Citation
Host Preferences in Trichogramma and How Understanding the Dynamics of a Farming System May Improve IPM Research

Material Information

Title:
Host Preferences in Trichogramma and How Understanding the Dynamics of a Farming System May Improve IPM Research
Creator:
PLUKE, RICHARD WILLIAM HAY ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Beans ( jstor )
Cabbages ( jstor )
Cassava ( jstor )
Crops ( jstor )
Eggs ( jstor )
Papayas ( jstor )
Parasite hosts ( jstor )
Parasitism ( jstor )
Pumpkins ( jstor )
Yams ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Richard William Hay Pluke. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
4/30/2005
Resource Identifier:
71266821 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

HOST PREFERENCES IN Trichogramma AND HOW UNDERSTANDING THE DYNAMICS OF A FARMING SYSTEM MAY IMPROVE IPM RESEARCH By RICHARD WILLIAM HAY PLUKE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Richard William Hay Pluke

PAGE 3

To my parents

PAGE 4

iv ACKNOWLEDGMENTS My sincere thanks go to my supervis or, Dr Gary Leibee. His unhesitating assistance, his professional gui dance and his personal trust made this research possible. I consider myself lucky to have been his stude nt. I would also like to thank my committee members (Dr John Capinera, Dr Peter Hildeb rand, and Dr Norman Leppla) for their support while I was at the Univ ersity of Florida. The shar ing of their experience and scientific insight was matched by their openness in allowing a nontraditional, offshore research project to run its course. My time in Puerto Rico was a personal and professional watershed and there are many people to thank. In particular, I would like to thank Dr Angel L. González and Dr Rosa Franqui for making it all possible. The success of the research was mostly thanks to Edna Pérez, Miguel Tirado, Luis Silva-Negrón, Flor Ortiz and Edgardo Vargas (from the Río Piedras ex perimental station); Juan Ortiz (from the Corozal experimental station); and Irma Cabr era and Alberto Vélez (from the Juana Díaz experimental station). The farming systems study would not have been possible without the help of extension agents Milagros Ali cea, David Matos, and Ramón Martínez; and the sixteen farmers and their families who shar ed their knowledge and friendship. Finally, I would like to thank my parents, my grandpare nts, and my brother and sister-in-law for their love and patience. It is not easy being so far awa y. What has made the distance bearable is a wonderful person, my fiancée, María Cristina Santos.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES..........................................................................................................xii LIST OF OBJECTS.........................................................................................................xvi ABSTRACT....................................................................................................................xvi i CHAPTER 1 INTRODUCTION........................................................................................................1 Global Cabbage Production..........................................................................................1 Cabbage Production in Flor ida and Puerto Rico..........................................................2 Cabbage Insect Pests in Puerto Rico............................................................................3 Diamondback Moth (DBM) and Development of Insecticide Resi stance in Puerto Rico and Florida.......................................................................................................3 Farming in the Central M ountains of Puerto Rico........................................................5 Historical Development of the Agri cultural Sector of Puerto Rico..............................8 Puerto RicoÂ’s Agricultural Intensifica tion Program...................................................14 2 LITERATURE REVIEW...........................................................................................15 Natural Enemies of the Diamondback Moth (DBM) and Soybean Looper (SL).......15 Diamondback Moth (DBM) Biology and Development............................................17 Soybean Looper Biology and Development...............................................................18 Trichogramma (Hymenoptera: Chalcidoidea)............................................................20 Trichogramma Host Preferences.........................................................................22 Trichogramma pretiosum (Riley)........................................................................25 Trichogrammatoidea bactrae (Nagaraja)............................................................25 Trichogramma minutum (Riley)..........................................................................26 Host Cues Mediating Trichogramma Searching........................................................27 Searching Techniques of Trichogramma Females and the Influence of Plant Characteristics........................................................................................................28 Parasitoid Learning.....................................................................................................30 Effects of Plant Structure on Egg Oviposition by Lepidopteran Hosts......................31

PAGE 6

vi Improving the Welfare of Small-Scale Farmers—the Role of Agricultural Research.................................................................................................................33 Farming Systems Research.........................................................................................36 Linear Programming...................................................................................................37 3 HOST PREFERENCES, PARASITISM LEVELS AND EFFECT OF PLANT DEVELOPMENT ON FIELD PARASITI SM OF DIAMONDBACK MOTH AND SOYBEAN LOOPER BY Trichogramma pretiosum .......................................40 Introduction.................................................................................................................40 Materials and Methods...............................................................................................43 Trichogramma pretiosum ....................................................................................44 Diamondback Moth (DBM)................................................................................45 Soybean Looper (SL)..........................................................................................46 Cabbage...............................................................................................................50 General Experimental Protocol...........................................................................51 Field cages....................................................................................................51 Insect releases...............................................................................................52 Egg counts and parasitism determination.....................................................53 Individual Experiment Characteristics................................................................54 Statistical Analysis..............................................................................................55 Results........................................................................................................................ .56 Egg Numbers and Quartile Positions..................................................................56 DBM eggs....................................................................................................57 SL eggs.........................................................................................................61 Comparing Egg-Laying Distribu tions of DBM and SL Eggs.............................66 Spread of Host Eggs among Quartiles.................................................................70 Field Parasitism of the Hosts’ Eggs.....................................................................71 Parasitism Levels.................................................................................................71 Differences in Percent Parasiti sm of the DBM and SL Eggs..............................75 Inter-Quartile Differences in Percent Parasitism.................................................75 Diamondback moth eggs..............................................................................75 Soybean looper eggs.....................................................................................78 Inter-Quartile Distribution of Parasiti zed Eggs in Comparison to the InterQuartile Distribution of all Eggs......................................................................80 Position of DBM Larvae in Referen ce to DBM Eggs Found on the Cabbage Plants................................................................................................................84 Discussion...................................................................................................................86 Host Egg Positioning...........................................................................................86 Parasitism of DBM and SL Eggs by Trichogramma pretiosum ..........................89 Trichogramma pretiosum’s parasitism patterns within the plants......................90 Use of T. pretiosum in the Control of DBM and SL in Cabbage........................92 Potential IPM Control Program for DBM in Puerto Rico...................................93 4 LABORATORY INVESTIGATIONS OF HOST PREFERENCE IN THE PARASITISM OF DIAMONDBACK MO TH AND SOYBEAN LOOPER BY Trichogramma ............................................................................................................97

PAGE 7

vii Introduction.................................................................................................................97 Materials and Methods...............................................................................................98 Trichogramma .....................................................................................................98 Soybean Looper (SL)..........................................................................................99 Diamondback Moth (DBM)..............................................................................101 Experimental Design.........................................................................................102 Basic Experimental Procedure..........................................................................103 Parasitism Preferences of Trichogramma pretiosum ........................................105 Parasitism Preferences of Conditioned Trichogramma pretiosum ....................105 Parasitism Preferences of 3 Trichogramma species..........................................107 Statistical Analysis............................................................................................108 Results.......................................................................................................................1 08 Parasitism Preferences of Trichogramma pretiosum with the Eggs of DBM & SL.....................................................................................................108 The Effect of ‘Conditioning’ on Parasitism Preferences of Trichogramma pretiosum when offered the Eggs of DBM and SL........................................111 Single-host diamondback moth eggs..........................................................112 Single-host soybean looper eggs................................................................113 Combined-host Petri dish results................................................................114 Dry eggs.....................................................................................................115 Parasitism Preferences of 3 Trichogramma Species when Offered DBM and Soybean Looper Eggs.............................................................................116 Diamondback moth eggs............................................................................116 Soybean looper eggs...................................................................................117 Diamondback moth and soybean looper eggs combined...........................118 Dry eggs.....................................................................................................120 Trichogramma emergence..........................................................................121 Discussion..........................................................................................................123 Host preference..........................................................................................123 Number of adults emerging from host eggs...............................................124 Dry host eggs..............................................................................................125 Conditioning of T. pretiosum .....................................................................126 Host preferences by the three species of Trichogramma ...........................127 5 CHARACTERIZING A FARMING SY STEM IN THE CENTRAL MOUNTAIN REGION OF PUERTO RICO: IPM COMPATIBLE?......................129 Introduction...............................................................................................................129 Methods....................................................................................................................132 Identifying the Study Farming System..............................................................132 Questionnaire Development..............................................................................134 Interviews..........................................................................................................134 Linear Programming..........................................................................................135 Determining the Parameters of an IPM Control Strategy for Diamondback Moth in Cabbage............................................................................................137 Results.......................................................................................................................1 43 System Characteristics.......................................................................................143

PAGE 8

viii Economic Considerations..................................................................................150 Incentives Program............................................................................................153 On-Farm Experiment.........................................................................................156 Linear Program..................................................................................................159 Manipulating the LP Model to Examine the Cabbage IPM Activity................164 Discussion.................................................................................................................167 Crop Characteristics and Incl usion in the LP Model.........................................167 Whole Farm Incomes, Land Use Patterns and Fluctuating Markets.................170 Reducing Inputs in IPM Systems—Labor.........................................................174 Reducing Inputs in IPM Systems—Cash..........................................................176 Organic Premiums and In creased Sales Prices..................................................177 Cabbage IPM Case Study..................................................................................178 Linear Programming Models and th eir Benefit to Farming Systems Research and to IPM Strategy Development.................................................184 APPENDIX A WIND SPEED DATA FOR THE FIELD EXPERIMENTS AT FORTUNA AGRICULTURAL RESEARCH SUBSTATION, 2002.........................................190 B QUESTIONNAIRES USED FOR CE NTRAL MOUNTAIN FARMS OF PUERTO RICO: INTERVIEW 1.............................................................................193 C QUESTIONNAIRES USED FOR CE NTRAL MOUNTAIN FARMS OF PUERTO RICO: INTERVIEW 2.............................................................................202 D QUESTIONNAIRES USED FOR CE NTRAL MOUNTAIN FARMS OF PUERTO RICO: INTERVIEW 3.............................................................................213 E MONITORING DIAMONDBACK MOTH AND SOYBEAN LOOPER IN AN EXPERIMENTAL PLOT OF CABBAGE MANAGED USING IPM TECHNIQUES........................................................................................................220 F LINEAR PROGRAMMING MODELS...................................................................223 G RESULTS FROM THE 8-YEAR LP MODEL SHOWING YEARS 4 AND 5 FOR ALL 16 FARMS I NVOLVED IN THE STUDY............................................224 LIST OF REFERENCES.................................................................................................247 BIOGRAPHICAL SKETCH...........................................................................................265

PAGE 9

ix LIST OF TABLES Table page 3-1 Growth stages of cabbage......................................................................................43 3-2 Details of the individual fi eld experiments at Juana Diaz.....................................55 3-3 Percent of SL eggs near leaf edge an d ratio of ‘central’ eggs to ‘edge’ eggs........65 3-4 Percentage of the hosts’ eggs f ound on the main vein of the leaves......................66 3-5 Two measures (t1 & t2) of the spread of the hosts’ eggs in the experimental plants and a comparison of th eir respective distributions......................................70 3-6 Percent parasitism of the DBM e ggs found in the DBM single-host field cages.......................................................................................................................72 3-7 Percent parasitism of the DBM eggs found in the combined-host field cages......72 3-8 Percent parasitism of the SL eggs found in the SL single-host field cages...........74 3-9 Percent parasitism of the SL eggs found in the SL combined-host field cages.......................................................................................................................74 3-10 Distribution profiles of parasitized eggs and all eg gs for DBM and SL in the combined host field cages of Experiment 1 (2001)...............................................81 3-11 Distribution profiles of parasitized eggs and all eg gs for DBM and SL in the combined host field cages of Experiment 1 (2002)...............................................81 3-12 Distribution profiles of parasitized eggs and all eg gs for DBM and SL in the combined host field cages of Experiment 3...........................................................82 3-13 Distribution profiles of parasitized eggs and all eg gs for DBM and SL in the combined host field cages of Experiment 4...........................................................82 3-14 Distribution profiles of parasitized eggs and all eg gs for DBM and SL in the combined host field cages of Experiment 5...........................................................83 3-15 Distribution profiles of parasitized eggs and all eg gs for DBM and SL in the combined host field cages of Experiment 6...........................................................83

PAGE 10

x 3-16 DBM egg and larval numbers in Experiment 3, broken down by plant quartile...................................................................................................................84 3-17 DBM egg and larval numbers in Experiment 5, broken down by plant quartile...................................................................................................................85 3-18 Distribution profiles by percentile for DBM eggs and larvae in the cabbage plants of Experiments 3 &5.....................................................................86 4-1 Percent parasitism of DBM eggs by ‘conditioned’ T. pretiosum females...........112 4-2 Percent parasitism of SL eggs by ‘conditioned’ T. pretiosum females................113 4-3 Percent parasitism of combined DBM & SL eggs by ‘conditioned’ T. pretiosum females............................................................................................115 4-4 Percent emergence and average number of Trichogramma adults emerging per egg for the conditioning experiments..................................................................122 4-5 Percent emergence and average number of Trichogramma adults emerging per egg for the Trichogramma species experiments..................................................123 5-1 The crops included in the LP model....................................................................136 5-2 Labor, costs and income per acre for the model’s crops......................................144 5-3 Farm characteristics, labor input s and annual net income for the 16 study farms...........................................................................................................147 5-4 Crops and fertilizers given as part of the ASDA incentive program...................154 5-5 Agrochemical products applied as pa rt of ASDA’s crop protection/weed control incentive program....................................................................................155 5-6 The crops chosen by the 8-year LP m odel for the three different scenarios........164 5-7 Scenario 1 (full incentives): change s that would result in the cabbage activities being included in the 8-year LP model.................................................165 5-8 Scenario 2 (no agrochemical incentive s): changes that would result in the cabbage activities being include d in the 8-year LP model..................................166 5-9 Scenario 3 (no incentives): change s that would result in the cabbage activities being included in the 8-year LP model.................................................167 A-1 Wind speed data for Experiment 1 (2002)...........................................................190 A-2 Wind speed data for Experiment 2 (2002)...........................................................190

PAGE 11

xi A-3 Wind speed data for Experiment 3 (2002)...........................................................191 A-4 Wind speed data for Experiment 4 (2002)...........................................................191 A-5 Wind speed data for Experiment 5 (2002)...........................................................192

PAGE 12

xii LIST OF FIGURES Figure page 1-1 Municipalities of Puerto Rico, with the farming systemsÂ’ study area outlined by a broad band.......................................................................................6 2-1 Soybean looper parasitized with Copidosoma floridanum .................................16 2-2 Looper parasitized with Voria sp........................................................................16 2-3 DBM parasitized by Cotesia spp........................................................................16 2-4 Diamondback moth life stages............................................................................18 2-5 Soybean life stages..............................................................................................19 2-6 Diamondback moth (left) and soybean looper (right) eggs either side of Pieris brassicae egg............................................................................................21 3-1 Preparing field cages for the experiments...........................................................50 3-2 Field cages used in Juana Diaz fieldwork...........................................................51 3-3 Sampling of the field plots and removal of cabbage...........................................53 3-4 Equipment used to find, remove a nd retain eggs for parasitism analysis...........54 3-5 Inter-quartile distribu tion of DBM eggs in th e cabbage plants of Experiment 1 (2001) and th e six experiments of 2002.......................................58 3-6 Number of DBM eggs, by quartile, found on the adaxial and ab axial sides of the leaves for the DBM single-host treatments...................................................59 3-7 Division of DBM eggs amongst the different morphological regions of a cabbage leaf........................................................................................................61 3-8 Inter-quartile distribution of soyb ean looper eggs in the single-host SL treatments............................................................................................................62 3-9 Number of SL eggs, by quartile, fo und on the adaxial a nd abaxial sides of the leaves........................................................................................................63

PAGE 13

xiii 3-10 Division of SL eggs amongst the di fferent morphological regions of a cabbage leaf........................................................................................................64 3-11 Distribution of the two host speciesÂ’ eggs by plant quartil es (chart) and by leaf number-derived percentiles (t able) for Experiment 1 (2001)......................67 3-12 Distribution of the two host speciesÂ’ eggs by plant quartil es (chart) and by leaf number-derived percentiles (t able) for Experiment 1 (2002)......................67 3-13 Distribution of the two host speciesÂ’ eggs by plant quartil es (chart) and by leaf number-derived percentile s (table) for Experiment 3..................................68 3-14 Distribution of the two host speciesÂ’ eggs by plant quartil es (chart) and by leaf number-derived percentile s (table) for Experiment 4..................................68 3-15 Distribution of the two host speciesÂ’ eggs by plant quartil es (chart) and by leaf number-derived percentile s (table) for Experiment 5..................................69 3-16 Distribution of the two host speciesÂ’ eggs by plant quartil es (chart) and by leaf number-derived quartiles (table) for Experiment 6.....................................69 3-17 Proportion of parasitized DBM eggs by quartile in Experiment 1 (2001)..........76 3-18 Proportion of parasitized DBM eggs by quartile in Experiment 1 (2002)..........76 3-19 Proportion of parasitized DBM e ggs by quartile in Experiment 3.....................76 3-20 Proportion of parasitized DBM e ggs by quartile in Experiment 4.....................77 3-21 Proportion of parasitized DBM e ggs by quartile in Experiment 5.....................77 3-22 Proportion of parasitized DBM e ggs by quartile in Experiment 6.....................77 3-23 Proportion of parasitized SL eggs by quartile in Experiment 1 (2001)..............78 3-24 Proportion of parasitized SL eggs by quartile in Experiment 1 (2002)..............79 3-25 Proportion of parasitized SL e ggs by quartile in Experiment 3..........................79 3-26 Proportion of parasitized SL e ggs by quartile in Experiment 4..........................79 3-27 Proportion of parasitized SL e ggs by quartile in Experiment 5..........................80 3-28 The distribution of DBM eggs and larv ae within the quartiles of the plant for Experiments 3 & 5.........................................................................................85 3-29 The weed Portulaca oleracea (purslane or verdolaga) bordering a field of cabbage...............................................................................................................95

PAGE 14

xiv 4-1 Petri dish design in the Trichogramma host choice experiments.....................103 4-2 T. pretiosum percent parasitism (separ ate hosts DBM & Looper)...............109 4-3 T. pretiosum percent parasitism (combined hosts DBM & Looper)............109 4-4 Parasitism levels of DBM and SL eggs in single host and combined host Petri dishes................................................................................................110 4-5 Total percent parasitism for single host and combined host Petri dishes.........111 4-6 Percent parasitism of DBM eggs by ‘conditioned’ T. pretiosum females........113 4-7 Percent parasitism of SL eggs by ‘conditioned’ T. pretiosum females.............114 4-8 Percent parasitism of combined DBM & SL eggs by ‘conditioned’ T. pretiosum females.........................................................................................115 4-9 Number of dry eggs in the combin ed-host Petri dish treatments and controls..............................................................................................................116 4-10 Percent parasitism of DBM eggs by the three Trichogramma species.............117 4-11 Percent parasitism of SL eggs by the three Trichogramma species.................117 4-12 Percent parasitism of combin ed DBM & SL eggs by the three Trichogramma species......................................................................................119 4-13 Percent parasitism means of DBM and SL eggs in the combined-host Petri dishes by the three Trichogramma species...............................................120 4-14 Number of dry DBM and SL eggs found in the combined host Petri dishes of the three Trichogramma species experiments...................................121 5-1 Seedbeds containing the cabbage seedlings......................................................138 5-2 Trichogramma release point and scouting flag in recently transplanted cabbage.............................................................................................................141 5-3 Field plan for on-farm IPM experiment............................................................142 5-4 Relative costs and cost br eakdowns for the crops grown.................................145 5-5 Relative labor requirements, divided by activity..............................................146 5-6 Farmer perceptions of crops grown in terms of profits ga ined, fluctuation of sales prices and fi nancial inputs needed.......................................................148

PAGE 15

xv 5-7 Harvest periods for the study crops, giving peak periods and approximate beginning and end times for the harvests..........................................................149 5-8 Crop budget for an acre of cabbage us ing the experimental IPM strategy for the control of DBM.....................................................................................157 5-9 Crop budget for an acre of cabbage us ing the regular insecticide-based strategy for the control of DBM........................................................................158 5-10 Years 4 and 5 of the 8-year LP mode l for Farm #8 with full incentives (Scenario 1).......................................................................................................162 5-11 Years 4 and 5 of the 8-year LP m odel for Farm #8 without agrochemical incentives (Scenario 2)......................................................................................162 5-12 Years 4 and 5 of the 8-year LP m odel for Farm #8 without agrochemical incentives or worker salary incentives (Scenario 3).........................................163 5-13 Pick-up truck taking planta in to market in San Juan........................................171 5-14 Cabbage grown in a multi crop system.............................................................181 5-15 Lizard found on one of the Trichogramma release stations.............................182 5-16 New housing developments in the central region of Puerto Rico.....................187 5-17 Images that represent ‘El Jíbaro ’, an important cultural icon of Puerto Rico........................................................................................................188 5-18 Farmers who collaborated in this farming systems study.................................189

PAGE 16

xvi LIST OF OBJECTS Object page 1 8 year model.........................................................................................................223 2 Model characteristics...........................................................................................223 3 Farmer configurations..........................................................................................223 4 Government incentives........................................................................................223 5 Output tables........................................................................................................223 6 2 year model.........................................................................................................223

PAGE 17

xvii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy HOST PREFERENCES IN Trichogramma AND HOW UNDERSTANDING THE DYNAMICS OF A FARMING SYSTEM MAY IMPROVE IPM RESEARCH By Richard William Hay Pluke May 2004 Chair: Gary Leibee Major Department: Entomology and Nematology Trichogramma pretiosum (Riley) (Hymenoptera: Chalcidoidea) parasitized both diamondback moth (DBM) (Linnaeus) ( Plutella xylostella , Lepidoptera: Plutellidae) and soybean looper (SL) ( Pseudoplusia includens ) (Hübner) (Lepidoptera: Noctuidae) eggs under field conditions in cabbage on the south coast of Puerto Rico. Parasitism levels varied and were affected by (a mong other factors) location of host eggs on the developing cabbage plants. Most of the eggs parasitized were found in the lower parts of the plants. Generally higher rates of parasitism were f ound in the DBM eggs than in the SL eggs. However, because of variation in parasitism levels, it could not be categorically said that T. pretiosum preferred the eggs of one species to the other, in this series of field experiments. Later experiments show ed that under laboratory conditions, T. pretiosum did prefer DBM eggs to SL eggs; and that pr ior exposure to one set of host eggs or the other did not have any effect on this host pr eference. This preference for DBM eggs to SL eggs was also exhibited by two other species of Trichogramma ( Trichogrammatoidea

PAGE 18

xviii bactrae (Nagaraja) and Trichogramma minutum (Riley)) that were examined. Of the three species, T. bactrae showed the highest percen t parasitism of SL eggs. A socioeconomic study was conducted in th e mountainous central region of Puerto Rico; and a linear programming (LP) model was developed to study the farming system of this region. A series of interviews a nd other data-gathering ex ercises gave detailed information on farm activities and the factors that influence the farming system. The LP model combined this information and gave a fair representation of the farming system. Most of the farm activity occurs during the Christmas period, while the summer months are normally a time of low activity. The governme ntÂ’s incentive scheme sustains many of the activities found; and leads to plantain being a dominant crop. Poor, unstable markets seem to be the biggest constraint to this sy stem. Agrochemical use is limited in many of the crops, especially roots and tubers. A ny introduced IPM technique would have to focus, primarily, on k eeping labor costs down.

PAGE 19

1 CHAPTER 1 INTRODUCTION Global Cabbage Production Cabbage is one of the vegetables in the Cruciferae family. These domesticated plants are thought to have originated ar ound 3,000 years ago from the wild cabbage ( Brassica oleracea L. subsp. oleracea ), indigenous to the co astal areas of Western Europe and Great Britain. Since then, this ancestral variety has been domesticated in different ways, giving rise to the group known as cole crops. Cole crops include cauliflower, broccoli, Brussels sprouts, kale , collards, and kohlrabi (Phillips and Rix 1993). These plants are now cultivated in most parts of the world. To tal world acreage of cabbage harvested in 2002 was 7,452,844 acres (Food and Agriculture Organization of the United Nations 2002). China had the biggest acreage at 3,631,668 acres; followed by India (691,895 acres); and the Russian Fe deration (444,790 acres). The FAO Production Yearbook (2002) lists 125 countri es with cabbage production. Cabbage was introduced to America in 1541-42, when Jacques Cartier brought it to Canada on his third journey (Boswell 1949). Sin ce then the plantÂ’s cu ltivation has spread greatly; and in 2002, 77,080 acres of cabbage were harvested in the United States of America (National Agricultural Statistics Service, US Dept . of Agriculture. 2003). This places it sixth on the USDAÂ’s list of larg est cabbage-producing nations (Economic Research Service, USDA 2003). Interestingly the FAO has the USA fourth on their list with 256,990 acres planted in 2002.

PAGE 20

2 Cabbage Production in Florida and Puerto Rico Cabbage ( Brassica oleracea L. var. capitata L.) is an important crop in Florida and in Puerto Rico. A total of 8,200 acres of cabbage was planted in Florida in 2001-02, most of which (7,700 acres) was planted betw een January and July. Yields were around 300 hundredweight per acre; and the total a nnual production value in the state was over $28 million for the year (Florida Agri cultural Statistics Service, 2003). In Puerto Rico, cabbage is a crop grow n in the small farms of the central mountains; and in the large, commercial farm s of the southern coast of the island. In 1997-1998, cabbage production contributed $145,000 to the IBA (Ingreso Bruto Agricola Gross Agricultural Income ) through the production of 9,000 hundredweight of cabbage, sold at a price of $16.08 per hundredweight (Agricultural Experimental Station, University of Puerto Rico, 1999). In 1998, around 210 cuerdas (1 cuerda = 0.971 acre) were planted (USDA Na tional Agricultural Stat istics Service, 1997). Although cabbage is important to many farm ers in Puerto Rico, its production is not large-scale, especially by US standards. In fact cabba ge production is declining in Puerto Rico—in 1987, 60,000 hundredweight we re produced. By 1997-98, the production figures were down to 9,000 hundredweight. One of the main reasons for this decline is the amount of cheap imported cabbage entering the island (mostly from the US). In 1998, 152,857 hundredweight of cabbage was importe d. This represents 88.1% of all the cabbage consumed in Puerto Rico (Agricu ltural Experimental Station, University of Puerto Rico, 1999). Another tr end is the move of cabbage production from small-scale mountain farms to larger, commercial farms on the south coast of th e island. In 1988-89, 95% of local production was in the mountains . By 1997, this had changed; and of the 210

PAGE 21

3 cuerdas planted on the island, 161 cuerdas we re found in the large farms on the south coast. Cabbage Insect Pests in Puerto Rico About 15 species of insect pests of cabbage in Puerto Rico have been recorded (Agricultural Experimental Station, Universi ty of Puerto Rico, 1999). These include the diamondback moth, whitefly, leafminers, a nd imported and gulf white cabbageworms and loopers. Currently, the diamondback moth ( Plutella xylostella ) (Linnaeus) is by far the most important pest. Both the cabbage looper ( Trichoplusia ni ) (Hübner) and the soybean looper ( Pseudoplusia includens ) (Walker) can be found, but the latter is the most common in the field. Loopers are the second most important pest of cabbage in the University of Puerto Rico’s Research Sta tion’s technological package on cabbage (1999), but with the increase of larger plantings of cabbage on the south coast, silverleaf whitefly ( Bemisia argentifolii Bellows and Perring) has become a bigger problem in the last few years, Diamondback Moth (DBM) and Development of Insecticide Resistance in Puerto Rico and Florida One of the reasons that farmers give for declining cabbage production in the mountains of Puerto Rico is DBM’s nega tive effect on profit margins. Diamondback moth is considered the most important pe st of cabbage on the island (Armstrong 1990). Apart from its inherent ability to do damage in a cabbage crop, it has also proven that it can develop resistance quickly to the inse cticides used agains t it (Shelton et al. 1993, Leibee and Capinera, 1995). This is one of the reasons why cabbage production has moved down to the large farms of the south coast. These large operations can better

PAGE 22

4 afford new insecticides such as spinosad (SpinTor, Dow AgroSciences) and indoxacarb (Avaunt, DuPont) recently available. The diamondback moth was identified for the first time in Puerto Rico by Herr Moschler (Rosario Perez 1984) from a specimen caught by Dr. Gundlach in 1890. In 1904, it was observed in cabbage, mustard, "nabos " and "coles.” In 1915, an illustration of leaves damaged by the larvae was made; a nd in 1918 illustrations were made of all the life stages. In terms of control practices, th e research stations in Puerto Rico were recommending Diazinon in the late 1960s, because the pest had become resistant to DDT and parathion. After 1972/73, and large outbreaks of the pest in the mountain farms, DBM was considered resistant to Diazi non. Methamidophos (Monitor) then became the recommended pesticide; which, along with e ndosulfan (Thiodan), ga ve control to the farmers again. In the 1980s, work on insecticid al resistance in the DBM of Puerto Rico showed that methamidophos, chlorpyrifos (L orsban), methomyl (Lannate), endosulfan, dimethoate (Cygon) and naled (Dibrom) were co mpletely ineffective in controlling this pest. Acephate (Orthene) and Bt were cons idered moderately effective (Armstrong 1983). Current use of Bt products such as Dipel 2X give adequate control in the mountain farms, while the newer compounds mentioned earlier ar e most frequently used in the farms of the southern coast. Leibee (1996) lists 31 insect pests of cabbage in Flor ida, but names DBM and the cabbage looper as the two main insect pests in the state. Like Puerto Rico, Florida has undergone similar cycles of new pesticid e releases, followed by development of resistance. An early citation of insecticid e resistance in Florida was by Wilson in 1957, who reported the failure of the insecticide T oxaphene in controlling cabbage insects in

PAGE 23

5 central Florida. The pyrethroids gave brief control during the early 1980s, but resistance to these insecticides soon developed (L eibee and Savage 1992a). During the 1990s, the insecticides of choice were th e Bt-based insecticides. Thes e, however, also came under threat, starting with developed resistance to the B. thurengiensis subspecies kurstaki (Btk) by DBM (Leibee and Savage 1992b, Shelton et al. 1993). This over-reliance on insect icides continues. One of the newest chemicals available for DBM control, spinosad (SpinT or, Dow AgroSciences), has already been circumvented by DBM through develope d resistance in Hawaii (Zhao et al . 2002). Diamondback moth is not unique in its developm ent of resistance to insecticides; cabbage looper (Genung 1957, Shelton and Soderlund 1983) and soybean looper (Boethel et al . 1992, Gianessi et al . 2002) have also developed resi stance to certai n insecticides. Insecticide resistance is a majo r stimulus to the efforts to develop alternative control measures. These measures, in part, look to re duce the selective pre ssure on insect pests caused by the over-reliance and over-use of in secticides. Integrated pest management strategies and the use of biol ogical control agents are part of this effort to reduce insecticide use. Farming in the Central Mo untains of Puerto Rico The farming systems study took place on the farms of three municipalities Barranquitas, Orocovis and Na ranjito (Figure 1-1).

PAGE 24

6 Figure 1-1. Municipalities of Puerto Rico, with the farming systemsÂ’ study area outlined by a broad band. According to the 1998 census (Puerto Rico Agricultural Statis tics Service 1998), Barranquitas ranked 9th in market value of agricult ural products produced by the 78 municipalities of Puerto Rico, with a total value of $14,227,929. The top five commodities by value of sales were poultry a nd poultry products, plantains, horticultural products, vegetables and melons, and root crops and tubers. Orocovis ranked 24th in market value of agricultural products, w ith a total value of $8,137,362. The top five commodities were coffee, plantains, vegetables and melons, cattle and calves, and bananas. Naranjito ranked 27th in market value of agricultu ral products, with a total value of $6,906,465. The top five commodities were po ultry and poultry products, plantains, cattle and calves, vegetables and melons, and horticultural produc ts. The top producing municipalities on the island are found on the northwest coast; where there is a specialization in the cattle and dairy industry. These three municipalities are among the seventeen municipalities that constitute what is known as the central region of Pu erto Rico. In 1992, this region accounted for

PAGE 25

7 54% of the island’s farms (Carro-Figuero a and Alamo-González 2002). Most of these farms (69%) were of 20 acres or less, with an average farm size of 28 acres. As with the rest of the island, the number of farms in the region is declining. Since 1950, the island has lost 65% of its agricultural land. Ownersh ip of land is dominated by private farm ownership. In 1992, 88% of the farms were pr ivately owned and opera ted; while only 6% of the farms were rented. The number of farm workers dropped in the central region between 1978 and 1992 (7%), but this was not as high as the rest of the island (32%). The age of the farmers increased during this same period, as did the level of schooling, with only 15% of the farmers attaining less than sixth grade schooling. Most of the farmers (58%) had been farming for more than 10 years; and in 1992, 65% of the farmers considered farming to be their principal occupation, with 49% of them stating that their biggest source of income was from farming. That only 52% of the farmers in the central region gained 25% or more of their inco me from farming is surprising; but may indicate th e level of support they get fr om federal and state welfare programs. Many of the patterns of crop production have changed in the last 25 years. In coffee, for example, yields have grow n considerably (95%) between 1978 and 1992, partly due to the shift from coffee grown under shade to coffee grown in the open. The haciendas that traditionally grew the bulk of the coffee gr own on the island disappeared in the 20th century; and coffee pr oduction declined significantly. Since the 1970’s, however, some bigger coffee producers have reemerged, (which, with agronomic changes, has led to increased production). Nearly all of the coffee production on the island occurs in the mountains.

PAGE 26

8 Plantain has always been an important crop in the central region. In 1978, 68% percent of the island’s plantain production was in this region. In the last few years, stiff competition has come from large producers on the south coast. By 1992, the central region produced only 54% of the island’s planta ins. A similar pattern is found in banana, where the central region’s pr oportion of output has declined from 85% in 1978 to 71% in 1992. In the last 25 years, root crops such as tanier (yautía) and yam (ñame) have become concentrated in the central region because of large decreases in acreage in other parts of the island. Historical Development of the Agri cultural Sector of Puerto Rico Like many of the Caribbean islands, the ea rly colonial agricu ltural sector was dominated by sugar cane, which was intr oduced by the Spanish around 1514 (Thompson and Taylor 1992). Another pervading difficulty was a relatively small workforce, which in the 17th century a governor tried to bolster by introducing th e death penalty for those who tried to leave the island for the more w ealthy South and Central American colonies. Puerto Rico held strategic importance for th e Spanish ships traversing the Caribbean Sea but had no real means of sustaining itself. In fact, for many years, the island’s economy relied on Mexican governmental yearly subsid ies (situado) that be gan the country’s long history of dependence (Trías 1997). Agricultur al trade was also not helped by an early Spanish royal decree stating that the island c ould only send exports to the port of Seville. Although trading restrictions were eased, disi nterest in human and capital investment pervaded the island’s economy. In the 18th century, things improved as trading restrictions were eased, and with the influx of skilled settlers that occurred, in part, because of the wars of independence on the mainland of South America. The next century saw further economic advances with

PAGE 27

9 the entry of large numbers of slaves ( 50,000+ slaves inhabited the island by 1846), and the developments of the sugar industry. Puerto Rico could now trade with countries other than Spain. The other important export crop was coffee, which reach ed a peak in the early 19th century and then again la te in the same century. Coffee was mainly exported to Europe because the North American market had captured much of BrazilÂ’s output; while sugarcane was mainly exported to North America because of the development of sugar beet in Europe. Slave labor was used to gr ow the sugarcane on the flat coastal plains while resident farmers grew coffee on smallholdings in the mountains. The coffee sector lacked infrastructure because of indebtedness of the farmers; and to the lack of investment by the immigrant merchants who had control of the credit and marketing facilities. Other than coffee and sugarcane, the island had historic ally produced tobacco, hides and ginger for export. In 1898, as part of the Spanish-American War, the US easily captured Puerto Rico, a few months after having captured Cuba. The Foraker Act was passed in May 1900; and it returned the island to civilian rule (Ameri can-led). The act described how Puerto Rico was to be governed; and its economic provi sions favored the US corporations. These provisions did not change with the Jones Act of 1917, which gave Puerto Ricans US citizenship. The economic article s of the Foraker Act were Puerto Rico couldnÂ’t sign trade ag reements with foreign countries Puerto Rico could not dete rmine itÂ’s own tariff system Puerto Rico had to use US shipping lines when trading with the US Puerto Rico must adopt the US monetary system

PAGE 28

10 Puerto Rico prohibited corpor ations but not individuals from owning or controlling more than 500 acres of land. The adoption of US currency had the gr eatest impact; and the circulation of currency decreased by 20%, which induced a severe credit shorta ge and large-scale selling of land. This credit shortage led to the entrance of US banks, which began offering loans to qualified borrowers. Coff ee producers generally didnÂ’t qualify. Another blow to the coffee producers access to credit arose indirectly from the effects of the devastating hurricane San Cicero of 1899, wher ein 80% of the coffee harvest and 60% of the coffee trees were lost. Legislation wa s passed to prevent fo reclosure on land for unpaid debts incurred from the hurricane. Un fortunately, this provision also meant that the land could not be used for collateral, e ffectively closing the credit market to small coffee producers. This (and the general inde btedness of the coffee sector) meant that there were further land sales. Although the Foraker Act meant that tari ffs against the US were dropped, this didnÂ’t help Puerto Rican coffee, because ot her Latin American countries already had no tariffs on their coffee being imported into the US. Puerto Rico also was beginning to lose its traditional markets. Spain now consider ed it a foreign country and imposed heavy duties on coffee exports. Cuba did the same as it tried to e xpand its local coffee production. The result of these changes meant that coffee, which in 1896 had represented 77% of all exports, was, by 1935, providing less th an one half of one percent in value of total exports. Sugarcane had a different fate under Am erican influence. Between 1900 and 1939, production increased tenfold; and by 1935, almo st 75% of the population acquired their living from the sugar industry (Thompson and Taylor 1992). The sugar industry was

PAGE 29

11 dominated by large American corporations, e xport quotas, and the influence of stateside refiners that minimized the export of valu e-added refined sugar (Weisskoff 1985). Much of the labor came from ex-slaves; and from peasants from mountain communities. The social relations of production that aros e between employers and workers greatly increased urbanization in Puerto Rico, wher e workers could be accessed for the seasonal work of sugarcane. The history of Puerto Rican agricultu re shows a lack of investment and infrastructural development; and describes a sector dominated by foreign powers set on serving their own needs. It is no surprise, then, that with the effect of the Great Depression on the sugar industry, and with the great social changes brought about after the Second World War, the agricult ural sector suffered a decline. Agricultural decline (1940-1980s). Between 1949 and 1978, the total number of farms decreased from 53,515 to 31,837; and the percent of all land in farms dropped from 81.8 to 48.1 (Dietz 1986, Weisskoff 1985). In 1987, less than 4% of the population worked in agriculture. A telling statistic is that local agri culture as a source for food consumption fell from 51.2% in 1949 to only 13.3% in 1978. This shor tfall in local food production in Puerto Rico was filled by food imports, which rose from 37.4% of all consumption to 45.6%. Much of the decline was fueled by polit ical and economic changes that were brought about to ameliorate the plight of th e growing number of landless laborers. Land reform movements led to the formation of the Popular Democratic party (PPD) in 1937 and their victory over the landed elite in the polls in 1940. Between the PPD’s “ Autoridad de Tierras ” and Franklin Roosevelt’s Puerto Rico Reconstruction

PAGE 30

12 Administration (PRRA), corporate land was bought to settle families; and subsistence farms established on lands that had been part of large plantations. This resettlement still did not provide enough employment; and after some initial experimentation by the PPD with import-substitution strategi es, they collaborated with the US federal government to introduce a developmental model called “indus trialization by invitation” (Dietz 1986, Pantojas-Garcia 1990, Picó 1998, Weisskoff 1985). In 1947, Operation Bootstrap was implemen ted in Puerto Rico. This program offered tax-free profit earning and repatria tion to US companies operating factories in Puerto Rico; and was supported by local technical advice, lo cation assistance, employment contracting, and financing. Be fore Operation Bootstrap, in 1941, Rexford Tugwell (the island’s governor) tried to prom ote public sector industrialization but it was not well received. Perversely, the new politi cal leaders had, with Operation Bootstrap, fallen into the traditional colonial role of managing profit-making endeavors for external economies. The original idea of the devel opment model had been to expand privatesector industrial employment; and to fome nt backward and forward linkages among industrial and agricu ltural sectors. Operation Bootstrap was in some re gards a success. By 1964, over 2,000 US firms had located factories in PR; an d some 100,000 jobs had been created. Unfortunately, during this peri od the agricultural sector wa s neglected, and many laborers had either gained jobs in the new factories or had left for jobs in th e United States (outmigration totaled some 750,000 between 1945 and 1953). Due to partial implementation of minimum wage legislation in the 1950s, Puer to Rico lost its adva ntage as a low-wage country; and this led to a reduction in labor-i ntensive factories and an increase in capital-

PAGE 31

13 intensive factories. High levels of out-mig ration prevented destabilizing unemployment rates; but in the early in 1970s , with industrial growth halte d by the oil crisis, something had to be done. In 1975, the federal Food Stamp program was extended to Puerto Rico; and by 1980, over half of the Puerto Rican populati on was receiving cash benefits from the program. Income derived from the Food Stamp program and other federal welfare programs accounted for 20% of Puerto RicoÂ’s GNP (Grosfoguel 1994). The industrialization of the 20th century and implementation of the Food Stamp program meant urbanization of the population and a shif t in their consumer patterns. Local food production had historically been marginalized to make room for sugar cane and coffee, subsisting in the mountains and around th e plantations. At the beginning of the 20th century, many farms had been sold to gain US currency; and during the industrialization process, many farmers had left their farms altogether. Now, the Food Stamps program had extended the monetary economy into peop leÂ’s lives, undercut ting the traditional system of barter and reciprocity. This imp acted on the local agriculture by redirecting food purchases from farmersÂ’ markets into supermarkets; and changed the way and form in which products were sold. This necessitate d an increased number of intermediaries and higher product volumes, which favored exte rnal imports. Local food producers were further marginalized. Problems facing the agricultural sector stretc h to the wider reality of Puerto RicoÂ’s existence. Weisskoff (1985: 59) describes how the economic initiatives and federal interventions of the 20th century brought about an economic reality in which many Puerto Ricans were disenfranchised to the bette rment of various US capital sectors

PAGE 32

14 In short, the US public underwrites the Puer to Rican people, while US corporations shift profits through their Pu erto Rican plants and back to the United States, tax free. The Puerto Rican family then buys it’s consumption needs, which consists for the most part of imported goods, shifting its public grant money back to the US private sector. As its own economy decomposes, Puerto Rico becomes the revolving door for funds flowing from the American public back to the American corporation. . . . Meanwhile, these corporations employ only a token Puerto Rican work force. The rest of the island’s people live peaceably on welfare, or not so peaceably in the ghettoes of American industrial cities. Puerto Rico’s Agricultural Intensification Program In 1978, the government of Puerto Rico put forward an intensification program to improve the productivity of the agricultu ral sector. The program was based on the application of capital-intensiv e technologies to many of the agricultural enterprises. Agricultural land was classified by soil ty pe, topography, and prevailing climate; and then divided by the types of crops that woul d best be grown on that land. Only in the assigned areas would the cultivation of a crop be fully supported. The production incentives were offered to farmers growi ng the designated crops at an “efficient” commercial scale that excluded many (Carro -Figueroa 1985). There was also a bias toward younger and better-edu cated farmers. The “traditi onal, low income” producers would receive limited subsidies “until retirement or other forms of attrition” (US Dept. of Commerce 1979: 302). Thus a system of in centives was set up that promoted the intensive use of agrochemicals and mechanizati on of farming practices. This in turn gave more support to the larger commercial farm s on the coastlands; and less to the smaller farms of the mountains. Univer sal to all, were reliance on agrochemicals and the singlecrop technological packages. Little thought was given to the complex, beleaguered farming systems of the mountains that continue d to find ways of feeding the island in the face of unregulated markets a nd changing social realities.

PAGE 33

15 CHAPTER 2 LITERATURE REVIEW Natural Enemies of the Diamondback Moth (DBM) and Soybean Looper (SL) Diamondback moth in the first half of the 20th century was not considered a major pest of cabbage, but just one of a comple x of caterpillars in cabbage (Leibee 1995). Cabbage looper (Hübner) ( Trichoplusia ni ) and to a lesser exte nt the soybean looper (Walker) ( Pseudoplusia includens ) were considered more dangerous pests. Part of the reason for DBM’s limited effect was the s uppression of its populations by natural enemies (Talekar and Shelton 1993). These na tural enemies in more recent times have not lowered DBM populations as before (because of inappropriate ins ecticide use) and so now DBM is considered a much more dange rous pest than the loopers (Leibee 1995; Mitchell et al. 1997). The most important hymenopteran pa rasitoid of DBM in Florida is Diadegma insulare (Cresson). In Puerto Rico, the most common natural enemies of DBM are Diadegma insulare , Cotesia plutellae (Kurdjumov) and Spilochalcis spp. (AbreuRodriguez and Cruz 1997). The loopers also ha ve their guild of natural enemies: these include the tachinids Chetogena scutellaris (Wulp) and Voria ruralis (Fallen) , and the braconids Meteorus autographae (Muesebeck) and Cotesia spp. In Puerto Rico, the main parasitoids found in soybean looper were tentatively identified as Voria ruralis , Cotesia spp. and Copidosoma floridanum (Ashmead) (personal observations) (Figures 2-1 through 2-3).

PAGE 34

16 Figure 2-1. Soybean looper parasitized with Copidosoma floridanum . A) Three dead looper larvae discolored by internal C. floridanum pupae. B) Copidosoma floridanum adults emerging from parasitized looper pupae. Figure 2-2. Looper parasitized with Voria sp. A) Four empty Voria sp. pupae in a parasitized looper larvae. B) Voria sp. adult. Figure 2-3. DBM parasitized by Cotesia spp. A) Parasitized DBM pupae. B) Cotesia sp. adult. A B A B A B

PAGE 35

17 Diamondback Moth (DBM) Biology and Development The diamondback moth (Linnaeus) ( Plutella xylostella , Lepidoptera: Plutellidae) is globally considered to be the most important pest of crucifers (Tal ekar and Shelton 1993) and its known presence in the USA dates from the 19th century. In tropical and subtropical regions, its life cycle can last as little as 20 days (Capinera 2000) and more than 12 generations can be completed in a year . Perez (1984) conducted work on DBM in Puerto Rico, and found that the eggs take 3 days to develop, the larvae 11.7 days, the pupa 3.6 days, and the adult 17.8 days. DBM e ggs are normally yellow in color and are relatively small (0.44 mm long and 0.26 mm wide). They lie flat to the leaf and are often found in depressions or folds. Often the e ggs are oviposited in large numbers, following the crease of a leaf, for example (Figure 2-4). There are 4 larval instars of DBM, and the first instar tends to mine the leaves. The other larval instars feed exte rnally and characteri stically eat the me sophyll layer of the cabbage leaves, leaving the transparent epid ermis and cuticle. This feeding behavior produces the ‘windows’ and hol es commonly seen with DB M damage. Once the fourth larval instar is complete, the insect spins a loose cocoon attached to the leaves. Before long, the adult emerges and soon begins to ma te. The adult is a slender moth, between 6– 8 mm long. The insect gets it name from the pattern seen on the dorsal side of the adult male when in repose. There are pale, triangul ar markings along the inner margin of the wings, which in rest form a row of three di amond-shape pale patches. The adult female starts ovipositing a few days after emergence and continues to do so for about another 10 days. The adults are relatively weak flyers a nd can be found close the plants they inhabit. In a row crop situation, the adults often can be found in relatively large numbers at the

PAGE 36

18 end of the rows and at the edges of the fields. Diamondback moth can be found throughout the year in Puerto Ri co when cabbage is planted. Figure 2-4. Diamondback moth life stages. A) healthy DBM eggs near leaf vein. B) 4th instar larvae. C) Diamondback moth cocoon. D) Diamondback adult with larval damage. Soybean Looper Biology and Development The soybean looper ( Pseudoplusia includens ) (Walker) (Lepidopter a: Noctuidae) is a generalist feeder and can be found on many different types of vegetable crops. Its potential to do damage is gr eat. Reid and Greene (1973) found that it took soybean looper larvae about 14 days from egg eclosion to pupati on in Florida. In this time, the larvae ate A D B C

PAGE 37

19 an average of ~82 cm2 of soybean leaves (at 29oC). Most of this feeding occurred in the last 3 stadia (only 3% of the leaf material was eaten in the first 3 stadia). The pupal stage lasts around 7 days before the adult moth emerges (Reid and Greene 1973, Mitchell 1967). The female begins oviposition about 4 days after emergence and does so on the underside of the leaves. The translucent wh ite eggs are 0.6 mm in diameter and 0.4 mm in height, and look like slightly deflated beach balls, affixed to the leaf by their flat base (Figure 2-5). Longitudinal ridges run down the egg chorion. The eggs are oviposited singly or in groups of 10 or mo re. Normally the eggs take ab out 5 days to hatch but can take as long as 10 days at 15º C (Jackson et al. 1969). Figure 2-5. Soybean life stages. A) Soybean looper egg. B) Soybean looper larva and cocoon. C) Final instar SL larva. D) Soybean looper adult. B A C DBy kind permission of John Himmelman

PAGE 38

20 Trichogramma (Hymenoptera: Chalcidoidea) Previous research on the natural enem ies of DBM has focused largely on hymenopteran, larval pa rasitoids such as Diadegma insulare (Cresson) and Cotesia plutellae (Kurdjumov) and there has been relatively litt le research into the use of egg parasitoids. Egg parasitoids ar e, however, inherently importa nt to the control of DBM. They have the ability to reduce pest populat ion levels early, prev enting feeding by newly emerged DBM larvae. Members of the family Trichogrammatidae (from now on referred to as Trichogramma ) are some of the best-known egg parasitoids. They are minute (from 0.2 to 1.5 mm in length) endoparasites and were first described by Westwood in 1833. They have been used extensively in the c ontrol of pests such as the corn earworm (Boddie) ( Helicoverpa ameriga ) and the sugarcane borer (Fabricius) ( Diatraea saccharalis ) since earlier this century (Meier 1941, Hirai 1991, Klemm and Schmutterer 1993, Sato et al. 1994, Wuehrer-Bernd and Ha ssan 1993). They also have proved successful in the field against DBM (Glass et al. 1981, Nguyen and Nguyen 1982, Vattanatangum 1988), and there is r ecord of collecting an indigenous Trichogramma species from DBM eggs in the Caribbean (Bennett and Yaseen 1972). Having said that, there has not been the same level of research on Trichogramma as there has been on parasitism of DBM by larv al parasitoids. It is thought by some that Trichogramma , as a generalist parasitoid, does not preferentially parasitize DBM e ggs due to their small size (Figure 2-6). If true, Trichogramma might be a bad candidate for use as a biological control agent of DBM in a multi-host situation.

PAGE 39

21 Figure 2-6. Diamondback moth (left) and soybe an looper (right) eggs either side of Pieris brassicae egg. Trichogramma is also known to parasitize l ooper eggs (Oatman et al. 1968, USDA 1973, Allen and Gonzalez 1974, Martin et al . 1976, Godin and Boivin 2000), although little natural parasitism in eastern North America is thought to occur (Lundgren 2002). Looper eggs do seem to have many of the preferred characteristics for Trichogramma parasitism such as relative size and lack of protective scales or hairs. Soybean looper eggs look very similar in size, shap e and color to corn earworm eggs, Helicoverpa zea (Boddie) (Lepidoptera: No ctuidae) and the eggs of the tobacco budworm, Heliothis virescens (Fabricius) (Lepidoptera: Noctuidae) , which are well-documented hosts of Trichogramma . The looper eggs are slightly more defl ated in their look than the eggs of the species mentioned above. Asid e from physical similarities, Trichogramma Â’s acceptance of SL eggs could depend on many other things such as internal chemical composition, chorion characteristics and ex ternal chemical residues for example. Trichogramma Â’s ability to recognize and use a va riety of host species is considered to be an evolutionary adaption to its multivol tine life cycle and limited capacity to control its dispersal in the field (Sachtleben 1929). Its polyphagous nature does not signify, however, that it is indiscrimina te in its choice of host eggs. Trichogramma carefully determines the number and the sex of the eggs laid into chosen hosts. It does so by an

PAGE 40

22 assessment of host size, age, nutritional suitab ility and previous parasitism of the host egg (Schmidt 1994). Many of these characteristics have been examined previously and some of the key references are given below Host curvature (De Jong and Pak 1984) Size and shape of host (Klomp and Teerink 1962) Host age (Pak 1986) Chorion thickness and architecture (Quednau 1855, Consoli et al. 1999) Interior chemical cues (Nettles et al. 1983) Egg surface chemical factors (Pak and De Jong 1987, Bin et al. 1993). Trichogramma Host Preferences Experiments have been conducted to quan tify the relative accept ability of DBM by Trichogramma over other hosts. Wuhrer and Hass an (1993) used the “contact and parasitism” method to esta blish the preference of 47 Trichogramma strains and 2 Trichogrammatoidea strains for acceptability of DBM over Sitotroga cerealla (Olivier). Trichogrammatoidea bactrae (Nagaraja), Trichogramma chilonis (Ishii), T. pintoi (Voegele) all showed a strong prefer ence in parasitizing DBM eggs over S. cerealla eggs (5, 3, 3 times as great, respectively). Trichogramma pretiosum showed no preference, although the species did visit DBM eggs w ith greater regularity. In experiments conducted by Klemm et al. (1992), however, T. pretiosum showed the highest percent parasitism of DBM and the highest searching capacity of all the 27 Trichogramma / Trichogrammatoidea species or strains studied. This, perhaps, alludes to the intraspecific variation found within Trichogramma species and the need to establish the efficacy of local populations. Host preference and suitability of the target pest is important to inundative release programs host mortality is primarily linked to the number of eggs the female adult parasitoid probes with her ovipositor (Be noit and Voegele 1979). She will only probe

PAGE 41

23 after a critical assessment of the egg’s suitability. If a host is not suitable, the female parasitoid will leave during her transectional walks over the egg, or even before that. For inoculative releases, success depends not onl y on the parasitism and host feeding levels of the initial natural enemies released, but al so on the actions of resulting progeny that develop. The efficiency of the resulting ge nerations thus depends on the reproductive potential expressed by the parasitoids in the field. The reproductive potential is greatly a ffected by host quality. Host quality for Trichogramma , as described before, is dependent on size, age, nutritiona l suitability and previous parasitism of eggs. DBM eggs are very small eggs that weigh on average 0.02 mg. This is smaller than the eggs of Ephestia kuehiella (Zeller) (0.03 mg—0.5 x 0.3 x 0.3 mm), a common factitious host, which often represents the small host egg size for Trichogramma species. It has been shown that Trichogramma wasp body size, fecundity, and longevity is dependent on larval f eeding (Charnov and Skinner 1985). Below a certain host egg volume, small adults emerge , as developing larvae are limited by nutrient and volume available for development. Ephestia kuehniella eggs can support the growth of one and sometimes two Trichogramma larvae (superparasitism is common in Trichogramma ) whereas a large host egg su ch as that belonging to Manduca sexta (L.) (1.36–1.45 mm in diameter) can support up to 40 larvae (usually 18–25 larvae). In general, most Trichogramma species are reported to prefer intermediate to large host eggs, often 0.8–1.8 mm in di ameter (Schmidt 1994). Miura and Kobayshi (1995) compared the reproductive properties of Trichogramma chilonis (Ishii) that were reared using DBM and E. kuehniella eggs at different temperatures. They found that the number of mature eggs in the ovary at

PAGE 42

24 emergence, and number of eggs laid on the initial day after emergence of T. chilonis raised on DBM was significantly smaller than for T. chilonis raised on E. kuehniella . This disparity increased with increasing temp erature. DBM would seem to limit the reproductive capacity of T. chilonis as compared to E. kuehniella . Would Trichogramma express this as a host preference for E. kuehniella were both host species found in the field? In their experiment, Miur a and Kobayashi did not give the T. chilonis a choice of hosts. Although the literature does seem to suggest that larger eggs ar e intrinsically more favorable to Trichogramma oviposition, there are many other factors that will eventually decide whether the female will oviposit. Mansfield and Mills (2002) showed that Trichogramma platneri (Nagarkatti) spent more time on heavier host eggs, but that the probability of successful parasitism was related to structural integrity of the chorion of the host egg. The chorion of an insect egg can vary greatly between different species: in thickness, architecture and chem ical or structural compos ition. Consoli (1997) found that the Trichogramma he was working with took differing amounts of time to drill eggs of different host species, even though they had the same chorion thickness. He found that eggs differed in the protein density a nd thickness of the exochorion layer. Vasquez et al. (1997) found that of the six Trichogramma species tested in the laboratory, T. pretiosum , T. minutum and T. bactrae gave greatest mortality of DBM eggs. They also found that T. bactrae and T. pretiosum gave the greatest level of parasitism (69–72%). The three species that gave greatest DBM mortality in the above study were chosen as the three Trichogramma species for this present work.

PAGE 43

25 Trichogramma pretiosum (Riley) This is thought to be the most common Trichogramma species found in the Western Hemisphere (Pinto et al. 1986, Zucchi et al . 1996) and it has been documented parasitizing the cotton leafworm ( Alabama argillacea ) (Hübner), velvet bean caterpillar ( Anticarsia gemmatalis ) (Hübner), sugarcane borers ( Diatraea spp.), Heliothis armigera (Hübner), cabbage looper ( Trichoplusia ni ), tomato pinworm ( Keiferia lycopersicella ) (Walshingham), Indian mealmoth ( Plodia interpunctella ) (Hübner) and others (Li-Ying 1994). Hassan (1993) documented that T. pretiosum was released commercially in the USA in cotton, corn and soybean. It was also re leased in Peru on cotton and in sugarcane, tomato and sorghum in France. It has been the only egg parasitoid found naturally in DBM eggs in North America (Oatman a nd Platner 1969). Wuehrer-Bernd and Hassan (1993) showed that T. pretiosum had a high egg-laying capacit y (53.7 eggs per female as compared to 40.5 eggs per female for T. bactrae ). Monje et al. (1999) showed that T. pretiosum preferred to parasitize the smaller Sitotroga cerealella (Lepidoptera: Gelechiidae) eggs to the larger Diatraea rufescens (Lepidoptera: Pyralidae), and Diatraea saccharalis (Lepidoptera: Pyralidae) eggs. Trichogrammatoidea bactrae (Nagaraja) Much of the interest in T. bactrae and the parasitism of DBM originates from work conducted in Asia. In one Thai study, T. bactrae’s parasitism of DBM ranged from 16.2 to 45.2% (Keinmeesuke et al. 1993). Trichogrammatoidea bactrae has also been found to be an indigenous egg parasitoid of DBM in China (Yurong 2002), and is considered the best candidate for further work (Yurong pers. comm .). Hassan (1993) documented that T. pretiosum was released commercially in the Phil ippines on sugarcane, and in Malaysia on cocoa. A weekly release of 50,000 adults per hectare was found to control DBM

PAGE 44

26 (Krishnamoorthy and Mani 1999, Krishnam oorthy et al. 2002). Wuehrer-Bernd and Hassan (1993) showed that T. bactrae was one of the three species of 47 strains and species of Trichogramma that parasitized DBM most eff ectively in the laboratory. Each female laid an average of 40.5 e ggs when parasitizing DBM eggs. Trichogrammatoidea bactrae also showed a strong preference for DBM eggs over Sitotroga cerealella eggs. Naranjo (1993) found that T. bactrae was well adapted to high temperatures and that its survival under high, fluctuating temperatures was similar to T. pretiosum . Klemm et al. (1992) found, however, that although T. bactrae parasitized well in the laboratory, it performed poorly in the field. Lundgren (2002) confirmed poor field impact of T. brassicae . Trichogramma minutum (Riley) Trichogramma minutum is an important natural enem y of the grape berry moth in northeastern USA (Nagarkatti et al. 2002). Hassan (1993) documented that T. minutum was released commercially in Honduras in banana . It is also an important parasitoid of the spruce budworm Choristoneura fumiferana (Clemens) in Canada (Quale et al. 2003), and Pinto and Stouthamer ( 1994) consider that the T. minutum complex includes what are the most widely used biological control agen ts in North America. Walcott (1948) called T. minutum a ‘tropicosmopolitan parasite’ and ha d found it on many kinds of butterflies and moths in Puerto Rico, although it was best known for its parasitism of the sugarcane borer, Diatraea saccharalis (F.). Walcott also mentioned that, as early as 1915, researchers had published work recognizing th e influence of rainfall and the non-burning of trash in sugarcane to pr eserve natural populations of T. minutum . He also noted that inundative populations of T. minutum released in sugarcane fared better in the warmer part of the year and in the more humid areas of the Puerto Rico.

PAGE 45

27 Host Cues Mediating Trichogramma Searching Parasitoids have to balan ce the finding of food with th eir reproductive requirements (Lewis et al. 1998). Their efforts are conduc ted at multiple spatial levels, from the ecosystem to the microhabitat. There has been increasing atten tion paid to trophic interactions between plant, host and parasitoid/predator (P rice et al. 1980, Karowe and Schoonhoven 1992, Dicke and van Loon 2000, Aldrich et al. 2003). There seems to be a complex chemical interaction between the three trophic levels that mediates how biological functions are best fulfilled. Parasitoids use both physical and chemical cues (kairomones) derived from the host species to find host eggs. Most Trichogramma have a broad range of host species and so in any particular habitat they could po tentially be confronted with a complex configuration of host species, necessitating a non-species spec ific searching technique. Paramount to foraging success is the use of hos t-derived cues. These can be chemical or physical. Chemical cues are very importan t and have been well documented. They can have a long-range effect, as when parasitoids respond to sex pheromones emitted by the host adults, (Noldus et al. 1990, Boo and Yang 2000), or a shorter-range effect such as responses to host frass (Lewis and Tum linson 1988, Ngi-Song and Overholt 1997) and wing scales (Shu and Jones 1989). These short-ra nge chemical elicitors are considered to function more as intensifiers of foraging and parasitizing behaviors rather than as attractants per se (Schmidt and Carter 1992, Frenoy et al. 1992). The physical cues are considered to be le ss important than the chemical host cues (Cloutier and Bauduin 1990), and th ey generally work at a more immediate level than the chemical host-cues. Nonetheless, physical cues are crucial to successf ul host location and acceptance. Both Völkl (2000) and Fischer et al. (2001) point out that parasitoid host

PAGE 46

28 location is a sequential process, structured in a hierarchical manner that depends on multi-sensory cues at different special and temporal levels. Physical cues such as vibrations, sound, color, shapes or movements have all been shown to elicit responses to hosts (Alphen van and Vet 1986, Vinson 1981, 1984), normally at the latter stages of host location and at the preliminary stage of host acc eptance. Calvin et al. (1997) show that with the lack of visual cues, Trichogramma can choose less suitable hosts in which to oviposit their eggs. Searching Techniques of Trichogramma Females and the Influence of Plant Characteristics It seems that Trichogramma generally shows a greater fidelity to a microhabitat than to taxon of host (Pinto and Stoutham er 1994) and so mastering the microhabitat, through efficient searching, is paramount to the reproductive success of the Trichogramma females. Number of offspring pr oduced per unit of searching time is usually considered the m easure of reproductive success (van Alphen and Vet 1986). Trichogramma use cues from habitats to identify preferred plants or habitat locations, which stimulate them to forage and parasiti ze. Altieri et al. (1981, 1982) showed that extracts of Amaranthus spp. and corn increased Trichogramma Â’s parasitism of H. zea eggs in a number of cropping systems. The reve rse is also true, wher e plant extracts can repel foraging Trichogramma (Bar et al. 1979, Cabello a nd Vargas 1985). Interesting tritrophic relationships exist where volat iles emitted from plants under attack from herbivorous insects are used as synomones for the attraction of parasitoids (Turlings et al . 1990, 1993). Trichogramma is known not to move far from th e point of release (Yu et al. 1984) and so the location of releas e has an influence on the resu lting pattern of parasitism

PAGE 47

29 within the plant (Hendricks 1967, Keller et al. 1985). Their lim ited ability to direct their direction of flight and to travel long distances means that many of the long-range chemical plant cues are not as important to host finding in Trichogramma as it is with other parasitoids. Researchers have shown, how ever, that the physical characteristics of the plants do have a great effect on Trichogramma Â’s ability to find and parasitize host eggs (Ables et al. 1980, Rome is et al. 1998). Lukianchuk and Smith (1997) showed that T. minutum found eggs faster on simple fo liage and that older female Trichogramma were more successful than younger females at finding spruce budworm eggs. They also found that wasps, which spent more time wa lking, had significantly less probability of finding an egg mass and were significantly sl ower at finding them. In their study on the effect of plant structure on searching efficiency of para sitoids, Andow and Prokrym (1990) reported that Trichogramma nubilale (Ertle and Davis) parasitized 2.9 times more Ostrinia nubilalis Hubner eggs and found them 2.4 times faster when foraging on simple artificial paper wax models than on comple x ones. Other plant f actors that affect a parasitoidÂ’s ability to find and parasitize hos t eggs include changes in plant size (Thorpe 1985) and changes in plant surface area (Need and Burbutis 1979, Kanour and Burbutis 1984). Gingras and Boivin (2002) described pl ant structure using three criteria: size, heterogeneity and connectivity. Size refers to the height or volume of the plant, heterogeneity refers to presence of different types of plant struct ures and connectivity refers to abundance of connections between the various parts of the plant. The cabbage plant is a plant that changes dramatically wi th maturation, not only in complexity but also in size or volume. One mitigating characteristic is that as the plant gets bigger, there is an increase in connectivity due to the lack of ve rtical extension of the internodes. Gingras et

PAGE 48

30 al. (2002) determined that in terms of size, cabbage, during its growth, remained ‘smaller’ than broccoli and Brussels sprouts; a nd in terms of heterogeneity, it remained at zero as nothing but vegetative plant structur es existed (unlike broccoli and Brussels sprouts). In terms of connectivity, it ranked hi gher than broccoli but lower than Brussels sprouts. The behavior of the parasitoids on the pl ants can be species specific. Some, like T. minutum and T. pretiosum, prefer the upper regions of the crop or stand (Smith 1988, Kot 1964 and Gonzalez et al. 1970). Others prefer the lower regions (Burbutis et al. 1977). There are other species that show no pref erence (Newton 1988). Martin et al. (1976) released T. pretiosum into large field cages that contai ned seven different crops (collards, cabbage, bell peppers, tomatoes, soybeans, bush beans and tobacco) infested with two species of Plusiinae and Heliothis spp. They found that T. pretiosum parasitized soybean looper eggs in all crops, incl uding cabbage, although at lowe r levels than the cabbage looper eggs. The Plusiinae eggs most heavil y parasitized were found on tomato. Apart from this difference, there was no real diffe rence found in parasitism levels between the other crops, even though they varied in plant height and architecture. Parasitoid Learning Generalist parasitoids such as Trichogramma can face great variability in host type and host location. A degree of plasticity in a parasitoid’s foraging abilities would improve its abilities to deal with such a variable e nvironment. This would be especially true if there was associative learning involved in the parasitoid’s adaption to a specific habitat. Papaj and Prokopy (1989) suggested three criter ia by which to define learning. These are that (1) an individual's behavior changes in a repeatable way as a consequence of experience, that (2) behavior changes gradua lly with continued expe rience, and that (3)

PAGE 49

31 the change in behavior accompanying experi ence wanes in the absence of continued experience of the same type, or as a cons equence of a novel experience or trauma. Bjorkstgen and Hoffmann (1995) found that in two strains of Trichogramma nr ivelae (Pang and Chen), both showed increased acceptance of host when there had been a prior ovipositional experience with that host. Nurindah et al. (1999) also found that prior ovipositional host experience of T. australicum (Girault) increased its efficiency of handling and acceptability of subsequent host eggs. Borkstgen and Hoffmann (1998) found that even in low ranked hosts, there wa s an increased acceptance of the host with prior ovipositional experience, even with a subsequent ovipositional experience with a highly acceptable host. In contrast to the findings above, Th omson and Stinner (1990) found that Trichogramma sp. near pretiosum , T. exiguum and T. minutum did not change host specificity to Ostrinia nubilalis , Heliothis zea and Manduca sexta when previously exposed to the scales of th ese host species. This would seem to make sense. An ovipositional experience that elicits an improved accepta nce and handling response would help the parasitoid handle eggs from th at species more efficiently in the future. Host scales are less deterministic in that the presence of the scales of one species does not preclude the presence of other host species. A parasitoid should be able to keep its options open as to which hosts it may find. Becoming more efficient in the handling of one hostÂ’s eggs, on the other hand, does not pr eclude a parasitoid from finding the eggs of anotherÂ’s. Effects of Plant Structure on Egg Oviposition by Lepidopteran Hosts There are numerous factors that determine ovipositional patterns of female phytophagous female insects on host plants. There are the abiotic factors such as

PAGE 50

32 temperature, light and humidity as well as th e biotic factors, such as leaf toughness and predator/prey avoidance. These vary within the various microhabitats of the plants. The physical and chemical characteristics of a pl ant are important in determining whether a female moth will oviposit on the plant or not. These characteristics are also influencial in determining which part of the plant a female insect will oviposit onto. It is known that volatile components from crucifers appear attractive to both DBM males and females (Palaniswamy et al. 1986). Gl ucosinolates, and in particular sinigrin, play a key role as cues for DBM ovipos ition (Renwick and Radke 1990, Justus and Mitchell 1996). Spangler and Calvin (2001) confirm that the chemical composition of plant tissue of certain regions may influence female ovipos ition. They also propose that the oviposition sites chosen by the female adult may be rela ted to subsequent vertical distribution of larval feeding sites. Udayagir i and Mason (1995) assert that the female European corn borer may also be using chem ical stimuli from regions of the plant to determine the best feeding site s for their offspring. Ulmer et al. (2002) discovered that Mamestra configurata (Walker) ovipositional preferences in cruciferous plants were related to amount of leaf material—the fema les preferred sections with the most leaf material. This would suggest that the decisi on was being made with larval feeding in mind. However, they also found that when the pl ants were full flowering, the females laid more eggs near the top of the plant, even t hough there was less leaf material there. Adult feeding and younger leaf tissue were thought to be the determini ng factor in this instance. The ovipositional preferences of in sects also depend on a number of interpopulational and interspe cific interactions of the insects present (Waldvogel and Gould 1990, Thompson and Pellmyr 1991). Pansera de Araújo et al. ( 1999) investigated

PAGE 51

33 the ovipositional patterns of four species of Noctuidae laying eggs on soybean plants in Brasil. Anticarsia gemmatalis (Hübner) and Spodoptera latifascia (Walker) eggs were found mainly on the medium height plant region s, although on leaflets of different sizes, while Rachiplusia nu (Guenée) and Pseudoplusia includens (Walker) eggs were found mainly in the upper and lower plant regions, respectively. These spec ies maintained their niches within the plant—even the two species th at laid eggs in sim ilar vertical positions on the plants. Improving the Welfare of Sm all-Scale Farmers—the Role of Agricultural Research In the 19th century, scie ntific progress in chem istry, biology, microbiology, and mechanics greatly enhanced the developm ent of agricultural practices. Informal experimentation had always been a part of agricultural development, but from the 18th century, there began concerted efforts to intensify production, spurred by great population increases in Europe, by the enclosur e of common lands and by the scientific awakenings of the Renaissance and the Age of Enlightenment in Europe (Kishlansky et al. 2001). The application of these agri cultural developments was tied to the education of the populace and to the expansion of educational institutes to disseminate the newfound knowledge. These same institutes today str uggle to balance the high productivity of modern agriculture with co mpetition for resources and the negative environmental and social impacts that intensive production tec hniques can bring. It was after the World Wars, with the threat of widespread famine in large parts of the developing world, that there was the move in developed countri es to try use modern technology for the betterment of the underrepresented and underf ed. The beginning of the Green Revolution is set at 1944, with the establ ishment of a plant-breeding st ation in northwestern Mexico

PAGE 52

34 where the goal was to boost grain yields. The team headed by Norman Borlaug, a plant breeder from the University of Minnesota, spectacularly achieved th ese goals. Since then, the blueprint of new varieties and altered agri cultural practices has been shared all over the world. Amidst huge yield increases, thes e transfers of technology were soon found to be lacking, (mostly resource-rich farmers adopt ed the technologies), as there were few ‘spill over’ effects to the poor, and th e practices brought socio-economic and environmental problems (Gibbon et al . 1995). Significant scientific achievements have not masked the fact that technologies relevant to the majority of small-scale farm ers have been in short supply. Part of the reason for this is that farmers have not been well integrated into the problem definition and technology generation process of agricult ural policy-making and research (Kassa and Gibbon 2003). Efforts to change this have led to the emergence of participatory research approaches that try to put people at the centr e of the research pro cess (Röling and Jiggins 1998). Two methodologies that have emerged from the reorganization of research paradigms are Farming Systems Research ( FSR) and Farmer Participatory Research (FPR). The former attempts to understand th e ‘whole farm’ and farmer’s conditions and priorities as the basis for planning research and extensi on activities (Collinson 2000). The latter methodology does the same but places much emphasis on the collaboration with farmers, originating from “F armer First” (Chambers et al . 1989) and Participatory Technology Development (Jiggins and De Zeeuw 1992) concepts that were first introduced in the late 1980s. FSR is consid ered less encompassing than FPR (Farrington 2000) but as methodologies evolve the diffe rences between the two are becoming less distinct.

PAGE 53

35 Much of the reason for these changes in research and extensi on attitudes derives from the heterogeneity of responses to ne w technologies that is attributed to characteristics of the farmer or the farm (Fed er et al. 1985). The i nherent capabilities of new technologies such as new seeds, fertilizer s or machinery to improve yields and labor productivity are unquestioned, and have been the reason for the unprecedented output growth experienced in the 20th century (Evenson and Gollin 2002). The heterogeneity of response, however, comes from those tec hnologies that are not embodied by physical inputs such as seed or fert ilizer, but instead by changi ng cultivation techniques and natural resource practices (Barrett and Moser 2003 unpubl .)—the knowledge-intensive technologies such as Integr ated Pest Management (IPM) , which rely on individuals. FSR and FPR methodologies have been criticized for only focusing on the agricultural system and not on the wider non-agricultural activities. These worries have led recently to the conceptual ization of a much wider appr oach called the Sustainable Livelihood Systems Approach or Livelihoods A pproach (LA). In this case, ‘livelihood’ refers to the capabilities and assets of peopl e and the activities they are engaged in to meet their basic needs (Ashley and Carney 1999, DFID 2000). It can also include other factors such as entitlement, security and quality of life (Sen 1999). Livelihoods are considered sustainable when they “are econom ically viable, can withstand stresses and shocks and can build up assets and capabi lities, while not undermining the natural resource base both now and in th e future” (Barrett and Moser 2003 unpubl .). The initial problems in using LA to study farming comm unities have been the large quantities of information generated and the difficulty in processing the data in an efficient and constructive manner.

PAGE 54

36 Farming Systems Research A farming system is defined by the FAO as “a population of individual farm systems that have broadly similar resour ce bases, enterprise patterns, household livelihoods and constraints, and for whic h similar development strategies and interventions would be appropriate”(F.A.O. 2003). Keating and McCown (2001) identify two key components of farming systems, name ly the biophysical ‘Pr oduction System’ of crops, animals, soil and climate together w ith certain physical input s and outputs and the ‘Management System’, made up of peopl e, values, goals, kn owledge, resources, monitoring opportunities and decision making. Th eir review of six types of farming systems analysis concludes that the challe nges and opportunities lie at the interface between the ‘hard’, scientific approaches to the analysis of the biophysical system and ‘soft’ approaches to intervention in social ma nagement systems. They also conclude that the use of models in farmer decision s upport systems has been disappointing and a way has to be found of making models releva nt to real world decision-making and management practices. This may not necessa rily be achieved by only making the models more accurate or more comprehensive. Part of the scientific process is the dec onstruction of problems so that individual elements can be identified, appraised, expe rimented on and understood. They are either left as they are, or altered in the hope of improvement. A weakne ss in the scientific process is that the deconstr uction removes the elements fr om their natu ral position, and contextually, this can lead to misinterpreta tions and oversights. The place of an element in a particular system is as important as the intrinsic characteristics of the element itself. Importantly, the producers themselves see thei r environment as a system (Dixon et al. 2001) and evaluate new technologies by the way in which they interact with the other

PAGE 55

37 elements of their environment. Success is ba sed on a perceived over-all betterment of the system. How can the totality of a system be unde rstood? Farmer Participatory Research (FPR) is a good way of studying communities a nd understanding their characteristics. It relies on human interactions and observations and is built on synergis tic collaborations of multi-disciplinary technical teams, community members and other stakeholders such as extension agents (Dlott et al. 1994, Biggs 1989, Chambers 1994, Cornwall and Jewkes 1995). Participatory research methods have beco me an integral part of Farming Systems Research (FSR), and they serve as an impor tant means of dialogue between participants and stakeholders. FSR is principally about technology generation, and its processes can be divided into four stages: descriptiv e (diagnostic), design, testing and extension (Norman 1980). Stakeholder feedback is cruc ial to all stages. The first stage is about understanding the livelihood system and generating research ob jectives based on identified problems or possibilities. The second stage determines how best the rese arch objectives can be met by planning an effective and efficient set of re search activities. The third stage is the execution of these activities. This stage is given validity by its inclusiveness, its relevance and its interactivity. The final stage disseminates results and implements new technologies. Linear Programming This is a form of modeling using an op timization matrix program that, for the purposes of FSR, looks to examine the interface of ‘hard’ scientific ap proaches and ‘soft’ approaches in social management. It does so by simulating and analyzing family farm livelihood systems by determining the optim al combination of farm and non-farm

PAGE 56

38 activities that is feas ible, given a set of fixed constrai nts (Cabrera 1999). LP models are not as exact in their simulation of production functions as some crop models, and are not as sophisticated as some economic models, but they do represent a robust and fairly simple means of characterizing farming system s. They also provide a way of numerating assessments of how alternativ e activities achieve household objectives. To construct a linear programming (LP) model, certain info rmation is needed. Hildebrand and Araújo (1997) state that the following is needed The farm and non-farm activities and op tions with their respective resource requirements and any constraints on their production The fixed resources and other maximum or minimum constraints that limit farm and family production Cash costs and returns of each activity A defined objective or objectives. With this information, LP models can be made to simulate the characteristics and activities found on farms that go beyond merely identifying the most productive strategies. The use of LP models in farm pl anning has its origins in the late 1950s, when whole farm planning was been developed. In 1958, Heady and Candler outlined the application of LP modeling to farm planning, and by 1963, its relevance to low-income agriculture had been demonstrated (Clayton 1963 ). Since then, it has been widely used to examine supply changes and policy shifts in agriculture (Hazell and Norton 1986). Its impact on improving livelihoods in developing countries, however, has never been great, in part due to the laborious data collecting pr ocess and to its lack of direct applicability (Collinson 2000). How then does LP modeling fit into FSR? Its primary use is in the first stage (description/diagnostic phase) and second stag e (research-planning) of FSR, but can also

PAGE 57

39 be used as an extension tool in stage 4. Norton et al . (1999) describe an approach to participatory IPM research that is bei ng implemented by the USAID-supported IPM Collaborative Research Support Program (IPM CS RP). One of the early activities in the IPM CSRP approach is the part icipatory appraisal stage that can take from between one to two weeks. It is at this stage; with disp arate sets of community data or on-farm data collected, that linear progra m (LP) modeling could be most useful. All the information gathered by the various participants could be distilled into a ma trix representing the enterprise activities (resou rce requirements and production functions), the farmÂ’s constraints and resources (la nd, labor, capital, costs on both spatial and temporal levels) and household objectives. Once a model has been validated, and it accurately reflects the farming systems under question, the designing of experiments (stage 2 of the FSR approach) can proceed. Alterna tively, validated models can be used to assess already existing technologies to see if they would be worth implementing into the farming system under study. Used properly, LP modeling can be a very useful tool to FSR practitioners.

PAGE 58

40 CHAPTER 3 HOST PREFERENCES, PARASITISM LEVELS AND EFFECT OF PLANT DEVELOPMENT ON FIELD PARASITISM OF DIAMONDBACK MOTH AND SOYBEAN LOOPER BY Trichogramma pretiosum Introduction Many studies have shown the potential of Trichogramma in the biological control of diamondback moth (DBM) (Oatman et al . 1968, Allen and Gonzalez 1974, Wuhrer and Hassan 1993, Godin and Boivin 2000). Most of these studies have been in the laboratory, and there is a feeling am ongst researchers who work with Trichogramma that more emphasis should be put on field research (Hassan pers. comm .). There is now decadesÂ’ worth of laboratory research to support field research. There is also the realization that not enough attention has been paid to the multi-trophic interactions that occur naturally in the field, which can ha ve a large impact on the efficacy of the parasitoids released in bi ological control programs. In this study, the initial objective was to find out whether Trichogramma parasitized DBM and SL to any significant de gree in a field situation in Puerto Rico. There have been a number of studies of the na tural enemies of these two pest insects in Puerto Rico but no mention of Trichogramma (Rosario-Perez 1984, Abreu-Rodriguez and Cruz 1997). Armstrong ( pers. comm .) who has studied the control of DBM for many decades in Puerto Rico, has never found Trichogramma parasitizing DBM eggs. The south coast of the island where the majority of the islandÂ’s cabbage is grown, is a hot, dry and scrubby environment. What potentia lly makes it worse for the use of Trichogramma are the strong winds that blow off the sea. Nevertheless, Trichogramma has been used on

PAGE 59

41 the south coast before (Walco tt 1948), and as recently as the late 1990Â’s, in experiments on a local tomato farm. For the use in toma to, windbreaks using stands of sugar cane were used to lessen the impact of high winds (Pantoja, unpubl .). One of the main reasons for this present study were the observations made in Florida th at late-season DBM population crashes might be attributable to Trichogramma activities; when the cabbage has been harvested and the use of pesticides has stopped (Leibee pers. comm .). If true, Trichogramma may have the potential to be a com ponent of an IPM strategy to control DBM in cabbage. Although DBM is by far the biggest thre at to cabbage production globally (Talekar and Shelton 1993), there are other inse ct pest species such as the soybean looper that can cause great damage (Agricultural E xperimental Station, University of Puerto Rico, 1999). How would the presence of the eggs of other insect pests affect parasitism of DBM by the generalist Trichogramma ? Would the eggs of other species be intrinsically better hosts than the smaller DBM eggs? Would the egg-laying distribution of other hosts mean that their eggs were found by Trichogramma more easily than DBM eggs? Would Trichogramma Â’s encounters with other eggs have a ny subsequent affect on the perceived suitability of DBM as a host? The worst-case scenario would be the exclusion of DBM as a host on the arrival of a more favorable host species. This present study was designed to examine the host preferences of T. pretiosum in cabbage where both DBM and SL eggs were present in as natu ral a setting as possible. Care has to be taken when using field percent parasitism data to determine the host preferences of a generalist parasitoid in a multi-host system. Observed percent parasitism is a combination of intrinsic suitability of a host egg and the ease and

PAGE 60

42 frequency in which it is encountered by the parasitoid in the field. In the laboratory, Trichogramma may prefer one speciesÂ’ eggs to anot her but because the preferred speciesÂ’ eggs are harder to find in the field, the less -suitable speciesÂ’ eggs may be parasitized more. Chapter 4 covers the laboratory expe riments that were conducted to determine intrinsic T. pretiosum preferences with DBM and SL eggs. Trichogramma parasitizing DBM or SL eggs in a cabbage field is a tritrophic system that includes the plant, phytopha gous lepidopteran hosts and hymenopteran parasitoid. All potentially have an influence on each other. The impact of plants on the parasitization process is well documented (Uetz 1991, Thorpe 1985, Geitzenauer and Bernays 1996, Maini and Burgio 1990, Smythe et al. 2003). This impact can be on the egg-laying behavior of the host species or on th e host-location activities of the parasitoid. Cabbage in Florida needs between 85 a nd 110 days from seed to maturity and between 70 and 90 days from transplanti ng to maturity, depending on the variety (Maynard et al. 1998). Cole and lettuce plants have three distinct gr owth phases: seedling development, a rosette period and heading (U niversity of California 1985). Andaloro et al. (1983) identified nine stag es of cabbage development (Table 3-1). As the plant develops, it creates a changi ng physical and chemical envi ronment for the phytophagous host insects seeking to exploit th e resources of the plant, and the parasitoids that seek to fulfill their reproductive needs.

PAGE 61

43 Table 3-1. Growth stages of cabbage. Growth stages Title Description 1 Cotyledons No true leaves present 2 Seedling Up to 5 true leaves 3 6-8 true leaves 4 9-12 true leaves Base of stem still visible from above 5 Precupping (Approximately 13–19 leaves) base of stem and of leaves are concealed from above. The innermost heat leaves grow in a more upright position 6 Cupping (Approximately 20–26 leaves) the upright innermost heart leaves are concealed by larger, older leaves 7 Early head formation (Approximately 2–4” diameter head) start of the formation of the ball-like structure of over-lapping leaves 8 Head fill (Approximately 3–8” diameter head) firm round head is visible within the wrapper leaves (the four outer leaves that touch the mature head) 9 Mature (Approximately 6–12” diameter head) no new visible leaf production will occur. Cabbage ready to harvest Source: (Andaloro et al. (1983). The objectives of these field experiments are: To determine whether T. pretiosum parasitizes DBM and SL eggs in the field. To determine whether T. pretiosum prefers one host to the other. To determine what the egg-laying distribu tion patterns of DBM and SL females are in cabbage plants at differe nt stages of development. To determine the parasitism patterns of T. pretiosum with respect to host egg location on the cabbage plants. Materials and Methods During the 2001 and 2002 field season, a to tal of eight field experiments were conducted. All followed the basic protocols de scribed below, alt hough there were some differences between individual experiments resulting from changes of necessity or

PAGE 62

44 improvement. The experiments were planned to be separate and independent entities, consequently, comparisons were made within individual experiments. Trichogramma pretiosum The T. pretiosum was obtained from Beneficial Insectaries, which sent Trichogramma egg cards on a weekly basis. The factitious host eggs used were Ephestia kuehniella (Zeller) (Lepidoptera: Pyralidae) eggs. On receipt of the Trichogramma , the egg cards were prepared immediately for the ne xt field experiment. To help synchronize the development of the wasps with the start of the experiment, th e vials containing the egg cards sometimes had to be pl aced in a refrigerator (6 ± 1oC, 20 ± 5% r.h.). The polystyrene box in which they were kept was lined with moist paper to maintain high humidity levels. Percent emergence of the Trichogramma was examined by punching a disk out of the egg square and counting the adults that emerged. The disks were 6 mm in diameter, and on average yielded 87 8 adults. The egg squares we re 18 mm by 40 mm and were supposed to yield 4,000 adults. C ounts indicated that emerge nce levels were about 56% of the potential maximum. It must be take n into consideration, however, that with a relatively small area of disk examined and with the mortalities of those eggs on the edge of the disk, the percent emergence ma y be higher. The sex ratio of the T. pretiosum supplied by Beneficial Insectaries was always greater than 90% females. To prepare the Trichogramma egg cards for the field experiments, 6 mm in diameter disks were punched out of the egg car ds and then cut into halves using a razor blade. Depending on the exact protocol of the experiment, a particular number of these halves were placed in empty diet cups, which were then closed and placed in the insect

PAGE 63

45 colony room (temperature 27 3oC, 65 5% r.h. and 16L:8D) to await the emergence of the adults. Normally, the adults started em erging during the scotophase of the third day after delivery. The diet cups were placed in to the field cages the same day that first emergence of the adults had been detected. In the first experiment I used five half disks, which yielded around 218 Trichogramma adults per field cage. This is a high er rate (195,000 per hectare) than that used by Krnjajic et al. (1997), who used ~75,000 Trichogramma evanescens in 1.5 ha. cabbage plots to decrease the damage of DB M by 60%. Some adjustments were made to the numbers of Trichogramma released in subsequent expe riments due to low percentage rates and changes in experimental design. Diamondback Moth (DBM) The diamondback moth were obtained from a culture maintained by Dr. Leibee at the Mid-Florida REC (University of Florida) in Apopka, Flor ida. Cocoons were shipped overnight. Groups of 30 pupae were placed into individual 25-dram, clear styrene tubes with tabbed caps (Bioquip catalogue # 8925). A total of ten tubes were prepared, corresponding to the experimental treat ments 5 DBM treatments and 5 DBM/SL treatments. An additional two styrene tubes were prepared as extras in case of accidents. Previously, it had been established that the sex ratio of the DBM sent from Dr Leibee's colony was 1:1 and that percent emergence of the adults was greater than 97% (C.I. = 1.1% = 0.05). With 30 pupae, we could thus rely on there being at least 10 of each sex. The DBM were released into the field cages 2 days after emergence. By this time they had mated and the females had begun laying eggs.

PAGE 64

46 Soybean Looper (SL) The soybean looper eggs used in the e xperiments came from a laboratory colony that had been established in the Fall of 2000 at the Univ ersity of Puerto Rico's agricultural research station in Río Piedras. They were reared using a combination of protocols (Jensen et al. 1974, Kogan and Cope 1974, Guy et al. 1985, Landolt 1997) and prior experience of researchers at the experime ntal station. The original insects had been collected as larvae and pupae from fields of various crops in Puerto Rico (soybean, tomato and eggplant being the principal hos t crops) and were later identified by Dr. Heppner (taxonomist, Department of Plant I ndustry, Florida Departme nt of Agriculture and Consumer Services) as s oybean looper. The insects were fed on the cabbage looper artificial diet (Bio-Serv cat alogue # F9282B). On emergence from the egg, groups of five first-instar soybean loop er larvae were placed into diet cups that had been quarter-filled with diet. Air-holes were punched into the caps of the diet cups and the trays of diet cups were placed in an insect rearing room (27 3oC, 75 5% r.h. and 16L:8D). After about 10 days, soybean looper larvae were transferre d, one to a cup, to fresh diet cups and placed back into the rearing room. The larvae were transferred to fresh diet a second time before pupation because of frass accumulation and some drying of the diet. The larvae were allowed to pupate in the diet cups, wh ich they did by spinni ng the cocoon on the top or sides of the cups. The adults emerged 6 to 7 days later and were placed in plastic storage boxes (40 cm 15 cm 10 cm) with other recently emerged adults. The number of boxes corresponded to the number of treatments and replicates in the experiments. Two diet cups (pill cups) were put in boxes, one with water and the other with 5% honey water.

PAGE 65

47 Dental cotton wicks were placed in the cups to soak up the excess liquid to prevent the adults from drowning. To augment humidity, a fo lded paper towel, sa turated with water, was also placed in the plastic box. All water used in the box was dist illed and autoclaved to prevent the development of bacterial, f ungal or viral diseases. The box was covered by light medical gauze and held in place with ¼ ”-wide band of sowing elastic. Once, early in the morning, and then again late afternoon, the gauze covering was sprayed with a fine mist of distilled, autoclaved water to main tain humidity. The water and honey-water diet cups and the paper towel were replaced ev ery day to minimize disease problems. Not all the soybean looper needed for th e experiments emerged on one day, and so those that emerged on any one day were caref ully distributed around the plastic boxes so that there were similar numbers of same-age d males and females in all the boxes. There was no more than 4 days’ difference between the oldest and youngest soybean looper adult in any one experiment. The soybean looper adults were always the limiting factor in the experiments, and the start of the e xperiments depended on when enough soybean looper adults had been accumulated. Experime nts were initiated once the soybean adults had begun to lay eggs, usually 4 days after emergence. Serratia marcescens infection in the soybean looper colony . Early in 2001, the laboratory colony of soybean looper in Rí o Piedras started showing unusually high mortality. The causal agent of this epizootic was eventually identified as the bacterium, Serratia marcescens. Serratia marcescens is an entomopathogenic bacterium and a member of the Enterobacteri aceae (facultative anaerobic, gr am-negative rods that are cytochrome oxidase negative and catalase posit ive). It usually produc es a characteristic

PAGE 66

48 red or pink pigment, although white to rose -red stains frequently occur (Inglis and Lawrence 2001). There were many symptoms associated w ith the disease, and these included cessation of feeding by the larvae, blooms of bacterial growth on the surface of the diet, bad odors and discoloration of the dying larvae . Some larvae managed to pupate but most turned black and failed to emerged. Those fe w that did emerge were deformed, did not produce eggs and died prematurely. No SL oviposition occurred in the second experiments of 2001 and of 2002 because of colony crashes due to the disease. In collaboration with Sr. Luis Silva-Negr ón (Associate Agricultural Scientist, Río Piedras Agricultural Experimental Station), meas ures were taken to reduce the spread of the disease and to try to erad icate the bacteria from the colony. One of these measures was to examine the diet being fed to the larv ae. The cabbage looper diet used to feed the soybean looper colony had the following ingredients: Sucrose 26 g/L Alfalfa herb powder 14 g/L Linseed oil 7 g/L Corn oil 5 g/L Vitamin mix 10 g/L Salt (mineral) mix 10 g/L Aureomycin 54 mg/L The preparation of the diet re quired that the agar was di ssolved in water, brought to the boil and then rapidly mixed together with the diet to preven t premature solidifying. This methodology meant that the antibiotic in th e diet was exposed to high temperatures, which dramatically reduced the effic acy of the antibiotic through denaturing. An experiment was run to see if introduc ing additional antibiotic, in such a way that the antibiotic was not exposed to high h eat, could slow or stop the epizootic killing

PAGE 67

49 the colony. Concentrations of 100 mg/L and 200 mg/L of streptomycin were added to freshly made diet that had been cooled to 50oF and stirred frequently to prevent solidification. There were 35 larvae placed on the 100 mg/L streptomycin diet and 36 larvae placed on the 200 mg/L streptomycin diet. Every day, the number of newly killed larvae was recorded. Of the 35 larvae placed on the 100 mg/L streptomycin diet, 17 had died by the tenth day and of the 36 larvae pl aced on the 200 mg/L streptomycin diet, 19 had died. The only effect the higher concentr ation had was to delay the deaths of the larvae by a few days. It was not felt that the addition of an tibiotic had made much of a difference. Aside from the inclusion of additional antib iotics, there were measures adopted to improve sanitation. Prior to the epizootic, the water used in the colony was only distilled. Afterwards, the distilled water was autoclav ed before being used in the colony. A UVlight cabinet was also procured and all cont ainers, instruments and other materials were surfaced sterilized. Dirty containers, trays and netting were washed in water containing a small amount of Clorox, rinsed with clean water and then placed to dry under the UVlight. Eventually the colony numbers stabilized and there was minimal evidence of S. marcescens . This was thought to be as a result of the sanitation measures applied, changes in insect diet and the development of some resistance in the colony against the bacteria. The latter cause was thought to be the most importa nt factor. Later additions of field soybean looper resulted in mortality epis odes that ended when most of the newly introduced larvae had died.

PAGE 68

50 Cabbage The variety Blue Vantage was used in all experiments, and the staff at the Fortuna agricultural research sub-sta tion maintained the plantings. The seed was bought from the Rupp Seed Company. The cabbage was sown at a spacing of 2 feet apart along the row and 3 feet between rows (Figure 3-1). Plants were irrigated by drip irrigation and sprayed once a week with Mattch (Ecogen Inc, La nghorne PA, USA), a Bt-p esticide blend of Cry1A(c) and Cry1C derived delta endotoxins of Bacillus thuringensis encapsulated in killed Pseudomonas fluorescens, until the experiments were ready to be conducted. This weekly insecticidal applicati on was not entirely successful at preventing damage from the ever-present native populations of DBM. Figure 3-1. Preparing field cages for the experiments

PAGE 69

51 General Experimental Protocol Field cages A few days before the experiment was conducted, the field cages were placed over the experimental plots of cabbage. The field cages were constructed of ½ ” Cresline PVC SCH 40 tubing, which could be dismantled and re sized when needed. The netting used in the field cages was Style # C 20A lumite ne tting from Synthetic I ndustries (Chattanooga, TN). The netting was held in place by plasti c bags filled with soil, and placed around the edges (See Figure 3-2 for an idea of the field cages). The cabbage plants were weeded before the field cages were placed over the experimental plots. Figure 3-2. Field cages used in Juana Diaz fieldwork As Figure 3-2 demonstrates, the experi ments were conducted using one row of cabbage; the other rows were later used in following experiments. The conditions under each of the cages were the same, and so a completely randomized design was adopted and the experimental plots were randomly di vided amongst the treatments and replicates. During all the experiments, there were nativ e populations of diamondback moth present. Soybean looper was also found but not nearly as frequently as the diamondback moth. To

PAGE 70

52 examine the possibility of there being native Trichogramma present at the research station, some of the cages were designated control cages. Insect releases When the SL adult females had begun to lay eggs, both SL and DBM adults were taken from the insect rearing room in the ag ricultural research stat ion in Río Piedras, driven to the field site at the agricultural research sub-station in Juana Diaz on the south of the island and released. Releases were ma de in the late afternoon and the cages were divided into three treatments and the contro l. The four treatments were as follows DBM only SL only DBM + SL together Control (DBM + SL together). The following day, the Trichogramma were taken to the field site and released under the field cages, 24 hours after the host sp ecies release. To try and mimic a more authentic large-scale situa tion, a point source release of a known amount of already emerged Trichogramma adults was rejected. Instead, punched circles of the egg cards were taken to the field when the first signs of adult emergence. It was assumed that similar amounts of adults would emerge as predicted by the preliminary percent emergence experiments. The Trichogramma were released from diet cups placed on the ground. Because of high wind conditions that exis t in this part of the southern coastal plain of Puerto Rico, it was thought best to release the wa sps from ground level, rather than from an elevated position. Appendi x A shows wind speeds, wind direction and maximum gust speeds for a five hour pe riod, late afternoon on the day of the Trichogramma release for five of the six experi ments conducted in 2002. There were no

PAGE 71

53 Trichogramma released in the control cages. Afte r a further 24 hours, the cages were opened and sample cabbage plants re moved for analysis (Figure 3-3). Figure 3-3. Sampling of the fiel d plots and removal of cabbage The plants were placed in plastic crates and taken back to Río Piedras where they were placed in a large, in dustrial refrigerator (4 ± 1oC, 80 ± 5% r.h.). Each crate held the cabbage plants of one replicate. Refrigerati on was felt necessary because of the amount of time necessary to process the samples. It was also used to prevent DBM egg hatch, which potentially could have happened the day after sampling of the cabbage. Egg counts and parasitism determination Individual cabbage plants we re taken apart leaf by leaf, starting from the bottom. Using a magnifying glass and hand lens, each leaf was carefully examined for eggs. Eggs were then isolated by cutting a leaf disk containing the eggs with a leaf auger and placing the disk into an egg tray (Figure 3-4). The egg tray was constructed as detailed in “The Trichogramma Manual” (Knutson 1998). All positiona l information relating to the eggs

PAGE 72

54 on the plants was carefully recorded, as was the location of individual eggs in the egg tray. Figure 3-4. Equipment used to find, remove and retain eggs for parasitism analysis The egg trays were placed in the insect colony room (temperature 27 3oC, 65 5% r.h. and 16L:8D) until signs of parasi tism could be discerned, which normally occurred 4–5 days later. The non-parasitized eggs normally hatched before any signs of parasitism were noticeable in the parasiti zed eggs. The first sign of parasitism was a graying of the egg, which was then followed by the development of silvery-black patches that eventually coalesced over the entire surface of the egg. Individual Experiment Characteristics As mentioned at the beginning of this chapter, the indivi dual experiments varied in the specifics of their design. These differe nces are tabulated below (Table 3-2). Differences were due to the health of the ins ect colonies, initial problems with the field cage netting and with the varying numbers of eggs laid and parasitism levels found.

PAGE 73

55 Table 3-2. Details of the individua l field experiments at Juana Diaz Expt. # No. of replicates No. of plants under cage No. of plants sampled # DBM released /cage # SL released /cage # T. pretiosum released /cage No. of leaves /plant 2001 1 5 20 3 30 14 ~220 10-14 2 5 20 3 30 5-11 ~200 22-25 2002 1 4 10 3-4 30 15 ~200 8-12 2 4 5 3 30 10 ~110 24-28 3 4 5 3 30 11 ~360 15-18 4 4 5 2 30 11 ~480 27-30 5 4 5 3-5 30 11 ~520 8-10 6 4 5 3 30 12 ~480 17-21 Statistical Analysis The data were analyzed using SAS softwa re version 8.01 (SAS Institute 1999). For the egg distribution amongst quartiles data, indivi dual pair-wise comparis ons were performed between quartiles, using Proc Genmod and the LSMEANS statement (the count data required the use of Poisson or negative bi nomial distributions). For comparisons of various egg position data between the two hos t species, a two-sample t test was used when the data were normally distributed. If the data were not normally distributed, transformations were applied. Normality was ch ecked using the Shapiro-Wilk test and the QQPLOT function in Proc Capability. The transformations used to achieve normality were Arcsine or Arcsine-root, transformations commonly used for percentage data. If the transformations were unsucce ssful, the Wilcoxon test was used. The Wilcoxon test in Proc NPAR1WAY was also used for compari ng the distribution of parasitized eggs and nonparasitized eggs for both host species. All di fferences were considered significant at the P<0.05 level.

PAGE 74

56 Results The two experiments of 2001 produced little in the way of results. In the first experiment there were low DBM egg numbers whilst in the second the SL adult females, already infected by S. marcescens , laid no eggs whatsoever. The second experiment was eventually abandoned because of flooding. A tota l of 3.35 inches of ra in fell the day after Trichogramma release, more than had fallen in the previous two months combined. Egg Numbers and Quartile Positions For the purpose of simplifying the results , the experimental plants were divided into quartiles, and egg numbers and parasiti sm levels are presented based on quartile position. Leaves were counted off, from the bo ttom, and placed into quartiles (i.e. each quartile contained a quarter of the total numbe r of leaves on the plant), with the first quartile of leaves being t hose closest to the ground. One of the main objectives of this study wa s to try and get an accurate description of the oviposition patterns of the two host species. For each egg found on the plant, the following information was recorded: leaf num ber (counted from the bottom of the plant up), adaxial/abaxial position, position on or off a vein (special note being taken of the main vein) and distance from the leaf edge (divided between those eggs found within 1.25 cm from the edge and thos e beyond that measurement). The one experiment of 2001 and the six experiments run during 2002 were conducted with plants of varying age and si ze. Experiment 1 of 2001, Experiment 1 of 2002 and Experiment 5 all had plants with between 8 and 12 leaves. Experiment 3 had plants with between 15 and 18 leaves and the pl ants in Experiment 6 were slightly larger with 17 to 21 leaves. The plants of Expe riment 2 had between 24 and 28 leaves while those of Experiment 4 were the largest with between 27 and 30 leaves. Using the cabbage

PAGE 75

57 growth stages as defined by Andaloro et al . (1983), Experiments 1 (2001 and 2002) and 5 fall into Stage 4 (“9–12 true leaves base of stem still visible from above”), Experiments 3 and 6 fall into Stage 5 (“precupping”) wh ilst Experiments 2 and 4 fall into Stage 7 (“early head formation”). DBM eggs For all of the completed 7 experiments th ere was a distinct pattern found for the DBM eggs laid on the plants. For most of the experiments the highest number of eggs were laid in the middle two quartiles of th e plants. Figure 3-5 shows the average number of single-host DBM eggs per quartile per plant for each of the 7 experiments.

PAGE 76

58 Figure 3-5. Inter-quartile distri bution of DBM eggs in the ca bbage plants of Experiment 1 (2001) and the six experiments of 2002. Different letters above the bars signify significant differences where P > 0.05. DBM was found to laid eggs on the stems of the plants. The eggs, however, were only laid on cabbage plant stems when the plants were relatively small (around 8–12 leaves). DBM eggs were found on stems in Experiments 1 (2002), 3 and 5. DBM eggs were also found in greater abundance on the adaxial side of the leaves than the abaxial side as shown in Figure 3-6. A verages for DBM (Expt.1) 0 1 2 3 4 5 6 7 8 stem1234 QuartilesAverage egg number DBM averages a b b a a a Averages for DBM (Expt. 1 2002) Averages for DBM (Expt.3) 0 1 2 3 4 5 6 7 8 9 10 stem1234 QuartilesAverage egg number DBM averages b b b a a A verages for DBM (Expt. 4) 0 1 2 3 4 5 6 7 8 stem1234 QuartilesAverage egg number DBM averages d c c b b a Averages for DBM (Expt.5) 0 0.5 1 1.5 2 2.5 stem1234 QuartilesAverage egg number DBM averages c c b b c a a Averages for DBM (Expt.2) 0 0.5 1 1.5 2 2.5 3 stem1234 QuartilesAverage egg number DBM averages d c c b b a Averages for DBM (Expt.6) 0 0.5 1 1.5 2 2.5 3 3.5 4 stem1234 QuartilesAverage egg number DBM averages d c c b a Averages for DBM (Expt.1 2001) 0 0.5 1 1.5 2 2.5 3 stem1234 QuartilesAverage egg number DBM averages n/aa a a a

PAGE 77

59 Figure 3-6. Number of DBM eggs, by quartile , found on the adaxial a nd abaxial sides of the leaves for the DBM single-host treatme nts. Above each bar is the ratio of number of eggs on the abaxial side to number on the adaxial side. The ratio given in the box, top right of each chart, is the ratio for the plant as a whole. Ratios in red signify greater numbers of eggs on the abaxial side and an asterisks shows significance differen ce of DBM egg numbers between the adaxial and abaxial side of the leaf where P > 0.05 Only in the fourth quartile were there sometimes more DBM eggs on the abaxial side than on the adaxial side. This reflects a different orientation of the newest leaves, which tend to be more vertical. In the more mature plants, these le aves fold upwards and over the apical bud. DBM female adults thus were consistent in prefer ring to lay eggs on the surfaces that faced upwards. Expt 1 (2002) DBM eggs 0 20 40 60 80 100 120 140 1234Quartile egg # Abaxial Adaxial *1 : 3.7 1 : 2.3 1 : 2.5 1 : 7.0 1 : 3.4 * Expt 2DBM eggs0 20 40 60 80 100 120 140 160 1234Quartile egg # Abaxial Adaxial *1.6 : 1 1 : 1.3 1 : 2.0 1 : 1.7 1 : 1.8 * Expt 3 DBM eggs0 50 100 150 200 250 300 350 1234 Quartile egg # Abaxial Adaxial 1 : 1.7 **1 : 4.3 *2.8 : 1 1 : 2.3 1 : 1.2 Expt 4 DBM eggs0 50 100 150 200 250 300 1234 Quartile egg # Abaxial Adaxial 1 : 5.9 1 : 7.8 1 : 5.9 1 : 4.2 * 1 : 1.4 * * * Expt 5 DBM eggs0 10 20 30 40 50 60 70 80 90 100 1234 Quartile egg # Abaxial Adaxial 1 : 1.5 1 : 3.2 1 : 2.4 *1.9 : 1 Expt 6 DBM eggs0 50 100 150 200 250 1234 Quartile egg # Abaxial Adaxial 1 : 5.8 * 1 : 1.3 1 : 8.1 1 : 6.9 * * Expt 1 (2001) DBM eggs 0 2 4 6 8 10 12 14 1234 Quartile egg # Abaxial Adaxial 1 : 1 1 : 4.5 1 : 2.6

PAGE 78

60 Female DBM adults showed no preference for any particular part of the leaf. Eggs were found on all morphologically distinct parts of the leaves that included the petiole, main vein and other smaller veins. Figure 37 shows the breakdown of how the eggs were distributed on the leaves for each of the six experiments of 2002 (leaf morphology data was not collected for Experiment 1 of 2001). Th e bar chart to the right of each pie chart shows the division of the DBM eggs between those found within 1.25 cm of the leaf edge and those found further away from the edge. A bar is presented for each the adaxial and abaxial sides of the leaf. Eggs were most likely to be found on the petioles when the plants were small, which, as for those eggs la id on the stems, must have been because of the more open canopy found in the smaller plants . The greatest proportion of eggs laid on the main vein also occurred with the sma ller plants (Experiments 1 and 5). Aside from this, the majority of the DBM eggs was eith er found on the secondary veins or in areas where there were no veins. As for DBM eggs being found close to the leaf edge, there was generally the same number of eggs found in the 1.25 cm band close to the leaf edge, as there was found away from the edge.

PAGE 79

61 Figure 3-7. Division of DBM eggs amongst the different morphological regions of a cabbage leaf. The bar charts on the right of the pie charts show the proportion of eggs found within half an inch of th e leaf edge or not for the adaxial (Ad.) and abaxial (Ab.) side of the leaf. SL eggs The inter-quartile distribution of looper e ggs was similar to that of the DBM eggs in that most of the eggs were found in th e middle two quartiles of the plants. Figure 3-8 DBM eggs (Expt.1) Petiole Main vein Vein No vein 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%Ad.Ab. > 0.5" < 0.5" DBM eggs (Expt.2) Petiole Main vein Vein No vein 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 0. 7 0. 8 0. 9 1Ad.Ab. > 0.5" < 0.5" DBM eggs (Expt.3) Petiole Main vein Vein No vein 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 0. 7 0. 8 0. 9 1Ad.Ab. > 0.5" < 0.5" DBM eggs (Expt.4) Petiole Main vein Vein No vein 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 0. 7 0. 8 0. 9 1Ad.Ab. > 0.5" < 0.5" DBM eggs (Expt.5) Petiole Main vein Vein No vein 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 0. 7 0. 8 0. 9 1Ad. Ab. > 0.5" < 0.5" DBM eggs (Expt.6) Petiole Main vein Vein No vein 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 0. 7 0. 8 0. 9 1Ad.Ab. > 0.5" < 0.5"

PAGE 80

62 shows the average number of soybean looper eg gs per quartile per plant for Experiment 1 (2001) and for five of the six experiments of 2002 (there were no SL eggs laid in Experiment 2 of 2002). No soybean looper eggs were found on the stems of plants in any experiment. Figure 3-8. Inter-quartile di stribution of soybean looper eggs in the single-host SL treatments. Different letters above th e bars signify significant differences where P > 0.05 Unlike the DBM females, the SL females showed a strong preference for laying their eggs on the abaxial side of the leaves . This is shown in Figure 3-9, where ratios of abaxial eggs to adaxial eggs ranged from 3.5:1 to 31.9:1. In all experiments, the percentage of SL eggs found on the abaxial side was significantly greater than the percentage of SL eggs found on the adaxial si de. Only once, in the fourth quarter in Experiment 6 was the percent number of eggs higher on the adaxial side. A verages for Looper (Expt.1 2002) 0 0.5 1 1.5 2 2.5 3 3.5 4 stem1234 QuartilesAverage egg number Looper averages a b b c n/a Averages for Looper (Expt.3) 0 0.5 1 1.5 2 2.5 3 3.5 4 stem1234 QuartilesAverage egg number Looper averages n/a c b b a Averages for Looper (Expt.4) 0 1 2 3 4 5 6 7 stem1234 QuartilesAverage egg number Looper averages n/a b b a a A verages for Looper (Expt.5) 0 1 2 3 4 5 6 stem1234 QuartilesAverage egg number Looper averages n/a c b b a Averages for Looper (Expt.6) 0 0.05 0.1 0.15 0.2 0.25 0.3 stem1234 QuartilesAverage egg number Looper averages b b a a n/a Averages for Looper (Expt.1 2001) 0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 20stem1234 QuartilesAverage egg number Looper averages b a n/a a a

PAGE 81

63 Figure 3-10 shows the distribution of the s oybean looper eggs on the leaves of the plants for five of the six experiments of 2002. In only two of the experiments were there eggs found on the petioles. Like DBM, most of the soybean looper eggs were found away from the main vein or petiole, either on the secondary veins or on the non-veined areas of the leaf. Figure 3-9. Number of SL eggs, by quartile, fo und on the adaxial and abaxial sides of the leaves. Above each bar is the ratio of nu mber of eggs on the abaxial side to number on the adaxial side. The ratio gi ven in the box, top right of each chart, is the ratio for the plant as a whole. Ratios in red signify greater numbers of eggs on the abaxial side and an aste risks shows significance difference of DBM egg numbers between the adaxial a nd abaxial side of the leaf where P > 0.05 Expt 3Looper eggs adaxial/abaxial position of eggs 0 1 0 20 30 40 50 60 70 80 90 1 00 1234Quartile egg # Abaxial Adaxial 4.0 : 1 **12.0 : 1 3.6 : 1*7.0 : 1*1 : 5.0 Expt 1 (2002)Looper eggs adaxial/abaxial position of eggs0 20 40 60 80 1 00 1 20 1 40 1 60 1 80 200 1234Quartile egg # Abaxial Adaxial 20.2 : 1 **8.3 : 1*22.2 : 1N/A N/A Expt 4Looper eggs adaxial/abaxial position of eggs 0 20 40 60 80 1 00 1 20 1 40 1 60 1 80 1234Quartile egg # Abaxial Adaxial 2.7 : 1 3.5 : 1 * 1.4 : 1* * * *6.5 : 1 2.9 : 1 Expt 5Looper eggs adaxial/abaxial position of eggs 0 20 40 60 80 1 00 1 20 1 40 1 60 1 80 1234Quartile egg # Abaxial Adaxial 14.1 : 1 * 1 : 4 17.0 : 1 77.0 : 1*20.8 : 1* * Expt 6Looper eggs adaxial/abaxial position of eggs 0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 20 1234Quartile egg # Abaxial Adaxial 5.5 : 1 *N/A N/A3.7 : 1 11.0 : 1* * Expt 1 (2001)Looper eggs adaxial/abaxial position of eggs 0 1 0 20 30 40 50 60 70 80 90 1 00 total12 Quartile egg # Abaxial Adaxial 31.9 : 1 * 12.0: 1 74.0: 1 13.2: 1* * *N/A

PAGE 82

64 Figure 3-10. Division of SL eggs amongst the different morphological regions of a cabbage leaf. Data is presented for five of the six 2002 experiments. The bar charts on the right of the pie ch arts show the proportion of eggs found within half an inch of the leaf edge or not for the adaxial and abaxial side of the leaf. SL eggs (Expt.6) Petiole Main vein Vein No vein 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%Ad.Ab. > 0.5" < 0.5" SL eggs (Expt.5) Petiole Main vein Vein No vein 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%Ad.Ab. > 0.5" < 0.5" SL eggs (Expt.4) Petiole Main vein Vein No vein 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%AdAb. > 0.5" < 0.5" SL eggs (Expt.3) Petiole Main vein Vein No vein 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%Ad.Ab. > 0.5" < 0.5" SL eggs (Expt.1) Petiole Main vein Vein No vein 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%Ad.Ab. > 0.5" < 0.5"

PAGE 83

65 The most striking characteristic shown in the bar charts of Figure 3-10 is the proportion of SL eggs that we re laid close to the edge of the leaf. As mentioned previously, most of the SL eggs were laid on the abaxial side of the leaves and of these, most were laid within 1.25 cm from the leaf’s edge. This was true even in the experiments with mature cabbage plants wher e the leaf edge was a smaller proportion of the total leaf area. Experiment 3 had the smallest percentage of eggs within the 1.25 cm band (77%), whilst Experiment 6 has the highe st percentage (95%). Table 3-3 shows the percent of abaxial SL eggs found within 1.25 cm of the leaf edge. Table 3-3. Percent of SL eggs near leaf edge and ratio of ‘central’ eggs to ‘edge’ eggs Experiment % of SL eggs near t value P value leaf edge (abaxial side) 1 92% 4.68 0.0002 3 77% 2.89 0.0202 4 91% 2.84 0.0162 5 91% 5.22 0.0003 6 95% 2.59 0.0251 The same analysis was done for DBM e ggs but there was no significant difference between the number of eggs found within the 1.25 cm band and those found outside of the band. This was true for both adaxial and abaxial sides. The only other difference between DBM and SL in leaf egg distribution was the percentage of eggs oviposited on the main vein. In all experiments the proportion of DBM eggs laid on the main vein was signifi cantly higher than the proportion of SL eggs laid on the main vein. This is shown in Table 3-4.

PAGE 84

66 Table 3-4. Percentage of the hosts’ eggs found on the main vein of the leaves. Note: Experiment 2 was not included as no SL eggs were found on the plants in that experiment. Comparing Egg-Laying Distribut ions of DBM and SL Eggs Although the general distributi on of the two hosts’ eggs was similar in that most eggs were laid in the middle section of the plants, there were differences found. An effective way of comparing distributions wa s to make a data set for each host species whereby each egg was identified by the number of the leaf on which it was deposited. So if there were eight eggs found on the leaf clos est to the ground, then the data set started with eight ‘1’s followed by a number of ‘2’s that equaled the number of eggs found on the second leaf up. This data set could then be used to determin e the corresponding leaf position for the 25th, 50th and 75th percentile of the egg batch on the plant. The average of these numbers gave a point reference for the ‘average’ leaf position for these egg distributions. The data set coul d also be used in an analys is to see whether the leaf positioning for the two host species’ eggs differed statistically. Figures 3.11–3.15 show the distributional differences between the tw o host species’ eggs within the cabbage plants of five of the six experiments of 2002 (Experiment 2 was excluded for lack of SL eggs). Experiment % of eggs on main vein Statistics DBM SL Z P value 1 32% 4% 18.395 < 0.0001 3 13% 3% 6.636 0.010 4 9% 4% 6.722 0.010 5 30% 6% 6.209 0.013 6 14% 8% 6.070 0.014

PAGE 85

67 Figure 3-11. Distribution of the two host spec iesÂ’ eggs by plant qua rtiles (chart) and by leaf number-derived percen tiles (table) for Experiment 1 (2001). Different letters following host species titles in the table denote significant differences in distributions. ( Z = 80.68, P < 0.0001) Figure 3-12. Distribution of the two host speciesÂ’ eggs by plant quartiles (chart) and by leaf number-derived percentiles (table ) for Experiment 1 (2002). Different letters following host species titles in the table denote significant differences in distributions. ( Z = 80.68, P < 0.0001) Percentiles DBM a SL a (leaf number) min 1 1 25th 3.75 5 50th 6 5 75th 8 7 max 12 12 ave 6.03 5.62 Percentiles DBM a SL b (leaf number) min 1 1 25th 2 4 50th 2 6 75th 4 7 max 8 9 ave 2.60 5.40 Expt.1 2001 (combined host) 0% 10% 20% 30% 40% 50% 60% 70% 1234Plant Quartilespercent of total # of eggs DBM averages Looper averages Expt 1 2002 (combined host)0% 10% 20% 30% 40% 50% 60% 1234Plant Quartilespercentage of egg total DBM averages Looper averages

PAGE 86

68 Figure 3-13. Distribution of the two host speciesÂ’ eggs by plant quartiles (chart) and by leaf number-derived percentiles (table) for Experiment 3. Different letters following host species titles in the ta ble denote significant differences in distributions. ( Z = 59.84, P < 0.0001) Figure 3-14. Distribution of the two host speciesÂ’ eggs by plant quartiles (chart) and by leaf number-derived percentiles (table) for Experiment 4. Different letters following host species titles in the ta ble denote significant differences in distributions. ( Z = 30.87, P < 0.0001) Percentiles DBM b SL a (leaf number) min 1 1 25th 7 6 50th 9 7 75th 11 9 max 16 14 ave 8.91 7.44 Percentiles DBM b SL a (leaf number) min 3 2 25th 13 11 50th 19 15 75th 22 19 max 29 29 ave 17.77 15.02 Expt 3 (combined host)0% 10% 20% 30% 40% 50% 1234Plant Quartilespercentage of egg total Looper averages DBM averages Expt 4 (combined host)0% 10% 20% 30% 40% 50% 1234Plant Quartilespercentage of egg total Looper averages DBM averages

PAGE 87

69 Figure 3-15. Distribution of the two host speciesÂ’ eggs by plant quartiles (chart) and by leaf number-derived percentiles (table) for Experiment 5. Different letters following host species titles in the ta ble denote significant differences in distributions. ( Z = 6.11, P = 0.0134) Figure 3-16. Distribution of the two host speciesÂ’ eggs by plant quartiles (chart) and by leaf number-derived quartiles (table) for Experiment 6. Different letters following host species titles in the ta ble denote significant differences in distributions. ( Z = 22.17, P < 0.0001). In five of the six experiments there were significant differences in the egg distributions. For Experiment 1 (2002), Expe riment 5 and Experiment 6, the DBM eggs were more likely to be found in positions cl oser to the ground. Two of these experiments (Experiment 1 (2002) and 5) had the smallest plants. Experiment 6 had plants at the prePercentiles DBM a SL b (leaf number) min 1 1 25th 3 5 50th 5 6 75th 8 8 max 9 10 ave 5.59 6.28 Percentiles DBM b SL a (leaf number) min 4 6 25th 7 10 50th 11 12 75th 13 14 max 20 17 ave 10.18 11.93 Expt 5 (combined host)0% 10% 20% 30% 40% 50% 1234Plant Quartilespercentage of egg total DBM averages Looper averages Expt 6 (combined host)0% 10% 20% 30% 40% 50% 60% 1234Plant Quartilespercentage of egg total DBM averages Looper averages

PAGE 88

70 cupping stage of development. For the other two experiments (Experiments 3 and 4) the egg-laying pattern was reversed, with SL eggs more likely to be found closer to the ground. Spread of Host Eggs among Quartiles Using the combined host field cage results, the evenness of distribution of the eggs amongst the quartiles of the plants was measured and compared for the experiments of 2002. Two indices were used in the comparis ons; the first was calculated by summing the number of eggs from the two adjacent quartiles with the most number of eggs and then dividing it by the total number of eggs (i .e., t2). The second was calculated by summing the number of eggs from the quartile with th e most number of eggs and dividing it by the total number of eggs (i.e. t1). Table 3-5 show s the percentages of spread of the two hosts’ eggs when using the above indices. A higher percentage indicates a less even vertical spread of host eggs within the plant. Table 3-5. Two measures (t1 & t2) of the spread of the hosts ’ eggs in the experimental plants and a comparison of th eir respective distributions. Experiment t2 P value t1 P value DBM SL DBM SL 1 88% 56% 0.0344 53% 26% 0.0242 3 79% 88% 0.6948 46% 56% 0.8870 4 75% 78% 0.6985 43% 37% 0.8973 5 62% 70% 0.4540 35% 52% 0.1594 6 86% 90% 0.1398 35% 56% 0.0292 It can be seen that there is little differe nce between the DBM eggs and the SL eggs in their spread pattern amongst the quartile divisions of the e xperimental plants. Only in Experiment 1 (2002) is one set of host’s e ggs more obviously ‘clumped’ than the other (t2: 2 = 4.473, P = 0.034 & t1: 2 = 5.078, P = 0.024). In this experiment, the DBM eggs were more heavily concentrated in the first two quartiles whereas the SL eggs were more

PAGE 89

71 evenly spread. The only other significant diffe rence is found in Experiment 6, where the t1 measurement shows more clumping in the SL eggs. Field Parasitism of the HostsÂ’ Eggs As mentioned in the introduc tion, there are a number of factors that influence the levels of parasitism found in the field. Asid e from basic suitability of the host eggs, the wasps have to deal with physical realities of the plant on which the host eggs are found. The wasps need to find an adequate number of eggs to parasitize or the plant may not be considered a suitable habitat for finding host eggs. The parasitism patterns found in this present study give some idea of how Trichogramma approaches the task of finding host eggs on a cabbage plant. Parasitism Levels Firstly, it must be said that no parasitis m was found in the control treatments and so no data from the controls will be covered in this results section. The parasitism levels found in the one experiment of 2001 and the si x of 2002 varied quite considerably. This was expected; there was variation among experi ments in the number of host eggs laid, differences in the size of cabbage plants in the field plots, and changes in the prevailing environmental conditions. Even within experime nts, there were sizeable differences in percent parasitism. This was largely because of differences in egg numbers per field cage. Table 3-6 shows the parasitism levels and egg numbers for the field cages with single-host DBM eggs. In Experiment 3, an average of 5% of the DBM eggs were parasitized in the field cages. In Experi ment 1 (2002) 48% of the DBM eggs were parasitized.

PAGE 90

72 Table 3-6. Percent parasitism of the DBM eggs found in the DBM single-host field cages Field cages Average ± SD 1 2 3 4 5 Expt 1 number of eggs 91 111 67 20 2 58.2 (2002) % parasitism 23 63 54 60 100 48% ±18.4% (per field cage) number of eggs 267 233 248 158 226.5 Expt 3 (2002) % parasitism 7 3 4 5 5% ±1.7% (per field cage) number of eggs 265 175 146 352 234.5 Expt 4 (2002) % parasitism 10 16 15 31 20% ±9.1% (per field cage) number of eggs 13 71 61 31 44 Expt 5 (2002) % parasitism 46 20 41 13 28% ±16.0% (per field cage) number of eggs 102 38 125 57 80.5 Expt 6 (2002) % parasitism 14 8 12 21 14% ±5.4% (per field cage) Expt 1 number of eggs 3 0 5 0 27 7 (2001) % parasitism 100 0 0 0 37 37% ±43.6% (per field cage) Table 3-7. Percent parasitism of the DBM e ggs found in the combined-host field cages Field cages Average ± SD 1 2 3 4 5 Expt 1 Number of eggs 66 157 25 60 49 71.4 (2002) % parasitism 45 74 68 37 77 60% ±18.0% (per field cage) Number of eggs 190 164 130 85 142.25 Expt 3 (2002) % parasitism 6 0 4 16 5% ±6.4% (per field cage) Number of eggs 179 206 365 194 236 Expt 4 (2002) % parasitism 16 15 10 6 11% ±4.6% (per field cage) Number of eggs 75 36 0 24 33.75 Expt 5 (2002) % parasitism 7 33 0 10 14% ±14.2% (per field cage) Number of eggs 163 130 63 117 118.25 Expt 6 (2002) % parasitism 9 12 19 30 16% ±9.3% (per field cage) Expt 1 Number of eggs 4 12 1 0 15 6.4 (2001) % parasitism 50 0 0 0 33 22% ±23.5% (per field cage)

PAGE 91

73 These parasitism levels were also mirrore d in the parasitism of the DBM eggs found in the combined host field cages (Table 3-7). Experiment 3 again had the lowest level of parasitism (5%) and Experime nt 1 (2002) had the highest (60%). For both single-host and combined-host field cage results, the results of Experiment 1 (2001) must be viewed with some caution because of the low number of DBM eggs present—an average of seven DBM eggs found on the three plants sampled in each field cage. Nevertheless, the results se em to indicate that Experiment 1 of 2001, Experiment 1 of 2002 and Experiment 5 of 2002 showed the greatest le vels of parasitism. These three experiments also contained the cabbage plants of smallest size. Table 3-8 shows the parasiti sm levels and egg numbers for the field cages with single-host SL eggs. Average percent parasitism varied from 0% in Experiment 6 to 27% in Experiment 4. The results from Experime nt 6 are questionable because of the low numbers of SL eggs found on the plants. St ill, a low percent pa rasitism was found for three of the other five experiments. Per cent parasitism of 5% or less was found in Experiments 1 (2002), 3 and 5. Aside from high levels of parasitism in Experiment 4, Experiment 1 (2001) was the only other expe riment to record significant levels of parasitism (24%).

PAGE 92

74 Table 3-8. Percent parasitism of the SL e ggs found in the SL single-host field cages Field cages Average ± SD 1 2 3 4 5 Expt 1 number of eggs 44 48 67 9 42 42 (2002) % parasitism 21 0 0 0 3 5% ±9.1% (per field cage) number of eggs 28 139 59 0 56.5 Expt 3 (2002) % parasitism 0 1 0 0 1% ±0.6% (per field cage) number of eggs 164 262 90 124 160 Expt 4 (2002) % parasitism 28 30 8 33 27% ±11.4% (per field cage) number of eggs 31 79 110 48 67 Expt 5 (2002) % parasitism 0 0 4 4 2% ±2.3% (per field cage) number of eggs 4 16 5 1 6.5 Expt 6 (2002) % parasitism 0 0 0 0 0% ±0.0% (per field cage) Expt 1 number of eggs 42 67 49 50 22 46 (2001) % parasitism 31 39 14 8 27 24% ±12.6% (per field cage) Table 3-9. Percent parasitism of the SL eggs found in the SL combined-host field cages Field cages Average ± SD 1 2 3 4 5 number of eggs 30 15 0 1 26 14.4 Expt 1 (2002) % parasitism 3 0 0 0 20 8% ±10.8% (per field cage) number of eggs 73 151 38 226 122 Expt 3 (2002) % parasitism 0 1 3 4 2% ±1.8% (per field cage) number of eggs 199 162 353 343 264.25 Expt 4 (2002) % parasitism 22 22 17 11 17% ±5.2% (per field cage) number of eggs 71 19 41 12 35.75 Expt 5 (2002) % parasitism 4 5 0 0 3% ±2.6% (per field cage) number of eggs 20 6 40 41 26.75 Expt 6 (2002) % parasitism 0 0 3 11 5% ±5.2% (per field cage) Expt 1 number of eggs 86 9 64 14 34 41.4 (2001) % parasitism 19 11 34 21 24 24% ±8.3% (per field cage)

PAGE 93

75 As with the parasitism of DBM eggs, the parasitism of SL eggs in the combinedhost field cages was similar to the parasitism of SL eggs in the single-host field cages. This is shown in Table 3-9 where again E xperiment 4 and Experiment 1 (2001) had the highest levels of parasitism. Levels of parasitism in all other experiments were under 10%. Differences in Percent Parasiti sm of the DBM and SL Eggs When percent parasitism of the two hostsÂ’ eggs was compared for each experiment by using percent parasitism of each host species eggs as recorded per field cage, statistical differences were onl y found in two of the experiments. Only Experiments 1 and 6 recorded signifi cant differences in percent parasitism between the host species (Experiment 1: t =7.225, df = 11, P > 0.0001 & Experiment 6: t = 2.383, df = 11, P = 0.0363). It is thought that the vari ation in numbers of eggs found between field cages and the subsequent per cent parasitism differences had much to do with there not being other significant di fferences found. Significant differences were expected in Experiment 5, wh ere parasitism in DBM eggs seemed much higher, and in Experiment 4, where higher parasitism of SL eggs was accompanied by large numbers of both hostÂ’s eggs and little variation in pe rcent parasitism figures. Experiment 1 (2001) had too few DBM eggs for there to be any meaningful comparison of percent parasitism between species. Inter-Quartile Differences in Percent Parasitism Diamondback moth eggs Figures 3-16 through 3-20 show how the parasitism of the DBM eggs was divided between the quartiles of the cabbage plants in the field experiments. There was no parasitism recorded in Experiment 2.

PAGE 94

76 Figure 3-17. Proportion of parasitized DB M eggs by quartile in Experiment 1 (2001) Experiment 1 (2002)0 10 20 30 40 50 60 70 80 90stem1234 plant region ( quartiles )egg numbers non-parasitized parasitized Figure 3-18. Proportion of pa rasitized DBM eggs by quartile in Experiment 1 (2002) Figure 3-19. Proportion of parasitized DBM eggs by quartile in Experiment 3 Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 0% 0% 1 38% 43% 2 27% 43% 3 0% 0% 4 20% 14% Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 75% 17% 1 94% 60% 2 40% 24% 3 0% 0% 4 0% 0% Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 25% 9% 1 5% 4% 2 3% 30% 3 4% 57% 4 0% 0% Experiment 1 (2001)0 2 4 6 8 10 12stem1234 plant region (quartiles)egg number non-parasitized parasitized Experiment 30 50 100 150 200 250 300 350 400stem1234 plant region ( quartiles )egg numbers non-parasitized parasitized

PAGE 95

77 Figure 3-20. Proportion of parasitized DBM eggs by quartile in Experiment 4 Figure 3-21. Proportion of parasitized DBM eggs by quartile in Experiment 5 Figure 3-22. Proportion of parasitized DBM eggs by quartile in Experiment 6 Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 0% 0% 1 23% 26% 2 18% 30% 3 8% 28% 4 8% 16% Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 19% 6% 1 62% 16% 2 55% 61% 3 15% 16% 4 0% 0% Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 0% 0% 1 40% 14% 2 19% 65% 3 11% 21% 4 0% 0% Experiment 40 50 100 150 200 250 300 350 400stem1234plant region ( quartiles )egg numbers non-parasitized parasitized Experiment 50 20 40 60 80 100 120 140stem1234 plant region ( quartiles )egg numbers non-parasitized parasitized Experiment 60 20 40 60 80 100 120 140 160stem1234 plant region ( quartiles )egg numbers non-parasitized parasitized

PAGE 96

78 In all experiments, the highest percent pa rasitism of eggs was found in the first quartile. The next highest percent parasiti sm was found in the second quartile. In all experiments apart from Experiment 1 (2002) th e greatest number of parasitized eggs was found in the second quartile, largely because of the fact that the middle two quartiles often contained the majority of the DBM eggs and that the levels of parasitism were higher in the lower levels of the plants. Th e fourth quarter, in all but Experiment 4, contained no parasitized eggs. Soybean looper eggs Figures 3-21 through 3-24 show how the parasitism of the SL eggs was divided between the quartiles of the cabbage plants in Experiments 1 (2001), 1 (2002), 3, 4 and 5. There was no parasitism recorded in Experiment 2 or Experiment 6. Figure 3-23. Proportion of pa rasitized SL eggs by quartile in Experiment 1 (2001) Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 0% 0% 1 13% 4% 2 33% 86% 3 10% 10% 4 0% 0% Experiment 1 (2001)0 20 40 60 80 100 120 140stem1234plant region (quartiles )egg number non-parasitized parasitized

PAGE 97

79 Figure 3-24. Proportion of pa rasitized SL eggs by quartile in Experiment 1 (2002) Figure 3-25. Proportion of parasitized SL eggs by qua rtile in Experiment 3 Figure 3-26. Proportion of parasitized SL eggs by qua rtile in Experiment 4 Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 0% 0% 1 30% 20% 2 11% 80% 3 0% 0% 4 0% 0% Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 0% 0% 1 0% 0% 2 1% 100% 3 0% 0% 4 0% 0% Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 0% 0% 1 15% 5% 2 18% 55% 3 9% 37% 4 2% 3% Experiment 1 (2002) 0 20 40 60 80 100 120stem1234 plant region ( quartiles )egg numbers non-parasitized parasitized Experiment 30 20 40 60 80 100 120stem1234 plant region ( quartiles )egg numbers non-parasitized parasitized Experiment 40 50 100 150 200 250 300 350stem 1 2 3 4plant region ( quartiles )egg numbers non-parasitized non-parasitized

PAGE 98

80 Figure 3-27. Proportion of parasitized SL eggs by qua rtile in Experiment 5 In all of the experiments, the majority of the parasitized SL eggs were found in the bottom two quartiles of the cabbage plan ts. The second quartile always contained the greatest number of parasitized eggs and, asid e from Experiment 1, the highest percent of parasitized SL eggs. Only in Experiment 4 we re there significant numbers of parasitized SL eggs found in the third and fourth quartile. Inter-Quartile Distribution of Parasitized Eggs in Comparison to the Inter-Quartile Distribution of all Eggs The figures 3-16 through 3-24 show that the parasitized eggs generally have a lower distribution within the plant than the eggs taken as a whole (i.e., both parasitized and non-parasitized eggs). Tables 3-10 th rough 3-14 compare the distributions of parasitized eggs to all eggs of that species, by showing the leaf number at which the 25th, 50th and 75th percentiles of the parasitized and total egg groups were found. Statistical differences in the distribution of paras itized and nonparasitized eggs are given. Quartile % parasitized QuartileÂ’s % of total # of parasitized DBM eggs stem 0% 0% 1 0% 0% 2 8% 83% 3 0% 0% 4 2% 17% Experiment 50 20 40 60 80 100 120 140stem1234 plant region ( quartiles )egg numbers non-parasitized parasitized

PAGE 99

81 Table 3-10. Distribution profiles of parasitized eggs and all eggs for DBM and SL in the combined host field cages of Experiment 1 (2001). Percentiles DBM a paras a SL a paras a (leaf number) (leaf number) min 1 3 1 2 25th 3.75 3 5 5 50th 6 5 5 5 75th 8 6 7 6 max 12 11 12 8 ave 6.03 5.29 5.62 5.18 Note: Different letters to the right of the ta ble headings indicate signi ficant differences in the distribution of the parasitized eggs a nd nonparasitized eggs. Analysis was done for each host species (DBM: Z = 0.4698, P = 0.4931; SL: Z = 1.7445, P = 0.1866) Table 3-11. Distribution profiles of parasitized eggs and all eggs for DBM and SL in the combined host field cages of Experiment 1 (2002). Percentiles DBM a paras b SL a paras a (leaf number) (leaf number) min 1 1 1 1 25th 2 1 4 3.75 50th 2 2 6 6 75th 4 3 7 6 max 8 8 9 6 ave 2.60 2.22 5.40 4.67 Note: Different letters to the right of the ta ble headings indicate signi ficant differences in the distribution of the parasitized eggs a nd nonparasitized eggs. Analysis was done for each host species (DBM: Z = 10.12, P = 0.0015; SL: Z = 0.74, P = 0.390).

PAGE 100

82 Table 3-12. Distribution profiles of parasitized eggs and all eggs for DBM and SL in the combined host field cages of Experiment 3. Percentiles DBM a paras b SL paras (leaf number) (leaf number) min 1 4 1 5 25th 7 4 6 5.5 50th 9 5 7 6 75th 11 7.5 9 11 max 16 10 14 12 ave 8.91 6.00 7.44 8.00 Note: Different letters to the right of the tabl e headings indicate signi ficant differences in the distribution of the parasitized eggs a nd nonparasitized eggs. Analysis was done for each host species (DBM: Z = 12.95, P = 0.0003; SL: not enough parasitized SL eggs to do the analysis). Table 3-13. Distribution profiles of parasitized eggs and all eggs for DBM and SL in the combined host field cages of Experiment 4. Percentiles DBM a paras b SL a paras b (leaf number) (leaf number) min 3 4 2 2 25th 13 8 11 8 50th 19 10.5 15 11 75th 22 13.5 19 16 max 29 27 29 25 ave 17.77 12.40 15.02 12.26 Note: Different letters to the right of the ta ble headings indicate signi ficant differences in the distribution of the parasitized eggs a nd nonparasitized eggs. Analysis was done for each host species (DBM: Z = 15.78, P < 0.0001; SL: Z = 12.00, P = 0.0005).

PAGE 101

83 Table 3-14. Distribution profiles of parasitized eggs and all eggs for DBM and SL in the combined host field cages of Experiment 5. Percentiles DBM a paras b SL a paras b (leaf number) (leaf number) Min 1 1 1 3 25th 3 2 5 3.75 50th 5 3 6 4 75th 8 4.25 8 4.5 max 9 5 10 6 Ave 5.59 3.15 6.28 4.25 Note: Different letters to the right of the ta ble headings indicate signi ficant differences in the distribution of the parasitized eggs a nd nonparasitized eggs. Analysis was done for each host species (DBM: Z = 18.18, P < 0.0001; SL: Z = 5.13, P = 0.0235). Table 3-15. Distribution profiles of parasitized eggs and all eggs for DBM and SL in the combined host field cages of Experiment 6. Percentiles DBM a paras b SL a paras a (leaf number) (leaf number) Min 4 4 6 7 25th 7 6 10 7 50th 11 7 12 11 75th 13 11 14 12 Max 20 14 17 17 Ave 10.18 8.41 11.93 10.80 Note: Different letters to the right of the tabl e headings indicate signi ficant differences in the distribution of the parasitized eggs a nd nonparasitized eggs. Analysis was done for each host species (DBM: Z = 14.98, P = 0.0001; SL: Z = 0.66, P = 0.4182) For all experiments, apart from Experiment 1 (2001), there were marked differences in distribution of parasitized DBM eggs as compared to DBM eggs taken as a whole. In all cases parasitized DBM eggs we re more likely to be found lower to the ground than DBM eggs taken as a whole. This pa ttern is repeated for SL eggs in all four

PAGE 102

84 experiments (Experiments 1, 4, 5 & 6) where th ere were sufficient numbers of parasitized SL eggs to make the comparison. Statistical differences were found in Experiments 4 & 5. Position of DBM Larvae in Reference to DBM Eggs Found on the Cabbage Plants It was noted in the experiments that when DBM larvae were present, they were commonly found higher in the plants than the DBM eggs. For Experiments 3 and 5, positions of first and second instar DBM larv ae were recorded—specifically leaf number from the ground and adaxial/abaxial position. Be cause of relatively high levels of native DBM present in the plots, there were always DBM larvae present to some degree on the plants. Tables 3-16 and 3-17 show the total number of DBM eggs and first and second instar larvae found in Experiments 3 and 5. Figure 3-28 is a graphical representation of these numbers. The vertical distribution of the eggs and larvae are divided by plant quartiles. Table 3-16. DBM egg and larval number s in Experiment 3, broken down by plant quartile. Includes the percent for each quartile of the to tal number of eggs or larvae present Quartile DBM egg # Quartile's % of total DBM larvae # Quartile's % of total 1 13 3% 4 1% 2 147 31% 38 10% 3 274 57% 133 34% 4 43 9% 211 55%

PAGE 103

85 Table 3-17. DBM egg and larval number s in Experiment 5, broken down by plant quartile. Includes the percent for each quartile of the to tal number of eggs or larvae present Quartile DBM egg # Quartile's % of total DBM larvae # Quartile's % of total 1 13 8% 0 0% 2 55 34% 3 2% 3 52 32% 41 27% 4 42 26% 110 71% Figure 3-28. The distribution of DBM eggs a nd larvae within the quartiles of the plant for Experiments 3 & 5 (Quartile 1 being the closest to the ground) Table 3-18 gives the distributi onal profile by leaf number of the DBM eggs and larvae in Experiments 3 & 5. Statistical differences were found in both experiments for the distribution of DBM eggs and larvae. Also included in the table is the ‘average’ leaf position of the eggs or larvae. Expt 3 (DBM eggs & larvae position)0% 10% 20% 30% 40% 50% 60% 70% 80% 1234Plant Quartilespercentage of plant total DBM eggs DBM larvae Expt 5 (DBM eggs & larvae position)0% 10% 20% 30% 40% 50% 60% 70% 80% 1234Plant Quartilespercentage of plant total DBM eggs DBM larvae

PAGE 104

86 Table 3-18. Distribution profile s by percentile for DBM eggs and larvae in the cabbage plants of Experiments 3 &5. Quartiles Expt. 3 Expt. 5 Eggs a Larvae b Eggs a Larvae b (leaf number) (leaf number) Min. 2 2 1 3 1st 7 11 4 6 2nd 9 13 5 8 3rd 10 14 7 8 Max. 16 16 9 9 Ave 8.74 12.21 5.25 7.35 Note: Different letters to the right of the “Eggs” or “Larvae” table headings indicate significant differences between the distributi on of the DBM eggs and DBM larvae (Expt. 3: Z = 273.03, P < 0.0001; Expt. 5: Z = 83.607, P = < 0.0001) The tables and figures presented abov e show that the DBM larvae were found higher in the cabbage plants for the two e xperiments where data were taken. A similar pattern was noted for other experiments wher e native DBM larvae were present. It must be said that the DBM larvae counted did not originate from the eggs laid by the experimental female moths. The larvae had emerged from eggs laid by native populations of DBM already present in the plots. The e ggs from which the larvae had emerged would have been laid around 2–4 days before the commencement of the experiment. Discussion Host Egg Positioning There were clear differences between DB M and SL oviposition on the plant. The most striking difference was how the SL fe males laid the most of their eggs on the abaxial side, close to the leaf margin, while DBM preferred the adaxial side of the leaf. Only DBM oviposited on the stems of the pl ants and only its eggs were found to any significant degree on the petioles of the leaves . Harcourt (1961) previously found similar

PAGE 105

87 results, although more DBM eggs were f ound on the stems and petioles in this study. Possibly, DBM’s smaller size allows the fema le to enter the plant to oviposit. That SL and other loopers predominantly oviposit on the underside of the leaves has previously been published (Harcour t 1962, Greene 1968, Pansera de Araújo 1999). Harcourt found 91% on the abaxial side while Pansera de Araújo et al. (1999) found 69% of the eggs on the abaxial side. Pansera de Araújo et al. (1999), also found that SL females tended to lay their eggs in the lowe r parts of soybean plants. Field observations seemed to suggest that SL adult females laid their eggs by positioning themselves close to the leaf margin on the adaxial side of th e leaf, bending their abdomen around the leaf edge and depositing the eggs on the abaxial side. From the field experiments, it also seem s that DBM females lay their eggs lower on the plant than SL when the plants are sm aller. In the experiments with the smaller plants (Experiments 1 (2001), 1 (2002) & 5) , the DBM eggs were found lower in the plants than SL eggs. For the more mature pl ants, this was reversed. In Experiments 3, 4 & 6, it was the SL eggs that were found lower in the plant. From personal observations in expe riments where DBM larvae had recently emerged, the larvae were always found higher in the plant than the egg remnants from which they emerged. It seems that the larv ae move up the plant, and feed on the newer leaves near the apical bud. Following the lo gic that DBM larvae tend to feed and develop in the upper reaches of the pl ant, the observed oviposition behavior could prove to be advantageous to survival; i.e. not ovi positing too close to the apical bud, where sometimes very large numbers of DBM larv ae are feeding and doing damage, and not too far from the apical bud, where the distance to the top is greater. This may explain why,

PAGE 106

88 for the bigger plants in the e xperiments, the median point of DBM egg-laying was further away from the ground, with the distance betw een the apical point and the main egglaying area remaining, more or less, the same. Th is type of vertical movement in the egglaying zone has been observed in other sy stems (Spangler and Calvin 2001), and was linked to chemical changes during the plantsÂ’ development and the optimization of larval feeding. An alternate reason as to why the egg-la ying zone of DBM moved with respect to cabbage plant growth could be linked to exis ting larval damage. Reed et al. (1989) found that allyl isothiocynates, metabolites from gl ucosinolates, are produced in large amounts in damaged crucifers. It has also been f ound that these compounds are repellent to DBM females at high concentrations, as with ca bbage plants with heavy feeding damage (Pivnick et al. 1994). It could be that the females, repelled fr om the areas near the apical point with larval damage, choose to lay their eggs below the larval feeding area. As the plants grew, the DBM eggs would be found higher in the plant. In many of the experiments, there was already significant f eeding damage from indigenous DBM field populations. Whether the existing damage could have produced enough metabolites to repel is not known. DBM females may lay eggs based on existing larval feeding rather than potential larval feeding. The SL eggs, on the other hand, tended to move less within the plantsÂ’ profile with cabbage plants of differing sizes. With longer egg emergence times and fewer native SL around, there was no chance of observing SL larval behavior. These larvae do grow considerably bigger than DB M larvae, and so perhaps ar e less constrained in their

PAGE 107

89 movements within a cabbage plant. If this was the case, then factors other than distance from apical growing point, may determin e where SL females lay their eggs. Parasitism of DBM and SL Eggs by Trichogramma pretiosum Field parasitism of the host spec ies’ eggs, with the numbers of Trichogramma used in these experiments, was generally lo w. Only once did average percent parasitism exceed 50% in an experiment, and this was with DBM eggs. DBM eggs were generally parasitized more than SL eggs 10 of th e possible 12 DBM experiments (single-host and combined-host) recorded an average of more than 10% parasitism. In the 12 SL experiments only 4 of the experiments recorded average parasitism of greater than 10%. In DBM there was generally higher parasitism in the experiments with smaller plants. This was not the case with SL. These experiments show that in different cabbage plantings, the distribution of host eggs in the DBM/SL host system cha nges—more of one host’s eggs can be found lower (or higher) in the plant than the eggs of the other host. In the three experiments (Experiments 1 (2002), 5 & 6) with high levels of DBM parasitism, the DBM eggs were found lower in the plant than the SL eggs. In Experiment 3, where DBM parasitism levels were lowest, SL eggs were generally found cl oser to the ground. In the other experiment where SL eggs were found lower in the plan t (Experiment 4), SL egg parasitism was higher than DBM egg parasitism. This would su ggest that whichever hosts’ eggs have the lower distribution in the plant, are more highly parasitized by Trichogramma If T. pretiosum had an intrinsic preference of on e host’s eggs over the other, how would that affect the parasitism of the two host species’ eggs with the relative changes in their egg distributions? With th e data gained from these experiments it is difficult to draw any real conclusions. What can be said is that, irrespective of which of the two host

PAGE 108

90 species is higher or lower on the plant, the le vel of parasitism for each host species is the same for the single-host treatments and for the combined-host treatments. This would suggest that the level of parasitism is not a ffected by what type of host egg is encountered first on the waspÂ’s way up the pl ant. Level of parasitism in a host species seems more to be affected by how soon that hostÂ’s eggs are encountered on the way up the plant. Trichogramma pretiosumÂ’s parasitism patterns within the plants For both hosts, there was greater parasitism in those eggs found in the lower parts of the plant. Only in one of the experiment s were there more parasitized DBM eggs found in the upper half of the plan t than in the lower half. Normally, about 80% of all parasitized DBM eggs were found in the lower half of the plant. This was even more pronounced for SL eggs, where around 90+% of the parasitized eggs, were found in the lower half of the plant. Furthermore, the di stribution pattern of the parasitized eggs as compared to the eggs taken as a whole was always lower in the plant, irrespective of other factors such as which quartile the bulk of the eggs were laid in, or how the two host speciesÂ’ eggs were positioned in relation to each other. This work supports the conclusions made by Gingras et al. (2003) that female Trichogramma parasitize more eggs at the base of cruciferous plants. The greater number of parasi tized eggs in the bottom ha lf of the plant may be due to T. pretiosum Â’s searching behavior. The Trichogramma used in the experiments hardly ever seem to fly, preferring to walk while e xploring their environment. This behavior has been observed before (Bigler et al . 1988). It seems that the Trichogramma did not to fly into the canopy of the plant, but instead walked up into the pl ant from the diet cups that had been placed in the field cages. (These Trichogramma had been examined previously and all had fully functioning wings). Circumstantial evidence for the Trichogramma

PAGE 109

91 walking up the stems, is the relatively high numbers of parasitized DBM eggs that were found on the stems of the plants (SL e ggs werenÂ’t found on the stems). If the Trichogramma did walk up the plant, searching for e ggs as they went, this would explain the pattern of parasitism found. Perhaps, after a certain amount of eggs were parasitized, the Trichogramma stopped parasitizing because of e xposure to strong winds, or just a decision on the suitability of th e top half of the cabbage plan ts as a habitat for finding and laying eggs. Fournier and Boivin (2000) found that when wind blew above 15 km/h for more than 4 hours in a day, there was significant reduction in Trichogramma dispersion. As Appendix A shows, these sort of wind speed s were experienced during the experiments and may well have had an effect on Trichogramma dispersion. Perhaps the high winds on the south coast originally forced the Trichogramma to walk, and not fly to the plants, which led to them searching the plants from the bottom up. The high winds may also have had some impact on how high the Trichogramma were willing to go up the plant. Alternatively, the Trichogramma, less protected from the winds higher up the plants (not protected by the sandbags, row banks or foliage ), may have been blown off the plants. Naranjo (1993) found that Trichogrammatoidea bactrae (Nagaraja) females laid most of their eggs within the first few hours after emergence (>90% we re laid within the first 12 hours). Others also found that Trichogramma laid their eggs early after oviposition (Fye and Larsen 1969, Pak and Oatman 1982). Knowing that an adult female parasitoidÂ’s life is a bala nce between finding food and fulfilling reproductive needs (Lewis et al. 1998), the work of NaranjoÂ’s and others would sugge st that there is a shift in priorities after the initial hour s of parasitism. The other alte rnative is that the female

PAGE 110

92 exhausts its complement of eggs after a short while (Naranjo found that maximum fecundity was 55 progeny per female for T. bactrae ). Whichever the case, the parasitism patterns encountered in the pr esent study may be explained by a change in biological priorities. Perhaps, during the first few hours after release, when parasitizing was a strong priority, the Trichogramma were still in the lower reaches of the plants. By the time the females were entering the higher parts of the plants, their need for nutrition had supplanted their need to find more host eggs. One other parasitism pattern observed was that in the larger plants (experiments 3, 4 & 6), the parasitized eggs were more spread out over the quartiles of the plants. This seems to go against the idea that in the bigger plants the Trichogramma would have to travel further to get to the top half of the plan t to parasitize. The results can perhaps be explained by the greater connectivity found in the larger cabbage plants, where leaves tend to droop down and touch those leaves be low. This follows the observations made by Gingras and Boivin (2002) who found that in their work, host-finding success in T. evanescens was influenced more by plant connectivity than plant size. Use of T. pretiosum in the Control of DBM and SL in Cabbage When one examines the literature of Trichogramma use in field experiment situations, one of the first things of note is the extent of variability found in the levels of parasitism. This is to be exp ected in field situations wher e the many factors that affect percent parasitism are never the same. This pr oved to be the case in the work presented above. What is interesting about the findings of this research is how the position of the hostÂ’s eggs within the plant can affect parasitism levels. Furthermore, it seems that in the case of DBM, the positioning of host eggs is linked to the development of the plant.

PAGE 111

93 The other conclusion of note is that the Trichogramma in these experiments only parasitized the host eggs in the lower parts of the cabbage plants. If this proves to be a common phenomenon, then knowing where the bul k of the pestÂ’s eggs are within the plant profile will be important to the effective use of Trichogramma . If, as the DBM data suggests, the bulk of host eggs moves higher up the plant as the pl ant develops, then some way has to be found to get the Trichogramma up higher. The effect of plant development on parasitism levels is then not only about increased surface area or increased architectural complexity, but al so the changing ovipositional response of the host species to this plant development (i.e., a tritrophic interaction). If future experiments show that T. pretiosum does indeed tend to walk to the base of the plants from release points at ground level, and then parasitize eggs as it moves up the plants, then some other type of release methodology may need to be found. What happens to Trichogramma released from elevated positions? How many fly into the crop and how many fly, drop or walk onto the ground before moving onto the plants? In terms of host preference, the presence of either host did not the affect the subsequent parasitism of the other hostÂ’s eggs . This suggests that it is not important for a crop manager to determine the mix of these tw o hosts in a field for biological control purposes when using T. pretiosum . The only thing that can be said is that T. pretiosum seems to prefer DBM over SL, and that it is more willing to continue parasitizing into the higher parts of the cabbage pl ant in search of DBM than with SL. This preference of hosts is examined in the laboratory expe riments described in the next chapter. Potential IPM Control Program for DBM in Puerto Rico From this work it seems that an IPM program, which incorporates the use of Trichogramma, could be successful. Th e primary candidate for use is T. pretiosum, as it

PAGE 112

94 has proved that it can parasitize DBM in the field. The next likely candidate would be T. bactrae , which has shown high parasitism levels of both DBM and SL in the field. Before the release of the Trichogramma, a windbreak should be established—stands of sugar cane, banana, plantain or maize probably working best. Trichogramma looks like it would have its greatest impact when the cabba ge plants were young, be fore the onset of cupping. Without strong winds and with rela tively simple plant architecture, the Trichogramma have the potential to find and paras itize most of the eggs laid by DBM. The quantities of Trichogramma released would have to be higher than those released in the field experiments (~195,000 per hectare) to ensure higher parasiti sm than that found in the experiments. The release points would have to be low to the ground if there was significant wind, with perhaps more releas e points upwind of the cabbage plants. Cabbage is a low plant, and there is the dange r of the wasps being carried away from the cropping area. The problem with ground releas es is that effective dispersal of the Trichogramma within the crop may be hindered by th e insect’s propensity to walk from ground release points. This would necessitate an increase in release points and added labor costs. There is also th e danger of predators such as fire ants, including the red imported fire ant, Solenopsis invicta, (Davis Jnr. et al . 2001). Fire ants we re present in the fields of the south coast and mi ght have preyed on the emerging Trichogramma . To maintain Trichogramma in the cropping area, there could be a type of banker plant system used. Field observations in the cabbage growing areas of the south coast of Puerto Rico showed that there were indigenous Trichogramma parasitizing the eggs of Zinckenia fascialis, a pyralid species comm only found in the weeds Trianthema portulacastrum (peseta or horse purslane) and Portulaca oleracea (purslane or

PAGE 113

95 verdolaga). Figure 3-29 shows these weeds bor dering a field of cabbage on a commercial farm on the south coast of Puerto Rico. Figure 3-29. The weed Portulaca oleracea (purslane or verdolaga) bordering a field of cabbage The pyralid did not attack the cabbage, and so could potentia lly be used to augment Trichogramma numbers in cabbage fields. The weeds may also provide additional nutrients for the Trichogramma adults. It is presently common practice to remove the weeds, but perhaps in certain ar eas they could be left alone. Intuitively a monoculture of cabbage, with little protection from wind a nd no flowers, would not seem a good environment for Trichogramma . Finding some way of retaining Trichogramma in the crop may increase parasitism levels. Trichogramma would complement the use of other parasitoids such as Diadegma insulare (Cresson) and Cotesia plutellae (Kurjumov) that have been shown to be effective (Muckenfuss 1992, Morallo-Rejes us and Sayaboc 1992, González-Rodríguez and Macchiavelli unpubl.) . González-Rodríguez and Macchiave lli’s work in Puerto Rico indicates that C. plutellae has more of an impact on DBM population levels in the later

PAGE 114

96 stages of the cropÂ’s development. Trichogramma could be effectively used in the earlier stages to maintain low levels of DBM in the field. Low levels of DBM at the crucial stage of pre-cupping is thought to be ke y to maintaining cont rol of the pest at the later stages of head formation, when minimal DBM damage is essential (Leibee pers. comm .). Scouting programs using sentinel plants and plant sa mpling would give important information on parasitism levels of the parasi toids being utilized. This info rmation can then be linked to actual plant damage and marketable yields. In other crop systems w ith different insect pests, the parasitism of eggs by Trichogramma did not result in impr oved saleable yields due to compensatory population shifts at othe r life stages of the pe st (Suh et al. 2000). This would need to be established for the cabbage/DBM/ Trichogramma system. To secure marketable yields, the judicious use of compatible insecticides, such as Bt products and some of the newer chemical compounds such as spinosad, emamectin benzoate and indoxacarb would most proba bly be used. Reducing the number of applications of these insectic ides through the use of the para sitoids would help conserve these products against the development of resistance.

PAGE 115

97 CHAPTER 4 LABORATORY INVESTIGATIONS OF HOS T PREFERENCE IN THE PARASITISM OF DIAMONDBACK MOTH AND SOYBEAN LOOPER BY Trichogramma Introduction The field experiments conducted on the sout h coast of Puerto Ri co (Pluke et al. unpubl .) give varying parasitism rates of soybean looper eggs ( Pseudoplusia includens ) and diamondback moth eggs ( Plutella xylostella ) by Trichogramma pretiosum on cabbage. The rates of parasitism found de pended on many factors that included plant size, T. pretiosum release rates and host egg positioni ng. It proved difficult to determine host preference of T. pretiosum as all other factors were not equal. For this reason a series of laboratory experiments were conducted to determine host preferen ces under laboratory conditions. With T. pretiosum showing a propensity to walk up the cabbage plants, and the host eggs being located in different vertical zones of the plant, it was thought that T. pretiosum would be exposed to the eggs of one host species before the e ggs of the other host species. How would this pr ior exposure influence Trichogramma Â’s subsequent host choices and parasitism rates? Nurindah et al. (1999) show ed that experience with host eggs increased the likelihood of Trichogramma australicum (Girault) responding to ensuing host eggs and decreased the handling time of these eggs. This improvement of parasitoid efficiency can be interpreted as -conditioning or the ga ining of experience (Vinson 1998). -conditioning is linked to encoun tering novel cues that trigger a genetically programmed response and can be called associative learning. The

PAGE 116

98 Trichogramma in the field experiments could have experienced both types of conditioning. The first, -conditioning, would have led to greater efficiency in the processing of additional eggs of that speci es (or perhaps of other species), while conditioning would have been a more species-s pecific associative res ponse that may have affected host preferences. Laboratory experi ments were designed to explore how prior exposure of T. pretiosum to one or other of the two host speciesÂ’ eggs would affect the host preferences expressed in the firs t series of laboratory experiments. The final series of laboratory experime nts introduced two other species of Trichogrammatids to examine host pref erences. The two species chosen, Trichogrammatoidea bactrae and Trichogramma minutum, were chosen for their compatibility to the Puerto Rican climate a nd for their potential in parasitizing DBM. Materials and Methods Trichogramma The three Trichogramma species were sourced from Beneficial Insectaries and Rincon Vitova, who sent Trichogramma egg cards on a weekly basis. Beneficial Insectaries sent the Trichogramma pretiosum while Rincon Vitova sent the Trichogramma minutum and Trichogrammatoidea bactrae . Both companies rear their Trichogramma in Ephestia kuehniella Zeller (Lepidoptera: Pyra lidae) eggs. Three days after delivery, the adults starte d to emerge, and it was these in sects that were used in the laboratory experiments. On receipt of the Trichogramma , the egg cards were cut into small strips and placed into 25-dram, clea r styrene tubes with tabbed caps (Bioquip catalogue # 8925) . These vials were then placed in the insect colony room (temperature 27 3oC, 65 5% r.h. and 16L:8D) to await the em ergence of the adults. Normally, the

PAGE 117

99 adults started emerging during the scotophase of the third day, and were ready for use on the fourth day after delivery. The females used in the experiments were all less than 24 hours old and had been in the presence of males on emergence, and therefore could have mated. The individual females were isolated by gently tapping the vial containing the Trichogramma adults onto the illuminated surface of a photographer's li ght table. The light table was used to minimize the number of Trichogramma escaping. An added advantage of the light table was that, with practice, one could determine the males from the females with the naked eye, using ‘hairiness’ of the antennae as th e defining characteristic. A ¼-dram shell vial was placed over each individual female and then, when she had made her way up the sides of the container, the vial was righted and the vial stoppered. Soybean Looper (SL) The soybean looper eggs used in the e xperiments came from a laboratory culture that had been established in the Fall of 2000 at the Univ ersity of Puerto Rico's agricultural research station in Río Piedras. The original insects had been collected as larvae and pupae from fields of various crops in Puerto Rico (soybean, tomato and eggplant being the principal host crops) and were later identified by Dr. Heppner (taxonomist, Department of Plant Industr y, Florida Department of Agriculture and Consumer Services) as soybean looper. The in sects were feed on the cabbage looper diet from BioServ (Bio-Serv catalogue # F9282B). On emergence from the egg, five firstinstar SL larvae were placed into each diet c up that had been quarte r-filled with diet. Airholes were punched into the caps of the diet cups and the trays of diet cups were placed in an insect rearing room (27 3oC, 75 5% r.h. and 16L:8D). After about 10 days, each

PAGE 118

100 larva was transferred to its own cup containing fresh diet and placed back into the rearing room. Because of frass accumulation and so me drying of the diet, the larvae were transferred to fresh diet one more time be fore pupation. The larvae were allowed to pupate in the diet cups. Cocoons were spun on the top or sides of the cups. The adults emerged after 6–7 days and were placed in plastic storage boxes (40 cm 15 cm 10 cm) with other recently emerged a dults. Two diet cups (pill cups) were placed into each box, one with water and the other with 5% honey-water. Dental cotton wicks were placed in the cups to soak up th e liquid to prevent the adults from drowning. To augment humidity, a folded paper towel, sa turated with water, was also placed in the plastic box. All water used in the box was distilled and autoclaved to prevent the development of bacterial, viral or fungal di seases. The box was covered by light medical gauze and held in place with ¼ ”-wide ba nd of sowing elastic. Once, early in the morning, and once again, late afternoon, the gauze was sprayed with a fine mist of distilled, autoclaved water to maintain hu midity. The water and honey-water diet cups and the paper towel were replaced everyday to minimize disease problems. The females started ovipositing about 4 days after placement in the plastic box. At this time the gauze was replaced by dark bl ue chiffon material placed over and down the insides of the box. Again this material was he ld together by the el astic sewing tape. The change in material was made necessary because it had proved impossible to remove the eggs from the gauze without damaging them. An advantage of the chiffon was that, by choosing a dark color, it was easy to find the eggs on the dark background. The female SL adults oviposited for a period of about a week, with maximum egg-laying occurring between the second and fift h day from commencement.

PAGE 119

101 When there were no experiments running, the egg-laden chiffon was removed and placed into another plastic container. The chi ffon was sprayed with a fine mist of water and a plastic top used to cl ose the container. In this way, nearly 100% humidity was maintained. First instars started to emerge 3–4 days later and were cultured as previously described. Diamondback Moth (DBM) DBM were obtained from a culture maintain ed at the Mid-Florida REC (University of Florida) in Apopka, Florida. Cocoons were shipped overnight. When the insects arrived, the plastic netting screens, onto whic h the diamondback moths had pupated, were placed into a Plexiglas cage (0.3 0.3 0.3 m) in the insect rearing room (temperature 27 ± 3oC; 75 ± 5% r.h. and 16L:8D). Two diet cu ps; one with water and the other with 5% honey water were placed into the Plexigla s cages. Dental cott on wicks measuring 2.5 cm in length were placed in the cups to so ak up the liquid to prevent the adults from drowning. To augment humidity a folded paper towel, saturated with water, was placed in the plastic cage. In general, ad ults started to emerge 2 days after arrival. DBM started to lay eggs a day or two after emergence and laid the majority of the eggs within the first 5 days. As with the SL eggs, it was difficult to develop a dependable method of acquiring undamaged DBM eggs for the experiments. Th e method that worked best is outlined below. Firstly, soluble paper (SolvyTM Original Water Soluble Stabilizer, Item No. 48608, SULKY of America, Harbor Heights, FL.) was cut into 20 cm by 25 cm sheets. A 10% solution of macerated cabbage leaf wa s then prepared and filtered into a handpumped mister bottle. A fine mist was then sprayed over the soluble paper sheets until they were covered in fine droplets. When dr y, the sheets exhibited a puckered surface that

PAGE 120

102 proved ideal for DBM oviposition. The sh eets were stored in a fridge at 4oC, with sheets of kitchen towel separating the individual so luble sheets to avoid them sticking. During DBM oviposition, the soluble sheets were hung in the cages by wetting the top edge and sticking them to the ceiling of the cage. Th is was done late afternoon. By morning the sheets were usually full of eggs. Experimental Design FalconTM Petri dishes (10 cm in diameter a nd 15 mm in depth) were used in all experiments. There were many factors to ta ke into consideration when designing the exact protocol. The Trichogramma are minute insects and eas ily desiccated by overly dry conditions. On the other hand, they can easily drown in condensation droplets that form around the edge of a humid, closed Petri dish. At various ti mes during protocol development, filter paper or greaseproof paper was used to line the bottom of the Petri dishes. It seems that the filter paper-lined Petri dish created too dry an environment for the Trichogramma as there was high Trichogramma mortality observed. The greaseproof paper also proved to be less than ideal. The pa per never sat flush to the bottom of the dish and so created an underside, separated from the eggs into which the Trichogramma could get lost. In the end Petri dish es without lining were used. Based on the literature, a food source wa s also deemed necessary during the 24hour experiment (Hassan and Guo 1991). Fr om preliminary experiments, it was determined that significantly less Trichogramma died when the 10% honey water was offered as a small droplet in the middle of th e Petri dish as opposed to it being soaked into a small filter paper, disk placed in th e middle of the dish. The drop had to be small (<5mm in diameter) and well sealed at its junction with the Petri dish floor so that the wasp couldn't easily break the surface tension and drown inside the drop.

PAGE 121

103 In all experiments, a female Trichogramma was placed into a Pe tri dish with four clumps of host eggs arranged around th e central honey water drop (Figure 4-1). Figure 4-1. Petri dish design in the Trichogramma host choice experiments Basic Experimental Procedure Early in the morning, on the day of setting up an experiment, the insect cages were checked for eggs. The chiffon in the looper c ontainers and the sol uble paper in the DBM cages were removed, having been placed in the container/cages the evening before. All eggs in the experiments were less than 24 hours old. The soluble paper was carefully dissolved in water and then the water was f iltered through chiffon squares held in place above plastic funnels by elastic bands. This had to be done carefully and slowly to prevent damage to the eggs, which accumula ted on the chiffon squares. A very fine, gentle spray of clean water was used to cl ean the DBM eggs of any chemical residue from the dissolved soluble paper. The chi ffon squares were then removed from the funnels and left to dry on a flat surface. host egg clusters small honey water dro p Petri dish (100 15 mm)

PAGE 122

104 The chiffon containing the looper eggs was sprayed with a fine mist of water and then, under a microscope using a brush and a teasing needle, indi vidual eggs were carefully prized off the chiffon. The same wa s done with the DBM eggs that had been filtered out onto the chiffon squares. They we re easier to remove, and a fine brush was generally sufficient. The eggs were then placed into the Petri dishes in arrangements dependant on the specific protocol of the experiment. Once all the eggs were placed in the Petri dishes, about 100 more looper eggs were taken off the chiffon and placed into another Petri dish these eggs were later used as replacement e ggs. All Petri dishes were placed in a plastic box with the bottom lined with soaked paper kitchen towel. The lid was placed on the container, and the container stored overnight in a refrigerator (4 ± 1oC, 20 ± 5% r.h.) overnight. The following morning, the experimental Petri dishes were removed from the fridge and checked for dried looper eggs, wh ich were removed and replaced by healthy eggs. The female Trichogramma adults were then isolated in to glass vials and droplets of honey water placed in the middle of the Petri di shes ready for the experiment. Pieces of 6” ½” of Parafilm® were cut in preparation. Usi ng a dissecting microscope, the Trichogramma was teased out of their vial onto a fi ne camel hair brush (size 0) and then placed into the Petri dish at a point equidi stant from two of the egg clumps. The lid was quickly placed on the Petri dish, Parafilm® wrapped around to seal the dish, and the dish labeled. This procedure was repeated until all expe rimental Petri dishes had been assigned one female Trichogramma each. The Petri dishes were th en placed back into plastic

PAGE 123

105 container. To maintain humidity it was im portant that the paper at the bottom of the container always remained soaked. The contai ner was placed in an insect rearing room (temperature 27 3oC, 65 5% r.h. and 16L:8D) for 24 hours. After this period, the Parafilm was removed, the Petri dishes opened and Trichogramma mortality recorded. The living Trichogramma were removed from the Petri dishes, which were then closed, returned to the plastic container a nd put back into the rearing room. After 5 days, the eggs in the Petri dish es were examined for parasitism and desiccation. After 5 days, those eggs not pa rasitized had hatched and had either been eaten by the emerging larvae or stood as empt y, transparent shells. The parasitized eggs were shiny black in color and easily identifie d. The dry eggs were brown or yellow and were in various states of de siccation. The number of all t ypes of eggs was recorded for later analysis. Parasitism Preferences of Trichogramma pretiosum Parasitism preferences of Trichogramma pretiosum when offered < 24 hoursold DBM and SL eggs . The experimental protocol was as described above. In total, three experiments were conducted. Each experiment was comprised of three treatments, and 5 replicates of each treatment. The three treatments were DBM eggs only in the four clumps of eggs around the honey water droplet SL eggs only in the four clumps of eggs around the honey water droplet Two of the four clumps of eggs around the honey water droplet were DBM eggs, the other two clumps were SL eggs. The two clumps of the same host were placed opposite each other. Parasitism Preferences of Conditioned Trichogramma pretiosum Parasitism preferences of Trichogramma pretiosum that have been 'conditioned' by prior to exposure to host eggs . The Trichogramma were 'conditioned'

PAGE 124

106 by dividing the newly emerged adults into tw o groups and putting each group into a 25 dram, clear styrene tube with tabbed cap c ontaining 50 of one or the other of the host eggs (< 24 hours old). The adult Trichogramma were left in the vials for four hours. Naranjo (1993) demonstrated that female Trichogramma lay many of their eggs within the first four hours of exposure to host e ggs, so four hours was seen as the maximum period for which the adults could be left before transferal to the experimental Petri dishes. The Petri dishes containing the host eggs were prepared as described previously. Because of the time-consuming work of is olating eggs, and because of the added experimental factor; i.e., host conditioning, it was impossible to run all of the host egg configurations at the same time. Thus, each experiment had only one of three egg configurations: DBM only, SL only or a co mbination of DBM and SL eggs. In each experiment there were 6 replicate Petri dishes for each treatment. For these experiments, contro l Petri dishes were added to the protocol. These were Petri dishes with host eggs where no Trichogramma were introduced. These controls were used to give base line data for determining what proportion of egg drying was caused by Trichogramma host feeding and what proportion was due to natural causes or stress from handling. There were 5 experiments where only DBM eggs were used, 3 experiments where only SL eggs were used, and 4 experiments where there was a combination of host eggs in the Petri dishes. The number of expe riments run depended on the success of the previous experiments. Success was determined by minimal Trichogramma mortality and low levels of egg drying caused by natural causes or stress from handling.

PAGE 125

107 The number of eggs parasitize d, dried or empty due to successful larval emergence was recorded. From Petri dishes of each conditioning type, 10 parasitized eggs were placed in a vial to determine Trichogramma emergence and the sex ratio. Parasitism Preferences of 3 Trichogramma species The three Trichogramma species used for these experiments were Trichogramma pretiosum , Trichogramma minutum and Trichogrammatoidea bactrae . These wasps were obtained from Beneficial Inse ctaries and Rincon Vitova In sectaries in California. The Petri dishes containing the host eggs were prepared as described previously. Because of the time-consuming work of is olating eggs, and because of the added experimental factor, i.e., three Trichogramma species, it was impossibl e to run all of the host egg configurations at the same time. T hus, each experiment had only one of three egg configurations: DBM only, SL only or a co mbination of DBM and SL eggs. In each experiment, there were 5 replicate Petri dish es for each treatment. As with the previous set of experiments, there was control Petri dishes included in the experimental design. In total, there were 5 experiments wher e only DBM eggs were used, 3 experiments where only SL eggs were used, and 4 experi ments where there was a combination of host eggs in the Petri dishes. The number of experiments conducted depended on the success of the previous experiments. Success was determined by minimal Trichogramma mortality and low levels of egg drying cause d by natural causes or stress from handling. Number of eggs parasitized, dried or empty due to succe ssful larval emergence was recorded. From Petri dishes of each Trichogramma species 10 parasitized eggs were placed in a vial to determine Trichogramma emergence and sex ratio.

PAGE 126

108 Statistical Analysis The data were analyzed using SAS soft ware version 8.01 (SAS Institute 1999). For the percent parasitism comparisons, Proc Mixed and the LSMEANS statement was used when the data was normally distributed. When it wasn’t and transformations would not normalize the data, Proc NPAR1WAY and the Wilcoxon test was used. The transformations used to achieve normality we re Arcsine or Arcsine-root. For the dry egg counts, comparisons were made using Proc Genmod and Poisson or negative binomial distributions. All differences were cons idered significant at the P<0.05 level. Results Parasitism Preferences of Trichogramma pretiosum with the Eggs of DBM & SL In the DBM-only Petri dishes, the average percent parasitism for the three experiments was 90% ( C.I. ±5.3%, = 0.05), 86% ( C.I. ± 10.8%, = 0.05) and 87% ( C.I. ± 7.0%, = 0.05). For the SL-only Petri dishes, the average percen t parasitism for the three experiments were 7% ( C.I. ± 3.4%, = 0.05), 26% ( C.I. ± 17.0%, = 0.05) and 7% ( C.I. ± 3.1%, = 0.05). See Figure 4-2. For the Petri dishes where the host eggs were combined, similar parasitism levels for the two hosts were observed. In the firs t experiment, the average percent parasitism for DBM eggs in the Petri dishes was 75% ( C.I. ±9.3%, = 0.05) whilst for SL eggs it was 19% ( C.I. ± 11.1%, = 0.05). For the second experiment, the percentages were 69% ( C.I. ± 10.4%, = 0.05) for DBM eggs and 36% ( C.I. ± 24.0%, = 0.05) for SL eggs. In the third experiment, the respective parasitism levels were 84% ( C.I. ± 13.6%, =0.05) for DBM eggs and 20% ( C.I. ± 9.3%, =0.05) for SL eggs (Figure 4-3).

PAGE 127

109 0 10 20 30 40 50 60 70 80 90 100Experiment 1 Experiment 2 Experiment 3% parasitism DBM parasitism Looper parasitism Figure 4-2. T. pretiosum percent parasitism (separ ate hosts DBM & Looper). 0 10 20 30 40 50 60 70 80 90 100Experiment 1 Experiment 2 Experiment 3% parasitis m DBM parasitism Looper parasitism Figure 4-3. T. pretiosum percent parasitism (combined hosts DBM & Looper). When the two charts above are examined, th ere seems to be less of a difference in parasitism levels between the two hosts in th e combined Petri dishes than in the single host treatments. This difference seems to or iginate from both a greater number of SL eggs being parasitized and from a slight d ecrease in the number of DBM eggs parasitized. This is demonstrated in Figure 4-4, which show s parasitism levels of the two hostsÂ’ eggs

PAGE 128

110 in the single host species Petri dishes against parasitism levels in the combination host species Petri dishes. There is a higher percent parasitism of DBM eggs in the single host Petri dishes than in the combined host Petri dishes ( Z = 8.864, P = 0.0029). This is reversed for the SL eggs where greater pa rasitism is found in the combined host Petri dishes ( Z = 9.450, P = 0.0021). The corresponding difference in parasitism of the host speciesÂ’ eggs is also given between the tw o Petri dish types (i .e., single host or combination) and as suggested above, there is less of a difference in parasitism levels between the two hosts in the combined Petri dishes than in the single host treatments ( Z = 5.700, P = 0.0170). 0 10 20 30 40 50 60 70 80 90 100DBM eggsSL eggs% difference in parasitism% parasitism Single host Petri dishes Combination host Petri dishb b a a a b Figure 4-4. Parasitism levels of DBM and SL eggs in single host and combined host Petri dishes Figure 4-5 shows the total level of per cent parasitism in th e two single host species Petri dishes and in the combined host Petri dish, without distinction of host species. The level of parasitism depends on which host speciesÂ’ eggs are in the Petri dishes and the differences are signi ficantly different (DBM vs. DBM+SL, t = 8.07, df =

PAGE 129

111 38, P < 0.0001; DBM vs. SL, t = 16.40, df = 38, P < 0.0001; DBM+SL vs. SL, t = 8.02, df = 38, P < 0.0001). 0 10 20 30 40 50 60 70 80 90 100Petri dish type% parasitism DBM egg Petri dishes combined host egg Petri dishes SL egg Petri dishesc b a Figure 4-5. Total percent parasitism for si ngle host and combined host Petri dishes The levels of dry eggs in all experiment s were at acceptable levels of less than 20%. During protocol development, one of the greatest problems was the drying of eggs, thought to have resulted from handling da mage. In preliminary experiments, the percentage of dry eggs often exceeded 70%, resulting in there being very few healthy eggs for the Trichogramma to parasitize. An arbitrary limit for accepting or rejecting an experiment was set at 30% dry eggs. These host preference experiments confir m what was suspected from the field experiments that T. pretiosum parasitizes DBM eggs to a gr eater extent than SL eggs, whether they are in single host treatments or combined. The Effect of ‘Conditioning’ on Parasitism Preferences of Trichogramma pretiosum when offered the Eggs of DBM and SL The four hours of exposure to eggs of one or the other of the two hosts species seemed not to have any great effect o n subse quent parasitism levels in the experiments.

PAGE 130

112 a a a b a a b a a a As with the first set of experiments, higher levels of parasitism were found in the DBM eggs than in the SL eggs. Conditioning, however , did not lead to significant differences in parasitism of eggs from the same host. Single-host diamondback moth eggs Table 4-1 and Figure 4-6 show the average pe rcent parasitism for four of the five experiments where DBM or SL-conditioned Trichogramma females were released into Petri dishes that only contained DBM eggs. The third experiment was discarded because of high Trichogramma mortality. Table 4–1 gives the overall averages for each experiment, and it shows that there was no ove rall difference in pa rasitism between the two Trichogramma groups (i.e., DBM-conditioned and SL-conditioned) ( t = 0.82, df = 42, P = 0.4186). This non-significance is in line with the diff ering results found between experiments. Table 4-1. Percent parasitism of DBM eggs by ‘conditioned’ T. pretiosum females Experiment % parasitism P value DBM-conditioned Looper-conditioned 1 65% 76% 0.1050 2 82% 69% 0.0373 4 83% 69% 0.0105 5 64% 66% 0.6508 Total 74% 70% 0.4186

PAGE 131

113 a a a b a a b a a 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Expt. 1 Expt. 2 Expt. 4 Expt. 5Total DBM-conditioned Looper conditioned Figure 4-6. Percent parasitism of DBM eggs by ‘conditioned’ T. pretiosum females (note: the letters denotin g significant difference on ly apply for individual experiments) Single-host soybean looper eggs No differences in parasitism levels of SL eggs were found for the two groups of conditioned Trichogramma females. Table 4-2 and Figure 4-7 show the average percent parasitism for the three experime nts where DBM or SL-conditioned Trichogramma females were released into Petri dishes that only c ontained soybean looper eggs. Table 4-2. Percent parasitism of SL eggs by ‘conditioned’ T. pretiosum females Experiment % parasitism DBM-conditioned Looper-conditioned P value 1 33% 33% 0.7867 2 11% 8% 0.3429 3 36% 28% 0.6743 Total 27% 23% 0.6053 a a a b a b a a a a

PAGE 132

114 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Expt. 1 Expt. 2 Expt. 3Total DBM-conditioned Looper conditioned Figure 4-7. Percent parasitism of SL eggs by ‘conditioned’ T. pretiosum females (note: the letters denoting significant difference only apply for individual experiments) Combined-host Petri dish results Conditioning had no effect on the parasitism levels of the two host species’ eggs found in Petri dishes when the host species ’ eggs were combined. Both groups of conditioned Trichogramma females parasitized more DBM eggs than SL eggs. When the effect of conditioning was examined for individual host species, there was no difference in percent parasitism (DBM eggs, t = -0.14, df = 45, P = 0.8873; SL eggs, t = 1.54, df = 45, P = 0.1294). There was also no conditioning effect on total percent of eggs parasitized ( t = 0.74, df = 91, P = 0.4616) or in percent paras itism differences of the two host species ( t = -0.76, df = 44, P = 0.4520). Table 4-3 and Figure 4-8 show the results of the four experiments conducted. a a a a a a a a

PAGE 133

115 Table 4-3. Percent parasitism of co mbined DBM & SL eggs by ‘conditioned’ T. pretiosum females Figure 4-8. Percent parasitism of co mbined DBM & SL eggs by ‘conditioned’ T. pretiosum females—(note: the letters de noting significant difference only apply to paired bars in the chart: DBM eggs P =0.8873; SL eggs P = 0.1294; Total P = 0.4616 & Difference P = 0.4520). Dry eggs For each of the two conditioning treatments , the number of dry eggs for the two host species was compared using the combin ed-host Petri dishes (Figure 4-9). The number of dry DBM eggs was higher in th e two conditioning treatments than in the control treatment where no Trichogramma had been introduced (DBM-conditioned vs. Experiment DBM-conditioned Looper-conditioned DBM eggs SL eggs Total Difference DBM eggs SL eggs Total Difference 1 69% 30% 50% 39% 65% 11% 38% 54% 2 77% 35% 56% 42% 71% 28% 49% 43% 3 53% 28% 41% 25% 59% 33% 46% 26% 4 64% 62% 63% 2% 73% 56% 65% 17% Total 66% 39% 53% 27% 67% 32% 50% 35% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% DBM eggsLooper eggsTotalDifference DBM-conditioned Looper conditioned a a a a a a a a

PAGE 134

116 Control, 2 = 4.36, df = 1, P = 0.0369 & SL-conditioned vs. Control, 2 = 0.2526, df = 1, P = 0.0040). There was no significa nt difference in the number of dry SL eggs found in the conditioning treatments as compared to the controls (DBM-conditioned vs. Control, 2 = 0.23, df = 1, P = 0.6340 & SL-conditioned vs. Control, 2 = 0.63, df = 1, P = 0.4262). The conditioning of the Trichogramma had no effect on the number of dry eggs found for either host egg type (DBM eggs, 2 = 0.89, df = 1, P = 0.3448 & SL eggs, 2 = 2.05, df = 1, P = 0.1525). Figure 4-9. Number of dry eggs in the combined-host Petri dish treatments and controls Parasitism Preferences of 3 Trichogramma Species when Offered DBM and Soybean Looper Eggs Diamondback moth eggs There were no significant di fferences in percent parasi tism of DBM eggs between the three Trichogramma species ( T. bactrae vs. T. minutum , t = -0.67, df = 32, P = 0.5077; T. bactrae vs. T. pretiosum , t = -0.45, df =32, P = 0.6529 & T. minutum vs. T. pretiosum , t = 0.23, df = 32, P = 0.8170). Two of the five experiments had to be discarded because of high Trichogramma mortality. Figure 4-10 shows the percent parasitism of DBM eggs by the three Trichogramma species. dry eggs (DBM eggs)0 5 10 15 20 conditioning type# dry eggs DBM conditioned Looper conditioned Control b a a dry eggs (SL eggs) 0 5 10 15 20 conditioning type # dry eggs DBM conditioned Looper conditioned Control a a a

PAGE 135

117 % parasitism of DBM eggs0 20 40 60 80 100 Expt 1 Expt 3 Expt 5 Average% parasitis m T. pretiosum T. minutum T. bactrae a a a a a a a a aa a a Figure 4-10. Percent parasiti sm of DBM eggs by the three Trichogramma species Soybean looper eggs Trichogrammatoidea bactrae parasitized a greater number of SL eggs than the other two Trichogramma species ( T. bactrae vs. T. minutum , t = 2.70, df = 32, P = 0.0111 & T. bactrae vs. T. pretiosum , t = 2.61, df = 32, P = 0.0136). Between T. pretiosum and T. minutum , no significant difference in parasitism levels of SL eggs was found ( t = -.03, df = 32, P = 0.9802). Figure 4.11 shows the results obtained from the three experiments and the average of these experiments. % parasitism of SL eggs 0 10 20 30 40 50 60 Expt 1 Expt 3 Expt 5 Average% parasitism T. pretiosum T. minutum T. bactrae a b a a a aa a a a b b Figure 4-11. Percent parasitism of SL eggs by the three Trichogramma species.

PAGE 136

118 Diamondback moth and soybean looper eggs combined Unlike the single host Petri dish experime nts, there was significant difference in the parasitism levels of DBM eggs by the three Trichogramma species in the combined host Petri dishes (Figure 4-12). Both T. bactrae and T. pretiosum parasitized a greater number of DBM eggs than T. minutum ( T. bactrae vs. T. minutum , Z = 6.1001, P = 0.0135 & T. pretiosum vs. T. minutum , Z = 5.5744, P = 0.0182). There were also differences in parasiti sm of SL eggs with T. bactrae parasitizing more SL eggs than T. minutum ( Z = 7.9165, P = 0.0049). There were no other significant differences recorded between pairs parasitizing SL eggs. These differences in parasitism levels ar e reflected in the total number of eggs parasitized in the combined Petri dish experiments. Here both T. bactrae and T. minutum parasitized a greater number of eggs than T. minutum ( T. bactrae vs. T. minutum , 2 = 15.29, df =1, P <0.0001 & T. pretiosum vs. T. minutum , 2 = 6.03, df = 1, P = 0.0140). No significant difference was found in the per cent parasitism differences of the two host species in the three Trichogramma species ( T. bactrae vs. T. minutum , t = -0.09, df = 49, P = 0.9299; T. bactrae vs. T. pretiosum , t = -1.24, df = 48, P = 0.2199 & T. minutum vs. T.pretiosum , t = -1.14, df = 48, P = 0.2594). The percent parasitism difference was calculated by subtracting the mean percent para sitism of SL eggs from the mean percent parasitism of DBM eggs

PAGE 137

119 0.0 20.0 40.0 60.0 80.0 100.0 DBMSLTotal% diff% parasitism T. bactrae T. pretiosum T. minutuma aa a b b a a aaa b b Figure 4-12. Percent parasitism of combined DBM & SL eggs by the three Trichogramma species (note: the letters de noting significant difference only apply to triplet bars in the chart). Although no significant differen ces were found when compar ing percent parasitism difference of each host species for the three Trichogramma species, T. pretiosum did show the highest differentiation in parasitism levels between the two host species eggs. This is shown by T. pretiosum being the only Trichogramma species to show a significant difference in percent parasitism between the two host species ( T. pretiosum , t = 3.07, df = 16, P = 0.0073) (Figure 4-13). Although ther e were no significant differences in parasitism of the two host species by T. bactrae and T. minutum, there was consistently higher parasitism of DBM eggs over SL eggs (T. bactrae , t = 1.55, df = 18, P = 0.1378 & T. minutum, t = 1.93, df = 17, P = 0.0703).

PAGE 138

120 0.0 20.0 40.0 60.0 80.0 100.0 T. bactraeT. pretiosumT. minutum% parasitism DBM eggs SL eggs b a a a a a Figure 4-13. Percent parasitis m means of DBM and SL eggs in the combined-host Petri dishes by the three Trichogramma species (note: the letters denoting significant differences only apply to the paired bars in the chart). Dry eggs The number of dry eggs found in the comb ined Petri dish experiments mirrored those found in the combined-host Petri dish es for the conditioning experiments. The number of dry DBM eggs was higher in the Trichogramma treatments than in the control ( T. bactrae , 2 = 11.63, df = 1, P = 0.0006; T. pretiosum , 2 = 17.79, df = 1, P < 0.0001, T. minutum , 2 = 14.99, df = 1, P = 0.0001), whilst for the SL dry eggs, there was no difference between the experimental treatments and the control, ( T. bactrae , 2 = 2.43, df = 1, P = 0.1189; T. pretiosum , 2 = 2.00, df = 1, P = 0.1569, T. minutum , 2 = 0.01, df = 1, P = 0.9254) (Figure 4.14). There was no difference in the numbers of dry eggs amongst the Trichogramma species treatments for DBM eggs ( T. bactrae vs. T. minutum , 2 = 0.37, df =1, P = 0.5403; T. bactrae vs. T. pretiosum , 2 = 1.18, df = 1, P = 0.2783 & T. minutum vs. T. pretiosum , 2 = 0.23, df =1, P = 0.6349) or for SL eggs ( T. bactrae vs. T. minutum , 2 =

PAGE 139

121 2.27, df = 1, P = 0.1317; T. bactrae vs. T. pretiosum , 2 = 0.01, df = 1, P = 0.9167 & T. minutum vs. T. pretiosum , 2 = 1.85, df = 1, P = 0.1737). Figure 4-14. Number of dry DBM and SL e ggs found in the combined host Petri dishes of the three Trichogramma species experiments. Trichogramma emergence For the conditioning experiments and the Trichogramma species experiments, Trichogramma emergence data was recorded. Percen t emergence (percent of eggs where a healthy Trichogramma adult emerged), average number of Trichogramma adults emerging per host egg and sex ratio were record ed. The sex ratio data was discarded. For the Trichogramma pretiosum the sex ratio was about 1:1. For the other Trichogramma, species there were nearly 1 female to 20 ma les in the emerged F1 generation. These sex ratios were similar to the sex ratios of the parental generation that emerged from the egg cards. The Trichogramma supply companies were contac ted about these extreme sex ratios—this ratio of males emerging would ruin any biological control effort. The explanation was that the Trichogramma being shipped for these experiments were what the insectaries had to maintain colonies ove r the winter in Canada and California. The companies were not taking the same care with the colonies because their operations were dry eggs (DBM eggs)0 10 20 30 40 trichogramma species or control# dry eggs T. bactrae T. pretiosum T. minutum Control b a a a dry eggs (SL eggs)0 10 20 30 40 trichogramma species or control# dry eggs T. bactrae T. pretiosum T. minutum Control a a a a

PAGE 140

122 not geared for commercial supply at this ti me. For this reason, the sex ratio data was discarded as it was seen more as an artifact of the rearing conditions in the insectaries than any experimental effect. Tables 4-4 & 4-5 show the emergen ce percentages and average number of emerged adults per egg for the conditioning and Trichogramma species experiments. The three Trichogramma species expressed solitary behavi or with DBM eggs and gregarious behavior with SL eggs. Although the Trichogramma were gregarious with SL eggs, it was more common for only one adult to emerge per SL egg. The highest average number of Trichogramma adults emerging from SL eggs was 1.37, which represents two adults emerging in one out of every three eggs. Table 4-4. Percent emergence and average number of Trichogramma adults emerging per egg for the conditioning experiment s (for single-host and combined-host Petri dishes). Given too is the standard deviation (SD). Host type DBM eggs SL eggs DBM cond. SLcond. DBM cond. SLcond. Single host % emergence 97% ± 18% 79% ± 18% 93% ± 17% 93% ± 11% ave # emerged trich. 1 ± 0 1± 0 1.08 ± 0.3 1.2 ± 0.4 Combined host % emergence 65% ± 48% 88% ± 33% 78% ± 42% 88% ± 33% ave # emerged trich. 1 ± 0 1 ± 0 1.1 ± 0.3 1.2 ± 0.41

PAGE 141

123 Table 4-5. Percent emergence and average number of Trichogramma adults emerging per egg for the Trichogramma species experiments (for single-host and combined-host Petri dishes). Given t oo is the standard deviation (SD). Discussion Host preference Under the laboratory conditions and prot ocols used for these experiments, Trichogramma pretiosum parasitized DBM eggs to a much greater level than the eggs of SL. These parasitism levels were also found when the host species’ eggs were combined, representing a marked preference for DBM eggs. In the combined-host Petri dishes, the level of parasitism of DBM eggs decreased whilst the parasitism level of SL eggs increased. It seems that the presence of DBM eggs stimulated the parasitism of SL eggs, whilst the reverse was true for the parasitism of DBM eggs. With only half the number of DBM eggs present in the combined treatments as compared to the single host treatments and with the strong preference for DBM eggs, one might expect the percent parasitism of DBM eggs to increase in the combined treatments (and the reverse for the SL eggs). It was somewhat surprising that SL was markedly not the preferred host in these experiments. Godin and Boivin (2000) showed that for many strains of Trichogramma , Trichoplusia ni was a highly suitable host specie s for parasitoid development, as expressed by progeny survival. One possible reason for the preference observed is the T. pretiosum T. minutum T. bactrae Host type DBM eggs SL eggs DBM eggs SL eggs DBM eggs SL eggs Single host % emergence 73% ± 45% 87% ± 35% 100% ± 0% 90% ± 31% 100% ± 0% 96% ± 19% ave # emerged trich. 1 ± 0 1.08 ± 0.27 1 ± 0 1.37 ± 0.56 1 ± 0 1.33 ± 0.48 Combined host % emergence 64% ± 49% 91% ± 29% 71% ± 46% 87% ± 34% 93% ± 27% 95% ± 22% ave # emerged trich. 1 ± 0 1.19 ± 0.40 1 ± 0 1.04 ± 0.19 1 ± 0 1.18 ± 0.39

PAGE 142

124 theory that Trichogramma reared on smaller-sized host eggs may be predisposed to parasitize equally small eggs, such as DBM eggs in this case. The Trichogramma in these experiments were reared on Ephestia kuehniella eggs. Ephestia kuehniella eggs are considered small hosts at around 0.5 0.3 0.3mm in measurement (Schmidt 1994). DBM eggs are similar in size to E. kuehniella eggs, possibly smaller at 0.44mm long and 0.26 mm wide. Nurindah et al. (1999) found that Trichogramma australicum reared on Sitotroga cerealella eggs (0.23 mm long and 0.19 mm wide ) attempted to drill smaller host model eggs than those reared from the larger Helicoverpa armigera (diameter 0.56 mm, volume 0.05 l). The former, accepted glass b eads of the size range 0.75–0.50 mm, whereas the latter accepted glass beads from 0.75 mm to 1.5 mm in diameter. The researchers suggested that cues from the rear ing host eggs could have been picked up at adult emergence, when the neonate females walked over and antennated the eggs they had just emerged from. The researchers also recognized that the Trichogramma could have been exposed to host cues before emergence as other researchers have found (Cortesero and Monge 1994, Bjorksten and Ho ffmann 1995). Whatever the mechanism, it could be argued that DBM eggs were the pref erred host in our experiments because of the host cues picked up from the sim ilarly sized hosts from which the T. pretiosum were reared. Number of adults emerging from host eggs Clutch size in a gregarious parasitoid depends on the female’s assessment of host size, nutritional suitabilit y, previous parasitization ev ents and other host-related characteristics (Schmidt 1994). Solitary pa rasitoids only have to make the decision whether to oviposit into an egg or not, usi ng the host suitability criteria mentioned above. The Trichogramma performed as solitary parasitoids with DBM eggs whereas with the

PAGE 143

125 SL eggs they showed gregarious behavior . This was similar to the study by Godin and Boivin (2000) where the Trichogrammatid species studied behaved as gregarious parasitoids for Trichoplusia ni (Hübner) (Noctuidae) and Pieris rapae (L.) (Pieridae) and as solitary parasitoids for DBM. In thei r study, they analyzed progeny allocation for 23 strains of Trichogrammatids in three host species, P. xylostella T. ni and P. rapae . In T. ni eggs, they found that the av erage clutch size varied between 1.14 and 2.98 per host. For P. rapae eggs, the average clutch size vari ed between 1.75 and 3.14. What they also found was that clutch size tended to decrease wi th host age. This was particularly true for T. ni eggs. The results from our study showed that clutch size in SL eggs was on the low side when compared to the T. ni of the aforementioned study, es pecially because the eggs used in this present study were very young and supposedly a more appropriate host for larger clutch sizes. Dry host eggs The conditioning experiments and the Trichogramma species experiments showed that the activity of the Trichogramma led to a greater number of dry DBM eggs. There was no difference found, however, in the numbe r of dry DBM eggs produced by the three Trichogramma species, contrary to the work of others (Vasquez et al. 1997). The number of dry eggs in the SL treatments was no different from the controls. Vasquez et al. (1997) used the term “direct para sitoid induced mortality” to describe host feeding and other parasitoid activities such as probing with out oviposition. This can be a very significant mort ality factor. In the study by Vasquez et al., T. minutum recorded host mortality figures greater than 90%, but only one third of this mortality was caused by parasitism. Host feeding has been well doc umented (Jervis and Kidd 1986, Jervis et al . 1996, McGregor 1997, Ueno 1997). It is consid ered particularly important with

PAGE 144

126 inundative biological control releases where ma ximum mortality of the host is achieved by the parasitoids being released and not by future generations of parasitoids. Why then, was there no signi ficant “direct parasitoid i nduced mortality” recorded in the SL eggs? It co uld be that when the Trichogramma females probed SL eggs, egg laying was always the result (sign of a very ac ceptable host or of one that is rejected before the probing stage of oviposition has been reached). Alternatively, it could be that the Trichogramma did not initiate host feeding on SL eggs. Perhaps after one or two host feeding attempts, the Trichogramma refuse to approach the SL eggs for host feeding. This may relate to the internal chemical com position of SL eggs. If this were the case, could internal chemical conditions be th e reason why SL was a less preferred host compared to DBM? Conditioning of T. pretiosum The conditioning treatment under which the Trichogramma adults were placed had no effect on subsequent parasitism levels of either DBM or SL eggs. The reason why the Trichogramma did not seem to ‘learn’ and impr ove the parasitism rates of the host species, or show a preference for one host ove r the other maybe due to the design of the experiment. The test duration, 24 hour s, is a relatively long time for Trichogramma , especially if it is the first 24 hours of their life. Naranjo ( 1993) showed that most of the parasitism occurs in the first few hours. Du ring the first few hours there may have been differences in how the ‘conditioned’ Trichogramma approached the host eggs. Perhaps acceptance and handling of the host eggs did im prove, leading to quicker parasitism that went unnoticed with this experimental design. It must be noted however that any increase in the speed of parasitism that may ha ve resulted from improved handling through Trichogramma ’s prior host experiences, did not lead to an increased number of

PAGE 145

127 parasitized eggs for that host, when compared to Trichogramma who had been conditioned to the other hostÂ’s egg. The conditioning of Trichogramma adults did not change the basic host preferences either. As mentioned in Chapter 2, a parasitoid that became more adept at handling the eggs of one species through prio r exposure should not be any less willing to accept and parasitize the eggs of another host sp ecies. It would not confer a reproductive advantage. In conclusion, conditioning in thes e experiments, did not lead to a change in the absolute number of eggs laid, or to a change in how the eggs were divided up amongst the two hosts. To determine eviden ce of types of conditioning/learning and improvements in handling times, these experime nts should be repeated with smaller time intervals. Host preferences by the three species of Trichogramma All three of the Trichogramma species showed the same levels of parasitism of single-host DBM eggs, this being perhaps becaus e of the large fluctuations in parasitism levels between experiments. Fo r single-host SL eggs however, T. bactrae did parasitized significantly more eggs th an the other two species. The combined-host experiments gave different results to the single-host experiments for DBM parasitism. Both T. bactrae and T. pretiosum parasitized a greater number of DBM eggs than T. minutum . For the SL eggs the only significant difference in parasitism levels was that T. bactrae parasitized more eggs than T. minutum . One interesting result was that, although all three species of Trichogramma parasitized more DBM eggs than SL eggs in the combined-host Petri dishes, only T. pretiosum showed significantly different parasitism levels of the two host speciesÂ’ eggs. Perhaps this lack of statistical difference is related to the phenomenon discussed in the T. pretiosum host

PAGE 146

128 preference experiments, whereby in the combin ed-host Petri dishes, the parasitism levels of the DBM eggs goes down while the parasi tism levels of the SL eggs goes up (as a result of exposure to the eggs of both host species), when compared to the single-host Petri dishes. Maybe the phenomenon is greater for T. bactrae and T. minutum resulting in there being no significant differe nce. Alternatively, it could be that the innate preference for DBM over SL is less marked in these two species. T. bactrae parasitized relatively high numbers of both species while for T. minutum , relatively low numbers of both hostsÂ’ eggs were parasitized. These experiments demonstrate that the Trichogramma species examined are potentially good parasitoids of the diamondback moth. Not only do they parasitize a high percentage of DBM eggs but their host-feeding or probing activities al so lead to further DBM egg mortality. An added be nefit is that these parasiti sm levels and host feeding activities are maintained even in a multiple -host situation with soybean looper which, using the criteria of parasiti sm levels and egg mortality, do es not represent as good a host for the Trichogramma species examined.

PAGE 147

129 CHAPTER 5 CHARACTERIZING A FARMING SYS TEM IN THE CENTRAL MOUNTAIN REGION OF PUERTO RICO: IPM COMPATIBLE? Introduction One of the main difficulties confronting the implementation of new integrated crop protection technologies is thei r successful on-farm transfer al. Often the technologies do not easily fit into the farming system, perhap s, for example, because they place excessive constraints on resources such as time and la bor. Control mechanisms in integrated pest management (IPM) programs are often not so immediate or obvious in their effect as those of synthetic insecticides, and this ma kes it difficult to engender confidence in the new techniques. How have these problems arisen ? Part of the problem has been a lack of understanding and appreciation by those that produce the new IPM technologies of the agro-ecological, cultural and socio-economic milieu in which the farmers work. In many cases, this has led to the excl usion of the farmer as both collaborator a nd beneficiary (Matteson et al. 1984). It is, perhaps, just as important to address this problem as it is to determine the biological parameters and impact of a biological agent such as Trichogramma, for example. The non-acceptance of IPM packages may not ju st be due to a lack of compatibility with the farming system, but also may be due to the reluctance on the part of farmers to abandon existing pest control practices. Economic realities ensure that many farmers are not willing to risk yields (and their subsis tence) by giving up the chemicals that have proven so successful at killi ng insect pests. This is ex acerbated, in some countries,

PAGE 148

130 because pesticide regulations are minimal and those regulations that do exist are poorly enforced. There is little incentive for farmers to change practices. So, how then does one ensure that IPM technologies have the best chance of adoption? The first step is to understand the parameters (i.e., the resources needed, the potential outputs and the i nherent limitations) of the relevant IPM strategy under consideration. From an economic perspectiv e, the costs of control practices are incorporated into economic thresholds. Fo r pesticide-based control programs, the calculation of costs is relati vely easy. For an IPM program, the calculations are more complicated because of the array of practices used. In addition to the economic costs, there should be some appraisa l of labor requirements. This includes information such as amount of labor used, levels of expertise required, (is there somebody specifically who will be doing the work), frequency of the work needed, and when in the year it all takes place. There are also subtler evaluations to be made, such as whether or not IPM practices have synergistic interactions with other elements of the farming system? Can crop residues of one crop be used as mu lch for another crop to improve the crop microclimate and generalist predator numbers? Can the scouting for one insect include the scouting for other ins ect pests on other crops? The second step in improving IPM adoption is to understand the characteristics of the farming system in which the IPM stra tegy is to be placed. Booij and Noorlander (1992), whilst examining a farming system, id entified three main cat egories of factors that affect predator occurrence at the farm le vel. The first category in cluded ‘crop effects’ and included the crop type. Each crop has its own structure, prey availability and husbandry practices. The second category, ‘system effects’ concerns th e general approach

PAGE 149

131 of farm management. This includes practices such as pesticide use, fertilization, soil tillage, weed management, etc. The final category comprised th e spatial pattern of fields, crops, field margins, hedges and the natu ral environment. This ‘agroecological infrastructure’ seems to be crucial to the abundance and recovery of many fieldinhabiting species. As developers of new IPM strategies we should be aware of how our strategies may interact with th e farming environment in question. If we understand the importance of charac terizing the farming system, how then does one proceed? One example is outlined by Ikerd et al. (1996), who evaluated two farming systems in Missouri for level of su stainability using environmental, economic and social criteria to evaluate the systems. They surmised that while not all logical environmental, economic and social criteria could be realistically assessed, it was important that explicit consideration was gi ven to all three areas. A well-planned Farming Systems Research (FSR) program will evaluate these domains of the farming system and will examine them at all category levels as defined by Booij and Noorlander. As mentioned in Chapter 2, linear program (LP) modeling can be a useful way of conjoining large amounts of data that then ca n be used to characterize the system under investigation. Bakker et al. (1998) point out that linear progra mming is most valuable in its ability to define priorities for agricultura l development. It does so by assessing where the development efforts could have the grea test impact on set object ives in light of existing constraints. “Linear Programming (LP) is a method for analyzing family farm livelihood systems by determining a combination of farm and non-farm activities that is feasible given a set of fixed farm constr aints and that maximizes (or minimizes) a particular objective or family goal” (Hild ebrand 1998). Using such systems analysis, an

PAGE 150

132 IPM program may be tailored to become a s ite-specific solution that encompasses the strengths and weaknesses of the particular fa rming system. Before this can be achieved, the information with which to build the model needs to be obtained. The purpose of the work conducted in the th ree municipalities of the central region of Puerto Rico was to form a better unde rstanding of how the farms function and how IPM practices could be incor porated. Specifically, the objective was to see how cabbage might be grown on these farms using an IPM methodology for the control of diamondback moth that included the use of Trichogramma . As discussed in Chapter 1, these mountain farms were never developed or supported in the same way as the larger farms of the coastal areas, a nd both the industrialization and agricultural intensification programs of the last century have had little positive imp act. They still remain key, however, to the identity and livelihoods of th e people who live in this region of Puerto Rico. It is for this reason that methods shoul d be identified that could help characterize the farming systems present and help prom ote sustainable development within these areas. Methods This work was not meant to be a compre hensive FSR study as it lacks many of the necessary components. It is how ever, an attempt to evaluate how FSR could help direct agricultural research and whether LP modeling could be a useful tool in this regard. Identifying the Study Farming System The first task was to identify the municipalities where the work was to be done. Barranquitas, Naranjito and Orocovis were eventu ally chosen because, in years past, they had collectively been the center of the islandÂ’s cabbage production (Agricultural Experimental Station, University of Puerto Rico 1999). With their relative proximity to

PAGE 151

133 the metropolitan area of San Juan, and with the benefits of the cooler mountain climate, they had become important suppliers of ca bbage to the local market. Most of the production of the cabbage was during the early months of the year, when North American production was restricted by the winter cold. Other re asons for choosing these municipalities were the homogene ity of the farming systems, the fact that family farms were the dominant farm type and the fact that the farms had a relatively low resource base as compared to the large coastal farm s. They continue with more ‘traditional’ farming practices such as the use of bulls for preparing the land. Once the three municipalities had been chosen, it was necessary to choose the farmers with whom to collaborate. With th e help of research personnel, the three extension agents of the region were contact ed and preliminary visits were made. To explain the objectives of the study a workshop wa s held with the extension agents. One of the objectives of the workshop was to determ ine whether the farms identified could be considered part of the same ‘recommendation domain’ (RD), or not. The RD concept was first developed by the Internat ional Maize and Wheat Improve ment Center (CIMMYT) in the 1970s and it defined an RD to be “a group of farmers operating the same system and for whom the same new technologies would be appropriate” (Collinson 2000). For the LP model to work, the information inputted needs to be from one RD to ensure accuracy and relevance. One of the identified farms in Oroc ovis was larger than the rest of the farms (~150 acres as opposed to usually less than 100 acres) and the principal crop was coffee. This set it apart from the other farms. Neve rtheless, it was include d because other crops were grown on the farm and, as big as it was, it was run by one family and had similar constraints and available resources.

PAGE 152

134 The extension agents chose six farmers from each municipality and letters were written to explain the work a nd to invite the farmers to a workshop. Although the farmer workshop was principally designed to in troduce the study, it also incorporated presentations on the principle pests of cabbage , the natural enemies of these pests and the new insecticides available for control in ca bbage. At the end of the farmer selection process there were 16 farmers who had agreed to participate in the study. Questionnaire Development Because of the large amount of data need ed, three questionnaires were prepared and three farm visits made to each farmer. The first questionnaire covered general farm data, the second covered the main agricultura l activities conducted on the farm and the third questionnaire pertained to the economics of the farm and family. The questionnaires can be found in Appendices B, C and D. The questionnaires were prepared in Spanish and checked for content and language by Dr. Hildebrand, University of Florida, and personnel at the agricultural experimental station in Río Piedras, Puerto Rico. The questionnaires then had to be translated in to English and sent to the University of Florida’s Institutional Review Board for evaluation and clearance. Interviews Almost all the interviews were conducte d with a Spanish-speaking colleague who assisted by taking down the farmers’ responses to the questions asked by the researcher. All interviews were conducted on the farms, wh ich gave the researcher an opportunity to walk the farms and meet the farmers’ families. On occasion, the extension agent for the municipality participated. Th e interviews took from between one hour to two and a half hours to complete. Often the farmer wa s assisted by his wife and family.

PAGE 153

135 Additional data was obtained through soci ologists, extension agents, agronomists and economists at the agricultural research stations of the University of Puerto Rico. Also approached were officials of the Puerto Ri can Department of Agriculture and USDA for information on subsidies and incentive progr ams. Information gathered for individual crops was compared with the technologica l packages produced by the University of Puerto Rico for each crop. In addition to technical data, time was spent gathering information on the economic and political development of Puerto Rico, and specifically for the central region of the island. All asp ects of life on the isla nd are affected by its association with the United States of America and by its status as th e oldest colony in the world. Chapter 1 covers the historical de velopment of the island and the impact of development programs of the 20th century. Linear Programming A linear programming model was develope d using the data collected from the questionnaires and other sources. The model maximized the discretionary cash produced at the end of the year after all basic family needs were satisfied. In the making of the model, several assumptions had to be made and activities and constraints defined: The year was divided into quartiles (Ja nuary-March, April-June, July-August and September-December) to reflect the different periods of the year. Christmas time is a time of elevated sales while the summer months are a time of little activity, and the model was designed to assimilate these differences. The tropical climate is such that many di fferent crops can be, and are, grown in these mountain farms. The crops included in the model were restricted to those crops that were of economic importance to the farms of the central region. Thirteen crops were eventually chosen (Table 5-1) . Many of the crops could potentially be grown at any time of the year. Crop appearan ces in the model were restricted to the times of the year when the crops were most commonly grown. Farmers accepted and used the governme nt incentives available to them.

PAGE 154

136 The work was not divided up between the farmer and the members of his family. A common family labor pool was used to complete the tasks associated with the activities. The labor available is a limite d resource and the unit of measurement was a day of work that consisted of 8 hours. Additional labor is available at a cost of $30 a day per laborer. Land is also a limiting resource. The farmer had savings available to hi m at the beginning of each quartile. The quantity of the money available was calcula ted based on size of farm and size of the workforce. The savings had to be recuperated by the end of the year. This money was carried over to the next year a nd made available as before. This gave the system the flexibility to chose when savings could be best used. Table 5-1. The crops incl uded in the LP model. Crop name English Crop name Spanish Scientific name Bananas Guineos Musa sapientum Beans (kidney, string) Habichuelas Phaseolus vulgaris Cabbage Repollo Brassica oleracea var. capitata Cassava Yuca Manihot esculenta Root celery, Celeriac Apio, Arracacha Arracacia xanthorrhiza Chayote Chayote Sechium edule Coffee Café Coffea spp. Ginger Jengibre Zingiber officinale Papaya, Pawpaw Papaya Carica papaya Plantains Plátanos Musa paradisiaca Pumpkin Calabaza Cucurbita moschata Taniers Yautia Xanthosoma spp. Yam Ñame Dioscorea spp. The linear model was constr ucted using Microsoft® Excel 2000 Professional and Microsoft® Visual Basic. The principal matrix was supported and defined by a number of input and output tables th at were also Excel documents. The solver used was the

PAGE 155

137 Premium Solver Platform with the XPRESS solver engine from Frontline Systems (Nevada, USA). Determining the Parameters of an IPM Co ntrol Strategy for Diamondback Moth in Cabbage Part of this work was to see how a novel crop protection strategy might be incorporated into an existing fa rming system found in the central region of Puerto Rico. To extend the entomological resear ch conducted in this doctoral work, it was decided to use the egg parasitoid Trichogramma (Hymenoptera: Chalcidoidea) in an IPM program against the diamondback moth (DBM) ( Plutella xylostella ; Lepidoptera: Plutellidae). The IPM program would include the use of scouting for insecticide application decisions and the use of IPM-compatible insecticides. The experiment was conducted on the farm of Sr. Ruben Ortiz, an experienced farmer from Orocovis who had grown cabbage for many years. It was agreed that he would plant and manage a quarter acre of cabbage using his normal production methods, the only difference being that the control program for DBM would follow a predetermined IPM methodology. Sr. Ortiz was given a notebook that was used as a diary to log all activities pertaining to the cabbage cr op. This included both the costs of materials and time taken for the work completed. Two thousand cabbage seeds were planted in seedbeds close to the house (Figure 51) on the 3rd of May, 2002. Half of the seeds were the Blue Vantage (Sakata Seed America, Inc.) variety and th e other half were the Río Verde (Syngenta Seeds, Inc.) variety. The planting medium was a mixture of soil from the farm and river sand, a technique that he had developed to reduce the damage to the roots at transplant. Because of the sandy nature of the soil, he watere d the seedbeds once a day. Seven times during

PAGE 156

138 the month that the plants were in the see dbed, he included one teaspoon of the fertilizer 20-20-20 to his spray. Weekly ap plications of Dipel DF, ( Bacillus thuringensis v kurstaki , Valent Biosciences Corporation) (1 teaspoon/gallon) was applied to the seedbeds to prevent DBM damage. While the seedlings were developing in the seedbeds, the land was prepared for transplant. Ve getation was sprayed and cut back using herbicides and machete. 750 lbs of calcium carb onate were applied to the land to increase pH. and later the soil plowed over using a pair of oxen. Banks were plowed into the field to help with drainage, to reduce soil humid ity and the threat of soil-borne diseases. Figure 5-1. Seedbeds contai ning the cabbage seedlings The cabbage seedlings were transplanted over a three-week period in June. One half of the field was transplanted with Blue Vantage transplants a nd the other half was planted with Río Verde transplants. The staggered planting was his normal practice, which, apart from dividing up the work, allowe d him to have cabbage ready for market for over a month. A handful of 8-8-12 (NPK) fertilizer was applied to every plant at transplant. One other application of this fe rtilizer was made dur ing the crop’s growth.

PAGE 157

139 Unlike other farmers, he did not designate a specific time for w eeding, although he did pull weeds up when he came across them during his almost daily visits to the cabbage. IPM Strategy. Imidacloprid (Admire 240, Bayer) was applied as a drench soon after transplant (1 teaspoon per 4 gallons of water). This insecticide was used to prevent root aphid damage (species unknown), a very re cent and very serious pest of cabbage in Orocovis. Admire 240 was not part of the fa rmer’s regular crop protection strategy. The Admire 240 was also applied to prevent th e build up of whitefly later on in the crop. The scouting method used was one developed by Dr. Leibee of the University of Florida. The sampling locations were determ ined by mentally dividing the field up into quarters and placing a marker flag at the centr e of each quarter. A fifth flag was placed in the middle of the field. Each week Sr. Ortiz scouted for DBM at each of the five marker flags. At each flag, four plan ts were randomly chosen and examined for DBM. Only the bud/head and the next four youngest leaves fo r each plant were examined. As the plant got older, this area of the pl ant began to represent the market able part of the plant. Two measurements were made for each plant. The first was the presence or absence of any of the life stages of the pest insect (i.e., DBM) and the sec ond was the presence or absence of DBM damage to the bud or head. In this case, the bud is represen ted by the vertical leaves and those that face inward. Each meas urement was recorded as a ‘+’ for presence and ‘-’ for absence. At the end of scouting th ere were 20 measurements (‘+’ or ‘-’ for each criteria), which could be used to give percentage DBM presence or DBM damage in the cabbage. Control decisions could be made using this information. The whole scouting exercise took no more than 15 minutes.

PAGE 158

140 Dipel DF was applied on a weekly basis as part of the IPM st rategy to control the DBM populations. This was one of the main in secticides used by Sr. Ortiz and he was happy with its effectiveness. The idea was to only use this ins ecticide against DBM during the life of the crop, and to use it at a lower frequency than normal. Sr. OrtizÂ’s normal management practice was to appl y Dipel DF once every 5 days. Other insecticides used by Sr. Ortiz and su rrounding farmers against DBM were Ambush (permethrin, Syngenta Crop Protection, Inc), Orthene (acephate, Ar vesta Corporation), Agree WG ( Bacillus thuringensis v aizawai , Certis USA), Monitor (O,S-Dimethyl phosphoramidothioate, organophosphate, Bayer) and Diazinon (diazinon, organophosphate, Syngenta). The Dipel was ap plied at two teaspoons (10 ml) per gallon and ten gallons were applied to the crop foliage. This worked out to be nearly a pound per acre. A regular knapsack sprayer was used to apply the insecticide mix, which also contained two teaspoons of Vel, a local washin g-up liquid that helped to spread the mix. Sr. Ortiz normally applied the insecticide la te in the afternoon. Based on scouting data, the weekly application of Dipel DF was to be increased to once every 5 days if DBM damage was found in over 30% of the plants. The T. pretiosum used in this experiment was obtain ed from Beneficial Insectaries, which sent the Trichogramma egg cards on a weekly basis. The factitious host eggs used were Ephestia kuehniella Zeller (Lepidoptera: Pyralidae) eggs. The wasps were flown overnight to Puerto Rico and were then take n the next day to Sr. OrtizÂ’s farm. A weekly delivery of the Trichogramma was planned. On receipt of the Trichogramma, Sr. Ortiz cut each egg card square into two and placed each piece into a diet cup that was then sealed with a lid. In this wa y, ten diet cups were prepare d. Based on previous calculations

PAGE 159

141 of Trichogramma emergence, 1,500–1,600 Trichogramma would be expected to emerge from each of the diet cups. This gave a rel ease rate of up to 64,000 wasps per acre. At the beginning of the experiment, after the met hods had been explained, Sr. Ortiz took the initiative and prepared release points for the Trichogramma . With wood, a zinc panel found on the farm, plastic cups and tape he made ten release stands for the Trichogramma (Figure 5-2). He had also t hought of placing holes into the bottom of the cups for water drainage. Figure 5-2. Trichogramma release point and scouting flag in recently transplanted cabbage On the day that the Trichogramma adults began to emerge, Sr. Ortiz took the Trichogramma to the field during late afternoon, and released them by taking off the diet cup tops and placing the cup and its top into the plastic cups attached to the release points. An advantage of the location of the experiment was that the field was sandwiched by two stands of banana plants, which reduced wind effects. The Trichogramma were

PAGE 160

142 normally released on a Monday or Tuesday, and Dipel DF applications were made later in the week. There was at least 2 days between Trichogramma release and application of the insecticide. The field plan for this experiment, with the Trichogramma release points and scouting positions, is shown in Figure 5-3. As ide from the varietal division of the cabbage planted, there was also a division ma de based on time of planting, with the older plants being found at the we stern edge of the field. Figure 5-3. Field plan fo r on-farm IPM experiment Although the agreement with the farmer wa s that he would sell whatever cabbage was produced, and that we were trying to produc e cabbage of the highest quality, this was not the objective of the expe riment. The principle objective was not to see whether this particular DBM control strategy worked or not (cabbage is not normally grown in the mountains at this time of year) but was to reco rd the requirements of the strategy in terms N Blue Vantage Río Verde First planting Second planting Scouting flag Trichogramma release point KEY

PAGE 161

143 of costs and labor. The work was done to collec t real data for use in the development of an alternative cabbage activity that could be placed and assessed in the LP model. Results System Characteristics Crop budgets for each of the 13 crops were developed from the questionnaire data and technical publications from the research station of the University of Puerto Rico (Table 5-2). The costs and incomes for both pl antain and banana were divided into two the cost and incomes of the first year and th e cost and incomes of the subsequent years (i.e., Year 1 and Year 2+). Plantain normally is left to produce for three to five years before being removed, whereas well-maintained banana is left for more than ten years. Chayote is left in production for around six years before being replanted. Coffee trees, on these farms, are usually kept in production for more than twenty years. Figure 5-4 represents how the overall production costs of one acre of each crop compares and also shows the breakdown of these costs. The costs are either the cost of an entire cropping cycle for an annual crop or the yearly produc tion costs of a biannual or perennial crop. For example, coffee is broken down into th e first year start-up costs and then, independently, the annual costs associated w ith the maintenance and harvest of mature coffee trees. Figure 5-5 is a similar represen tation, but it symbolizes the comparison of overall labor requirements and the breakdown of these requirements for each crop. Table 5-3 gives a profile of each of the fa rms studied in this work. As mentioned previously, apart from farm #3 (the large, coffee-dominated farm), the other farms are very similar in most respects. Table 5-3 gi ves a breakdown of the 16 farms detailing the area under cultivation, labor sources availabl e, predominant farming activities and approximate annual incomes.

PAGE 162

144 Table 5-2. Labor, costs and income per acre for the modelÂ’s crops. Crop activity Labor (days) Cash input ($) Income ($) Net income ($) Banana (Yr 1) 41 $1,291 $1,600 $310 Banana (Yr 2+) 24 $848 $1,600 $752 Beans 26 $924 $1,226* $302* Cabbage 59 $1,888 $3,380 $1,492 Cassava 54 $916 $2,400 $1,484 Root celery, Celeriac 37 $1,109 $2,700 $1,591 Chayote 371** $15,811** $36,648** $20,837** Coffee 83+ $2,582+ $5,400+ $2,818+ Ginger 59 $1,315 $3,500 $2.185 Papaya, Pawpaw 78 $3,175 $7,560 $4,385 Plantains (Yr 1) 46 $2,087 $4,080 $1,993 Plantains (Yr 2+) 17 $1,471 $2,400 $929 Pumpkin 34 $1,170 $2,000 $830 Taniers 32 $869 $2,400 $1,531 Yam 43 $1,478 $4,800 $3,322 Note: (* half sold shelled, other half sold in their pods, ** over 6 years, + over 8 years)

PAGE 163

145 Figure 5-4. Relative costs and cost breakdowns for the crops grown Celeriac $1,109 Banana yr 1 $1,291 Banana yr 2 $848 Beans $924 Cabbage $1,888 Cassava $916 Chayote yr 2+ $1,262 Coffee yr 1 $672 Coffee mature $290 Ginger $1,315 Papaya $3,175 Plantain yr 1 $2,087 Plantain yr 2 $1,471 Pumpkin $1,170 Yam $1,478 T a ni e r $869 Labor Other Insurance Agrochemicals KEY

PAGE 164

146 Figure 5-5. Relative labor re quirements, divided by activity Celeriac 37 days Banana yr 1 41 days Pumpkin 34 days Coffee (harvest) 6 years 67 days Coffee (pre-harvest) 2 years 16 days Chayote 6 years 371 days Cabbage 59 days Cassava 54 days Banana yr 2 24 days Beans 26 days Ginger 59 days Papaya 78 days Plantain yr 2 17 days Plantain yr 1 46 days Yam 43 days Tanier 32 days Agrochemical applications Harvest Other KEY

PAGE 165

147Table 5-3. Farm characteristics, labor inputs and annual net income for the 16 study farms Farm number # of acres # of acres (cultivated) FarmerÂ’s # hours/week # full-time workers # part-time workers Type of farm Annual net income of farm 1 6 4 20 2 (12hrs/wk) Mixed $5-6,000 2 53 42 70 5 (40hrs/wk) Mixed $0-10,000 3 149 ? 25 14 (40hrs/wk) 20 (coffee) Mainly coffee $0-10,000 4 27 24 30 (wife, 10hrs/ wk coffee) Mixed $0-10,000 5 90 29 30 (wife, 15hrs/ wk coffee) 1 (8hrs/wk) Mixed $0-10,000 6 43 25 35 Mixed $20-30,000 7 40 12 35 3 (30hrs/wk Oct & Nov) Mixed ~$10,000 8 56 40 70 (wife, 30hrs/wk) 3 (40hrs/wk) Mixed & poultry $20-30,000 9 124 35 50 (wife, 15hrs/wk) 2 (40hrs/wk) 2 (for papaya harvest) Mixed $20-30,000 10 30 25 60 4 (40hrs/wk) 2 (4hrs/wk Dec & Feb) Mainly root crops $0-10,000 11 100 60 65 6 (8hrs/wk) Mainly plantain $10-20,000 12 33 15 40 1 (40hrs/wk) 1 (20hrs/wk) Mainly plantain ~$20,000 13 25 25 40 3 (40hrs/wk) Mixed + fighting cocks ? 14 33 15 42 1 (40hrs/wk) 1 (24hrs/wk 9 months) Mainly plantain $10-20,000 15 69 40 12 5 (40hrs/wk) 3 (coffee hrvst, Oct.-Dec.) Mixed $10-20,000 16 70 40 40 5 (40hrs/wk) 2 (planting cabbage) Mixed $10-20,000

PAGE 166

148 Figure 5-6 is compiled from farmersÂ’ answ ers when they were asked to compare the different crops based on th ree criteria: income generate d, amount of fluctuation in sale prices and the amount of inputs needed. The charts are based on answers that used a scale of one to four to determine farmersÂ’ perceptions. Figure 5-6. Farmer perceptions of crops grown in terms of profits gained, fluctuation of sales prices and financial inputs needed. The numbers given above the columns are the rankings for the individual crops. RELATIVE PROFITS GAINEDBananaBeansCabbageCassavaCeleriacChayoteCoffeeGingerPapayaPlantainPumpkinYamYautia FLUCTUATING PROFITSBananaBeansCabbageCassavaCeleriacChayoteCoffeeGingerPapayaPlantainPumpkinYamYautiaFINANCIAL INPUTSBananaBeansCabbageCassavaCeleriacChayoteCoffeeGingerPapayaPlantainPumpkinYamYautia 0 1 2 3 463511310821037910 HIGH HIGH LOW LOW 0 1 2 3 410551071319112348 LOW LOW HIGH HIGH 0 1 2 3 45631279981311132 HIGH LOW LOW HIGH

PAGE 167

149 Figure 5-7 is compiled from questionnaire results and from information given on the Puerto Rico Agricultural Statistics Service website that was sourced from the Administración de Servicios y Desarrollo Agropecuario (ASDA), Puerto Rico Department of Agriculture (http://www.nass .usda.gov/pr/). It shows that most of the harvesting and maturation of crops occurs at th e end of the year and into the beginning of the next. The only crops that can be harves ted throughout the year are banana, papaya, plantain and pumpkin. Figure 5-7. Harvest periods for the study cr ops, giving peak periods and approximate beginning and end times for the harvests Almost all of the farmers relied on a mix of crops to generate income during the year. There were three farmers that only plan ted plantain. Plantain was the major crop for the majority of farmers (13 out of the 16 fa rmers). This crop produces the year round, is easy to grow and has government support in terms of incentives and a guaranteed market. When asked, the farmers mentioned easy mana gement, good markets, government help and ease of agrochemical application as thei r reasons for growing th is crop. Other crops Products Banana Beans Cabbage Cassava Celeriac Chayote Coffee Ginger Papaya Plantain Pumpkin Tanier Yam KEY: peak periodsbeginning and end of harvest SeptemberOctoberNovemberDecember MayJuneJulyAugust JanuaryFebruaryMarchApril

PAGE 168

150 that were grown in significant quantity on many of the farms were coffee, celeriac, ginger, banana, tanier, pumpkin, cassava and yam. Some crops, such as papaya, beans, cabbage and chayote were restricted to a fe w numbers of farms, which specialized, to some extent, on these crops. Coffee, the crop most subsidized by the government, was only grown on significant acreage on three of th e sixteen farms. All three of these farms are in the municipality of Orocovis, whic h falls into the designated coffee production zone that is supported by government incentive schemes. Economic Considerations Throughout the year, farmers had payments to make and income to be found for their household needs. When asked about how these needs fluctuated during the year, most farmers said that there was no particular time of year that had substantially higher financial needs and that most of the fluctua tions were covered by the savings they had. Nevertheless, when their payments and hous ehold financial needs were broken down by quartile, generally the time when money was mo st needed was in the final quarter or in the first quarter of the year. This relates to the time with most activity on the farm and to the time when there were many family and community events related to the Christmas/Three Kings period. All the farmers had savings that were either kept in a cooperative or in a bank. Only one farmer had savings of greater than $20,000, (he had owned and then sold a hardware store). The other farmers had $10,000 and less in savings. One of the farmers who relied solely on plantain mentioned that normally he had over $10,000 in savings but because of over-production and depressed prices for planta in in the last few years, his savings had diminished to below $5,000. For this reason he was looking to diversify on his farm. Out of the sixteen families interviewed, five rece ived food stamps as part of the federal

PAGE 169

151 “Programa de Asistencia Nutricional” (PAN) program. Others received social security payments, a government pension or medical insurance payments. All the farmers with full-time workers received th e government incentive payment, which paid for half of the workers’ wages. Most of the farmers receiv ed free fertilizers from the government and participated in the government programs of subsidized agricultural machinery and agrochemical applications. To receive thes e incentives, the farmers were visited by government officials and were required to f ill in numerous forms during the year. Doing the voluminous paperwork and laborious bookkeeping was often the greatest direct contribution to the running of the farm made by the farmer’s wife. As shown in Table 5-2, annual incomes are not very high for most families. Of the two families that make over $20,000 a year, a significant proportion of their income comes from sources other than the field crops they grow. The rest of the farmers normally produce less than $20,000 a year. Surprisingly, no ne of the farmers and very few of their spouses have any formal off-farm work. Work begins on the farm at six in the morning and finishes around noon for lunch. The afternoon s are a more flexible time with regards to work—often the farmers wait for the midday heat to pass before heading back to the fields. During the busy periods of the year th e afternoons are used for work, while at less busy times of the year, afternoons are spent doing chores around the house, making social or work-related trips, or receiving visits from ‘intermediarios’ (people who buy the harvested produce), extension workers or inqu isitive graduate students. The farmers’ wives, although not working in the fields on a regular basis, are an integral part in the functioning of the farm. Many farmers have to pick up their farm workers in the morning and then provide them with breakfast (prepa red by the farmer’s wife) before going into

PAGE 170

152 the fields. Normally the wife will help at ha rvest time (especially wi th coffee) and will also help with farm-related chores around the house (such as shelli ng beans and preparing planting material). Most of the older wives spent the day looking after grandchildren while their own children were away at wor k. Of the sixteen families, only three had children who worked on the farm or who had in terest in agriculture. Part of the economic changes on the island has meant that many pe ople leave the central region early each morning to go to work in the San Juan me tro area. The new career objectives of the younger generations, and the education opport unities provided by state and federal programs, have led many of the younger people away from agriculture. Lifestyle changes that have, in part, developed from exposure to the American way of life has meant that it is now necessary for both partners in a rela tionship to have a job, which necessitates the use of family for child care duties. Social changes have also meant that labor is more expensive and more difficult to obtain in the central region. People are less willing to accept laborer positions on farms where the work is arduous and not well pai d. There is a minimum wage limit of $4.50 an hour, but farmers usually have to provide fringe benefits such as transport to and from the farm, and meals, to entice workers to sta y. Most farm laborers only work the morning hours. All but one of the farmers in the study had five or less full-time workers; the one who had more, had the large coffee farm a nd needed the extra labor. A common strategy is to have part-time or piece-meal workers that can fill in during the busy times of the year. The summer months are commonly known as ‘el tiempo de muerto’ (literally ‘the time of death’), which derives from the sugarc ane-growing days, when there was little to

PAGE 171

153 do and it was economically difficult to make ends meet. Of the sixteen farms, eleven used part-time or piecemeal workers. Incentives Program The extensive incentives program offered by the government sustains much of the agriculture in the central re gion of Puerto Rico. The progr am is run by A.S.D.A. (La Administración de Servicios y Desarrollo Agro pecuario) of the Puerto Rican Department of Agriculture, and it serves to promote a nd develop both crop and livestock production in the country. As mentioned in Chapter 1, the incentives program is restricted to specific crops in delineated zones, where the ag ricultural activ ities are conducted on an appropriately “efficient” commercial scale (V icente-Chandler 1978). Possibly the biggest incentive in the program is the worker sa lary incentive, which guarantees a minimum wage of $4.50, of which the government pays $2.25. The government portion of the payment is made to the farmer every three months, who in turn has the responsibility of paying the workers their full wage. Farms are visited on a regular ba sis to confirm the number of workers and what work is be ing done. Farm workers are also given a government bonus at the end of the year, equi valent to 4% of their annual income. Another commonly utilized in centive is the provision of fertilizer. Table 5-4 gives information on the crops included, the quantitie s of fertilizer provided, restrictions and types of fertilizers distributed. Coffee is a crop that receives more incentives than all other crops. Fertilizers are given based on production and type of coffee. For Arabica coffee, an incentive of two hundredweight of fert ilizer (analysis: 12-5-15-3) is given for each hundredweight of green coffee beans produced, up to a maximum of 15 hundredweight per acre. For Robusta coffee, it is two hundredweight of a different fertilizer (analysis: 10-5-15) for each hundred weight of green coffee beans produced, up

PAGE 172

154 to a maximum of 10 hundredweight per acre . Each hundredweight of green coffee beans also gets a cash bonus of $22. Table 5-4. Crops and fertilizers given as part of the ASDA incentive program Crop Type Incentive Maximum # of acres Type of fertilizer provided (amount per acre) to which incentive applies (analysis) Plantain 6 hundredweight 5 acres 10-5-20-3 Bananas 6 hundredweight 5 acres 10-2-30-3 Root and 6 hundredweight 5 acres 12-5-15-3 or tuber crops 8-8-12 Pasture and 50% of the cost o f 5 acres 15-5-10 forage crops 6 hundredweight Citron 9 hundredweight 5 acr es 20-5-20-3 & 10-5-15-3 Passion fruit 4 hundredweight 5 acres 15-15-15-3 Note: other crops include West Indian ch erry, coffee and the herb culantro (recao) New plantings of coffee are supported with a payment of $1,300 per acre to cover the cost of plants, fertilizer, lime and the application and costs of necessary herbicides and insecticides. Purchase of equipment is fu rther supported by payments of up to 50% of the costs, up to a maximum of $8,000. Other incentives include purchase of all coffee produced and the payment of transport costs fo r the farmer/worker to and from the farm. Another commonly used incentive is the subsidized agricultural machinery, which is used for clearing land and preparing the fields. Although cheaper, the farmers complain of long waits and uncertain arrival of the m achinery, which means that they frequently resort to using the more expensive, privat ely owned machinery. Also popular is the crop protection and weed management assistance provided by ASDA. Herbicides, insecticides and fungicides are applied by crop protect ion ‘brigades’ who provide the labor,

PAGE 173

155 equipment and chemicals so as to mi nimize accidents and the misuse of the agrochemicals. The services are offered for a wide variety of crops. Table 5-5 gives the crops included and the agrochemicals applied. The weed control service is not limited to specific crops and ASDAÂ’s services can be solicited for general weed control around the farm. Table 5-5. Agrochemical products applied as part of ASDAÂ’s crop protection/weed control incentive program Crop Product name Active ingredient Type of agrochemical/use Coffee Di-Syston 15G Disulfoton Insecticide Temik 15G Aldicarb Insecticide/nematicide Roundup Ultra Glyphosate Herbicide Goal 2XL Oxyfluorfen Herbicide Horticultural mineral oils Petroleum-based Combating sigatoka Manzate DF Mancozeb Combating sigatoka Plantain and Banana Ametryn (Evik) Herbicide Gramaxone Paraquat Herbicide Roundup Ultra Glyphosate Herbicide Vydate L Oxamyl Insecticide/nematicide Mocap 10G Ethoprop Insecticide/nematicide Nemacur 15G Insecticide/nematicide Fungicide mix Fruit crops (Copper sulfate or Manzate + Mancozeb Ditano +Malathion 57% EC+ Diethyl succinate Volk oil spray) Fungicide Temik 15G Aldicarb Insecticide Cassava M-Pede Potassium salts Insecticide/miticide/nematicide Yam Temik 15G Aldicarb Insecticide Copper sulfate + Manzate Mancozeb Fungicide Pumpkin Copper sulfate Fungicide Sevin 80 Carbaryl Insecticide Lannate Methomyl Insecticide General Amdro Hydramethylnon Combating fire ants Siege Pro Fire Ant Bait Hydramethylnon Combating fire ants

PAGE 174

156 On-Farm Experiment Using the IPM strategy described in th e methods section of this chapter, diamondback moth damage in the experiment al cabbage plot was kept to acceptable levels. A Microsoft Excel sheet documenti ng scouting results and photographs of the developing crop can be viewed in Appendix E. At no time during the crop did the DBM damage require an increase in the weekly sp rays of Dipel DF or weekly releases of Trichogramma pretiosum . Two of the weekly Trichogramma releases were not realized due to unforeseen problems with the shippers of the insects. From the diary kept by the farmer, it was possible to determine the cash and labor requirements of the IPM strategy used against the diamondback moth and to integrate this information into a crop budget that was then used in the linear model. Figure 5.8 shows the crop budget for one acre of the ‘IPM ca bbage’; while figure 5.9 gives the crop budget for one acre of the regularly managed cabbage.

PAGE 175

157 Figure 5-8. Crop budget for an acre of cabbage using the experimental IPM strategy for the control of DBM Unitqntycostqntycostqntycostqntycost$/unittotal cost Material CostsSeedslb0.34$6 9 200$6 9 Land preparation eitherbull s cuts 2 $5 0 $2 5 $5 0 ormachine s hours 0 $ 0 $2 3 $ 0 calciu m ton 2 $2 0 $1 0 $2 0 Fertlzr #1(8-8-12) seedbe d lb20$ 5 $0.24$ 5 transplntlb132 0 $317$0.24$31 7 Fertlzr #2(20-20-20)lb1 5 $1 6 5 $5.20$1.04$21 Pestcde #1(Diazinon)gallon0.11$ 6 $5 5 $ 6 Pestcde #2(Dipel)lb 8 $1404$70.00$17.50$21 0 Herbcde #1(Gramazone)gallon0.031$ 2 $4 9 $ 2 Trichogramm a square33 6 $33 6 168$168.00$1$504 NAP crop insuranceacre1$100$100$10 0 Labor costsda y 20.56 5 $61710.1 7 $305.10$3 0 $92 2 TOTAL$1,05 9 $24 3 $1,303 Laborqnty(days)qnty(days)qnty(days)qnty(days)days/unitdays Preparing seedbed10.40.40.4 & planting seeds12.32.32.3 Maintaining seedbe d 1 5 5 5.0 Land preparation either bull s cuts 2 2 12.0 ormachine s cuts1 0 0 0.0 calciu m application11 11.0 Transplanting acre112. 5 12. 5 12. 5 Fertilzr #1(8-8-12) seedbe d application10.10.10.1 trnsplntapplication11.3 3 1.331.3 Fertlzr #2(20-20-20)application 3 1.9 8 10.6 6 0.6 6 2. 6 Pestcde #1(Diazinon)application 2 0.6 6 0 0.330.7 Pestcde #2(Dipel)application 8 5.3 6 42.6 8 0.678.0 Herbcde #1(Gramazone)application10. 5 0. 5 0. 5 Scoutingscout 8 0.840.40.11.2 Trichogramm a application 8 3.241. 6 0.44.8 Manual weeding pass14 44.0 Harvesting acre 11 5 1 5 15. 0 TOTAL41.1320.3 4 61.47 Yieldqntyincomeqntyincomeqntyincomeqntyincome$/unittotal incomeINCOMEhdwt. $ 0 20 8 $2,704$ 0 $ 0 $13.0 0 $2,704 Net Income:$1,401Period IPeriod IIPeriod IIIPeriod IV

PAGE 176

158 Figure 5-9. Crop budget for an acre of cabba ge using the regular insecticide-based strategy for the control of DBM The release of Trichogramma and the weekly scouting exercises did not greatly increase labor requirements, partly because th e procedures were simple and fitted into the existing management practices, and partly b ecause of the reduced number of insecticide applications. However, costs did increase si gnificantly, because of the costs of the Unitqntycostqntycostqntycostqntycost$/unittotal cost Material CostsSeedslb0.34$69200$69 Land preparation eithe r bullscuts2$50$25$50 o r machineshours0$0$23$0 calciu m ton2$20$10$20 Fertlzr #1(8-8-12) seedbedlb20$ 5 $0.24$ 5 transplntlb1320$317$0.24$317 Fertlzr #2(20-20-20)lb1 5 $1 6 5 $5.20$1.04$21 Pestcde #1(Diazinon)gallon0.11$ 6 $55$ 6 Pestcde #2(Dipel)lb12$210 6 $105.00$17.50$315 Herbcde #1(Gramazone)gallon0.031$2$49$2 NAP crop insuranceacre1$100$100$100 Labor costsda y 19.905$5979.84295.2$30$892TOTAL$79 3 $110.20$904 Laborqnty(days)qnty(days)qnty(days)qnty(days)days/unitdays Preparing seedbed10.40.4$0 & planting seeds12. 3 2. 3 2.3 Maintaining seedbed1 5 5 5.0 Land preparation either bullscuts2212 o r machinescuts1000 calciu m application1111.0 Transplantingacre112. 5 12. 5 12. 5 Fertilzr #1 (8-8-12) seedbedapplication10.10.10.1 trnsplntapplication11.331.331.33 Fertlzr #2(20-20-20)application 3 1.9 8 10.6 6 0.6 6 2. 6 Pestcde #1(Diazinon)application20.6 6 0.3 3 0.7 Pestcde #2(Dipel)application128.04 6 4.020.6 7 12.1 Herbcde #1(Gramazone)application10. 5 0. 5 0.5 Manual weedingpass1444.0 Harvestingacre 11 5 1 5 15.0TOTAL39.8119.6 8 59.5 Yieldqntyincomeqntyincomeqntyincomeqntyincome$/unittotal incomeINCOMEhdwt. 0208$2,70400 1 3 $2,704 Net Income:$1,800Period IPeriod IIPeriod IIIPeriod IV

PAGE 177

159 Trichogramma . There are no local producers of biologi cal control agents in Puerto Rico, and the Trichogramma had to be shipped from Californi a, which proved to be expensive. Aside from cost, there is the management issu e of dealing with the suppliers in a foreign language and the uncertainties and comp lications of multi-carrier, long distance shipments. These additional constraints are only represented in the model by costs and labor, but it should be acknowle dged that the constraints a ssociated with the use of Trichogramma in this system have another dimens ion. On the positive side, there was the acceptance and willingness of the farmer to work with this new technology. Attempts were made to determine whether the Trichogramma was active in the experimental field using sentin el plants loaded with DBM e ggs. The plants were left in the field for 24 hours before being examined in the laboratory. Unfortunately, the sentinel plants wilted within the first few hours in th e field due to an unsuitable potting mix and the plants lay prone in the pots. No evidence of egg parasitism was found. Later on in the experiment hardier sentinel plants were placed in the field, and out of the eight plants put in the field, seven did not have any parasitized eggs. In contrast, nearly all the eggs of the eighth plant were parasitized. Although this plant was placed in the cabbage near the stand of plantain that threw so me shade on this part of the plot, more work would have to be done to explain the parasiti sm pattern found. Importantly, the Trichogramma did parasitize DBM eggs in this field setting. Linear Program With the crop budgets and the informati on gained on farm characteristics and family economics, an eight-year linear pr ogramming (LP) model was created. Ideally, with the perennial crops such as coffee a nd banana, a long-term model of over 20 years would have been preferable. Unfortunately, with fourteen different cr ops (the original 13

PAGE 178

160 crops plus the ‘IPM cabbage’), modeling an increased number of years had to be balanced with model size and solver capabil ity. The Excel file with the eight-year model can be found in Appendix F. A smaller two-year model that can be run using the standard Excel solver is also included in Appendix F. This smaller model should be used to help familiarize the reader with the functioning of LP models and should not be taken as an accurate representation of the farm ing system under investigation. In the model, each year was divided into four quarters (i.e., 3 months per quarter) to improve accuracy. In each quarter, the inputs and outputs needed or produced for each activity, and the financial requirements of th e family, are given. The objective function of the model was set to maximize the discretio nary cash produced. The sum of each year’s discretionary cash produced was used as the ob jective function cell. It was hoped that the model’s answer would give a similar mix of ac tivities as found on the farms in the central regions of Puerto Rico. In addition, the mode l was used to see how the crop mix would change with two alterations of the incentives program. The first change was the removal of all agrochemical incentives (Scenario 2) and the second cha nge was, in addition to the removal of the agrochemicals, the removal of the worker salary assistance (Scenario 3). Because of the dynamics of an LP model, the first few years and the last few years of the model may not accurately represent the farm sy stem. For this reason stability patterns are looked for in the middle years, and a couple of these years are then chosen to represent the model. The 8-year model was run, with th e three different scenarios, for each of the sixteen farms. The results for each farm can be found for Years 4 and 5 in Appendix G. The results show the crops chosen by the mode l and also show the percentage of the land used, the number of workers hired and how th e savings were used and recuperated during

PAGE 179

161 the year for each farm. Between farms there were broad similarities across all three scenarios, although there were some farms wh ere the model could not find a feasible solution with the incentives removed. With the agrochemicals incentive removed (Scenario 2), one farm model was unable to find a feasible solution. With all incentives removed (Scen ario 3), there were five of the sixteen farms without feasible solutions. The resu lts from Farm #8 were fairly typical of how changes in the incentives pr ogram affected crop mixes given by the LP model. The results from Years 4 and 5 for Farm #8 are thus used in this results section to demonstrate these crop mix changes. Figure 5-10 gives the results of there being no change to the incentives program (Scenario 1). Plantain dominates all farm solutions with this configuration. Chayote and celeriac almo st always accompany the plantain, usually on a smaller scale. Figure 5-11 shows the crop mix of Farm #8 when the agrochemicals are removed (Scenario 2). There is a greater mi x of crops in this scenario and also more labor hired, in part because of the lack of government-provided labor that normally applies agrochemicals. Figure 5-12 gives th e crop mix for Farm #8 when there are no incentives offered (Scenario 3). Aside from the reduced income there is also a reduction in the diversity of crops grown. The model does not support the broad mix of crops of Scenario 2 and instead chooses crops th at, although less profitable, do not have prohibitively high labor or cash requirements.

PAGE 180

162 Figure 5-10. Years 4 and 5 of the 8-year LP model for Farm #8 with full incentives (Scenario 1). Figure 5-11. Years 4 and 5 of the 8-year LP model for Farm #8 without agrochemical incentives (Scenario 2). maximized income: $634,799Year 4Year 5 Celeriac 34 . 18274881 Celeriac 11 . 78480674 Banana 0 Banana 0 Beans 17 . 07983012 Beans 33 . 45608171 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 10 . 75911155 Chayote 10 . 75911155 Coffee 0 Coffee 0 Ginger 18 . 25318468 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 1 . 935876861 Pumpkin 7 . 971996794 Yam 0 Yam 0 Tanier 25 . 05182691 Tanier 37 . 26889166 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%6.68100%0%I100%9.35100%100% II100%6.060%0%II100%0.530%0% III100%0.000%0%III100%0.000%0% IV100%12.400%100%IV100%15.490%0% Year 4 crops in the ground0510152025303540 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $785,379 Y ear 4 Y ear 5 Celeriac 0 Celeriac 16 . 14112453 Banana 0 Banana 0 Beans 0 Beans 0 . 287446285 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 12 . 89477725 Chayote 12 . 89477725 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 43 . 10522275 Plantain 42 . 81777646 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 . 287446285 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%2.68100%100%I100%1.30100%100% II100%0.920%0%II100%3.330%0% III100%0.060%0%III100%0.440%0% IV100%0.270%0%IV100%1.910%0% Year 4 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 181

163 Figure 5-12. Years 4 and 5 of the 8-year LP model for Farm #8 without agrochemical incentives or worker salary incentives (Scenario 3). Table 5-6 shows which crops are chosen for the three incentive scenarios. It also shows the frequency of the crop choices in those farms where a feasible solution was given. In Scenario 1, only plantain, chayote a nd celeriac are chosen regularly. In Scenario 3, there are also only three crops chosen w ith regularity: celeriac, pumpkin and tanier. Scenario 2, as mentioned previously, has the greatest diversity of crops. There are five crops that are chosen on over half the farms to fulfill the requirements of the model. Celeriac, chayote, pumpkin and tanier appear in all three scenario solutions, although only the first two are chosen in any quantity in Scenario 1. There are some crops that are never selected—banana, cabbage, IPM cabbage , coffee, papaya and yam. Reasons for this will be discussed later. maximized income: $433,403 Y ear 4 Y ear 5 Celeriac 33 . 07760505 Celeriac 16 . 13936298 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 6 . 030985207 Pumpkin 12 . 1342135 Yam 0 Yam 0 Tanier 41 . 28056777 Tanier 49 . 78272733 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I98%7.43100%0%I100%7.58100%100% II100%0.850%0%II100%0.850%0% III95%0.000%0%III100%0.000%0% IV46%3.500%100%IV39%3.410%0% Year 4 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground0102030405060 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 182

164 Table 5-6. The crops chosen by the 8-year LP model for the three different scenarios. The frequency of those choices is also given. Scenario Crop Number of farms with crop/total number of farms Plantain 16/16 Full incentives (Scenario 1) Chayote 14/16 Celeriac 12/16 Tanier 1/16 Pumpkin 1/16 Beans 1/16 Cabbage 1/16 Tanier 15/15 Beans 15/15 Agrochemicals removed (Scenario 2) Celeriac 14/15 Chayote 13/15 Pumpkin 8/15 Ginger 7/15 Celeriac 10/11 Tanier 10/11 No incentives (Scenario 3) Pumpkin 10/11 Chayote 1/11 Plantain 1/11 There is a steady decrease for all farm s in the money (“maximized income”) produced by the model from Scenario 1 through to Scenario 3. This is to be expected with the removal of the incen tives. The quantity of the money produced in the model is much higher than was found on the farms. Ther e was also not such a high percentage of the land area utilized on the farms, as the area used in the model. Manipulating the LP Model to E xamine the Cabbage IPM Activity An LP model can be used to see what it w ould take for activities to be included in the final solution. This can be achieved by decreasing the cash or labor inputs or by increasing the yield or the sales price. The information generated by the model can help direct research, by using the changes that gi ve the new technology th e best opportunity of being included in the model and by extension, being successfully adopted by the farmers.

PAGE 183

165 The inputs and sales prices were manipulat ed for both the regularly managed cabbage and for the cabbage that incorporated the IP M strategy. For each incentives scenario, the cash requirements, the labor and the sales pr ice were adjusted incrementally until the cabbage activity was included in the 8-year model. This was done for three farms (Farm #6, Farm #8 & Farm #16) and Tables 5-7, 58 and 5-9 show what changes would be needed for the cabbage activities to be included in the model. Table 5-7. Scenario 1 (full incentives): ch anges that would result in the cabbage activities being included in the 8-year LP model. Cabbage Cabbage (IPM) Alteration Alteration Selling price Selling price Farm #6 15% increase Farm #6 30% increase Farm #8 15% increase Farm #8 30% increase Farm #16 15% increase Farm #16 30% increase Production costs Production costs Farm #6 Oct-Dec: 40% decrease Farm #6 Oct-Dec: 70% decrease Jan-Mar: N/A Jan-Mar: N/A Farm #8 Oct-Dec: 40% decrease Farm #8 Oct-Dec: 70% decrease Jan-Mar: N/A Jan-Mar: N/A Farm #16 Oct-Dec: 40% decrease Farm #16 Oct-Dec: 70% decrease Jan-Mar: N/A Jan-Mar: N/A Labor Labor Farm #6 Oct-Dec: 20/40 days decrease Farm #6 Oct-Dec: N/A Jan-Mar: N/A Jan-Mar: N/A Farm #8 Oct-Dec: 20/40 days decrease Farm #8 Oct-Dec: N/A Jan-Mar: N/A Jan-Mar: N/A Farm #16 Oct-Dec: 20/40 days decrease Farm #16 Oct-Dec: N/A Jan-Mar: N/A Jan-Mar: N/A Note: “N/A” signifies that even a 100% re duction in these input requirements wouldn’t favor the inclusion of cabbage.

PAGE 184

166 Table 5-8. Scenario 2 (no agrochemical incen tives): changes that would result in the cabbage activities being include d in the 8-year LP model. Cabbage Cabbage (IPM) Alteration Alteration Selling price Selling price Farm #6 5% increase Farm #6 20% increase Farm #8 5% increase Farm #8 20% increase Farm #16 5% increase Farm #16 20% increase Production costs Production costs Farm #6 Oct-Dec: 5% decrease Farm #6 Oct-Dec: 45% decrease Jan-Mar: 30% decrease Jan-Mar: N/A Farm #8 Oct-Dec: 10% decrease Farm #8 Oct-Dec: 45% decrease Jan-Mar: 40% decrease Jan-Mar: N/A Farm #16 Oct-Dec: 10% decrease Farm #16 Oct-Dec: 45% decrease Jan-Mar: 40% decrease Jan-Mar: N/A Labor Labor Farm #6 Oct-Dec: 5/40 days decrease Farm #6 Oct-Dec: 30/41 days decrease Jan-Mar: 5/20 days decrease Jan-Mar: N/A Farm #8 Oct-Dec: 5/40 days decrease Farm #8 Oct-Dec: 30/41 days decrease Jan-Mar: 5/20 days decrease Jan-Mar: N/A Farm #16 Oct-Dec: 5/40 days decrease Farm #16 Oct-Dec: 30/41 days decrease Jan-Mar: 5/20 days decrease Jan-Mar: N/A Note: “N/A” signifies that even a 100% re duction in these input requirements wouldn’t favor the inclusion of cabbage. Fewer adjustments had to be made in Scenario 2, (where there was worker salary assistance but no agrochemical incentive), for th e cabbage activities to be chosen in the model. Scenario 3, with no incentives available, needed the greatest changes in inputs or sales price for cabbage to be included. As expected, the ‘IPM’ cabbage needed greater reductions in inputs and higher sales prices than the regularly managed cabbage for inclusion in the model. These figures only show what it would take with changes in only one aspect of input requirements or sales price to achieve inclusion in the model. For this reason, in some instances, even with a 100% reduction in the input, the cabbage was still not seen as a viable optio n in the model’s solution.

PAGE 185

167 Table 5-9. Scenario 3 (no incentives): change s that would result in the cabbage activities being included in the 8-year LP model. Cabbage Cabbage (IPM) Alteration Alteration Selling price Selling price Farm #6 30% increase Farm #6 50% increase Farm #8 15% increase Farm #8 35% increase Farm #16 20% increase Farm #16 40% increase Production costs Production costs Farm #6 Oct-Dec: 80% decrease Farm #6 Oct-Dec: N/A Jan-Mar: N/A Jan-Mar: N/A Farm #8 Oct-Dec: 55% decrease Farm #8 Oct-Dec: 85% decrease Jan-Mar: N/A Jan-Mar: N/A Farm #16 Oct-Dec: 65% decrease Farm #16 Oct-Dec: 95% decrease Jan-Mar: N/A Jan-Mar: N/A Labor Labor Farm #6 Oct-Dec: 20/40 days decrease Farm #6 Oct-Dec: 35/41 days decrease Jan-Mar: 20/20 days decrease Jan-Mar: N/A Farm #8 Oct-Dec: 15/40 days decrease Farm #8 Oct-Dec: 30/41 days decrease Jan-Mar: 15/20 days decrease Jan-Mar: N/A Farm #16 Oct-Dec: 20/40 days decrease Farm #16 Oct-Dec: 30/41 days decrease Jan-Mar: 20/20 days decrease Jan-Mar: N/A Note: “N/A” signifies that even a 100% re duction in these input requirements wouldn’t favor the inclusion of cabbage. Discussion Crop Characteristics and Inclusion in the LP Model The LP model is inherently objective and c hooses those activities that have the best mix of low cash/labor inputs and high yield/cas h outputs. This cost benefit analysis is performed within the framew ork of time divisions, wher e changing availability of resources and changing economic needs, dete rmine the final mix of activities chosen. Tanier, chayote, pumpkin and celeriac were the crops chosen in all incentives scenarios. Tanier is a crop with low labor requireme nts, low cash inputs and relatively good cash returns. In the model, the tanier planted in March-May was chosen over that planted in

PAGE 186

168 April-June, perhaps because profits for this planting were made during the labor/inputintensive end-of-year months. Chayote, anot her crop found in all scenarios, has a very high start-up cost (the tre llis structure) but very good returns. The crop produces throughout the year. Also, once established, is easy to manage, has few pests and has low labor demands. Pumpkin does not have particul arly low labor or cas h requirements and its net income is not all that high. Its advant ages are its quick retu rns (four months to harvest followed by one to two months of harves ting) and the fact that it can be cultivated throughout the year. Celeriac , like pumpkin, does not have particularly low input requirements or particularly high economic return s. It can be cultivat ed for much of the year and profits can be made from April and into the summer months, where little other crop activity exists. The crops that were never represented in the model were banana, cabbage, IPM cabbage, coffee, papaya and yam. Coffee and banana have relatively low net incomes, while cabbage, papaya and yam had prohibitivel y high costs. The high costs were usually associated with labor (and agrochemicals in the case of cabbage and papaya, which do not fall under the governmental agrochemical incentive scheme). As the results show, changing the incentives offered affects which crops best fulfill farm objectives. Plantain is only included in the model when there ar e full incentives given. PlantainÂ’s main costs derive from relatively high labor requirem ents (year-round harvesting and maintenance) and regular agrochemical applications. Bo th of these are am eliorated through the governmentÂ’s incentives. Once the agrochemical incentives are removed (Scenario 2), other crops emerge within the model. Ginger, beans and pumpki n, crops not chosen in Scenario 1, become

PAGE 187

169 more cost-effective than the previously subsid ized crops and so are selected in the model (ginger and beans do not form part of the agrochemical incentives scheme). Plantain drops out, not because of the increased costs of the agrochemicals, but because of the higher labor costs. Without the incentives, th e farmer has to cover the labor associated with the regular agrochemical applications . When all incentives are removed (Scenario 3), the crop mix again changes. The most significant government incentive is the assistance with worker salaries. All crops w ith significant labor requirements drop out of the model when this incentive is remove d, and those crops that have low labor requirements and reasonable economic return s in the model replace them. These crops do not have the same economic returns as the s ubsidized crops and so over-all farm income declines. These changes in labor costs are reflec ted by the LP model in the number of laborers hired as part of the final solution. The greatest number of laborers used in the model is found in Scenario 2, where the agro chemical incentive has been removed, thus necessitating the use of a grea ter number of laborers to apply agrochemicals. This increase in labor is sustained because the worker salary incentive remains in place and labor is relatively cheap. Scen ario 1, which also has the worker salary incentive, has the second highest number of workers employed. Interestingly, without the agrochemicals incentive (Scenario 2) or w ithout the worker salary he lp (Scenario 3), there is proportionally more labor hired in Periods I and IV (January-March & OctoberDecember respectively) than in the other two periods of th e year. This is because the farmer has to hire the labor that normally is supplied by the government to apply the agrochemicals needed during the busy e nd-of-year/beginning-of-year period. Also,

PAGE 188

170 without the plantain in the m odel, there are more crops repr esented in the model that are only cultivated in Periods I and IV. The hi gher labor needs in Pe riods I and IV are mirrored in Scenario 3, although the hiring le vels are lower because of the increased labor costs. These results indicate that w ithout the incentives, there would be greater likelihood of more part-time or temporary wo rkers to fulfill Periods I and IV labor requirements and a reduction in th e number of full-time workers. Whole Farm Incomes, Land Use Pa tterns and Fluctuating Markets The two glaring differences between the m odel output and the real-life situations on the farms are the amounts of money made, and land used in the modelÂ’s solutions. The two are linked in that the added acreage under cultivation brings greater profits. As results from the interviews show, the farms normally ma de less than $20,000 a year, whereas net income from the models fr equently exceeded $100,000. Interview results also show that the percentage of land used was, in all bu t one case, less than 100%. In fact, the average percent of the farm under cu ltivation was 60%. Why then was there this discrepancy between the model and reality? One crucial aspect of agriculture in th e central region of Puerto Rico is the difficulty that farmers face in selling their products. Markets are small, demand fluctuates, transport links are not sufficient, and there is an incompatibility between the quantity and form of the agricultural products that come from the farms and the quantity and form of the products desired by the mark ets. As mentioned in Chapter 1, consumer patterns have shifted on the island and most people buy from supermarkets, which obtain most of their produce from outside the country or from the larger farms on the south coast. There are no large supermarket trucks visiting farms in the central mountains, or visiting local collection poi nts in the towns. Nearly all trade is done with

PAGE 189

171 “intermediarios”, small, independent middlem en who link the mountain farmers with the island’s markets. Figure 5-13 shows how much of the plantain reaches the markets. Figure 5-13. Pick-up truck taking pl antain to market in San Juan The supermarkets also prefer their produce in processed and packaged form. Nearly all of the farmers mentioned that the superm arkets carry roots and tubers from other countries that include Costa Rica and Dominican Republic. The reason they give is that these products are washed, processed and packaged. For example, cassava perishes quickly after harvest and needs to be protec ted by dipping it into wax or paraffin to extend its shelf life. There is no facility in th e mountains to treat harv ested cassava in this manner and so, paradoxically; most of the cassa va sold in Puerto Rican supermarkets comes from abroad. Another issue in the rise of supermarkets is the quantities of produce that these outlets need. Gene rally, one farm from the central region cannot alone supply the needs of a supermarket. Thus, accum ulating enough produce from these mountain farms becomes a time-consuming and risk-prone task; something that most supermarkets are not willing to tolerate.

PAGE 190

172 The rise of the large commercial farms on the south coast has been another big blow to the farms of the central region. Th ese farms can provide the quantities needed by the supermarkets. There are also good transpor t links in the form of a modern highway between the south coast and the major market s of the San Juan metropolitan area. In addition to this, there are farms on the south coast that hire marketing specialists to approach the supermarkets and find buyers for their produce. The speci alists earn a profit on the sales. A method utilized by some of the mountain fa rmers is direct sale of the produce to the “plazas”, government-sponsored markets in city areas. The problem with this method is the time and costs involved with taking the produce out of the mountains and then marketing the produce directly to the buyers. One farmer mentioned that stallholders were unwilling to buy his produce at a fair pr ice because they had informal agreements with the “intermediarios” not to buy direct fr om farmers. Other farmer efforts include the development of local farmer markets, whic h unfortunately do not provide much income. Local people in the central region already buy much of their produce informally from the farmers, while city people are unwilling to make the hour’s trip up the mountains on a regular basis to buy the produce. Other farm ers have tried to set up marketing groups, such as the group from Orocovis, which are setting up a chayote processing and retail cooperative. The local and federal governments ar e assisting the farmers by buying their produce as part of federal funding projects such as the WIC (Women, Infants and Children nutritional assistance program), the Farmers’ Market Nutrition Program, and the National School Lunch program. The local govern ment has a network of 15 centers that

PAGE 191

173 form part of the marketing program offe red by ASDA, where farmers can go buy and sell produce. The local government also buys agricult ural produce that includes crops such as plantain, celeriac, yam and ginger (usually at times of surplus). Coffee has its own marketing incentive scheme run by ASDA. ASDA and the local Department of Agriculture have recently established a progr am that identifies and promotes locally grown produce. A “del pais” (“ of the country”) s ticker identifies all marketed produce and there are advertising cam paigns that exhort buyers to support local agricultural producers, providers of healthy, fresh, Puer to Rican produce. The government is also trying to establish farmer-led production nucle i that can supply governmental institutions with fresh produce and that can support the federal programs mentioned earlier. In spite of all this governmental help, most farmers report losses of some kind through unsold produce. Some crops are worse th an others in terms of percent lost. The crops most often mentioned as having hi gh losses were cassava, papaya, ginger and plantain. Plantain was also, however, one of the crops that was mentioned in terms of having good markets. There ar e good markets during the cooler, winter months, when yields are suppressed and pr ices are high. During the su mmer months there is overproduction and wastage. The government buys so me excess plantain, but there is much lost. Whereas the model assumes that all the pl antain produced is sol d, the reality of the summer markets means that the full profits are not gained. The crops mentioned by the farmers as ha ving a good market were beans, cabbage, celeriac chayote, coffee and yam. The differe nces in markets could explain not only the reduced income found on the farms in comparison to the model, but also how crops such as coffee and yam, which have governmental market support, are found on the farms but

PAGE 192

174 not in the LP model solution. The relatively po or markets also help explain why not all of the land on the farms is cultivated. Another reason for this is the unsuitability of some of the land on these mountain farms for any sort of cultivation, whether it is because of steep slopes or inaccessible riverbanks. Reducing Inputs in IPM Systems—Labor To improve the likelihood of an IPM stra tegy being adopted by a farmer, associated inputs have to be minimized. In the farming system of the central mountains of Puerto Rico, there is little family labor to draw on, and so labor has to be brought in at a cost. Because of minimum wages and the higher social expectations of the general populace, labor is a costly resource. Ways have to be found of either reducing the amount of labor used, or making more efficient use of the labor already hired. One obvious way to reduce labor is to stop calendar sprays. One character of this farming system, however, is that there is not much use of insec ticides on many of the crops. Management of most of the root and tuber crops, co ffee, plantain, banana and chayote use very little, if any, insecticide applications. Cabbage, pumpkin and papaya are the biggest recipients of in secticide applications, and th ese crops are subjected to calendar sprays. For these crops, there is the possibility of eliminating calendar sprays and implementing quick and simple scou ting techniques and threshold-derived treatments. Treatments against sporadic pests would result in the greatest reductions in labor. Pests such as the diamondback moth, with their ubiquitous presence and short life cycles, result in frequent sprays that would negate th e potential labor reductions associated with scouting exerci ses and reduced applications. As previously mentioned, much of the wo rk is concentrated at the beginning and end of the year, with the summer months bei ng particularly slow. This is one of the

PAGE 193

175 reasons why there are numerous part-time or temporary workers. The agricultural census of 1998 states that of the 40,799 workers reco rded, only 16,584 worked for five months or more (http://www.nass.usda.gov/census/cen sus97/puertorico/tbl29.pdf). Nevertheless, it may be useful to determine whether there are more full time workers than perhaps there should be. Many of the full-time workers ar e part of the extended family, and so decisions to hire them are not purely economic. Also, with a ll the paperwork, it is easier to keep full-time workers on the books, receiv ing government salary incentives, than it is to keep re-establishing salary assistance for part-time workers. The only time in this study where farmers did not claim salary assistan ce was when temporary or part-time workers were hired. For these reasons and others, ther e could be more full-time workers in the system than would be warranted on purely economic grounds. This could be particularly true in the summer months. If this phenomenon was true, then, perhap s research efforts could look to design IPM activities that utilize the ‘excess’ labor of the summer months . Farm sanitation and the removal of crop residues is one idea. Ot her ideas include the pl anting of trap crops, banker plants and landscape alterations that would decrease a pest’s likelihood of finding or damaging a crop sown later in the year. Th ese alterations would al so give beneficial organisms a greater chance of successful esta blishment. Another use of labor in the summer months could be the provision of trai ning in IPM techniques or other pertinent subjects for the workers. In the summer m onths, the farmer could involve himself in trying to develop markets that may give a premium for his produce (such as the organic markets). The successful use of ‘slow’ pe riods in the year depends on changing the mindsets of people away from a purely reac tive use of pesticides towards a more

PAGE 194

176 interactive, holistic approach of the In tegrated Pest/Crop Management (IPM/ICM) concepts. Can anything be done now to lower pest threats later? Farmers, extension agents and scientists all play a part in answering this question. Other means of decreasing labor is the implementation and use of appropriate technology and equipment. This may not necessa rily relate specifica lly to IPM-related technology or techniques, but if there is a decrease in overall labor demands, the acceptance of IPM strategies may be easier to bear. Reducing Inputs in IPM Systems—Cash Successful implementation of IPM practices will reduce pesticide applications, and so reduce cash inputs by loweri ng pesticide costs and labor co sts. In terms of releasing biological control agents, costs are prohibi tively high because of the lack of local producers of beneficial organisms. This is an avenue to explore; th e problem may be that local markets for beneficial organisms are sm all and uneconomical. The alternative is the conservation of existing natural enemies to reduce pest species’ numbers. This may effectively be achieved through modifications of the farm landscape and farm activities, especially if these activities utilize the labor of full-time workers in the slow, summer months. The use of compatible insecticides or rational use of less compatible insecticides would be an important way of supporting these efforts. If local rearing or the importation of natural enemies were thought feasible, then there would have to be supporting scientific work done to identify the species mo st likely to successfu lly control the pest species without having a negative impact on the island’s ecosystem. Unlike labor, there were no times in th e year where the crop activities led to fluctuations in cash availabi lity. In the busy Periods I a nd IV, the improved markets, better prices and higher profits offset the increased need fo r cash input requirements. In

PAGE 195

177 the summer, the situation is reversed. Any cash shortages there may be are covered by farmer savings or by the regular federal ec onomic assistance. The farms most likely to suffer are those that rely on onl y one or two crops, such as those farms that only grow plantain. Without the incentives scheme or surplus-buying program run by the local government, these farmers might well face cash s hortages in the summer months. If this were the case, then any design of IPM stra tegies would have to take this into consideration. Organic Premiums and Increased Sales Prices From an economic perspective, it is not rational to adopt a measure that is more costly to implement when there is no financial gain to be made on the outcome. This is especially true when there is no enforcemen t of legislation mandating these changes. In some countries there is enough social drive, so cial pressure or leve l of economic security that agricultural practiti oners are willing to accept th e potential economic losses associated with the adoption of pest cont rol strategies, which although less reliant on agrochemicals, can often lead to higher co sts or labor requirements. One obvious way of ensuring the adoption of IPM crop protection pr actices is the existence of premiums for sustainable/organic crops. As mentioned before, many of the crops grown in the central region of Puerto Rico are not exposed to insecticides. Possibly the most difficult management practice to change would be the use of herbicides. The farmers in the study did not use insecticides in the production of celeriac, cassava, ginger or tanier. None of these crops were exposed to fungicide trea tments and tanier, ginger and cassava were weeded manually. Thus there is the possibili ty of marketing these products as organic, and gaining higher prices for them.

PAGE 196

178 Puerto Rico offers good potential for the development of organic agriculture. The decline of agriculture during the last century, and the disenfranchiseme nt of much of the local agriculture with the intensive, agroch emical-based production systems promoted by the government in the late 70’s, means that it would not take much to shift land into organic production. Puerto Rico also has a good land-grant univ ersity system that could support such moves and the island has unfette red access to the rapidly increasing organic markets of the United States. The country has had a long tradition of environmental activism (Concepción 1995) and there have be en active organic pr oduction efforts on the island for the last two decades. Unpublishe d research by Guptil l (2001) shows that despite all these advantages and even with the existence of a number of organizations that promote sustainable agriculture, organic produc tion in Puerto Rico is negligible. Her conclusions point to the political influe nce and economic domination by the business class whose power derives from the economic developments of the last century, and whose interests lie in maintaining high cons umer levels of imported products and ideas on the island. Thus there is no great drive to develop the orga nic sector, even while other Caribbean nations are developing this opti on and taking advantage of its benefits. Cabbage IPM Case Study Cabbage is one of the crops grown in the mountains that require regular insecticidal applications. Without it, there is significant economi c damage caused by the ever-present DBM moth. Cabbage was an impor tant crop in the centr al region but since the 1980’s its importance has greatly decr eased. Reasons for this were DBM’s development of resistance to insecticides and the competition from cabbage farmers on the south coast. Figures from the 8-year LP model show that, based purely on a cost/benefit analysis, cabbage is economically not the best crop to grow in this farming

PAGE 197

179 system. This is especially true for cabbage that uses the experimental IPM strategy for controlling DBM. The increase in selling price and the re ductions in inputs needed for cabbage’s inclusion in the model varied depending on th e incentives scheme offered in the model. With the incentives offered presently by the local government (Scenar io 1), it would take a lot of changes to make cabbage economica lly viable. The model can be of help in deciding how best this may be achieved. From the results, it seems that no amount of change in labor availability would bring IPM cabbage into contention and that the reductions in production costs required are also unrealistic. Increasing the selling price is the best way of including cabbage in the model, although the 30% increase needed may be considered high. The regular price for 100 lbs of cabbage is around $13 and this would have to increase to $16.90 for the model to include the IPM cabbage in the final solution. A 50 lb. bag of cabbage normally contains 8–12 cabbage heads, which, with a 30% increase in selling price, w ould come out as an extra 16–24 cents per cabbage head. This may well be acceptable to people willing to pay a premium. The scenario that incorporated cabbage in to the model with the least amount of changes was Scenario 2, where the worker sa lary assistance rema ined in place but the agrochemical incentives had been removed. Cabba ge is not one of the crops that benefits from the agrochemical incentive scheme. In this case, the selling price of IPM cabbage would only have to increase by 20% (11–16 cen ts increase per cabbage head). The production costs would have to be halved in Pe riod IV or the labor requirements for this period lowered to 11 days per acre for the IPM cabbage to be included at the regular sales

PAGE 198

180 price. Input reductions in th e January-March period (Period I) would still not lead to inclusion of IPM cabbage into the model. For Scenario 3, increases in sales price and reductions in inputs would have to be higher than in the other two scenarios for cabba ge to be included. Interestingly, it seems that without the incentives, ther e is more variation in what is needed to include cabbage into the model between the different farms. Th is is probably evidence that the individual characteristics and resources of a farm have more of a direct impact on what constitutes the best mix of crops on a farm when the in centives are removed. The incentives scheme removes much of the natural variability in the system and lessens its ability to evolve and make use of new opportunities. How then could researchers bring IPM to farmers growing cabbage in the central region of Puerto Rico? The LP models show that considerable care would have to be taken not to increase labor or cash inputs sign ificantly. Insecticide sprays would remain a pillar of the DBM control strategy, but c ould be made less costly through a reduced number of sprays. Decisions to do so could be based on an effective scouting strategy. Numbers of DBM could perhaps be reduced through landscape alterations. The farmers in the mountains have an advantage in how and where their cabbage is planted. They are not restricted by the necessity of agricultur al machinery or by management of large acreage. The cabbage can be grown amongst ot her crops, which may help diminish its profile within the farm, making it more di fficult to be found by the DBM. Ruben Ortiz often planted beans and corn along the margins of the cabbage (Figure 5-14) but this was not commonly observed on other farms.

PAGE 199

181 Figure 5-14. Cabbage grown in a multi crop system Banks of plantain or banana could be us ed as windbreaks for natural enemies or could be used make it more difficult for DBM to move into the cabbage. If, as has been suggested, DBM does move towards the edges of a stand of cabbage, could a brassica trap crop (weed plant?) be sown around the edges of the cabbage in an attempt to ‘remove’ the edges of the cabbage crop and to draw the DBM away from the cabbage and into the surrounding plants? One interesting phenomenon noticed during the work on Ruben Ortiz’ farm was the presence of lizards in the ca bbage field. As Figure 5-15 show s, there was initial concern that the lizards were predating on the Trichogramma emerging from the release stations.

PAGE 200

182 Figure 5-15. Lizard found on one of the Trichogramma release stations. This may well be the case. It may also be th at the lizards were just using the release stations as vantage points from which to hunt their prey. Having more vertical crops like corn close-by may improve the impact of generalist predators by giving them a better chance to find their prey. The building of stone placements or the use of organic mulch might also give shelter to gene ralists such as spiders. Some of these activities could be done during the summer. If beneficial insects were in corporated as part of the c ontrol strategy, then some of these landscape alterations ma y be directed at improving pa rasitoid retention in the cropping area or augmenting natu ral populations of the organism s that already exist. With cabbage only grown for a few months of the y ear, it may be useful to focus the work on native generalists such as Cotesia plutellae Kurdjumov (Hymenoptera: Braconidae:

PAGE 201

183 Microgastrinae) that have the best chan ce of establishing themselves on the other lepidopteran species in the area. Alternat ively, innundative releas es of a biological control agent may work best for cabbageÂ’s short cropping cycl e; but then some way has to be found to provide natural enemies at a r easonable price. It ma y also be useful to examine the possibilities of employing ento mopathogenic organisms. The mountains have higher humidity while the coastal farm s have higher temperatures. Pheromone traps might be used in a trap-and-kill strategy or mi ght be used to attract adults to sources of insect pathogens. Probably the most practical way of improving existing DB M control is by improving existing methodologies and by inco rporating some simple, non-chemical techniques such as the scouting strategy us ed in this work. Work could be done on improving insecticide delivery to all parts of the cabbage plant, something that becomes increasingly difficult with plant developmen t. The biology of DBM is such that a few missed larvae represent a significant sour ce for future damage. Wetting agents and sticking agents can all help in the delivery, as can the manipulation of planting patterns. All sides of the plant should be accessible to the spray noz zle and calibrations should be accurate to ensure that the dosage per plant is right. Choosing the right insecticide is also important. Rotating insecticides to avoid the development of resistance and using insecticides with limited negative effect on beneficial organisms is important. Improving the effectiveness of insecticide use is probabl y the most cost effective way of maintaining low DBM populations in these central mountain farms.

PAGE 202

184 Linear Programming Models and their Bene fit to Farming Systems Research and to IPM Strategy Development The gathering and collating of the data needed for an LP model is in itself a useful exercise. It is a very thorough process that touches on most aspects that influence the farming system under investigation. Having said that, the data gathering process is very time consuming and should be shared amongs t a research team. A multi-disciplinary team conducting a farming systems research ( FSR) project would be ideally suited to such a task. Once the data has been configur ed, the actual model does not take that much time to construct and is only limited by the modelerÂ’s creativity or by the processing capabilities of the spread sheetÂ’s solver . Multi-year models can be enormous. Once the model is working, has been validat ed, and accurately reflects the realities of the farming system, then it can be a very us eful tool. LP models are used in industry to decide on the most profitable use of resources . The same, to a large extent, occurs when farmers make decisions on what to grow. The model generated in this work was, for the most part, accurate although it did not incorporate the influe nce of markets. With more work the model could include th e effect of markets. A comple tely accurate representation of the central region farming system was not the primary objective of this work. The objective was to see if LP modeling could be used to help IPM research become more effective. The integrated management of insect pests is multi-faceted and draws on numerous potential strategies. Which strategies and which combinations should be implemented? Agricultural research is a means by which th is question is answered. Research, however, can be a painstaking, drawn-out process th at sometimes builds on its own momentum rather than following those avenues that have the greatest benefit. FSR is one attempt to

PAGE 203

185 use the input of different stakeholders to understand the farming system, define research needs and then produce answers. LP modeling provides a framework that can be used to appraise research ideas and that can serve as a focus for the stakeholders in an FSR project. Scenarios can be examined and development projects that extend beyond the realm of pure agricultura l research can be assessed. The impact of the local government’s incentive schemes had a very big impact on the functioning of the farms in the central region of Puerto Rico, which in turn affect ed the likelihood of IPM strategies succeeding. In this study, the LP model highlighted the fact that labor costs are central to defining the cropping activities unde rtaken. There is little family labor, and so farm labor has to be hired. The local government helps fa rmers by paying for half of the labor costs and by supplying the labor needed to apply agrochemical crops on selected crops. IPM can be labor-intensive and extr a thought would be needed to reduce IPM labor needs. The model again helps in pointing out that the su mmer months are a time when there are less labor needs in the system, something that could be utilized by IPM planners. The information gathered for the model also highlights those crops that have high agrochemical labor requirements and which are not part of the agrochemicals incentive scheme—papaya, chayote, pumpkin and cabbage. Serious insect pests threaten all these crops and so it would be sensible to begin IP M research with these crops. The other crops either do not have real insect problems (i .e., many of the roots and tuber crops) or are covered by the government’s agrochemical in centives. It would be difficult to introduce any meaningful or economically beneficial IPM scheme into these crops with the present incentives program and lack of real pest problems.

PAGE 204

186 The LP model also shows that operati onal cash is important in defining the systemÂ’s output but that there are no particular times of the year that are more needful than others. A significant pr oportion of the operational cost s are due to the hiring of labor. FarmerÂ’s generally have the money they need for running the farm thanks to their savings and government assistance programs. Perhaps more operational money would be needed if the farmers expanded their operati ons and farmed larger areas, necessitating larger sums of money at particular times of the year. FarmerÂ’s do not look to increase acreage greatly, partly due to the poor, unstable markets and partly because of the lack of infrastructure such as irrigation schemes in these mountain farms. Lack of infrastructure is linked to the historical negligence of the agricultural sector in the central region of Puerto Rico. Improving sales price seemed the easiest way for a crop to become part of the modelÂ’s solution. This is not something an ag ricultural researcher can directly change. This knowledge helps in collaborative development/research exercises by making researchers aware of the larger picture. The LP model may not be entirely accurate but it does present a means by which farmers, extension officers and researchers can explore together the various components of a system. Using the model, ideas can be tested and research objectives refined. Efforts to impr ove the farming system can be made in a collaborative and coordinated way, with the most efficient use of resources. Data generated by the model can even help in defining future agricultural policy and legislation. Policy makers are beginning to unde rstand that a new appr oach is needed in the mountains of Puerto Rico. All of the ma jor rivers arise from the central mountains and environmental pollution is an ever-inc reasing problem. Federal programs for soil

PAGE 205

187 conservation and the payments for land under fallow have begun to be adopted (The USDA Farm Service’s Conservation Reserve Pr ogram, Environmental Quality Incentives Program and Stewardship Incentive Program ar e all implemented). Unfortunately, not all developments are helping agri culture in the mountains. It has been government policy over the last few years to move people out of the congested San Juan metro area and into the underutilized rural areas (F igure 5-16). Apart from the environmental damage of such construction projects there are also fears of the social upheavals with few jobs in the mountains. Figure 5-16. New housing developments in the central region of Puerto Rico. The central region of Puerto Rico is an area undergoing transformations. The quintessential Puerto Rican farmer, ‘El Jíbaro ’, is one of the most important cultural symbols on the island and is honored in a number of ways (Figures 5-17a, b & c). There

PAGE 206

188 is great pride in the simp le lives once led by these mountain farmers and many people still have family or holdings in these areas and regularly go back to ‘connect’ with a fading way of life. The government is suppor ting the agriculture in the mountains, and without this support ag riculture might not be a feasib le endeavor. A new agricultural paradigm is needed, one that relies on the development of niche markets, some governmental support and the acc eptance of environmental responsibility. All these moves would benefit from the advancement of IPM strategies in these mountain farms. Figure 5-17. Images that represent ‘El Jíbaro’, an important cultural icon of Puerto Rico. A) ‘Monumento al Jíbaro’, Luis Ferré Highway, Cayey. B) the red ‘pava’ or straw ‘jíbaro’s’ hat; symbol of th e Popular Democratic party.C) ‘El Pan Nuestro de Cada Día’ (Our Daily Br ead) by the Puerto Rican painter Ramón Fradé (1875–1907). IPM techniques could assist crops to reach organic status and also could ensure that agrochemical use was kept to a minimum. This is not an easy task and any useful tool, such as LP modeling, s hould be incorporated. In this way a more sustainable and economically viable agriculture could benefit the active farmers still living in the central region of Puerto Rico (Figure 5-18) A B C

PAGE 207

189 Figure 5-18. Farmers who colla borated in this farming syst ems study. A) Sr. Aponte. B) Sr. Ortiz. A B

PAGE 208

190 APPENDIX A WIND SPEED DATA FOR THE FIELD EXPERIMENTS AT FORTUNA AGRICULTURAL RESEARCH SUBSTATION, 2002 Table A-1. Wind speed data for Experiment 1 (2002) Table A-2. Wind speed data for Experiment 2 (2002) Time Wind Direction Ave. Wind Speed Max. Wind Gust (km/hour) (km/hour) 4.00 pm S 12 40 4.30 pm S 11 30 5.00 pm S 6 14 5.30 pm SW 16 47 6.00 pm SW 8 21 6.30 pm SW 4 11 7.00 pm SW 4 6 7.30 pm SW 5 12 8.00 pm W 7 10 8.30 pm W 4 5 9.00 pm W 5 6 Note: 26th March 2002 from noon, the average winds were above 16 km/hr Time Wind Direction Ave. Wind Speed Max. Wind Gust (km/hour) (km/hour) 4.00 pm S 19 55 4.30 pm SW 21 55 5.00 pm S 14 44 5.30 pm SW 9 28 6.00 pm W 19 43 6.30 pm W 10 23 7.00 pm W 9 16 7.30 pm SW 8 20 8.00 pm SW 6 10 8.30 pm SW 6 8 9.00 pm SW 6 9 Note: 22nd February 2002 from 9.00 am, the average winds were above 16 km/hr

PAGE 209

191 Table A-3. Wind speed data for Experiment 3 (2002) Table A-4. Wind speed data for Experiment 4 (2002) Time Wind Direction Ave. Wind Speed Max. Wind Gust (km/hour) (km/hour) 4.00 pm S 14 55 4.30 pm S 18 56 5.00 pm S 14 53 5.30 pm S 21 60 6.00 pm SW 13 37 6.30 pm SW 10 23 7.00 pm W 8 14 7.30 pm SW 13 45 8.00 pm SW 5 8 8.30 pm W 6 13 9.00 pm W 8 16 Note: 22nd April 2002 from 10 am to 6 pm winds were generally greater than 16 km/hr Time Wind Direction Ave. Wind Speed Max. Wind Gust (km/hour) (km/hour) 4.00 pm S 11 42 4.30 pm S 11 37 5.00 pm S 6 13 5.30 pm S 7 20 6.00 pm SW 6 13 6.30 pm S 6 12 7.00 pm SW 3 6 7.30 pm W 3 4 8.00 pm NW 1 2 8.30 pm W 3 4 9.00 pm N 1 2 Note: 12th May 2002 less than 16 km/hr all day

PAGE 210

192 Table A-5. Wind speed data for Experiment 5 (2002) Time Wind Direction Ave. Wind Speed Max. Wind Gust (km/hour) (km/hour) 4.00 pm S 10 40 4.30 pm S 8 27 5.00 pm S 8 26 5.30 pm S 13 56 6.00 pm S 6 24 6.30 pm S 6 18 7.00 pm S 6 18 7.30 pm SW 5 10 8.00 pm SW 5 13 8.30 pm SW 5 10 9.00 pm W 5 6 Note: 26th May 2002 from 10 am to 2 pm the average winds were above 16 km/hr Dr. Raúl Zapata, Assistant Dean of Acad emic Affairs at the Mayaguez campus of the University of Puerto Rico kindly o ffered the data used to produce these wind information tables. Dr Zapata runs the w eather station at the Fortuna Agricultural Research Station. The wind data is collected at the weather station is collected at a height of 10 meters. It is assumed that wind speed s near the ground are lower in velocity. No data was available for Experiment 6 (2002). The wind data presented above was taken on the day that the Trichogramma were released into the field cages. The Trichogramma were released at around 4.30 pm. The tables above show 30-minute wind data record ings for a five-hour period around the time of Trichogramma release.

PAGE 211

193 APPENDIX B QUESTIONNAIRES USED FOR CENTRA L MOUNTAIN FARMS OF PUERTO RICO: INTERVIEW 1 PART I Date: INTERVIEW GUIDE Interviewee(s): Interviewee # ____________ Postal Address: Location: Telephone number: Starting time: _____________ Finishing time:__________________ INTRODUCTION Good day. Before anything, thank you very much for granti ng me this interview. This study is about the production of cabbage in Puerto Rico. I am a doctoral student in entomology from th e University of Florid a and I am here in Puerto Rico for two years working on my thesis. I work in the Crop Protection Department in the Agricultural Research Sta tion of Río Piedras. My telephone number is 528 4920. You can contact me there if you have any questions in the future. The information that you give will be used to cons truct computer models that study the most important factors in the cabbage productio n system. These models form part of my doctoral disertation where I will evaluate th e usefulness of some methods of biological control against certain pests. I hope that the computer models help in improving the management of pests and the impact of other activities found on the farm.

PAGE 212

194 I have some questions about the administrative activities of your farm. Most of these questions concern labor matters and the co sts associated with the production on your farm. This information with some other questio ns of a socioeconomic nature will help me evaluate the efficiency of bi ological control methods in a cabbage farm. The information that you give me will be made anonymous, I am not going to use your name, (only the Interviewee number), and no other will use this information either. Moreover, I am going to amalgamate the information from all of the interviews into my computer models so that the models do not represent one person or his/her farm. There will be no repercussion from your participation in this study. I would like to divide the interview into three parts. It would be better to have them on separate days. Each part will take between 2 to 3 hours. I hope that you are comfortable with that and remember, you do not have to answer any of the questions if you do not want to. There will not be a problem if you wish to withdraw from this study. Do you have any questions before we start the interview?

PAGE 213

195 PART I 1. BASIC INFORMATION First I have some questions about your family and then some about the farm.. DEMOGRAPHICS Civil Status : Single Married Have pa rtner Div/Sep. Widowed Your work: ¿Do you work off the farm? Yes No ¿Full time or part time? Full Part ¿How many hours a week? _______ ¿All the year? Yes No ¿Which months? _______ The family : Companion ¿Off-farm work? Yes No ¿Full time or part time? Full Part ¿How many hours a week? _______ ¿All the year? Yes No ¿Which months? _______

PAGE 214

196 ¿do they help on the farm? Yes No Children for each one: # Male Female Age Where do they live Do they go to school Work on the farm? Hours weekly Off-farm work? Hours weekly * 1,2 or 3? 1 2 3 4 5 6 7 8 * If the child lives off the farm: 1. do you support them financially 2. they give you some financial support 3. is financially independent of you Other family Are there other persons that live on the farm or depend in part on the farm's activities? Who? Do any of them work on the farm? ¿Do any of them work off the farm? What type of job?

PAGE 215

197 Are there other person, not of the family, th at live or are dependent on the farm? FARM DATA Size of the farm: __________ cuerdas number of cuerdas: owned ________________ rented _______________ Total area under cultivation: ___________ cuerdas GENERAL ACTIVITIES Who does most of the housework? The person Hours per week ______________________________ _________________ ______________________________ _________________ ______________________________ _________________ Mention the major crops on the farm: ____________________________________ ____________________________________ ____________________________________ ____________________________________ ____________________________________ ____________________________________

PAGE 216

198 Other crops Product Major use For sale/for use in house Product Major use For sale/for use in house Coffee Sweet potato Plantain Yam Banana Tanier Oranges Beans Cintrón Pigeon pea Other citrus Tomato Avocado Lettuce Mango Other vegetables and melons Papaya Cooking herbs Chayote Honey Passion fruit Eggs Soursop Candied fruit Pineapple Malanga Cassava Apio Do you have any livestock activities on the farm? Yes No Mention the most important livestock activities on the farm: ____________________________________ ____________________________________ ____________________________________ ____________________________________ ____________________________________ ____________________________________ Do you have any non-agricultural activities on your farm? (e.g. a shop, art and craft sales, mechanic shop etc…..

PAGE 217

199 Yes No Mention the non-agricultural activitie s that you have on your farm. Activity Who manages the activity? ____________________________________ ______________________________ ____________________________________ ______________________________ ____________________________________ ______________________________ ____________________________________ ______________________________ ____________________________________ ______________________________ ____________________________________ ______________________________ Of all your activities which are th e most important to you? Why? 1. ___________________________________________________________________ 2. ___________________________________________________________________ 3. ___________________________________________________________________ 4. ___________________________________________________________________ 5. ___________________________________________________________________ FARM MAINTENANCE What are the mantenance activities on the fa rm? For example, fixing the fences or the paths, trails or roads. Month Activity Jan Feb Mar Apr

PAGE 218

200 May Jun Jul Aug Sep Oct Nov Dec Maintenance activities without specific times in the year _________________________________________________________ _________________________________________________________ _________________________________________________________ _________________________________________________________ For each one: Type of work How many times a year? How much time do you spend on the activity (days/hours)? Who does the work? Book Keeping Do you keep records of the activities on the farm?

PAGE 219

201 Yes No What type of information do you record? Who does the work? EMPLOYMENT DATA How many people do you employ on your farm Full time employees _______________ Part time: every month __________ specific times ___________ Have you had problems acquiring labor? Yes No Problems:

PAGE 220

202 APPENDIX C QUESTIONNAIRES USED FOR CENTRA L MOUNTAIN FARMS OF PUERTO RICO: INTERVIEW 2 PART II Before anything, thank-you again for granting me this interview. This document is the second part of the questionnaire. The first part concerned general information of the farm and family. Now I have questions that are mo re specific to the individual crops and the system of production. 2. INDIVIDUAL ACTIVI TIES ON THE FARM THE CROPS We will use the most important crops that you identified in the first part of the questionnaire: Question Crop 1 2 3 Using a classification between 1 and 5 (whe re "1" refers to "nothing/never" and "5" refers to ”a lot/always"), answer the three que stions that follow, in the three columns above. I am interested in the com parisons you make between the crops: #1. Do you make much profit from this crop?

PAGE 221

203 Almost nothing 1 2 3 4 5 A lot #2. Do you make a (good) profit from this crop every year? Almost never 1 2 3 4 5 Always #3. Do you need much capital investment for the production of this crop? Almost nothing 1 2 3 4 5 A lot

PAGE 222

204

PAGE 223

205CROPSName of the crop:______________________________ _ General How many cuerdas do you sow How many cuerdas are suitable of this crop in a typical year?for the growth of this crop? What are the major advantages of this crop A . Market B . Easy to gro w C. Do not have many pests/diseasesmark them in order of importance D. Governmental support E. Other Have you constantly sown grown this crop in the last 5 years? Y es No How many times in 5 years? _________ _ To who do you sell the products? A . Supermarket B. Intermediar y C. Wholesaler D . Plaza/Plcitamark them in order of importance E. Other, indicate Financial information Do you receive government incentives for growing this crop? Yes No A . Government B. Bank C. Others Would it be viable to grow this crop without incentives? Y es No

PAGE 224

206 COSTS/BENEFITS Operational costs: Seeds Quantity of seed (or planting material) needed for one cuerda or for the total area under cultivation of this crop: _____________________ (one cuerda/______ cuerdas) Where did you acquire the seed s (or planting material)? Free seeds : A. previous harvests/stored seeds B. other farmers C. government D. other ______________________ Bought seed (from the government, ot her farmers, seeds companies etc): Supplier Cost per unit ($) Unit Percent of your seeds that you buy from this supplier. Total cost ($) E. F. G. H. Type of soil preparation: Cost:

PAGE 225

207 Irrigation system: System Cost When Fertilizer: Product How do you apply it and how much time does it take? * How many times a crop What quantity do you use every time? * How much does it cost? (units) * note if it is for one cuerda or for the total area ( ______ cuerdas)

PAGE 226

208 Labor: Hours and costs of labor for one cuerda or for the total area under cultivation of this crop. for one cuerda or for the total area ___________ cuerdas Person Activity Cost per unit ($) Unit Quantity Total cost ($) Soil preparation days Preparation or purchase of planting material days Maintenance of seed bed days Planting days Weeding days Irrigation days Application of chemicals days Harvesting days Storage( maturation, classification, packing) days You Delivery days Soil preparation days Preparation or purchase of planting material days Maintenance of seed bed days Planting days Weeding days Irrigation days Application of chemicals days Harvesting days Storage( maturation, classification, packing) days Laborers Delivery days Soil preparation days Preparation or purchase of planting material days Maintenance of seed bed days Planting days Weeding days Irrigation days Application of chemicals days Harvesting days Storage( maturation, classification, packing) days 1 Family member Who? _______ Delivery days 2 Soil preparation days

PAGE 227

209 Preparation or purchase of planting material days Maintenance of seed bed days Planting days Weeding days Irrigation days Application of chemicals days Harvesting days Storage( maturation, classification, packing) days Family member Who? _______ Delivery days Soil preparation days Preparation or purchase of planting material days Maintenance of seed bed days Planting days Weeding days Irrigation days Application of chemicals days Harvesting days Storage( maturation, classification, packing) days 3 Family member Who? _______ Delivery days Crop Protection: Indicate the problems most common in this crop. Insects: Diseases: Weeds:

PAGE 228

210 Application of insecticides Do you use insecticides on a calender basis, or does it depend on when you see the pest? For one cuerda or fo r the total area ___________ cuerdas Product How many times do you apply the insecticide during the life of the crop? (During a week?) How much does the insecticides cost? (units) How many units do you apply each time? Target insects? How many hours for each application?

PAGE 229

211Application of herbicides Do you use herbicides on a calender basis, or does it depend on when you see the pest? For one cuerda or fo r the total area ___________ cuerdas Product How many times do you apply the herbicide during the life of the crop? (During a week?) How much does the herbicide cost? (units) How many units do you apply each time? Target weeds? How many hours for each application?

PAGE 230

212 Production: ¿How much do you harves t per cuerda or by any other un it more appropriate from this crop? Quantity Unit Area What do you do with the harvest? Quantity (percent?) Price A. Sell it: To whom? i.______________________ ii ______________________ iii _____________________ iv _____________________ v ______________________ B. Use by the family: C. Wasted: D. Other:

PAGE 231

213 APPENDIX D QUESTIONNAIRES USED FOR CENTRA L MOUNTAIN FARMS OF PUERTO RICO: INTERVIEW 3 PART III Now, I have some questions for you concerning the finances of the farm. I would like to learn more about the costs and income from your farm. Also how these change during the year. 3. FINANCIAL DATA GENERAL What percent of your income is derived from the farm? 0-25% 25%-50% 50-75% 75%-100% ¿How are the finances managed on the farm? A. contracted out B. done yourself C. organized by family D. other How are your funds/accounts organized?: A. onr fund for all the activities of the farm , family and home. B. family accounts seperated plus one fund for all the econoic activities. C. family accounts seperated plus one fund for the home and another for the farm activities. D. family accounts seperated plus multiple funds for numerous activities in the house and farm.

PAGE 232

214 E. other If there are multiple funds, what are they and who manages them? A shop; animal rearing and craft projects ar e examples of activiti es that could have separate funds. Fund Person responsible ____________________________ ___________________________________ ____________________________ ___________________________________ ____________________________ ___________________________________ ____________________________ ___________________________________ ____________________________ ___________________________________ COSTS Please mention the frequency (i.s. weekly/monthly/annually) Rent for the farm or house/mortgage: (weekly/monthly/annually) Vehicle payments : gasoline: ________________ (weekly/monthly/annually) Payments for equipment or for farm improvements: (weekly/monthly/annually) Water: (weekly/monthly/annually) Light: (weekly/monthly/annually) Telephone: (weekly/monthly/annually)

PAGE 233

215 Taxes: (weekly/monthly/annually) 1. ________________________________________________________________ 2. ________________________________________________________________ 3. ________________________________________________________________ Servicing loans or debts: (weekly/monthly/annually) Maintenance of family and home (fo od, clothes etc.): $ ________ monthly/annually Food _________ monthly/annually Clothes _________ monthly/annually Education _________ monthly/annually Medical costs _________ monthly/annually Vacations _________ monthly/annually Others _________ monthly/annually Others _________ monthly/annually Others _________ monthly/annually Financial help that you give to family off the farm: (weekly/monthly/annually) _____________________________________________________________ _____________________________________________________________ _____________________________________________________________ _____________________________________________________________ Other costs: (weekly/monthly/annually)

PAGE 234

216 INGRESOS Mention the frequency (i.e . weekly/monthly/annually) I have written below, the income from the major activities that you mentioned to me in the previous interviews. Pl ease verify these figures. Crop: Income __________________________________ _______________________ __________________________________ _______________________ __________________________________ _______________________ __________________________________ _______________________ __________________________________ _______________________ __________________________________ _______________________ Livestock: Income __________________________________ _______________________ __________________________________ _______________________ __________________________________ _______________________ Non-agricultural activities: Income __________________________________ _______________________ __________________________________ _______________________ __________________________________ _______________________ Income from property or land leased to others: (weekly/monthly/annually) Income from equipment leased to others: (weekly/monthly/annually)

PAGE 235

217 ¿Could you give more information about the wo rk of family members who work off the farm. With their wages, include percent available for use in farm activities: Person Wage Percent (weekly/monthly) available ____________________________ ______________ ____________ ____________________________ ______________ ____________ ____________________________ ______________ ____________ ____________________________ ______________ ____________ ____________________________ ______________ ____________ Federal/state financial suppor t (support from programs of the Federal government or from the government of Puerto Rico. It does not have to be cash only. Do they help with the maintenance of machinery or vehicles?): 1. __________________________________________________________________ 2. __________________________________________________________________ 3. __________________________________________________________________ 4. __________________________________________________________________ 5. __________________________________________________________________ Do you receive help with the cost of labor? Yes No What percent of the cost do you pay? What are the conditions of this suppor t? (e.g. type of cultivation, management practices etc….)

PAGE 236

218 Help from family off the farm: (weekly/monthly/annually) Income from a cooperative: (weekly/monthly/annually) Other incomes: (weekly/monthly/annually) DYNAMICS OF YOUR BUDGET If you can, note the dynamics of your funds ava ilable for farm activit ies during the period of a year. Note the times and the reasons belo w; (if it is a problem of low income or high costs. For example during Christmas or th e summer for reason of the holidays or increased costs). Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Times Reasons

PAGE 237

219

PAGE 238

220 APPENDIX E MONITORING DIAMONDBACK MOTH AND SOYBEAN LOOPER IN AN EXPERIMENTAL PLOT OF CABBAGE MANAGED USING IPM TECHNIQUES The experimental cabbage plot plan is s hown above, indicating the scouting points and Trichogramma release points. On the next two page is a bitmap image of the Microsoft Excel sheet that displays the monitoring r ecord for the diamondback moth and soybean looper. Cabbage plot NBLUE VANTAGE RIO VERDESCOUTING FLAG TRICHOGRAMMA RELEASE POINT First planting Second planting

PAGE 239

221

PAGE 240

222

PAGE 241

223 APPENDIX F LINEAR PROGRAMMING MODELS The 8-year and 2-year model are containe d in Microsoft® Ex cel 2000 SR-1 files. They can be downloaded by clicking the links below. The 8-year model has one f ile containing the matrix (“ 8 year model.xls ”), three input files: (“ Model characterisitics.xls ”, “ Farmer configurations.xls ” and “ Government incentives.xls ”) and one output file (“ Output tables.xls ”). They all need to be open to be a part of the functioning model. The regular Microsoft® Excel 2000 solver cannot be used to resolve the model because of the size of th e matrix. The Premium Solver Platform with the XPRESS solver engine from Frontline Systems was found to work. The 2-year model gives an idea of how an LP model functio ns. The input and output tables are found on different sheets in the same file (“ 2 year model.xls ”). The regular Microsoft® Excel 2000 solver can be us ed in this case, as long as the solver option has been added in. (Note: the 8-year model contains Visual Basic programming. As a consequence, you may need to enable macros if the software asks about it). (Use Microsoft Explorer 4.01 or above).

PAGE 242

APPENDIX G RESULTS FROM THE 8-YEAR LP MOD EL SHOWING YEARS 4 AND 5 FOR ALL 16 FARMS INVOLVED IN THE STUDY

PAGE 243

225 Farm 1 6 acres (4 acres under cultivation) Farmer & family hours: 20 hrs/week Full-time workers: 0 Part-time workers: 2 (12hrs/week) Annual income: $5-6,000 Scenario 1 : full incentives Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives NOT FEASIBLE Maximized income: $56,559Year 4Year 5 Celeriac 0 . 064436519 Celeriac 1 . 283665104 Banana 0 Banana 0 Beans 0 Beans 0 . 60141757 Cabbage 0 . 046428702 Cabbage 0 . 046428702 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 1 . 574379244 Chayote 1 . 574379244 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 4 . 425620756 Plantain 3 . 268527695 Pumpkin 0 . 463694258 Pumpkin 0 . 636660386 Yam 0 Yam 0 Tanier 0 Tanier 1 . 04622784 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gained I100%0.38100%62%I100%0.24100%0% II100%0.010%0%II100%0.100%100% III100%0.010%0%III100%0.000%0% IV100%0.000%38%IV100%0.450%0% Year 4 crops in the ground012345 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground012345 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Maximized income: $42,886Year 4Year 5 Celeriac 5 . 388307041 Celeriac 2 . 758575051 Banana 0 Banana 0 Beans 0 Beans 2 . 132610966 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 . 970510183 Chayote 0 . 970510183 Coffee 0 Coffee 0 Ginger 2 . 903820841 Ginger 0 . 254401094 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 . 747315722 Pumpkin 1 . 378353254 Yam 0 Yam 0 Tanier 1 . 181873128 Tanier 3 . 396735468 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gained I100%0.35100%0%I100%0.72100%0% II100%0.550%0%II100%0.000%0% III100%0.000%0%III100%0.000%0% IV100%0.670%100%IV100%1.260%100% Year 4 crops in the ground0123456 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground012345 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 244

226 Farm 2 53 acres (42 acres under cultivation) Farmer & family hours: 70 hrs/week Full-time workers: 5 Part-time workers: 2 (12hrs/week) Annual income: $0-10,000 Scenario 1 : full incentives Maximized income: $858,390Year 4Year 5 Celeriac 0 Celeriac 11 . 29703039 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 20 . 63208327 Chayote 20 . 63208327 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 32 . 36791673 Plantain 32 . 36791673 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavin g s spentsavin g s g ainedPeriod% land used# persons hiredsavin g s spentsavin g s g ained I100%5.58100%100%I100%3.74100%100% II100%3.350%0%II100%5.050%0% III100%2.080%0%III100%2.390%0% IV100%2.510%0%IV100%3.550%0% Year 4 crops in the ground05101520253035 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520253035 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 245

227 Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives Maximized income: $752,894Year 4Year 5 Celeriac 0 Celeriac 0 Banana 0 Banana 0 Beans 36 . 22784095 Beans 36 . 22784095 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 16 . 77215905 Chayote 16 . 77215905 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 36 . 22784095 Tanier 36 . 22784095 Period% land used# persons hiredsavin g s spentsavin g s g ainedPeriod% land used# persons hiredsavin g s spentsavin g s g ained I100%12.64100%0%I100%10.78100%100% II100%1.970%100%II100%1.970%0% III100%0.190%0%III100%0.070%0% IV100%16.800%0%IV100%16.800%0% Year 4 crops in the ground010203040 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $704,813 Y ear 4 Y ear 5 Celeriac 0 Celeriac 14 . 08688917 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 53 Plantain 53 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%0.78100%100%I100%0.94100%100% II100%0.960%0%II100%3.120%0% III100%0.400%0%III100%0.940%0% IV100%0.610%0%IV100%1.890%0% Year 4 crops in the ground0102030405060 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground0102030405060 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 246

228 Farm 3 149 acres Farmer & family hours: 25 hrs/week Full-time workers: 14 Part-time workers: 20 (12hrs/week) Annual income: $0-10,000 Scenario 1 : full incentives Scenario 2 : only workersÂ’ pay incentive ( no agrochemicals) NOT FEASIBLE Scenario 3 : no incentives NOT FEASIBLE maximized income: $1,521,940 Y ear 4 Y ear 5 Celeriac 55 . 38854864 Celeriac 0 Banana 0 Banana 0 Beans 0 Beans 55 . 38854864 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 93 . 61145136 Plantain 93 . 61145136 Pumpkin 33 . 89625604 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 55 . 38854864 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%29.96100%0%I100%18.65100%100% II100%10.030%0%II100%5.970%0% III100%16.720%0%III100%4.060%0% IV86%7.360%100%IV100%30.000%0% Year 4 crops in the ground020406080100 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground020406080100 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 247

229 Farm 4 27 acres (24 acres under cultivation) Farmer & family hours: 30 hrs/week Full-time workers: 0 Part-time workers: 0 Annual income: $0-10,000 Scenario 1 : full incentives Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) maximized income: $394,566 Y ear 4 Y ear 5 Celeriac 0 Celeriac 14 . 57305068 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 11 . 20182395 Chayote 11 . 20182395 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 15 . 79817605 Plantain 15 . 79817605 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%3.52100%100%I100%2.15100%100% II100%2.000%0%II100%4.190%0% III100%1.680%0%III100%1.180%0% IV100%1.440%0%IV100%3.100%0% Year 4 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $336,022 Y ear 4 Y ear 5 Celeriac 2 . 560350151 Celeriac 0 Banana 0 Banana 0 Beans 18 . 07018304 Beans 18 . 07018304 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 8 . 929816959 Chayote 8 . 929816959 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 18 . 07018304 Tanier 18 . 07018304 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%6.08100%0%I100%5.99100%100% II100%2.490%0%II100%1.260%0% III100%0.760%0%III100%0.310%0% IV100%9.080%100%IV100%8.640%0% Year 4 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 248

230 Scenario 3 : no incentives Farm 5 90 acres (29 acres under cultivation) Farmer & family hours: 30 hrs/week Full-time workers: 0 Part-time workers: 1 Annual income: $0-10,000 Scenario 1 : full incentives maximized income: $201,406 Y ear 4 Y ear 5 Celeriac 5 . 42466595 Celeriac 2 . 709733823 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 . 057441253 Ginger 0 . 057441253 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 2 . 714932127 Pumpkin 2 . 709733823 Yam 0 Yam 0 Tanier 24 . 22762662 Tanier 24 . 29026618 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%5.73100%69%I100%5.72100%100% II100%0.270%0%II100%0.280%0% III100%0.000%0%III100%0.000%0% IV20%2.960%31%IV10%2.380%0% Year 4 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $1,354,345 Y ear 4 Y ear 5 Celeriac 0 Celeriac 18 . 82927052 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 25 . 11988187 Chayote 25 . 11988187 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 64 . 88011813 Plantain 64 . 88011813 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%11.06100%100%I100%7.47100%100% II100%6.890%0%II100%9.730%0% III100%5.160%0%III100%5.980%0% IV100%5.740%0%IV100%7.490%0% Year 4 crops in the ground010203040506070 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040506070 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 249

231 Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives maximized income: $1,156,702 Y ear 4 Y ear 5 Celeriac 15 . 01947663 Celeriac 0 Banana 0 Banana 0 Beans 59 . 09436303 Beans 59 . 09436303 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 30 . 90563697 Chayote 30 . 90563697 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 59 . 09436303 Tanier 59 . 09436303 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%20.29100%0%I100%22.43100%100% II100%13.310%0%II100%6.140%0% III100%5.440%0%III100%2.960%0% IV100%32.790%100%IV100%30.210%0% Year 4 crops in the ground010203040506070 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040506070 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $675,380 Y ear 4 Y ear 5 Celeriac 5 . 429864253 Celeriac 2 . 714932127 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 2 . 714932127 Pumpkin 2 . 714932127 Yam 0 Yam 0 Tanier 87 . 28506787 Tanier 87 . 28506787 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%22.26100%58%I100%22.26100%100% II100%1.850%0%II100%1.850%0% III100%0.000%0%III100%0.000%0% IV6%10.050%42%IV3%9.470%0% Year 4 crops in the ground 020406080100 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground 020406080100 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 250

232 Farm 6 43 acres (25 acres under cultivation) Farmer & family hours: 45 hrs/week Full-time workers: 3 Part-time workers: 3 Annual income: $20-30,000 Scenario 1 : full incentives Scenario 2 : only workersÂ’ pay incentive (no agrochemicals) maximized income: $582,375 Y ear 4 Y ear 5 Celeriac 0 Celeriac 10 . 81969012 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 10 . 3938645 Chayote 10 . 3938645 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 32 . 6061355 Plantain 32 . 6061355 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%4.86100%100%I100%2.56100%100% II100%2.320%0%II100%3.980%0% III100%1.680%0%III100%1.890%0% IV100%1.740%0%IV100%2.960%0% Year 4 crops in the ground05101520253035 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520253035 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $474,241 Y ear 4 Y ear 5 Celeriac 17 . 14913694 Celeriac 0 Banana 0 Banana 0 Beans 22 . 12453734 Beans 33 . 69825201 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 9 . 301747986 Chayote 9 . 301747986 Coffee 0 Coffee 0 Ginger 11 . 57371467 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 22 . 12453734 Tanier 33 . 69825201 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%7.21100%0%I100%10.16100%100% II100%6.180%0%II100%1.350%0% III100%1.340%0%III100%0.000%0% IV100%12.870%100%IV100%15.360%0% Year 4 crops in the ground0510152025 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 251

233 Scenario 3 : no incentives Farm 7 40 acres (12 acres under cultivation) Farmer & family hours: 35 hrs/week Full-time workers: 0 Part-time workers: 3 Annual income: ~$10,000 Scenario 1 : full incentives maximized income: $252,104 Y ear 4 Y ear 5 Celeriac 12 . 53275218 Celeriac 6 . 430650593 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 . 853244096 Yam 0 Yam 0 Tanier 24 . 88179293 Tanier 38 . 11882251 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I73%5.77100%0%I100%9.46100%0% II76%1.060%0%II100%0.500%0% III72%0.000%0%III100%0.000%0% IV14%1.800%100%IV6%2.940%100% Year 4 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $503,725 Y ear 4 Y ear 5 Celeriac 0 Celeriac 0 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 1 . 873091986 Chayote 1 . 873091986 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 38 . 12690801 Plantain 38 . 12690801 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%2.84100%100%I100%1.47100%100% II100%1.430%0%II100%1.740%0% III100%0.860%0%III100%1.290%0% IV100%1.240%0%IV100%1.220%0% Year 4 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 252

234 Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives NOT FEASIBLE Farm 8 56 acres (40 acres under cultivation) Farmer & family hours: 70 hrs/week Full-time workers: 3 Part-time workers: 0 Annual income: $20-30,000 Scenario 1 : full incentives maximized income: $380,113 Y ear 4 Y ear 5 Celeriac 16 . 31344282 Celeriac 3 . 167420814 Banana 0 Banana 0 Beans 23 . 68655718 Beans 36 . 83257919 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 4 . 41876081 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 8 . 727261197 Pumpkin 3 . 167420814 Yam 0 Yam 0 Tanier 26 . 85397799 Tanier 36 . 83257919 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%6.94100%0%I100%8.91100%100% II100%2.070%0%II100%0.530%0% III100%1.700%0%III100%0.000%0% IV100%12.780%100%IV100%16.170%0% Year 4 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $785,379 Y ear 4 Y ear 5 Celeriac 0 Celeriac 16 . 14112453 Banana 0 Banana 0 Beans 0 Beans 0 . 287446285 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 12 . 89477725 Chayote 12 . 89477725 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 43 . 10522275 Plantain 42 . 81777646 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 . 287446285 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%2.68100%100%I100%1.30100%100% II100%0.920%0%II100%3.330%0% III100%0.060%0%III100%0.440%0% IV100%0.270%0%IV100%1.910%0% Year 4 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 253

235 Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives maximized income: $634,799 Y ear 4 Y ear 5 Celeriac 34 . 18274881 Celeriac 11 . 78480674 Banana 0 Banana 0 Beans 17 . 07983012 Beans 33 . 45608171 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 10 . 75911155 Chayote 10 . 75911155 Coffee 0 Coffee 0 Ginger 18 . 25318468 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 1 . 935876861 Pumpkin 7 . 971996794 Yam 0 Yam 0 Tanier 25 . 05182691 Tanier 37 . 26889166 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%6.68100%0%I100%9.35100%100% II100%6.060%0%II100%0.530%0% III100%0.000%0%III100%0.000%0% IV100%12.400%100%IV100%15.490%0% Year 4 crops in the ground0510152025303540 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $433,403 Y ear 4 Y ear 5 Celeriac 33 . 07760505 Celeriac 16 . 13936298 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 6 . 030985207 Pumpkin 12 . 1342135 Yam 0 Yam 0 Tanier 41 . 28056777 Tanier 49 . 78272733 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I98%7.43100%0%I100%7.58100%100% II100%0.850%0%II100%0.850%0% III95%0.000%0%III100%0.000%0% IV46%3.500%100%IV39%3.410%0% Year 4 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground0102030405060 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 254

236 Farm 9 124 acres (35 acres under cultivation) Farmer & family hours: 50 hrs/week Full-time workers: 2 Part-time workers: 2 Annual income: $20-30,000 Scenario 1 : full incentives Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) maximized income: $1,674,231 Y ear 4 Y ear 5 Celeriac 0 Celeriac 0 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 7 . 806925866 Chayote 7 . 806925866 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 116 . 1930741 Plantain 116 . 1930741 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%10.20100%100%I100%5.68100%100% II100%5.680%0%II100%6.230%0% III100%3.440%0%III100%5.190%0% IV100%4.740%0%IV100%4.650%0% Year 4 crops in the ground020406080100120140 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground020406080100120140 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $1,331,051 Y ear 4 Y ear 5 Celeriac 19 . 27312806 Celeriac 4 . 440061606 Banana 0 Banana 0 Beans 104 . 7268719 Beans 113 . 9387474 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 5 . 621190998 Chayote 5 . 621190998 Coffee 0 Coffee 0 Ginger 9 . 211875457 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 4 . 440061606 Yam 0 Yam 0 Tanier 109 . 1669335 Tanier 113 . 9387474 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%27.86100%0%I100%29.91100%0% II100%6.430%0%II100%2.580%0% III100%0.000%0%III100%0.000%0% IV100%49.830%100%IV100%50.820%100% Year 4 crops in the ground020406080100120 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground020406080100120 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 255

237 Scenario 3 : no incentives Farm 10 30 acres (25 acres under cultivation) Farmer & family hours: 60 hrs/week Full-time workers: 4 Part-time workers: 2 Annual income: $0-10,000 Scenario 1 : full incentives maximized income: $487,212 Y ear 4 Y ear 5 Celeriac 16 . 48861387 Celeriac 29 . 50675584 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 2 . 153947217 Pumpkin 0 Yam 0 Yam 0 Tanier 39 . 75692648 Tanier 59 . 33878308 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I31%5.40100%0%I45%9.28100%0% II34%2.880%0%II50%4.910%0% III32%0.000%0%III46%0.230%0% IV16%1.450%100%IV26%4.070%100% Year 4 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040506070 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $519,857 Y ear 4 Y ear 5 Celeriac 0 Celeriac 12 . 95308769 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 15 . 93703235 Chayote 15 . 93703235 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 14 . 06296765 Plantain 14 . 06296765 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%3.34100%100%I100%1.77100%100% II100%1.570%0%II100%3.510%0% III100%0.950%0%III100%0.490%0% IV100%0.860%0%IV100%2.340%0% Year 4 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 256

238 Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives maximized income: $464,400 Y ear 4 Y ear 5 Celeriac 4 . 733324925 Celeriac 0 Banana 0 Banana 0 Beans 13 . 40517241 Beans 13 . 40517241 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 16 . 59482759 Chayote 16 . 59482759 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 13 . 40517241 Tanier 13 . 40517241 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%5.16100%0%I100%5.30100%100% II100%3.570%100%II100%1.310%0% III100%0.860%0%III100%0.040%0% IV100%7.260%0%IV100%6.450%0% Year 4 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $349,690 Y ear 4 Y ear 5 Celeriac 8 . 179634766 Celeriac 7 . 884556414 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 5 . 437596551 Chayote 5 . 437596551 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 4 . 109790424 Pumpkin 4 . 069844341 Yam 0 Yam 0 Tanier 25 . 34302902 Tanier 21 . 56826303 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%4.44100%100%I100%3.89100%100% II100%1.330%0%II100%1.130%0% III100%0.000%0%III100%0.000%0% IV62%2.060%0%IV48%1.780%0% Year 4 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground0510152025 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 257

239 Farm 11 100 acres (60 acres under cultivation) Farmer & family hours: 65 hrs/week Full-time workers: 0 Part-time workers: 6 Annual income: $10-20,000 Scenario 1 : full incentives Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives NOT FEASIBLE maximized income: $1,262,851 Y ear 4 Y ear 5 Celeriac 0 Celeriac 0 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 100 Plantain 100 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%4.00100%100%I100%4.30100%100% II100%4.220%0%II100%4.960%0% III100%2.700%0%III100%4.270%0% IV100%4.120%0%IV100%3.390%0% Year 4 crops in the ground020406080100120 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground020406080100120 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $942,419 Y ear 4 Y ear 5 Celeriac 49 . 87380731 Celeriac 3 . 619909502 Banana 0 Banana 0 Beans 57 . 73469004 Beans 96 . 3800905 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 38 . 64540046 Pumpkin 3 . 619909502 Yam 0 Yam 0 Tanier 53 . 7461022 Tanier 96 . 3800905 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%16.17100%0%I100%24.44100%100% II100%6.520%0%II100%1.960%0% III100%10.840%0%III100%0.000%0% IV100%31.010%100%IV100%42.920%0% Year 4 crops in the ground010203040506070 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground020406080100120 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 258

240 Farm 12 33 acres (15 acres under cultivation) Farmer & family hours: 40 hrs/week Full-time workers: 1 Part-time workers: 1 Annual income: $20,000 Scenario 1 : full incentives Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) maximized income: $463,437 Y ear 4 Y ear 5 Celeriac 0 Celeriac 12 . 13269318 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 10 . 16183383 Chayote 10 . 16183383 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 22 . 83816617 Plantain 22 . 83816617 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%3.19100%100%I100%1.99100%100% II100%1.800%0%II100%3.620%0% III100%1.310%0%III100%1.240%0% IV100%1.300%0%IV100%2.560%0% Year 4 crops in the ground0510152025 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground0510152025 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $383,963 Y ear 4 Y ear 5 Celeriac 5 . 120153905 Celeriac 0 . 857599653 Banana 0 Banana 0 Beans 22 . 60700257 Beans 24 . 64470172 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 7 . 497698632 Chayote 7 . 497698632 Coffee 0 Coffee 0 Ginger 2 . 037699142 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 . 857599653 Yam 0 Yam 0 Tanier 23 . 46460223 Tanier 24 . 64470172 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%6.97100%0%I100%7.06100%100% II100%2.270%0%II100%0.930%0% III100%0.230%0%III100%0.000%0% IV100%11.080%100%IV100%11.150%0% Year 4 crops in the ground0510152025 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 259

241 Scenario 3 : no incentives Farm 13 25 acres (25 acres under cultivation) Farmer & family hours: 40 hrs/week Full-time workers: 3 Part-time workers: 0 Annual income: $? Scenario 1 : full incentives maximized income: $231,744 Y ear 4 Y ear 5 Celeriac 8 . 815438322 Celeriac 8 . 681897647 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 . 965436023 Pumpkin 3 . 255983979 Yam 0 Yam 0 Tanier 33 . 74896585 Tanier 30 . 62652255 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%6.14100%0%I100%5.86100%100% II100%1.380%0%II100%1.340%0% III100%0.000%0%III100%0.000%0% IV32%2.090%100%IV29%2.820%0% Year 4 crops in the ground0510152025303540 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520253035 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $329,077 Y ear 4 Y ear 5 Celeriac 0 Celeriac 10 . 62197195 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 8 . 173070732 Chayote 8 . 173070732 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 16 . 82692927 Plantain 16 . 82692927 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%2.40100%100%I100%1.41100%100% II100%1.270%0%II100%2.860%0% III100%0.940%0%III100%0.770%0% IV100%0.860%0%IV100%2.000%0% Year 4 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 260

242 Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives maximized income: $267,635 Y ear 4 Y ear 5 Celeriac 7 . 240845066 Celeriac 1 . 414747929 Banana 0 Banana 0 Beans 14 . 20252046 Beans 18 . 44627325 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 5 . 138978825 Chayote 5 . 138978825 Coffee 0 Coffee 0 Ginger 4 . 243752787 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 1 . 414747929 Yam 0 Yam 0 Tanier 15 . 61726839 Tanier 18 . 44627325 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%4.69100%0%I100%5.06100%100% II100%1.950%0%II100%0.580%0% III100%0.050%0%III100%0.000%0% IV100%7.500%100%IV100%8.240%0% Year 4 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $70,375 Y ear 4 Y ear 5 Celeriac 8 . 76690645 Celeriac 8 . 741645564 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 1 . 700685189 Pumpkin 1 . 603197461 Yam 0 Yam 0 Tanier 6 . 800338944 Tanier 8 . 773050676 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I45%1.00100%0%I48%1.09100%0% II47%0.080%0%II52%0.400%0% III43%0.000%0%III48%0.000%0% IV24%0.320%100%IV29%0.310%100% Year 4 crops in the ground0246810 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground0246810 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 261

243 Farm 14 33 acres (15 acres under cultivation) Farmer & family hours: 42 hrs/week Full-time workers: 1 Part-time workers: 1 Annual income: $10-20,000 Scenario 1 : full incentives Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) maximized income: $447,636 Y ear 4 Y ear 5 Celeriac 0 Celeriac 5 . 106583164 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 6 . 044431784 Chayote 6 . 044431784 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 26 . 95556822 Plantain 26 . 95556822 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%3.82100%100%I100%1.45100%100% II100%1.280%0%II100%2.090%0% III100%0.800%0%III100%1.130%0% IV100%0.870%0%IV100%1.560%0% Year 4 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $359,615 Y ear 4 Y ear 5 Celeriac 19 . 80212307 Celeriac 1 . 20726416 Banana 0 Banana 0 Beans 11 . 97136524 Beans 25 . 34113023 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 6 . 451605612 Chayote 6 . 451605612 Coffee 0 Coffee 0 Ginger 13 . 36976498 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 1 . 20726416 Yam 0 Yam 0 Tanier 13 . 1786294 Tanier 25 . 34113023 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%4.25100%0%I100%7.44100%100% II100%5.670%0%II100%0.820%0% III100%1.000%0%III100%0.000%0% IV100%8.500%100%IV100%11.370%0% Year 4 crops in the ground0510152025 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 262

244 Scenario 3 : no incentives Farm 15 69 acres (40 acres under cultivation) Farmer & family hours: 12 hrs/week Full-time workers: 5 Part-time workers: 3 Annual income: $10-20,000 Scenario 1 : full incentives maximized income: $196,366 Y ear 4 Y ear 5 Celeriac 9 . 172889893 Celeriac 11 . 41105979 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 1 . 627742427 Pumpkin 0 . 87971087 Yam 0 Yam 0 Tanier 18 . 32910014 Tanier 25 . 81148265 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I62%2.89100%0%I96%5.81100%0% II65%0.870%0%II100%1.140%0% III62%0.000%0%III95%0.000%0% IV27%0.890%100%IV19%1.980%100% Year 4 crops in the ground05101520 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground051015202530 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $952,498 Y ear 4 Y ear 5 Celeriac 0 Celeriac 10 . 01533232 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 14 . 70797376 Chayote 14 . 70797376 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 54 . 29202624 Plantain 54 . 29202624 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%11.22100%100%I100%5.33100%100% II100%4.990%0%II100%6.610%0% III100%3.950%0%III100%4.480%0% IV100%4.010%0%IV100%5.550%0% Year 4 crops in the ground0102030405060 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground0102030405060 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 263

245 Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives NOT FEASIBLE Farm 16 70 acres (40 acres under cultivation) Farmer & family hours: 40 hrs/week Full-time workers: 5 Part-time workers: 2 Annual income: $10-20,000 Scenario 1 : full incentives maximized income: $781,153 Y ear 4 Y ear 5 Celeriac 5 . 689432731 Celeriac 0 Banana 0 Banana 0 Beans 57 . 95599656 Beans 57 . 95599656 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 11 . 04400344 Chayote 11 . 04400344 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 57 . 95599656 Tanier 57 . 95599656 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%16.80100%0%I100%17.70100%100% II100%5.800%0%II100%3.080%0% III100%1.960%0%III100%1.030%0% IV100%28.320%100%IV100%27.350%0% Year 4 crops in the ground010203040506070 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground010203040506070 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $1,027,369 Y ear 4 Y ear 5 Celeriac 0 Celeriac 30 . 93823854 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 25 . 79935028 Chayote 25 . 79935028 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 44 . 20064972 Plantain 44 . 20064972 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 0 Tanier 0 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%9.27100%100%I100%6.13100%100% II100%5.720%0%II100%10.370%0% III100%4.750%0%III100%4.060%0% IV100%4.460%0%IV100%7.850%0% Year 4 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 264

246 Scenario 2 : only workersÂ’ pay incent ive (no agrochemicals) Scenario 3 : no incentives maximized income: $874,097 Y ear 4 Y ear 5 Celeriac 9 . 551095566 Celeriac 0 Banana 0 Banana 0 Beans 45 . 95008294 Beans 45 . 95008294 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 24 . 04991706 Chayote 24 . 04991706 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 0 Yam 0 Yam 0 Tanier 45 . 95008294 Tanier 45 . 95008294 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%15.95100%0%I100%16.77100%100% II100%8.920%0%II100%4.360%0% III100%3.490%0%III100%1.870%0% IV100%24.720%100%IV100%23.080%0% Year 4 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground01020304050 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation maximized income: $485,749 Y ear 4 Y ear 5 Celeriac 7 . 36436401 Celeriac 11 . 73212746 Banana 0 Banana 0 Beans 0 Beans 0 Cabbage 0 Cabbage 0 IPM Cabbage 0 IPM Cabbage 0 Cassava 0 Cassava 0 Chayote 0 Chayote 0 Coffee 0 Coffee 0 Ginger 0 Ginger 0 Papaya 0 Papaya 0 Plantain 0 Plantain 0 Pumpkin 0 Pumpkin 2 . 345739143 Yam 0 Yam 0 Tanier 74 . 36776345 Tanier 67 . 65426086 Period% land used# persons hiredsavings spentsavings gainedPeriod% land used# persons hiredsavings spentsavings gaine I100%15.72100%0%I100%15.44100%100% II100%3.040%0%II100%2.810%0% III100%0.000%0%III100%0.000%0% IV17%5.750%100%IV17%7.330%0% Year 4 crops in the ground01020304050607080 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation Year 5 crops in the ground020406080 Celeriac Banana Beans Cabbage IPM Cabbage Cassava Chayote Coffee Ginger Papaya Plantain Pumpkin Yam TanierCropsArea under cultivation

PAGE 265

247 LIST OF REFERENCES Ables, J. R., D. W. McCommas, S. L. Jones, and R. K. Morrison. 1980. Effects of cotton plant size, host egg lo cation, and location of parasi te release on parasitism by Trichogramma pretiosum . Southwest. Entomol. 5: 261–264. Abreu–Rodriguez E., and C. Cruz. 1997 . Enemigos naturales de Plutella xylostella (L.) (Lepidoptera: Plutellidae) en Puerto Rico. J. Agric. Univ. P.R. 81(1–2): 83–85. Agricultural Experimental Statio n, University of Puerto Rico, 1999. Conjunto technologico para la produccion de repo llo. Publicación 158. Impreso en la Oficina de Publicaciones, Estación Experimental Ag rícola, Colegio de Ciencias Agrícolas, Recinto Universitario de Mayagüez, Universidad de Puerto Rico. Aldrich, J. R., R. J. Bartlet, J. C. Dickens, A. L. Knight, D. M. Light, and J. H. Tumlinson. 2003. Insect chemical ecology rese arch in the United States Department of Agriculture – Agricultural Research Service. Pest Manag. Sci. 59: 777–787. Allen, J. C., and D. Gonzalez. 1974 . Spatial attack patterns of Trichogramma pretiosum around release sites compared with a ra ndom diffusion model. Environ. Entomol. 3(4): 647–652. Altieri, M. A., S. Annamalai, K. P. Katiyar, and R. A. Flath. 1982 . Effects of plant extracts on rates of parasitization of Anagasta kuehniella (Lep: Pyralidae) eggs by Trichogramma pretiosum (Hym.: Trichogrammatidae) under greenhouse conditions. Entomophaga. 27: 431–438. Altieri, M. A., W. J. Lewis, D. A. Norlund, R. C. Gueldner, and J. W. Todd. 1981 . Chemical interactions between plants and Trichogramma wasps in Georgia soybean fields. Prot. Ecol. 3: 259–263. Andaloro, J. T., K. B. Rose, A. M. Shelton, C. W. Hoy, and R. F. Becker. 1983. Cabbage growth stages. New York’s F ood and Life Sciences Bulletin. Number 101. Andow, D. A., and D. R. Prokrym. 1990. Plant structural comp lexity and host–finding by a parasitoid. Oecologia. 82:162–165. Armstrong, A. 1990 . Susceptibility of the diamondback moth, Plutella xylostella (L.) (Lepidoptera: Yponomeutidae), to insecticides sprays on cabbage. J. Agric. Univ. P.R., 74 (2): 153–166.

PAGE 266

248 Ashley, C., and D. Carney. 1999 . Sustainable Livelihoods: Lessons from early experience. Department for In ternational Development. UK. Bakker, E. J., H. Heng sdijk, and K. Sissoko . 1998. Sustainable land use in the Sudano– Sahelian zone of Mali: Exploring economica lly viable options using multiple goal linear programming. Neth. J. Agr. Sci. 46 (1): 109–122. Bar, D., D. Gerling, and Y. Rossler. 1979. Bionomics of the principal natural enemies attacking Heliothis armigera in cotton fields in Isr ael. Environ. Entomol. 8: 468– 474. Bennett, F. D., and M. Yaseen. 1972 . Parasite introductions for the biological control of three insect pests in the Lesser Antill es and British Honduras. Pans 18(4): 468–474. Biggs, S. 1989 . Resource–Poor Farmer Participa tion in Research: A Synthesis of Experiences from Nine National Agri cultural Research Systems, OFCOR Comparative Study Paper 3. Internationa l Service for National Agricultural Research, The Hague, The Netherlands. Bigler, F., M. Bieri, A. Fritschy, and K. Seidel. 1988 . Variation in locomotion between laboratory strains of Trichogramma maidis and its impact on parasitism of Ostrinia nubilalis in the field. Entomol. Exp. Appl. 49: 283–290. Bin, F., S. B Vinson, M.R. Strand, S. Colazza, and W. A. Jones. 1993 . Source of an egg kairomone for Trissolcus basalis , a parasitoid of Nezara viridula . Physiol. Entomol. 18: 7–15. Bjorkstgen, T. A., and A. A. Hoffmann. 1995. Effects of pre– adult and adult experiences on host acceptance in choice and non–choice tests in two strains on Trichogramma. Entomol. Exp. Appl. 1995. 76 (1): 49–58. Bjorkstgen, T. A., and A. A. Hoffmann. 1998 . Persistence of experience effects in the parasitoid Trichogramma nr. brassicae . Ecol. Entomol. 23 (2) 110–117. Boethel, D. J., J. S. Mink, A. T. Wier, J. D. Thomas, B. R. Leonard, and F. Gallardo. 1992 . Management of insecticide resistant soybean loopers ( Pseudoplusia includens ) in the southern United States, pp. 66–87. In L. G. Copping, M. B. Green, and R. T. Rees [eds.], Pest management in soybean. Society of Chemical Industry, Essex, England. Boo, K. S., and J. P. Yang. 2000 . Kairomones used by Trichogramma chilonis to find Helicoverpa assulta eggs. J. Chem. Ecol. 26 (2): 359–375. Booij, C. J. H., and J. Noorlander. 1992. Farming systems and insect predators. Agr. Ecosyst. Environ.: 40: 125–135. Boswell, V. 1949 . Our Vegetable Travelers. Nationa l Geographic Magazine. 96 (2): 145– 217.

PAGE 267

249 Burbutis, P. P., G. D. Curl, and C. P. Davis. 1977 . Host searching behavior by Trichogramma nubilale on corn. Environ. Entomol. 6 400–402. Cabello, T., and P. Vargas. 1985 . Estudio con olfactomometro de la influencia de la planta y del insecto huesped en la actividad de busqueda de Trichogramma cordubensis Vargas Y Cabello y de T. sp. p . buesi (Hyme. Trichogrammatidae). Boletin del Servicio de Defensa contra Plagas e Inspección F itopatológia. 11: 237– 241. Cabrera, V. E. 1999 . Farm problems, solutions, and extension programs for small farmers in Cañete, Lima, Peru. M. S. Thesis. University of Florida. 29 Nov. 2003. http://etd.fcla.edu/etd/ uf/1999/amj9816/cabrera.pdf Calvin, D. D., J. E. Losey, M. C. Knapp, and F. L. Poston. 1997. Oviposition and development of Trichogramma pretiosum (Hymenoptera: Tr ichogrammatidae) in three age classes of southw estern corn borer eggs. Environ. Entomol. 26 (2): 385– 390. Capinera, J. L. 2002 . Diamondback moth, Plutella xylostella (Linnaeus) (Insecta: Lepidoptera: Plutellidae). Featured Cr eatures EENY–119. University of Florida Department of Entomology and Nema tology. Florida Cooperative Extension Service. 29 Nov. 2003. http://edis. ifas.ufl.edu/pdffiles/IN/IN27600.pdf Carro–Figueroa, V. 1985 . Farming systems research and the small farmers in Puerto Rico: possibilities and constr aints of the perspective as a developmental approach. In Proceedings “Caribbean farming system s and alternatives for development”. Université Antilles–Guyane, Développe ment Agricole Caraibe. 9–11 May 1985. Carro Figueroa, V., and C. Alamo González. 2002 . Cambio en la estructura agrícola de la región central de Puerto Rico: tipos de finca y características de sus operaciones. University of Puerto Rico, Mayaguez cam pus, College of Agricultural Sciences, Agricultural Experimental Station, San Juan, Puerto Rico. Bulletin 302. Chambers, R. 1994 . The origins and practice of part icipatory rural appraisal. World Develop. 22, 953–969. Chambers, R., A. Pacey, and L. A. Thrupp. 1989 . Farmer First: Farmer Innovation and Agricultural Research . London: IT Publications. Charnov, E. L., and S. W. Skinner. 1985 . Complementary approaches to the understanding of parasitoid oviposition decisions. Environ. Entomol. 14: 383–391. Clayton, E. 1963 . Economic planning in peasant agri culture. Department of Economic Monographs. University of London. Wye College. UK. Cloutier, C., and Baudin F. 1990 . Searching behavior of the aphid parasitoid Aphidius nigripes (Hymenoptera: Aphidiidae) foraging on potato plants. Environ. Entomol. 19: 222–228.

PAGE 268

250 Collinson, M. 2000 . Introduction, Chapter 1, pp 1–4. In M. Collinson [ed.], A History of Farming Systems Research. FAO/ CABI. London, New York, Rome.\ Concepción, C. M. 1995 . The Origins of Modern Environmental Activism in Puerto Rico in the 1960s. Int. J. Urban Regional. 19: 112–28. Consoli, F. L. 1997 . Criaco in vitro de Trichogramma galloi Zucchi, 1988 e T. pretiosum Riley, 1879 (Hym.: Trichogrammatidae): dese nvolvimento de um ovo artificial e aprimoramento de dietas artificiais. Tese (Doutorado). Escola Superior de Agricultura `Luiz de Queiroz', Universida de de Sao Paulo, Piracicaba, Brasil, p. 153. Consoli, F. L., E. W. Kitajima, and J. R. P. Parra. 1999 . Ultrastructure of the natural and factitious host eggs of Trichogramma gallao Zucchi and Trichogramma pretiosum Riley (Hymenoptera: Trichogrammatidae). Int. J. Insect Morphol. 28 (3): 211–231. Cornwall, A., and R. Jewkes. 1995. What is participatory research? Soc. Sci. Med. 41: 1667–1676. Cortesero, A. M., and J. P. Monge. 1994 . Influence of pre–emergence experience on response to host and host plant odour s in the larval parasitoid Eupelmus vuilleti . Entomol. Exp. Appl. 72: 281–288. Davis Jnr., L. R., R. K. Vander Meer, and S. D. Porter. 2001 . Red imported fire ants expand their range across the West I ndies. Florida Entomol. 84 (4): 735–736. De Jong, E. J., and Pak G. A. 1984. Factors determining differential host–egg recognition of the two host species by different Trichogramma spp. Mededelingen van de Faculteit Landbouwwetenscha ppen Rijksunivers iteit Gent 49, 815–825. Department for International Development. 2000 . Sustainable Livelihoods Guidance Sheets. DFID. London, UK. Dicke, M., and J. J. A. van Loon. 2000. Multitrophic effects of herbivore–induced plant volatiles in an evolutionary contex t. Entomol. Exp. Appl. 97 (3): 237–249. Dietz, J. L. 1986. Economic History of Puerto Rico: In stitutional Change and Capitalist Development. Princeton, NJ. Pr inceton University Press. Dixon, J., A. Gulliver, and D. Gibbon. 2001 . Global Farming Systems Study: Challenges and Priorities to 2030. The Global Farming Systems Study. FAO. Pgs 98. Dlott, J. W., M. A. Altieri, and M. Masumoto. 1994. Exploring the theory and practice of participatory resear ch in US sustainable agriculture : a case study in insect pest management. Agriculture and Human Values. Spring–Summer: 126–139.

PAGE 269

251 Economic Research Service, USDA 2003 . World Table 1––Selected vegetable production in leading countries and the world, 1993–2002. 2003 Yearbook. Economic Research Service, United Stat es Department of Agriculture. 29 Nov. 2003. http://www.ers.usda.gov/publica tions/vgs/tables/world.xls. Evenson, R. E., and D. Gollin. 2002 . Crop Variety Improvement and its Effect on productivity: the Impact of Interna tional Research. Wallingford, UK: CAB International. Farrington, J. 2000 . The development of diagnostic methods in FSR. pp 59–67. In M. Collinson [ed.], A History of Farming Systems Research. Food and Agriculture Organization of the United Na tions and CABI Publishing. Feder, G., R. Just, and D. Zilberman. 1985 . Adoption of agricultural innovations in developing countries: a survey. Economic Development and Social Change. 32 (2): 255–298. Fischer, S., J. Samietz, F. L. Wäckers, and S. Dorn. 2001 . Interaction of vibrational and visual cues in parasitoid host location. J. Comp. Physiol. A. 187: 785–791. Florida Agricultural Statistics Service . 2003 . Vegetables—Acreage, Production and Value. 29 Nov. 2003. http://www.nass.usda.gov/fl/rtoc0cr.htm Food and Agriculture Organizatio n of the United Nations, 2002 . Production Yearbook (FAO, Rome, 2002). 29 Nov. 2003. http://apps.fao.org/lim500/nph– wrap.pl?Production.Crops.P rimaryandDomain=SUA Food and Agricultural Organizat ion of the United Nations. 2003 . Farming Systems and Poverty. Analysis of farming systems. 29 Nov. 2003. http://www.fao.org/farmingsystems/description_en.htm Frenoy, C., C. Durier, and N. Hawlitzsky. 1992 . Effect of kairomones from egg and female adults stages of Ostrinia nub ilalis (Hubner) (Lepidop tera, Pyralidae) on Trichogramma brassicae Bezdenko (Hymenoptera, Trichogrammatidae) female kinesis. J. Chem. Ecol. 18 (5): 761–773. Fye, R. E., and D. J. Larsen. 1969. Preliminary evaluation of Trichogramma minutum as a released regulator of lepidopterous pests of co tton. J. Econ. Entomol. 62: 1291–1296. Geitzenauer, H. L., and E. A. Bernays. 1996. Plant effects on prey choice by a vespid wasp, Polistes arizonensis. Ecol. Entomol. 21: 227–234. Genung, W. G., 1957 . Some possible cases of insect resistance to insecticides in Florida. Proc. Florida State Hort. Soc. 70:148–152.

PAGE 270

252 Gianessi, L. P., C. S. Silvers, S. Sankula, and J. E. Carpenter, 2002. Plant biotechnology: Current and pot ential impact for improving pest management in US agriculture, an analysis of 40 case studi es – Insect Resistant Soybean. National Center for Food and Agricultur al Policy, June 2002. 29 Nov. 2003. http://www.ncfap.org/ Gibbon, D., A. Lake, and M. Stocking. 1995 . Sustainable Development: A Challenge for Agriculture.Chapter 2. Pp 31–68. In S. Morse and M. Stocking (eds.) People and Environment. UCL Press. England. Gingras, D., and G. Boivin. 2002 . Effect of plant structure, host density and foraging duration on host finding by Trichogramma evanescens (Hymenoptera: Trichogrammatidae). Environ. Entomol. 31 (6): 1153–1157. Gingras, D., P. Dutilleul, and G. Boivin. 2002 . Modeling the impact of plant structure on host–finding behavior of para sitoids. Oecologia. 130: 396–402. Gingras, D., P. Dutilleul, and G. Boivin. 2003 . Effect of plant st ructure on host finding capacity of lepidopterous pests of cruc ifers by two Trichogramma parasitoids. Biological control: theory and applica tions in pest management. 27 (1): 25–31. Glass, P., Smits P., Vlaming P., and Van Lenteren J. 1981. Biological control of lepidopteran pests in cabbage crops by means of inundative release of Trichogramma species ( Trichogramma evanescens and Trichogramma cacoecine ): A combination of field and laborator y experiments. Med. Fac. Landbounet. Rijksuniv. Gent. 46, 487–498. Godin C., and Boivin G. 2000 . Effects of host age on parasitism and progeny allocation in Trichogrammatidae. Entomol. Exp. Appl. 97: 149–160. Gonzalez, D., G. Orphanides, R. van den Bosch, and T. F. Leigh. 1970 . Field cage assessment of Trichogramma as parasites of Heliothis zea : development of methods. J. Econ. Entomol. 63: 1292–1296. Greene, G. L. 1968 . Distribution of Tricholplusia ni eggs and larvae on cabbage plants as a basis for sampling efficiency. J. Econ. Entomol. 61: 1648–1650. Grosfoguel, R. 1994 . World Cities in the Carribean: the Rise of Miami and San Juan. Review—The Fernand Braudel Center fo r the study of Economies, Historical Systems, and Civilizations. 17 (3): 351–81. Guptill, A. E. 2001 . “Barriers to Organic Agricultu re in Puerto Rico Under US Colonialism”. Unpublished. 12 Oct. 2003. h ttp://people.cornell.edu/pages/aeg9/ Guy, R. H., N. C. Leppla, and J. R. Rye. 1984 . Trichoplusia ni . In P. Singh and R.F. Moore [eds.]. Handbook of Insect Rearing Vol. II. Elsevier Science Publishers, Amsterdam.

PAGE 271

253 Harcourt, D. G. 1961 . Design of a sampling plan for studies on the population dynamics of the diamondback moth. Plutella maculipennis (Curt.) (Lepidoptera: Yponomeutidae). Can. Entomol. 93: 820–831. Harcourt, D. G. 1962 . Design of a sampling plan for studies on the population dynamics of the imported cabbageworm, Pieris rapae (L.) (Lepidoptera: Pieridae). Can. Entomol. 94: 849–859. Hassan, S. A. 1993. The mass rearing and utilization of Trichogramma to control lepidopterous pests: achievements and outlooks. Pestic. Sci. 37: 387–391. Hassan, S. A., and M. F. Guo. 1991 . Selection of effective stra ins of egg parasites of the genus Trichogramma (Hym., Trichogrammatidae) to control the European corn borer Ostrinia nubilalis Hb. (Lep., Pyralidae). J. Appl. Ent. 111: 335–341. Hazell, P. B. R., and R. D. Norton. 1986 . Mathematical Programming for Economic Analysis in Agricultur e. McMillan. New York. Heady, E. O., and W. Candler. 1958 . Linear Programming Methods. Iowa State University Press. Hendricks, D. E. 1967 . Effects of wind on dispersal of Trichogramma semifurnatum . J. Econ. Entomol. 60. 1367–1373. Hildebrand, P. 1998 . Introduction to Linear Progra mming – Economic Analysis in Small Farm Livelihood Systems. Class notes . University of Florida. Food and resource Economic Department. Hildebrand, P. E., and A. A. Araújo. 1997 . Introduction to linear programming. Gainesville, USA: The Florida Book Store. Hirai, K. 1991. Trichogramma parasitising Pieris rapae crucivora Boisduval. Colloques de INRA 56: 71–72. Ikerd, J., G. Devino, and S. Traiyongwanich. 1996. Evaluating the sustainability of alternative farming systems: A cas e study. Am. J. Alternative Agr.: 11 (1): 25–29. Inglis, G. D., and A. M. Lawrence. 2001. Effects of Serratia marcescens on the F1 generation of laboratory–reared Heliothis virescens (Lepidoptera: Noctuidae). J. Econ. Entomol. 94 (2): 362–366. Jackson, C. G., G. D. Butler, Jr., and D. E. Bryan. 1969 . Time required for development of Voria ruralis and its host, the cabbage looper, at different temperatures. J. Econ. Entomol. 62:69–70 Jensen, R. L., L. D. Newsom, and J. Gibbens. 1974. The soybean looper: effects of adult nutrition on oviposition, mating freque ncy, and longevity. J. Econ. Entomol. 67 (4): 467–470.

PAGE 272

254 Jervis, M. A., B. A. Hawkins, and N. A. C. Kidd. 1996 . The usefulness of destructive host feeding parasitoids in classical bi ological control: th eory and observation conflict. Ecol. Entomol. 21 (1): 41–46. Jervis, M. A., and N. A. C. Kidd. 1986 . Host–feeding strategies in hymenoptera parasitoids. Biol. Rev.. 61 (4): 395–434. Kanour, W. W., and P. P. Burbutis. 1984. Trichogramma nubilale (Hym. Trichogrammatidae) field releas es in corn and a hypotheti cal model for control of European corn borer (Lep.: Py ralidae). Econ. Entomol. 77: 103–107. Karowe, D. N., and L. M. Schoonhoven. 1992 . Interactions among three trophic levels: the influence of host plant on performance of Pieris brassicae and its parasitoid, Cotesia glomerata . Entomol. exp. appl. 62 (3). 241–251. Kassa, H., and D. Gibbon. 2003 . Does the Sustainable Livelihoods Approach need a more explicit systems perspective? Syst ems dynamics modeling to facilitate entry points to smallholder farming systems. AGRIPPA – Peer Reviewed Electronic Journal – Farming Systems. FAO. 29 Nov. 2003. http://www.fao.org/Agrippa/ Keating, B. A., and R. L. McCown. 2001 . Advances in farming systems analysis and intervention. Agr. Syst. 70: 559–579. Keinmeesuke, P., A. Vattanatangum, O. Sarnthoy, B Sayampol, T. Miyata, T. Saito, F. Nakasuji, and N. Sinchaisri. 1993. Life table of diamondback moth and its egg parasite Trichogrammatoidea bactrae in Thailand, pp 309–315. In Talekar [ed.], Management of diamondback moth and othe r crucifer pests: Proceedings of the second international workshop. Asian Ve getable Research and Development Center, Shanhua, Taiwan. Keller, M. A., W. J. Lewis, and R. E. Stinner. 1985 . Biological and practical significance of movement by Trichogramma species: a review. Southwest. Entomol. 8: 138–155. Kishlansky, M., P. Geary, and P. O’Brien. 2001 . A Brief History of Western Civilization, Vol. 1: The Unfinish ed Legacy. Pearson Longman. Pgs 432. Klemm, U. 1992. Selection of Effective Species or Strains of Trichogramma Egg Parasitoids of DBM. Chapter 35 In : Talekar N.S. [ed.] Diamondback Moth and Other Crucifer Pests "Proceedings of th e Second International Workshop, Tainan, Taiwan, 10–14 December 1990" 603 pp. Asian Vegetable Research and Development Center. Klemm, U., and Schmutterer H. 1993 . Effects of neem preparations on Plutella xylostella L. and its natural enemies of th e genus Trichogramma. Zeitschrift fur Pflanzenkrankheiten und Pflanzenschutz 100 , 113–128.

PAGE 273

255 Klomp, H., and Teerink B. J. 1962. Host selection and number of eggs per oviposition in the egg parasites Trichogramma embryophagum Htg. Nature 195, 1020–1021. Kogan, M., and D. Cope. 1974 . Feeding and nutrition of insects associated with soybeans. 3. Food intake, utilization, and growth in the soybean looper, Pseudoplusia includens . Ann. Entomol. Soc. Am. 67 (1): 66–72. Kot, J. 1964 . Experiments in the biology and ecology of species of the genus Trichogramma Westw. And their use in plant protection. Ekol. Pol.–Pol. J. Ecol. 12. 243–303. Krishnamoorthy, A., and M. Mani. 1999 . Biological control of cabbage diamondback moth, Plutella xylostella by an egg parsitoid, T richogrammatoidea bactrae Nagaraja. Paper presented at the Inte rnational Seminar on Integrated Pest Management held at IICT, Hyderabad between 8 and 9th October. 1999. Krishnamoorthy, A., N. Rama, and M. Mani. 2002. Biological control of Diamondback moth, Plutella xylostella L on cabbage using an egg parasitoid, Trichogrammatoidae bactrae and mustard as a trap crop. J. Biol. Control. (In Press). Krnjajic, S., N. Dimic, P. Peri c, M. Vuska, and M. Cvekovic. 1997. Biological control of cabbage pests. Acta Hort. 462: 119–124. Landolt, P. J. 1997 . Cabbage looper (Lepidoptera: Noct uidae) fecundity maximized by a combination of access to water and food and remating. Ann. Entomol. Soc. Am. 90 (6): 781–789. Leatemia, J. A., J. E. Laing, and J. E. Corrigan. 1995 . Effects of adult nutrition on longevity, fecundity, and offspring sex ration of Trichogramma minutum Riley (Hymenoptera: Trichogrammatidae). Can. Entomol. 127 (2): 245–354. Leibee, G. 1996 . Insect pest management in crucifers in Florida. Pest Management in the Subtropics. Integrated pest Manage ment – a Florida Perspective. 413–422. Leibee, G., and J. L. Capinera. 1995 . Pesticide resistance in Florida insects limits management options. Fl. Entomol. 78(3): 386–399. Leibee, G., and K. E. Savage. 1992a. Evaluation of selected in secticides for control of diamondback moth and cabbage looper in cabbage in central Florida with observations on insecticide re sistance in the diamondback moth. Florida Entomol. 75: 585–591. Leibee, G., and K. E. Savage. 1992b. Observations on insecticide resistance in diamondback moth, pp. 41–46 In Seminar Proceedings: Global Management of Insecticide Resistance in the 90s. A bbot Laboratories. Sept. 15–17, 1992. Lake Bluff, IL.

PAGE 274

256 Lewis, W. J., J. O. Stapel, A. M. Cortesero, and K. Takasu. 1998 . Understanding how parasitoids balance food and host needs: Importance to biological control. Biol. Control. 11 (2) 175–183. Lewis, W. J., and J. H. Tumlinson. 1988 . Host detection by chemically mediated associative learning in a parasi tic wasp. Nature (London). 331: 257–259. Li–Ying, L. 1994 . Worldwide use of trichogramma fo r biological control on different crops: a survey, pp. 37–54. In Wajnberg E. Hassan S. (eds.) “Biological Control with Egg Parisitoids”. CAB International, Oxon. Lukianchuk, J. L., and S. M. Smith. 1997 . Influence of plant structural complexity on the foraging success of Trichogramma minutum Riley: A comparison of search on artificial and foliage models. Entomol. exp. appl. 84: 221–228 . Lundgren, J. G., G. E. Heimpel, and S. A. Bomgren. 2002 . Comparison of Trichogramma brassicae (Hymenoptera: Trichogramma tidae) augmentation with organic and synthetic pesticides for cont rol of cruciferous Lepidoptera. Environ. Entomol. 31 (6): 1231–1239. Maini, S., and G. Burgio. 1990. Biological control of the European corn borer in protected pepper by Trichogramma maidis Pint. and Voeg. and Bacillus thuringiensis Berl, pp. 213–216. In E. Wajnberg and S. B. Vinson [eds.], Trichogramma and other egg parasitoids. Proceedings, 3rd International Symposium, Sept. 23–27, 1990, San Antonio, TX. Les Colloques, No 56. Institut National de la Recherche Agronomique (INRA), Paris. Mansfield, S., and N. J. Mills. 2002. Host egg characteristics, physiological host range and parasitism following inundative release of Trichogramma platneri (Hymenoptera: Trichogrammatidae) in walnut orchards. Environ. Entomol. 31 (4): 723–731. Martin, P. B., P. D. Lingren, G. L. Greene, and R. L. Ridgeway. 1976 . Parasitization of two species of Plusiinae and Heliothis spp. after releases of Trichogramma pretiosum in seven crops. Environ. Entomol. 5(5): 991–995. Matteson, P. C., M. A. Altieri, and W. C. Gagne. 1984. Modification of small farmer practices for better pest management. Ann. Entomol. Soc. Am. 29: 383–402. Maynard, D. N., Hochmuth, G. J., Vavrin a, C. S., Stall, W. M., Kucharek, T. A., Stansly, P. A., Taylor, T. G., Smith, S. A., and Smajstrla, S. A. 1998. Cole Crop Production in Florida. University of Flor ida, Institute of Food and Agricultural Sciences. 29 Nov. 2003. http://edis.ifas.ufl.edu/CV122. McGregor, R. 1997 . Host feeding and oviposition by pa rasitoids on hosts of different fitness value: influences of egg load a nd encounter rate. J. Insect Behav. 10 (3): 451–462.

PAGE 275

257 Meier, N. 1941 . Trichogramma: Ecology and Results for Utilisation for the Control of Injurious Insects. Selkhozgiz , Moscow. 175 pp. Mitchell, E. R. 1967 . Life history of Pseudoplusia includens (Walker) (Lepidoptera: Noctuidae). J. Georgia Entomol. Soc. 2:53–57. Mitchell, E. R., Hu G. Y., and Okine J. S. 1997 . Diamondback moth (Lepidoptera: Plutellidae) infestati on and parasitism by Diadegma insulare (Hymneoptera: Ichneumonidae) in collards and adjacen t cabbage fields. Florida Entomol. 80 (1) 54–62. Miura, K., and Kobayashi M. 1995 . Reproductive properties of Trichogramma chilonis females on Diamondback Moth eggs. Appl. Entomol. Zool. 30 (3): 393–400. Monje, J. C., C. P. W. Zebi tz, and B. Ohnesorge. 1999. Host and host age preference of Trichogramma galloi and T. pretiosum (Hymenoptera: Trichogrammatidae) reared on different hosts. J. Econ. Entomol. 92 (1): 97–103. Morallo–Rejesus, and Sayaboc 1992 . Management of diamondback moth with Cotesia plutellae : Prospects in the Philippines. Pr oceedings of the 2nd International Workshop, AVRDC, Tainan, Taiwan. 279–286. Muckenfuss, A. E., Shepard B. M., and Ferrer E. R., 1992. Natural mortality of diamondback moth in coastal South Carolin a. Proceedings of the 2nd International Workshop, AVRDC, Tainan, Taiwan. 27–36. Nagarkatti, S., A. J. Muza, M. C. Saunders, and P. C. Tobin. 2002. Role of the egg parasitoid Trichogramma minutum in biological control of the grape berry moth, Endopiza viteana. BioControl 47: 373–385. Naranjo, S. E. 1993 . Life history of Trichogrammatoidea bactrae (Hymenoptera: Trichogrammatidae), an e gg parasitoid of Pink bollworm (Lepidoptera: Gelechiidae), with emphasis on perfor mance at high temperatures. Environ. Entomol. 22: 1051–1059. National Agricultural St atistics Service, US Dept. of Agriculture. 2003 . “Vegetables 2002 Summary”. Pgs 86. 29 Nov. 2003. http://usda.mannlib.cornell.edu/re ports/nassr/fruit/pvg–bban/vgan0103.pdf Need, J. T., and P. P. Burbutis. 1979. Searching efficiency of Trichogramma nubilae. Environ. Entomol. 8: 224–227. Nettles, W. C. Jr., R. X. Morrison, Z. N. Xie, D. Ball, C. A. Shenkir, and S. B. Vinson. 1983. Effect of cations, anions and salt concentrations on oviposition by Trichogramma pretiosum Riley in wax eggs. Entomol. Exp. Appl. 33: 283–289.

PAGE 276

258 Newton, P. J. 1988 . Movement and impact of Trichogrammatoidea cryptophlebiae Nagaraja (Hymenoptera: Trichogrammatidae) in citrus orchards after inundative releases against the false codling moth, Cryptophlebia leucotreta (Meyricj)(Lepidoptera: Torticidae ). B. Entomol. Res. 78: 85–99. Ngi–Song, A. J., and W. A. Overholt. 1997 . Host location and acceptance of Cotesia flavipes Cameron and C. sesamiae (Cameron) (Hymenoptera: Braconidae), parasitoids of African gramineous stembor ers: role of frass and other host cues. Biol. Control. 9 (2):136–142. Nguyen, I., and Nguyen S. 1982. The use of Trichogramma in Vietnam. Zaschchita Rastenii. 1: 52. Noldus, L. P. J. J., W. J. Lewis, and J. H. Tumlinson. 1990 . Beneficial arthropod behavior mediated by airborne semioche micals. IX. Differential response of Trichogramma pretiosum , and egg parasitoid of Heliothis zea , to various olfactory cues. J. Chem. Ecol. 16 (12): 3531–3544. Norman, D. W. 1980 . The Farming Systems Approach: Relevancy for the Small Farmer. Michigan State University Rural Development paper no. 5. Pgs 26. Norton, G. W., E. G. Rajotte, and V. Gapud. 1999 . Participatory research in integrated pest management: lessons from the IPM CRSP. Agric. Human Values. 16: 431– 439. Nurindah, B., W. Cribb, and G. Gordh. 1999 . Experience acquisition by Trichogramma austalicum Girault (Hymenoptera: Trichogrammatidae). Aust. J. Entomol. 38 (2): 115–119. Nutson, A. 1998 . The Trichogramma Manual. Texas Agri cultural Extension Service, The Texas A and M University System. Bulletin B–6071. 29 Nov. 2003. http://entowww.tamu.edu/ext ension/bulletins/b–6071.html Oatman, E R, and G. R. Platner. 1969. An ecological study of insect populations on cabbage in southern Calif ornia. Hilgardia 40:1–40. Oatman, E. R., G. R. Platner, and P. D. Greany. 1968 . Parasitization of imported cabbageworm and cabbage looper eggs on cabbage in southern California with notes on the colonization of Trichogramma evanescens . J. Econ. Entomol. 61 (3):724–730. Pak, G. A. 1986 . Behavioural variations among strains of Trichogramma spp.: a review of the literature on host–age select ion. J. Appl. Entomol. 101: 55–64. Pak, G. A., and E. J. de Jong. 1987 . Behavioural variati ons among strains of Trichogramma spp: Host recognition. Neth. J. Zool. 37 (2): 137–166.

PAGE 277

259 Pak, G. A., and Oatman E. R. 1982. Biology of Trichogramma brevicapillum . Entomol. Exp. Appl. 32: 61–67. Palaniswamy, P., C. Gillott, and G. P. Slater. 1986. Attraction of diamondback moths, Plutella xylostella (L.) (Lepidoptera: Plutellidae), by volatile compounds of canola, white mustard, and faba bean. Can. Entomol. 118: 1279–1285. Pansera de Araújo, M. C. G., I. B. M. Da Cruz, M. Cavalheiro, and A. K. de Oliveira. 1999 . Placement of Noctuid eggs (Lepidoptera) on soybean plants. Ann. Entomol. Soc. Am. 92 (5): 702–706. Pantojas–Garcia, E. 1990 . Development Strategies as Ideology: Puerto Rico’s Export– led Industrialization Experience. Lo ndon: Lynne Reinner Publishers. Papaj, D. R., and R. J. Prokopy. 1989 . Ecological and evolutionary aspects of learning in phytophagous insects. Annu. Rev. Entomol. 34:315–350. Phillips, R., and M. Rix. 1993 . “The Random House Book of Vegetables", Random House, NY. Picó, F. 1998. Historia General de Puerto Rico. Rio Piedras, PR. Ediciones Huracán. Pinto, J. D., E. R. Oatman, and G. R. Platner. 1986 . Trichogramma pretiosum and a new cryptic species occurring sympatri cally on southwestern North America (Hymenoptera: Trichogrammatidae). Ann. Entomol. Soc. Am. 79: 1019–1208. Pinto, J. D., and R. Stouthamer. 1994 . Systematics of the Trichogrammatidae, pp 1–36. In Wajnberg E. Hassan S. [eds.] “Biologi cal Control with Egg Parisitoids”. CAB International, Oxon. Pivnick, K. A., B. A. Jarv is, and G. P. Slater. 1994 . Identification of olfactory cues used in host–plant finding by diamondback moth, Plutella xylostella (Lepidoptera: Plutellidae). J. Chem. Ecol. 20: 1407–1427. Price, P. W., C. E. Bouton, P. Gross, B. A. McPheron, Thompson J. N., and A. E. Weis. 1980 . Interactions among three trophic le vels: Influence of plants on interactions between insect herbivores and natural enemies. Ann. Rev. Syst. 11: 41–65. Procházka, P. 2001. Czech Horticulture and the European Union, pp. 454–459, volume 2. In Proceedings, 9th International Conference of Horticulture, 3–6 September 2001, Lednice, Czech Republic. Puerto Rico Agricultural Statistics Service 1998. County level data / informacion a nivel de municipio. Fact Finders for Agri culture. National Agricultural Statistics Service, United States Department of Agriculture and The Office of Agricultural Statistics, Puerto Rico Depart ment of Agriculture. 29 Nov. 2003. http://www.nass.usda.gov/pr/county.htm

PAGE 278

260 Quayle, D., J. Régnière, N. Cappuccino, and A. Dupont. 2003 . Forest composition, host–population density, and parasitis m of spruce budworm Choristoneura fumiferana eggs by Trichogramma mi nutum. Entomol. Exp. Appl. 107 (3):215– 227. Quednau, W. 1955. Uber einige neue Trichogramma– Wirte und ihre Stellung im Wirt– Parasit–Verhaltnis. Ein Beitrag zur Anal yse des Parasitismus bei Schlupfwespen. Nachrichten blatt des Deut schen Pflanzenschtzdienstes Systematic, Brauschweig 7 , 145–148. Reed, D. W., K. A. Pivnick, and E. W. Underhill. 1989 . Identification of chemical oviposition stimulants for the diamondback moth, Plutella xylostella , present in three species of Brassicaceae. Entomol. Exp. Appl. 53: 277–286. Reid, J. C., and G. L. Greene. 1973 . The soybean looper: pu pal weight, development time, and consumption of soybean fo liage. Florida Entomol. 56 (3): 203–206. Röling, N. G., and J. Jiggins. 1998. The Ecological Knowledge System, pp. 283–331. In Röling N.G and M.A.E. Wagemakers [eds] Facilitating Sustainable Agriculture. Cambridge University Press. New York. Romeis, J., T. G. Shanower, and C. P. W. Zebitz. 1998 . Physical and chemical plant characters inhibiting the searching behaviour of Trichogramma chilonis . Entomol. Exp. Appl. 87 (3): 275–284. Rosario–Perez, C. 1984 . "Interaccion de las Malezas y dos Insecticidas en las poblaciones de Plutella xylostella (L.) (Pepidoptera : Yponomeutidae) y su biologia". Masters Thesis, University of Puerto Rico, Recinto de Mayagüez, Mayagüez, Puerto Rico. Sachtleben, H. 1929 . Die Forleule. Monographien zum Pflanzenschtz. Springer Verlag, Berlin. SAS Institute Inc. 1999 . SAS/STAT user's guide. Vers ion 8.01: SAS Institute, Inc., Cary, NC. Sato, K., Mocida O., and Kikuchi A. 1994 . Control of the diamondback moth ( Plutella xylostella ) (Lepidoptera: Ypnomeutidae) by egg parasitoids (Trichogramma). Relationships among the density of Trichogramma chilonis (Hymenoptera: trichogrammatidae) released, percent of parasitism, and mortality due to host feeding. Annual Report of the Society of Plant Protection of North Japan: 45, 168– 170. Schmidt, J. 1994. Host recognition and acceptance by Trichogramma, pp. 166–217. In Wajnberg E. Hassan S. [eds.] “Biologi cal Control with Egg Parisitoids”. CAB International, Oxon.

PAGE 279

261 Schmidt, J. M., and M. H. Carter. 1992 . The locomotory response of the egg parasitoid Trichogramma evanescens Westwood to hexane extr acts of eastern spruce budworm. Can. J. Zool. 70: 941–949. Schu, S. and R. L. Jones. 1989 . Kinetic effects of a kairomone in moth scales of the European corn borer on Trichogramma nubilale Ertle and Davi s (Hymenoptera: Trichogrammatidae). J. Insect Behav. 2 (1): 123–131. Sen, A. 1999 . Development as Freedom. Oxford University Press, Oxford and New York. Shelton, A. M., and D. M. Soderlund. 1983 . Varying susceptibility to methomyl and permethrin in widely separated cabba ge looper (Lepidoptera: Noctuidae) populations within eastern North Am erica. J. Econ. Entomol. 76: 987–989. Shelton, A. M., J. A. Wyman, N. L. Cushin g, K. Apfelbeck, T. J. Dennehy, S. E. R. Mahr, and S. D. Eigenbrode. 1993 . Insecticide resistance of diamondback moth (Lepidoptera: Plutellae) in North America. J. Econ. Entomol. 86:11–19 Smith, S. 1988 . Patterns of attack on spruce budworm egg masses by Trichogramma minutum (Hymenoptera: Trichogrammatidae) rel eased in forest stands. Environ. Entomol. 17. 1009–1015. Smyth, R. R., M. P. Hoffmann, and A. M. Shelton. 2003 . Larval performance in relation to labile ov iposition preference of Crocidolomia pavona [F.} (Lepidoptera: Pyralidae) among phonological stages of cabbage. Environ. Entomol. 32 (4): 765– 770. Spangler, S. M., and D. D. Calvin. 2001 . Vertical distribution of European corn borer (Lepidoptera: Crambidae) egg masses on sweet corn. Environ. Entomol. 30 (2): 274–279. Suh, C. P.–C., D. B. Orr, and J. W. Van Duyn. 2000 . Trichogramma releases in North Carolina cotton: why releases fail to s uppress Heliothine pests. J. Econ. Entomol. 93(4): 1137–1145. Talekar, N. S., and A. M. Shelton. 1993 . Biology, ecology and management of the diamondback moth. Annu. Rev. Entomol. 38:275–301. Thompson, J. N., and O. Pellmyr. 1991 . Evolution of oviposition behavior and host preference in Lepidoptera. Annu. Rev. Entomol. 36: 65–89. Thompson, P., and T. Taylor. 1992 . A Historical Overview of the Agricultural Sector in Puerto Rico. International Working Pape r Series; Food and Resource Economics Department (IW92–18); Institute of Food and Agricultural Sciences.

PAGE 280

262 Thomson, M. S., and R. E. Stinner. 1990. The scale response of Trichogramma (Hymenoptera: Trichogrammatidae): Variati on among species in host specificity an the effect of conditioning. Entomophaga. 35 (1): 7–22. Thorpe, K. W. 1985. Effects of height and habita t type on egg parasitism by Trichogramma minutum and T. pretiosum (Hymenoptera:Trichogrammatidae). Agric. Ecosyst. Environ. 12: 117–126. Trías Monge, J. 1997. Puerto Rico: the Trials of the Oldest Colony in the World. Yale University Press. 288 pp. Turlings, T. C. J., P. J. McCall, H. T. Alborn, and J. H. Tumlinson. 1993 . An elicitor in caterpillar oral secreti ons that induces corn seedlings to emit chemical signals attractive to parasitic wasp s. J. Chem. Ecol. 19: 411–425. Turlings, T. C. J., J. H. Tumlinson, and W. J. Lewis. 1990 . Exploitation of herbivore– induced plant odors by host–seeking parasi tic wasps. Science (Washington). 250: 1251–1253. Udayagiri, S., and C. E. Mason. 1995. Host plant constituents as oviposition stimulants for a generalist herbivore: European corn borer. Entomol. Exp. Appl. 76: 59–65. Ueno, T. 1997 . Effects of superparasitism, larval competition, and host feeding on offspring fitness in the parasitoid Pimpla nipponica (Hymenoptera: Ichneumonidae). Ann. Entomol. Soc. Am. 90 (5): 682–688. Uetz, G. W. 1991. Habitat structure and predator–prey interactions in vegetated aquatic systems, pp. 325–348. In S. S. Bell, E. D. McCoy, and H. R. Mushinsky [eds.], Habitat structure, the physical arrangemen t of objects in space. Chapman and Hall, New York. Ulmer, B., C. Gillott, and M. Erlandson. 2002 . Oviposition preferences of Berth armyworm Mamestra configurata Walker (Lepidoptera: Noctuidae) on different crucifer cultivars and growth stag es. Environ. Entomol. 31 (6): 1135–1141. University of California. 1985 . Integrated Pest Management for Cole Crops and Lettuce. University of California, Statewide integr ated pest management project, Division of Agriculture and Natural Resources Publications 3307. (USDC) US Dept. of Commerce. 1979. Economic Study of Puerto Rico. Washington D.C. USGPO. (USDA) US Department of Agriculture. 1973 . Host plant preferences of caged populations of the tobacco budworm, cabba ge looper, and soybean looper : and effect of supplemental releases of Trichogramma pretiosum and Chrysopa carnea on these three pests. Southern Region, Agri. Res. Serv., USDA and Agricultural Research and Education Center, IFAS, Un iversity of Florida. The Service, 1973 viii, 78 leaves : ill. ; 28 cm

PAGE 281

263 (USDA) US Department of Agriculture, Na tional Agricultural Statistics Service, 1997 . Census of Agriculture. 29 Nov. 2003. http://www.nass.usda.gov/census/cen sus97/puertorico/tbl13.pdf van Alphen, J. J. M., and L. E. M. Vet. 1986. An evolutionary approach to host finding and selection, pp. 23–61. In J. Waage and D. Greathead [Eds.], Insect parasitoids. Academic, London. Vasquez, L., A. M. Shelton, M. Hoffman, and R. Roush. 1997. Laboratory evaluation of commercial trichogrammatid pr oducts for potential use against Plutella xylostella (L.) (Lepidoptera: Plutellid ae). Biol. Control. 9 (2) 143–148. Vattanatangum, A. 1988 . Country review paper: Thaila nd. Informal expert consultation on IPM in major vegetable crops in Asia. 14–165 November, RAPA (FAO), Bankok 5p. Vincente–Chandler, 1978 . Concepts, Plan and Program for a Modern Agriculture in Puerto Rico: Summary. Vinson, S. B. 1981 . Habitat location, pp.. 51–77. In “Semiochemicals, their role in pest control’. [eds] D.A. Norlund, R.L., W.J. Lewis. Wiley and Sons, New York. Vinson, S. B. 1984 . How parasitoids locate their hosts : a case of insect espionage. Pgs. 325–348. In “Insect communication”. ed T. Lewis. London: Royal Entomol. Soc. Vinson, S. B. 1998. The general host selection behavi or of parasitoid hymenoptera and a comparison of initial strategies utilized by larvaphagous and oophagous species. Biol. Control. 11: 79–96. Völkl, W. 2000 . Foraging behaviour and sequential multisensory orientation in the aphid parasitoid, Paueisa picta (Hym., Aphidiidae) at differe nt spatial scales. J. Appl. Ent. 124: 307–314. Walcott, G. N. 1948 . Insects of Puerto Rico: Chalcido idea, Trichogrammidae. Jour. Agr. Univ. P.R. 32 (4): 772–775. Waldvogel, L. T., and F. Gould. 1990. Variation in ovipos ition preference of Heliothis virescens in relation to macro evolutionary patterns of heliothine host range. Evolution. 44: 1326–1337. Weisskoff, R. 1985 . Factories and Food Stamps – the Puerto Rican Model of Development. The John Hopkins Univer sity Press, Baltimore and London. 190 pgs. Westwood, J. O. 1833 . Descriptions of several new British forms among the parasitic Hymenopterous insects. The London, Ed inburgh and Dublin Philosophical Magazine and Journal of Science. 2: 443–445.

PAGE 282

264 Wilson, J. W. 1957 . Abundance of aphids and caterpill ars in commercial cabbage fields in the vicinity of Sanford, Florida. Florida Entomol. 40: 95–101. Wuehrer–Bernd, G., and Hassan S. 1993 . Selection of effective species/strains of trichogramma (Hym., Trichogrammatidae) to control the diamondback moth Plutella xylostella L. (Lep., Plutellidae). J. Appl. Entomol. 116 (1) 80–89. Yu, D. S. K., J. E. Laing, and E. A. C. Hagley. 1984 . Dispersal of Trichogramma spp. (Hymenoptera: Trichogrammatidae) in an apple orchard after inundative releases. Environ. Entomol. 13: 371–374. Yurong, H., C. Kewei, and P. Xiongfei. 2002 . Egg Parasitoids of Plutella xylostella in South China. in Seminar Proceedings: Improving Biocontrol of Plutella xylostella . CIRAD/USDA/INRA Laborat ories. Oct. 21–24, 2002. M ontpellier, France. Zhao, J. – Z., Y., –X. Li, H. L. Collins, L. Gusukuma–Minuto, R. F. L. Mau, G.D. Thompson, and A. M. Shelton. 2002 . Monitoring and characterization of diamondback moth (Lepidoptera: Plutellid ae) resistance to Spinosad. J. Econ. Entomol. 95 (2): 430–436. Zucchi, R. A., J. D. Pinto, and R. C. Monteiro. 1996 . Some records on the Trichogramma species associated with Diatraea in the New World (Hymenoptera: Trichogrammatidae – Lepidoptera: Pyralid ae), pp. 639. Proceedings of the XX International Congress of Entomo logy; 25–31 August 1996, Florence, Italy.

PAGE 283

265 BIOGRAPHICAL SKETCH Richard was born in Gweru, Zimbabwe, on August 15, 1970. He lived there until 1984, when he and his family moved to Woodbr idge, Suffolk, in the United Kingdom. He graduated with a bachelorÂ’s degree in biolog ical sciences from Durham University in 1992 and then followed this with a masterÂ’s degree in tropical agronomy from the University of Nottingham in 1994. In 1995, he was accepted as a Voluntary Service Overseas (VSO) volunteer and spent 3 years in Guyana, South America, working for the Inter-American Institute for Cooperation on Agriculture (IICA). During this time, he participated in the regional Carambola Frui t Fly Eradication Progr am and also worked with local farmers to produce a book on Integr ated Pest Management (IPM) and the use of botanical insecticides. In 1998 he started his Ph.D. at the Univer sity of Florida and spent 3 years in Puerto Rico conducting his research.