks showed that the pre presen sence ce of suc such h pr preda edator torss on bu buckw ckwhea heatt and sun sunflo flower wer pr produ oduced ced an increase in the density of predators in the cover-cropped vineyards. Such greater densities of predators were correlated with lower leafhopper numbers, and this relationship was much more clear-cut in the case of the Orius-thrips interaction. The mowing experiment suggests a direct ecological linkage, as the cutting of the cover cover-crop -crop vegetation forced the movement of Anagrus and predators harbored by the flowers, both years resulting in a decline of leafhopper numbers on the vines adjacent to the mowed cover crops. This study coincides with Boller (1992), who reported that when understory summer vegetation is present in the vineyards, a highly complex habitat containing multiple strata is developed. Boller’s group summarized twelve years of investigations conducted in vineyards yar ds of nor northe thern rn Sw Switze itzerla rland. nd. In a sur surve vey y of twe twenty nty-on -onee vin viney eyard ardss of varying flower flower richness these authors found a considerable increase i ncrease in “neutral herbivores” and of beneficial entomophagous species with increasing numbers of plant species. Their results confirmed that the proportion of flowering flowering perennial dicot plants was responsible for the increase i ncrease in beneficial arthropod taxa. They also found that in vi vine neya yard rdss ex exhi hibi biti ting ng fl flor oraa wi with th a la larg rgee nu numb mber er of pl plan antt sp spec ecie ies, s, th thee tendency in most species (grape moths, spider mites, eirophyd mites, thrips, and noctuid larvae) was to fluctuate much less at significantly lower density levels than pest populations in botanically poor vineyards. yar ds. In par partic ticula ularr the they y men mentio tion n tha thatt flo flower wer-ri -rich ch vin viney eyard ardss ex exhib hibited ited higher hig her par parasi asitiza tizatio tion n of gra grape pe mot moth h eg eggs gs by Trichogramma cacoeciae and higher populations of the predatory mite Typhlodromus pyri.
Chapter 7
Influence Adjacent TheThe Influence of Adjacentof Habitats on InsectHabitats Populations in Crop Fields on Insect Populations in Crop Fields
The biogeographic region rather than the single homogeneous fie ield ld ma may y of often ten be th thee ap appr prop opri riat atee un unit it fo forr pe pest st-m -man anag agem emen entt re rese sear arch ch (Levins and Wilson, 1979). According to Rabb (1978), an agroecosystem should be conceived as an area large enough to include those uncultivated landscapes that influence crops through intercommunity interchanges of organisms, materials, and energy. Our single-commodity approach approach in organizing research on pests often allows us to ignore associations with other crops, host plants, and adjacent plant communities that are of critical importance in the life systems of pests. The ve veget getatio ational nal com compon ponent ent of agr agroec oecosy osyste stems ms can be vie viewed wed as a mosaic of annual and perennial crop fields, forest patches, pasturelands, fallow fields, orchards, swamps, old fields, and tree plantations. The agricultural landscape consists of (1) the agricultural field (consisting usually of a single crop and any weeds present, but sometimes including additional crops or a cover); (2) native and/or weedy vegetation that may be present on its borders; (3) the surrounding agricultural fields; and (4) the vegetation occurring in native or uncultivated habitats in the surrounding area. The composition of the agricultur cul tural al lan landsc dscape ape det determ ermine iness the pre presen sence ce of ov overw erwint interi ering ng sit sites es and the ability of an insect to locate appropriate habitats and food resources over the course of its lifetime (Perrin, 1980). Although it is useful to consider agroecosystems as a crop “island” subject to colonization from several sources, a regional perspective is necessary for predi pr edicti cting ng the mo move vemen mentt pat patter terns ns of pes pests ts and nat natura urall ene enemie miess acr across oss agricultural landscapes. Rabb (1978) argues that an insect population’ss performance and survival is related to the large-unit ecosystem tion’ heterogeneity, especially when the risks assumed by the population in moving between different sites are considered. Root (1975) dis-
cusses “compound ecosystems” and the variable responses of herbivores to the dispersion and size of resource patches within habitats. Many predators tend to feed on several different species of prey and to distribute themselves on vegetation in response to the availability of prey rather than the plant species. On a broad scale, just as the heterogeneity of distributions of plant populations can influence the effects of herbivores on plants, so the heterogeneity of herbiv herbivore ore distribution patterns can influence the effects of predators and/or parasites on the herbivores. Several studies suggest that the vegetational settings associated with particular crop fields influence the kind, abundance, and time of arrival of herbivores and their natural enemies (Price, 1976). Large populations of certain pests, especially polyphagous, univoltine univoltine species, migrate en masse from alternate hosts in the vicinity to newly established (and presumably, vulnerable) crop monocultures (Duelli et al., 1990). The scale and intensity of this phenomenon depends, of course, upon the vagility of the insects involved (Andow, 1983b). Inse In sect ctss ar aree mo mobi bile le sp spec ecie ies; s; th thee sc scal alee of th thei eirr ho home me ra rang ngee di difffe fers rs ac ac-cording to their method of locomotion (walking or flying) and dispersal. Highly mobile organisms may use different fields or uncultivat ated ed ar area eass du duri ring ng th their eir li life fe.. St Stud udie iess of th thee ec ecol olog ogy y of su such ch or orga gani nism smss must mu st co cons nsid ider er pr proc oces esse sess no nott on only ly on in indi divi vidu dual al si sites tes but als also o at th thee re re-gional or landscape level. The diversity of the farmland mosaic, defined by the variety of crops and wild plants, and their spatial arrangement, for example, the size of fields and the heterogeneity of their spatial distribution, play key roles in determining the abundance, dan ce, di dive versi rsity ty,, and dis disper persio sion n of ins insect ect spe specie ciess (Ba (Baudr udry y, 198 1984). 4). The movement movem ent of individual insects will respond to a wide range of landscape factors including the scale of the habitat (plant, field, and landscap sc apee le leve vel) l),, ha habi bita tatt pe perm rmea eabi bilit lity y, pa patc tch h si size ze an and d sh shap ape, e, an and d de degr gree ee of isolation (Figure 7.1). Landscape structure influences microclimate and crop growth, as well as other factors that affect the movement patterns of insects. Stud St udie iess in No Nort rth h Ca Caro roli lina na ha have ve sh sho own th that at fiel ield d si size ze an and d ge geog ogra raph phica icall distribution of fields can affect the movement and location of Mexican bean bea n bee beetle, tle, Epilac infest estatio ation n (St (Stinn inner er et al. al.,, 198 1983). 3). Sin Since ce Epilachna hna varives varivestis, tis, inf this insect overwinters along the edges of woods and reproduces much faster on garden beans than on soybean, the beetles exhibit a seasonal movement from woody edges to beans to soybean and back to woods.
FIGURE 7.1. Insect FIGURE Insect mov movement ement in respon response se to landscape struct structure ure on arab arable le farms shown at three scales: (a) the scale of the individual plants; (b) the field scale; and (c) the landscape scale. The size, shape, and spatial pattern of patches will be important across all of these scales (after Fry, 1995).
Significan Signif icantt mo move vemen mentt aga agains instt the dir directi ection on of pre preva vailin iling g win winds ds is pos possisiblee by str bl stron ong g fl flier ierss or si simp mply ly by al alig ight htin ing g un unti till th thee wi wind ndss ch chan ange ge fr from om th thee prevailing direction. Johnson, Turpin, and Bergman (1984) describe an examp ex ample le of sou southe therly rly mo move vemen mentt by Leptinotar thee Co Coll Leptinotarsa sa deceml decemlineata, ineata, th orado potato beetle, in the same area that Empoasca fabae moves from spring breeding grounds in the lower Mississippi basin to Wisconsin crop fields with the prevailing air currents.
CROP EDGES AND INSECT PESTS
Several species of weeds present in areas adjacent to crops may serv se rvee as al alte tern rnat ativ ivee ho host stss fo forr cr crop op pe pest stss (V (Van an Em Emde den, n, 19 1965 65b) b).. Mo Most st of thes th esee in inse sect ct pe pest stss te tend nd to fe feed ed on wi wild ld pl plan ants ts bo bota tani nica cally lly re rela late ted d to th thee
crops. crop s. Ov Over er 20 200 0 pe pest stss of ce cere real alss ut util iliz ized ed wi wild ld gr gras asse sess th that at ar aree pa part rtic icuularly abundant and ubiquitous in arable areas. Insect movement between uncultivated land and crops can be related to the natural disdis persal of the pest, lack of suitable food in one of the habitats, host alte al tera rati tion on,, or ma majo jorr di dist stur urba banc nces es,, su such ch as th thee us usee of he herb rbici icide des. s. Sm Small all flying insects such as aphids and thrips may be concentrated on crop edges by displaced air currents attributable to nearby windbreaks (Lewis, 1965). Van Emden (1965a) found a heavy initial edge infestation of alate cabbage aphids caused by shelter to windward; however, because of increased mortality and a decreased reproductive rate probably caused by physical factors, the density of aphids was soon half that of the crop center. Adjacent habitats, such as shelter belt be lts, s, ca can n be us used ed as ove verw rwin inte teri ring ng si sites tes by pe pest sts. s. In th thee no nort rthe hern rn ro roll ll-ing plains of Texas, problems with the boll weevil (Anthonomus grandis) are linked to the planting of shelterbelts which provide litter for the adult boll weevil to overwinter in a state of diapause (Slosser and Boring, 1980). Certain hedgerow plants provide sources of several species of pests and predators that may move into adjacent apple orchards (Solomon, 1981). In England, the winter moth Operophtera brumata feeds on wild Prunus spp., beech, and oak, in addition to apple. The larvae can disperse by drifting in the wind, so hedgerow and woodland la nd tr tree eess ca can n be si sign gnif ific ican antt lo loca call so sour urce cess of th this is pe pest st.. Th Thee un unde derl rlyi ying ng herb layer of a woody border may include plants attractive to crop pests. pes ts. Wain ainho house use and Coa Coake kerr (19 (1981 81)) fou found nd tha thatt the dis distri tribu butio tion n of the perennial stinging nettle, Urtica dioica, explained the abundance of carrot fly, Psila rosae, in noncrop borders and have suggested a strategy eg y to sim simpli plify fy fi field eld bo boun undar daries ies in or order der to min minimi imize ze fly inf infest estatio ations. ns. On the basis of his survey of crop-field borders, Dambach (1948) concluded that the more nearly border vegetation is related, botanically,, to the adjacent crop plants, the greater is the danger of its servcally ing in g as a po pote tent ntia iall so sour urce ce of in infe fest stat atio ion n by in inju juri riou ouss in inse sects cts.. Th Thus us,, le less ss crop cr op-p -pes estt ri risk sk is in invo volv lved ed in th thee us usee of wo wood ody y bo bord rder er veg eget etat atio ion n in ar ar-eas where the predominant crops are grain, vegetables, and forage plants. In temperate regions, an increasingly common alternative to hedgerows around orchard margins is the single-species planted windbreak. Poplar ( Populus spp.), willow ( Salix spp.), and some conifers aree us ar used ed,, but th thee mo most st wi wide desp spre read ad is al alde derr ( Alnus spp. p.). ). No None ne of th thes esee Alnus sp
windbreak trees provides an important source of phytophagous insects or mites that feed on apple, so they pose no threat to orchardpest management (Solomon, 1981).
FIELD BOUNDARIES AND NATURAL ENEMIES
There is clear evidence that plants outside or around the cultivated field provide important resources to increase the abundan abundance ce and impact of natural enemies. Habitats associated with agricultural fields may provide resources for beneficial arthropods that are unavailable in the crop habitat, such as alternate alt ernate hosts or prey, food and water resources, shelter, favorable favorable microclimates, overwin overwintering tering sites, mates, and refuge from pesticides (Dennis and Fry, 1992). Genera Gen erally lly,, hed hedges ges sup suppo port rt a ric richer her ins insect ect com commun munity ity tha than n adj adjace acent nt crop fields (Lewis, 1965), and the presence of certain hedges can enrich the insect population nearby for approximately a distance to leeward of three to ten times its height and to windward of equal to or twice its height. Since the studies of Dambach (1948), it is known that the shelter provided by edge vegetation is important in encouraging natural enemies. There are still, however, many questions that need further research (Wratten, 1987; Kajak and Lukasiewicz, 1994): 1. To what extent do beneficial insects depend on hedges, ditch bank ba nks, s, ol old d fiel ields ds,, an and d fo fore rest stss fo forr th thei eirr co cont ntin inue ued d ex exis iste tenc ncee in ag ag-ricultural areas, particularly during winter? 2. Do these borders and other abrupt transitions between one ecosystem sys tem and ano anothe therr (ec (ecoto otones nes)) inf influe luence nce the spe specie ciess di dive versi rsity ty and abundance abundan ce of entomophagous insects in adjacent crop fields? 3. Which attributes of the boundary are important for the natural enemies? 4. Can existing natural refuges within boundaries be improved improved or can new refuges be created? Several researchers have shown that vegetation in adjacent areas can provide the alternative food and habitat essential to perpetuate certain natural enemies of pests in crop fields. Many beneficial insects find overwintering quarters in the litter develo developed ped in shrub and
osage (Maclura pomifera) borders of crop fields (Dambach, 1948). Van Emd Emden en (1 (1965 965a) a) and Pol Pollar lard d (1 (1968 968)) sho showed wed tha thatt the pro propor portio tion n of predaceous insects increased with reduced hedgerow hedgerow management. In hi hiss st stud udy y, Po Poll llar ard d (1 (196 968) 8) di divi vide ded d a 4 m hi high gh ha hawt wtho horn rn he hedg dgee ad ad- jacent to a cereal field into six 30 m lengths and removed the bottom flora from three 30 m sections with a paraquat-diquat mixture, a total weed killer. Both sides of the hedge were treated, and the treatment continued for three seasons. This removal of shelter (as well as, of course, herbaceous plants harboring alternative prey) significantly reduced the predatory fauna in the bottom 1.5 m of the hedgerow. Anthocoris nemorum was one predator affected, and so were spiders and th thee ca cara rabi bid d be beetl etles es Bembi Bembidion dion guttula and Agonum dorsa dorsale. le. These predators were particularly likely to colonize the adjacent fields. In more detailed studies with Agonum dorsale, Pollard (1968) was able to show that this predator overwintered in the hedge bottom. By dissecting females and examining their ovaries, he showed that A. dorsale captured prey 54 m into pea and wheat crops, as it probably invaded the fields from the hedge. Severall other studies indicate that the abundan Severa abundance ce and diversity of entomophagous entomophago us insects within a field are related closely to the nature of the surrounding vegetation. In northern Florida, predator density and diversity were greater in maize plots surrounded by annually burned pinelands and complex weedy fields than in those plots surrounded by sorghum and soybean fields (Altieri and Whitcomb, 1980) 19 80).. Nei Neighb ghbori oring ng ve veget getatio ation n can als also o det determ ermine ine the rat rates es of col colon oniization zat ion and pop popula ulatio tion n gra gradie dients nts of nat natura urall ene enemie miess wit within hin a par partic ticula ularr crop field (Altieri and Todd, 1981). A study of the dispersal of carabi car abid d and sta staph phyli ylinid nid adu adults lts int into o cer cereal eal fi field eldss fro from m fi field eld bou bound ndari aries es (Coombes and Sotherton, 1986) showed that beetles could be recovered up to 200 m into the fields and that two patterns of dispersal could be distinguished. One pattern typical of the carabids A. dorsale and Tachyporus hypnorum showed decreasing numbers, with progres gr essi siv vely la late terr pe peak akss al alon ong g a tr tran anse sect ct fr from om th thee bo boun unda dary ry to th thee ce cent nter er of the field (Figure 7.2). Similarly, in southern Georgia, predators were more abundant in the edges of soybean fields adjacent to pea fields and weedy tracts than in soybean edges adjacent to vegetation-free fields (Altieri and Todd, 1981). In Georgia, predator numbers in soybean rows declined sharply the farther the rows were from a weedy ditch bank and a for-
FIGURE 7.2. Density of the ground beetle (Demetrias atricapillus) at atricapillus) at different distances from a field boundary (after Wratten, 1987).
est edge adjacent to the field (Figure 7.3). Van Emden (1965a) found that syrphid predators of the cabbage aphid, Bre Brevicoryne vicoryne brassicae, were distributed in crop edges near flowerin flowering g weeds; predation kept pest densities at the crop borders to below half the level found at the center of the crop area. In Hawaii the presence of nectar-sou nectar-source rce plants in sugar cane field margin mar ginss allo allowed wed pop popula ulatio tion n le leve vels ls to ris risee and inc increa reased sed the ef effi ficie cienc ncy y of the sugar cane weevil parasite Lixophaga sphenophori (Topham and Beardsley, 1975). The authors suggest that the effective range of thee pa th para rasi site te wi with thin in ca cane ne fie ield ldss is li limi mite ted d to ab abou outt 45 to 60 in inch ches es fr from om nectar sources present in the field margins. Continuous herbicidal elimin eli minatio ation n of fi field eld-ma -marg rgin in nec nectar tar-so -sourc urcee pla plants nts had a det detrim riment ental al ef ef-fect on populations of Lixophaga and, therefore, led to a decrease in the efficiency of the parasite as a biocontrol agent of the weevil (Table 7.1).
FIGURE 7.3. Abundance FIGURE Abundance gradient gradient of insect predat predators ors across a soyb soybean ean field adjacent to two different types of plant communities in Georgia (after Altier i and Todd, 1981).
Maier (1981) observed higher parasitization rates of apple maggot (Rhagoletis pomonella) by braconids in northern Connecticut apple and hawthorn orchards where plants such as blueberry ( Vaccinium spp.), dogwood ( Cornus spp.), and winterberry (Ilex cillata) commonly grew nearby. These plants support populations of several frugivorous frugiv orous tephritids that serve as alternate hosts to the braconids. In Norway’s apple orchards, the numbers of the key pest Argyresthia conjuge conjugella lla is largely dependent on the amount of available food, i.e., the number of berries of the wild shrub Sorbus aucuparia that develop each year. Since only one larva develops in a single berry, the number of Argyresthia can never be higher than the total numb nu mber er of be berr rrie ies. s. Th Thus us,, in ye year arss wh when en Sorbus has no be berrri ries es in a ce cerrtain ta in ar area ea,, no Argyresthia lar larva vaee are pro produc duced, ed, and con conseq sequen uently tly the there re will be no parasites (the braconid Microgaster politus) in the area. Entomologists have suggested plantings of Sorbus which produce an abund ab undant ant and re regul gular ar cro crop p ev every ery yea yearr. Argyresthia alw always ays fi finds nds eno enoug ugh h food fo od to ma main inta tain in it itss po popu pula lati tion on at a re reas ason onab ably ly hi high gh le leve vel. l. Un Unde derr su such ch conditions, Microgaster and other natural enemies will also operate and reproduce sufficiently every year to regulate their host below the level where Argyresthia is forced to emigrate. Hence, the apple avoids infestation (Edland, 1995). In a two-year study, Landis and Haas (1992) found higher parasitism of O. nubilalis larvae by the parasitoid Eriborus terebrans (Hymenoptera: Ichneumonidae) Ichneumonidae) in edges of cornfields than in field interiors; and in the second year, they observed significantly higher parasitism by E. terebrans in edges of corn plantings adjacent to
TABLE 7.1. Parasitization of R. obscuru obscurus s Grubs by Lixopha Lixophaga ga spheno sphenophori phori Before and After Application of Herbicide to Field Margin Percent of recovered grubs parasitized Time of Test Before herbicide application
Immediately after herbicide application
35 days after herbicide application
Distance Infield from Margin
HerbicideTreated Field
Untreated Control Field
margin
95.0
100.0
50 ft.
100.0
85.6
100 ft.
100.0
87.5
150 ft.
94.4
100.0
200 ft.
26.4
100.0
TOTAL
80.7
95.0
margin
79.0
89.0
50 ft.
86.5
71.4
100 ft.
83.4
89.0
150 ft.
92.8
70.0
200 ft.
23.6
100.0
TOTAL
76.5
83.6
margin
23.6
60.0
50 ft.
5.3
71.4
100 ft.
0.0
62.5
150 ft.
16.7
50.0
200 ft.
5.6
73.0
TOTAL
10.0
64.9
Source: after Source: after Topham and Beardsley, 1975.
woo oode ded d ar area eass th than an in th thos osee ar area eass ne near ar no nonw nwoo oode ded d ar area eass or in fie ield ld in in-teriors. In Germany, parasitism of rape pollen beetle was about 50 percent at the edge of all fields. Toward the center of the field it dropped significantly significantly to 20 percent (Thies and Tscharntke, 1999). In their studies in Illinois, Mayse and Price (1978) found that the mean number of both herbivore and predator/parasitoid species per habitat spac sp acee in so soyb ybea ean n fie ield ldss was hi high gher er at th thee ed edge ge th than an in th thee ce cent nter er of th thee field. The presence of relatively complex vegetation in the crop bor ders was an important factor in such trends.
Carabids have been studied by many researchers in central Europe who concluded that important ground beetle predators of crop pests exploit the shelter of hedges during field cultivation and during winter. Pollard (1968) argues that in many sparsely forested countries much of the carabid crop fauna is of woodland origin and that many species are now dependent to a large extent on hedges for their con tinued existence in agricultural areas. Agonum dorsale is a carabid that exhibits seasonal migration between field and edge. Sotherton (1984) observed that different predatory Carabidae and Staphylinidae preferred different field-boundary types but that hedge banks or shelterbelts were more attractive to most polyphagous predators than grass banks or grass strips. Wallin (1985), also investigating the spatial and temporal distribution of Carabidae in cereal fields and adjacent habitats, suggested that certain carabid species seemed to prefer field edges at different times of their lives. The field edges serve as important shelter habitats at certain times of the season and function importantly as overwintering sites for various carabid species. More recently,, Varchola and Dunn (1999) con recently convincingly vincingly showed that both simple and complex roadside vegetation bordering cornfields was important to the carabid communities, especially before canopy closure in cornfields. Such habitats apparently provide carabids with necessary resources and functions, especially overwintering and breeding sites that are unavailable in relatively bare crop fields. In the Rhineland, Thiele (1977) observed that following the grain harvest on August 5, the catches of A. dorsale on the edges rose rapidly during August from the initial initi al minimum values in July. Based on several studies of ground beetle movements and beetle habitat requirements, conceptual models of beetle movement between boundaries and ad jacent crop fields have been proposed (Figure 7.4). Chiverton and Sotherton (1991) studied the effects of excluding herbicides from cereal-crop field edges; they were able to demonstrate that these edges had greater abundance of nontarget arthropods and thus supplied predators with ample prey. All these earlier studies of carabid habitat preferences and relative abundance abundance led to research directed at conserving and enhancing carabid populations in and around annual crop fields. The same applies to other arthropods, and LeSar and Unzicker (1978) have proposed the establishment of grass or legume strips along field margins to enhance the colonization of soybean fields by spiders.
-
FIGURE 7.4. A conceptual model of carabid beetle movement between fields and hedges (after Thiele, 1977).
The proximity of forest edges and hedgerows that serve as hibernation sites has a fundamental effect on the occurrence of coccinellids in agr agricu icultu ltural ral are areas. as. In Cze Czecho choslo slov vaki akia, a, an app apple le or orcha chard rd sur surro round unded ed by deciduous forests had a tenfold higher abundance of Coccinella
quinpuepunctata, because the neighboring forest provided a dor-
mancy manc y si site te fo forr th thee be beet etle, le, wh whic ich h hi hibe bern rnate atess in th thee lit litte terr of fo fore rest st ed edge gess (Hodek, 1973). In England, researchers found higher populations of coccinellids in bean plots surrounded by patches of nettles, as opposed to bean plots surrounded by trees and buildings (Burn, 1987). Severall other authors have reported that the presence of alternate Severa hosts or prey on weeds growing in field margins increases parasitism and/or predation of specific pests within crops. Since the life cycles of many parasitoids and predators are not synchronized with those of their the ir ho hosts sts/pr /prey ey,, som somee nat natura urall ene enemie miess mus mustt rel rely y on alte alterna rnati tive ve sou source rcess to maintain establishment within a community. This may be especially important when the pest species have become scarce in the fie ield ld.. Th Thee wi wide desp spre read ad pl plan anti ting ng of al alde derr ( Alnus spp spp.) .) in sou southe theast ast Eng Eng-land has established a considerable reservoir of the predacious mirid Blepharidopt Bleph aridopterus erus angulatu angulatus, s, a regulator of the mite Panonychus ulmi. On these trees, B. angulatus feeds on aphids and leafhoppers; and when the numbers of these prey decline in August, they move on to nearby orchards, thus controlling spider mite populations (Solomon, 1981). Flowering willows, Salix caprea, support high populations tio ns of the pr preda edacio cious us ant anthoc hocor orids ids Antho Anthocoris coris nemoru nemorum m and A A.. ne nemmoralis in early April. At this time, apple aphids and apple psylla are jus j ustt be begi ginn nnin ing g to ha hatc tch, h, an and d an anth thoc ocor orid idss fr from om wi willo llow w ma may y co colo loni nize ze in response to high numbers of these species. Aveling (1981) showed that A. nemorum was an abundant predator, migrating into hop gardens from spring populations in adjacent trees and hedgerows, especially from nettles, Urtica dioica, infested with nettle aphid. DESIGNING AND MANAGING BORDERS
Klinger (1987) provided margin strips of Sinapsis arvensis and Phacelia tanacetifolia and found that these led to higher densities of polyphagous predators in the strips and in adjacent fields than in wheat plots without strips. Also, syrphid adults occurred at higher densities in the strips than in the field, presumably because the flies foraged on P. tanacetifolia and S. arvensis. The impact of different predatory groups on aphid populations was not quantified in this work, although there was a trend toward lower aphid densities in the field with adjoining strips. Sengonca and Frings (1988) found syrphid adults to be more abundan abundantt in sugar beet plots with P. tanacetitan aceti-
folia margin strips than in sugar beet monocultures. Borders of Phacelia have also been explored in cabbage where syrphid numbers
increased increa sed,, and aph aphid id po popul pulati ations ons dec declin lined. ed. Sp Speig eight ht (19 (1983) 83) cit cites es wo work rk reporting that strips of dill and coriander in eggplant fields led to enhanced predator numbers ( Coleomegilla maculata and Chrysoperla carnea), increased consumption rates of Colorado potato beetle (Leptinotarsa tinotar sa decemlineata) egg masses, and decreased larval survivorship. In many cases, weeds and other natural vegetation around crop fields harbor alternate hosts/prey for natural enemies, providing seasona so nall re reso sour urce cess to br brid idge ge th thee ga gaps ps in th thee lif lifee cy cycl cles es of en ento tomo moph phag agou ouss insects and crop pests. A classic case is that of the egg parasitoid wasp Anagrus epos whose effectiveness in regulating the grape leafhopper, Erythroneur Erythroneura a elegantula, was increased greatly in vineyards near areas invaded by wild blackberry ( Rubus sp.). This plant supports an alternative-host leafhopper (Dikrella cruentata), which breeds in its leaves in winter (Doutt and Nakata, 1973). Recent studies have shown that prune trees planted next to vineyards also allow early-season buildup of Anagrus epos. After surviving the winter on an alternate host, the t he prune leafhopper, leafhopper, Anagrus wasps move into the vineyard in the spring, providing grape leafhopper control up to a month earlier than in vineyards not near prune tree refuges (Murphy, Rosenh Ros enheim eim,, and Gr Grane anett, tt, 19 1996) 96).. Mur Murph phy y and col collea leagu gues es (19 (1998) 98) com com-pleted ple ted a rig rigor orous ous ev evalu aluati ation on of the ef effec fecti tive venes nesss of Fre French nch pru prune ne tre trees es in increasing control of the grape leafhopper. leafhopper. Results from this study indi in dica cate te th that at th ther eree is a co cons nsis iste tent nt an and d si sign gnif ifica icant nt pa patte ttern rn of hi high gher er pa parrasitism in grape vineyards with adjacent prune tree refuges than in vineyards lacking refuges. Researchers now recommend that trees should always be planted upwind from the vineyard but otherwise can be managed as a typical commercial prune orchard; they also suggest planting as many trees as is economically feasible, since the more trees there are, the more productive the refuge is likely to be. By monit monitorin oring g rubi rubidiumdium-labele labeled d Anagrus, Cor Corbet bettt and Ros Rosenh enheim eim (1996) found that Anagrus colonizing the study vineyards from external sources consistently exhibited a distinct spatial pattern: low abundance in the first vine row downwind of French prune trees; a large increase at the third vine row downwind; downwind; and a gradual decline from this peak with increasing distance from the refuge. It is likely that th at a wi wind ndbr brea eak k ef effe fect ct is op oper erat atin ing g in th this is sy syst stem em:: Anagrus emerging
from overwintering habitats external to the vineyard-French prune system syst em are colonizing at a higher-than-average higher-than-average rate immediately downwind win d of the refuge as a result of the turbulence generated by French prun pr unee tr tree ees. s. Fr Fren ench ch pr prun unee tr tree ee re refu fuge gess ar aree th thus us ha havi ving ng tw two o im impa pacts cts on early-season abundance of Anagrus: (1) directly contributing Anagrus that have overwintered in the refuge and (2) increasing the colonization rate by Anagrus having overwintered in habitats external to the French prune-vineyard prune-vineyard system (Figure 7.5). The amount of additional addit ional colonization generated by a windbreak effect of refuges is dependent on the proximity and size of external overwintering overwintering habitats, because must st be di disp sper ersi sing ng in la larrge nu numb mber erss in th thee wi wind nd st stre ream am fo forr a Anagrus mu windbreak effect to cause increased colonization. Thus, refuges that are close to riparian habitats would generate high colonization, whereas refuges that are many kilometers away might generate no noticeable windbreak-induced windbreak-in duced colonization.
FIGURE FIGUR E 7.5 7.5.. Hyp Hypoth othesi esized zed sou source rces s of Anagrus Anagrus col coloniz onizing ing vine vineyar yards ds earl early y in the season. Anagrus Anagrus colonize colonize vineyards from adjacent French prune tree refuges. Anagrus also Anagrus also colonize from external overwintering sites. The windbreak effect generated by prune trees causes increased colonization by external Anagrus immediately downwind of refuges (after Corbett and Rosenheim, 1996).
At a more regional level, level, one of the unique cases exploring the relationship between landscape, vegetation diversity, and insect pests comes from a twenty-year experiment conducted near Waco, Texas, from 1929 to 1949 by the Bureau of Entomology and Plant Quarantine in cooperation with the Soil Conservation Service. Entomologists measured the effects of new farming and soil conservation methods on populations of beneficial and pest insects in cotton (De Loach, 1970). About 600 acres of upland farmland was divided into adjacent areas of 300 acres each, designated Y and W (Figure 7.6).. In both, 7.6) both, the old farming farming practic practices es were conti continued nued for for fou fourr year years, s, through 1942, while pretreatment counts were made. New conservation methods then t hen were begun in Y, Y, while the old ol d practices practice s were continued in W, where cotton occupied the largest acreage, followed closely by corn. There was also a substantial acreage of oats and pas-
FIGURE 7.6. Diversification of agricultural landscapes in Texas for soil conservation purposes that led to improved control of insect pests of cotton (after De Loach, 1970).
ture and a little l ittle sorghum; these crops occupied nearly 100 percent of the land area. By the new practices, several acres of clover were plan pl anted ted al alon onee or ove vers rsee eede ded d in th thee oa oats ts,, so some me gr gras asss ar area eass we were re ad adde ded, d, and the land was terraced. Insecticides were not used in i n any fields in the two areas during the experiment, so that the effect of cultural methods alone could be measured. The new conservation practices resulted in a reduction in numbers of cotton pest insects and a reduction in the percentage of damaged squares and bolls. No attempt was made at elucidating the mechanisms explaining such reductions, although it is assumed that natural enemies were highly favored favored by the new landscape designs, thus resulting in enhanced pest mortality. More recent research conducted in the mid-1990s suggests that within agricultural fields the diversity of parasitoids and the intensity of parasitism is generally greater at field edges where crops are adjacent to later-successional plant communities than in field interiors or along field edges with crop-crop or crop-early successional interfaces. Also, on a larger spatial scale, the diversity of parasitoids and the intensity of parasitism should be greater in agricultural landscapes sca pes emb embedd edded ed in a mat matrix rix of lat later er-su -succe ccessi ssiona onall pla plant nt com commun munitie itiess (old fields, hedgerows, woodlots) than in simple agricultural landscapes composed primarily of field crops. Therefore, on both the small sma ll wit within hin-f -fiel ield d and the lar large ge bet betwee ween-f n-field ield sca scales les,, a hig highly hly di dive verse rse landsc lan dscape ape str struct ucture ure may pr prov ovide ide the gr greate eatest st pot potent ential ial fo forr the bio biolog logiical suppression of pests by their natural enemies. Marino and Landis (1996) compared maize fields of small size embedded in a landscape of abundant hedgerows hedgerows and woodlots to a simple landscape of large-size fields embedded in a landscape with few hedgerows and woodlots to determine the influence of overall landscape diversity on parasitoid communities of the armyworm (Pseudaletia unipunctata). They found that parasitism was greatest in the complex landscape. Meteorus wasps were the most abundant parasitoids attacking armyworm, and the presence of alternative hosts in the complex landscape explained its increased abundance. In a st stud udy y in no nort rthe hern rn Ge Germ rman any y, Th Thies ies an and d Ts Tsch char arnt ntke ke (1 (199 999) 9) fo foun und d th that at structural simplicity in agricultural landscapes was correlated with large amounts of plant damage caused by the rape pollen beetle (Meligethes aeneus) and small amounts of larval mortality caused by three ichneumonid parasitoids.
These studies give credence to emerging approaches that suggest the importance of the landscape as a level of organization of processes such as dispersion of plants, arthropod movement, and nutrient flow (Paoletti, Stinner, and Lorenzoni, 1989). Since agriculture is a major force shaping landscape structure and dynamics, it is useful to examine the relationships between arthropods and vegetation patterns at the landscape ecological level, especially in regions dominated by large-scale monocultures, which represent highly fragmented landscapes. Much concern has been expressed regarding regarding the effects of these fragmented landscapes on the survival of a variety of beneficial entomofauna. entomofauna. As habitats become more fragmented, a variety of species that require relatively large areas of suitable habitat have a more difficult time surviving in the increasingly smaller fragments; populations may become extinct in these fragments, and they may never return. Reversing these effects through agricultural diversification is a key challenge.
CASE STUDY 1: EXCHANGE OF ARTHROPODS AT THE INTERFACE OF APPLE ORCHARDS AND ADJACENT WOODLANDS
In northern California, apple orchards are distributed within a matrix of natural vegetation that provides abundant opportunities to study arthropod colonization and interhabitat exchanges of arthropods. Altieri and Schmidt (1986a) conducted comparative comparative studies on the ecology of arthropod communities in four ecologically different dry-f dr y-farm armed ed app apple le orc orchar hards: ds: (1 (1)) an “ab “aband andon oned” ed” orc orchar hard d no nott man manage aged d or disturbed for twenty-five years, (2) two “organic” (not sprayed with wit h syn synthe thetic tic pes pestici ticides des)) orc orchar hards, ds, one cle clean an cul culti tiva vated ted and the oth other er with a mixed grass-legume cover crop, and (3) a “commercially” managed orchard (clean cultivated and subjected to chemical fertilizer and pesticide treatments). These ecologically different orchards constitute a “cultural evolution continuum.” In the abandoned orchard, stable relationships between arthropods and the local vegetation have developed, probably because they are not disturbed. In the commercial orchard, high-energy inputs are substituted for some plant-insect interactions. The organic orchards combine characteris-
tics of both systems. All orchards have at least one bordering edge with multilayered communities of wild vegetation. Colonization of Orchards
The magnitude of the exchange of predators and parasitic Hymenoptera between the orchards and the t he wild vegetation edges are shown in Figure 7.7. Except for Syrphidae, considerably more individuals moved from edge to orchard in unsprayed (organic) (organic) than in sprayed orchards. Little exchange seemed to take place between abandoned orchards and woodlands. Figure 7.8 shows the temporal dynamics of colonization from the hedges to the t he sprayed and organic orchards by Coccinellidae. D-Vac D-Vac samples taken from the shrub and herb layer of the edges revealed that edges of the organic orchards supported considerably more natural enemies than the t he edge of the sprayed orchard. In th thee ea earl rly y se seas ason on,, co cons nsid ider erab ably ly mo more re ap aphi hids ds in inv vad aded ed th thee sp spra raye yed d or or-chards than the organic and abandoned orchards.
FIGURE 7.7. Seasonal mean number of natural enemies (Hym = parasitic Hymenoptera; Syr = Syrphidae adults; Pip = Pipinculidae; Raph = Raphididae; Chry = Chrysopidae; Hem = Hemiptera; Cocc = Coccinellidae; and Can = Cantharidae) caught in malaise traps in the interface interface between apple orchards (A = abandoned, O = organic, and S = sprayed) and wild vegetation edges in northern California (after Altieri and Schmidt, 1986a).
P A R T E S I A L A M / E A D I L L E N I C C O C . . O N
May
June
July
Aug
Sept
SAMPLING DATES organic sprayed
FIGURE 7.8. Mean number of adult Coccinellidae caught in malaise traps placed in the interface of woodlands and sprayed and organic apple orchards (after Altieri and Schmidt, 1986a).
Ground-Dwelling Predators
The ant species collected in the edges of the managed orchards also al so we were re fo foun und d in th thee bo bord rder erss an and d ce cent nter ers, s, su sugg gges esti ting ng th that at pa part rt of th thee ant communities living in the wild margins colonized the planted orchards. Comparisons of mean ant abundance between edge, border, and center revealed major within-site differences. Throughout the season, more ants were caught in the borders or vegetation-free centers (Figure 7.9). Abundance gradually declined from the edge to t he center of the orchards. In the abandoned orchard, however, however, ant species ci es co comp mpar arit itio ion n an and d pa patt tter erns ns of an antt ab abun unda danc ncee we were re mo more re or le less ss un uniiform from edge to center from mid-May on. These trends can be explained by the structural similarities of the centers and edges. The vegetational complexity of the center of the abandoned system was similar to the edge, unlike the centers of the clean-cultiv clean-cultivated ated systems that lacked the diversity of grasses and herbs characteristic of the edges.
P A R T L L A F TI P R E P S T N A F O R E B M U N
WEEKLY SAMPLING DATES adjacent vegetation edge orchard border row orchard center row
FIGURE FIGU RE 7. 7.9.Mean 9.Mean nu numb mber er of ant ants s ca caug ught ht in pit pitfa fall ll tr trap aps s pl plac aced ed in th the e ce cent nter er an and d border tree rows of an organic apple orchard and in the adjacent vegetation edge (after Altieri and Schmidt, 1986a).
More ants were found in the wild vegetation edges of the managed orchards than in the edges of the abandoned orchards. Ants were most abundant in the abandoned edge early in the season, thereafter ant catches declined but then stabilized. Conversely, ant catches in the sprayed orchard’s edge gradually increased and surpassed the abundance levels of ants in the organic edge from July on. Spider catches were significantly higher in the edges of the abandoned orchards than in the organic or sprayed orchard edges. From May 15 through July 15, more spiders were caught in the organic edge than in the sprayed edge. Carabidae behaved differently than ants or spiders. Pitfall catches were higher in the center and border rows of apple trees than in the edges. This seems to be a normal pattern, as many Carabidae (i.e., Agonum dorsale) exhibit seasonal migration between field and edge.
Predation Pressure
Predation pressure (mainly by ants), as measured by the removal rates of potato tuberworm (Phthorimaea operculella) larvae from cardboard sheets placed on the orchard floor, was highest in the center of the t he abandoned orchard, followed by the organic orchards and thee sp th spra raye yed d or orch char ard d (T (Tab able le 7. 7.2) 2).. Pr Pred edat atio ion n wa wass gr grea eater ter in th thee ce cent nter er of the organic-cover orchard than in the center of the clean-cultivated one. On average, significantly more larvae were removed from the edges edg es of the clea clean-c n-cult ultiv ivated ated or organ ganic ic and spr spraye ayed d orc orchar hards ds tha than n in the TABLE 7.2. Removal of Potato Tuberworm Larvae, Phthorimaea operculella (Placed on the Ground), and of Mediterranean Flour Moth Eggs, Anagasta kuehniella (P kuehniella (Plac laced ed on th the e Tre rees es), ), by Pr Pred edat ator ors s in th the e Ce Cent nter ers s an and d Ed Edge ges s of Vari ari-ous Northern California Apple Orchards Orchard System
% Eggs Removed1
% Larvae Removed2
Sprayed Center
21.0
5.2
17.0
6.2
Border
26.0
8.0
Edge
33.0
10.2
32.7
7.6
Center
25.0
8.0
61.5
12.3
Border
34.1
12.1
Edge
43.1
10.2
70.5
13.9
Center
38.0
9.2
86.9
14.5
Border
42.0
10.7
Edge
36.0
4.2
86.4
15.2
Organic
Abandoned
Source: after Source: after Altieri and Schmidt, 1986a. 1Predation
pressure on the eggs was estimated on four occasions by hanging twenty twe nty-fi -five ve 8.5 11. 11.0 0 cm paper cards cards (with (with fifty fifty moth moth eggs each) each) from from the branches of each of five trees in the center, border, and edge of each orchard. 2Means of three sampling dates. Larvae removal data were obtained ob tained by placing on the ground ground forty forty 22 22 cm cardboa cardboard rd sheets sheets (twen (twenty ty in the center center and and twent tw enty y on th the e ed edge ge)) ea each ch co cont ntai ainin ning g tw twen enty ty glu glued ed fo four urth th ins insta tarr lar larvae vae.. Pr Pred edat ation ion pressure was measured by determining removal of larvae in an eighteen-hour period.
centerss of the center these se sam samee orc orchar hards. ds. Rem Remov oval al rat rates, es, ho howe weve verr, wer weree sim simila ilarr in the wild vegetation edges and centers of the organic-cover and abandoned orchards. Cultivation and insecticide application probably disrupted ant communities in the centers of the clean-cultivated and sprayed orchards, confining ant foraging to the t he edges. On the trees, predation of artificially placed Anagasta kuehniella eggs was consistently higher in the edges than in the border or center trees of the managed orchards. No differences in predation were observed between edge and center in the abandoned system. A clear gradient in predation pressure was observed in the centers of the orchards, declining from abandoned to sprayed.
MANIPULATING CROP-FIELD BORDER VEGETATION
It is clear then from these studies that the t he potential for pest regulation by certain natural enemies may be related to our ability to exercise some degree of control over the habitats surrounding crop fields. Could the herbivore-predator assemblages of an agroecosystem be manipu man ipulate lated d by cha chang nging ing the ve veget getatio ational nal com compo positi sition on and oth other er fea fea-tures of surrounding edges and habitats? Some work conducted in England in recent years is providin providing g some key information for management of field boundaries to increase natural enemy abundance and efficiency efficie ncy.. A strategy implemented by the UK Game Conservancy Trust Trust and resear res earche chers rs of Sou Southa thampt mpton on Un Univ ivers ersity ity con consis sists ts of ex exper perime imenta ntally lly reducing field size by creating new, nonwoody, predator overwintering refuges. Raised banks (created by careful plowing) have have been sown with grasses such as Lolium, Dactylis, Agrostis, and Holcus. Simply switching off the herbicide spray during normal cereal operations (plus the creation of adjacent sterile strips) will effectively provide new refuges hundreds of meters from existing orthodox field boundaries. Small (about 10 ha) and large (about 40 ha) fields have been used, and “peninsular” boundaries have been created which will reach the respective field center. center. This type of bank can also be sown with pollen-bearing and nectar-bearing plants in order to attract Hymenoptera and Syriphidae.
Workers at the Game Conservancy Trust have designed simple “minih “mi nihedg edgero erows, ws,”” wh which ich can be acc accomm ommoda odated ted at fi field eld mar margin ginss wit within hin the width occupied by simple post-and-wire-strand fencing. These minihedgerows are nothing more complicated than a raised narrow strip, planted with suitable vegetation. The grasses Dactylis glomerata and Holcus lanatus seem particularly suitable plants for the beetles, but there is also scope for sowing flowering plants as adult food sources for other natural enemies. A further development development is that such “predator conservation strips,” running parallel with the crop rows (these banks are 0.4 m high, 1.5 m wide, and 300-400 m long and across the centers of the field), can be drilled at intervals across the crop to enhance natural enemy populations over the whole field area (Thomas and Wratten, 1990). These strips can be created afresh each ea ch ye year ar if th thee fa farm rmer er wi wish shes es to ch chan ange ge th thee di dire rect ctio ion n of pl plo owi wing ng.. Re Re-search showed that high predator densities (nearly 1,500 predators/m2) could be achieved in two years. An economic evaluation has shown that gains from the increased predator efficiency efficiency could more than repay the costs of labor and the expected 0.5 percent loss in crop yield, which together are less than any aphid-induced losses of about 5 percent or spraying costs equivalent to about 2.5 percent of crop yield. Research in Sweden by Chiverton (1989) showed that increased densities of cereal aphid predators (caradids such as Bembidion lampros, rove beetles of the genus Tachyporus, and several species of linyphiid spiders) were found overwintering on grassy banks, starting as soon as one year after bank establishment. As suggested by studies in northern Florida, the species composition of weed communities of uncultivated land surrounding crop fields can be modified by plowing the land at different times of the year (Altieri and Whitcomb, 1979a). It was found that by increasing certain weeds experimentally, the diversity and numbers of herbivorous and predaceous insects associated with these weeds increased. Coccinellids were most abundant in plots plowed in December because these treatments enhanced the abundance of goldenrod ( Solidago sp.) and Mexican tea (Chenopodium ambrosioides), which in turn provided suitable food (aphids and other herbivores) and habitat for the coccinellids and other predators (Figure 7.10). Hence, in t his case, the manipulation of a particular predator colonizing cornfields cornfields depended upon the type and abundance of vegetation present around the fields as determined by the time of plowing.
FIGURE FIGUR E 7.1 7.10. 0. Food web for formed med by the maj major or art arthro hropod pod spec species ies ass associ ociate ated d wit with h gol goldenr denrod od (Solida (Solidago go altiss altissima) ima) in in nor north th Florida (after Altieri and Whitcomb, 1979).
CASE STUDY 2: BIOLOGICAL CORRIDORS IN VINEYARDS
As mentioned, research by Kido et al. (1981) established that French prunes (Prunus domestica) adjacent to vineyards could also serve as overwintering sites for the parasitic wasp A. epos, and Murphy, Rosenheim, and Granett (1996) detected higher grape leafhoppe ho pperr par parasi asitism tism in gra grape pe vin viney eyard ardss wit with h adj adjace acent nt pru prune ne tre treee ref refuge ugess than in vineyards lacking refuges. Corbett and Rosenheim (1996), however, determined that the effect of prune refuges was limited to few vine rows downwind, and A. epos exhibited a gradual decline in vineyards with increasing distance from the refuge. This finding poses an important limitation to the use of prune trees, as the colonization of grapes by A. epos is limited to field borders leaving the central rows of the vineyard void of biological-control protection. To overcome this limitation, Nicholls, Parrella, and Altieri (2000) tested whether an established vegetational corridor enhanced movement of beneficial insects beyond the “normal area of influence” of adjacent habitats or refuges. The study was conducted in northern Califo Cal iforni rniaa in 199 1996 6 and 199 1997 7 and in invo volv lved ed tw two o adj adjace acent nt vin viney eyard ardss sur sur-rounded on the north side by riparian forest vegetation vegetation;; the main difference between the two vineyards is that vineyard A was penetrated and dissected by a five-meter-wide and 300-meter-long vegetational corridor composed of sixty-five different species of flowering plants. Vineyard B had no corridor. In both years in vineyard A, adult leafhoppers exhibited a clear dens de nsit ity y gr grad adie ient nt re reac achi hing ng lo lowe west st nu numb mber erss in vi vine ne ro rows ws ne near ar th thee co corr rriidor and forest and increasing in numbers toward the center of the field, away from the adjacent vegetation. The highest concentration of leafhoppers occurred after the first twenty to twenty-five rows (30 (3 0 to 40 m) do down wnwi wind nd fr from om th thee co corr rrid idor or.. Su Such ch a gr grad adie ient nt was no nott ap ap-parent in vineyard B, where the lack of the corridor resulted in a uniform dispersal pattern of leafhoppers (Figures 7.11 and 7.12). Nymphal populations behaved similarly, reaching the highest numbers in the center rows of block A in both years. Generalist predators in the families Coccinellidae, Chrysopidae, Nabidae, and Syrphidae exhibited a density gradient in vineyard A, indicating that the abundance and spatial distribution of these insects was influenced by the presence of the corridor which channeled dis-
FIGURE 7.11. Seasonal patterns (numbers per yellow sticky trap) of adult leafhopper E. elegant elegantula ula in in bl bloc ock k A, as in infl flue uenc nced ed by th the e pr pres esen ence ce of th the e co corr rrid idor or (P < 0.05; Mann-Whitney U-test) (Hopland, California, 1997) (after Nicholls, Parrella, and Altieri, 2001).
persal pers al of th thee in inse sect ctss in into to ad adjac jacen entt vi vine ness (F (Fig igur ures es 7. 7.13 13 an and d 7. 7.14 14). ). Pr Pred ed-ators were more homogeneously (but reaching lower overall abundance) distributed in vineyard B, as no differences in spatial pattern in predator catches was observed between bare edge and central rows. Anagrus Anagr us epos colonized vineyards from the corridor and forest throughoutt the sampling area, exhibiting higher densities in late July throughou and throughout August of both years in the central vineyard rows where leafhoppers were most abundant. By following the abundance patterns of leafhoppers, the Anagrus wasp did not display the distributional response exhibited by predators. For this reason, these researchers concluded that predator enhancement near the vegetational corridor explained the lower populations of leafhoppers and thrips i n thee fir th irst st tw twen enty ty-f -fiv ivee ro rows ws.. Su Such ch su succ cces essf sful ul im impa pact ct of pr pred edat ator orss ca can n be assumed because fewer adults and nymphs of leafhoppers and thrips
Bare edge Center
Block B 1997
Near Forest
p
800 700
ar t/ r e b
sr m p
600
e u p
500
a
hf
400
M
L
e
300 200
n o n a e
100 0
Near Forest Center y n a u M - J - 5 0 2 1
n u J - 6 2
Date
l u J - 9
l u J - 4 2
l u J - 1 3
Bare edge g u A 8
g u A 4 1
g u A 1 2
g u A 8 2
FIGURE 7.12. Seasonal patterns (numbers per yellow sticky trap) of adult leafhopper E. elegant elegantula ula in in block B without the presence of the corridor but with an adjacent forest (P (P < < 0.05; Mann-Whitney U-test) U-test) (Hopland, California, California, 1997) (after Nicholls, Parrella, and Altieri, 2001).
were caught near the corridor than in the middle of the vineyards. Overall abundance of predators was higher in vineyard A than B throughout the season (Figure 7.15). The corridor provided a constant supply of alternative food for predators, effectively decoupling predators from a strict dependence on grape herbivores and avoiding a delayed colonization of the vineyard. This complex of predators continuously circulated into the vineyard interstices, establishing a set of trophic interactions leading to lower numbers of leafhoppers and thrips in the border rows of the vineyard. Findings from this study also suggest that the creation of corridors across vineyards can serve as a key strategy to allow natural enemies emerging from riparian forests to disperse over large areas of other wisee mon wis monocu ocultu lture re sys system tems. s. Su Such ch cor corrid ridors ors sho should uld be com compo posed sed of locally adapted plant species exhibiting sequential flowering periods, which attract and harbor an abundant diversity of predators and
FIGURE FIGUR E 7.1 7.13. 3. Sea Season sonal al pat pattern terns s of pre predat dator or cat catche ches s (nu (numbe mbers rs per ye yello llow w sti stick cky y trap) in block A, as influenced by the presence of forest edge and the corridor (P P < < 0.0 0.05; 5; Man Mann-W n-Whit hitney ney U-t U-test est)) (Ho (Hopla pland, nd, Cal Calif ifornia ornia,, 1997 1997)) (af (after ter Nic Nichol holls, ls, Pa Parrrella, and Altieri, 2001).
parasitoids and increase biodiv biodiversity ersity.. Thus, these t hese corridors or strips, which may link various crop fields and riparian forest remnants, can create a network that would allow many species of beneficial insects to disperse throughout whole agricultural regions transcending farm boundaries (Baudry, 1984).
CASE STUDY 3: STRIP MANAGEMENT TO AUGMENT PREDATORS
As a way to enhance predator abundance in cereal fields, researchers in Switzerland introduced vegetation edges as successional strips into the field. One 8 ha winter cereal field was subdivided by five wide-strips leaving cereal spaces of 12, 24, and 36 m between the strips (Lys and Nentwig, 1992). Significantly higher recapture rates, indicating higher predator activity, were found in the strip-managed
FIGURE FIGUR E 7.1 7.14. 4. Sea Season sonal al pat pattern terns s of pre predat dator or cat catche ches s (nu (numbe mbers rs per ye yello llow w sti stick cky y trap) in block B without the corridor but with an adjacent forest (P ( P < < 0.05; 0.05; MannWhitney U-test) (Hopland, California, 1997) (after Nicholls, Parrella, and Altieri, 2001).
area than in the control area, especially for carabid beetles such as Poecilus cupreus, Carabus granulatus, and Pterostic Pterostichus hus melanarius. Several observations led to the conclusion that this higher activity was generally due to a prolongation of the reproductive reproductive period in the t he strip-managed area. Besides the marked increase in activity and density, a large increase in the diversity of ground beetle species was observed. The most marked increase in number of species was found in the first year ye ar.. Th Thee ve vege geta tatio tion n st stru ruct ctur uree of th thee ce cere real al fi fiel eld d wa wass en enri rich ched ed fo foll llo owing use of weed strips. After three years of research the authors conclud cl uded ed th that at we weed ed st stri rips ps of offe ferr no nott on only ly hi high gher er fo food od av avai aila labi bili lity ty but al also so more suita suitable ble ov overwin erwinterin tering g sites. In addition, addition, these weed strips strips offer refuges during field disturbance or during unfavorable unfavorable climatic conditions, such as droughts. Weed strips increase the chance of survival of many carabid species in arable ecosystems, thus counteracting the faunal impoverishment trends promoted by monocultures. Nentwig
8 7 6 5 4 3 2 1 0
FIGURE 7.15. Comparison of abundance of generalist generalist predators (numbers per yellow sticky trap) between block A (with a corridor) and block B (without a corridor) (P (P < < 0.05, Wilcoxon’s signed rank test) (Hopland, California, 1996) (after Nicholls, Parrella, and Altieri, 2001).
(1998) (199 8) fo foun und d si simi mila larr ef effe fects cts wi with th 3- to 99-mm-wi wide de so sown wn we weed ed st stri rips ps di di-viding large fields into small parts so that t hat the distance between strips does not exceed 50 to 100 m. A favorite plant to be used as strips with wi thin in or ar arou ound nd fiel ields ds is Phaceli (Hollan lland d and Tho Thomas mas,, Phacelia a tanaceti tanacetifolia folia (Ho 1996). In reviewing these studies Corbett and Plant (1993) argued for the need to develop a mechanistic framework to evaluate and predict the respo res ponse nse of nat natura urall ene enemie miess to ve veget getati ationa onall arr arrang angeme ements nts in agr agroec oecoosystems. Using a hypothetical field with 10 m wide strips interplanted at 100 m intervals (Figure 7.16), they assumed that these strips are used solely as an overwintering refuge by three natural enemy species: (1) a predatory mite having very low mobility (a diffusion coefffic ef icie ient nt of 1 m2 /da /day); y); (2) a pre predat dator ory y coc coccin cinelli ellid d bee beetle tle ha havin ving g mode moderrate mobility (10 m2 /day); and (3) a highly mobile parasitoid (100 m2 /day). Once the crop has germinated, the strips do not provide resources in any greater abundance than the crop, nor do they provide a
FIGURE 7.16. Diagram of hypothetical diversified agroecosystem. Interplanted strips are placed 100 m apart within a crop. The model predicts natural enemy abundance along a transect through the field (after Corbett, 1998).
more favorable favorable physical habitat. The mobility (i.e., the probability of making a move in a given time period) is therefore the same in the strips as it is in the crop. Natural enemies overwinter overwinter in the strips at a density of ten individuals per square meter. The spatial patterns in abundance predicted by the model for these three hypothetical natural enemies are illustrated in Figure 7.17. The model predicts that the natural enemies will spread from the strips, resu re sulti lting ng in hi high gher er ab abun unda danc ncee in th thee cr crop op th than an wo woul uld d ha have ve oc occu curr rred ed in a crop monoculture. The distance to which they are enhanced varies substantially, however. For the predatory mite, enhancement is confined to the region immediately adjacent to interplanted strips, producing a steep gradient in density with increasing distance. The highly mobile parasitoid, on the other hand, is enhanced throughout throughout the crop—there is no spatial pattern to suggest that strips influenced abundance. abundan ce. As a result, natural enemies with low mobility exhibit no enhancement beyond 20 m from strips, while more mobile natural enemies are enhanced fourfold. Corbett and Plant (1993) proposed a second scenario using the same field and natural enemies. In this scenario, however, however, the interplanted vegetational zones are not overwin overwintering tering refuges: natural enemies must colonize the agroecosystem from external sources. The
FIGURE 7.17. Spatial patterns predicted by model for field with interplanted strips that serve solely as an overwintering refuge. Peak abundance occurs at interplanted strips. Patterns are shown for for hypothetical natural enemies of three different mobilities: low mobility (1 (1 m2 /day), moderate mobility (10 m 2 /day), high mobility (100 m2 /day) (after Corbett, 1998).
strips do, however, provide more resources than the crop. Therefore, thee pr th prob obab abil ilit ity y of ma maki king ng a mo mov ve in a gi giv ven ti time me pe peri riod od is lo lowe werr in th thee strips than in the crop. The resources in interplantings are assumed to be “substitutable,” to some degree, for resources that occur in the crop. They could be either (1) alternate prey or hosts (“supplementary” tar y” res resou ource rces) s) or (2) flo floral ral res resour ources ces tha thatt are an imp imperf erfect ect sub substit stitute ute for the preferred host but that benefit the natural enemy when available (“complementary” resources). The model predicts that natural enemy mobility would dramatically affect the observed enhancement due to increased diversity (Fig (F igur uree 7. 7.18 18). ). Na Natu tura rall en enem emie iess th that at ar aree hi high ghly ly mo mobi bile le wo woul uld d sh sho ow li litttle enhancement when plots are 50 m wide because predators are dispersing among all plots in the experimental field. The observed enhancement increases with plot size since as strips are farther apart their effect becomes more detectable. However, even plots 200 m in size do not detect the enhancement that would occur in a diver diversified, sified, commercial-scale field. For less mobile natural enemies, enhancement me nt is ob obse serv rved ed wh when en pl plot otss ar aree sm smal all, l, bu butt th thee ob obse serv rved ed en enha hanc ncem emen entt
FIGURE 7.18. Effect of mobility on the abundance of natural enemies on crop vegetation in a diversified agroecosystem. “Relative Abundance” is the ratio of natural enemy abundance predicted for the diversified system to that predicted for a crop monoculture. Relative abundance is calculated only for crop vegetation more than 20 m from interplantings. interplantings. Effect Effect of diver diversifica sification tion is shown for three different situations: where the interplantings act solely as an overwintering overwintering refuge; where they provide additional food food resources but no overwintering overwintering sites; and where interplantings provide both (after Corbett and Plant, 1993).
decreases with increasing plot size. This is because such predators are enhanced only in the area adjoining strips. The model also predicts that the abundan abundance ce of the three natural enemies is higher in the interplanted strips than in the crop vegetation. This accumulation of natural enemies in the strips is due to the lower tendency for movement movement there and results in the strips acting as a sink for the natural enemies. The parasitoid exhibits a spatially uniform density in the crop vegetation and the greatest accumulation in the strips. This sink effect results in abundance on the crop that is 60 percent of what it would be in an undiversified field. The other natural enemies exhibit some spatial patterning within the crop and a milder sink effect.
Chapter 8
The Dynamics ofinInsect Pests The Dynamics of Insect Pests Agroforestry Systems in Agroforestry Systems
Agroforestry is an intensive land-management system that combines trees and/or shrubs with crops and/or livestock (Nair, 1993). Many of the benefits of agroforestry are derived from the increased diversity diver sity of these systems compared to corresponding monocultures of crops or trees. Despite the fact that little research has been conducted on pest interactions within agroforestry systems, agroforestry has been recommended to reduce pest outbreaks usually associated with monocultures. Although the effects of various agroforestry designs on pest populations can be of a varied nature (microclimatic, nutr nu trit itio iona nal, l, na natu tura rall en enem emie ies, s, et etc. c.), ), re regu gula latin ting g fa fact ctor orss do no nott ac actt in is isoolation from one another. The few reviews on pest management in agroforestry (Schroth et al., 2000; Rao, Singh, and Day, 2000) expect that high plant diversity protects agroforestry systems to some extent from pest and disease outbreaks. These authors use the same theories advanced by agroecologists to explain lower pest levels in polycultural agroecos ec osys yste tems ms as di disc scus usse sed d in Ch Chap apter ter 3. Th Thes esee au auth thor orss als also o ca caut utio ion n th that at the use of high plant diversity as a strategy to reduce pest and disease risks in agroforestry systems involves considerable technical and economic difficulties. difficulties. Whereas a farmer is free to t o cultivate crops either on separate fields or in association, the choice of the crops themselves (and thus the overall crop diversity of the farm) is strongly influenc flu enced ed by the av availa ailabil bility ity of mar marke kets ts fo forr the res respec pecti tive ve pr produ oducts cts and the needs of the household. household. The select selection ion of timber and fruit trees also al so ha hass to re resp spec ectt lo loca call ma mark rket et co cond nditi ition ons, s, al alth thou ough gh mo more re fr free eedo dom m of choi ch oice ce ma may y ex exis istt fo forr “s “ser ervi vice ce”” tr tree ees, s, fo forr ex exam ampl ple, e, tr tree eess gr gro own fo forr bi bioomass, shade, or wind protection.
THE EFFECTS OF TREES IN AGROFORESTRY ON INSECT PESTS
The del delibe iberat ratee ass associ ociatio ation n of tre trees es wit with h agr agrono onomic mic cro crops ps can res result ult in insect-management benefits due to the structural complexity and permanence of trees and to their modification of microclimates and plant apparency within the production area. Individual plants in annual cropping systems are usually highly synchronized in their phenology and are short lived. The lack of temporal continuity is a problem for natural enemies because prey availability is limited to short periods of time and refugia, and other resources are not available consistently. consistently. The addition of trees of variable phenologies or diverse age structure through staggered planting can provide refuge and a more constant nutritional supply to natural enemies because resource availability through time is increased. Trees can also provide alternate hosts to natural enemies, as in the case of the planting of prune trees adjacent to grape vineyards to support overwintering populations of the parasitoid A. epos (Murphy et al., 1998). Shade from trees may markedly reduce pest density in understory intercrops. Hedgerows or windbreaks of trees have a dramatic influence on microclimate; almost all microclimate variables (heat input, wind speed, soil desiccation, and temperature) are modified downwind of a hedgero hedgerow w. Tall intercrops or thick ground covers can also alter the reflectivity, temperature, and evapotranspiration of shaded plan pl ants ts or at th thee so soil il su surf rfac ace, e, wh whic ich h in tu turn rn co coul uld d af affe fect ct in inse sect ctss th that at co collonize according to “background” color or that are adapted to specific microclimatological ranges (Cromartie, 1981). Both immature and adult insect growth rates, feeding rates, and survival survival can be dramatically cal ly af affec fected ted by cha change ngess in moi moistu sture re and tem temper peratu ature re (Pe (Perri rrin, n, 197 1977). 7). The effect of shade on pests and diseases in agroforestry has been studied quite intensively in cocoa and coffee systems undergoing transformation from traditionally shaded crop species to management in unshaded conditions. In cocoa plantations, insufficient overoverhead shade fav favors ors the develo development pment of numerous herbivorous herbivorous insect specie spe cies, s, inc includ luding ing thr thrips ips (Seleno and d mi miri rids ds (Sahl(Selenothrips thrips rubr rubrocinctu ocinctus) s) an bergella sp., Distantiella sp., etc.). Even in shaded plantations, these insects concentrate in spots where the shade trees have been destroyed, such as by wind (Beer et al., 1997). Bigger (1981) found an
increase in the numbers of Lepidoptera, Homoptera, Orthoptera, and the mirid Sahlbergella singularis and a decrease in the number of Diptera and parasitic Hymenoptera from the shaded toward the unshaded part of a cocoa plantation in Ghana. In co cofffe fee, e, th thee ef effe fect ct of sh shad adee on in inse sect ct pe pest stss is le less ss cl clea earr th than an in co co-coa, as the leaf miner (Leucoptera meyricki) is reduced by shade, (Hypothenem thenemus us hampe hampei) i) may incre where wh ereas as the cof coffee fee ber berry ry bor borer er (Hypo increase ase under shade. Similarly, unshaded tea suffers more from attack by thrips and mites, such as the red spider mite (Oligonychus coffeae) and the pink mite (Acaphylla theae), wher whereas eas heav heavily ily shad shaded ed and moist plantations are more damaged by mirids ( Helopeltis spp.) (Guharay et al., 2000). Although in Central America coffee berry borer appears to perform equally well in open sun and managed shade, naturally occurring Beauveria bassiana (an entomopathogenic fungus) multiplies and spreads more quickly with greater humidity, and fungus applications should coincide with peaks in rainfall (Guharay et al., 2000). After a study of how the microclimate created by multistrata shade management affected herbivores, herbivores, diseases, weeds, and yields in Centrall Ame tra Americ ricaa cof coffee fee pla planta ntatio tions, ns, Sta Stave verr and col collea league guess (2 (2001 001)) def define ined d the conditions for minimum expression of the pest complex. For a low-elevation dry coffee zone, shade should be managed between 35 to 65 percent, as shade promotes leaf retention in the dry season and reduces Cercospora coffeicola, weeds, and Planococcus citri. Obviously ou sly,, the op optim timum um sha shade de con condit dition ionss fo forr pes pestt sup suppr press ession ion dif differ ferss wit with h climate, altitude, and soils. The selection of tree species and associations, density, and spatial arrangements, as well as shade-managementt re men regim gimes, es, are cri critica ticall con consid sidera eratio tions ns for sha shadede-str strata ata des design ign (Fi (Figgure 8.1). The complete elimination of shade trees may have an enormous impact on the diversity and density of ants. Studying the ant community in a gradient of coffee plantations going from plantations with high density of shade to shadeless plantations, Perfecto and Vandermeer (1996) reported a significant decrease in ant diversity. Although the relationship between ant diversity and pest control is not well understood, we can speculate that a diverse ant community can offe of ferr mo more re sa safe fegu guar ards ds ag agai ains nstt pe pest st ou outb tbre reak akss th than an a co comm mmun unit ity y do domminat in ated ed by ju just st a fe few w sp spec ecie ies. s. In Co Colo lomb mbia ia,, pr prel elim imin inar ary y re repo port rtss po poin intt to lower levels of the coffee borer, the main coffee pest in the region, in shaded coffee plantations. There are more indications that a non-
FIGURE 8.1. Conceptual graph depicting the relative importance of yield-reducing factors in a low, dry coffee zone in Nicaragua. Effects are shown to be additive with with the effect effect of each successiv successive e pest repre represented sented by by the area betwe between en the lines. The lowest line indicates the accumulated potential for yield reduction at diffe dif fere rent nt sh shad ade e le leve vels ls.. Sin Since ce th the e yy-ax axis is is ne nega gati tive ve,, th the e ra rang nge e of le leas astt yie yield ld re redu ducction is 35 to 65 percent (after Guharay et al., 2000).
dominant domin ant sma small ll ant spe specie ciess is res respo ponsi nsible ble for the con contro trol. l. App Appare arentl ntly y, this species does not live in unshaded plantations. Cocoa is another crop that is traditionally cultivated under shade trees. The ant species that have been so successful in controlling pests in cocoa are all species that flourish under shaded conditions. One of the most obvious consequences consequences of pruning or shade elimination, with regard to the ant community, is the change in microclimatic conditions. In particular particular,, microclimate becomes more variable with more extreme levels of humidity and temperature. A recent study documented changes in the composition of the ant community with shade and leaf-litter manipulation similar to those that occur after plowing (Perfecto and Vandermeer, 1996). Chemical cues used by herbivores herbivores may be altered in an agroforestry system. Trees may exhibit a dramatically different chemical pro-
file than annual herbaceous plants intercropped in the system, masking or lessening the impact of the chemical profile produced by the annual crop. Several studies hav havee demonstrated olfactory deterrence as a factor in decreasing arthropod abundance (Risch, 1981). The attractiveness of a plant species for the pests of another species can be usefully employed in agroforestry associations in the form of trap crops which concentrate the pests or disease vectors, a plac pl acee wh wher eree th they ey ca caus usee les lesss da dama mage ge or ca can n be mo more re ea easi sily ly ne neut utra rali lize zed d (e.g., by spraying or collecting). Such trap crops are an interesting option when they attract pests from the primary crop within the field (local attraction) but not when they attract pests from areas outside the field (regional attraction). Nascimento, Mesquita, and Caldas (1986) demonstrated the strong attraction of the Citrus pest Cratosomus flavofasciatus by the small tree Cordia verbenacea in Bahia, Braz Br azil, il, an and d re reco comm mmen ende ded d th thee in incl clus usio ion n of th this is tr tree ee at di dist stan ance cess of 10 100 0 to 150 m in Citrus orchards. They speculated that pests of several other fruit crops could similarly be trapped by this tree species. In certain agroforestry systems, such as alley cropping or systems with perennial crops and leguminous shade trees, relatively large quan qu anti titi ties es of NN-ri rich ch bi biom omas asss ma may y be ap appl plie ied d to cr crop ops. s. In ca case sess of lu luxxury consumption of N, this may result in reduced pest resistance of the crops. The reproduction reproduction and abundance abundance of several several insect pests, especially Homoptera, are stimulated by the high concentration of free nitrogen in the crop’s foliage. In studies assessing the effects of nine hedgerow species on the abundance abundan ce of major insect pests of beans and maize, and associated predatory/parasitic anthropods, Girma, Rao and Sithanantham (2000) found that beanfly ( Ophiomyia spp.) infestation was significantly higher in the presence of hedgerows (35 percent) than in their absence (25 percent). Hedgerows did not influence aphid (Aphis fabae) infestation of beans. In contrast, maize associated with hedgerows exper ex perien ienced ced sig signif nifica icantl ntly y lo lower wer sta stalk lk bor borer er ( Busseola Busseola fusca and Chilo spp.) and aphid (Rhopalosiphum maidis) infestations than pure maize, thee ma th marg rgin in of di difffe fere renc ncee be bein ing g 13 pe perc rcen entt an and d 11 pe perc rcen entt re resp spec ecti tiv vely for the two pests. Ladybird beetles closely followed their prey, aphids, with significantly higher catches in sole-cropped plants than in hedgerow plots and away from hedgerows. Activity of wasps was signif sig nifica icantl ntly y gr great eater er clo close se to hed hedger gerow owss tha than n aw away ay fro from m the them. m. Spi Spider der catches during maize season were 77 percent greater in the presence
of hedgerows than in their absence, but catches during other seasons were similar between the two t wo cropping systems. In on onee of th thee fe few w st stud udie iess of th thee in infl flue uenc ncee of tem tempe pera rate te ag agro rofo fore rest stry ry practices on beneficial arthropods, Peng and colleagues (1993) confirmed the increase in insect diversity and improved natural enemy abundance in an alley cropping system over that of a monoculture crop system. Their study examined arthropod diversity in control plots sown to peas ( Pisum sativum var. Sotara) versus peas intercropped with four tree species (walnut, ash, sycamore, and cherry) and hazel bushes. They found greater arthropod abundance in the alley-cropped plots compared to the control plots, and natural enemies were more abundant in the tree lines and alleys than in the controls.. The authors attributed trols attributed the incre increase ase in natur natural al enem enemies ies to the greater availability of overwintering sites and shelter in the agroforestry system. In fact, Stamps and Linit (1997) argue that agroforestry holds promise for increasing insect diversity and reducing pest problems because the combination of trees and crops provides greater niche diversity and complexity in both time and space than does polyculture of annual crops.
DESIGNING NATURAL SUCCESSIONAL ANALOG AGROFORESTRY SYSTEMS
At the heart of the agroecology strategy is the idea that an agroecosystem should mimic the diversity and functioning of local ecosystems, thus exhibiting tight nutrient cycling, complex structure, and enhanced biodiversity. biodiversity. The expectation is that such agricultural mimics, like their natural models, can be productive, pest resistant, and conservative of nutrients and biodiversity. Ewel (1986) argues that natural plant communities have several traits (pest suppression among amon g them) that would be desirable to incorporate into agroecosystems. In order to test this idea, he and others (Ewel et al., 1982) studied prod producti uctivity vity,, gro growth, wth, resili resilience, ence, and reso resourceurce-utiliz utilization ation char characteracteristics of tropical successional plant communities. These researchers contend that desirable ecological patterns of natural communities should be incorporated into agriculture by designing cropping systems that mimic the structural and functional aspects of secondary succes suc cessio sion. n. Thu Thus, s, the pr prev evale alent nt coe coevo volv lved ed nat natur ural al sec secon ondar dary y pla plant nt as-
sociations of an area should provide the model for the design of multispecies crop mixtures (Soule and Piper, 1992). This succession analog method requires a detailed description of a natura nat urall eco ecosys system tem in a spe specif cific ic en envir vironm onment ent and the bot botani anical cal cha charac rac-terization of all potential crop components. When this information is available, ava ilable, the first step is to find crop plants that are structurally and funct fu nction ionall ally y sim simila ilarr to the pla plants nts of the nat natura urall eco ecosys system tem.. The spa spatia tiall and chronological arrangement of the plants in the natural ecosystem are then used to design an analogous crop system (Hart, 1980). In Costa Cos ta Ric Rica, a, res resear earche chers rs con conduc ducted ted spa spatia tiall and tem tempor poral al rep replac laceme ements nts of wil wild d spe specie ciess wit with h bot botani anicall cally y, str struct uctura urally lly,, and eco ecolog logical ically ly sim simila ilarr cultivars. cultiv ars. Thus, successional members of the natural system such as Heliconia spp., cucurbitaceous vines, Ipomoea spp., legume vines, shru sh rubs bs,, gr gras asse ses, s, an and d sm small all tr tree eess we were re re repl plac aced ed by pl plan anta tain in,, sq squa uash sh varieties, and yams. By years two and three, fast-growing tree crops (Brazi (Br azill nu nuts, ts, pea peach, ch, pal palm, m, ros rosew ewood ood)) may for form m an add additio itional nal str stratu atum, m, thus maintaining continuous crop cover, avoiding site degradation and nutrient leaching, and providing crop yields throughout the year (Ewel, 1986). Under a scheme of managed succession, natural successional stages are mimicked by intentionally introducing plants, animals, practices, and inputs that promote the development of interactions and connections between component parts of the agroecoystem. Plan Pl antt sp spec ecie iess (b (bot oth h cr crop op an and d no nonc ncro rop) p) ar aree pl plan ante ted d th that at ca capt ptur uree an and d re re-tain nutrients in the system and promote good soil development. These plants include legumes, with their nitrogen-fixing bacteria, and plants with phosphor phosphorus-trapping us-trapping mycorrhizae. As the system develops, increasing diversity, food-web complexity, and level of mutualistic interactions all lead to more effecti effective ve feedback mechanisms for pest and disease management. The emphasis during the development process is on building a complex and integrated i ntegrated agroecosystem with less dependence on external inputs. Ther Th eree ar aree ma many ny wa ways ys th that at a fa farm rmer er,, be begi ginn nnin ing g wi with th a re rece cent ntly ly cu culti lti-vated field of bare soil, can allow successional development to proceed beyond the early stages. One general model is to begin with an annual monoculture and progress to a perennial tree crop system.
1. The farmer begins by planting a single annual crop that grows rapi ra pidl dly y, ca capt ptur ures es so soil il nu nutr trien ients ts,, gi giv ves an ea earl rly y yi yiel eld, d, an and d ac acts ts as a pioneer species in the developmental process. 2. As a next step (or instead of the previous one), the farmer can plant a polyculture of annuals that represent different components of the pioneer stage. The species would differ in their nutrient needs, attract different insects, have different rooting depths, and return a different proportion of their biomass to the soil. One might be a nitrogen-fixing legume. All of these early species would contribute to the initiation of the recovery pro cess, and they would modify the environment so that noncrop plants and animals—especially the macroorg macroorganisms anisms and microorganisms necessary for developing the soil ecosystem—can also begin to colonize. 3. Following the initial stage of development, short-lived perennial crops can be introduced. Taking advantage of the soil cover created by the pioneer crops, these species can diversify the agroecosystem in important ecological aspects. Deeper root systems, more organic matter stored in standing biomass, and greater habitat and microclimate diversity all combine to advance the successional development of the agroecosystem. 4. Once soil conditions improve sufficiently, the ground is prepared par ed for pla planti nting ng lon longer ger-li -live ved d per perenn ennial ials, s, esp especi ecially ally orc orchar hard d or tree crops, with annual and short-lived perennial crops maintained in the areas between them. While the trees are in their early growth, they have limited impact on the environment arou ar ound nd th them em.. At th thee sa same me ti time me,, th they ey be bene neffit fr from om ha havi ving ng an annu nual al crops around them, because in the early stages of growth they are oft often en mo more re sus suscep ceptib tible le to int interf erfere erence nce fro from m agg aggres ressi sive ve wee weedy dy species that would otherwise occupy the area. 5. As the tree crops develop, the space between them can continue to be managed with annuals and short-liv short-lived ed perennials. 6. Eventually, once the trees reach full development, the end point in th thee de deve velo lopm pmen enta tall pr proc oces esss is ac achi hieeve ved. d. Th This is las lastt st stag agee is do domminated by woody plants, which are key to the site-restoring growerss of fallow vegetation because of their deep, permanent grower root systems.
Once a successionally developed agroecosystem has been created, the problem becomes how to manage it. The farmer has three basic options: • Re Retu turn rn th thee en enti tire re sy syst stem em to th thee in init itial ial st stag ages es of su succ cces essi sion on by in in--
troducing a major disturbance, such as clear-cutting the trees in the perennial system. Many of the ecological advantages that have been achieved will be lost and the process must begin anew. • Maintain the system as a tree-crop-based agroecosystem. • Reintroduce disturbance into the agroecosystem in a controlled and an d lo loca cali lize zed d ma mann nner er,, ta taki king ng ad adv van antag tagee of th thee dy dyna nami mics cs th that at su such ch patchiness introduces into an ecosystem. Small areas in the system te m ca can n be cl clea eare red, d, re retu turn rnin ing g th thos osee ar area eass to ea earl rlie ierr st stag ages es in su succcession, and allowing a return to the planting of annual or shortlived crops. If care is taken in the disturbance process, the belowground belowgr ound ecosystem can be kept at a later stage of developdevelopment, whereas the aboveground system can be made up of highly hig hly pro produc ducti tive ve spe specie ciess tha thatt are av availa ailable ble fo forr har harve vest st rem remov oval. al. According to Ewel (1999), the only region where it pays to imitate natura nat urall eco ecosys system tems, s, rat rather her tha than n str strug uggle gle to imp impose ose sim simpli plicit city y thr throug ough h high inputs in ecosystems that are inherently complex, is the humid tropical lowlands. This area epitomizes enviro environments nments of low abiotic stress str ess bu butt ov overw erwhel helmin ming g bio biotic tic int intric ricac acy y. The ke keys ys to agr agricu icultu ltural ral suc suc-cess in this region are to (1) channel productivity productivity into outputs of nutritional and economic importance, (2) maintain adequate vegetation ti onal al di dive vers rsity ity to co comp mpen ensa sate te fo forr lo loss sses es in a sy syst stem em si simp mple le en enou ough gh to be horticulturally manageable, (3) manage plants and herbivores to facil fa cilita itate te as asso soci ciat atio iona nall re resi sist stan ance ce,, an and d (4 (4)) us usee pe pere renn nnia iall pl plan ants ts to ma main in-tain soil fertility, guard against erosion, and make full use of resources. To many, many, the ecosystem-analog approach is the basis for the promotion of agroforestry systems, especially the construction of forestlike agroecosystems that imitate successional vegetation and exhibit low requirements for fertilizer, high use of available nutrients, and high protection from pests (Sanchez, 1995).
THE NEED FOR FURTHER RESEARCH
The effects of agroforestry plants (or techniques) on pests and diseases eas es can be di divid vided ed int into o bio biolog logica icall (sp (speci ecieses-rel relate ated) d) and ph physi ysical cal ef ef-fects of components (e.g., microclimate). The former is highly specific for certain plant-pest or plant-disease combinations and must be stud st udie ied d on a ca case se-b -byy-ca case se ba basi sis. s. Th Thee la latte tterr ar aree ea easi sier er to ge gene nera raliz lize, e, but even they depend on the regional climatic conditions. Based on results sul ts fro from m int interc ercro roppi pping ng stu studie dies, s, agr agrofo ofores rester terss ex expec pectt tha thatt agr agrofo ofores resttry systems may provide opportunities to noticeably increase arthropod po d di dive versi rsity ty and lo lower wer pes pestt pop popula ulatio tions ns com compar pared ed to the pol polycu ycultu lture re of annual crops or trees by themselves. However, more work is needed in specific areas of research such as studies of the differences in arthropod populations between agroforestry and traditional agronomic systems, research into the specific mechanisms behind enhancement of pest management with agroforestry practices, and basic research into the life histories of target t arget pests and potential natural enemies. An understanding of what aspects of trees modify pest populatio ula tions— ns—she shelter lter,, foo food, d, or ho host st res resou ource rcess for nat natura urall ene enemie mies, s, tem tempo po-ral continuity, continuity, microclimate alteration alt eration or apparency—should help in determining future agroforestry design practices (Rao, Singh, and Day, 2000). Well-designed agroforestry techniques can reduce crop stress by providing providin g the right amount of shade, reducing temperature extremes, sheltering off strong winds, and improving soil fertility, thereby improv pr ovin ing g th thee to tole lera ranc ncee of cr crop opss ag agai ains nstt pe pest st an and d di dise seas asee da dama mage ge,, wh while ile at the same time influencing the development conditions for pest and disease organisms and their natural enemies. Poorly designed systems, on the other hand, may increase the susceptibility of crops to pests.
Chapter 9
Designing Pest-Stable Designing Pest-Stable Vegetationally Vegetationally Diverse Agroecosystems Diverse Agroecosystems
MONOCULTURES AND THE FAILURE OF CONVENTIONAL PEST-CONTROL APPROACHES
Until about four decades ago, crop yields in U.S. agricultural systems depended on internal resources, recycling of organic matter, built-in biological control mechanisms, and rainfall patterns. Agricultural yields were modest but stable. Production was safeguarded by growing more than one crop or variety in space and time in a field as in insu sura ranc ncee ag agai ains nstt pe pest st ou outb tbre reak akss or se sev ver eree we weat athe herr. In Inpu puts ts of ni nitr troogen were gained by rotating major field crops with legumes. In turn, rotations suppressed insects, weeds, and diseases by effectively breaking in g th thee lif lifee cy cycl cles es of th thes esee pe pest sts. s. A ty typi pica call co corn rn-b -bel eltt fa farm rmer er gr grew ew co corn rn rotated with several crops, including soybeans, and small-grain production was intrinsic to maintain livestock (USDA, 1973). Most of the labor was done by the family with occasional hired help, and no specialized equipment or services were purchased from off-farm sources (Altieri, 1995; Gliessman, 1999). In the developing world, small farmers developed even more complex and biodiverse farming systems syst ems guided by indigenous knowledge knowledge that has stood the test of time (Thrupp, 1997). In these types of farming systems, the link between agriculture and ecology was quite strong, and signs of env enviironmental degradation were seldom evident. As agricultural modernization progressed, however, however, the ecologyfarming linkage was often broken as ecological principles were igno ig nore red d an and/ d/or or ov over erri ridd dden en.. As pr prof ofit it ra rath ther er th than an pe peop ople le’’s ne need edss or en en-vironmental concerns shaped the modes of agricultural production, agribusiness agribusin ess interests and prevailing policies favored favored large farm size, specialized production, crop monocultures, and mechanization.
Toda oday y, mon monocu ocultu ltures res ha have ve inc increa reased sed dra dramat matica ically lly wo world rldwid wide, e, mai mainly nly through the geographical expansion of land annually devoted to single crops. Thus, monoculture has implied the simplification of biodiversity diver sity,, the end result being an artificial ecosystem requiring constant human intervention in the form of agrochemical inputs which, in addition to temporarily boosting yields, result in a number of undesirable environmental environmental and social costs. As long as the structure of monocultures is maintained as the structural base of modern agricultural systems, pest problems will continue to be the result of a negative treadmill that reinforces itself (Figure 9.1). The exacerbation of Monoculture
Natural vegetation displacement
Intensive use of fertilizers
Massive supply of host plants
Destruction of beneficial habitat fauna
Nutritional imbalance in crops
Host plant switch
Reduction in biodiversity
Higher vulnerability to pests
Recruitment of herbivores
Disruption of natural control
Intensive use of pesticides Herb He rbic icid ides es
Inse In sect ctici icide des s
Resistance to insecticides Elimination of natural enemies Elimination of trap crops and insectary plants harboring natural enemies
In situ completion of life cycles
Pest resurgence More severe pest problems
Secondary pests
Treadmill Effect Lower effectiveness of insecticides Higher uses of insecticides Higher costs of production Yield decline in the long term
FIGURE 9.1. The ecological consequences of monoculture with special reference to pest problems and the agrochemical treadmill.
pest problems due to vegetational simplification has been the subject of this book, but pest problems can also result from excessive use of chemic che mical al fer fertili tilizer zerss and pes pestici ticides des (Ph (Phela elan, n, Mas Mason on,, and Sti Stinn nner er,, 199 1995). 5). Thee su Th subs bsta tant ntia iall yi yiel eld d lo loss sses es du duee to pe pest sts, s, ab abou outt 20 to 30 pe perc rcen entt fo forr most crops (similar pest-loss levels were observed forty years ago) despite the increase in the use of pesticides (about 4.7 billion pounds of pesticides were used worldwide in 1995, 1.2 billion pounds in the United States alone) is a symptom of the environmental environmental crisis affecting agriculture. Cultivated plants grown in genetically homogeneous monocultures do not possess the necessary ecological defense mechanisms ani sms to tol tolera erate te the imp impact act of pes pestt out outbre breaks aks.. Mo Moder dern n agr agricu icultu lturis rists ts have ha ve se sele lect cted ed cr crop opss ma main inly ly fo forr hi high gh yi yiel elds ds an and d hi high gh pa palat latab abil ility ity,, ma makking them more susceptible to pests by sacrificing natural resistance for productivity. productivity. Consequently, as modern agricultural practices reduce du ce or eli elimin minate ate the res resou ource rcess and opport opportuni unitie tiess for nat natura urall ene enemie miess of pes pests, ts, the their ir nu numbe mbers rs dec declin line, e, dec decrea reasin sing g the bio biolog logical ical sup suppre pressi ssion on of pests. Due to this lack of natural controls, an investment investment of about $40 billion in pesticide control is incurred yearly by U.S. farmers, which is estimated to save approximately $16 billion in U.S. crops. However, the indirect costs of pesticide use to the environment and publ pu blic ic he heal alth th ha have ve to be ba bala lanc nced ed ag agai ains nstt th thes esee be bene neffits its.. Ba Base sed d on th thee avail av ailabl ablee dat data, a, the en envir viron onmen mental tal cos costt (im (impac pacts ts on wil wildli dlife, fe, pol pollin linato ators, rs, natural enemies, fisheries, water, and dev development elopment of resistance) and social costs (human poisonings and illnesses) of pesticide use reach about $8 billion each eac h year (Conway and Pretty, 1991). What is worrisome so me is th that at pe pest stic icid idee us usee is st stil illl hi high gh an and d st stil illl ri risi sing ng in so some me cr crop oppi ping ng systems. Data from California show that from 1991 to 1995 pesticide use increased from 161 to 212 million pounds of active ingredient. This increase was not due to increases in planted acreage, as statewide crop acreage remained constant during this period. Much of the increase was due to intensification of some crops (grapes, strawberries) and higher uses of particularly toxic pesticides, many of which are linked to cancer (Liebman, 1997). On the other hand, increasing evidence suggests that crops grown in chemically fertilized systems are more susceptible to pest attacks than crops grown in organic-rich and biologically active soils. Many studies suggest that the physiological susceptibility of crops and path pa thog ogen enss ma may y be af affe fect cted ed by th thee fo form rm of fe fert rtili ilize zerr us used ed (o (org rgan anic ic ve verrsus chemical fertilizer). Studies documenting lower densities of sev-
eral insect herbivores in low-input systems have partly attributed such reduction to a low nitrogen content in organically farmed crops (Altieri, 1995). Given these findings, findings, a major challenge for scientists and farmers advocating for more ecologically based pest management (EBPM) is to fi find nd str strate ategie giess to ov overc ercome ome the eco ecolog logica icall lim limits its imp impose osed d by mon monoocultures (lack of diversity and foliage nutritional imbalances), by converting con verting such systems into diver diversified sified agroecosystems dependent on internal resources and above aboveground ground and belowground synergism rather than on high external inputs. This chapter provides some ideas and principles of agroecosystem design that may lead to a more optimal biological pest regulation.
TOWARD SUSTAINABLE AGRICULTURE
The search for self-sustaining, low-input, diversified, and energyefficient agricultural systems is now a major concern of many researchers, farmers, and policymakers worldwide. More sustainable food-production systems seek to make the best use of nature’s goods and services while not damaging the environment (Altieri, 1995, 1999; Thrupp, 1997; Conway, 1994; Pretty, 1995, 1997; Pretty and Hine, 2000). This can be done by integrating natural processes such as nutrient cycling, nitrogen fixation, soil regeneration, and natural enemies of pests into food-production processes. processes. A sustainable agriculture also minimizes the use of nonrenewable inputs (pesticides and fertilizers) that damage the environment or harm the health of farmers and consumers by encouraging regenerative and resourceconserving technologies. It makes better use of the knowledge and skills of farmers, improving their self-reliance. It also seeks to make productive use of people’s capacities to work together to solve common management problems, such as pest problems at a regional level. Most people involved involved in the promotion of sustainable agriculture aim at creating a form of agriculture that maintains productivity in th thee lo long ng te term rm by do doin ing g th thee fo follo llowi wing ng (P (Pre rett tty y, 19 1995 95;; Al Alti tier eri, i, 19 1995 95): ): • Optimizing the use of locally available resources by combining
the different components of the farm system, i.e., plants, ani-
•
•
•
•
•
mals, soil, water, climate, and people, so that they complement one another and have the greatest possible synergetic effects Reducing the use of off-farm, external, and nonrenewable inputs with the greatest potential to damage the environment or harm the health of farmers and consumers, and a more targeted usee of th us thee re rema main inin ing g in inpu puts ts em empl ploy oyed ed wi with th a vi vieew to mi mini nimi mizi zing ng variable costs Relying mainly on resources within the agroecosystem by replacing external inputs with nutrient cycling, better conservation, and an expanded use of local resources Improving the match between cropping patterns and their productive potential and environmental constraints of climate and landscape to ensure long-term sustainability of current production levels Working to value and conserve biological diversity, both in the wild wi ld an and d in do dome mest stic icat ated ed la land ndsc scap apes es,, an and d ma maki king ng op opti tima mall us usee of thee bi th biol olog ogic ical al an and d ge gene neti ticc po poten tenti tial al of pl plan antt an and d an anim imal al sp spec ecies ies Taking full advantage of local knowledge and practices, including innovative approaches not yet fully understood by scientists although widely adopted by farmers
As it has been emphasized in this book, a key strategy in sustainable agriculture is to t o restore agricultural diversity of the agricultural landscape (Altieri, 1987). Diversity can be enhanced in time through crop rotations and sequences and in space in the form of cover crops, intercropping, agroforestry, agroforestry, crop-liv crop-livestock estock mixtures, and manipulation tio n of ve veget getati ation on out outsid sidee the cro crop p are area. a. Veg egeta etatio tion n di dive versi rsifi ficat cation ion no nott only results in pest regulation through restoration of natural control but also produces optimal nutrient recycling, soil conservation, energy conservation, conservation, and less dependence on external inputs.
REQUIREMENTS OF SUSTAINABLE AGROECOSYSTEMS
The basic tenets of a sustainable agroecosystem are conservation of renewable resources, adaptation of the crop-animal combinations to the environment, and maintenance of a moderate but sustainable level of productivity. To emphasize long-term ecological sustain-
ability rather than short-term productivity, the production system must do the following following:: 1. Reduce energy and resource uses and regulate the overall energy input so that the output:input ratio is high 2. Reduce nutrient losses by effectively effectively containing leaching, runoff, and erosion and improve nutrient recycling through the use of legumes, organic manure, compost, and other effective recycling mechanisms 3. Encourage local production of crops adapted to the natural and socioeconomic setting 4. Sustain a desired net output by preserving the natural resource base that exists by minimizing soil degradation or genetic erosion 5. Red Reduce uce cos costs ts and inc increa rease se the ef effi ficie cienc ncy y and eco econo nomic mic via viabil bility ity of small and medium-sized farms, thereby promoting a diver diverse, se, potentially resilient agricultural system (Altieri, 1987) As shown in Figure 9.2, from a management viewpoint, the basic components of a sustainable agroecosystem include 1. Veg egeta etati tive ve co cove verr as an ef effec fecti tive ve soi soill- and wat water er-co -conse nservi rving ng mea mea-sure, met through the use of no-till practices, mulch farming, use of cover crops, etc. 2. Reg Regula ularr sup supply ply of or organ ganic ic mat matter ter thr throu ough gh re regul gular ar add additio ition n of or or-gani ga nicc ma matte tterr (m (man anur ure, e, co comp mpos ost) t) an and d pr prom omot otio ion n of so soil il bi biot otic ic ac ac-tivity 3. Nutrient-recycling mechanisms through the use of crop rotations, crop-livestock mixed systems, agroforestry, and intercropping systems based on legumes, etc. 4. Pest regulation assured through enhanced activity activity of biological control agents achieved through biodiversity manipulations and by introducing and/or conserving natural enemies DESIGNING HEALTHY AGROECOSYSTEMS
As shown in the previous chapters, diversified cropping systems, such as those based on intercropping and agroforestry or cover cropping pi ng of or orch char ards ds,, ha have ve re rece cent ntly ly be been en th thee tar targe gett of mu much ch re rese sear arch ch.. Th This is
OBJECTIVES
Diversified in Diversified time and space
Dynamically stable
Productive and food self-sufficient
Conservation and regeneration of natural resources (water, soil, nutrients) germplasm
Economic potential
Socially and culturally acceptable technology
Self-promoting and self-help potential
MODEL SUSTAINABLE AGROECOSYSTEM PROCESSES
Soil cover
Nutrient recycling
Sediment capture, water harvest, and conservation
Productive diversity
Crop protection
Ecological “order”
METHODS
Crop systems:
Polyc Po lycult ulture ures: s:
Living Liv ing and nonliving barriers:
Regional diversity:
Genetic diversity:
Agroecosystem design and reorganization:
polycultures fallow rotation crop densities mulching cover cropping no tillage selective weeding
use of residues rotation with legumes zonification of production improved fallow manuring alley cropping
selective weeding terracing no tlilage zonification contour planting
forest enrichment crop zonification crop mosaics windbreaks, shelterbelts
species diversity cultural control biological control
mimicking natural succession agroecosystem analysis methodologies
Diversity within the agroecosystem: polycultures agroforestry crop-livestock association variety mixtures
FIGURE 9.2. Methods to diversify agroecosystems to ensure a series of ecological processes and sustainability objectives.
interest is largely based on the new emerging evidence that these systems are more stable and more reso resource urce cons conservin erving g (Vande (Vandermeer rmeer and Perfecto, 1995). Many of these attributes are connected to the higher levels of functional biodiversity associated with complex farming systems. The link between biodiversity and ecosystem function has been a main focus of this book. The various studies and the designs used to test the t he effects that plant diversity has on the regulation of insect herbivore populations populations are a key source of information for implementin men ting g str strate ategie giess tha thatt enh enhanc ancee the ab abun undan dance ce and ef effi ficac cacy y of ass associ oci-ated natural enemies (Altieri and Letourneau, 1984). The inherent self-regulation characteristics of natural communities ti es ar aree lo lost st wh when en hu huma mans ns mo modi dify fy an and d si simp mplif lify y th thes esee co comm mmun uniti ities es by breaking the fragile thread of community interactions. Agroecologists maintain that this breakdown can be repaired in agroecosystem sys temss by res restor toring ing com commun munity ity hom homeos eostas tasis is thr throug ough h the add additi ition on or enhancement of biodiversity (Altieri and Nicholls, 2000). Thee fir Th irst st st step ep is to id iden enti tify fy th thee ro root ot ca caus uses es of th thee in inst stab abil ilit ity y or “l “lac ack k of immunity” of agroecosystems (Box 9.1). The second step is to encourage management practices that will optimize key agroecosystem processes, which underlie agroecosystem health (Box 9.2). All such practices should lead to enhancement of aboveground and belowground functional biodiversity biodiversity which in turn play ecological roles in restoring the productive capacity of the system. An important step is to also identify the type of biodiversity that is desi de sira rabl blee to ma main intai tain n an and/ d/or or en enha hanc ncee in or orde derr to ca carr rry y ou outt ke key y ec ecol olog og-ical services, services, and then to deter determine mine the best practices practices that will encourage the desired biodiv biodiversity ersity components. Figure 1.5 showed that there are many agricultural practices and designs that have the potential to enhance functional biodiversity and others that negati negatively vely affect it. The idea is to apply the best management practices in order to enhance or regenerate the kind of biodiversity that subsidizes the BOX 9.1. Causes of Agroecosystem Dysfunction • Excess pesticides
• Monoculture
• Excess fertilizers
• Low functional biodiversity
• Lo Low w so soilil or organ ganic ic ma matt tter er co cont nten entt
• Ge Genet netic ic un unif iform ormit ity y
• Lo Low so soil il bi biol olog ogic ical al ac acti tivi vity ty
• Nut Nutri rien entt de defi fici cien enci cies es • Moistu Moisture re imbalanc imbalances es
BOX 9.2. Routes to Agroecosystem Health and Mechanisms to Improve Agroecosystem Immunity Routes to agroecosystem health • Strengthen the immune immune system (proper functioning of natural pest control) • Decrease toxicity toxicity through elimination of agrochemicals • Optimize metabolic function (organic matter matter decomposition and nutrient cycling) • Balance regulatory systems (nutrient cycles, cycles, water balance, energy flow, flow, population regulation, etc.) • Enhance conservation and regeneration of soil-water resources and biodiversity • Incre Increase ase and sustain sustain long-term long-term productivity productivity Mechanisms to improve agroecosystem immunity • Incre Increase ase plant species species and genetic diversity diversity in time and space space • Enhance functional biodiversity (natural enemies, antagonists, etc.) • Enhance soil soil organic matter matter and biological biological activity activity • Increase soil cover and crop competitive ability • Eliminat Eliminate e to toxic xic inputs inputs
health and sustainability of agroecosystems by providing ecological services such as biological pest control, nutrient cycling, water and soil conservation, etc. (Gliessman, 1999). As depicted in Figure 9.3, crop health can be achieved by establishing mechanisms that aid in the regulation of insect pests through two routes: (1) enhancing the rich natural enemy biodiversity harbored by a diversified agroecosystem, which has been explored in this book, and (2) encouraging a healthy soil rich in organic matter and a diverse soil biota.
HEALTHY SOILS—HEALTHY PLANTS
The fields of integrated pest management (IPM) and integrated soil-ferti soilfertility lity manag management ement (ISF (ISFM) M) hav havee proc proceeded eeded separa separately tely without without realizi rea lizing ng tha thatt lo low-i w-inp nput ut agr agroec oecosy osyste stems ms rel rely y on syn syner ergie giess of pla plant nt di di-versity and the continuing function of the soil microbial community and its relationship with organic matter to maintain the integrity of the agroecosystem. It is crucial for scientists to understand that most
Agroecological Principles
Agroecosystem Design
Habitat Management and Diversification
Organic Management
Soil Quality
Plant Health
Agroecosystem Health
FIGURE 9.3. The pillars of egroecosystem health.
pest-management methods used by farmers can also be considered soil-fertility management strategies and that there are positive interactions between soils and pests that, once identified, can provide guidelines for optimizing total agroecosystem function (Figure 9.4). Increasingly,, research is showing that the ability of a crop plant to reIncreasingly sist or tolerate insect pests and diseases is tied to optimal physical, chemical, and especially biological properties of soils. Soils with high organic matter and active biological activity generally exhibit good soil fertility, as well as complex food webs and beneficial organisms that prevent infection. On the other hand, farming practices that cause nutrition imbalances can lower pest resistance (Phelan, Mason, and Stinner, 1995). Yet it is for this reason that the crop-div crop-diversification ersification strategies emphasized in this book must be complemented by regular applications of organic amendments (crop residues, animal manures, and composts) to maintain or improv improvee soil quality and productivity. productivity. Much is
Fertilizers Cover crops Green manures
Enhanced Soill Fertility Soi
Mulching Compost Rotations, etc. Interactions (+,
SYNERGISM
HEALTHY CROP
Healthy Agroecosystem
Crop diversity Cultural practices Pesticides
Enhanced Pest Regulation
Habitat modification
FIGURE FIGUR E 9.4 9.4.. Int Inter eract actions ions of soi soill and pes pestt man manage agemen mentt pra practi ctices ces use used d by fa farmer rmers, s, som some e of whi which ch ma may y res result ult in syn synerg ergism ism leading to a healthy and productive crop.
known know n abo about ut the ben benef efits its of mu multis ltispec pecies ies ro rotati tations ons,, co cove verr cro crops, ps, agr agrooforestry, and intercrops. Less well known are the multifunctional effects of organic amendments beyond the documented effects on imim proved soil structure and nutrient content. Well-aged manures and composts can serve as sources of growth-stimulating substances, such as indole-3 acetic acid and humic and fulvic acids (Hendrix et al., 1990). Beneficial effects of humic acid substances on plant growth are mediated by a series of mechanisms, many similar to those tho se res result ulting ing fro from m the dir direct ect app applica licatio tion n of pla plantnt-gr grow owth th re regul gulato ators. rs. The ability of a crop plant to resist or tolerate pests is tied to optimal physical, chemical, and biological properties of soils. Adequate moisture, good soil tilth, moderate pH, correct amounts of organic matter and nutrients, and a diverse and active community of soil organisms all contribute to plant health. Organic-rich soils generally exhibit good soil fertility as well as complex food webs and beneficial organisms that prevent infection by disease causing organisms such as Phthium and Rhizoctonia (Hendrix et al., 1990). On the other hand, farming practices such as high applications of nitrogen fertilizer can create nutrition imbalances and render crops susceptible to diseases such as Phytophtora and Fusarium and stimulate outbreaks of Homopteran insects such as aphids and leafhoppers (Campbell, 1989 19 89). ). In fa fact ct,, th ther eree is in incr crea easi sing ng evi vide denc ncee th that at cr crop opss gr gro own in or orga gani niccrich and biologically active soils are less susceptible to pest attack. Many studies suggest that the physiological susceptibility of crops to insect pests and pathogens may be affected by the form of fertilizer used (organic versus chemical fertilizer). The literature is abundan abundantt on the benefits of organic organic amendments thatt enc tha encou ourag ragee res reside ident nt ant antago agonis nists ts whi which ch enh enhanc ancee bio biolog logica icall con contro troll of plant diseases. Several bacteria species of the genus Bacillus and Pseudomonas, as well as the fungus Trichoderma, are key antagonists that suppress pathogens through competition, lysis, antibiosis, or hyperparasitism (Palti, 1981). Studies documenting lower abundance of sever several al insect i nsect herbivores in low-input systems have partly attributed such reduction to a low nini trogen content in organically farmed crops (Altieri, 1985). In Japan, density of immigrants of the planthopper Sogatella furcifera was signifi ni fica cant ntly ly lo lowe werr wh while ile th thee se settl ttlin ing g ra rate te of fe fema male le ad adul ults ts an and d th thee su surv rviv ival al rate of immature stages of ensuing generations were lower in organic rice fields. Consequently, Consequently, the density of planthopper nymphs and adults
in the ensuing generations decreased in organically farmed fields (Kajimura, 1995). In England, conv conventional entional winter wheat fields developed a larger infestation of the aphid Metopolophium dirhodum than its organic counterpart. This crop also had higher levels of free protein amino acids in its leaves during June, which were believed to have have resulte su lted d fr from om a ni nitr trog ogen en to top p dr dres essin sing g of th thee cr crop op ea earl rly y in Ap Apri ril. l. Ho Howe wev ver er,, the difference in the aphid infestation between crops was attributed att ributed to the aphid’s response to relative proportions of certain nonprotein to protein amino acids in the leaves at the t he time of aphid settling on crops (Kow (K owals alski ki and Viss isser er,, 19 1979) 79).. In gre greenh enhou ouse se ex exper perime iments nts,, whe when n gi give ven na choice of maize grown on organic versus chemically treated soils, EuEu ropean corn borer females preferred to lay significantly more eggs in chemically fertilized plants (Phelan, Mason, and Stinner, 1995). 1995).
RESTORING DIVERSITY IN AGRICULTURAL SYSTEMS
Most regions in the world have many types of agricultural systems determined by local variations in climate, soil, economic relations, social soc ial str struct ucture ure,, and his histor tory y. Cle Clearl arly y, the these se sys system temss are alw always ays cha chang ng-ing in size, land-tenure assignments, technological intensity, population shifts, resource availability, environmental degradation, economic growth or stagnation, political change, and so forth. Farmers adapt to some of these changes by responding through technological innov inn ovatio ation n to va varia riatio tions ns in the ph physi ysical cal en envir viron onmen ment, t, pri prices ces of inp inputs uts,, etc. A logical outcome to present environmental and socioeconomic constraints is the desire by farmers for more sustainable agricultural methods. The basic concepts of a self-sustaining, low-input, low-input, diversified, diversified, and efficient eff icient agricultural system must be synthesized into practical alteral ternative systems to suit the specific needs of farming communities in different agroecological regions of the world. A major strategy in sustainable agriculture is to restore agricultural diversity in time and space through crop rotations, cover crops, intercropping, crop-livestock mixtures, etc. (Altieri, 1987). As seen in Figure 9.5, different options to diversify cropping systems are available depending on whet wh ethe herr th thee cu curr rren entt mo mono nocu cultu lture re sy syst stem emss to be mo modi diffied ar aree ba base sed d on annual or perennial crops. Diversification can also take place outside of the farm, for example, in crop-field boundaries boundaries with windbreaks,
Modern Agricultural Systems Sys tems Annual Crop Based
Row Crop Monocultures
Small Grains
Perennial Crop Based
Vegetables
Alfalfa
Fruit Orchards
Vineyards
Diversification Diversific ation Strategies crop sequences and rotations multiple cropping (inter, strip, etc.) no tillage living mulches managed fallows mulch farming windbreaks win dbreaks Agro-Pastoral Systems
Agrifore stry Agriforestry (combined crop-tree production)
Crop-Livestock Mixed Farming
cover cropping perennial polycultures and mixed orchards intercrop w/annual crops mulching shelterbelts surrounding vegetation manipulation manip ulation Agro-Silva-Pastoral Agro-Silva-Pastoral Systems
FIGURE 9.5. Diversification strategies for for annual crop-based or perennial crop-based modern agroecosystems.
shelterbelts, and living fences or corridors, which can improv improvee habitat for wildlife and beneficial insects, provide sources of wood, organic matter, and resources for pollinating bees, and, in addition, modify wind speed and the microclimate (Altieri and Letourneau, 1982; Kemp and Barrett, 1989). In this book many alternative diversification strategies that exhibit beneficial effects on soil fertility, crop protection, and crop yields have been explored. Key ones include the following: 1. Crop Rotations: Temporal diversity incorporated into cropping systems, providing crop nutrients, and breaking the life cycles of several insect pests, diseases, and weeds 2. Polycultures: Complex cropping systems in which two or more crop species are planted within sufficient spatial proximity to result in competition or complementation, thus enhancing yields (Francis, 1986; Vandermeer, 1989) 3. Agroforestry Systems: An agricultural system where trees are grown together with annual crops and/or animals, resulting in enhanced complementary relations between components, increasing multiple use of the agroecosystem (Nair, (Nair, 1993) 4. Cover Crops: The use of pure or mixed stands of legumes or other annual plant species under fruit trees for the purpose of improving soil fertility, enhancing biological control of pests, and modifying the orchard microclimate (Finch and Sharp, 1976) 5. Animal Integration: In agroecosystems this aids in achieving high biomass output and optimal recycling If these alternative technologies are used, the possibilities of complementar men tary y int intera eractio ctions ns bet betwee ween n agr agroec oecosy osyste stem m com compo ponen nents ts are enh enhanc anced, ed, resulting in one or more of the following effects: 1. Continuous vegetation cover for soil protection 2. Constant production of food, ensuring a varied diet and several marketing items 3. Closing of nutrient cycles and effective effective use of local l ocal resources 4. Soi Soill and wat water er con conser serv vatio ation n thr throu ough gh mu mulch lching ing and win wind d pro protec tec-tion
5. Enhanced biological pest control through diversification which provides resources to beneficial biota 6. Increased multiple-use capacity of the landscape and/or sustained crop production without relying on environmentally degrading chemical inputs In sum summar mary y, ke key y eco ecolog logica icall pri princi nciple pless fo forr des design ign of di dive versi rsifi fied ed and sustainable agroecosystems include the following following:: 1. Increasing species diversity, as this promotes fuller use of resources (nutrients, radiation, water, etc.), pest protection, and compensatory growth. Many researchers have highlighted the impo im port rtan ance ce of var ario ious us sp spat atial ial an and d te temp mpor oral al pl plan antt co comb mbin inat atio ions ns to facilitate complementary resource use or to provide intercrop advantage such as in the case of legumes facilitating the growth of cereals by supplying extra nitrogen. Compensatory growth is anothe ot herr de desir sirab able le tr trait ait:: as on onee sp spec ecie iess su succ ccum umbs bs to pe pest sts, s, we weat athe herr, or harvest, another species fills the void and maintains full use of available resources. Crop mixtures also minimize risks by creating the sort of vegetati vegetative ve texture that controls specialist pests. 2. Enhance longevity through the addition of perennials that contain a thick canopy, thus providing continual cover that can also protect the soil. Constant leaf fill builds organic matter and allows uninterrupted nutrient circulation. Dense, deep root systems of long-lived woody plants are an effective mechanism for nutri nu trient ent cap captur ture, e, of offse fsettin tting g the ne negat gativ ivee los losses ses thr throu ough gh lea leachi ching ng.. 3. Impose a fallow to restore soil fertility through t hrough biologically mediated mechanisms and to reduce agricultural pest populations as life cycles are interrupted with forest regrowth. regrowth. 4. Enhance additions of organic matter by including high biomassproducing plants. Accumulation of both “active” and “slow fraction” organic matter is key for activating soil biology, imim proving soil structure and macroporosity, and elevating the nutrient status of soils. 5. Increase landscape diversity by having in place a mosaic of agroecosystems representative representative of various stages of succession. Risk of complete failure is spread among, as well as within, the various cropping cropping systems. Improved pest control is also linked to spatial heterogeneity at the landscape level.
ENHANCING SURROUNDING BIODIVERSITY
As discussed in Chapter 7, manipulation of vegetation adjacent to agricultural fields can be key in providing overwintering overwintering sites and alternative food sources for entomophagous arthropods. The impact of such border vegetation or refuges is dependent on its plant composition and the spatial extent of its influence on natural enemy abundance, which is determined by the distance to which natural enemies disperse into the crop (Corbett and Rosenheim, 1996). Other authors have emphasized the importance of landscape complexity, such as large fallows, riparian forests, or other features near crop fields. Research has shown that natural enemy abundance and efficiency increases with landscape heterogeneity while pest damage increases as the percentage of noncrop area in the t he landscape decreases (Thies and Tscharntke, 1999). Such observations observations suggest that in order to maximize the impact of natural enemies through habitat management, agroecologists must look beyond the immediate confines of agricultural lands to include the uncultivated habitats that separate or surround cultivated cultivated fields. Landis and Marino (1996a) argue that in order to effectively conserve natural enemies in early-successional agricultural landscapes the creation and management of mid- to latesuccessional habitats may be required. In essence, this is a process of refragmenting highly disturbed landscapes by adding a network of more stable habitats of varying successional stages. These habitats should serve multiple functions as cross-wind trap strips, filter strips, ripari rip arian an bu buff ffer er zon zones, es, or agr agrofo ofores restry try pro produc ductio tion n sys system temss (La (Land ndis is and Marino, 1996b). There is sufficient information on certain forms of cultural and biological control applicable to some crop pests of known biology. Based on such information, Perrin (1980) advanced a series of env enviironmental management proposals to improve the control of insect pests affecting the cereal-rape system in southern England. Although Perrin suggests some important changes in the design of cereal-rape systems, the protocol does not address some important dilemmas, such su ch as wh whet ethe herr it is de desi sira rabl blee to ha have ve he hedg dger ero ows re remo move ved d or st stop op ae ae-rial spraying of insecticides. Nev Nevertheless, ertheless, Perrin’s Perrin’s proposal is a step in the right direction in that it advances a regional approach where land la ndsc scap apee di dive vers rsit ity y is ma mani nipu pula late ted d in a co coor ordi dina nate ted d ma mann nner er by al alll ag ag-ricultural sectors involved. involved. The possibilities of such cooperation are
not enc not encou ourag raging ing wh when en ant antago agonis nistic tic pr produ oductio ction n and con conser serv vatio ation n vie views ws prevail. prev ail. Such conflicts are well illustrated by the debate on hedges in England, where on the one hand their removal increases the efficiency with which the land can be farmed with modern machinery, but on the other may decrease the local diversity and abundance of birds, insects, and plants. In these cases, an agroecological approach must be developed so that economic, social, and environmental environmental goals are defined by the local rural community, and low-input technologies are implemented to harmonize economic growth, social equity, and environmental preservation (Figure 9.6).
CASE STUDY 1: DIVERSIFICATION OF AN ONION AGROECOSYSTEM IN MICHIGAN
In order to optimize the mortality of the major onion insect pest (onion maggot) in Michigan, a functionally diverse onion agroeco-
FIGURE 9.6. The role of agroecology in satisfying social, environmental, and economic goals in rural areas (after Altieri, 1995).
system was designed. This design was derived from quantitative models describing the relationships among components in the system. From an understanding of these quantitative interactions, designs incorporating diseases, weeds, insects, etc., can be derived as long as the relationships that are used in the construction of these “free-body” models are structure independent or incorporate aspects of structure as a variable. The alternative design of the onion agroecosystem shown shown in Figure 9.7 stresses planned or functional diversity. The cow pasture and weed we edy y bo bord rder erss pr pro ovi vide de al alte tern rnat atee ho host st an and d ne nect ctar ar fo forr th thee on onio ion n ma magg ggot ot parasite Aphaereta pallipes (Groden, 1982). The cow pasture also provides a rich resource for earthworms, thereby potentially maximizing the densities of the tiger ti ger fly predator of onion flies. The long, narrow strips of onions minimize the distance from any point in the onion field and the weedy borders and cow pasture. This is important since A. pallipes numbers decline exponentially from weedy borders and an d co cow w pa past stur ures es in into to th thee on onio ion n fiel ield d (G (Gro rode den, n, 19 1982 82). ). Th This is is al also so tr true ue of on onio ion n fl flies ies in infe fect cted ed wi with th di dise seas asee ca caus used ed by Entomop Entomophthor hthora a muscae muscae.. Weedy field borders are not mowed so that attachment sites for diseased flies are provided. Narrow weedy borders maximize the probability of E . muscae spores encountering healthy flies by crowding in g to toge geth ther er re rest stin ing g an and d at atta tach chme ment nt si site tess fo forr he heal alth thy y fl flies ies du duri ring ng mi middday. By mowing some of the weedy border, this crowding effect can be increased. The planting of radishes adjacent to onions provides an alternate host and thus a continuous food supply for the rove beetle Aleochara bilineata. A number of plantings should be used in order to pr prov ovid idee a se seas ason on-l -lon ong g fo food od re reso sour urce ce fo forr th thee ca cabb bbag agee ma magg ggot ot,, an and da number of different planting dates of onions should be incorporated i ncorporated into the design (Groden, 1982). Groden also showed that earlyplanted onions adjacent to late-planted onions serve as a highly attractive trap crop resulting in a concentration of the onion maggot population in the early planting. Because the later plantings go largely untouched, untouched, the early planting can be positioned near the radish interface so that the hosts pool for A. bilineata is concentrated, thereby making prey search more efficient. In or orde derr to de deal al wi with th th thee pr prob oble lem m of em emer ergi ging ng fl flie iess af afte terr on onio ions ns ar aree harvested, management of cull onions becomes a major issue. A diversification-manag diver sification-management ement option inv involves olves sowing a fall rye or oat cover crop immediately after harvest so that in a week the cover crop
Cow Pasture
Weedy Border
Early Onions Radish Planting 1 Radish Planting 2 Radish Planting 3 Late Onions
Weedy Border
Early Onions Radish Planting 1 Radish Planting 2 Radish Planting 3 Late Onions
Weedy Border
FIGURE 9.7. Sustainable agriculture planting for minimizing the impact of onion maggot and the need for the use of insecticides insecticides for control of this pest (after Groden, 1982).
hidess th hide thee cu cull ll on onio ions ns in th thee fi fiel eld, d, ma maki king ng it di difffi ficu cult lt fo forr th thee on onio ion n fl flie iess to find the culls. A modification is not to harvest a small section of onio on ion n ro rows ws,, an and d th then en,, wh whil ilee so sowi wing ng th thee co cov ver cr crop op,, th thee to tops ps of th thee on on-ions can be cut off and left on the ground. These cut tops are very at tractive to the onion flies (more attractive than cull onions); however, the immature onion fly cannot survive on them because they dry up
before befo re in inse sect ct de deve velo lopm pmen entt ca can n be co comp mple lete ted. d. Th Thus us,, th thee cu cutt to tops ps se serv rvee to keep the onion flies from laying on the culls until the cover crop comes up, and then the searching efficiency of the female flies is drastically reduced. In addition, crop rotation significantly reduces the number of flies colonizing an onion field in the spring (Mortinson, Nyrop, and Eckenroad, et al., 1988).
CASE STUDY 2: A DIVERSIFIED SMALL FARMING SYSTEM IN CHILE
Since 1980, the Centro de Educacion y Tecnologia Tecnologia (CET), a Chilean nongovernmental organization, has engaged in a rural development program aimed at helping resource-poor peasants to achieve year-round food self-sufficiency and rebuild the productive capacities of their small land holdings. The CET’s CET’s approach has been to establish several 0.5 ha model farms where most of the food requirements for a family with scarce capital and land can be met. In this farm system, the critical factor in the efficient use of scarce resources is diversity. Thee CE Th CET T fa farm rm is a di div ver ersi siffie ied d co comb mbin inat atio ion n of cr crop ops, s, tr tree ees, s, an and d an aniimals. In an attempt to t o maximize production efficiencies, efficiencies, the components are structured to minimize losses to the system and promote positivee interactions. Thus, crops, animals, and other farm resources positiv are managed to t o optimize production efficiency efficiency,, nutrient and organic matter recycling, and crop protection. The principal components include the following: • • • • • • •
Vegetables: spinach, cabbage, tomatoes, lettuce, etc. Intercropped maize-beans-potatoes and peas-fava beans Cereals: wheat, oats, and barley Forage crops: clover, alfalfa, and ryegrass Fruit trees: grapes, oranges, peaches, apples, etc. Nonfruit trees: black locust, honey locust, willows, etc. Domestic animals: chickens, pigs, ducks, goats, bees, and dairy cow
The animal and plant components are chosen according to (1) the crop or animal nutritional contributions, (2) their adaptation to local
agroclimatic conditions, (3) local peasant consumption patterns, and (4) market opportunities. The design is also based on cropping patterns, crops, and management techniques practiced by local campesinos (small farmers). In Chile, campesinos typically produce a grea gr eatt var arie iety ty of cr crop opss an and d an anim imal als. s. It is no nott un unus usua uall to fin ind d as ma man ny as five or te ten n tr tree ee cr crop ops, s, te ten n to fif ifte teen en an annu nual al cr crop ops, s, an and d th thre reee to five an aniimal species on a single farm. The physical layout of these model farms varies depending on local conditions; however, most vegetables are produced in heavily composted raised beds located in the garden section, each of which can ca n pr prod oduc ucee up to 83 kg of fr fres esh h vege geta tabl bles es pe perr mo mont nth. h. Th Thee re rest st of th thee vegetables, cereals, legumes, and forage plants are produced in a sixyear rotational system (Figure 9.8). This rotation was designed to provide the maximum variety of basic crops in six plots, taking advanta va ntage ge of the soi soil-r l-resto estorin ring g pro proper pertie tiess of rot rotatio ations. ns. Eac Each h plo plott rec recei eive ved d the following treatments during the six-year period (Figure 9.9): Year 1:
Summer: corn Summer: corn,, bean beans, s, and potat potato o Winter: peas and fava beans Year 2: Summer: Summ er: tomat tomato, o, onio onion, n, and squa squash sh Winter: Wi nter: supp supplemen lementary tary pastu pasture re (oats (oats,, clov clover er,, rye ryegrass grass)) Year 3: Summer: Summ er: soy soybean bean,, peanu peanuts, ts, or sunf sunflow lower er Winter: wheat companion—planted with pasture Year 4-6: Permanent pasture: clover clover,, alfalfa, and ryegrass ryegrass In each plot, crops are grown in several temporal and spatial designs, such as strip cropping, intercropping, mixed cropping, cover crops, and living mulches, optimizing the use of limited resources and enhancing the self-sustaining and resource-conser resource-conserving ving attributes of the system. An important consideration in the rotational design was the stability of the cropping system in terms of soil-fertility maintenance and pest regulation. It is well accepted that a rotation of grains and leguminous forage crops provides significant additional inputs of nitrogen and much higher yields of the subsequent crop of grain obtained with continuous grain monocropping. The output of grain depends on the efficiency of the legumes in supplying nitrogen. Studies in Chile, and elsewhere, have shown that legumes such as swee sw eett clo clove verr, al alfa falf lfa, a, an and d ha hair iry y ve vetc tch h ca can n pr prod oduc ucee be betw twee een n 2. 2.3 3 an and d 10 tons/ha of dry matter and fix between 76 and 367 kg of nitrogen/ha.
FIGURE 9.8. Model design of a self-sufficient farming system, based on a sixyear rotational scheme adaptable to Mediterranean environments (adapted from Altieri, 1987).
FIGURE FIGUR E 9.9. Crop sequence in a six-year rotation design for central central Chile (after Altieri, 1987).
This is sufficient to meet most of the nitrogen requirements of agronomic and vegetable crops. The rotational scheme provides nearly continuous plant cover that aidss in the con aid contro troll of ann annual ual wee weeds. ds. Inc Incorp orpora oratin ting g leg legume ume co cove verr cro crops ps in annual crops, such as corn, cabbage, and tomato, through overseeding and sod-based rotations, has reduced weeds significantly. significantly. In addition, these systems reduce erosion, conserve moisture, and, therefore, offer offer great potential for hillside farmers. The crop rotations scheme also had a profound impact on insect pest populations. For example, the corn rootworm ( Diabrotica spp.) consistently reached higher levels in a continuous corn monoculture than in cornfields following following soybean, clover, alfalfa, or other crops. The presence of alfalfa in the rotational scheme enhanced the abundance and diversity of insect predators and parasites on the farm. Strip cutting of alfalfa forced movement of predators from alfalfa to other crops. Cutting and spreading alfalfa hay, containing high numbers of beneficial insects, throughout the farm also increased natural enem en emy y po popu pula lati tion ons. s. Ce Cere real al re resi sidu dues es us used ed as st stra raw w mu mulc lche hess in th thee su succceeding crops significantly reduced whitefly populations, the principal vector of several viruses, by affecting their attraction to host crops. Spiders, ground beetles, and other predators also were enhanced by the alternative habitat provided by the mulch. CET personnel have closely monitored the performance of this integrated farming system. Throughout Throughout the years, soil fertility has improved (P 2O5 levels, which were initially declining, increased from 5 to 15 ppm) and no serious pest or disease problems have have been noticed. The fruit trees in the t he garden orchard and those surrounding surrounding the rotational plots produce about 843 kg of fruit/year (grapes, quince, pears, plums, etc.). Forage production reaches about 18 tons/0.21 ha per year, milk production averages average s 3,200 L per year, and egg production reaches a level of 2,531 units. A nutritional analysis of the system based on the production of the various plant and animal components (milk, eggs, meat, fruits, vegetables, honey, etc.) shows that the system produces a 250 percent surplus of protein, 80 percent and 550 percent surplus of vitamin A and C respectively, and a 330 percent surplus of calcium. A household economic analysis indicates that given a list of preferences, the balance between selling surpluses and buy uyin ing g pr pref efer erre red d it item emss is a ne nett in inco come me of US US$7 $790 90.. If al alll th thee fa farm rm ou outt-
put is so put sold ld at wh whol oles esal alee pr pric ices es,, th thee fa fami mily ly co coul uld d ge gene nera rate te a ne nett in inco come me of US$1,635, equivalent equivalent to a monthly income of US$136, 1.5 times greater than the monthly legal minimum wage in Chile (Yurjevic, 1991).
Conclusion Conclusion Typical commercial-production agriculture has resulted in the simplification of cropping systems in general. The expansion of monocultures has decreased abundance and activity of natural enemies due to the removal of critical food resources and overwintering sites (Corbett and Rosenheim, 1996). Many scientists are concerned that with accele acceleratin rating g rates of habit habitat at remo removal val,, the contr contribu ibution tion to pest suppression by biocontrol agents using these habitats will decline (Fry, 1995; Sotherton, 1984). For this reason, many researchers cited in this book have proposed options at the field and landscape level to rectify this decline by increasing the vegetational diversity diversity of agroecosystems and surrounding areas. This Th is bo book ok ha hass ex exam amin ined ed hu hund ndre reds ds of st stud udie iess wh whic ich h sh sho ow th that at co commplementary interactions occur between crops and/or between crops and weeds grown in polycultures and between adjacent cultivated and uncultivated vegetational components of agroecosystems. These intera int eracti ctions ons can ha have ve po posit sitiv ivee or ne negat gativ ive, e, dir direct ect or ind indire irect ct ef effec fects ts on thee bi th biol olog ogica icall co cont ntro roll of sp spec ecif ific ic cr crop op pe pest sts. s. Th Thee ex expl ploi oitat tatio ion n of th thes esee interactions in real situations involves agroecosystem design and manage man agemen mentt and req requir uires es an und unders erstan tandin ding g of the num numero erous us rel relatio ationnships among plants, herbivores, and natural enemies (Altieri and Letourneau, 1982). Clearly, the emphasis of this approach is to restore sto re nat natur ural al con contro troll mec mechan hanism ismss thr throu ough gh the add additio ition n of sel select ectiv ivee di di-versity,, rather than forcing the establishment of biological control in versity simplified environments environments (such as monocultures) where the essential ecolog eco logica icall ele elemen ments ts are lac lackin king g to allo allow w for opt optimu imum m per perfor forman mance ce of natural enemies (Van Driesche and Bellows, 1996). A key strategy in agroecology is to exploit the complementarity and synergy that result from the various combinations combinations of crops, trees, and animals in agroecosystems that feature spatial and temporal arrangements such as polycultures, agroforestry systems, and croplivestock mixtures. This implies identifying the type of biodiversity that is desirable to maintain and/or enhance in order to carry out ecological services, and then to determine the best practices that will en-
courage the desired biodiv biodiversity ersity components. Many agricultural practices and designs have the potential to enhance functional biodiv biodiversity ersity and others negatively negatively affect it. The idea is to apply the best managemanagement practices in order to enhance or regenerate the kind of biodiversity diver sity that can best subsidize the sustainability of agroecosystems by providing ecological services such as biological pest control, nutrient cycling, water and soil conservation, and so forth. The role of agroecologists should be to encourage those agricultural practices that increase increa se the abundance and diversity of aboveground and belowground organisms, which in turn provide key ecological services to agroecosystems (Altieri and Nicholls, 2000). In order for this diversification strategy to be more rapidly implemented, it is necessary to develop a much better understanding of the ecology of parasitoids and predators within and outside of the cultivated habitat, identifying those resources that are necessary for their survivorship and reproduction (Gurr, Van Emden, and Wratten, 1998). It is al also so im impo port rtan antt to de dete term rmin inee to wh what at ex exte tent nt po popu pula latio tions ns wi with thin in th thee crop cro p con contri tribu bute te to the ov overa erall ll nat natura urall ene enemy my met metapo apopul pulati ation on in sub subsesequent qu ent yea years. rs. If the these se con contri tribu butio tions ns are min minor or,, the then n in inve vestm stment entss in hab hab-itat management should be oriented specifically to increasing the sour so urce ce po popu pula lati tion onss ou outs tsid idee th thee cr crop op to en ensu sure re a gr grea eate terr nu numb mber er of im im-migrants each year, an action parallel to increasing the dosage of a chemic che mical al bio biocid cide. e. Ho Howe weve verr, if the sub subpo popul pulati ations ons wit within hin the cro cropp ppin ing g system contribute significantly to the year-to-year metapopulation dynamics, then habitat modifications should consider not only tactics fostering immigration into the crop but also those augmenting the proba pr obabil bility ity of suc succes cessio sional nal emi emigra gratio tion n wh when en thi thiss hab habita itatt bec become omess un un-suitable. Such actions could include the addition of plant species to provide alternate hosts and/or food sources, the addition of habitats as sui suitab table le ov overw erwint interi ering ng site sitess or the pr prov ovisi ision on of cor corrid ridors ors wit within hin the cropping system to facilitate movemen movementt between the subcomponents of the metapopulation. In su summ mmar ary y, th ther eree ar aree fo four ur ke key y is issu sues es to co cons nsid ider er wh when en im impl plem emen entting habitat management: 1. The selection of the most appropriate plant species and their spatial/temporal deployment 2. The predator-parasitoid behavioral mechanisms which are inin fluenced by the manipulation
3. The spatial scale over which the habitat enhancement operates 4. The potential negative negative aspects associated with adding new plants to the agroecosystem (Landis, Wratten, and Gurr, 2000) Obviously,, proposed habitat-management techniques Obviously t echniques must fit existing cropping systems and adapt to the needs and circumstances of farmers. Prokopy Prokop y (1994) has cautioned that all diversification interactions should be evaluated within the context of a broader integrated management program of the agricultural crop. The reason for caution is that potential benefits may be less than unforeseen costs. For example, although blackberry plants in California serve as a host plant for altern alt ernati ative ve hos hosts ts of the par parasi asitoi toid d A thes esee sa same me pl plan ants ts co coul uld d be a A.. ep epos os th reservoir for the bacterium responsible for Pierce’ Pierce’ss disease, a serious disease of grapes transmitted by the blue-green sharpshooter. sharpshooter. Thus, an ac actio tion n ta take ken n to in incr crea ease se th thee ef efffic icac acy y of na natu tura rall en enem emie iess co coul uld d in incu curr losses through increased levels of disease. However, actions to reduce Pierce’s disease, such as removal of sharpshooter host plants within wit hin rip ripari arian an hab habitat itatss as rec recomm ommend ended ed in Cali Califo forni rnia’ a’ss Nap Napaa Valle alley y, can in turn lead to a reduction of natural enemies of grape leafhoppers, compounding compounding a minor pest problem. Clearly then, it is important to provide the right kind of diversity. At times the addition of one or two plants is all it may take. Recent studies conducted in grassland systems suggest that there are no simple links between species diver diversity sity and ecosystemic stability.. What is apparent is that functional characteristics of component ity spec sp ecie iess ar aree as im impo port rtan antt as th thee to tota tall nu numb mber er of sp spec ecie ies. s. Re Rece cent nt ex expe perriments with grassland plots conclude that functionally different roles represented by plants are at least as important as the total number of specie spe ciess in det determ ermini ining ng pro proces cesses ses and ser servic vices es in eco ecosys system temss (T (Tilm ilman, an, Wedin, and Knops, 1996). This latest finding has practical implications for agroecosystem management. If it is easier to mimic specific ecosystem processes rather than duplicate all the complexity of nature, then the focus should be placed on incorporating a specific biodiversity biodiversity component that plays a specific role, such as a plant that fixes nitrogen, nitrogen, provides cover for soil protection, or harbors resources for natural enemies. In the case of farmers without major economic and resource limitations who can afford a certain risk of crop failure, a crop rotation or a sim-
ple crop association may be all it takes to achieve a desired level of stability. In the case of resource-poor farmers, where crop failure is intolerable, highly diverse polyculture systems would probably be the best choice. The obvious reason is that the benefit of complex agro ag roec ecos osys yste tems ms is lo low w ri risk sk;; if a sp spec ecies ies fa fall llss to di dise seas ase, e, pe pest st at atta tack ck,, or weather, another species is available to fill the void and maintain a full use of resources. The cen centra trall iss issue ue in sus sustain tainabl ablee agr agricu icultu lture re is not ach achie ievin ving g max maximu imum m yield—it is long-term stabilization (Reganold et al., 2001). Sustaining agricultural productivity productivity will require more than t han a simple modification of traditional ad hoc techniques. The dev development elopment of self-sufficient, diversified, economically viable agroecosystems comes from novel designs of cropping and/or livestock systems managed with technologie og iess ad adap apte ted d to th thee lo loca call en envi viro ronm nmen entt th that at ar aree wi with thin in th thee fa farm rmer ers’res’resources. Energy and resource conservation, environmental quality, public health, and equitable socioeconomic development should be considered in making decisions on crop species, rotations, row spacing, fertilizing, pest control, and harvesting. Many farmers will not shift to alternative systems unless there is a good prospect for monetary gain, brought about by either increased output or decreased production costs. Different Different attitudes will depend primarily on farmers’’ per ers percep ceptio tions ns of the sho shortrt-ter term m and lon longg-ter term m eco econo nomic mic ben benef efits its of sustainable agriculture. Restoration of natural controls in agroecosystems through vegetavegetation management not only regulates pests but also helps to conserve energy, improve soil fertility, minimize risks, and reduce dependence on external resources. The ultimate goal of agroecological design is to integrate components so that overall biological efficiency is improved, prov ed, biodiversity is preserved preserved,, and the agroecosystem productivity and its self-sustaining capacity is maintained. The goal is to design a quilt of agroecosystems within a landscape unit, each mimicking the structure and function of natural ecosystems, that is, systems that include the followin following: g: 1. Veg egeta etati tive ve co cove verr as an ef effec fecti tive ve soi soill- and wat water er-co -conse nservi rving ng mea mea-sure, met through the use of no-till practices, mulch farming, and use of cover crops and other appropriate methods 2. A regular supply of organic matter through the regular addition of or orga ganic nic mat matter ter (ma (manu nure, re, pla plant nt bio biomas mass, s, com compos post, t, and pr promo omo-tion of soil biotic activity)
3. Nutrient-recycling mechanisms through the use of crop rotations, crop-livestock crop-livestock systems based on legumes, etc. 4. Pest regulation assured through enhanced activity activity of biological control agents achieved by introducing and/or conserving through vegetationall designs, natural enemies, and antagonists vegetationa This is particularly important i mportant in underdeveloped underdeveloped countries where sophis ph istic ticat ated ed in inpu puts ts ar aree ei eith ther er no nott av avai ailab lable le or ma may y no nott be ec econ onom omic ical ally ly or environmentally advisable, especially to resource-poor farmers. Further research in this area should provide an ecological basis for the design of diverse, pest-stable, self-sustained agroecosystems. These systems are urgently needed worldwide in an era of deteriorating environmental quality, a worsened energy situation, and escalating in g in inpu putt co cost sts. s. Th This is ap appr proa oach ch to ag agri ricu cult ltur uree wi will ll be pr prac acti tica call on only ly if it is ec econ onom omica icall lly y se sens nsib ible le an and d ca can n be ca carr rrie ied d ou outt wi with thin in th thee co cons nstr trai aint ntss of a fairly normal agricultural-management system. Howe However ver,, given the trend toward large-scale, specialized farm production throughout the world, objectively there is not much room left for a fair implementation of a regional insect-habitat management program. Emerging biotechnological approaches such as transgenic crops deployed in more than 40 million hectares in 2000 are leading agriculture toward further specialization, and the potential effects of transgenic crops on nontarget beneficial organisms is of concern to biological control practitioners (Rissler and Mellon, 1996; Hilbeck et al., 1998; Altieri, 2000). Long-term maintenance of diversity requires a management strategy that considers regional biogeography and landscape patterns, as welll as des wel design ign of en envir vironm onment entall ally y sou sound nd agr agroec oecosy osyste stems ms abo above ve pur purely ely economic concerns. This is why several authors have repeatedly questioned whether the pest problems of modern agriculture can be ecologically alleviated within the context of the present capitalintensive structure of agriculture. Buttel (1980) suggests that many problems of modern agriculture are rooted within that structure and calls for the consideration of major social change, land reform, redesign of machinery, machinery, research, and extension reorientation in the agricultural sector to increase the possibilities of improv improved ed pest control throu thr ough gh ve veget getatio ation n man manage agemen ment. t. Whe Whethe therr the po poten tentia tiall and spr spread ead of ecologically based pest management is realized will depend on policies, attitude changes on the part of researchers and policymakers,
existence of markets for organic produce, and also farmer and consumer sum er mo move vemen ments ts tha thatt dem demand and a mo more re hea health lthy y and via viable ble agr agricu icultu lture. re. It is cr cruc ucia iall th that at sc scie ient ntis ists ts in invo volv lved ed in th thee se sear arch ch fo forr su sust stai aina nabl blee ag ag-ricultural technologies be concerned about who will ultimately benefit from them. Thus, what is produced, how it is produced, and for whom it is produced are key questions that need to be addressed if a socially equitable agriculture is to emerge. When such questions are exami ex amined ned,, iss issues ues of lan land d ten tenur ure, e, lab labor or,, app appro ropri priate ate tec techno hnolog logy y, pub public lic health, and research policy unavoidably unavoidably arise. Undoubtedly these are key challenges to deal with when developing a sustainable agriculture in the twenty-first century. Such issues must be urgently addressed by committed scientists and farmers in a true partnership.
References References
Adams, Adam s, J. B. an and d M. E. Dr Drew ew.. 19 1965 65.. Gr Grai ain n ap aphi hids ds in Ne New w Br Brun unsw swic ick. k. II II.. Ap Aphi hid d po poppulations in herbicide-treated oatfields. Canadian Journal of Zoology 43: 789794. Adams, M. W., A. H. Ellingbae, and E. C. Rossineau. 1971. Biological uniformity and disease epidemics. BioScience 21: 1067-1070. Aldrich, R. J. 1984. Weed-crop ecology—Principles in weed management. Breton Publishers, North Scituate, MA. Alghali, A. M. 1993. Intercropping as a component in insect pest management for Insectt Scien Science ce and Appli Applicatio cations ns 14: 49grain gra in cow cowpea pea pro produc ductio tion n in Nig Nigeri eria. a. Insec 49-54. 54. Ali, A. D. and T. E. Reagan. 1985. Vegetation Vegetation manipulation impact on predator and prey populations in Louisiana sugar-cane ecosystems. Journal of Economic Entomology 7-8: 1409-1414. Ali, Al i, M. I. an and d M. A. Ka Kari rim. m. 19 1989 89.. Th Thee us usee of tr trap ap cr crop op in ma mani nipu pula lati ting ng po popu pula lati tion on of the cotton jassid on cotton. Bangladesh Journal of Zoology 17: 159-164. Allen, W. W. and R. F. Smith. 1958. Some factors influencing the efficiency of Apanteles medicagines Muesebeck (Hymenoptera: Braconidae) as a parasite of the alfalfa caterpillar, Colias philodice eurytheme Boisduval. Hilgardia 28: 1. Altier Alt ieri, i, M. A. 198 1984. 4. Pat Patter terns ns of ins insect ect div divers ersity ity in mon monocu ocultu ltures res and pol polycu ycultu ltures res of brussels sprouts. Protection Ecology 6: 227-232. Altieri, M. A. 1987. Agroecology: The scientific basis of alternative agriculture. Westview Press, Boulder, CO. Altieri, M. A. 1991a. Ecology of tropical herbivores in polycultural agroecoPlant-anima -animall intera interaction ctions: s: Evolut Evolutionary ionary ecology in tropic tropical al and systems. In Plant temperate regions, pp. 607-617. P. W. Price, T. M. Lewinsohn, G. W. Fernandez, and W. W. Benton, eds. Wiley, New York. Altier Alt ieri, i, M. A. 199 1991b. 1b. How bes bestt can we use bio biodiv divers ersity ity in agr agroec oecosy osyste stems? ms? Outlook on Agriculture 20: 15-23. Altieri, M. A. 1991c. Increasing biodiversity to improve insect pest management in agro-ecosystems. In Biodiversity of microorganisms and invertebrates: Its role in sustainable agriculture, pp. 165-182. D. L. Hawksworth, ed. CAB Int., Wallingford, UK. Altieri, M. A. 1991d. Traditional farming in Latin America. The Ecologist 21: 93-96. Altieri, M. A. 1995. Agroecology: The science of sustainable agriculture. Westview Press, Boulder, CO.
Altier Alt ieri, i, M. A. 199 1999. 9. The eco ecolog logica icall rol rolee of bio biodiv divers ersity ity in agr agroec oecosy osyste stems. ms. Agriculture, Ecosystems and Environment 74: 19-31. Altieri, M. A. 2000. The ecological impacts of transgenic crops on agroecosystem health. Ecosystem Health 6: 13-23. Altieri, M. A. and J. D. Doll. 1978. Some limitations of weed biocontrol in tropical crop ecosystems in Colombia. In Proceedings IV International Symposium on Biological Control of Weeds, pp. 74-82. T. E. Freeman, ed. University of Florida, Gainesville, FL. Alti Al tier eri, i, M. A. A.,, C. A. Fr Fran anci cis, s, A. Sc Scho hoon onho hove ven, n, an and d J. Do Doll ll.. 19 1978 78.. A re revi view ew of in inse sect ct preval pre valenc encee in mai maize ze ( Ze L.)) an and d be bean an (Phaseo polycultura lturall Zea a may mayss L. Phaseolus lus vulga vulgaris ris L.) polycu systems. Field Crops Research 1: 33-49. Altieri, M. A. and S. R. Gliessman. 1983. Effects of plant diversity on the density and herbivory of the flea beetle, Phyllotreta cruciferae Goeze, in California collard (Brassica oleracea) cropping systems. Crop Protection 2: 497-501. Altieri, M. A. and S. B. Hecht. 1991. Agroecology and small farm development. CRC Press, Boca Raton, FL. Altieri, M. A. and D. K. Letourneau. 1982. Vegetation management and biological control in agroecosystems. Crop Protection 1: 405-430. Altieri, M. A. and D. K. Letourneau. 1984. Vegetation diversity and insect pest outbreaks. CRC Critical Reviews in Plant Sciences 2: 131-169. Altieri, M. A., W. J. Lewis, D. A. Nordlund, R. C. Gueldner, and J. W. Todd. 1981. Chemical interactions between plants and Trichogramma wasps in Georgia soybean fields. Protection Ecology 3: 259-263. Altier Alt ieri, i, M. A. and M. Lie Liebma bman. n. 198 1988. 8. Wee Weed d man manage agemen ment: t: Eco Ecolog logica icall gui guidel deline ines. s. In 183-218. 8. Weed manage management ment in agroec agroecosyste osystems: ms: Ecolo Ecological gical approa approaches, ches, pp. 183-21 M. A. Altieri and M. Z. Liebman, eds. CRC Press, Boca Raton, FL. Altieri, M. A. and C. I. Nicholls. 2000. Applying agroecological concepts to development of ecologically based pest management systems. In Proceedings of a Workshop “Professional societies and ecological based pest management systems,” pp. 14-19. National Research Council, Washington, DC.
Altieri, M. A. and L. L. Schmidt. 1985. Cover crop manipulation in northern California orchards and vineyards: Effects on anthropod communities. Biological Agriculture and Horticulture 3: 1-24. Altieri, M. A. and L. L. Schmidt. 1986a. The dynamics of colonizing arthropod communities at the interface of abandoned organic and commercial apple orchards and adjacent woodland habitats. Agriculture, Ecosystems and Environment 16: 29-43. Altieri, M. A. and L. L. Schmidt. 1986b. Population trends and feeding preferences of flea beetles ( Phyllotreta cruciferae Goeze) in collard-wild mustard mixtures. Crop Protection 5: 170-175. Altieri, M. A. and L. L. Schmidt. 1987. Mixing cultivars of broccoli reduces cabbage aphid populations. California Agriculture 41: 24-26.
Bach, C. E. 198 Bach, 1980a. 0a. Eff Effect ectss of pla plant nt den densit sity y and div divers ersity ity on the pop popula ulatio tion n dyna dynamic micss of a specialist herbivore, the striped cucumber beetle, Acalymma vittata (Fab.). Ecology 61: 1515-1530. Bach, C. E. 1980b. Effects of plant diversity and time of colonization on an herbivore-plant interaction. Oecologia 44: 319-326. Baliddawa, C. W. 1985. Plant species diversity and crop pest control: An analytical review. Insect Science and Applications 6: 479-487. Bantilan, R. T., M. C. Palada, and R. R. Harwood. 1974. Integrated weed management. I. Key factors affecting crop weed balance. Philippine Weed Science Bulletin 1: 14-36. Conservatio vation n biolog biological ical contro control. l. Aca Barbosa Bar bosa,, P., ed. 199 1998. 8. Conser Academ demic, ic, San Die Diego, go, CA. Barney, R. J., W. O. Lamp, E. J. J . Armbrust, and G. Kapusta. 1984. Insect predator community and its response to weed management in spring-planted alfalfa. Protection Ecology 6: 23-33. Baudry Bau dry,, J. 198 1984. 4. Eff Effect ectss of lan landsc dscape ape str struct ucture ure on bio biolog logica icall com commun muniti ities: es: The cas casee of heterogenous network landscapes. In Methodology in landscape ecological research and planning, Volume I, pp. 55-65. J. Brandt and P. Agger, eds. Roskilde University Center, Roskilde, Denmark. Beer, J., R. Muschler, D. Kass, and E. Somarriba. 1997. Shade management in coffee and cacao plantations. American Society of Agronomy. (Symposium on Tropical Agroforestry Indianapolis, Indiana, November 5, 1996.) In Forestry sciences: Directions in tropical agroforestry research, pp. 139-164. P. K. R. Nairr and C. R. Lat Nai Latt, t, eds eds.. Klu Kluwer wer Aca Academ demic ic Pub Publis lisher hers, s, Dor Dordre drecht cht,, Net Nether herlan lands. ds. Beet Be ets, s, W. C. 19 1990 90.. Ra Raisi ising ng and sust sustain aining ing pro produc ductiv tivity ity of sma small ll hol holder der far farmin ming g systems in the tropics. AgBe Publishing, Alkmaar, Holland. annota otated ted bib biblio liogra graphy phy of wee weeds ds as rese reserrBendi Be ndixe xen, n, L. E. an and d D. J. Ho Horn rn.. 19 1981 81.. An ann voirs for organisms affecting crops. III. Insects. Agricultural Research and Development Center, Wooster, OH. Bentley, S. and J. B. Whittaker. 1979. Effects of grazing by a chrysomelid beetle, Gastrophysa viridula, on competition between Rumex obtusifolius and Rumex crispus. Journal of Ecology 67: 79-90. Bhatnagar, V S. and J. C. Davies. 1981. Pest management in intercrop subsistence farming. Proceedings Int. Workshop on Intercropping. ICRISAT, Patancheru, India. Bigger, M. 1981. Observations on the insect fauna of shaded and unshaded Amelonado cocoa. Bulletin of Entomological Research 71 (1): 107-119. Bobb, M. L. 1939. Parasites of the oriental fruit moth in Virginia. Journal of Economic Entomology 32: 605. Boller, E. F. 1992. The role of integrated pest management in integrated production of viticulture in Europe. In Brighton Crop Protection Conference, pp. 499-506. British Crop Protection Council, Brighton, England. Stat atee of th thee Brow Br own, n, L. R. an and d J. E. Yo Youn ung. g. 19 1990 90.. Fe Feed edin ing g th thee wo worl rld d in th thee ni nine neti ties es.. In St world, pp. 59-78. L. R. Brown et al., eds. W. W Norton & Co., New York.
Brush, S. B. 1982. The natural and human environment of the central Andes. Mountain Research and Development 2: 14-38. Buchanan, G. A. 1977. Weed biology and competition. In Research methods in weed science, pp. 25-41. B. Truelove, ed. Southern Weed Science Society, Auburn, AL. Bugg, R. L. and J. D. Dutcher. 1989. Warm-season cover crops for pecan orchards: Biologica ogicall Agric Agriculture ulture and HortiHorticultu Horti cultural ral and entom entomologi ological cal impli implicatio cations. ns. Biol culture 6: 123-148. Bugg Bu gg,, R. L. an and d C. Wa Wadd ddin ingt gton on.. 19 1994 94.. Us Usin ing g co cove verr cr crop opss to ma mana nage ge ar arth thro ropo pod d pe pest stss of orchards: A review. Agriculture, Ecosystems and Environment 50: 11-28. Burn, A. J. 1987. Cereal crops. In Integrated pest management, pp. 209-256. A. J. Burn, T. H. Coaker, and P. C. Jepson, J epson, eds. Academic Press, London. Integrated rated pest managem management. ent. Burn Bu rn,, A. J. J.,, T. H. Co Coak aker er,, an and d P. C. Je Jeps pson on,, ed eds. s. 19 1987 87.. Integ Academic Press, London. Buschman, L. L., H. N. Pitre, and H. F. Hodges. 1984. Soybean cultural practices: Effect Eff ectss on pop popula ulatio tions ns of geo geocor corids ids,, nab nabids ids,, and oth other er soy soybea bean n art arthro hropod pods. s. Environmental Entomology 13: 305-317. Buttel, F. H. 1980. Agricultural structure and rural ecology: Toward a political economy of rural development. Sociologia Ruralis 20: 44-62. Campbell, R. 1989. Biological control of microbial plant pathogens. Cambridge University Press, Cambridge, UK. Chiverton, P. A. 1989. The creation of within-field overwintering sites for natural enemies of cereal aphids. In Brighton Crop Protection Conference—Weeds, pp. 1093-1096. British Crop Protection Council, Farn, Surrey. Chiverton, P. A. and N. W. Sotherton. 1991. The effects on beneficial arthropods of the exclusion of herbicides from cereal crop edges. Journal of Applied Ecology 28: 1027-1039. Churnakova, B. M. 1960. Supplementary feeding as a factor increasing the activity Vsesoyuznogo uznogo Nauchn Nauchno-issl o-issledova edovatelsc telscogo ogo of par parasi asites tes of har harmfu mfull ins insect ects. s. Trudy Vsesoy Instituta Zashchity Rastenii 15: 57-70. Coll, M. 1998. Parasitoid activity and plant species composition in intercropped systems. In Enhancing biological control. C. Pickett and R. Bugg, eds. University of California Press, Berkeley. Coll, M. and D. G. Botrell. 1994. Effect of nonhost plants on an insect herbivore in diverse habitats. Ecology 75: 723-731. Coll Co ll,, M. an and d D. G. Bo Botr trel ell. l. 19 1996 96.. Mo Move veme ment nt of an in inse sect ct pa para rasi sito toid id in si simp mple le an and d di di-verse plant assemblages. Ecological Entomology 21: 141-149. Collins, Colli ns, F. L. and S. J. Johnson. 1985. Reproduct Reproductive ive respon response se of caged adult velve velvettbean caterpillar and soybean looper to the presence of weeds. Agriculture, Ecosystems and Environment 14: 139-149. Journal rnal for Farm Farming ing Conway, Conw ay, G. R. 1994 1994.. Sus Sustai tainabi nabilit lity y in agr agricul icultura turall dev develo elopme pment. nt. Jou Systems and Research Extension 4: 1-14.
Conway, G. R. and J. N. Pretty. 1991. Unwelcome harvest: Agriculture and pollution. Earthscan Publications, London. Coombes, D. S and N. W. Sotherton. 1986. The dispersal and distribution of polyphagous predatory Coleoptera in cereals. Annals of Applied Biology 108: 461-474. Corbett, A. and R. E. Plant. 1993. Role of movement in the response of natural enemies to agroecosystem diversification: A theoretical evaluation. Environmental Entomology 22: 519-531. Corbett, A. and J. A. Rosenheim. 1996. Impact of a natural enemy overwintering refuge and its interaction with the surrounding landscape. Ecological Entomology 21: 155-164. Cost Co stel ello lo,, M. J. an and d M. A. Al Alti tier eri. i. 19 1994 94.. Li Livi ving ng mu mulc lche hess su supp ppre ress ss ap aphi hids ds in br broc occo coli li.. California Agriculture 48: 24-28. Costello, M. J. and M. A. Altieri. 1995. Abundance, growth rate and parasitism of Brevicoryne brassicae and Myzus persicae (Homoptera: Aphididae) on broccoli Agriculture, lture, Ecosyst Ecosystems ems and Environ Environment ment 52: 187 grown grow n in livi living ng mul mulche ches. s. Agricu 187-196 -196.. Costello, M. J. and K. M. Daane. 1998. Influence of groundcover on spider populations in a table grape vineyard. Ecological Entomology 23: 33-40. Croft, B. A. 1975. Tree fruit pest management. In Introduction to insect pest management, pp pp.. 47 4711-50 507. 7. R. L. Me Metc tcal alff an and d W. H. Lu Luck ckma mann nn,, ed eds. s. J. Wi Wile ley y & So Sons, ns, New York. Croft, B. A. and S. C. Hoyt. 1983. Integrated management of insect pests of pome and stone fruits. J. Wiley & Sons, New York. Cromartie, W. J. 1981. The environmental control of insects using crop diversity. In CRC handbook of pest management in agriculture, Volume 1, pp. 223-251. D. Pimentel, ed. CRC Press, Boca Raton, FL. Daane, K. M. and M. J. Costello. 1998. Can cover crops reduce leafhopper abundance in vineyards? California Agriculture 52: 27-32. Daane, K. M., M. J. Costello, G. Y. Yakota, and W. J. Benteley. 1998. Can we manipulate leafhopper densities with management practices? Grape Grower 30: 18-36. Dambach, C. A. 1948. Ecology of crop field borders. Ohio State University Press, Columbus, OH. De Loach, C. J. 1970. The effect of habitat diversity on predation. Proceedings of Talll Tim Tal Timber berss Con Confer ferenc encee on Eco Ecolog logica icall Ani Animal mal Con Contro troll by Hab Habita itatt Man Manage agemen ment t
2: 223-241. Demp De mpst ster er,, J. P. 19 1969 69.. So Some me ef effe fect ctss of we weed ed co cont ntro roll on th thee nu numb mber erss of th thee sm smal alll ca cabbbage white ( Pieris rapae L.) on brussels sprouts. Journal of Applied Ecology 6: 339-405. Dempster, J. P. and T. H. Coaker. 1974. Diversification of crop ecosystems as a means of controlling pests. In Biology in pest and disease control, pp. 106-114. D. P. Jones and M. E. Solomon, eds. Wiley & Sons, New York.
Francis, C. A. 1990. Sustainable agriculture in temperate zones. Wiley & Sons, New York. Francis, C. A., C. A. Flor, and S. R. Temple. 1976. Adapting varieties for intercropped systems in the tropics. In Multiple cropping, pp. 235-254. ASA Special Publication No. 27. G. B. Triplett, P. A. Sanchez, and R. I. Papendick, eds. American Society of Agronomy, Madison, WI. Frank, T. and W. Nentwig. 1995. Ground dwelling spiders (Araneae) in sown weed strips and adjacent fields. Acta Oecologia 16: 179-193. Fry,, G. 199 Fry 1995. 5. Lan Landsc dscape ape eco ecolog logy y of ins insect ect mov moveme ement nt in ara arable ble eco ecosys system tems. s. In Ecology and int integr egrate ated d far farmin ming g sys system tems, s, pp pp.. 17 1777-20 202. 2. D. M. Gl Glen en,, M. P. Gr Grea eaus us,, an and d H. M. Anderson, eds. John Wiley & Sons, Briston, UK. Fye, A. E. 1983. Cover crop manipulation manipulation for building pear psylla (Homo (Homoptera ptera:: Psyllidae) predator population in pear orchards, Journal of Economic Entomology 76: 306-310. Garcia, M. A. and M. A. Altieri. 1992. Explaining differences in fleabeetle Phyllotreta cruciferae Goeze densities in simple and mixed broccoli cropping systems as a function of individual behavior. Entomologia Experimentalis et Applicata 62: 201-209. Girma, Gir ma, H., M.R M.R.. Rao Rao,, and S. Sit Sithan hanant antham ham.. 200 2000. 0. Ins Insect ect pes pests ts and ben benefi eficia ciall art arthro hro-pod populations under different hedgerow intercropping systems in semiarid Kenya. Agroforestry Systems 50 (3): 279-292. Gliessman, S.R. 1999. Agroecology: Ecological processes in agriculture. Ann Arbor Press, Ann Arbor, MI. Gliessman, S. R. and M. A. Amador. 1980. Ecological aspects of production in traditional agroecosystems in the humid lowland tropics of Mexico. In Tropical ecology ecolo gy and devel development opment,, pp. 601 601-60 -608. 8. J. I. Fur Furtad tado, o, ed. IST ISTE, E, Kua Kuala la Lam Lampur pur.. Gold, C. S. 1987. Crop diversificatio diversification n and tropical herbivores: herbivores: Effects of inter inter- Aleurotrac rotrachelus helus social socialis is croppi cro pping ng and mix mixed ed var variet ieties ies on the cas cassav savaa whi whitef teflie lies, s, Aleu Rialeurode eurodess varia variabilis bilis in Colomb and Rial Colombia. ia. Unpubl Unpublished ished disser dissertatio tation. n. Unive University rsity of California at Berkeley. Gold, C. S., M. A. Altieri, and A. C. Bellotti. 1989a. The effects of intercropping and mixed varieties of predators and parasitoids of cassava whiteflies (Hemip Bulletin in of Entomol Entomological ogical Researc Research h 79: 115 tera: ter a: Ale Aleyro yrodida didae) e) in Colu Columbi mbia. a. Bullet 115-121 -121.. Gold, C. S., M. A. Altieri, and A. C. Bellotti. 1989b. Relative oviposition rates of Erinnys ys ello (Le the cas cassav savaa hor hornwo nworm, rm, Erinn (Lep.: p.: Sph Sphing ingida idae) e) and acc accomp ompany anying ing par paraa sitism by Telenomus sphingis (Hym.: Scelionidae) on upper and lower surfaces of cassava. Entomophaga 34: 73-76. Gold, C. S., M. A. Altieri, and A. C. Bellotti. 1990. Direct and residual effects of short duration intercrops on the cassava whiteflies Aleurotrachelus socialis and Dialeurodes variabilis (Homoptera: Aleyrodidae) in Colombia. Agriculture, Ecosystems and Environment 32: 57-67. Goodman, D. 1975. The theory of diversity-stability relationships in ecology. Quarterly Review of Biology 50: 237-266.
Gould, F. 1986. Simulation models for predicting durability of insect-resistant germplasm: A deterministic diploid, two-locus model. Environmental Entomology 15: 1-10. Grigg, D. B. 1974 . The agricultural systems of the world. An evolutionary ap proach. Cambridge University Press, Cambridge, MA. Groden, E. 1982. The interactions of toot maggots and two parasitoids, Aleochara bilineata (Gyll.) and Aphaereta pillipes (Say). MS thesis. Michigan State University, East Lansing, MI. Guharay, F., J. Monterrey, D. Monterroso, and C. Staver. 2000. Manejo integrado deplagos en el cultivo de café. CATIE, Managua, Nicaragua. Gurr, G. M., H. F. Van Emden, and S. D. Wratten. 1998. Habitat manipulation and natural enemy efficiency: Implications for the control of pests. In Conservation biological control, pp. 155-183. P. Barbosa, ed. Academic Press, New York. Gut, L. J., C. E. Jochums, P. H. Westigard, and W. J. Liss. 1982. Variation in pear psylla (Psylla pyricola Foerster) densities in southern Oregon orchards and its implications. Acta Horticulture 124: 101-111. Hagen, K. S. 1986. Ecosystem analysis: Plant cultivars (HPR) entomophagous species and food supplements. In Interactions of plant resistance and parasitoids and predators of insects, pp. 151-195. D. J. Boethal and R. D. Eikenbary, eds. Ellis Harwood, Chichester, England. Harlan, J. R. 1975. Our vanishing genetic resources. Science 188: 618-622. Hart, R. D. 1974. The design and evaluation of a bean, com and manioc polyculture cropping system for the humid tropics. PhD dissertation. University of Florida, Gainesville, FL. Hart, R. D. 1980. Agroecosistemas. CATIE, Turrialba, Costa Rica. Harwood, R. R. 1974. Farmer-oriented research aimed at crop intensification. In Proceedings of cropping systems workshop, IRRI, pp. 7-12. International Rice Research Institute, Los Banos, Philippines. Harwood, R. R. 1979. Small farm development—Understanding and improving farming systems in the humid tropics. Westview Press, Boulder, CO. Hasse, V. and J. A. Litsinger. 1981. The influence of vegetational diversity of hostfinding and larval survivorship of the Asian corn borer, Ostrinia furnacalis Guenee. International Rice Research Institute Saturday Seminar, Department of Entomology. IRRI, Philippines. Haynes, Hayne s, R. J. 1980. Influe Influence nce of soil management management practice on the orchard agroecosystem. Agroecosystems 6: 3-32. Helenius, J. 1989. The influence of mixed intercropping of oats with field beans on the abu abunda ndance nce and spa spatia tiall dis distri tribut bution ion of cer cereal eal aph aphids ids (Ho (Homop mopter tera: a: Aph Aphidi ididae dae). ). Agriculture, Ecosystems and Environment 25: 53-73. Helenius, J. 1991. Insect numbers and pest damage in intercrops vs. monocrops: Concepts and evidence from a system of faba bean, oats and Rhopalosiphum padi (Homoptera, Aphididae). Journal of Sustainable Agriculture 1: 57-80.
Helenius, J. 1998. Enhancement of predation through within-field diversification. In Enhancing biological control, pp. 121-160. E. Pickett and R. L. Bugg, eds. University of California Press, Berkeley, CA. Hendrix, P. F., D. A. Crossley, J. M. Blair, and D. C. Coleman. 1990. Soil biota as Sustainable nable agricu agricultura lturall system systems, s, components compo nents of sustai sustainable nable agroe agroecosyst cosystems. ems. In Sustai pp. 637-654. C. A. Edwards, ed. Soil and Water Conservation Society, Akeny, IO. Hilbec Hil beck, k, A., M. Bau Baumga mgartn rtner, er, P. M. Fri Fried, ed, and F. Big Bigler ler.. 199 1998. 8. Eff Effect ectss of tra transg nsgeni enicc Bacillus thuringiensis corn-fed prey on mortality and development time of immature Chrysoperla carnea (Neuroptera: Chrysopidae). Environmental Entomology 27 (2): 480-487. Hodek, I. 1973. Biological of Coccinellidae. Junk NX Publishers, Academia, The Hague, The Netherlands. Hokkanen, H. M. T. 1991. Trap cropping in pest management. Annual Review of Entomology 36: 119-138. Holland, J. M. and S. R. Thomas. 1996. Phacelia tanacetifolia flower strips: Their effect eff ect on ben benefi eficia ciall inv invert ertebr ebrate atess and gam gamebi ebird rd chi chick ck foo food d in an int integr egrate ated d far farmming system. Acta Jutl. 71: 171-182. Hooks, C. R. R., H. R. Valenzuela, and J. Defrank. 1998. Incidence of pests and arthropod natural enemies on zucchini grown with living mulches. Agriculture, Ecosystems and Environment 69: 217-231. Horowitz, M. T., T. Blumdel, G. Hertzlinger, and N. Hulin. 1962. Effects of repeated applications of ten soil-active herbicides on weed populations. Weed Research 14: 97-109. Hove Ho vela land nd,, C. S. S.,, G. G. Bu Buch chan anan an,, an and d M. C. Ha Harr rris is.. 19 1976 76.. Re Resp spon onse se of we weed edss to soi soill phosphorus and potassium. Weed Science 24: 194-201. Huffaker, C. B. and P. S. Messenger. 1976. Theory and practice of biological control. Academic Press, New York. Idris, A. B. and E. Grafius. 1995. Wildflowers as nectar sources for Diadegma insulare (Hyme (Hymenopte noptera: ra: Ichne Ichneumonid umonidae), ae), a parasi parasitoid toid of diamon diamondback dback moth (Lepidoptera: Yponomeutidae). Environmental Entomology 24: 1726-1735. Igzoburkie, M. U. 1971. Ecological balance in tropical agriculture. Geographical Review 61: 519-529. Johnson, T. B., F. T. Turpin, and M. K. Bergman. 1984. Effect of foxtail infestation on corn rootworm larvae (Coleoptera: Chrysomelidae) under two corn-planting dates. Environmental Entomology 13: 1245-1248. Kajak, Kaj ak, A. and J. Luk Lukasi asiewi ewicz. cz. 199 1994. 4. Do sem semi-n i-natu atural ral pat patche chess enr enrich ich cro crop p fie fields lds wit with h predatory epigean arthropods? Agriculture, Ecosystems and Environment 49: 149-161. Kajimu Kaj imura, ra, T.Y T.Y.. 199 1995. 5. Eff Effect ect of org organi anicc ric ricee far farmin ming g on pla planth nthopp oppers ers.. Rep Reprod roduct uction ion of the white-backed planthopper, Sogatella furcifera (Homoptera Delphacidae). Research on Population Ecology 37: 219-224.
Kareiva, Kareiv a, P. 198 1983. 3. The inf influe luence nce of veg vegeta etatio tion n tex textur turee on her herbiv bivore ore pop popula ulatio tions: ns: Re Re-source concentration and herbivore movement. In Variab Variable le plants and herbivores in natural and managed systems, pp. 259-289. R. F. Denno and M. S. McClure, eds. Academic Press, New York. Kareiva, P. 1986. Trivial movement and foraging by crop colonizers. In Ecological theory and integrated pest management practice, pp. 59-82. M. Kogan, ed. Wiley & Sons, New York. Kemp Ke mp,, J. C. an and d G. W. Ba Barr rret ett. t. 19 1989 89.. Sp Spat atia iall pa patt tter erni ning ng:: Im Impa pact ct of un uncu cult ltiv ivat ated ed co corr ridors on arthropod populations within soybean agroecosystems. Ecology 70: 114-128. Khan, Z. R., J. A. Pickett, J. van der Berg, and C. M. Woodcock. 2000. Exploiting chemical ecology and species diversity: Stemborer and Striga control for maize in Africa. Pest Management Science 56: 1-6. Kido, H., D. L. Flaherty, C. E. Kennett, N. R. McCalley, and D. E. Bosch. 1981. Seeking the reasons for differences in orange tortrix infestations. California Agriculture 35: 27-28. Klinger, K. 1987. Effects of margin-strips along a winter wheat field on predatory arthropods and the infestation by cereal aphids. Journal of Applied Entomology 104 (1): 47-58 Kloen, H. and M. A. Altieri. 1990. Effect of mustard (Brassica hirta) as a non-crop plant on competition and insect pests in broccoli (Brassica oleraceae). Crop Protection 9: 90-96. Kowalski, R. and P. E. Visser. 1979. Nitrogen in a crop-pest interaction: Cereal aphids. In Nitrogen as an ecological parameter, pp. 67-74. J. A. Lee, ed. Blackwell Scientific Publishers, Oxford, UK. Landis, D. A. and M. J. Haas. 1992. Influence of landscape structure on abundance and within field distribution of European corn borer (Lepidoptera: Pyralidae) larval parasitoids in Michigan. Environmental Entomology 21: 409-416. Land La ndis, is, D. A. an and d P. Ma Mari rino no.. 199 1996a 6a.. Ef Effe fect ct of la lands ndsca cape pe str struc uctu ture re on pa para rasit sitoi oid d di diver ver-sity and parasitism in agroecosystems. Ecological Applications 6 (1): 276-284. Landis, D. A. and P. C. Marino. 1996b. Parasitoid communities of pest lepidoptera in agricultural landscapes: Theory, reality and implications for biological control. Bulletin of the Ecological Society of America 77 (3 Suppl. Part 2): 250. Annual Combined Meeting of the Ecological Society of America on Ecologists/Biologists as Problem Solvers, Providence, Rhode Island, August 10-14, 1996. Landis, D. A., S. D. Wratten, and G. A. Gurr. 2000. Habitat management to conserve natural enemies of arthropod pests in agriculture. Annual Review of Entemology 45: 175-201. Lasa La sack ck,, P. M. an and d L. P. Pe Pedi digo go.. 19 1986 86.. Mo Move veme ment nt of st stal alk k bo bore rerr la larv rvae ae (L (Lep epid idop opte tera ra:: Noctuidae) from noncrop areas into corn. Journal of Economic Entomology 79: 1697-1702.
Lashomb, J. H. and Y. S. Ng. 1984. Colonization by Colorado potato beetles, Leptinotarsa decemlineata (Say) (Coleoptera: Chrysomelidae), in rotated and non-rotated potato fields. Environmental Entomology 13: 1352-1356. Latheef, M. A. and R. D. Irwin. 1980. Effects of companionate planting on snap bean insects, Epilachna varivestis and Heliothis zea. Environmental Entomology 9: 195-198. Leius, K. 1967. Influence of wild flowers on parasitism of tent caterpillar and codling moth. Canadian Entomologist 99: 444-446. LeSar, C. D. and J. D. Unzicker. 1978. Soybean spiders: Species composition, population densities and vertical distribution. Illinois Natural History Survey. Biological Notes No. 107. Ubana, IL. Lest Le ston on,, D. 19 1973 73.. Th Thee an antt mo mosa saic ic-t -tro ropi pica call tr tree ee cr crop opss an and d th thee li limi miti ting ng of pl plot otss an and d di disseases. PANS 19: 311. Letour Let ournea neau, u, D. K. 198 1983. 3. The eff effect ectss of veg vegeta etatio tional nal div divers ersity ity on her herbiv bivoro orous us ins insect ectss and associated natural enemies: Examples from tropical and temperate agroecosystems. PhD dissertation. University of California, Berkeley, CA. Letourneau, D. K. 1987. The enemies hypothesis: Tritrophic interaction and vegetational diversity in tropical agroecosystems. Ecology 68: 1616-1622. Letourneau, D. K. and M. A. Altieri. 1983. Abundance patterns of a predator Orius tristicolor (Hemiptera: Anthoconidae) and its prey, Frankliniella occidentalis (Thysanopte (Thysa noptera: ra: Thripi Thripidae): dae): Habitat attra attraction ction in polyc polyculture ulturess versus monocultures. Environmental Entomology 122: 1464-1469. Levine, E. 1985. Oviposition by the stalk borer, Papailpema nebris (Lepidoptera: Journa urnall of Eco Eco-Noct No ctui uida dae) e) on we weed eds, s, pl plan antt de debr bris is,, an and d co cove verr cr crop opss in ca cage ge te test sts. s. Jo nomic Entomology 78: 65-68. Levins, R. and M. Wilson. 1979. Ecological theory and pest management. Annual Review of Entomology 25: 7-29. Lewis, T. 1965. The effects of shelter on the distribution of insect pests. Scientific Horticulture 17: 74-84. Liang, W. and M. Huang. 1994. Influence of citrus orchard groundcover plants on arthropod communities in China: A review. Agriculture, Ecosystems and Environment 50: 29-37. Liebman, J. 1997. Rising toxic tide: Pesticide use in California, 1991-1995. Pesticide Action Network, San Francisco, CA. Liebman, M. and E. R. Gallandt. 1997. Many little hammers: Ecological management of crop-weed interactions. In Ecology in agriculture, pp. 291-343. L. E. Jackson, ed. Academic Press, San Diego, CA. Lincoln, C. and D. Isley. 1947. Corn as a trap crop for the cotton bollworm. Journal of Economic Entomology 40: 437-438. Litsinger, J. A., V. Hasse, A. T. Barrion, and H. Schmutterer. 1991. Response of Ostrinia furnacalis (Guenee) (Lepidoptera: Pyralidae) to intercropping. Environmental Entomology 20: 988-1004.
Palada, M. C., S. Ganser, R. Hofstetter, B. Volak, and M. Culik. 1983. Associations of interseeded legume cover crops and annual row crops in year-round cropping systems. In Experimentally sound agriculture, pp. 193-214. N. Lockeretz, ed. Praeger, New York. Cultur tural al pra practi ctices ces and inf infect ectiou iouss cro crop p dise disease ases. s. Sprin Palti, Pal ti, J. 198 1981. 1. Cul Springer, ger, New York. Paoletti, M. G., B. R. Stinner, and G. G. Lorenzoni. 1989. Agricultural ecology and environment. Elsevier, Amsterdam. Peng Pe ng,, R. K. K.,, L. D. In Inco coll ll,, S. L. Su Sutt tton on,, C. Wr Wrig ight ht,, an and d A. Ch Chad adwi wick ck.. 19 1993 93.. Di Dive vers rsit ity y of airborne arthropods in a silvoarable agroforestry system. Journal of Applied Ecology 30 (3): 551-562. Peppers, B. B. and B. F. Driggers. 1934. Non-economic insects as intermediate hosts of parasites of the oriental fruit moth. Annals of the Entomological Society of America 27: 593-598. Perfect, T. J. 1991. Biodiversity and tropical pest management. In The Biodiversity of microo microorganism rganismss and invert invertebrate ebrates: s: Its role in sustain sustainable able agriculture, agriculture, pp. 145-148. D. L. Hawksworth, ed. CAB International, Wallingford, UK. Perfecto, I. 1995. Biodiversity and the transformation of a tropical agroecosystem: Ants in coffee plantations. Ecological Applications 5 (4): 1084-1097. Perfecto, I. and J. H. Vandermeer. 1996. Microclimatic changes and the indirect loss of ant diversity in a tropical agroecosystem. Oecologia 108: 577-582. Perr Pe rrin in,, R. M. 19 1975 75.. Th Thee ro role le of th thee pe pere renn nnia iall st stin ingi ging ng ne nett ttle le Urtica dioic rese serrdioica a as a re voir of beneficial natural enemies. Annals of Applied Biology 81: 289-297. Perrin, R. M. 1977. Pest management in multiple cropping systems. Agroecosystems 3: 93-118. Perrin, R. M. 1980. The role of environmental diversity in crop protection. Protection Ecology 2: 77-114. Perr Pe rrin in,, R. M. an and d M. L. Ph Phil illi lips ps.. 19 1978 78.. So Some me ef effe fect ctss of mi mixe xed d cr crop oppi ping ng on th thee po popu pu-lation dynamics of insect pests. Entomologia Experimentalis et Applicata 24: 385-393. Peterson, P. 1926. Oriental fruit moth damage in cultivated and uncultivated orProcee ceedin dings gs of the Ann Annual ual Mee Meetin ting g of the New Jer Jersey sey Sta State te Hort Horticu icultu ltural ral chards. Pro Society 3: 83-86. Phelan, P. L., J. F. Mason, and B. R. Stinner. 1995. Soil-fertility management and host preference by European corn borer, Ostrinia nubilalis (Huebner), on Zea mays L.: A comparison of organic and conventional chemical farming. Agriculture, Ecosystems and Environment 56 (1): 1-8. Enhancing ncing biolo biological gical contr control: ol: Habita Habitatt manage manage-Picket Pic kett, t, C.H C.H.. and R.L R.L.. Bug Bugg. g. 199 1998. 8. Enha ment to promote natural enemies of agricultural agr icultural pests. University of California Press, Berkeley, CA. insec sectt en enemi emies es of th thee co cott tton on Pier Pi erce ce,, W. D. D.,, R. A. Cu Cush shma man, n, an and d C. E. Ho Hood od.. 19 1912 12. The in boll weevil. Bureau of Entomology Bulletin No. 100, USDA, Washington, DC. Piment Pim entel, el, D. 196 1961. 1. Spe Specie ciess div divers ersity ity and ins insect ect pop popula ulatio tion n out outbre breaks aks.. A Ann nnal alss of En En-tomological Society of America 54: 76-86.
Risc Ri sch, h, S. J. 19 1980 80.. Th Thee po popu pula lati tion on dy dyna nami mics cs of se seve vera rall he herb rbiv ivor orou ouss be beet etle less in a tr trop opii cal agroecosystem: the effect of intercropping corn, beans and squash in Costa Rica. Journal of Applied Ecology 17: 593-612. Risch, S. J. 1981. Insect herbivore abundance in tropical monocultures and polycultures: An experimental test of two hypotheses. Ecology 62: 1325-1340. Risch, Ris ch, S. J. 198 1983. 3. Int Interc ercrop roppin ping g as a cul cultur tural al pes pestt con contro trol: l: Pro Prospe spects cts and lim limita itatio tions. ns. Environmental Management 7: 9-14. Risch, S. J., D. Andow, and M. A. Altieri. 1983. Agroecosystem diversity and pest control: contr ol: Data, tenta tentative tive concl conclusions, usions, and new resea research rch direc directions. tions. Environmental Entomology 12 (3): 625-629. Rissler, J. and M. Mellon. 1996. The ecological risks of engineered crops. MIT Press, Cambridge, MA. Robinson, R. A. 1996. Return to resistance: Breeding crops to reduce pesticide resistance. Ag. Access. Davis, CA. Robinson, R. R., J. H. Young, and R. D. Morrison. 1972. Strip-cropping effects on abunda abu ndance nce of pre predat datory ory and har harmfu mfull cot cotton ton ins insect ectss in Okl Oklaho ahoma. ma. Environmental Entomology 1: 145-149. Root, R. B. 1973. Organization of a plant-arthropod association in simple and diverse habitats: The fauna of collards (Brassica oleracea). Ecological Monographs 43: 95-124. Root, R. B. 1975. Some consequences of ecosystem texture. In Ecosystem analysis and prediction. S. A. Levin, ed. Penn State University Press, Philadelphia, PA. Rosset, P., J. Vandermeer, M. Cano, P. G. Varela, A. Snook, and C. Hellpap. 1985. El Frijol como cultivo trampa para el combate de Spodoptera sunia Guenee (Lepidoptera: Noctuidae) en plantulas de tomate. Agronomia Costarricense 9: 99-102. Russe Ru ssell ll,, E. P. 198 1989. 9. En Enem emie iess hy hypot pothe hesi sis: s: A re revi view ew of th thee ef effec fectt of ve veget getat atio ional nal di diver ver- Environmental nmental Entomol Entomology ogy 18: 590sity sit y on pred predato atory ry inse insects cts and par parasi asitoi toids. ds. Enviro 590-599. 599. Sanchez, P. A. 1995. Science in agroforestry. Agroforestry Systems 30: 5-55. Sche Sc hell llho horn rn,, N. A. an and d V. L. So Sork rk.. 19 1997 97.. Th Thee im impa pact ct of we weed ed di dive vers rsit ity y on in inse sect ct po poppulation dynamics and crop yield in collards, Brassica oleracea (Brassicaceae). Oecologia 111 (2): 233-240. Schoonhoven, A. von, C. Cardona, J. Garcia, and R. Garzon. 1981. Effect of weed Empoasca sca kraeme kraemeri ri Ross and Moo covers cov ers on Empoa Moore. re. Pop Popula ulatio tions ns and dry bea bean n yie yields lds.. Environmental Entomology 10: 901-907. Schr Sc hrot oth, h, G. G.,, U. Kr Krau auss ss,, L. Ga Gasp spar arot otto to,, J. Ag Agui uila lar, r, A. Du Duar arte te,, an and d K. Vo Vohl hlan and. d. 20 2000 00.. Pests and diseases in agroforestry systems of the humid tropics. Agroforestry Systems 50 (3): 199-241. Sengonca, C. and B. Frings. 1988. The influence of Phacelia tanacetifolia to pests and beneficial insects in sugar beet plots. Pedobiologia 32 (5-6): 311-316. Settle, W. H., L. T. Wilson, D. L. Flaherty, and G. M. English-Loeb. 1986. The variegated leafhopper, as increasing pest of grapes. California Agriculture 40: 30-32.
Wratten, S. D. 1987. The effectiveness of native natural enemies. In Integrated pest management, pp. 89-112. A. J. Burn, T. H. Coaker, and P. C. Jepson, eds. Academic Press, London. Wratten, S.D. and H.F.Van Emden. 1995. Habitat management for enhanced activity of natural enemies of insect pests. In Ecology and integrated farming sysGlen, M. P. Greaves, and H. M. Anderson, eds. John tems, pp. 117-145. D. M. Glen, Wiley and Sons, Chichester, UK. Wrubel, R. P. 1984. The effect of intercropping on the population dynamics of the arthropod community associated with soybean (Glycine max). MS thesis. University of Virginia, Blacksburg, VA. Wyss, Wy ss, E. E.,, U. Ni Nigg ggli li,, an and d W. Ne Nent ntwi wig. g. 19 1995 95.. Th Thee im impa pact ct of sp spid ider erss on ap aphi hid d po popu pula la-tions in a strip-managed apple orchard. Journal of Applied Entomology 114: 473-478. Yan, Y. H., Y. Yi, X. G. Du, and B. G. Zhao. 1997. Conservation and augmentation of natural enemies in pest management of Chinese apple orchards . Agriculture, Ecosystems and Environment 62 (2-3): 253-260. Yurjevic, A. M. 1991. The assessment of agroecological bottom up development. PhD diss dissert ertati ation. on. Lat Latin in Ame Americ rican an Stu Studie dies, s, Uni Univer versity sity of Cal Califo ifornia rnia,, Ber Berkel keley, ey, CA. Zandstra, B. H. and P. S. Motooka. 1978. Beneficial effects of weeds in pest management—A review. PANS 24: 333-338. Weed-crop crop compet competition ition—A —A revie review. w. Intern Zimd Zi mdah ahl, l, R. L. 19 1980 80.. WeedInternation ational al Plan Plantt Prote Protecction Center, Corvallis, OR.
Index
Index
Page num Page number berss fol follo lowed wed by the let letter ter “f” ind indica icate te fi figur gures; es; tho those se fol follo lowed wed by the let letter ter “t” indicate tables. Adaptation, 175 Additions, enhancing, 186 Additive polyculture, 31t Adjacent vegetation, 26 Ageratum conyzoides, 114 Agonum dorsale, 132, 136 Agricultural diversity, erosion of, 4 Agricultural systems, seven types of, 19 Agrochemical technology, 5 Agrochemical treadmill, 172, 172f Agroecologist, role of, 13, 15f Agroecology, role of, 188f Agroecosystems biodiversity in, 10-15 classification of, 17-18 diversification of an onion agroecosystem in Michigan, case study, 188-191 dominant, classification of, 21, 21f dysfunction, causes of, 178 healthy, 176, 178-179 immunity, 179 optimizing total function, 180, 181f patterns of contrasting, 20, 20f pest-stable vegetationally diverse, 171-196 pillars of health, 179, 180f sustainable, requirements of, 175-176, 177f traditional, 6-9 vegetational component of, 127 weeds as sources of insect pests in, 48-49
Agroforestry systems, 185 designing, 166-169 further research, need for, 170 insect pests in, 161-170 traditional, 7 Agropyron repens, 49 Alder, 130, 138 Aleochara bilineata, 189 Alfalfa, 195 Allen, W.W., 112 Allylisothiocyanate, 64, 87 Alternative-host leafhopper, 139 Altieri, M.A. and brussels sprouts, 90 and ecology of arthropods, 143 and the enemies hypothesis, 88 and experimental design, 65 and flea beetles, 102 and green cover, maintaining, 123 and living mulches, 97 and maize with fava bean, 93 and normal area of influence, 151 and parasitization rates of Heliothis zea eggs, 53 and Phyllotreta cruciferae, 43 and reduced pest incidence, 79 and rural development programs, Latin America, 8 and the velvetbean caterpillar, 74 and visual-inspection sampling, thrips, 105 Amaranthus dubius, 64 Amazonian Kayapo, 7 Amblyseius fallacis, 114 Anagasta kuehniella, 148
Anagrus epos
decline in vineyards, 151, 152 early-season buildup of, 139 egg parasitism rates, 124, 125f and grape leafhopper, 52 overwintering, 162 Anagrus sp., 92 Andow, D.A., 30, 42, 68, 79, 85, 88 Animal integration, 185 Annual polycultures, insect dynamics in, 25-26 Ants, 145-146, 146f, 163 Apanteles glomeratus, 52 Aphaereta pallipes, 189 Aphelinus mali, 113 Aphidius rhopalosiphi, 98 Aphidius spp., 98 Aphidophagous predators, 117 Aphidophagous syrphids, 50 Aphis gossypii, 48 Aphytis proclia, 52 “Apparency,” plant, 35 Apple maggot, 134 Apple orchards in California, case study, 118-121 “Appropriate/inappropriate landings,” 36 Arable irrigation systems, 19 Argyresthia conjugella, 134 Armyworm, 39 Aromatic plants, 107 Arrhenatherum elatius, 49 Arthropod populations, 50 Arthropod species, 31t Associated biodiversity, 11-12, 11f Associational resistance, 31-32 Attractant effect, 88 Availability, crop, 22-23, 23f Aveling, C., 138
Bach, C.E., 35, 102-104 “Background” color, 162 Bacteria, 182 Baliddawa, C.W., 55
Bare-ground monocultures, 96 “Bare-soil” cultivation, 36 Bean yellow dwarf virus (BYDV), 49 Beanfly, 165 Beauveria bassiana, 163 Beetle-movement behavior, 99 Bell bean, 118 Bergman, M.K., 129 Bigger, M., 162 Biochemical input, reliance on, 17 Biodiverse farms, 6-9 Biodiversity definition of, 3 ecological role of, 9-10 enhancing, 187-188 global threats to, 3 nature of in agroecosystems, 10-15 Biological control, 23-28, 56t-59t Biological corridors in vineyards, case study, 151-154 Black pecan aphid, 121 Blackmargined aphid, 121 Blepharidopterus angulatus, 138 Blueberry, 134 Boll weevil, 130 Boller, E.F., 126 Bora Indians, 7 Borders, designing and managing, 138-143 Botrell, D.G., 106 Brassica campestris, 74 Brassica kaber, 43 Brevicoryne brassicae, 35, 96 Bromegrass, 39 Bromus spp., 48 Brussels sprouts, 65-66, 90 Buchanan, G.A., 70, 71 Buckhorn plantain, 70 Buckwheat, 122 Bugg, R.L., 114, 122 Bureau of Entomology and Plant Quarantine, 141 Buttel, F.H., 201
Cabbage aphids, 85, 133 Cabbage root fly, 96 Cabbage worm, 66 Caldas, R.C., 165 Camphorweed, 108-109 Cantharidae, 119 Carabid beetles, 132, 136, 137f, 155 Carabidae, 136, 146 Carrot fly, 32, 48, 130 Carrot rust fly, 98 Case studies apple orchards in California, 118-121 biological corridors in vineyards, 151-154 cassava intercrops and pest incidence, 93-95 diversification of an onion agroecosystem in Michigan, 188-191 a diversified small farming system in Chile, 191-196 exchange of arthropods at the interface of apple orchards and adjacent woodlands, 143-148 maize intercrops and pest attack, 91-93 pecan orchards in Georgia, 121-123 reducing stemborers in Africa, 95 strip management to augment predators, 154-159 summer cover crops in vineyards, 123-126 Cassava intercrops and pest incidence, case study, 93-95 Cassava whiteflies, 43, 43f Cassia sp., 53 Cates, R.G., 35 Cauliflower plants, 87 Centipede grass, 122 Centro de Educacion y Tecnologia (CET), 191 Cephus cinctus, 39 Cereals, pests, 48-49, 130 Chemical constituents, soil, 70
Chemical cues, herbivores, 164 Chemical defense, plant, 35 Chilo partellus, 39 Chiverton, P.A., 136, 149 Chrysomelid beetles, 85 Chrysopidae, 119 Circulifer tenellus, 48 Clean-cultivated monocultures, 87 Clover-soybean, 89 Coaker, T.H., 96, 130 Coccinellid beetle, 104, 137-138 Coccinellidae, 54, 119, 144, 145f Codling moth, 119, 120, 120t Coffee, 163 Coffee berry borer, 163 Coffee senna, 70 Coleomegilla, 104 Coll, M., 79, 105, 106 Collards, 66, 74 Collier, R.H., 36 Colonization of orchards, 144 Colorado potato beetle, 129, 139 Commercial seedbed preparation, 5 Compensatory growth, 186 Complementarity, 197 Complementary interactions, 197 “Compound ecosystems,” 128 Condylostylus sp., 92 Conservation, 175, 185 Contrasting agroecosystems, patterns of, 20, 20f Corbett, A., 22, 42, 139, 151, 156, 157 Corn earworm, 65 Corn rootworm, 195 Corn-belt farmer, 171 Costello, M.J., 97 Cover crop, 185 Cover-crop management, orchards under, 111-126 apple orchards in California, case study, 118-121 effects on farming systems, 112f pecan orchards in Georgia, case study, 121-123 summer cover crops in vineyards, 123-126
Cowpea aphid, 122 Cowpea-maize, 82 Cratosomus flavofasciatus, 165
Creeping bent grass, 96 Cressa sp., 70 Crimson clover, 122 Critical composition period, weed, 72, 74, 76 “Critical period,” 72 Crop colonization stage, 84 Crop-diversification, 180. See also Diversification Crop edges, 129-131 Crop-field border vegetation, 148-149 Crop pest population management, 61t-63t Crop rotations, 185 Crop sequence, six-year rotation design, 194f Crop-weed competition, 67-68, 69f Crop-weed-insect interaction studies, 25 Crop-weed management, 68-76 Cross-wind trap strips, 187 Croton sp., 53 Cucumber beetle, 85 Cultural diversity, 9 “Cultural evolution continuum,” 143 Curly dock, 70
Dalbulus maidis, 35
Diffusion models, 35 Direct sowing, 71 Diversification, crop, 23-28, 175, 197-198 diversification of an onion agroecosystem in Michigan, case study, 188-191 a diversified small farming system in Chile, case study, 191-196 increasing, 186 landscape, 186 long-term maintenance of, 201 restoring, 183-186, 184f selective, 197 Diversification of an onion agroecosystem in Michigan, case study, 188-191188-191 A diversified small farming system in Chile, case study, 191-196 Dock sawfly, 48 Dogwood, 134 Dominant agroecosystems, classification of, 21, 21f Dominant predators, 50 Driggers, B.F., 112 Dunn, J.P., 136 Dutcher, J.D., 122 D-Vac machine, 118 Dysaphis plantaginea, 48
E. indica, 63 E. kraemeri, 92
Dambach, C.A., 130, 131 Date of planting, 107-108 Early-successional agricultural Defrank, J., 96 landscapes, 187 Dempster, J.P., 96 Ecological effects, weed diversity, Desmodium sp., 53 67-68 Destructive biota, 11 Ecological services, 12-13, 14f Developing countries, biodiversity in, 7 Ecological theory, plant diversity, Diabrotica balteata, 59, 63, 92 29-39 Diamondback larvae, 66 Ecologically based pest management (EBPM), 2, 174 Diamondback moth, 32 Edge vegetation, 131 Diapause, 130 Diaphania hyalinata, 89 Eleusine indica, 71 Diatraea lineolata, 93 “Emigration effect,” 34
Empoasca fabae, 129 Empoasca kraemeri, 59, 63, 71
“Enemy hypothesis,” 36, 37t Energy requirement, production stability and, 19, 19f Enhancement, predatory mite, 157 Entomofauna, beneficial, 50 Entomophagous insects, 132 Entomophthora muscae, 189 Environmental simplification, process of, 4 Eotetranychus willamette, 55 Eriborus terebrans, 134 Erigone spp., 90 Erythroneura, 44 European corn borer, 49, 88 Ewel, J.J., 166, 169 Exchange of arthropods at the interface of apple orchards and adjacent woodlands, case study, 143-148
Fall armyworm (FAW), 65, 92-93 Fallow, 186 Farmland mosaic, diversity of, 128 Fast-growing tree crops, 167 Feeny, P., 35 Fertilizer, 5 Festuca spp., 48 Field boundaries, 131-138 Filter strips, 187 Finch, S., 36 Flaherty, D., 90 Flea beetle and broccoli, 102, 103f and cauliflower plants, 85 collards interplanted with tomato on, 32 feeding on Brassica campestris, 64 and weed growth in collard crop cycle, 74, 75t and wild mustard, 87 Flor, C.A., 78 Flower thrips, 82
Flowering weeds, 25, 51-54 Flowering willows, 138 Food production, 185 Food web, goldenrod, 149, 150t Forage, 195 Francis, C.A., 78 Frankliniella occidentalis, 89 French prune tree, 140, 140f, 151 Frings, B., 138 Fulvic acid, 182 Functional biodiversity, assembling a, 12, 13f Functional diversity, 28 Functional grouping, 17 Fusarium, 182 Fusarium oxysporum, 5 Fye, A.E., 114
Garcia, M.A., 102 Generalist predators, 33, 153, 156f Genetic diversity, maintaining, 7 Genetic uniformity, vulnerability of, 4-5 Geranium carolinianum, 70 Girma, H., 165 Gold, C.S., 43 Goldenrod, 109, 113, 149 Granett, J., 151 Grape leafhopper, 139 Grasses, 111, 149 Grass-weed borders, 63, 64f Grazing systems, 19 “Green bridge,” 49 Green Revolution rice, Bangladesh, 4 Groden, E., 189 Ground beetle, 132, 133f Ground-cover and herbivore dynamics, 26 vegetation, 112 Ground-dwelling predators, 145-146
Haas, M.J., 134 Habitat, influence of, 127-159
Habitat management, 123, 198-199 Habitat removal, 197 Hairy vetch, 122 Harpalus pennsylvanicus, 65 Hasse, V., 83, 85 Hawthorn hedge, 132 Hedgerow plants, 130, 131, 165 Helenius, J., 31, 79, 97 Heliothis zea eggs, 53, 55t Hemiptera, 92 Herbicide use, 70-71 Herbivore and chemical cues, 164 mediating crops and weeds, 47 specialized, 30, 30f trends in polycultures, 85, 87-91 Herbivore dynamics, 21, 26 Herbivore-enemy interaction, 36 Heterogeneity, 21 Hierarchy of systems, agriculture as, 17, 18f High-density cropping, 35 High-yield varieties (HYV), 4 “Hit-and-miss” approach, 28 Homogenization, 5 Homopteran, 182 Hooks, C.R.R., 96 Hornworm, 94 Host-parasitism rates, recording, 105 Host-plant finding, 32 Huang, M., 114 Huastec Indians, Mexico, 7 Humic acid, 182 Hymenoptera, 120, 144f Hypothetical field, 156, 157f
ICA Pijao, 92 Ichneumonidae, 51 Immunity, agroecosystem, 178, 179 In situ repositories, native crop diversity, 9 Indole-3 acetic acid, 182 Insect stability, and plant diversity, 29-45
Insecticide, 142 Insects and agroforestry systems, 161-170 and crop edges, 129-131 influence of adjacent habitats, 127-159 locomotion, 128, 129f management, and multiple-cropping systems, 77-109 manipulation, 47-76 in orchards, 111-126 and polycultures, 78-85, 80t-82t, 84f, 99-106 stability in agroecosystems, 29-45 and weed-diversified crop systems, 55-66, 56t-59t, 60t, 61t-63t, 64f weeds as sources of, 48-49 Integrated pest management (IPM), 1, 122, 179 Integrated soil-fertility management (ISFM), 179 Intercropping, 83, 84, 85t, 108 Interplanted vegetation, 22 Intracrop diversity, 35 Irwin, R.D., 83 Isolation, extent of, 10
Jimson weed, 71 Johnson, T.B., 129 Johnsongrass, 90, 115
Kareiva, P., 100-102 Kentucky bluegrass, 96 Kido, H., 151 Kienegger, M., 36 Klinger, K., 138 Knotgrass, 48
Lacebug, 94 Lacewing, 50-51 “Lack of immunity,” 178
Lady beetles, 93 Lagopsis supina, 114
Lamb’s quarter, 113 Land equivalent ratio (LER), 109 Landis, D.A., 134, 187 Landis, D.L., 142 Landscape, as level of organization, 143 Latheef, M.A., 83 Leaf blight, 5 Leaf miner, 163 Leafhoppers alternative-host, 139 and apple orchards in California, 119, 119f ground cover impact on, 115 and maize-bean polyculture, 92 nymph, 124, 124f seasonal patterns of, 151, 152f, 153f western grape, 124 Legumes, 111 Leius, K., 51, 113 Leptinotarsa decemlineata, 129 Leptochloa filiformis, 63, 71 LeSar, C.D., 136 Letourneau, D.K., 79, 89, 105 Lewis, T., 83 Ley farming, 19 Liang, W., 114 Liebman, M., 79 Linit, M.J., 166 Litsinger, J.A., 27-28, 79, 83, 85 Living mulch, 96-98 Lixophaga sphenophori, 133 Local-attraction probability, 39, 41f Locomotion, insects, 128, 129f Lolium multiflorum, 49 Longevity, enhancing, 186 Long-term stabilization, 200 Low pest potentials, 27 Lucerne, 98 Lupinus spp., 64 Lydella grisescens, 52 Lygus bugs, 114
Macrocentrus ancylivorus, 113 Macrodactylus sp., 64
Maier, C.T., 134 Maintenance, productivity, 175 Maize, intercropping of, 6 and pest attack, case study, 91-93 Management commercial, 122 cover-crop, 111-126 crop-weed, 68-76 habitat, 123, 198-199 insects in multiple-cropping systems, 107-109 intensity of, 10 minimal, 122 pest, 17-28 population, crop pests, 61t-63t soil-fertility, 180 strip, 117, 154-159 understory-, 122 vegetation-, 26 Manure, 182 Marino, P.C., 142, 187 Markov processes, 100 Mayse, M.A., 67, 135 McClure, M., 115 Mechanical input, 17 Mechanized planting, 5 Mediterranean flour moth, 118 Mesquita, A.L.M., 165 Metapopulation dynamics, 198 Metaseiulus occidentalis, 90, 115 Meteorus sp., 92, 142 Metopolophium dirhodum, 98, 183 Mexican bean beetle, 128 Mexican tea, 108, 149 Microgaster, 134 Minihedgerows, 149 Mirids, 162, 163 Mixed cropping, 84t Mobility, natural enemies, 158, 159f Molasses grass, 95 Monocropping, 192
Monocultures expansion of, 197 and pest-control approaches, 171-174 Monophagous pests reduction of, 31 susceptibility to crop diversity, 79 Moody, K., 27-28, 79 Morning glory, 70 Mowing, 115 Mulching, 185 Multiple-cropping systems, 6, 77-109, 80t-82t Multiple-use capacity, 186 Multispecies rotations, 182 Murphy, B.C., 139, 151 Myzus festucae, 98
Onion maggot, 188-191, 190f Orchards, insect ecology in, 111-126 apple orchards in California, case study, 118-121 colonization of, 144 exchange of arthropods at the interface of apple orchards and adjacent woodlands, case study, 143-148 pecan orchards in Georgia, case study, 121-123 Organic amendments, 180 Organic matter, 200 Orius sp., 90, 105, 126 Orius tristicolor, 89 Osage borders, 132 Ostrinia nubilalis, 49, 52, 134
Napier grass, 95 Nascimento, A.S., 165 Natural enemies field boundaries and, 131-138 hypothesis, 32-34 mobility of, 158, 159f role of weeks in ecology of, 49-54 Navel orangeworm, 115 Neighboring vegetation, 132 Nentwig, W., 66, 117, 155-156 “Neutral herbivores,” 126 Nicholls, C.L., 123, 151 Niggli, U., 117 Nitrogen, low, 182 Nonrandom pest movement, 100-102, 101f Nontarget food sources, manipulation of, 24-25 “Normal area of influence,” 151 Nutrient cycles, closing, 185 Nutrient-recycling mechanisms, 201 Nutrition imbalances, 180
P. cruciferae, 43
O’Connor, B.A., 113 Okra, 108
Pacific mite, 115 Panonychus ulmi, 114, 138 Papaipema nebris, 52
Paraquat, 63 Parasitic wasps, soybean fields, 53, 54f Parasitoid populations, 34 Parrella, M.P., 123, 151 Peas, 166 Pecan orchards in Georgia, case study, 121-123 Peng, R.K., 166 Peninsular boundaries, 148 Peppers, B.B., 112 Perennial crop systems, 19 Perennial orchards, 45 Perfect, T.J., 105 Perfecto, I., 163 Permanence, crop, 10, 22 Permanent rain-fed cropping systems, 19 Perrin, R.M., 84, 187 Pest-control approaches, 171-174 Pesticide use, 1, 5, 173 Pest management and agricultural habitat, 17-23 and crop diversification, 23-28
Pest-regulation, 201 Price, P., 36, 38, 67, 135 Pest-stable agroecosystems, 2, 171-196 Prickly sida, 71 Peterson, P., 112 Productive biota, 10 Phacelia tanacetifolia, 113, 138, 139, Prokopy, R.C., 199 156 Prokrym, D.R., 42 Phillips, M.L., 84 Pruning, 164. See also Shade Phthium, 182 Pteridium, 70 Phyllotreta cruciferae, 43, 74, 87f, 96 Push-pull system, Trans Nzoia, 95 Phytophagous insects, 50 Puvuk, D.M., 49, 65 Phytophtora, 182 Pieris rapae, 35, 55, 96 Pillars, agroecosystem health, 179, “Quality” of plant diversification, 28 180f Pitfall catches, 146 Planned biodiversity, 11-12, 11f Plant, R.E., 22, 42, 156, 157 Rabb, R.L., 127 Plant diversity Ragweed, 52, 112 and insect stability, 29-45 Raised banks, 148 narrowing of cultivated, 3-4 Rao, M.R., 165 Plantains, 48 Rape pollen beetle, 142 Planthopper, 182 Red fescue, 96 Plants, healthy, 179-183 Red spider mite, 163 Pollard, E., 132, 136 Reducing stemborers in Africa, case Pollen, 52 study, 95 Polyculture, 77-78, 185 Regional-attraction probability, 39, 41f herbivore trends in, 85, 87-91 insect abundance in, 78-85, 80t-82t, Release-recapture experiments, 106 “Resident vegetation,” 114 84f Resource biota, 11 insect dynamics in, 99-106 Resource concentration, 34-35, 36, 37t predator-prey relationships in, 90, Resource-conserving, 192 91f Resource-poor farmers, 7-8, 8f, 24, Polyphagous pest 200, 201 reduction of, 31, 32t Rhinoceros beetle, 113 susceptibility to crop diversity, 79 Poplar, 130 Rhizoctonia, 182 Population management, crop pests, Rhoades, D.F., 35 61t-63t Rhopalosiphum maidis, 93, 104 Potato tuberworm, 118, 147, 147f Rhopalosiphum padi, 49 Predaceous beetle, 88 Riparian buffer zones, 187 Predation pressure, 147-148 Risch, S.J., 30, 35, 88, 99, 104 Predator catches, seasonal patterns of, Root, R.B., 31, 33, 127 154f, 155f Rosenheim, J.A., 139, 151 Predator conservation strips, 149 Rosset, P., 39 Predator-parasitoid behavioral Rosy apple aphid, 119 mechanisms, 198 Predatory mite, enhancement of, 157 Russell, E.P., 39-40
Salinization, 5 San Jose scale, 52, 113 Scaphytopius acutus, 115 Schellhorn, N.A., 66 Schmidt, L.L., 43, 97, 143 Scolie dejeani, 113 Searching behavior, 59 Seasonal patterns leafhopper, 151, 152f, 153f predator catches, 154f, 155f Self-regulation, 178 Self-sufficient farming system, 193f Self-sustaining, 192 Semipermanent rain-fed cropping systems, 19 Sengonca, C., 138 Sesbania exaltata, 122 Shade, 99-100, 162-167 Shade-strata design, 163, 164f Sheehan, W., 39, 40-42 Shelter belts, 130 Shifting cultivation systems, 19 Shoot fly, 94 Showy crotalaria, 70 Sicklepod, 70 Sight counting, 104 Silverleaf maize, 95 Sinapsis arvensis, 138 Single-species planted windbreak, 130 Single versus complex crop habitats, 25-26 Sink, refuge as, 22 Sithanantham, S., 165 Sitobium avenae, 49, 98 “Slow fraction,” 186 Sluss, R.R., 113 Small-grain production, 171 Smartweed, 112 Smith, J.G., 65 Smith, R.F., 112 Sod-strip intercropping, 96 Sogatella furcifera, 182 Soil, healthy, 179-183 Soil conservation, 141, 141f Soil Conservation Service, 141 Soil disturbance, 71-72, 73t
Soil erosion, 5 Soil pH, 70 Soil-fertility maintenance, 192 Soil-fertility management, 180 Sorbus aucuparia, 134 Sorghum, 39, 83 Sork, V.L., 66 Sotherton, N.W., 136 Southern green stink bug, 74 Southwood, T.R.E., 28 Soybean, 89, 132-133, 134f Space, crop arrangement in, 27-28, 28f Spatial dispersion, 34-35, 72, 108, 198 Spatial patterns, 157, 158f Spatial scale, 199 Specialist predator, 34 Speight, M.R., 139 Spiders, 119, 123, 132, 146 Stabilization, long-term, 200 Stalk borer, 165 Stamps, W.T., 166 Staphylinidae, 136 Staver, C., 163 Stinging nettle, 54, 108, 130 Stinner, B.R., 49, 65 Strip cropping, 93, 108 Strip management, 117 to augment predators, case study, 154-159 Strip vegetation, interplanted, 42 Striped cucumber beetle, 102, 103 Substitutive polyculture, 31t Successional development, 167-169 Sudan grass, 95 Sugar cane field margins, 133 Summer cover crops in vineyards, case study, 123-126 Sustainable agriculture, 174-175 Synergism, 181f, 197
Tahvanainen, J.O., 31 Tall plants, 107, 162 Tashkent Laboratory, 52, 113 Telenga, N.A., 113
Temperate regions, 130 Temple, S.R., 78 Temporal continuity, lack of, 162 Temporal/spatial patters, crop, 22 Tetranychus urticae, 93, 114 Thiele, H., 136 Thies, C., 142 Thresh, J.M., 48 Thrips, 105, 124, 162 Time, crop arrangement in, 27-28, 28f Tobacco plants, 83 Tomato plants, 83 Tonhasca, A., 93 Transgenic crops, 201 Trap-cropping effect, 38-39, 40t, 41f Trapping, 105-106 Tree growth, affected by ground-cover, 112 Trees, and agroforestry, 162-166 Trichogramma cacoeciae, 126 Trichogramma spp., 113 Trifluralin, 70 Trophic level density-dependence of, 25 interactions in a community of four, 38t Tropical lowlands, 169 Trujillo-Arriaga, J., 93 Tscharntke, T., 142 Turpin, F.T., 129 Typhlodromus pyri, 126 UK Game Conservancy Trust, 148, 149 Undergrowth, floral, 112 Undersowing, 98 Understory-management approach, 122 Unzicker, J.D., 136 Urtica dioica, 48 Valenzuela, H.R., 96 Van Emden, H.F., 48, 51, 130, 132, 133 van Huis, A., 93 Vandermeer, J., 38, 39, 109, 163 Varchola, J.M., 136 Vegetation, diversity of, 10, 21
Vegetation cover, 185, 200 Vegetation-management strategies, 26 Velvetbean caterpillar, 74 Velvetleaf, 71 Venice mallow, 71 Vineyards biological corridors in, case study, 151-154 summer cover crops in, case study, 123-126 Visual camouflage, 89 Visual-inspection sampling, thrips, 105
Waddington, C., 114, 122 Wainhouse, D., 130 Wallin, H., 136 Walnut aphid, 113 Warm-season grasses, 122 Way, M.J., 28 Weed-crop diversity studies, occurrence of pests in, 60t Weed-free maintenance, 74 Weeds abundance of, 108 attaining desirable level of, 69 crop-weed management, 68-76 insect dynamics in, 55-66, 56t-59t, 60t, 61t-63t, 64f isolating the ecological effects of, 67-68 and natural enemies, 49-54 as source of insect pests, 48-49 Western flower thrips, 124 Western grape leafhoppers, 124 Wetzler, R.E., 104 Whitcomb, W.H., 65 Wild blackberry, 139 Wildflowers, 51 Wild mustard, 87-88, 87f, 88t Williamette mite, 90, 115, 116f Willow, 130 Wilson, R.C., 97 Windbreak-induced colonization, 140 Windbreaks, 183
Wind protection, 185 Winterberry, 134 Wood, B.J., 113 Wrubel, R.P., 89 Wyss, E., 117 Yan, Y.H., 114 Yellow pecan aphid, 121
Yellow Sigatoka, 5 Yellow wavelength, 98
Zimdahl, R.L., 72 Zoecon codling moth, 118 Zonocerus variegatus, 93