rCMATO PIi-JWCFM, KEIFSRIA LYCOPERSICELLA (WALSINGKAiM) : POPULATION BYNAI'IICS AND ASSESSMENT OF PLANT INJURY IN SOUTHERN FLORIDA By JORGE E. PENA A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLiMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF FHILCSCPHY UNIVERSI'TY OF FLORIDA 1983 ACKNOWLEDGMENTS I thank Dr. Van Waddill, my advisor and chairman, for his encouragement, support and friendship, but most of all for his valuable suggestions and allowing me freedom to conduct my research. I would also like to express my appreciation to the following people: Dr. S.K. Karr for his interest in teaching me the art of communication and also for his help in solving administrative problems during my studies. Dr. J.L. Stimac for his help and constructive criticism, as well as his ideas to improve the quality of this study. Dr. K.H. Pohronezny for his constructive criticism, suggestions and for reviewing this manuscript. Dr. D.J. Schuster for supplying material for my research as well as his interest in this study. The Agricultural Research and Education Center, Homestead, Florida, and to Dade Agricultural Council for providing the grantmanship and scholarship to support my studies. The staff of AREC, Homestead, for their cooperation, espe- cially Jonnie Csterholdt, Carolyn Reitman, Susan Housley, Leslie Sawyerlcng, Rodney Chambers, Linda Douthit, and Wilbur Dankers for their help during data collection. Mrs. Sheila Eldridge and Mrs. Barbara Hollien for kindly typing this manuscript. Drs. R.E. Litz and S.K. O'Hair for their friendship, encourage- ment and support during the past years . Mr. Ben Gregory for his honest friendship and willingness to share ideas in research. Ms. Annie Yao, Mr. A. Bustillo, Mr. W. Chongrattanameteekul , Mr. K. Patel and Mr. C. Ho for their friendship and support. I am indebted to my family for their love and encouragement, the Litz family, Eleanor Merritt and Bunny Hendrix for their friend- ship and support. TABLE OF CONTENTS Page ACKNOWLEDGIffiNTS ii LIST OF TABLES vi LIST OF FIGURES xi ABSTRACT XV INTRODUCTION 1 CHAPTER I: LITERATURE REVIEW 6 Family Gelechiidae 5 Studies on Keiferia ly coper sicella 7 Tomato Plant Phenology and Measurement of TPW Dispersion and Economic Damage 19 Environmental Factors Affecting TPW Population .... 21 CHAPTER II: DESCRIPTION OF TOflATO PLANT PHENOLOGY AND EVALUATION OF TOMATO PINWORM FOLIAR DPJ'IAGE ASSESSMENT 25 Introduction 25 Materials and Methods 26 Results and Discussion 30 Conclusions and General Discussion 51 CHAPTER III: SPATIAL DISPERSION OF TOMATO PINWORM EGGS ON TOMATOES. 55 Introduction 55 Materials and Methods 56 Results and Discussion 58 Conclusions and General Discussion 87 CHAPTER IV: SPATIAL PATTERNS OF DISPERSION OF TOMATO PIIW0R2>I LARVAE IN TOMATOES 89 Introduction 89 Materials and Methods 90 Results and Discussion 91 Conclusions and General Discussion 117 CHAPTER V: TOMATO PINWORM ARTIFICIAL INFESTATION: EFFECT OF FOLIAR AND FRUIT INJURY ON GROUND TOiVlATOES 119 Introduction 119 Materials and Methods 120 Results and Discussion 122 Conclusions and General Discussion 15 2 CHAPTER VI: ADULT DISPERSION AND COLONIZATION OF TOMATO FIELDS BY THE TOMATO PINWORM 154 Introduction 154 Materials and Methods 155 Results and Discussion 157 Conclusions and General Discussion 169 CHAPTER VII: EGG AND LARVAL PARASITISM OF TOMATO PINWORM IN SOUTHERN FLORIDA 171 Introduction 171 Materials and Methods 172 Results and Discussion 174 Conclusions and General Discussion 189 CHAPTER VIII: EFFECTS OF RAINFALL AND RELATIVE HUMIDITY ON IMMATURE STAGES OF THE TOMATO PINWORM UNDER GREENHOUSE AND FIELD CONDITIONS 190 Introduction 190 Materials and Methods 190 Results and Discussion 194 Conclusions and General Discussion 211 CHAPTER IX: INFLUENCE OF POST-HARVEST TOMATO FIELDS ON THE POPULATION DYNAMICS OF THE TOMATO PINWORM 213 Introduction 213 Material and Methods 213 Results and Discussion 217 CONCLUSIONS AND GEl^IERAL DISCUSSION 225 REFERENCES . 230 APPENDIX EXPLANATORY TABLES FOR QIAPTERS II AND III 243 BIOGRAPHICAL SKETCH 265 LIST CF TABLES Table Page 1 Lar^/al parasites of Keiferia lycopersicella reported from U.S.A. and South America until _ 1981 2 Classification of tomato pinwoirm leaf damage on 'Flora-Dade' tomatoes, based on greenhouse and field observations. Homestead, Florida, 1930 31 32 Leaf area and reproductive plant struct\ires in tomatoes, cv Flora-Dade' , planted on 5 dates in Homestead, Dade County, Florida during 1980- 19S1 Stage of development description for tomato cv Flora-Dade. Description is based on the average of observations from tomato plants grown during Fall 1980 through Winter 1981. Homestead, 42 Florida Tomato leaf weight and leaf area consumed by different larval instars of Keiferia lycopersicella under greenhouse conditions; T 24_+3°C, 75+^2% RH. . . Percentage of tomato pinworm larval occurrence in foliar injuries with different phenological charac- teristics Comparison of different sample sizes for tomato pinworm eggs. Homestead, Dade County, Florida, 1980 Mean number of tomato pinworm eggs per plant by planting date for 8 tomato plantings in Homestead, Florida, 1979-1981 Ovipositional preference of tomato pinworm for upper and lower surfaces of tomato leaves from plants grown under greenhouse and field conditions 45 52 60 64 Table Page 10 Mean number of tomato pinwcrm eggs in 2 plant strata (upper and lower halves) per plant at different sampling dates. Homestead, Dade County, Florida, 1980 ^^ 11 Mean number of tomato pinworm eggs per plant in 6 strata: upper, middle and lower external; upper, m.iddle and lower internal canopy of the tomato plant. Homestead, Florida, 1981 67 o 12 Relationship between daily mean temperature ( C) and TPW oviposition in 5 tomato plant strata. Homestead, Florida, 1981 71 13 Percentage distribution of TPW eggs for each stratum of tomato plants in 5 tomat plantings. Homestead, Dade County, Florida, 1980 75 14 TPW egg sample allocation for 6 plant strata during 3 different plant stages: second vegetative (TR2) ^ first reproductive (TR ) , and second reproductive stage (TR^) ■^'^ 15 Mean tomato pinworm eggs on tomato leaves from different strata of 45 day-old plants. Homestead, Florida, 1980 79 16 TPW egg sample allocation on tomato leaves numbered from bottom to top. Plants 45 days old 81 17 Relationship between frequency of occurrence of TPW eggs per leaflet as dependent variable and distance among eggs and leaflet area as independent vari- ables 18 TPW oviposition on tomato at different plant stages. Homestead, Florida, 1981 82 86 19 Sample size and relative net precision (RNP) for sampling injuries at low and high population densities. Homestead, Dade County, Florida, 1980. . 93 20 Mean number of TPW foliar injuries and standard error on different sampling units at specified date. Crop planted in Nov., 1979. Homestead, Florida 96 Table Page 21 Mean number of TPW foliar injuries and standard error on different sampling units at specified date. Crop planted in Jan., 1980. Homestead, Florida 97 22 Sample size and relative net precision (RNP) for sampling TPW larval injuries on upper and lower plant canopy. Homestead, Florida, 1980 100 23 Sample statistics : Mean tomato pinworm larval in- juries per plant in 8 tomato plantings. Homestead, Florida, 1979-81 102 24 Mean tomato pinworm (TPW) foliar larval injuries at 2 different plant levels for 3 different plantings. Homestead, Dade County, Florida, 1980. . . 103 25 Mean tomato pinworm (TPW) larval injuries in 6 plant strata for 5 plantings. Homestead, Dade County, Florida, 1981 105 26 TPW larval injury sample allocation for 6 plant strata at 3 different plant stages: second repro- ductive (TR2) , third reproductive (TR_) and senescent (S-, ) 113 27 Mean n\miber and standard error of tomato pinworm (TPW) injuries in 5 different plantings at speci- fied date and plant growth stage 115 28 Tomato fruit damaged in the upper and lower plant canopy, after a single artificial infestation with K. lycopersicella larvae, on ground tomatoes .... 124 29 Marketable value for tomato fruit damaged in the lower and upper plant canopy after a single arti- ficial larval infestation of K. lycopersicella on ground tomatoes 125 30 Tomato fruit damaged in the lower and upper plant canopy after a double artificial infestation of K. lycopersicella larvae on ground tomatoes 125 31 Marketable value for the tomato fruit damaged in the lower and upper plant canopy after a double infestation of K. lycopersicella on ground tomatoes 128 Table Page 37 Parasitism of the tomato pinworm larvae in tomato fields in southern Florida, Dade County, 1981 39 Number of Keiferia lycopersicella eggs col- lected from two strata and percent of para- sitism by Trichogramma pretiosum 40 Distribution of normal and parasitized tomato pinwo2rm eggs in 2 tomato fields 41 Parasitism of tomato pinworm eggs by Tricho- gramma pratiosum in 2 fields with different host densities 42 Plant water content in five tomato plantings related to oviposition by the tomato pinworm 43 Effect of simulated rainfall on foliar larval injuries caused by the tomato pinworm Keiferia lycopersicella on plants grown under greenhouse conditions 149 32 Effect of planting time on fruit injured by K. lycopersicella larvae to ground tomatoes, cv 'Flora-Dade' during 1981 33 Differences in cost and relative net precision between sampling 6 plants per row and 1 random plant per row 34 Differences in mean fruit injured by K. lycopersicella in pruned and not pruned tomato plants 35 Effect of hedges and edgerows on tomato pinworm field infestation at three fields in Home- stead, Florida, 1981 ^^^ 36 Parasitism of the tomato pinworm larvae in tomato fields in southern Florida, Dade County, 1980 ^"^^ 177 38 Keiferia lycopersicella eggs parasitism by 2 strains of Trichogramma pretiosum in the laboratory. T25+1°C; 75+2% RH 1"^^ 44 Mean percentage of tomato pinworm adults emerged by day after pupal treatment with different simulated rainfall regimes . . 180 181 183 203 204 208 Table Page 45 Effect of crop age of post-harvested tomato plants on volunteer plants and number of tomato pinworm larval injuries 220 46 General effect of cultural practices on volunteer tomatoes and infestation by tomato pinworm 222 47 Effect of planting age and cultural practices on volunteer tomato plants and nmnber of TPW injuries 223 48 Tomato pinworm egg frequency distributions determined on tomato plants during 1981 244 49 Tomato pinworm foliar injury frequency distributions determined on tomato plants during 1980 249 50 Tomato pinworm foliar injury frequency distributions determined on tomato plants during 1981 253 51 TPW egg allocation sample for 6 plant strata. Planting 8, 1981. Age: 38 days. Stage of development TV 258 52 TPW egg allocation sample for 6 plant strata. Planting 8, 1981. Age: 46 days. Stage of development TR 259 53 TPW egg allocation sample for 6 plant strata. Planting 7, 1981, Age: 68 days. Stage of development TR 260 54 TPW egg allocation sample for 5 plant strata. Planting 4, 30 Oct. 1980. Age: 77 days. Stage of development TR 261 55 TPW larval injury sample allocation for 6 plant strata. Planting 5, 1981. Age: 78 days. Stage of development TR 262 56 TPW larval injury sample allocation for 6 plant strata. Planting 5, 1981. Age: 108 days. Stage of development TR 263 57 TPV7 larval injury sample allocation for 6 plant strata. Planting 4, 1981. Age: 120 days. Stage of development TR 264 LIST OF FIGURES Figiire Page 1 Illustration of tomato cv Flora-Dade growth at 2 stages of development. TV2=second vegetative stage; TR =early reproductive stage; a=primary leaf; b=lateral develop- ment 29 2 Influence of time on leaf area (dm ) expansion, flower- ing and tomato fruit numbers of cv Flora-Dade grown on 'Rockdale' soil under southern Florida conditions .... 39 Stages of development of tomato. TV =early vegetative stage; TV =late vegetative stage; TR , TR , TR^=repro- ductive stages; S =senescent stage 41 Linear relationship between tomato pinworm (Keiferia lycopersicella) larval head capsule width (ram) and foliar injury length, r^=0.47 47 Linear relationship between tomato pinworm (Keiferia lycopersicella) larval instars and visual leaf damage scale, r"=^=0.677 50 Average number of tomato pinworm eggs per plant stratum during 5 different sampling dates in 2 to- mato plantings at different growth stages . A) Planting 7: Jan. 30, 1981. B) Planting 8: Feb. 23, 1981. TR =second reproductive stage of develop- ment; TV =sec5nd vegetative stage of development. Plant strata: 1, 2, 3: upper, middle, lower external, 4, 5, 6: upper, middle, lower internal '^3 Frequency of tomato pinworm eggs at different distances (cm) between eggs when mean eggs were A) 2 eggs per leaf- let and B) 5 eggs per leaflet ^^ Percentage of tomato pinworm (TPW) larval injuries in 2 sampling units from different plant portions, related to number of injuries in the whole plant: 1) Ist planting, Nov. 3, 1979; 2) 3rd planting, Jan. 8, 1980 99 Figure Page 9 Percentage of tomato pinworm (TPW) foliar injuries found at upper, medium and lower stratum in 4 tomato plantings: 1) Oct. 30, 1980; 2) Nov. 25, 1980; 3) Dec. 30, 1980; and 4) Jan. 30, 1981. Bars followed by different letters were significantly different according to Duncan's Multiple Range Test (P=0.05). Percentages were previously transformed to arc sine. Percentages are expressed as actual n\:mibers before transformations 108 10 Percentage of larval injuries at the external and internal canopy evaluated from 5 tomato plantings. Plantings 4, 5 and 6 planted in Oct., Nov., and Dec, 1980; Plantings 7 and 8 planted in Jan. and Feb., 1981. Homestead, Florida, 1980-81 Ill 11 Relationship between number of tomato pinworm larvae per plant and nvimber of injured fruits and leaves in the lower plant canopy by a single artificial infestation with TPW larvae. Homestead, Florida, 1980-81 130 12 Relationship between niimber of tomato pinworm larvae per plant and number of injured fruits and leaves in the upper plant canopy by a single infestation of TPW larvae. Homestead, Florida, 1980-81 132 13 Relationship between number of leaves injured in upper and lower canopy and number of fruits injured in upper and lower canopy by a single artificial infestation with TPW larvae. Home- stead, Florida, 1980-81 135 14 Relationship between number of larvae per plant and number of injured fruits and leaves (upper and lower canopy) after a do\ible artificial in- festation with TPW larvae. Homestead, Florida, 1980-31 137 15 Relationship between number of leaves injured in upper and lower canopy and number of fruits injured in upper and lower canopy by a double artificial infestation with TPW larvae. Home- stead, Florida, 1980-81 140 16 Regression of percent of yield reduction against infestation densities per plant of tomato pinworm larvae 144 Figure Page 17 Regression of percent yield reduction of to- mato fruit against TPW number of foliar injuries per plant 1^^ 18 Abundance of tomato pinworm male moths at three different field sites. A) Field 1, 1980; B) Field 2, 1980; C) Field 3, 1981. Mean number of moths at each site corresponds to the average from 4 pheromone traps in north, south, west and east directions. Homestead, Florida, 1980-81 159 19 Mean number of tomato pinworm larval injuries occurring in 4 commercial fields at 0, 30, and 120 m from the field border. Bars with the letter "a" denote statistical differences at 0.05% dif- ference level for a particular date 164 20 Field plan: Fields 1, 2, and 3, respectively, each field was divided into 8-9 quadrats. The shaded areas represent higher insect populations . Home- stead, Florida, 1981 -^^ 21 A-B. Relationship of Keiferia lycopersicella egg density to percent parasitism by Trichogramma pretiosum in 2 fields -'-^^ 22 A-C. Seasonal occurrence of Keiferia lycopersicella eggs and parasitization by T. pretiosum in tomato fields located at the (a) northern, (b) middle, and (c) southern areas of Dade County, Florida 188 23 Seasonal abundance of TPW eggs in experimental fields, related to temperature and rainfall regimes during A) 1980, and B) 1981, in Home- stead, Florida ^^^ 24 Seasonal abundance of TPW larvae in tomato fields, related to temperature and rainfall regimes during 1980-81, in Homestead, Florida 200 25 Mean nianber of TPW injuries per plant during 9 days of simulated rainfall under greenhouse con- ditions, avg. daily temperature 25+2°C 206 26 Percentage of TPW adult emergence under greenhouse conditions after treatment of pupae with 3 regimes of artificial rainfall (200, 100, 50 and 0 ml water), temperature 24+3°C 210 Figure Page 27 Tomato field status following the main harvest under S . Florida conditions . Home- stead, Florida, 1980 216 28 Number of tomato plants and TPW injuries per m in 2 post-harvested tomato fields. Homestead, Florida, 1980 219 Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TOMATO PINWOPM, KEIFERIA LYCOPERSICELLA (WALS INGHAM) : POPULATION DYNAMICS AND ASSESSMENT OF PLANT INJURY IN SOUTHERN FLORIDA By JORGE E. PENA April 1983 Chairman: Dr. V.H. Waddill Major Department: Entomology and Nematology Experiments were conducted in Homestead, Florida, during 1979-1981 to describe tomato plant phenology, tomato pinworm (TPW) , Keiferia lyco- persicella (Walsingham) , dispersion patterns , economic damage to tomato and the effects of parasitoids, edgerows, rainfall and cultural practices on TPW population dynamics. Tomato cv. Flora-Dade phenology was described. Six stages were designated based on the number of leaves, flowers, fruits and physiological plant characteristics. This description can be of use in making pest management decisions. Based in the relative variation (RV) and sampling costs, sampling units for TPW egg and larval stages were determined. Eggs were generally (51%) found in the upper plant canopy, and larvae (50%) in the lower plant canopy. Larger sampling units were allocated to the upper and lower plant canopy for eggs and larvae, respectively. An economic injury level was determined to be 1 larva per plant. Yield can be reduced 10-40% when 1-12 larvae are attacking 45 day-old plants. The results indicated a correlation between nimiber of foliar injxiries in the lower plant canopy and fruit damage. In southern Florida, higher TPW infestation occurred during March-May, 1980 and March-April, 1981, compared with other months (Jan., Feb.). Trichogramma pretiosum Riley caused 33-73% TPVJ egg mortality during May- July, 1981. TPW larval parasitism fluctuated between 39-42% during 1S80-1981. The most abundant larval parasite was Apanteles spp., followed by Sympiesis stigmatipennis and Temelucha spp. TPW ad^xLt dispersion and effects of field edges on TPW dispersion and field colonization were evaluated. Field areas surrounded by edge- rows had higher TPW damage than areas surroiinded by pastures . The use of artificial rainfall demonstrated that when plant foliage was irrigated there was a behavioral change in larval feeding which res\ilted in 50% reduction of larval injuries compared to injuries on soil-irrigated plants. TPW adult emergence was reduced 86% when high levels of water were applied to pupae in the soil. The effect of cultural practices on the TPW oversummering popu- 2 lation was evaluated. The mean number of injuries per m was 28 times higher in crops planted later (December, 1980)^ than in crops planted earlier (October-November, 1980) . Lower ni:imbers of injuries were found in crops disced and mowed than in abandoned fields. Parasitoids, cultural practices, and southern Florida climatological patterns can have an impact on TPW population levels . INTRODUCTION The tomato, Lycopersicon esculentum Mill., is one of the most popular and important vegetables in the world (Pxirseglove 1968) . Tomato produc- tion in the U.S.A. is concentrated in California, Florida, Texas, New York, New Jersey, Michigan and Virginia (Thompson and Kelly 1957) . Florida tomato acreage was 31360 ha during 1980-1981. Tomato produc- tion is considered to comprise 28.31% of the total vegetable acreage in Florida (Anonymous 1982) . The tomato growing areas in Florida are divided into 4 major districts: Palmetto-Ruskin, Pomp£ino Beach-Fort Pierce, Dade Co\inty and Immokalee-Naples (Anonymous 1981). Dade County has 18.3% of the total state tomato production and supplies most of the winter (De- cember through Febiniary) vegetables for the U.S.A. The cultivation of fresh market tomatoes demands a high monetary in- vestment from farmers. The cost of producing tomatoes in Dade County during 1980 was $5123.25 per ha, which represents an increase of 1.44 times over the production cost of 1975 (Greene et al. 19SC) . Expansion of tomato acreage in Florida resulted in changes of agro- nomic practices to maximize tomato production (Geraldson 1975) . Changes in horticultural practices also established an agro-ecosystem with ento- mological characteristics commcn to monocultures. From 1950-1975 insect control in tomatoes was almost exclusively chemical. To help growers avoid problems with insecticides such as insecticide resistance, secondary pest outbreaks and objectionable pesticide residues, an integrated pest man- agement program was established in Dade County on tomatoes (Pohronezny et al. 1978) . This program goal was to develop economically, technically and ecologically sound systems of integrated pest management. This approach had some constraints, however, such as the high crop value which reduces the use of pest management tactics (Bottrell 1979) . More- over, the fruit quality standards for fresh tomatoes cause undue emphasis on chemical control measures in order to prevent contamination of fruit by insects and to prevent cosmetic damage to the fruit (Lange and Bronson 1981) . Accordingly, insect pests in tomatoes can be categorized as direct pests and indirect pests. Direct pests attack the product and directly destroy a significant part of its value. Indirect pests attack plant parts other than the saleable product but may reduce yield of the product (Ruesink and Kogan 1975) . Among direct pests of tomato in Florida are the corn ear- worm, Heliothis zea Boddie, the southern armyworm, Spodoptera eridania (Cramer), and the tomato pinworm, Keiferia lycopersicella (Walsingham) . Indirect pests are the serpentine leafminer complex, Liriomyza spp, the tobacco hornworm, Manduca sexta (Joh.), and the granulate cutworm, Feltia siibterranea (Fab) (Poe 1972). The tomato pinworm (TPW) can be either a direct or indirect pest of tomatoes. The larva of this insect feeds in the mesophyll of the leaves causing a seirpen tine- type mine during the first 2 larval instars. In the latter instars the larvae can cause a blotch-type mine or they tie leaflets together. The larvae also bore into fruit, providing an entrance for plant pathogens which cause major damage to fruit. The importance of the TPW as one of the most serious pests that affect tomato production in semitropical areas of Florida has been docu- mented by Poe (1974a) and Wolfenbarger et al. (1975). Tomato pinworm incidence was noted in Florida as early as 1932 (Watson and Thompson 1932) with serious outbreaks occurring during 1942, and from 1970 through 1973. Several factors have been mentioned by Poe et al. (1975) as caus- ing these outbreaks, i.e., type of insecticide used, change of tomato pro- duction practices, and harmful effect of pesticides on natural enemies. Other factors such as weather have been overlooked. Current practices for TPW control in Florida have been almost exclusively chemical (Waddill 1975) , although emphasis has also been given to breeding tomatoes for TPW-resistance (Schuster 1977a) and less to TPW biological control (V.H. Waddill, personal communication). The effects of several factors, e.g., rainfall and cultural practices, that influence the life system of the TPW are still not iinderstood. To develop effective integrated pest management for tomato, the interrelationships among the crop (plant physiology, phenology) , pests (arthropods, weeds, pathogens) and environment (climate, natural enemies, horticultural practices) must be carefully studied. It is necessary to understand TPW ecology and basic biology by studying the role of several factors that cause seasonal and annual changes in pest populations. The ability to assess the presence and abundance of the pest by accurate sampling techniques would pe,rmit a reliable study of TPW potential for inflicting economic damage. By evaluating the role of extrinsic factors, e.g., weather, natural enemies and agronomic practices, it may be possible to reduce the TPW problem. This study was initiated to answer these and related questions. The specific objectives of research were: 1) to describe different stages of development of the tomato plant. 2) to evaluate techniques for tomato pinworm damage assessment. 3) to discuss sampling techniques for tomato pinworm immature stages under southern Florida conditions and to describe TPW spatial distribution. 4) to evaluate the importance and population dynamics of TPW natural enemies . 5) to evaluate the effect of hedges and edgerows on TPW dispersion and field colonization. 6) to determine a way to assess yield losses in ground tomatoes due to TPW. 7) to determine the influence of rainfall on TPW population. 8) to study post-harvest field managment practices that influence TPW survival . Therefore, the first chapter is a general literature review of studies on K. lycopersicella and addresses the effects of biotic and abiotic factors on the population dynamics of this insect. The second chapter is a study of tomato plant phenology and also covers the evaluation of foliar damage assessment techniques. In chapters III and IV, I address sampling techniques and dispersion patterns of tomato pinworm eggs and larvae. The fifth chapter deals with the effect of tomato pinworm infestation on upper and lower parts of the plant. In the same chapter I state the relationship between TPW population index and yield losses. In chapter six I address the distribution of male moths and larval stages in tomato fields, and the effect of edgerows in such distribution. In chapter VII, I deal with the abxmdance of egg and larval natural enemies of the tomato pinworm. The interaction of rainfall and TPW is presented in chapter VIII. Finally, I evaluated the data regarding horticultural practices and the relationship between changes of tomato agroecosystem and oversunmering populations of tomato pinworm (chapter IX) . CHAPTER I LITERATURE REVIEW Family Gelechiidae The family Gelechiidae is one of the largest of the microlepidoptera (about 580 North American species) . Larvae vary in habits- Some are leafminers, a few form leaf galls, many roll or tie leaves, and one species, Sitotroga cerealella Olivier, is an important, pest of stored grains (Borror et al. 1976) . Studies on crop pests in this fam.ily have been concentrated on pests of high economic importance, such as the pink bollworm (Pectinophora gossypiella Saunders) , the potato tuber^vorm (Phthorimaea operculella Zeller) , the angoumois grain moth (S_. cerealella) , and Keiferia lycopersicella Walsingham, the tomato pinworm. The pink bollworm and the potato tuberworm are generally considered good colonizers with highly mobile behavior within and between fields (Stern 1979, Van Steenwick et al. 1978); however, many experts considered these moths weak fliers which move great distances by being carried pas- sively by air cixcrents (Kaae et al. 1977) . They are capable of having several generations per year, with the last generation showing a strong dispersal tendency (Kaae et al . 1977). The potato tuberworm is perhaps the most closely related to the tomato pinworm in patterns of behavior and plant selection (Kofmaster 1949) . Several authors (Shelton and Wyman 1979, Meisner et al . 1974, Traynier 1975) have studied factors influencing oviposition of potato tuberworm and the relationship between populations of the pest and the host plant. Their studies were used as a base in this research to compare with K. lycopersicella population dynamics. Studies on Keiferia lycopersicella (Walsingham) The tomato pinworm (TPW) , K. lycopersicella (Wals) , is frequently confused with other species (Povolny 1977) , particularly with Scrobipal- pula absoluta (Meyr.) and Phthorimaea operculella (Zell.) (Doreste and Nieves 1968) , since they are also considered pests of potato and tomato (Povolny 1973) . K. lycopersicella and S^. absoluta are apparently iso- lated from each other geographically and ecologically. K. lycopersicella apparently avoids the cordillerian territory of the northern and southern part of South America (Garcia et al. 1974, Mallea et al. 1972, Quiroz 1976) The range of K. lycopersicella is in the eastern part of the American continent and penetrates into Central America, Mexico (Povolny 1973) and the U.S.A. (Elmore and Rowland 1943). Phthorimaea operculella has been reported on tomatoes in Venezuela, (Doreste and Nieves 1968 )_, Bermuda (Grooves 1974), and Egypt (Abdel-Salam et al. 1971). In the U.S.A. K. lycopersicella is considered a key pest of tomatoes in western California (Oatman 1970), Texas, Florida, Pennsylvania and Hawaii (Swesey 1928, Thomas 193 3) . The tomato pinworro was first recog- nized as a pest of tomatoes by Morrill (1925) , and was later reported by Elmore (1937) and Thomas (1933) . In Florida, the TPW has been primarily studied by Watson and Thompson (1932) , Swank (1937) , and recently by Poe (1973) . The seasonal history of the TPW was reported by Elmore and Rowland (1943) in California where it appears first during March and April after overwintering in the pupal stage at or near the surface of the soil. Later studies of the seasonal occurrence of TPW in Cali- fornia showed that larval populations increased abruptly in September and October (Oatman et al. 1979) and in April- June (Oatman 1970) . Batiste et al. (1970b) reported that there is no evidence for diapause in this insect. Destruction of the tomato plants shortly after harvest may prevent the insect from surviving the winter and infesting the crop during the following season. Poe (1974a) reported that on the west coast of Florida, severely infested fields occurred in the spring crop (February- May) with less damage on plants during the autimin. Early infestations in greenhouses also lead to heavy losses in the field. Host Plants of Keiferia lycopersicella Elmore and Rowland (1943) reported that tomato and potato are pre- ferred hosts of TPW. Several solanaceous plants, e.g. eggplant iSolanum melongena (L) ] and nightshade (Solanum nigrxmi L.) , also are known hosts for the TPW (Batiste et al. 1970b, Elmore and Rowland 1943, Swank 1937, Thomas 1933) . Batiste and Olson (1973) demonstrated that K. lycopersicella preferred tomato for oviposition over 12 other solanaceous plant species. TPW could be reared on Solaniim melongena L., S_. dulcamara L., S_. nigrum, and S. elaegnifoliiom Cav. but not on S_. nodiflorum Jacq., S_. douglasi Dunal, Datura mete lo ides A. , D. stramonium L. , D. ferox L. , Nicotiana biglovii (Torr.) and N. glauca Grah. The same author suggests that in California, Solanum melongena , S_. dulcamara and S_. elaegnifoliimi may play a role in the population dynamics and distribution of TPW. Life Cycle of Keiferia lycopersicella Accounts of the life histoiry and behavior of K. lycopersicella have been reported by Elmore and Rowland (1943) , Swank (1937) , and Poe (1973) . Poe (1973) found that eggs are laid singly or in groups of two or three on the host plant foliage. Elmore and Howland (1943) described the egg as ellipsoid, 0.37 by 0.23 mm, light yellow when first deposited, grad- ually darkening to a light orange before hatching. Eggs hatch 4-9 days after deposition (Swank 1937) at 20.68°C and after 4-4.5 days at 27-29 C (Elmore and Rowland 1943) . Weinberg and Lange (1980) determined that eggs hatch in a range of 3.5 +0 days at 35°C and 7.8 +0.2 days at 20°C. Keiferia lycopersicella has four larval instars (Elmore and Howland 1943, Swank 1937). Head capsule width of the larval instars are 1st in- star 0.14-0.157 mm; 2nd instar 0.23-0.28 mm; 3rd instar 0.36-* ^,39 mm; 4th instar 0.52-0.61 mm (Elmore and Howland 1943). Newly hatched larvae averaged 0.85 mm in length. The head capsule is dark brown and the re- mainder of the body is a yellowish gray common to many newly hatched lepidopterous larvae. The mature larvae are 5.8-7.9 mm in length and characterized by an ash gray color with dark purple spots (Elmore and Howland 1943) . Larvae of K. lycopersicella characteristically possess a pale prothoracic shield with conspicuous dark fuscous shading along lateral and posterior margins (Capps 1946) . Duration of the leaf mining (lst-2nd instars) stage ranges between 4.7-5.8 days. The leaf folding stage lasts between 5.6-16 days for a range of temperatures of 13-29 C (Elmore and Howland 1943) . Weinberg and Lange (1930) found that egg hatching to pupation times range from 8 j^O.9 to 18 +1.6 days when o rearea at 35 C. 10 The pupae are initially green, later turning to a brown typical of lepidopterous pupae commonly found in the soil (Elmore and Rowland, 1943) . Before pupation the larvae form a loose pupal cell of sand grains at a depth of 0.25-1.5 inches beneath the soil surface (Poe 1973). Wein- berg and Lange (1980) recorded that pupation requires 11.3 +^0.5 at 20 C, and 5,1 + 0.2 days at 35°C. The length of the pupal stage was 38.7 and 11.4 days at temperatures of 12.65 and 26.4 C (Elmore and Howland 1943) . Swank (1936) obtained a range of 7-17 days with an average of 11 days for the pupal stage at 26 C. Adults are characterized by an alar expanse of 9-12 mm. Labial palpi have a short forrowed brush on the underside of the second joint, a terminal joint somewhat thickened with scales, and are compressed with the extreme tip pointed. The head and thorax are mottled with dark brown. Forewings are elongate ovate, the hind wings have a pointed apex, a strong pencil of hair scales, are dilated at tip of costa in females, and dilated from base of costa in the males; the abdo- men is dark fuscous above with basal joints slightly ochreous, the underside is light ochreous sprinkled with dark fuscous spots. Adult longevity is 7 days (24 + 2°C) when they are fed on water and 8.5 days at 24 + 2°C when fed a 10% honey solution. At temperatures of 10 and 13 C the respective longevities were 20.5 and 22.8 days (Elmore and Howland 1943) . Insect Behavior Elmore and Howland (1943), reported that copulation occurs within 24 to 48 hrs after moth emergence, and McLaughlin et al. (1979) stated that 11 sexual activity such as female calling was greatest during the 1st hr of darkness. Very little copulation occurred after the 3rd hr. Males ran or walked in their approach to calling females . Approach was gener- ally from behind or at ca. 90 to the female and was accompanied by rapid wing fanning. The copulatory strikes of the males were made laterally beside the females. Moths remained in copula from 30 min to 2 hr. Elmore and Howland (1943) and Poe (1973) described the behavior of larval stages of K. lycopersicella . Newly eclosed larvae disperse briefly from the hatched egg before entering the leaf. First instar larvae spin a tent of silk over themselves and tunnel into the leaf. Further feeding results in a blotch-like mine. The 3rd and 4th larval instars feed from within tied leaves, folded portions of a leaf, or they may enter stems or fruits. The 3rd instar appears to be the most mobile and several types of behavior may occtir (Poe 19731. This stage larvae can draw 2 leaves together, may tunnel into stems or fruits at the calyx, but usually the larvae form leaf folds on the upper leaf surface. The four instars can cause injury to 3-6 leaves during develop- ment (Poe 1973) . Elmore and Howland (1943) demonstrated that larvae that have mined calyx lobes and nearby tissues enter the fruit instead of folding leaves. Usually, the larvae enter the fruit beneath the calyx lobes or fruit stems, but in heavily infested fields about 50% of the injured fruit may be damaged in other places as well. The damaged areas caused by shallow feedi.ng just beneath the skin of the fruit appear as blotches. Larvae that enter the fruit penetrate to a depth ranging from 0.9-1.9 cm. Differences in the phenology of larval injuries were studied by Batiste et al. (1970), who found that mines of the early stage larvae 12 superficially resembled the serpentine type mines produced by dipterous leafminers of the genus Liriomyza. The mines could be distinguished easily, because the dipterous leafminer leaves a trail of frass within the mine, whereas the TPW larvae deposits nearly all the frass in a single mass at the tunnel entrance. Tomato Plant Resistance to TPW Breeding for resistance work with tomatoes has largely been con- cerned with pathogens, but currently there is a renewed em.phasis on insect resistance as part of integrated pest management (Lange and Bronson 1981) . Resistance to many tomato insects does occiir and includes resis- tance to the fruitworm, Heliothis zea (Cosenza and Green 1979); leaf- miners, Liriom.yza spp. (Schuster et al. 1979) ; tomato pinworm, K. lyco- persicella (Schuster 1977a); hornworms, Manduca spp (Kennedy and Henderson 1978) , Colorado potato beetle, Leptinotarsa decemlineata Say (Schalk and Stoner 1976; potato aphid, Macrosiphum euphorbiae (Thomas); flea beetles; white flies (Aleyrodidae) ; spider m.ites (Acarina) and many others (Lange and Bronson 1981) . The mode of resistance in tomato is complex and may involve many factors including antibiosis, preference, phenological devel- opment (such as flowering time, time of fruiting, etc.), morphological characteristics, presence or absence of foliage pigments, foliage vol- atiles, and physiological incompatibility. Resistance to the tomato pinworm has been studied by Schuster (1977a)., Schuster et al. (1979) , and Kennedy and Yamamoto (1979) . Schuster (1977a) found that accessions of Lycopersicon sculentum Mill x L. pimpinelli folium were more susceptible, while those of L_. peruvianum (L) Mill, L. 13 peruvianum var. huitiifusum Mill., L^. esculentum x L^. peruvianum, L. cheesmani f. minor (Hook F) Mull., and h. glandulosxmi Mull., were less susceptible than the commercial cultivar 'Walter' (L^. esculentimi Mill.) . Selections of L. hirsutum Huiab and L. hirsutum f. glabratum Mull, were more resistant and had 25-50% and 50-75% less damage respectively than 'Walter' . In laboratory studies the same author found that mine lengths after 2 days were significantly shorter' for PI niombers 129157 (L. hirsutum f glabratum) and 298933 (L. peruvianum) . Schuster et al. (1979) stated that levels of resistance to tomato pinwoirm and vegetable leafminer appeared to be intermediate and the varieties PI 12930 and PI 1404403 of L. esculentum were found moderately resistant to both insects. Kennedy and Yamamoto (1979) found an extractable toxic factor in the foliage of L. hirsutum f . glabratima affecting Manduca sexta , H^. zea, K. lycopersicella , Aphis craccivora, A. gossypii , and Myzus persicae. Schuster (1977b) reported that tomato varieties 'Pennorange E 160 A' and 'Pearson' had less fruit damage by K. lycopersicella and armyworms, primarily Spodoptera eridania (Cramer), than did the 'Walter' variety. Chemical Control of TPW Chemicals are widely used to control tomato pests. The need for insecticides varies from year to year and from one area to another (Lange and Bronson 1981) . Chemical control of TPW was obtained in 1943 by Elmore and Rowland (1943) who recommended synthetic cryolite and talc dust (50% sodium fluoaminate) . In California, several insecticides were evaluated by Middlekauff et al. (1963) and reevaluated by Batiste et al. (1970a) . The latter authors reported little or no control of larvae by 14 insecticides applied as soil treatments under greenhouse conditions . These same authors stated that methyl parathion was the most effective material in greenhouses, and also recommended parathion, methidathion, phosphamidon, mexacarbamate and methamidophos . Spray deposits of para- thion were fourid by the same authors to be significantly less effective against eggs or early stage larva than was toxaphene-DDT. Poe and Everett (1974) presented results of experiments to control TPW in 2 locations in Florida. They reported that granular insecticides in general did not perform as well as most spray materials for reduction of the TPW mines and larvae in tomato transplants . They recommended acephate, diazinon, endosulfan, and methomyl to keep seedlings nearly mine free. Chlordimeform was coMsidered phytotoxic to seedlings but when sprayed alone or combined with Bacillus thuringiensis Berliner on older plants gave good control of TPW larvae without plant toxicity. Poe and Everett (1974) recommended highly residual insecticides to maintain a crop free of damaged fruit. Waddill (1980) reported that certain insecticides used on demand for tomato pinworm in Homestead, Florida, significantly reduced TPW damage below that in the untreated check. Permethrin + Bacillus thuringiensis were applied least often and were among the best treatments. The author also showed that when used on demand a low rate (0.225 lbs ai) of methomyl resulted in significantly more damage than the same rate plus 0.5 lbs Bacillus thuringiensis . Schuster (1977b) reported that when measured by the number of damaged fruit, the degree of control of the TPW and southern aimiyworm with Bacillus thuringiensis WP and chlordimeform was significantly depen- dent on the tomato cultivar. The contact toxicity of 4 synthetic 15 pyret±iroids and methomyl to some adult parasites of tomato pests indi- cated that fenvalerate was generally the least toxic to parasites com- pared to permethrin, burethrin, and NRD1C49 (j^) -d-cyano-m phenoxybenzyl (+) cis, trans-3-(2,2 dichlorovinyl) -2-dimethyl-cyclo-propanecarboxylate) as well as methomyl (Waddill 1980) . Fenvalerate was judged the most promising candidate for use in a pest management program in tomatoes for integrated control of the TPW and the vegetable leafminer. Lindquist (1975) obtained the best control of TPW with synergized pyrethrins (MGK pyrethrins) and endosulfan. Emergence of K, lycopersicella and Apanteles spp from pupae and soil treated with insect growth regulators (IGR's) resulted in 23% suppression of pinworm adult emergence when applied directly to the TPW pupae but was ineffective when applied to the soil. The IGR's caused a reduced emergence of the parasite Apanteles spp from 51% to 0% (Poe 1974b) . Prada and Gutierrez (1974) reported some results on microbial insecticide control of Scrobipalpula absoluta, the South American pinworm. Seventy five to eighty percent control of the pest was obtained within 5-100 days after treatment at the rate of 500-200 Neoplectana carpo- capsae Weiser nematodes per plant or with Bacillus thuringinesis (150- 500 g/ha) . Schuster (1982) demonstrated that a mixture of B. thuringiensis and Coax® (454 g + 1.8 kg product/378 Its) when applied to TPW infested tomato seedlings, increased TPW mortality up to 42.2%. Cultural Practices for TPW Control According to Lange and Bronson (1981) , the mechanization of pro- duction of processing tomatoes has not only revolutionized the industry but has altered many control techniques and as a result, a few formerly 16 major pests have been reduced to a secondary position. Elmore and How- land (1943) considered some cultural practices as undesirable because of their adverse impact on TPV7 control. These include failure to destroy aban- doned plantings, careless disposal of infested culled fruit, and use of infested seedlings. In Florida, Swank (1937) recommended that all mate- rial remaining in the field after the crop is harvested be carefully plowed under. He suggested that the carelessly abandoned fields could become a reser^/oir for infestation of a nearby or succeeding crop. Poe (1973) stated that the best control for TPW is based on several cultural practices: use of non-infested seedlings, destruction of plant debris, use of light traps for adults in small areas, and destruction of plants growing from seeds in compost heaps. Price and Poe (1977) reported that staking and artificial mulching of tomato plants reduced damage caused by K. lycopersicella and other pests. Biological Control of TPW Employment of biological control measures for insect and mite pests of row crops has been limited, and the poor record probably relates largely to the short-lived row crop environment, which presumably does not permit establishment of the effective host-natural enemy relation- ships which often characterize more stable environments (van den Bosch et al. 1976). Modern-day biological control techniques have not been fully exploited in tomato under field conditions (Lange and Bronson 1981) . They have been widely accepted in European glasshouse tomato pro- duction, however. Reports on parasitism of K. lycopersicella were made by Elmore and Howland (1943), Swesey (1928), Thomas (1933), Oatman et al. (1979) and Poe (1973) (Table 1). 17 3 S 0 C) m n ro K CN ro ro (Ti 0^ crv t; T5 7-t H rH c C > U} cn 0) dJ Q> 11 (11 M M tn fi Fi n aj o n g :? j:: i-l f— 1 w H H W w n m n ro m m ti ro ro ^ ^ ^ -* ■q* "d" i UH o n o •rl M 0 ^ •rl Ul MH m HH P, ro c •iH ■rl -rl n •=5 c u r-i rH rH fa rH ro OJ 0) 0) 0) '0 c c ro ro ro ro •H ro c ro ro ro ro ro ro ro ro ■rl 0 0 Tl ■n TI T) rH Ti 0 T3 Ti 03 T1 'O t! 13 TD £ a •H ■H ■H ■H ro •rl s -ij -d -H ■H ■rl -H "d -^ ro p 3 c c C c Fi c 3 -S -5 c c: c j: •5 -fi h (U tn m u ■'^ ■rl (1) 0) ' ■h n •H fi ro ro -? c m •rl rH M ii ■P rH tn 3 ■H 0 (1) 3 XJ t-n fi 4-' fi ^ 3 ■rl 3 tn o Ih D4 t) •H ro in 0) 0) (/) -n D m W tn •H 0) •r4 iH OJ ro T) •H ro h •H > Fi •H £ 0 •rl iH CJ OJ U 0 (1) c f) -fi rH Q) ^ 4J OJ ro 4-1 .fi m ^ O4 tn Hi i/i 3 X (1) 0 U5 CI) H 0 3 i-t n) ro C 04 4-1 r-t 0 n G 0 ^ Oi ro 4J OJ H 04 ro £ ro < u u u < ro O4 T3 Q4 -H Ul m • l •H p P -G 0 u a: cn o ro cn (1) •H rj 0 01 u < 4-) ro N i-H 0) rtJ ^3 M a) 4-1 4J o (C iH m II (U ,Q m m ••h -p -rH -P > O C) 3 Tl 0 n 4J s^ Its ft c <1) C) II H -P cd (0 H !m M (U 29 C o E a o > Q CM > o (A CO E o H 30 scale of 0-4.5 (Table 2) was devised, based on personal obseirvations and the damage descriptions of Batiste et al. (1970a). The leaf injury length (cm) was also measured and larval head capsule width recorded. Finally, the presence or absence of TPW larvae in different types of leaf injury was determined. A simple linear regression model was used to examine the relationship between the head capsule width and injury length, and between larval ins tar and the damage rating scale. The eval- uation of the different methods of damage assessment was discussed with regard to the practicality of their use for scouting programs . Results and Discussion The Tomato Crop A STommary of leaf area and n\imber of flowers and fruits is shown in Table 3. Tomatoes planted during October, 1980 began to flower 61 days after plant emergence and fruit set occurred at 73 days. Maximum leaf area was reached at 134 days. Tomatoes planted during November, 1980 started blooming at 54 days, and fruit set occurred at 68 days. Maximum leaf area occurred 88-112 days after plant emergence. Tomatoes planted during December, 1980 and January-February, 1981 had a shorter vegetative period, with flowering at 42-62 days and fruiting at 49-62 days. Leaf area reached a maximum at 63-89 days. The total leaf area during these plantings was lower than that produced from fall plantings . Under southern Florida conditions, average temperature changes drastically from autumn to early spring (Mitchell and Ensign 1928) . In this area, the effect of planting date determines growth and tomato 31 Table 2. Classification of tomato pinwojon leaf damage on 'Flora-Dade' tomatoes, based on greenhouse and field observations. Home- stead, Florida, 1980. Degree of Damage Description 0 No damage 1-1.5 Mining of leaves, ca. 0.50 cm or less in length; mine narrow and elongate; tissue transparent; mine on any part of the leaf- let; some leaves attacked by more than 1 larva; small larvae present. 2-2.5 Mining of leaves ca. 0.51-0.68 cm; 1/4 of the mine is necrosed, but changing to a raised area or oblong to ovoid blotch; frass accimiu- lation at the bottom of the injury. 3-3.5 Blotching of leaves; blotch necrosed over 50% of the injury; no holes indicating lar- val exit; size 1-2 cm; epidermis of the leaf opaque to chlorotic due to larval injury to midvein; construction of silk tent in epi- dermis. 4-4.5 Leaf folded; fold can occur at any lobe of the leaflet. Necrosis extended to 75-80% of the leaf; extensive frass accumulation on blotch or fold; injury length 2-4 cm. 32 > U o 0) s ^ 0 Ul rH a) fr. 0 ni H 0 S CO 3 S-( !ji ro 3 G >1 0 -rH +J ■H +J C -P U 3 cn iC 3 0 >, c T! U 0) -H 0 Q S i-i 0) •• U aT3 0) d) 0) fl cn O u a < M T} " OJ C TJ p rC (T3 U-l O sC '3 JJ 0) cn M 0) IT3 5 0 4-1 K o o, rH , H, r-i, ^, ^, f^. .-^ + 1 + 1 + 1 + 1 + 1 + 1 + 1 •4-1 + \r\ O o o o r- ^ ^ O CN CN 00 CN -i, rH, O + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 CO 00 00 ^ in ^ CM CN 00 0 ro r~ o cn n ^ in -d- in r-- cn CO CO n D H CN iH o. iH . H .^ a) tn (0 +j M 9) > ■rH iTJ 01 4-1 0) 0) m tjl ifl a) +j > tn >. 4J rH c U CI) fl u o en m Ul b 0) o m V n R C) u. P C) T! rH f1 <1J Sm > (-1^ (1) (IJ T) H ^ II O B^ H Cfl (1) ^ rji o. 01 Di -p H - Pi r^ H 41 ,uip.j ea J V * B 9T 42 Table 4. Stage of development description for tomato cv Flora-Dade. Description is based on the average of observations from tomato plants grown during Fall 1980 through Winter 1981. Homestead, Florida. Plant Stage romato Plant Description Vegetative TV, TV Reproductive TR TR TR Plants 1-15 days old. Complete formation of 2-3 primary leaves; loss of cotyledons; plant height ca. 5-7 cm. Plants 15-35 days old; plant erect (12-16 cm) ; 5-7 leaves, development of laterals; plant with only 1 main stem. Plant 35-40 days old; development of laterals from nodes 1-5; at leaf 4-5 the stem bifurcates producing another stem as vigorous as the first main stem; production of floral clusters at node 5 and second main stem; height 50 cm. Plants 67-70 days old; fruit set; plant postrated; yellowing of primary leaves. Plant 109-135 days old; 90% fruit ripe; post- harvest maturity; at least 60% of the primary leaves necrosed, development of secondary laterals at nodes 3-5; plant totally postrate; height ca. 32-57 cm. Table 4 — continued. 43 Senescence Plant 140-200 days old; dead leaves on main stem and second main stem; regrowth of plant from auxiliary buds at nodes 1 , 2 and produc- tion of up to 3 floral clusters may occur; possible fruit development. 44 The principal application of this nomenclature system is to deter- mine the amount of yield reduction produced by damage inflicted at given stages of plant development. As an example, if I use Keularts' (1980) data from his experiment in tomato defoliation, 20% defoliation of lower plant leaves at stages IV^ through TR2 did not alter mean yield per plant. However, 20% defoliation of upper plant leaves at TR2 stage caused yield reduction. The nomenclature system can apply to single plants or entire crops. It would be worthwhile to apply this system to other tomato cultivars. Methods of Damage Assessment for TPW Larvae Average leaf area and weight consumed by TPW larvae. The data from this experiment demonstrated the complexity of measuring TPW foliar dam- 2 age. The average leaf weight (mg) and leaf area (cm ) consumed by larvae of a determined instar are shovm in Table 5 . Average leaf area consumed ranged from 0.5 to 1.57 cm for 1st to 4th instar. First and fourth instar larvae consumed 5 and 13.42 mg of leaf, respectively. Variance of leaf weight measurements was large suggesting that many uncontrolled factors influence feeding of individual larvae in the field. Either method might be used for laboratory and greenhouse experiments where the researcher would have more control of the factors influencing variability (e.g., leaflet size, leaf age). Length of Foliar Injury and Use of Damage Scale Length of foliar injury and TPW head capsule width were cor- related (r=0.&3; P=0.001, F=39.33) (Fig. 4). Furthermore, there was 45 r-i U c Q> t^ U m 0) G u-l C) tw -H -r-l JJ Ti -H T) >1 C 3 3 D tfl 0 m c ^ a 0 c (C O Q) u 14-1 Ti ffi C a) 3 H (V) T) rH c H (0 OJ o JJ •H ^ «3 en !j s o 1-1 n3 •H 0 ^ -p CJ fl M-t Fi ■H (1 (U H i^ + 1 + 1 IX in in rH 'I' i-i r- o o iH n cN tn 00 00 CM r~- ^3 C ■H OJ O o o Pi + 1 + 1 03 (1) CN "T CNI CM M rr o M CO iJ 00 CN iH O O + 1 + o ^ rf in (71 in r-i o O ■n e C) x: — en a ■H 'O Q) TI Q) OJ u (1) S § < MH !/] 'U U1 .■a c 113 c 0) 0 CU n J u J u e (> C) -p II (N c u 0) (1) •.. :^ j: ■p jj Cl) rri M c o a H H V) & c 3 o •r~i u O at u. ( ui3) ((}3u 91 Xjniu I 48 a significant relationship between TPW (Fig. 5) larval instar and the degree of damage observed (r=0.79). Both techniques suggest the possi- bility of prediction of damage level in the tomato plant at stages TV-, - TR, . Such a prediction may be influenced by other factors such as plant stages and larval density. The use of larval instars to determine injury length has a reduced bias compared to use of TPW damage degree scale. Foliar injury measure- ment is only advisable for research experiments (e.g., plant resistance, pesticide screening) in which the time frame available to determine the dependent variable is not a constraint. Other aspects to be considered for further study are larval preference for larger or smaller leaflets, as well as presence of different larvae in the same leaflet. The use of TPW damage scale is perhaps less precise than the tech- nique mentioned above. Damage scale technique may introduce personal error in measurement of larval instar in relation to degree of damage. It is possible, however, to use this technique as an adjunct aid to the population index (niomber of injuries per plant) . As an example, using the equation y=0.80 + 0.795x, where y = the leaf injury damage scale and x = the tomato pinworm larval instar; if the value of x equals 3, the average degree of damage in the plant will be 3.18. This information will help to determine the effect of the insect in economic terms, once the economic threshold is reached for plant stages TV - TR . At this point there is no information available for TPW EIL values for plants in these early stages. Therefore, further studies will be necessary to indicate that the presence of a particular larval instar is capable of producing a determined economic damage. c tn o o •H Ul +J (0 0) in 3 tn a) -H M > P ■H c 50 -^ -n — pi o. c o o E o 51 The ratio of percentage of larvae present to percentage of larvae absent (Table 5) in the observed injuries was 4:1 for the folded necrosed injuries, 31:1 for the folded with no necrosis, 1:3 for blotches with necrosed tissue, and 3:1 for transparent blotches. Consequently, the use of necrosed blotches will indicate that ca. 77% TPW larvae will be absent from the observed blotches. If a high niomber of injuries per plant falls in this category, the probability of not measuring larval presence in each injury is increased. We can deduce that necrosed tissue generally indicates that larvae are already attacking the fruit or other leaves, or have left the canopy to pupate. In a crop such as tomato where the margin of profit is great, expensive methods of control are usually dictated. The use of a system that will predict the damage level to the plant requires a high level of accuracy. It is suggested that the method described here is advisable for plants during stages TV to RV, . Conclusions and General Discussion Studies of tomato growth in different cropping seasons are useful to determine effect of planting time on plant development. Tomatoes, 7 cv Flora-Dade, planted later in the winter have less (ca. 117 dm^) 2 leaf area than those planted early in the fall (ca. 253 dm leaf area) . Thus, those crops planted in October-November may be able to support more damage than those planted in January-February. The proposed system divides the plant stages inro 2 vegetative stages (TV^ - TV2) , 3 repro- ductive stages (TR-,, TR and TR^) , and a senescent stage (S ) . The description of the developmental stages of tomato can aid in using pest management tactics. Definition of shorter developmental stages with 52 Table 6. Percentage of tomato pinworm larval occurrence in foliar injuries with different phenological characteristics. Percentage of TPW Larval Occurrence Damage Description Present Absent Transparent blotch 72.5 27.5 Necrosed blotch 23.3 76.6 Folded, no necrosed leaf 96.87 3.12 Folded, necrosed leaf 81.25 18.75 53 with more subdivisions would enhance phenological plant description. This may allow better pest monitoring when plant development is in the TR stage. 2 Results on leaf weight consumed (mg) and leaf area consumed (cm ) provided information on increments of those parameters for each larval instar. Standard error and confidence intervals demonstrated a high variability for both methods. Further research is necessary to deter- mine if such variability is caused by larval behavior or by use of different leaflet sizes and leaf area. I consider the leaf weight method promising in such areas as plant resistance and behavioral chemicals (deterrents) evaluation. Damage assessment based on the leaf area mined by TPW is not considered appropriate for monitoring TPW density because of inherent variability in insect behavior and plant morphology. Injury length has proven useful in evaluating plant resistance (Schuster 1977a) . The relationship between larval head capsule 2 and injury length was intermediate (r =0.47) . The regression equation developed in this study can be used by plant resistance evaluators to determine feeding inhibition at a given instar. This technique has to be carefully used, however, since it is dependent on the type of leaflet consiimed. Larvae that attack small leaflets might develop as well as one in a large leaflet but the injury length will be smaller. Data gathered from the visual damage classification proved to be useful to evaluate damage inflicted by TPW. Since TPW instars have a distinct behavior as leaf blotchers and leaf tiers, it will be easier to develop knowledge in which the average larval instar will determine the damage degree in a plant. 54 Scouts should use different techniques at the same time if possible. A population index, degree damage scale and a survey deter- mining the real presence of the larvae in the foliage will provide a better estimate than a single technique. More research is needed to evaluate these techniques together. Evaluation should be based on time expended and reliability of the methods . Further study of the relation- ship between several types of foliar damage and direct damage to the tomato fruit is needed. CHAPTER III SPATIAL DISPERSION OF TOMATO PINWORM EGGS ON TOMATOES In troduc tion Tomato pinworm (TPW) is one of the most important pests of tomato Lycopersicon esculentum (Mill.) (Watson and Thompson 1932, Oatman 1970, Poe et al. 1974) . Little is known, however, about ovipositional pat- terns of this pest on tomato plants under field conditions. There is some indication that caged moths under laboratory conditions deposit eggs indiscriminately on all parts of the plant including the upper leaves (Elmore and Rowland 1943). Wellik et al. (1979) indicated that lower portions of the plant should be examined in the field for both larvae and eggs of the TPW. Studies of TPW egg dispersion are necessary because this knowledge affects the sampling program as well as the method of analyzing the data. Furthermore, dispersion patterns can be used to give a measure of population size as well as to describe the factors that may affect the condition of the population. This paper (1) describes the spatial distribution of TPW eggs on field-grown tomato plants under varying levels of TPW infestation, (2) presents an evaluation and discussion of factors affecting this distribution and (3) discusses sampling strategy. 55 56 Materials and Methods Experimental Plots To test for a possible relationship between oviposition of TPW and different leaf strata of tomato cv Flora-Dade, 8 plantings (Oct. 3, 1979; Dec. 5, 1979; Jan. 8, 1980; Oct. 30, 1980; Nov. 25, 1980; Dec. 30, 1980; Jan. 30, 1981; Feb. 28, 1981) of non-staked tomatoes were evaluated at the Agricultural Research and Education Center, University of Florida, Homestead, Florida. Each planting (ca. 450-947 plants) was direct-seeded in raised beds (3-5) (ca. 45 m long) of Rockdale soil, and mulched with light colored plastic. The seedbed's midlines were 182 cm apart. Plants were spaced 38 cm apart. Sampling Methods Sample size was selected by a preliminary random sampling of 50 plants on 2 dates. The method described by Elliott (1979) was adopted. The relative variation (SE/x) x 100) was calculated to compare sam- pling methods over a variety of sampling units (Hillhouse and Pitre 1974, Ruesink 1980) . Ten to twenty plants in each planting were randomly selected on a weekly basis from February 7, 1980, through May, 1980, and from Jan. 27, 1981, through May, 1981. Whole leaves of the plant were first examined to determine differences in ovi- position on lower and upper leaf surfaces (Plantings 1-3) and to detect differences in oviposition in different plant strata (Plantings 1-8) . A plant was divided into upper half and lower half in the first 3 plantings (1979-80) and divided serially into 6 sections (upper, middle and lower of each of the external and internal canopies) (1980-1981) . 57 External canopy was defined as extending from the periphery to 5-15 cm into the plant interior- The variance was stabilized by fitting the niimber of eggs obtained to a suitable model (Poisson and negative binomial) and transforming to logarithm (x+1) or x+0.5 (Elliott 1979) depending upon the original frequency distribution of the counts. The mean counts of eggs in upper and lower strata were compared by student's t-test for plantings 1-3. Egg densities in the 6 strata for plantings 4-8 were compared by analysis of variance (ANOVA) . Means were grouped by Duncan's Multiple Range Test (P=0.05) . When tests indicated significant differences in egg densities between strata of plantings (4-8) , optim\am sample allocation among strata was determined for each planting date (Cochran 1977) . Population Distribution Related to Leaf Position To test differences in oviposition of TPW related to the vertical distribution of the leaves with respect to the main axis, 17 randomly sampled plants, each of which were 45 days old, were observed in a commercial field. Leaves were numbered from bottom to top and the num- ber of eggs recorded. Data were analyzed by ANOVA and means were separated by use of Duncan's Multiple Range Test (P=0.05). When t-tests indicated significant differences in egg densities between leaves , optimxim sample allocation among leaves was determined (Cochran 1977) . Distances Between Eggs and Effects on Distribution To determine if TPW egg distribution pattern is influenced by leaf- let size and egg density, the frequency of egg deposition on each 58 leaflet was recorded. Then, distance between eggs on each leaflet was counted on 40 middle leaves collected from plants located in the same field mentioned before. Several authors (Cottam and Curtis 1956) proposed methods to evaluate randomness in spatial distribution of the population by measurement of distances between individuals. In this experiment, distance between eggs was checked by measuring the shortest straight line between nearest neighbors with a metric ruler. Distance accuracy was 0.05-0.25 cm. The frequency of occurrence of each dis- tance was evaluated for egg densities. Also, simple linear regression was applied to determine any relation between egg density and leaflet Oviposition Related to Plant Age To determine if plant age affects oviposition, the number of eggs on each plant was counted on 60-80 plants ranging in age from 2 to 21 weeks. Plants in this experiment were in the same field as previously described tests (plantings 4-8) . Plants were inspected weekly during April and May, when the highest peaks of oviposition occurred. Data were subjected to ANOVA, and means separated by use of Duncan's Multiple Range Test (P=0.05) . Results and Discussion Selection of Number of Sample Units The main objective of planning a survey should be to obtain the required information with a minimum amount of labor. To achieve this, it is necessary to select a number of sample units that are in agreement 59 with the desired degree of precision and cost. This requirement is difficult to meet in practice. First, an acceptable index of precision (SExlOO) is 25% (Barfield 1981). Secondly, the actual cost of sampling X tomatoes is 7 dollars per acre (Table 7) . Sample size was selected by a preliminairy random sampling of 50 plants in 2 dates (Table 6) . Three major criteria were followed to select sample size. First, following the criteria outlined by Elliott (1979) , a suitable sample size was selected when the mean value ceased to fluctuate. It is observed (Table 5) that with an increase in sample size from 10 to 25 (at low egg density) , the resultant mean (x) fluctuates around 0.4-1.5 eggs/plant. Also, at higher egg density (2-7 eggs/plant) the n\jmber of selected sampling lonits is 20-25. Second, the use of index of precision (SExlOO) over different sampling units is X a more adequate technique to select sample unit size. Accordingly, the lower index of precision (Ip) was obtained when the number of samples equals 50. Therefore, the percentage of the standard error of the mean can be 34% if the TPW egg density per plant is low (0.4-1.5 eggs/plant). This percentage is not good enough to make pest management decisions. The index of precision can be 20% if the density is higher (2-7 eggs/plant) . Third, the sample number does not reconcile with the actual budget per acre. Cost of sampling eggs is 1.4-77 dollars (Table- 5) more expensive than the actual sampling cost per acre. The number of samples for a fixed level of precision (random sampling) was calculated. A random eag distribution was assumed, n=(—^) where. Ex n=number of samples required, s=standard deviation, x=mean, and E=predetermined standard error (e.g., 0.25). For instance, at 60 en -H O rH u a, X o CO rH "^r m -^ ^^ o r- vD in in n IX CM \ w C r-i E 0 • d ai ~ 0 CT •H 03 a ■^ T3 fl -H a^ M g 0 0 H U Cn >1 ^ CO tN v£l rr 00 CN ^D •^ •sj ^ CM CO iH cN r^ o ■^ fH CO 10 in n O rH O O O O O •'3' ON r-- CO "d* i£> CO in r- n H •^ O CO CO 'T CN rH rH O O O O i-l rH rH 1-1 O r-- n ro CN CN (N in o in o Lfi o r-H iH CN CN in in o Lfi o in o en rri ■H rr. 01 0) (11 Xi c m r~- -i 3 rH Ti () (1) 4J i-H J) M c It! fl 0) (ti O •? TI r-l •H 04 tn 4-J tn X IT1 c c tn ;? fl o m iH () m fn Oi 0) G m •H ^ rrt c iH 0) ■5 m a, o^ (1) >1 (T3 4-1 4J tn c -H c Q) in tu U-l 'A c o Ui T! :5 4-1 0) j: Ul b tn m o ■H ryi •rH u +J iij x; 61 endemic levels of TPW egg population (0.4-1,33), t±ie niamber of samples to be taken, being S=2.79, E=0.25, x=1.15 will be 94, with a cost of 158 dollars per acre. If the TPW egg population is epidemic (2.16-7), the number of samples to be taken will be 28, being s=3.82, x=2.9 and E=0.25. The cost of sampling will be 47 dollars per acre. Accordingly, under low TPW egg densities, increasing sample precision as the sample size increases is not worth the work required in taking larger samples. Consequently, I selected sample sizes of 10-20 which gave the best practical results per unit of work expended ($16.8-25.2 dollars per acre). It is considered that sampling TPW eggs is not a practical method to make spray decisions . Statistical Description of TPW Egg Spatial Distribution The use of statistical methods, e.g., t- tests, analysis of vari- ance, involves several conditions described by Snedecor and Cochran (1967) . One of them is that data must follow a normal distribution. The distributions of density measurements on plant samples are sum- marized for each planting in the Appendix. These statistics (Table 8) support the hypothesis that TPW eggs are clustered on plants. This clustering was more apparent when TPW egg densities on each plant ranged from 0.302-1.3. As mean densities increased, variance also in- creased except for planting 6. Values of the negative binomial parameter (k) (Elliott 1979) range from 0.451-0.013 for my sampling. Thirty-two percent of the weekly counts for each planting were fitted to the nega- tive binomial distribution (see Appendix) . For plantings with higher 62 ■p m c O (tJ •P 1-1 u ft 3 y. >1 Xi V c tn (a tn r-i OJ ft c :s e r-\ (T3 63 population densities (average 0.302-1.30), kurtosis and skewness decreased as the mean increased. Skewness values were all positive. This indicates that egg distribution tails off among higher counts. This information in conjunction with the data indicating clumping can aid in sampling design. Distribution of Eggs on the Upper and Lower Siirfaces of Leaves Statistically significant differences (P=0.01) were found for egg numbers on lower and upper leaf surfaces. Eighty-nine percent of the total eggs found per plant were on the lower surface (Table 9) . These results and the results from the greenhouse contrast with those found in caged plants by Elmore and Rowland (1943) , who detected 45% of all egg deposition on the upper surface of the leaves. Insect pre- ferences for oviposition on the underside might be correlated with dif- ferences in pubescence of the 2 leaf surfaces. The average nxmber of trichomes on the underside was 1441 per leaflet as opposed to 469 on the upper surface. This may also indicate preference to avoid egg desiccation, or to avoid higher light intensities during oviposition (Hinton 1981) . Distribution of Eggs on Upper and Lower Halves of the Plant Statistically significant differences (P=0-05) were found in the nimiber of eggs deposited on the upper half of the plant vs the lower half of the plant for the 3 sam.ple dates in the first planting (Table 10) . The upper part of the plant had more eggs on 13 of the 15 sampling dates. There were no significant differences between upper and lower halves 64 Table 9. Ovipositional preference of tomato pinworm for upper and lower surfaces of tomato leaves from plants grown under greenhouse and field conditions. Mean Number of Eggs of TPW Greenhouse Leaf Side Field^ Caged Plants Upper 0.851^ 0.10 Lower 7.3763 2.77 Mean based on counts from 80 plants. b Mean based on counts from 60 plants. c Numbers were significantly different at P=0.001. 65 ,^ o Ul O) C) (Ti > rH t-i (rt ^ ^ (T, T! ^ ■H a) S>< 15 0 () rH 1-t u< T! « C >l irt +J c in p (U 0 a U ft 3 0) '-' 'C ■-I rt o n o o -J o f-1 c ^ tN r^ vD O rt c C C e •H u M o ft ^ fi c (t1 •H cn ft -U 0 c -!-l (1) (T3 :m g 0) o M-l +J in •H 4-1 Ti C) JJ M fl (1) = ! r. 4J fO to ^ (U in IT! k s a 66 in the second planting (Dec, 1979). Analysis of the data from the third planting (Jan., 1980) indicated significant differences in 6 of the 11 sampling dates. The upper half of the plant had more eggs except for 2 dates. In general, when numbers of eggs were higher in the lower strata, this coincided with younger plant age (40-60 days after germin- ation) . Numbers of eggs were higher in the upper strata when plants were in reproductive or older age (75-80 days after germination. These data indicated that for 'Flora-Dade' ground tomatoes, ovipositional preferences existed based on the level of the plant. Because of the low econom.ic threshold for TPW in tomatoes, it may be necessary to reduce the sampling unit to detect major differences in internal and external parts of the plant when populations are low. Consequently, smaller sampling units were tested in subsequent experiments. Distribution of TPW Eggs by Sampling Six Plant Strata Statistically significant (P=0.05) (Table 11) differences were not detected among tlie strata for the 4th (Oct., 1980) and 5th (Nov., 1980) plantings possibly due to the relatively low mean egg niimbers per plant. Hov/ever, the highest number of eggs oviposited was obtained in the upper external canopy for planting 4 and in the middle internal canopy for planting 5. Tliere was an increase in eggs for the lower internal canopy in planting 4 during January and Febrioary, when nocturnal temperatures were lowest (2°C) , and an increase toward the upper external portion of the plants when temperatures were fluctuating between 17-29°C (April-May) . 67 (T3 (3 e g 0) (U ^ -C +j 3 U-l • • o m -p >i (ti a k 0 C -P to c rH -H 04 M i^ c a ^ ^ n r-* S z :i ■73 -2 :i g "3 *- 0 3 3 3 3 70 Perhaps moths protect themselves from the cold temperatures by staying close to the ground in the lower canopy. Despite these assvraiptions, when number of eggs found per stratimi was regressed (Table 12). against temperature, there was no evidence of a relationship between the two variables. Significant differences in niombers of eggs per stratum were detected for the 6 (Dec, 1980), 7 (Jan., 1981) and 8 (Feb., 1981) plantings. There was no significant variation among the six strata during juvenile plant stages. Most of the significant differences were observed (Fig. 6) during the mature stages (TR) of the plant. Concen- trations of eggs in the upper external strata varied slightly among plantings. In plantings 4 and 5, eggs generally occurred on the top and middle external canopy during the last weeks of sampling (April and May) , and on all strata during the first weeks (juvenile stages) in March and April. When mean numbers of eggs in the external and internal canopy were added to reduce the strata to 3 (upper, middle and lower) , no statis- tical differences were observed despite the stratum reduction. This agrees with the results expressed when 2 strata (upper and lower) were sampled, indicating that differences in oviposition tend to be masked if the units are widened. In general, the upper external strat-um had the highest number of eggs, followed by the middle external and internal strata, during most of the sampling dates. At plantings 6 to 8 , the TPW eggs occurred in greatest abundance on the upper and middle external strata dur- ing all growth stages. More eggs (44-68%) were deposited within the upper external canopy of the plant than in any other stratum (Table 13) . Four to twenty eight percent of the eggs were laid in the next (middle external stratum) . The lower external stratum had the lowest range (1-11%) ; however. 71 Table 12. Relationship between daily mean temperature ( C) and TPW oviposition in 6 tomato plant strata. Homestead, Florida, 1981. Independent Variable Dependent Variable No. Eggs/stratimi bo bl' Temperature upper internal upper external middle internal middle external lower internal lower external 0,05 0.22 0.13 0.36 0.10 0.31 0.02 0.14 0.01 0.10 0.13 0.36 -0.007 0.005 0.46 0.04 0.10 0.01 0.02 0.0056 0.10 -0.02 -0,11 0.01 Coefficient of determination. Correlation coefficient. Intercept of y axis. Slope. -P 3 +J • c 0 c c 0) rH OJ (T5 g i^ 1^ a ^ 0) 0 0) U-l rH ■-I M-l r-~ 0) T! ■H > fO T! Cn OJ ■H C TJ g vi) •H -P M-l •k Oi C 0 !-l c «. (0 en ■H (N M 0) +J -P cn U ^ w 3 r-i -p tn 0 *• c i^ 10 (C ^ a +J H +J 0) (tl Oi § u -p !-( i^ T3 en (1) 01 c ft o 4J 4J o c • CO c 0) (0 rH 01 0) tn iH d cn M II eu c 0) 0) CN u l^ Pi (U g U-l Eh +J p ■H 4-) c 0 TJ c •H s • (U c -P r-l g u ■H fS CO a d) a 03 •H CO 0) « g 4.) CN TI OJ 0 C iH -P ta « 14-1 T! iH ^ 0 T) M-l a 0) •H 0 0 fa 0) 01 g M ■p (0 ^ (U OJ 00 -p H (= en (JJ ^ b 01 a 3 +j C 0) a c ■H > 3 CN -P •H a) c -P Oi fi (0 ca VD (« •H H -P !-4 a^ 0) >^ dj cn 01 in > 0) ,,— s (1) < -p rr-l m > 'a c '^ kO 01 iH 0 r^ c CD ej fl <; fa en m OJ 73 in} e J IS/ $33^ ueaM 74 Lrt CSI uiniOJ4S/<3Q3 uoay^ 75 Table 13. Percentage distribution of TPW eggs for each stratvm of tomato plants in 5 tomato plantings. Homestead, Dade County, Florida, 1980. Planting Date Oct. 30 Nov. 25 Dec. 31 Jan. 30 Feb. 28 Stratum 1980 1980 1980 1981 1981 Upper external 68 32 44 46 51 Middle external 4 15 23 28 21 Lower external 1 3 10 11 10 Upper internal 0 3 7 0.6 5 Middle internal 9 43 11 10 9 Lower internal 16 0. 8 4 1 2 76 growth of the plant upwards and outwards can mislead my interpretation of actual ovipositional preference. The percentage of eggs found per internal stratum ranged from 0-7% in the upper internal, 9-43% in the middle internal, and 10-37% in the lower internal. TPW oviposits mainly in the upper external canopy when egg populations range from 0.75-1.5 and when the plant was in its reproductive stage. A lower proportion of eggs was found in all other strata. Sampling 6 plant strata demonstrated that TPW tends to oviposit in the middle and upper canopy. It is necessary to use sample allocation (n ) , as outlined bv Cochran (1977) to minimize sampling cost or vari- h 2 ance (s ) . I assumed equal sampling cost for each stratum. Sample allocation was estimated on dates in which statistical differences in oviposition were detected. In general, more samples should be allocated to the upper and middle external strata (Table 14) . Because TPW eggs are clumped in the upper and middle canopy, these strata had the highest variance (see Appendix) (Tables 51-54). For a fixed total cost, n = (C-C )S h h where (C-C )= L ^N, S / C ^ .h h h , Z C n , nh= n W S n NnS, Thererore as S,- increases so does nh. 1 1 h h = h. n i=l ^W, S SN, S h h h h The average number (n=20) for all planting dates was 6, 5 and 3 samples from upper, middle and lower external canopy, and 1, 4 and 1 from upper, middle and lower internal canopy. Allocation ranged from 5-10 samples for the upper external canopy (see Appendix) , and ranged from 2-8 samples for the middle external canopy. I considered this sample allocation to to be the best, because standard error (SE) of the sample mean was more C 3 M o 77 a 0) H fl) ^-^ m m H) +j m to nj -P 4J (n c «i 0) 0 rr> u c 0) ■H [(1 !-< 3 t: T! c V.0 3 -O ^ o n u M-l ti (!) c !h CJ ■H +J +J M IT! >-( () •H 0 U4 78 constant through time (range: 0.20-0.66). There were exceptions for these sample allocations. For instance, during the month of February (planting 4) , more numbers of samples were allocated to the lower inter- nal stratum (see Appendix) (Table 54) . Another aspect that requires more understanding is the relation between phenologic'al stages and sample allocation. As an example, it was observed (see Appendix) that when plants were in vegetative stage (TV) , more samples (n^=18) , should be allocated for the upper external and middle internal canopy. When plants are in first reproductive stage (TR^ ) , more samples (n^=6) , are allocated for upper external stratum. Finally, when plants reach the second reproductive stage (TR^) , all strata had similar sample allo- cations except for lower internal canopy (n, =0) . Egg Distribution Influenced by Leaf Position During hea^/y oviposition (avg 21.94 eggs per plant) on 45 day-old tomato plants, the highest number of eggs was observed on leaf number 4 (Table 15) . The numbers of eggs on leaves 3 and 5 were statistically equal to those found on leaf 4. The number of eggs decreased sharply on leaves adjacent to the apical point toward the bottom of the plant (leaves 1-2) . Tjese results indicated that middle leaves of 45 day-old plants under conditions of high egg oviposition (1-5.5 eggs per leaf) have 65% of the total egg population. These data differ from those obtained in experiment 1. The higher number of eggs per plant indicates that the insect tends to oviposit in the upper-middle canopy, avoiding the 2 top and bottom leaves of the plant. Several factors may influence the ovipositional pattern. First, these results agree with Hinton (1981) 79 Table 15. Mean tomato pinworm eggs on tomato leaves from different strata of 45 day-old plants. Homestead, Florida, 1980. Mean No. No. Leaf Position Eggs/Leaf Leaflets/Leaf Eggs/Leaflet 1 bottom 2.20^* 7 0.31 2 3.20^ 8 0.40 3 middle 5.40^ 11 0.49 4 5.50^ 11 0.50 5 5.00^ 11 0.45 6 top 2.00^ 8 0.25 7 1.00^ 7 0.14 Numbers followed by different letters were significantly differ- ent statistically at P=0.05 according to Duncan's Multiple Range Test. 80 who stated that species that lay eggs on plants have a marked preference for laying a certain height above the ground. Secondly, the insect may be avoiding overcrowding in the smaller top leaves and competition of foliar consumption by TFW larvae on the lower leaves. The highest sample (n=17) allocation was for leaves in the middle canopy (Table 16) . 2 Higher variance (s =44.4) was found for eggs deposited on those leaves, as was a high mean (x=6.5) . This is caused by egg clumping in the canopy. The fourth leaf had the highest allocation sample (n, =5) , followed by the fifth leaf (n, =4) . The lowest allocation was for the bottom leaf (n, =1) . The standard error of the mean sample was lowest (SE/x=0.21), for the third leaf and slightly higher (SE/x=0.24) for the fourth leaf. There- fore, when higher density and large variance are found, the leaves selected should be the middle ones . Sample allocation was reduced for bottom and top leaves. These leaves had smaller variance and smaller density than the middle ones. Distances Between Eggs per Leaflet and Effect on Distribution In the present study, the results indicated that TPW egg density was not related to leaflet area (Table 17) . The coefficient of deter- 2 mination (r =0.025) indicated that females tend to oviposit different egg densities disregarding leaflet size. Therefore, any leaflet can be selected as the sampling unit. Frequency of egg occurrence per leaflet was not related to distance between eggs. Low coefficients of deter- mination (r^=0. 19-0.23) between frequency of occurrence at different egg densities (2, 5 and 10 eggs/leaflet) and egg distances indicate lack of linear relationship between these variables. The slope (bl) obtained 81 w O en tn 0) nj Eh 'T 82 Table 17. Relationship between frequency of occurrence of TPW eggs per leaflet as dependent variable and distance among eggs and leaflet area as independent variables. 2* *** Dependent Independent r r b h-^ Variable Variable No eggs Leaflet area 0.026 0.16 1.99 0.04 a b 2 TPW eggs Distance among 0.23 0.48 1.75 -0.15 eggs 5 TPW eggs Distance among 0.19 0.44 1.47^ -0.12 eggs 10 TPW eggs Distance among 0.23 0.48 1.38^ -0.22^ eggs __ r =coefficient of determination. ** r=correlation coefficient. *** bQ=intercept of y asis. tb, =slope. numbers were highly statistically significant (P=0.01). n\mbers were statistically significant (P=0.05). 83 for an^ egg density was negative and highly significant (P=0 001) . This can be explained in Fig. 7 , where the frequency of egg occurrence at dis- tances higher than 3 cm was as low as 5%. The average distance between eggs was 0.5-0.75 cm. The average number of eggs found on each leaflet was 2-3. These results agree with those expressed by Poe (1973) ; in the present study the number of eggs on each leaflet was as high as 11. Eggs tended to be laid more uniformly in some parts of the leaflet. Per- haps the female lays 2 eggs successively on a certain part of the leaf- let, but is likely to move away after oviposition. The arrangement of eggs may also be a reflection of heterogeneity of conditions among parts of a leaflet such as pubescence and leaf venation. From the practical standpoint, these results can be used to determine use of single leaflets as less variable sampling units compared to the whole plant. A more detailed study of female behavior is necessary to determine the role of leaf factors (e.g., pubescence) affecting oviposition. Differences in Oviposition Related to Plant Age The relationship between oviposition and stage of plant development was determined during the study of plantings 4-8. Statistical differ- ences were detected among these plantings (Table 18) , when plantings were 19, 15, 11, 7 and 3 weeks old (stages S^, TR^, TR^, TN^, TV^ respec- tively. The largest number of eggs was detected in planting 7, when this planting was in the TR - TR stages. At the same time, egg num- 1 ^ bers decreased for planting 6 after the 10th week of plant growth. The mean nimiber of eggs in planting 8 increased slightly from week 3 (T^/2^ ' through 7 (TR^ ) . These data indicate that there may be several factors. en 0) c (U B (U (U u c (fl +J (n •H T) • ■P -M 0) C iH (li 4-i k (0 ■P (rt 01 m m m m di m (U LD c , p pa O 3 T) C c •H m u< -p o 0) -p iH (Tl U-J s fl o CI) +J H M-l u 0 0) U< >1 1) rn c m m 01 0) S en tji •H 01 fa (1) 85 [: - lO CM o C4 I d 3 « CQ u e IB (fl 15 /f 3uanbai J 86 ■-t 01 O ^ rt -I o J ^ SJ Z 2 — 87 such as presence of inflorescence and water content associated with plant age, which account for frequency of TPW oviposition. The effect of plant water content in oviposition will be discussed in chapter 3. General Discussion and Conclusions Tomato pinworm egg distribution is one of the least studied aspects of this pest. In this research useful data were gathered about the spatial pattern of TPW eggs in the tomato plant, effect of plant age on oviposition, and use of these data for TPW egg sampling allocation. The tomato pinworm tends to select certain leaf sides, plant strata and plant age for oviposition. TPW prefers ovipositing on the leaf under- side. Ninety percent of the total eggs were laid in the leaf underside. Tomato pinworm oviposited mainly in the upper and middle plant canopy. Fifty-one percent of the eggs were laid in the upper canopy. There were exceptions to this rule. For instance, a crop planted during Oct. 30, 1980, in which sampling was done during January and February, had a higher than usual percent (17%) , of eggs laid in the internal canopy. The information described before was used to develop a sampling plan. First, 10-20 plants/acre can be used as sample size if a practical equi- librium between sampling cost per acre (33) dollars, and index of precision (54-31%) , is desired. It was found that proportional sampling can be allocated for the upper external stratum (n =6) , followed by the middle internal stratum (n, =5) , and 3, 1,4 and 1 samples for lower external, upper, middle and lower internal canopy, respectively. Experiments also showed that under high density (22 eggs/plant) the insect prefers the middle leaves of the plant. These results were expected because TPW egg 88 population is considered clumped in the plant (k=0.01-0.45) . The strata that has more eggs had higher variance than the strata with less ovi- position. Therefore, more samples were allocated (n =4) to the middle h stratum. Also, the shortest range for a confidence interval of the mean was in the upper external stratum (0.51-1.7). At this time it is not known which is the economic threshold based on egg counts. The larval economic threshold is considered to be 1 larvae per plant (Chapter V) , therefore, if no egg mortality is expected 1 egg per plant will be the economic threshold. The strata in which confidence intervals are less fluctuating through time are the upper and middle external canopies. Then, these strata are the ones to be selected for egg sampling in the field. As expected, oviposition increased as plant age increased. More eggs were found on plants 4 weeks old (TV ) than on plants 2 weeks old (TV ) . Egg numbers increased for plants 8-16 weeks old (TR^ - TR ) , then decreased for plants 17-21 weeks old (TR -S) . More research is nececessary to evaluate attraction for female oviposition based on physical and chemical qualities of the plant. 2 Leaflet area and egg density were unrelated (r =0.026) . Distance 2 between eggs was not influenced by egg density (r =0.19-0.23). The average distance between eggs per leaflet ranged from 0.5 to 0.75 cm. These results indicated TPW tendency to oviposit eggs at a common distance despite egg density and leaflet area. CHAPTER IV SPATIAL PATTERNS OF DISPERSION OF TOMATO PINWORM LARVAE IN TOMATOES Introduction Dispersion patterns of Keiferia lycopersicella (Wals.) , tomato pin- worm (TPW) , larvae in tomatoes have not been studied in great detail. Knowledge of these patterns is necessary to develop a better sampling procedure. Several techniques for population estimates such as ab- solute, relative estimates and population indices are used to determine insect distribution (Southwood 1978) . The TPW larval population intensity method has been used by Wellik et al. (1979) to determine sampling accuracy. However, TPW population indices have only been used to measure economic damage (Wolfenbarger et al. 1975) and plant resistance (Schuster 1977a) . Because leaf mining lepidopterous larvae do not move from, a given plant to neighboring ones (Dethier 1959, Nishijima 1960, Schoonhoven 1972) population indices can also be used to detect distribution pat- terns (Gomez and Bernardo 1974, Henson and Stark 1959, Condrashoff 1964) . I used TPW damage index to (1) estimate dispersion patterns of TPW larvae during different plant stages, (2) determine an appropriate sample size and sample unit for TPW larval injuries, and (3) discuss sampling strategy. 89 90 Materials and Methods Sampling Methods To determine the relationship between damage by TPW on different strata of the tomato plant, 8 plantings of nonstaked tomatoes cv. 'Flora-Dade' (Nov. 3, 1979; Dec. 5, 1979; Jan. 8, 1980; Oct. 30, 1980; Nov. 25, 1980; Dec. 30, 1980; Jan. 30, 1981; and Feb. 28, 1981) were evaluated at the University of Florida Agricultural Research Center, Homestead, Florida. Each planting (ca. 450-947 plants) was direct- seeded in raised beds (3-5) (ca. 45 m long) of Rockdale soil, and mulched with light colored plastic. Seedbeds midlines were 182 cm apart. Plants were spaced 38 cm apart. Sample size was selected by a preliminary random sampling of 50 plants on 2 dates. The method described by Elliott (1979) was adopted in this selection. The sample size was chosen at the point when TPW mine density variance stabilized. The percentage error that can be tolerated in the estimation of the population mean was expressed as the standard error of the mean. Finally, the cost of sampling within the plant (C ) (Southwood 1978) and the cost of moving from 1 plant to another (C ) were considered to estimate the relative net precision P 100 — value (RNP = -— ) , RV being the ratio SE/x and C = C + C . RV X C ^ u p s u ^ Selection of the sampling unit for each plant was made by randomly collecting 1 leaf/plant, 2 leaves/plant and inspection of the whole plant (upper and lower canopy) . The survey was made on 11 dates on 91 2 crops planted during Oct., 1979, and Dec, 1979. The mean per sample and standard error of the mean were estimated. The proportion of the true population collected per sample was determined by dividing the niimber of individuals per sample unit by the number on the whole plant and by comparing the RNP per sample unit. After this, 20 plants in each planting were sampled by use of the simple random sample technique (Cochran 1977) . Sampling was done on a weekly basis from Feb. 7, 1980, through May 29, 1981. The whole leaves of the plant were first examined to determine differences in larval injuries on lower and upper leaf surfaces (plantings 1-3) and then to detect differences in larval injuries in different plant strata (plantings 1-8) . The plant was divided into 2 sections (upper and lower) for the first 3 plantings (1979-1980) and divided serially into 6 sections (upper, middle, and lower part of the external and internal canopy) for the last 5 plantings. After values of bilateral asymmetry of frequency distribution (kurtosis) have been obtained, data can be normalized (Sokal and Rohlf 1969) . Data were transformed by replacing the value (x) by logarithm (x+1) (Elliott 1979) . The counts were assessed by a t-test for results from plantings 1-3, and by analysis of variance (plantings 4-8). Means were separated by Duncan's Multiple Range Test (P=0.05) . Results and Discussion Sample Size Two goals were determined for sample size in tomatoes. One of them was an index of precision (I =25%) and the other was the cost of tomato 92 sampling in southern Florida ($7/acre) . To accomplish this, sample size was selected based on 3 major criteria. First, following the criteria outlined by Elliott (1979) , a suitable sample size can be selected when the mean ceases to fluctuate (Table 19) . It is observed that when average larval injuries run as low as 2.12 or as high as 11.05, injuries, variances and means tended to stabilize at ca. 20 plants. Second, the use of index of precision (I =SE/x) demonstrated that P for low populations (0.2-2.12), SE/x ratio was between 91-24 when 20-50 plants were sampled (Table 19) . It fluctuated between 20-29% for the 20-15 plants at populations above 10.93 injuries per plant. These results demonstrate that for lower insect populations it is necessary to increase the number of plants up to 50 and to reduce it to 15 when the population is as high as 11 injuries per plant. Third, the sample number does reconcile with the actual budget. Cost of sampling larval injuries is 1.26 dollars less than the actual budget per acre if the sample size selected is 20. The use of relative net pre- cision (RNP = — — ) can also be used for sample size selection. The u larger the RNP of a sampling method, the greater the precision for the same cost. RNP values when population is low can be selected for an acceptable I_. Therefore, RNP values can be accepted for 30-50 plants under such conditions. RNP values when the population is high can be selected for 25-50 plants. Nevertheless, to reconcile precision with cost, sampling 20 plants gave an acceptable RNP for high or low TPW populations. 93 e Sm 0 s • lO i:; O a •H 00 u a cTi iH 0 4J » (0 tS e Ti 0 -H nj 4-1 ^4 en 0 U a > £ -p ^ to 3 0 u ( ) n m fl) Tl ^—^ irt til ri K >k. ^-^ T3 fl C 0) 0 +J •H 01 tn CI) •H e () () c ■iH (1) JJ -n m iH a o n M •H +J T! m C H 01 3 a <1) r. N ■H Ul £ rri (U ■H i-l j:: p. R Tl fil c UJ (tj X n ■r^ > tn u c (Tl (1) h4 U 94 95 Sampling Unit Data from different numbers of leaves per plant in the upper and lower portions of the plant are shown in Tables 20-21. In Fig. 8 it is shown that it is necessary to select 2 leaves in the lower canopy if the number of injuries is as low as 0.03-0.55. This is 20-30% of the total number of injuries per plant. When the insect population increased, 2 leaves from the upper and lower parts were also necessary to obtain 32- 34% of the total population. Selection of 1 leaf per plant from the upper portion gave as low percentages of the total damage as 1-2%, but increased during later sampling dates up to 9%. The number of injuries in 1 leaf per plant selected from the lower mid portion of the plant remained stable at about 10% with the highest being 16%. Another aspect to consider for sample unit evaluation is the confi- dence interval of sample mean (CI=x+t S~) . According to the results expressed in Tables 20-21, at low larval injury densities it is more appropriate to sample the whole plant. The confidence interval remained stable (0.24, -.08) for the whole plant until larval density was higher than 1. When density increased above 1, 2 leaves per plant gave a more stable confidence interval through time. It is necessary to evaluate sample unit based on cost of sampling. Cost per sampling unit is shown in Table 22. Relative net precision was considered lowest when the whole plant was inspected. Two leaves from the lower canopy gave an acceptable RNP during 5 of the sampling 96 n . u (Ti m r- TJ (T> c <-t m jj ^ m • > Ti 0 a a o c o, o ■H O o u 3 • (T3 4-1 13 O OJ 97 - I ■u m c T3 (1) -H M u (1) n MH i-i m fa -H T! c in 0 a) +j M rn 0 (I) u H u 0 (0 00 ■rH -T! U JP was obtained from sampling 1 leaf. A sound sampling program requires precision and depends on resources. A balance must be struck between the two to keep variance minimal for fixed costs. Emphasis must always be given to practical considerations (Ramsany 1980) . In general, 2 leaves from the lower canopy should be used as a sampling unit when a stable RNP is desired. Statistical Distribution of TPW Larval Injuries Different procedures were used to detect data normality. The sta- tistics summarized in Table 23 suggest a larval aggregation of TPW in the tomato plant foliage. In general, as the mean increases, variance also increases. The variance to mean ratio, or index of dispersion (I) will approximate unity if there is agreement with a Poisson series. The (I) values obtained were far from unity. Values ranged from 2.45 to 6.05. Values of skewness and kurtosis were all positive. This means that fre- quency distribution of larval counts tails off among the higher counts. Values of the k from negative binomial distribution are considered more clumped when k approaches zero. The lowest value found in this data was k=0.21. The Poisson distribution, however, fitted (P=0.005) 34% of the sampling dates (see Appendix). In general, TPW injuries were considered clumped in the tomato plant. Distribution of Injuries on Upper and Lower Portions of the Plant Statistically significant differences (P=0.05) were detected for the injuries located in the lower part of the plant for the 5 sampling dates in the first planting (Table 24) . Statistical differences were 102 m iH •H a tn 0 M ■p 0) u a 3 1^ M (U •H M 03 3 M •n 0) C C •H s 0) rH ^ Id w a n] 103 •H fa S^ 3 •k. ■n >1 C 4J •H c p iH o (0 n > u !1) rfl Tl iH (0 a ;^ m «> •H T! iH (11 CJ <1) IW 4-1 CO ..•^ (1) ■^ f; a^ o o C 15 •H C +J -H C !^ rrt .-( 0 a ■P n3 +j g G C5 (1) +J IM C 'T m <3 s 01 X OJ !I! ^ CO ~D " T r- CO "^ o J^ X o a 1 ■2 5 ^ t% c r- '^. = -; ro ° "^ ~ o ^ 1 ro r; o T ^ o o o o ^ ^ N c-j a" c c o fl ■^ - Tj m o CO o o = T O c c 3 1j . o c c ^" c H rs; 2 k (!.' t: l- o "1 T c '"'. '- c 0 ^ o c c c c ^ " % ~ i5 ^ T •" ^ 0 r-l =. ■^ o >■ TJ 3 3 C q 0 0 104 only detected once for the second planting. In general, the lower part of the plant had 60% more injuries than the upper half during the first 4 sampling weeks for the first planting (Nov., 1979). Also, the upper half had zero injuries during the first 4 sampling weeks. When plantings 1, 2, and 3 were 16 (TR ) , 12 (TR ) , and 8 (TR ) weeks old, respectively, there were similar numbers of injuries in the lower (55%) and upper (45%) parts of the plant in planting 1. In planting 2, 75% of the injiiries occurred in the lower half and planting 3 had 67% in the lower half. Nevertheless, the number of injuries in each stratum was not significantly different. There were more injuries in the upper part of the plant when the plantings were 21 (S ) , 17 (TR ) , and 13 (TR ) weeks old. This may indicate more active lairval consxjmption in the lower half of younger plants and fewer injuries in that level in older plants. These results disagree with the Florida results of Wolfenbarger et al. (1975) , but are in agreement with those of Wellik et al. (1979) in Texas. Distribution of TPW Foliar Injuries in 6 Plant Strata Statistically significant differences were detected for 3 of the 16 sampling dates for planting 4 (Nov. , 1980) (Table 25) . Greater numbers of larval injuries were found in the middle internal, lower external, and lower internal canopies. When data were converted into percentages and analyzed, there were significant differences between the lower portion of the plant and other strata (Fig. 9-10) . The number of foliar injuries for planting 5 (Oct. , 1980) showed statistically significant differences in 1 sampling date. Greater damage was recorded from the 1 ! 105 ^ 00 C» r-H Eh o 0 ■p u c C) m Tl C;) m X Q CD 5 s :5 o CO c5 ^ 5 5 ^ I 5 ro ,5 '-'•■ o = " = = = = c c = = = - c 1r CO 1! C (^ 1 j; ^ ^ 5 rr •5 10 s ^. c o c o c = o c - o = = = = T ^ 5 ra ^ ■»o n ;i; 5 J3 ^ rO 5 o •^ c c c o c c o c O = O O a c - ~i o r: .3 ^ iS s 3 ^ r^ r^ o "io o ■^ o o = o o o = c = = o 3 = ^ CO ^ 5 ^ ro T s s IT! 5 o 5 o o ^■ c c o o o o - c o o o o o o — -: J ^ 3 C - C O 3 O O -4 o c 106 c -i i ^ o IT! ^ O O O c ^ r^ :/l C r-1 O -^ tN O -I r-l O " rN CO .-< -H C — O ^ -^ C I 2 n CO c cn f (0 rH 1 tn +J Ul c •. •• C c 0 p O OJ (0 •H p n u M 4-1 •H (U +J ra T3 • 4-( e 1 i4 g l 0 c (U en l-l •H (fl a 4-1 > iH ft •* c 0) (71 c P J-l ,_« 0) c 3 i 4 H 0 c 0 CJ h H (0 (t) 4H 108 a -o 3 Q, ai 0 3 S -J D 'liiriii' jjjjjlllil ?^>?'SSJJS?^S$iS^-:-:-: idii liilii ■■■^littfl ilpt . ■■iiiiiiiil ' " ' ailiiiiiiiij ii^_ 1.- uinjBJjs/ saiintui iBijoj /v\dl X 109 a a D rail ..i^_ tijljiiljjjB'jIlj'jB liHIii {■■■iri'ii'iiiii[Mi _ ■■K'iii'iliH ¥.- -jililijjlljjll' ^. <"■■■■ di iiiiiliii rMi _ 00 mm^iis/saiinlui ieijo; /v\dl >1 > .. n -a () z Id c (I) m ^ +j u • tn +j Q) r-i o g rt) r) n c K ^ c <.[) •H ■p c r} rH •H 0) 00 4J tn t.1 c rH c (Tl (0 a ' rH ^ (« 1.0 •H c u 4-) m irt c rH (0 r-l r~- 14^ tL, « n IT> o C C.) +J •H rti (rt 4-) rtl H C ■p n Ifl G +j r-t fj Pt u u-| i-i • *■ (1) e o fc O CO O T! ^ rH 0 . - 4J O (0 OJ <0 OJ "O Sh 3 Q -H :3 r-i u tri (3 'd o ■H > C rH Ill ' I fa B '^ c ilUll 1 ■I 00 E 9 Z 09 C e IS o o o 00 o o o C4 S3{jnfui Jciioj 112 lower external canopy. During the 14-16 weeks of plant development, the lower, middle, and upper canopy contained 91, 88, and 88% of larval injuries, respectively. Larvae tended to occupy the upper part of the plant in older plantings (18-20 weeks) . Statistical differences for the crop planted on Dec. , 1980 were observed in 4 of the sampling dates. More extensive larval injury was found in the lower external and middle external canopy. When plants were 10-15 weeks old, higher proportions of injuries (99-100%) again were found in the middle and lower canopy. Statistical differences for the crop planted in Jan. , 1981 were observed in 3 of the 7 sampling dates. Higher percentages of numbers of injuries were found (98-100%) when the crop was 9-15 days old. Larger numbers of injuries were recorded for the upper external part of the plant for planting 8 (Feb., 1981) during 1 sampling date. Sampling 5 strata demonstrated that TPW larval injuries are signifi- cantly higher in the middle and lower canopy. Sample allocation (n^^) as 2 outlined by- Cochran (1977) is necessary to minimize variance (S ) . Sample cost was considered equal for each stratum. Sample allocation was estimated on dates in which statistical differences were detected. In general, more samples should be allocated to the middle and lower strata (Tabic 26) . The average numbers (n=20) for all plantings were 2, 3, and 5 samples for upper, middle and external canopy, and 0.09, 5, and 3 from upper, middle, and internal canopy. Allocation ranged from 0-20 samples for the lower external canopy, and ranged from 0-9 samples for the lower internal canopy. I considered this sample allocation to be the best, because standard error (BE) of the sample mean was mere constant through time (range :0. 20-0.66) . 113 Table 26. TPW larval injury sample allocation for 6 plant strata at 3 different plant stages: second reproductive (TR ) , third re- productive (TR ) and senescent (S ) . Stratum External Internal \ h 2 TR„ 2 TR„ 2 TR^ 2 TR^ TR3 TR TR^ Avg. Stage Upper Middle Lower Upper Middle Lower h 4* 3 6 0 4 0 1 3 1 0 5 4 0 5 9 0 6 0 0 0 20 0 0 0 2 2 5 0 . 5 2 0 0 0 0 12 4 4 5 4 0 5 2 2 1 2 0 2 9 2 4 5 1 4 2 1 2 5 0 5 4 5 5 2 0 2 6 2 3 5 0.09 5 3 n = 'J S h ( 'h h )n; n = 20, N = 947. n n 114 There were exceptions to these sample allocations. Allocation increased for the upper canopy when plants were in the second repro- ductive stage (TR ) and during the senescent stage (S ) . It is necessary to correlate this sample allocation with fruit damaged in different parts of the plant canopy. Knowledge of this relationship will help in better prediction of yield losses. Larval Injuries at Different Plant Growth Stages The relationship between TPW foliar larval injuries and stage of plant development was determined during the study of plantings 4-8 (Table 27). For the earlier planting (Oct., 1980), the mean number of TPW injuries appeared during the second reproductive stage (TR ) and peaked during the senescent (S ) stage. On later plantings (Nov. -Dec, 1980) TPW larval infestation was not observed xintil the second repro- ductive stage (TR ) , when plants were 14 and 11 weeks old, respectively. Higher foliar infestation was obtained when plants were in the late reproductive stage (TR ) and also in the senescent stage. For the winter season plantings (Jan. -Feb., 1981) injuries were first observed during the second reproductive stage. The reason for differences in larval infestation in different crops can be related to several factors. First, there was a low oviposition rate in the earlier planting (Oct.- Dec). Secondly, the earlier fall crop (Oct., 1980) had a dense leaf canopy and approached the senescent stage later than the other plantings (Table 2). Thus, infestations in this planting were probably influenced by a complex of factors, including increased food consumption by the pest during the senescent stage due to reduction of food quality. 115 g ^ o » c ■r-l u, o 4-1 (rt b (1) o fn 4J ifl +J M-l tn 0 ^ U jj 0 ■t !-J o ^1 >^ 0) en Tl 4-> i^ c m fl T) -H ^ Tl s 0) p •H c IM -H c O (t! (!) nj tin 2 m 115 g I I -^ M C3 01 -jU 117 General Discussion and Conclusions In this research useful data were gathered about the use of TPW damage index to estimate TPW larval patterns during different plant stages and about selecting an appropriate sample size and xmit. The population index used (TPW larval injuries/plant) indicated that during plant stage TV , TPW damage range (0.0-0.05) is lower than that found in reproductive stages (TR =0-0.55, TR =0-5.05, TR =0.8-10-15) and lower than the estimates from the senescent stage (S =0.6-6.5). Further information is needed to detect if nutritional changes in the plant are the major factor for higher or lower number of injuries per plant. The information is also necessary for detailed plant resistance studies. The information supplied by sample size can be used to develop a sampling plan. For instance, 20 plants/acre can be used as sample size when a practical equilibrium between index of precision (I ) and cost (RNP) is desired. The sampling unit to be selected is 2 leaves/plant from the lower canopy when an acceptable (RNP=0.25 or less) is found. This information can be used for commercial sampling for detection of foliar inj\iries. One of the problems involving this selection is that TPW foliar injury does not necessarily mean fruit injury. Therefore, establishment of a relationship between fruit and leaf damage is needed. Sample allocation resulted in a major proportion of samples allocated to the middle and lower canopy. When total sample size equals 20, n^=8 118 samples should be taken from the middle as well as from the lower canopy. An n^=2 should be allocated for the upper canopy. These results were expected because TPW injuries were considered clumped in the tomato plant. The values obtained from the index of s2 dispersion (±=— ) ranged between 2.45-6.05. This departure from unity X means TPW larval aggregation in the plant. The data can be used to transform larval counts in order to obtain normalization. If sequential sampling is projected, it is necessary to find a common k value for TPW injuries. The population index should be incorporated into a more sophis- ticated damage and popiilation evaluation method. For instance, techniques such as degree of damage (Chapter 2) would be useful to- gether with the population index. The combination of these techniques will help the scout in detection of larval populations approaching economic thresholds. CHAPTER V TOMATO PINWORM ARTIFICIAL INFESTATION: EFFECT OF FOLIAR AND FRUIT INJURY ON GROUND TOMATOES Introduction Economic losses resulting from insect injury to the foliage and fruit are difficult to measure. One of the problems in determining an economic threshold of a pest species is to distinguish between its mere presence in a crop and the density that will cause an unacceptable loss quality or quantity (Stern 1973) . Damage to the tomato fruit by the tomato pinworm, Keiferia lycopersicella (Wals.), has been evaluated by Poe and Everett (1974), Wolfenbarger et al. (1975) and Wellik et al. (1979). The results were contrary in the three studies. Poe and Everett (1974) found no correlation between leaves mined and presence of larvae and fruit loss. Wolfenbarger et al. (1975) determined that TPW damage to the 3 top leaves could be associated with fruit injury. Wellik et al. (1978) indicated that damaged lower leaves and large fruit are the best for estimating K. lycopersicella infestations. Waddill (1975) and Schuster and Everett (1982) show estimates for losses from TPW natural infestations. In this study, tvro techniques for estimation of TPW fruit loss are assessed: (1) use of TPW larval artificial infestation to measure yield reduction and (2) use of the population index (TPW foliar injuries) and 119 120 its relation to yield losses. In addition, fruit damage related to time of planting is examined, and the effect of plant pruning on TPW fruit damage is evaluated. Materials and Methods First Experiment Reductions in 'Flora-Dade' tomato yield caused by different popu- lation levels of TPW were measured during 1980 and 1981 at the Agri- cultural Research Center, Homestead, Florida. Crops were direct-seeded on December 5, 1979; January 8, December 1980; December 30, 1980; and January 30, 1981. Plots were thinned to one plant every 0.30 m. Seed- bed's midlines were 182 cm apart. Each crop was sprayed weekly with fenvalerate 2.4EC at a rate of 0.064 kg ai/ha. Insecticide application was discontinued 25 days before artificial infestation with TPW second instar larvae. Larvae were reared on tomato plants in a greenhouse at o 23-H2 C and 75+4% RH. There were 2 replications per treatment m a randomized complete block design. Artificial infestation was as follows: First, plants were inspected to remove any TPW eggs and larvae. Then, 10 plants 40-45 days old were infested once with different numbers of larvae (1, 2, 4, 8, 12, 14 per plant). Plants were inspected 1 day after infestation and larvae replaced if lost. In a 2nd experiment, plants were infested twice with larval levels mentioned above. An uninfested control was sprayed with fenvalerate to keep it TPW free. A second control without insecticide or arti- ficial infestations was also used to compare with treatments. When foliar and fruit injuries occurred in the unsprayed control, these numbers were siobtracted from the numbers found in the infested plots . 121 At harvest, counts of the niombers of leaves and fruits injured were recorded together with the total number of fruits per plant from the upper and lower canopy of the plant. Fruits were graded according to USDA grading standards (Anonymous 1982) . In order to reflect market standards, values of the fruit were modified for statistical analysis by multiplying each fruit grade by a number. Numbers of extra large fruit were multiplied by 5, large by 4, medium by 3, small by 2, and very small by 1. Results were also analyzed before multiplication for fruit size. Data were subjected to analysis of variance; treatment means were separated by Duncan's Multiple Range Test at P=0.05. Regression analysis was used to establish a possible relationship between the number of larvae and the nximber of leaves and fruit damaged. Data from leaf injuries in the upper plant canopy ( Y ) , leaf injuries in the lower canopy (Y^^) , number of fruits injured in the upper (F ) , and number of fruits injured in the lower (F ) canopy were regressed separately on one independent variable: number of larvae per plant (x) . The number of fruits injured in the upper (F ) and lower (F, ) canopy were also regressed on Y and Y , resoectively. u 1 "- Finally, 2 regression models were used to find a statistical model that related percent yield loss to the number of larvae per plant and to number of foliar injuries. The main objective was to find a simple model for infestation levels which could be used to establish economic injury levels. The regression model used was the form y =a+bx and y= 2 '^ a+bx+cx . For the simple linear regression, y=estimated value of y, a=the y intercept of the regression line, b the regression coefficient and X the sample estimate. For the curvilinear regression, c=the second 122 regression coefficient preceding the second power of x. Results were compared with the previous work by Poe and Everett (1974) , Wolfenbarger et al. (1975) and Wellik et al. (1979). Second Experiment To determine the effect of planting date on damage to the tomato firuit by TPW larvae, natural infestations with this insect were studied in those plantings mentioned in Chapter 3. Crops were direct-seeded during October, November and December, 1980 and January and February 1981. Number of injuries per plant and number of tomato fruits damaged and not damaged were recorded for each planting. A total of 20 randomly selected plants was chosen and numbers of damaged tomatoes were compared. Secondly, comparisons in damage estimates obtained from the customary sampling method (6 contiguous plants per row) vs randomly selected plants per row were made. Results were compared by use of relative net precision (RNP=100/RVxCu) , for both systems. Thirdly, to determine the effect of pruning laterals on tomato fruit damaged by TPW, laterals from a total of 20 randomly chosen plants (45 days old) were removed and yield was com- pared to that from 20 unpruned plants. Treatments were replicated 4 times. Analysis of variance and Duncan's Multiple Range Test were used to compare treatments. Results and Discussion First Experiment Single artificial infestation. After a one-time infestation of plants with TPW larvae, the number of fruits damaged in the lower plant 123 rranopy was significantly different from the sprayed control (P=0,05). In the upper canopy average fruit damaged did not differ statistically among infestation levels (Table 28) . The number of fruit damaged in the lower canopy when plants were infested with 1 larva was 10.5, 24.5 and 29.5% less than that found at 8, 12, and 14 larvae, respectively (Table 28) . In general, damage from 4 larvae per plant and 14 larvae per plant did not differ statistically. Results may indicate that 4 larvae per plant represents the upper practical limit for fruit damage for 'Flora-Dade' tomatoes. The numbers of injuries per fruit may increase with increased larval infestation levels. Once populations reach ca. 4 TPW larvae per plant, multiple injuries to the same fruit may become very common. There is more fruit in the lower than in the upper part of the plant. Because of larval positive geotaxis, larval activity seemed concentrated on the lower plant parts. Consequently, fruit infestation in the lower canopy is higher. Marketable value of fruits on each canopy was significantly differ- ent for the infestation levels when different values were assigned to fruit (Table 29). Values of fruit damaged by 1, 2 and 4 larvae per plant differed from damage by 8, 12 and 14 larvae per plant. Eight, twelve and fourteen larvae caused 1.63, 1.76 and 2.2 times more damage than 1 larvae. Double artificial infestation. After a doiible infestation with TPW larvae, tlie nxmber of fruits damaged in the lower canopy at all levels of infestation v/as again significantly different from the uninfested con- trol. Tomato fruit damaged in the upper canopy did not differ statis- tically among infestation levels (Table 30) . There were no significant 124 0^ g 0 o C +J U 13 ■U P C 0 (fl »^ iH 0) H a< O a () 3 ■rH to (1) V< c^ Q) +J tt 0 c I) H >i >H e ^ (13 -U > * U( c () C) fi m ■p ( ) c (rt u H (1) ft •s o Vl ►-I (T) 0) g H-l 0 c (U H -H (C > 4-1 !m C • ifl .-a CO J M CN a< u 0) (U !-l rH ^ (U + 1 o + 1 04 (N n n^ + 1 + 1 + 1 + O o o o o + 1 + 1 +1 o + 1 + 1 +1 c ■P cu rri nj 03 r^ 0) C w S-l -H +J ■P 4-1 0 g 3 c O 0 e n 125 0 'T3 c c U 0 Q4 OJ 3 o c C ■r-l C5 •H Tl +J 0) (tl tn -P (tj (fl P fl !-l > IW u fl n rH j-i fl T-H e fl f) •H 4-1 U !^ M-l C> ■H U-l -P (!) fl 3 ,-( (I) fl i-< > 0) •rH H !/5 • ^ (1) fl fl (I) +J 0 (U U 4J M <\) fl U JJ Fr fl U-l n a fl 4-1 Q) fl u c 0 •• fl o [fl o 13 OJ fl M £ fl 4J 4-> rs C c c (U fl -H Vl 4-1 0) tn TJ U-l i en 0) rH 1 P 4-1 0) iH c I^ fl fl •H > O W -H 4J M-l • -H ■H < C C • D 0> Q ■H Ul W rH • If 4J D rH O r-( C C fl o e (U • M S-l ^ -a fl in (U >i o en M i-i • fl Q) 0) o Xi > 4J II e ro (1) tn H (1) fl H tn (!) a ryi x: C 4-1 fl (S >i M cu rH n ft (1) ■H :? 4J (5 rH r-l a iH 7'. vT 4-1 Ui en a 0) fl 3 o iH c fl 3 > n 126 n n T! 4-1 (0 m R o u ■p aj +j -fl M-l c n1 3 O >i ^^ ft 171 n c C (Tl C) U (1> -P fl c > (rt IH iH rtl ft rH U ifll i 0 •-i 0) ^1 +J 1+-! >i ft o +J c c fl d u eu i-l 0) u :? 1) 3 0) to > +J (1. i en C ^ c 0 0) td U Sh TJ CtJ ■H 3 4-1 C Q ■p e t) -H •rH O n) ^ O 3 V4 tn 4J 4J !H >4H c -■H 14H •H S tr -a c t3 0) +J « -H OJ c 3 ^ T) en -rf O a; w (0 (« £ ■9 0 S -P 4J § 1 i= ^ n ja iU 3 0 ^ (U (1) 4J rH U-( 43 01 01 4-1 >i 0) ill ^ 0 i4 c CC CO Fi >| 01 U4 a n c M-l 01 0 U CI) u 3 (I) i-H a, 01 ft > p a; 01 > 4-1 ^4 c 01 01 hJ .-1 II •H m 4-1 C/) c r-l >i •H 3 rH T) ^4 .11 U M-l c O 03 O OJ !) t7^ Cfl (0 S4 ■r4 01 £ +J r-i 4J C 0) 01 c; Sq 4-) •H 0) M-l >< Tl 1+4 CI) 0) •H O a) c M X en (1) u ■w 01 03 g Fi (5 4-) C) ^ 0 • iH C '- h II LD C) r-l 0) O u rH H • CW cO .•0 o g II -n W ^4 &I 03 M 4-1 4-1 c oj Pi 4-1 g CP -H 03 -D g (1) S (0 0) 0) d Figure 11. Relationship between number of tomato pinworm larvae per plant and num±>er of injured fruits and leaves in the lower plant canopy by a single artificial infestation with TPW harvae. Homestead, Florida, 1980-1981. 130 a o c o u *■ 7 f O ^ 5.0_ a c ... 2.5 3 Y =1.97+ 0.20 X r2. 0.3 7 F=6.02 0.0J_ T r a. o lOj c o u u « 7.5 0 0, 5.0J 3 Y= 2.30+ 0.4 IX rl 0.511 RlO.46 :2.5-. 0.0 J. 1 — \ 1 1 1 — 12 4 8 12 Number T PW Larvae 14 Figure. 12. Relationship between niimber of tomato pinworm larvae per plant and number of injured fruits and leaves in the upper plant canopy by a single infestation of TPW larvae. Homestead, Florida, 1980-81. 132 10 a o c (0 u a a 7.5 Fu.-0.85^0 014x+0.002x^ F--2.3 r 2. 0.33 Pr7F 0.15 'J) .r 2.5 3 u. 7.5- T3 5 . YU"3.34+1.03x-0.046x2 r^.38 Fi2.83 Pr^F 0.11 2.5 0 1 Number TPW larvae 12 -1 — 14 133 (x) was best expressed by the following significant (P<0.03) linear 2 2 regression. The r value (r =0.37) was less than intermediate. The correlation coefficients for the lower canopy were intermediate, indi- cating that 51-37% of the variation in foliar and firuit injuries was due to larval numbers infesting plants. When data from the upper canopy were regressed against the number of larvae per plant, curvilinear regression had a better fit than linear regression. 'There was no significant relationship found between injuries in the upper canopy (Y ) and number of larvae (x) infesting the u plant (Fig. 12) . Again, no significant relationship was found between injured fruits in the upper canopy (F ) and the number of larvae per plant. The lack of significant relationship between larval niimbers and TPW injuries in the upper canopy indicate that upper canopy counts can not explain number of larvae present in the plant at TR stage. The relationship between fruits injured in the lower canopy (F^) and foliar injuries in the lower canopy (Y ) was best expressed by a signifi- 1 2 2 cant (P=0.024) quadratic equation F =1.09 + 0.46Y -0.02Y and r =0.56. (Fig. 13) . F^ began to decrease at a level of about 7.5 injuries per plant. This may be an artifact, or it may be that beyond this level of larval infestation of leaves, multiple injuries to fruit may be more common than infestation of undamaged ones. Regression analysis of numbers of fruit injured in the upper canopy and foliar injuries in the same plant part did not indicate a significant relationship (Fig. 14) . Also, regression of fruit injured on lower canopy and the total plant injuries was not significant. Therefore, TPW sampling by scouts is probably best done in the lower plant canopy. u ^ m c 1-1 -a ;? a il >i^ ft -M 0 -H a ? m o 0 s^ •H 0) 4J S (0 0 4J ■H T) m c c (d •H )^ H 0) (0 ft -H ft U 3 ■H C •H •H 4J T1 ITI di U 0) 3 H •r-i rp C c •H •H cn tn (1) (rt > Id >i 0) XI H >. t-l ft O B M (C (U C 0 fi iH 0) _o CJ ha S c 7.5 FU2.51l4.0.586x-003x F* 1.32 r^-0.22 Pr>F 0.31 i 2-5 1 r 7.5 0 c U i 5 0) u 'c 2.5 YI-2.06+0.91X-0.02X F=18.69 r^-O.SOS Pi>F-- 0.0006 0 1 ~^ — 12 14 Number TPW larvae 138 Relationship Between Leaf, Fruit Injury and Larval Infestation Levels Double infestation. The relationship between leaves injured in the lower canopy (Y ) and number of TPW larvae (x) was best fitted to the highly significant (P=0.0006) curvilinear regression, Y =2.06+0.91x - 0.02x^, F=18.69, r^=0.805 (Fig. 14) . The relationship between fruit injured in the lower canopy (F ) and the niamber of lairvae (x) was not sig- nificant {F=1.32, P=0.31) . When data from upper canopy were regressed against number of larvae (two-time infestation) per plant, there was no significant regression found between injuries in the upper canopy (Y ) and number of larvae (Fig. 141. No significant relationship was found between injured fruits in the UDper canopy (F ) and number of larvae per u plant. The relationship between fruits injured in the lower canopy and foliar injuries in the lower canopy was not significant, F=1.21, P=0.341, 2 r =0.212 (Fig. 15). The regression analysis between fruits injured m the upper canopy (F ) and leaves injured in the upper canopy (Y ) was u ^ not significant, F=0.98, P=0.33, r^=0.04 (Fig. 15). The reasons why there was no relationship between fruits injured and foliar injuries caused by a double TPW infestation may be due to an artifact, but it may be that a double TPW larval infestation causes more injuries to the same fruit, which will result in no increment of the total fruit damaged per plant. Yield Loss vs Density of TPW Larvae in Tomatoes The second order model y=a+bx+cx , (y=percsnt of yield loss, x= number of TPW larvae per plant) , was fitted to the data on mean percent U (U (U (0 •9 > B ^ 3 Id C H c "^ lO fl ^■^ 0*-*J 0 -H c ? (fl 0 e 0 M -H (1) -p » 10 0 -P i-H 0) 0) T) M-l c c (d -rl M rM 0) n] a-H 04 U 0 -H IW C -H ■M -P ^ TJ (0 03 U (U D H •^-y ■H 0 -O 01 (U Id > (d >i (U A H >1 M-f D( o o c ^ (d (U u "i ^ 3 0) c 3 0 C rH 0) 0) T) S c +J CO (U rH ^ !m CO i CM FH ? VO ^. • I ^ 1- '• 1 3 / "■ , • , 1 in O in ^i /idouBD jaddn pajnfui sjinjj ON o 6 X CO cs o d a a 3 V _3 'c /idouBD j3mo| pajnfui sjinjj 141 >. a o c G u k 0) a a 3 "O 4) a e 7.5. 5.0 _ Y = U4-0.031x-0.02x2 . r2.0.24FrI.21 n.s. 2.5_ 0.0 _L Y=2.47 039X r2.0.29R4.15 n.s. Number TPW Larvae 142 yield losses for 3 tomato plantings attacked by TPW. The relationship 2 was highly significant, P=0.001 and had an intermediate r value of 0.64 (Fig, 16) . The coefficient estimates b and c were significant. A positive increment in yield losses was observed until 12 larvae were infesting the plant; in contrast, beyond this infestation level, the percentage of yield losses decreased. Yield Losses vs Number of Plant Injuries A significant curvilinear (y=8,76+4. 21x - 0.14x^) regression was found when mean percent yield losses combined for 3 plantings were plotted vs TPW injuries per plant (P=0.001) (Fig. 17) for numbers of foliar injuries up to 27 per plant. Regression analysis indicated a positive relationship between yield losses and 10-15 injuries per plant. The relationship turned negative when more than 15 injuries are found 2 per plant. Based on r value, the fitness of the model was intermediate 2 (r =0.608), but higher than that found with a simple linear regression 2 model (r =0.322) . Actual yield losses may vary, depending on time of planting and southern Florida environmental conditions. A more robust model relating density and yield losses should be considered with lower larval infes- tation levels per plant. The models developed here provide valuable information on the TPW-plant interactions. TPW larvae may bore into the fruit for different reasons, yet high attraction to the fruit was not found by Swank (1937) . Thus, larval density, contact between injured leaves and fruits, or positive geotaxis observed when the larva suspends itself by the thread produced from the spineret, may account for the fruit damage especially on the lower canopy. Figure 16. Regression of percent of yield reduction against infestation densities per plant of tomato pinworm lairvae. 144 Y: 7.594 5.71X 0,26X' F = 18.14 P>F0.0001 r2= 0.64 Q. a« 0 12 4 8 12 14 16 TPW Larval Density c 0) o u (1) a 0 c o H ro • Ul +J 0) c M m m iH OJ Ui 3 3 CJi -r-i •H C 146 e^ «» X ^ ^ ■• u. oo 1 A 0. X fSJ ^^ ^ CNI «* 9> H- II »-< u. lueid/uojionpay p|3!A% 147 Estimates of economic injury levels for TPW larvae may be made by inspecting the equations describing the relationship between fruit damage and larval density. The high market price of the tomato crop and cost of controlling TPW larvae indicate again that infestation levels lower than 1 larva per plant should be used. The minimijm economic injury level damage detejrmined for green cloverworm in soybeans by Stone and Pedigo (1972) and modified by Hopkins et al. (1982) and Hall and Teetes (1982) is defined as EIL= v/(C/P)/b, where c=total cost of controlling the insect pest/ha, p=price or market value of the crop per ha, and b=the regression value from the regression equation used. Gain threshold (Stone and Pedigo 1982) = (control cost/market value) x 100. For example, if the cost of con- trolling TPW is $50 per ha, and the market crop value of tomato is $6629.5/113, the percent gain threshold would be 0.0075x100=0.754. The economic injury level for an infestation of TPW could be calculated by using the regression coefficient (b=1.67) from the linear regression equation y=15.08 + 1.67 x; r^=0.45; F=17.72 P>F=0.004: EIL= ^/ 0. 75/1 .67 = 0.67 tomato pinworm larvae per plant. Similarly, the economic injury level based on the population index (number of injuries per plant) can be calculated using the regression coefficient (b=1.07) from the linear regres- sion equation y=16.82 + 1.07x; r^=0.32; EIL= J 0.75/0.07 = 0.83 tomato pinworm larval injuries per plant. There are constraints for these economic injury levels. They are based on results from one phenological stage (TR.) • Therefore, it is not known if they can be used for earlier stages (TV - TR ) . Planting time will also affect these results as demonstrated in the next experiment. 148 Second Experiment Effect of planting time on yield. Earliest plantings (October, 1980) had the lowest TPW fruit damage (Table 32) . Significant differences in number of fruit damaged per plant were found between those crops planted during October-December and crops planted during January-February. Largest infestations occurred in the latest planting (Februairy, 1981) . The amount of TPW damage in southern Florida depends on planting date with the greatest threat during the winter-spring crops. Planting time might be considered as part of an integrated control program for TPW. These re- sults call for further research on economic injury levels related to planting time. For instance, TPW population peaks occur during March- May (Chapter 7). At this time, those crops planted during the fall have already been harvested or are in the second reproductive stage (TR ) . The use of an earlier planting date will be a common sense approach to pest avoidance. Effect of plant pruning and number of plants per row related to fruit damage . Pruning plants had no effect on fruit damage (Table 33) . Trends in the data suggest that further studies should be done with more levels of pruning in the experimental design. Differences in relative net precision using 6 contiguous plants/row opposed to random plants per row did not show any statistical difference (Table 34) . Number of total fruits per row in those plants selected at random was slightly higher but did not lead to better RNP values. Therefore either method can be used for selecting tomato plant for TPW assessment. 149 >i 00 X! CTi rH T! (U 171 !-l C 3 •H •r-i 1-1 C 3 •H ts +J _ •H !1) 3 T) U (TJ U-l Q C ITl 0 u B fo tn () c H •* ■P m c dJ m 0 H -u a rtl s 4-1 o o 4-1 •p Tl u C 150 T) C V (Tl c m 4J r-( U) u^ O C) g 0 c -o 0) TJ S-( C M-1 g H O 5 s o + 1 U-l Cn c m ■rH c T, •H U ^ 0 rH O (11 U s IT3 TD ■P R >. Cli M Xi ^.^ C vl) ■H ^ T) JJ 4-1 (1) m Tl O :5 a) (IJ c f) 4-1 Ul in rH 1 1/1 fH -P (1) o O u jj Uh cn '^ - X •M U) (1) ^ !h c d (1) OJ G rH ^ TS c Ol P -iH 3 4J H rH w 151 Wl 3 c ^-1 (0 M-l iH c C \ (0 tn OJ +j e ■rH 3 c • S-l •H U3 fa 4-1 tn c iH OJ (ti 113 U rH -P c a 0 0) &H J-1 o (U -P M-i td S -P 152 Conclusions and General Discussion The infoirmation generated in this study is useful to demonstrate the complex effect of different TPW densities on tomato plants. First, a single infestation of 1, 2, 4, 8, 12 and 14 larvae per plant resulted in a 10.50 - 40% fruit loss. The result contrasted with 22.5 - 49% fruit loss obtained when larval levels were doubled. This demonstrated that only one generation of TPW is necessary to cause severe (40%) losses, and also, data suggest that in spite of low larval levels (1 larva/plant) , yield losses are twice as much if two generations occur. Second, regression analysis between number of firuit injured and niomber of TPW larvae demonstrated the importance of plant stratum selection for sampling TPW. Generally, the niimber of TPW larvae (x) 2 was better related (r =0.51) to the number of leaves injured in the lower canopy than to the number of leaves injured in the upper canopy 2 (r =0.38) . Also, number of leaves injured in the lower canopy (x) 2 and fruits injured in the lower canopy were better related (r =0.56) than leaves and fruits in the upper canopy. These results were expected because higher infestations occurred on middle and lower canopies where most of the tomato fruit is located. The relationship between number of TPW larvae per plant (x) 2 and percent yield loss was intermediate (r =0.64). Data from a 2 simple linear regression (r =0.45) was used to determine EIL. EIL for TPW was 0.67 larvae per plant. The relationship between ntmiber of injuries per plant (x) and percent yield losses was fitted to a signi- 2 fxcant curvilinear regression (y=8.76 + 4.21x - 0.14x ). The fitness 2 of the model was intermediate (r =0.608). Data from a simple linear 153 2 regression (r =0.322) was used to determine EIL based on the damage index. The EIL was 0.83 TPW larval injuries per plant. There are, however, some constraints for these economic injury- levels. The EIL is based on results from one phenological stage (TR2) ; therefore, it is not known if they apply to other stages (TV TR ) . Planting time will affect these results. Results concerning effect of planting time suggest once again that the southern Florida farmers should avoid planting during the later winter season because of higher TFW population peaks during March-May. CHAPTER VI ADULT DISPERSION AND COLONIZATION OF TOMATO FIELDS BY THE TOMATO PINWORM Introducuion Dispersion and field colonization should be among the first factors considered with regard to insect distributional patterns (Price 1976) . Knowledge of spatial patterns is considered basic to design of appro- priate sampling plans and provides insight into the biology of the species in question (Shepard and Camer 1976) . Colonization patterns of members of the family Gelechiidae have been studied mainly for the pink bollworm (Pectinophora gossypiella Sa\anders) . The Gelechiidae are considered weak fliers whose patterns of dispersion are strongly influenced by wind, patchiness of the environ- ment, and also voltinism (Shelton and Wyman 19 79a, b; Kaae et al. 1977; Stern 1979) . Little has been done on the study of the factors affecting patterns of dispersion of the TPW in commercial fields. It has been observed that a common feature of TPW is its aggregation at the edges of fields (V.H. Waddill, personal communication). The purpose of this study was to determine the colonization pattern of TPW and to define the effect of edges and hedgerows on the insect's distribution. 154 155 Materials and Methods Dispersal of TPW Male Moths in the Field This study was conducted during February through May, 1980, and May-June, 1981, in order to determine short range pinworm dispersal within the field. TPW was monitored by placing Pherecon Ic® sticky traps 60 cm above the ground and baiting them with sex attractant (95.3%-(E)-4-tridecen-l-01 Acetate; 4.5%- (Z) -4-Tridecen-l-Ol Acetate) (personal communication, R. Heath, USDA, Gainesville, Fla.). In all 0.5-3 ha tomato fields of this study, three traps were placed in each of the major compass directions (north, west, east, and south) ; the trap lines and distances between them were 12, 60, and 300 m from the center of the first field, and 12, 30, and 48 m for the remaining ones. Trapped TPW male moths were recorded and removed daily from each trap during February-March and April-May. This survey deteinnined male moth activity from the border to the center of the field. Colonization Pattern of Tomato Fields by TPW Tomato pinworm egg distribution can be considered an indicator of female moth activity in a field. Because of low TPW lairval motility within the fields, immature stages were also taken as an index of popu- lation dispersion and colonization in the tomato fields. Four fields ca. 1-3 ha located near Homestead, Florida were the study sites. The fields had been under cultivation for several years in a corn- tomatoes- bean- or squash rotation. No insecticides were applied during the 156 sampling time. Three sampling sites were located at rows 0, 30, and 120 m from the edge to the center of the field. Twenty randomly selected plants from each row site were selected weekly, until the faiming practices (tillage, discing, or insecticide application) interfered with sampling. The sampling unit used was larval injuries per plant. Data from each field were analyzed by analysis of variance and means from each site were separated by Duncan's Multiple Range Test (P=0.G5). Effect of Edges and Hedgerows of TPW Distribution Three post-harvested tomato fields abandoned after the 2nd com- mercial harvest were sampled on May 4-22, 1981. Field sizes fluctuated 2 between 0.88-1.32 ha. These were divided into (ca. 1100 m ) quadrats. Two middle leaves per plant were taken as the sampling unit from 20 randomly selected plants per quadrat, to detect the presence of TPW larval injury. The first field was bordered to the northwest by mango trees (Mangifera indica L.), to the north and northeast by tomatoes, and to the south by a main road and grassland. The second field was bor- dered to the northwest by grassland, to the south by a hedgerow or windfall of Casuarina equisetifolia Forst, and to the east and north by a com field and a main road, respectively. The third field was separated to the north and south by a windfall of C. equisetifolia, and bordered to the east by a sweet potato (Ipomoea batata (L.)) field, and to the west by a main road and grassland. For each field, mean numbers of TPW foliar injuries were averaged per quadrat. Each quadrat was considered a treatment. Differences among quadrats were defined by use of analysis of variance and statistical differences were determined by use of Duncan's Multiple Range Test (P=0.05). 157 Results and Discussion Dispersal of TPW Male Moths in the Field The results presented in Fig. 18 reveal a patteim of K, lycopersi- cella male distribution which indicate that males are mainly foxind in the borders of the field. Data indicate male distribution only, al- though I hypothesize (without evidence) that female distribution follows the same pattern. Fewer males (10-50) were captured during the early part of the tomato growing season, February-March, than later, May-June (75-395) . The data probably reflect several successive generations being produced in the field. During February-March, 1980, males were more abundant (62%) in traps located farther from the center of the field (300 m) than in those located closer (12-26%) . The same negative trend toward the center of the field (17-42%) was observed during May- June, 1980-81. The predominant wind direction during February-March was NE, and the proportion of moths captured corresponded to those traps oriented N, W, E, and S, respectively. The behavior of TPW males in- dicated that there may be a dispersal tendency toward areas close to the edges of the field. If the trapped moths were not part of the natal population, data may indicate that TPW is a good colonizer which con- centrates mainly in the field edges, having a slower dispersion within the field. Perhaps the migratory TPW population initiates a fast field colonization at the borders, but later generations colonize the field slower, since there is no need to cover long distances inside the tomato field if food is available. S -H O .-I (U - 159 O X ON c U •a iZ o oo o evj so CO 1-1 a u V) 0 in o a (0 paan^deo sq^o^ 160 c ~ = U TS U : s o E 00 0 ; : ON k> CM E : t T3 cc o U. 1 1 - • »i . IQ •V pajn»dB3 sqiow 161 u c U "O 00 73 u. I a (S (A o a •H -r^ 0 fri TS a 168 OBD a Q Q o "■■■■■ 1 V 0 S 0 H 1 . < saoiBuiox 169 influence of edges on dispersion of TPW adults and that aggregation of TPW is also a function of density. TPW gregariousness was evident in resource exploitation. Conclusions and General Discussion In this research useful data were gathered about the pattern of colonization and dispersion of TPW in tomato fields, as well as data about the effects of edges and hedgerows on TPW population dispersion. TPW adult male moths were generally (62%) more abxindant on borders than in inner regions of tomato fields (12.4%). TPW foliar injuries were generally significantly higher at the field borders than at 30 and 120 m from the border of the field. The infestation rate progressed toward inner areas depending on initial infestation of the border and sampling date. For instance, fields infested during January had lower infesta- tion than those sampled during March to May. Three fields showed significantly higher populations on edges surrounded by windfalls, roads, or pastures, and lower populations on edges surrounded by the same crop or other crops. The accumulation of TPW moths, eggs, and larval injuries on field edges indicate the importance of intensifying sampling in those areas in order to apply control measures to stop infestation progress. There are still many questions unsolved about TPW dispersal patterns . One of them is: Do these moths have a similar dispersal pattern to the one found for P_. gossypiella by Flick and Noble (1961)? It is possible that TPW generations are divided into colonizers , those that migrate at 170 higher altitude locating new food soxirces, and local individuals whose movement inside fields is characterized by a low altitude and erratic flight. CHAPTER VII EGG AND LARVAL PARASITISM OF TOMATO PINWORM IN SOUTHERN FLORIDA Introduction The tomato pinworm (TPW) , Keiferia lycopersicella (Walsingham) has been a serious pest of tomatoes in southern California (Oatman 1970) , Texas (Wellik et al. 1979) and in Florida (Poe 1974). In southern Florida, TPW is a more serious pest during spring when populations build up as the crop matures. Tomato pinworm larval parasites have been studied by Cardona and Oatman (1971) , and seasonal occtirrence of para- sites has been investigated in southern California by Oatman et al. (1979) . Hymenopterous larval parasites attacking TPW in Florida are Apanteles spp. (Poe 1974b) , Temelucha spp. , Sympiesis stigmatipennis Girault, Zagrammosoma multilineatum (Ashmead) , and Parahornius palli- dipes Ashmead (Krorabreii et al. 1979) . The only egg parasites reported have been Trichogramma spp. in southern California (Oatman et al. 1979) and in Colombia attacking the South American pinworm Scrobipalpula absoluta (Meyr.) (Garcia et al. 1974). Reported here are investigations of 1) parasitism of TPW larvae in tomato fields during 1980-81; 2) the parasitism of TPW eggs in the laboratory by the naturally-occurring T. pretiosum (Homestead strain) and by a laboratory reared strain of T. pretiosum (Texas strain) ; 3) the effect of host distribution and host density on the parasitization by 171 172 Trichogramma spp.; and 4) the seasonal occurrence of the Trichogramma spp. in several tomato fields. Materials and Methods Larval Parasitization Twelve to fifteen commercial fields were surveyed for larval parasitism from January 14 through September 5, 1980 and from January 23 through July, 1981. Mined leaves were examined in the fields and taken to the laboratory. Occupied TPW mines were held for adult parasite or moth emergence in an ice cream carton which had about 5 g of white sand in the bottom. No additional food was provided. Therefore, most early TPW instars probably died of starvation. Hosts and parasites were recorded and identified. Percent parasitism was calculated from the numbers of adult moths and parasites which emerged. Egg Parasitism Laboratory studies. The parasitism of tomato pinworm eggs by 2 strains of T. pretiosum was studied in a laboratory at ca. 25 °C, 75 + 2 RH, and scotophase of 11 h. 'Flora-Dade' tomato plants were exposed to oviposition by TPW adults in an insectary. Eight plants with a total of 405 TPW eggs were exposed to ca. 300 females of T. pretiosum (Texas strain) in a cage (24 x 24 x 24 cm) for 24 h. At the same time, excised leaves with a total of 51 eggs were placed in 6 petri dishes (100 x 25 mm) which contained moist filter paper and ca. 3 field-collected females of T. pretiosum (Homestead strain) per dish. Eggs were removed after 24 h. exposure and placed (1 egg per dish) in smaller petri dishes (50 x 9 mm) . 173 The eggs were examined initially to determine whether they were para- sitized and then observed daily for parasite adult emergence. Field studies. Three tests were conducted from May to July, 1981 in abandoned tomato fields near Homestead, Florida, to determine ef- fectiveness of naturally occurring Trichogramma spp. In the first test, I studied the spatial distribution on the plant of both parasitized and nonparasitized TPW eggs. Only 2 leaves from the upper 1/3 and middle 1/3 portions of the plant were sampled since preliminary sampling indicated few eggs were present on the lower leaves. Fifteen plants were collected randomly from a field which had an avg. of 35.2 eggs/plant. Possible differences in egg density and percentage of parasitism at each level were determined by a student's t-test. The second test evaluated the effect of host distribution on the level of parasitism, and the relationship between host density and percent parasitism. Two fields were selected as sites for this study: (Field 1) had an avg. of 3.10 eggs per 2 leaves per plant and (Field 2) had an avg. of 0.70 eggs per 2 leaves per plant. The fields were di- vided into 3 (2 border and 1 middle) sections lengthwise of 0.44 ha each. Two leaves were collected from the middle portions of 80 plants randomly selected per section. Egg parasitism was determined in the laboratory by examining the eggs under a dissecting microscope. Fre- quency distributions of parasitized and nonparasitized TPW eggs were compared with 3 types of distribution (Poisson, Negative Binomial, and 2 Positive Binomial) and tested by x for goodness of fit. Data were transformed to log (x+1) prior analysis. Differences in egg density among sections of each field were determined by the use of Duncan's 174 Multiple Range Test. Correlation between egg density and parasitism was analyzed following arc sine transformations. The third test was conducted to determine the extent of TPW egg paraby T. pretiosinn in Dade County. Fourteen tomato fields were sanpled weekly by collecting 2 middle leaves from 10 randomly selected plants per field. Six of the fields were located in the northern part of the farming area and the rest in the southern portion. Samples were taken to the laboratoiry to determine degree of parasitism. The survey was suspended as the plantings were destroyed by discing or mowing. Results and Discussion Larval Parasitism Parasitization of tomato pinworm larvae ranged from an avg. of 39.34% in 1980 to 46.26% in 1981. During the study, parasitization averaged ca. 2.5, 22, 51, 40, and 48.27% in January, February, March, April, and May, 1980, respectively. Parasitization averaged ca. 49, 46, 53, and 46% in April, May, J\ane, and July, 1981, respectively. Although parasitism ranging from 40-60% was common during 1980-81, there was not a consistent corresponding increase in parasitism with an increase in host density (Table 36) . Of 3 primary parasites reared from tomato pinworm during 1980, Apanteles spp. was the most important, parasitizing 10-66.66% of the larvae. Temelucha spp. and Sympiesis stigmatipennis were the next most abundant parasitoids. During 1981, (Table 37) 4 parasites were reared. Apanteles spp. was again the most important, parasitizing 46% of the larvae, followed by Parahormius pallidipes , Sympiesis stigmatipennis and Chelonus phthorimae in order of importance. 175 e o^ e Eh o OT H (rt M *"> (S -H >l rtJ O 01 4J 4-1 (U rt C 0 n^ )-i 3 Eh -H (d 0 0 ft U O 176 177 5^ e CT* ^ rH +J >1 (tf u -p +J (1) c O M 3 E-t rH 0 0 o U >1 3 3 178 Egg Parasitism Laboratory studies. The results of the laboratory experiment (Table 38) demonstrated that both strains of T. pretios-um can oviposit and develop in TPW eggs. Although the percent parasitism and parasite emergence from eggs parasitized by the Texas strain were 1.2 times greater and 71% less, respectively, than from eggs parasitized by the Homestead strain, these differences may have been due to differences in egg density and searching area rather than differences between strains. It is known that kairomones play an important role in successful para- sitism by Trichogramma (Seabrok 1977) , and this could also account for the observed differences since the Texas strain was reared on Angoumois grain moth (Sitotroga cerealella) eggs. The sex ratio was 50:50 (males and females) for the Texas strain and ranged between 50:50-60:40 (males and females) for the Homestead strain. Field studies. The mean number of TPW eggs and percent parasitism were significantly higher on the middle leaves than upper leaves in test 1 (Table 39) . When the egg density per plant was regressed on percentage 9 parasitism from both levels, the r" obtained was 0.0181 which indicates lack of correlation. This may indicate a complex of hosts other than K. lycopersicella (including Heliothis zea [Boddie] ) . It may also in- dicate that searching capacity of T. pretiosum is a limiting factor in parasitism. The results obtained from the interaction between host distribution and level of parasitism by T. pretiosum is reported in Table 40. The 179 Table 38. Keiferia lycopersicella eggs parasitism by 2 strains of Trichogramma pretiosum in the laboratory. T 25+1 °C; 75+2% RH. ~ % % Days to Parasite No. Eggs Parasitism Emergence Emergence Strain Exposed X + SE X + SE X + SE Texas 405 68.93+2.91 27,47+1.58 8.33+0.862 Homestead 51 57,96+5.92 93,10+1.36 8.50+2,67 180 Table 39. Number of Keiferia lycopersicella eggs collected from two strata and percent of paras itism by- Trichogramma pretiosimi. Leaf Location Total Eggs Avg. per 21 Leaves Range % of Parasitization x/21 Leaves Range Upper Middle 104 281 6.93 a* 18,66 b 0-19 0-50 39.73 a 57.80 b 0-100 26-100 Values followed by different letters are significantly different (P<0.05) by Students t-test. 181 Table 40. Distribution of normal and parasitized tomato pinworm eggs in 2 tomato fields . No. Plants Total TPW eggs Parasitized TPW eggs 80 Field 1 3.09 12.76 2.16 5.45 1.43 1.65 19.39 11.97 Total TPW eggs Parasitized TPW eggs 80 Field 2 0.659 1.35 0.42 0.74 0.526 3.12 0.45 5.46 ^X value of 19.39 is not below the 5% point of 19.67 (V=9 , a=0.05). Therefore, the model is not a good fit to the original counts and agreement with a negative binomial distribution is not accepted at the 95% probability level. X values of 11.97, 3.12 and 5.46 were well below the 5% point of 19.67 (V=9, a=0.05). Agreement with a negative binomial distribu- tion accepted at the 95% probability level. 182 distribution of total TPW eggs only fit the negative binomial for the second field. Parasitized eggs in both firled were fit to the negative binomial distribution. Thus, there is an aggregative distribution of TPW eggs and a concomitant distribution of T^. pretiosum only when low egg densities per plant are found. These results agree with Morrison and Strong (1980) . There were significant differences in host density among the 3 areas of Field 2 but not in Field 1, despite a higher number of host eggs (x=4.04) (Table 41) . The major difference between the fields was that Field 1 was apparently more completely colonized by TPW than Field 2. Egg parasitism ranged from 0-55% for the first field and 0-70% for the second field, despite the lack of correlation (Fig. 21 A 2 2 and B) (r =0.026 and r =0.018, respectively) between the percentage of parasitism and the mean egg density for both fields. There was a negative trend in parasitism for egg density higher than 3 eggs/2 leaves (Field 1) , and a positive trend when TPW egg density was less than 3 (Field 2) . Spatial variations in host density (Morrison and Strong 1980) may result in spatial variations in parasite activity. These results indicate that the distribution of T. pretiosxim followed the same distribution as the TPW eggs; however, there was not a consistent increase in percent parasitism with increased host density. Since TPW eggs are not the only hosts for T. pretiosum, the response of the para- site might be the result of spatial differences in the densities of alternate hosts. The limitation of searching capacity, if a reality, or the response to kairomones, if any, may also play important roles in the degree of successful parasitism. The field survey indicated higher parasitism in the fields located in the northern part of Dade County (Fig. 22A) compared to the fields 183 Table 41. Parasitism of tomato pinworm eggs by Trichogramma pretiosum in 2 fields with different host densities. Area No. eggs in sample x of eggs/2 leaves % Parasitism Field 1 Border 1 363 3.5 a Middle 424 5.0 a Border 2 367 Field 2 3.5 a Border 1 51 0.6 a Middle 22 0.3 a Border 2 91 1.00 b 55 39 30 58 57 70 Values followed by different letters are significantly different at P^.05, Duncan's Multiple Range Test. c en 0) u (u a o (11 - O) CO CM O E 3 o o o CO o >0 o o cs J (UIS3JD)UJ SUjSDJ DJ a6D4U9 3Jaj 0) ^ n +J c k (0 0 c (n rr> ^.^ en ft) (D ^^ 01 (I) iH j:; iH -p a) • o ■p m ■H (fl Tl tn •M S>4 -o u 0) aj n u. -p rH 0 nj (X4 l> o >1 0 k iH ■H IT m tn c •H -a 3 >^ tH 0 (U Q) u <4-l •H ■H 14-4 (1) 0) 01 !>^ n ttl +j u M-l fl O fa ^w 0 0 (I) 4J t) tn c C (fl 0) •H (I) u u u C m 3 2 o tn c o n M 0 •H (1) -P iH a) -P ifl ij 3 c a, O 0 tn (i1 tn H ,^ (1) t) w T1 • C C U 1 0 -H ta sC +J •k n1 (1) CN N r-t IN ■H T1 +J Ti (1) •f-i -rH u tn Fi 3 m m !-4 „„-^ ■H (rt .0 Ui [1. 188 W SI lis V H Vd % < o o z -i a. < I/) S90a Mdl NVIW 189 located in the southern part (Fig. 22C) . Parasitism of TPW eggs ranged from 30% in May to 65% in July for the northern fields. In the southern fields parasitism ranged from 0% in May to 15% in July. The average parasitism for both areas for the study period was 26.9%. Although parasitism ranging from 33-73% was not uncommon when density was higher than 5.2 eggs/2 leaves, there was not a consistent corresponding in- crease in parasitism with an increase in host density except for 1 field surveyed (Fig. 22B) . It is known that the frequency with which a para- site finds hosts is a product of several components. Among them, Burnet (1958) and Hassell (1966) found that the nature of host distribution has considerable influence on the response of entomophagous insects. Studies of the incidence of T. pretios\im on other hosts (Heliothis spp.) should be conducted in tomato fields. This information is needed to understand better how T. pretiosum can be manipulated for maximum benefit of the tomato producer. General Discussion and Conclusions Data obtained from research on parasitism of TPW eggs and larvae in southern Florida indicate that Trichogramma pretiosum, as an egg para- site, and Apanteles spp., as a larval parasite, are the most abundant natural enemies. The parasitoid population increased during May-June, when most southern Florida tomato fields are at post-harvest. This may also indicate insecticide resistance of the parasite species from continuous insecticide applications. Further research is suggested on releases of these parasites on post-harvested fields to reduce TPW resurgence during the following season. CHAPTER VIII EFFECTS OF RAINFALL AND RELATIVE HUMIDITY ON IMMATURE STAGES OF THE TOMATO PINWORM UNDER GREENHOUSE AND FIELD CONDITIONS Introduction Some information concerning the effect of temperature and various environmental factors on Keiferia lycopersicella (Walsingham) , the tomato pinworm (TPW) , is available in the literature (Weinberg and Lange 1980, Poe 1974b, McLaughlin et al. 1979). Little has been p\iblished, however, concerning the study of rainfall and hiomidity influencing the rate of population increase of this insect. Effects of rainfall and humidity on pests of the same family have been studied by Simmons and Ellington (1933) , Hof master (1949) , Warren (1956) , and Clayton and Henneberry (1982) . In general, they found that rainfall and humidity may reduce or increase abundaince of different stages of the Gelechidae species studied. This paper describes a study of the relationship between rainfall and survival of the egg, larval, and pupal stages of TPW the laboratory and in the field. Materials and Methods Seasonal Occurrence of TPW in Experimental Plots To find a possible relationship between temperature and rain- fall and egg deposition by the TPW, oviposition was studied on 8 non-staked tomato plantings cv. Flora-Dade (Nov. 3, 1979; Dec. 5, 1979; Jan. 8, 1980; Oct. 30, 1980; Nov. 25, 1980; Dec. 30, 1980; Jan. 30, 1981; 190 191 Feb. 28, 1981) . Tomatoes were seed-planted at the Agricultural Research and Education Center, University of Florida, Homestead, Florida. Each planting (ca. 450-947 plants) was set in raised beds (3-5) (ca. 45 m long) of Rockdale soil and mulched with light colored plastic. Plants were spaced 38 cm apart. Oviposition per plant was averaged weekly for 3 plantings during 1980 and for 5 in 1981. Each weekly sample consisted of 10-20 plants selected at random. Temperature and rainfall regimes during February through May, 1980 and January through May, 1981 were determined by the Climato logical Weather Station at the Agricultural Research and Education Center, Homestead, Florida. Seasonal Occurrence of TPW in Southern Florida Studies were conducted in 12-15 commercial fresh market tomato plantings in Homestead, Dade County, Florida during 1980-81. The fields were part of a survey program. Every crop differed in size and prac- tices such as insecticide application, but horticultural practices were similar. Each planting during 1980 was systematically sampled for larval injuries by inspecting the whole plant. During 1981, I limited the sampling to inspection of 2 of the middle leaves on each plant. A total of 20 plants per field was sampled. Each planting was sampled before and after harvest, and sampling was suspended when the grower disced or mowed the crop. Surveys continued when new plants emerged in those fields. The survey during 1980 was concluded during October, when the fall tomato crop was planted. The survey during 1931 was concluded in July, 1981. During 1981, pheromone sticky traps (Pherocon Ic®) baited with pheromone (95 . 3%- (E) -3-Tridecen-l-Ol Acetate; 4.5%-(Z)-4- Tridecen-1-01 Acetate) were placed in 5 fields. Male adults trapped 192 were recorded weekly, and the mean niamber compared with immature stage infestation levels in the field. Again, a possible relationship between population increase and environmental factors is expressed. Effect of Plant Water on Oviposition Ovipositional preference related to water content was studied by collection of 2 tomato plants of 5 different ages. Plants were natu- rally infested with TPW eggs in the experimental field mentioned above. The plantings were 6, 5, 4, 3, and 2 months of age. Every week, plants were pulled and taken to the laboratory where the numbers of eggs were recorded. After this, the fresh weight of plant leaves was measured (g) , and 2 days later, the dry weight measured. Water content per plant was estimated as the difference between fresh weight and dry weight. The experiment lasted 6 weeks until plants approached 7.5, 6.5, 5.5, 4.5, and 3.5 months old. Plant water content and oviposition with respect to planting time were analyzed and compared by Duncan's Multiple Range Test at the 0.05 level. Effect of Simulated Rainfall on Larvae An average of 5 second instar larvae per plant, reared on 30 day- old potted tomato plants, was used to determine the effect of rainfall on larval mining. Each plant had 8 +_ 1 leaves. The first treatment consisted of a simulated continuous drizzling (mist) for 24 h. Total water sprayed per plant was 100 cc per day. Drizzling was simulated by use of an automatic mister that sprayed fine droplets for 1 minute every 5 m.inutes over the foliage. The second treatment consisted of spraying 193 the foliage with 50 ml of water 2 times a day. Rainfall was simulated by use of a manual sprinkler. The third treatment was soil irrigation with 100 CO of water per day. The number of injuries per leaf was counted 2 days after the experiment was set, and counts countinued every other day for during 9 days . Treatments were replicated 3 times . The mean numbers of injuries per plant were obtained and compared by analysis of variance (ANOVA) . Fresh weight of leaves consumed was recorded (5 days after the experiment started) from randomly selected infested leaves , following the procedure explained in Chapter I , in order to determine leaf consumption iinder different water regim.es. The larval head capsule width was measured to determine larval instar. Effect of Simulated Rainfall on Pupae Newly formed pupae of TPW were subjected to the following condi- tions: a layer of coarse white sand was placed in a series of boxes (21 X 30 X 23 cm) , constructed of dry lumber, each fitted with a solid lid and a screen (1 mm diam) in the bottom to help water drainage. Boxes were surro\anded with Tanglefoot® to avoid predation by ants. The boxes were held in the greenhouse at 24 +_ 3°C. The pupae used in this study were taken from a culture of 1st generation insects. Insects were reared in a 12:12 light-dark cycle at a 24 + 3°C and 75 + 2% RH. A total of 20 pupae was placed in each box. Two hundred ml (97 + 2% RH) , 100 ml (80 +1% RH) , 50 ml (60 + 10% RH) , and 0 ml (27 + 3% RH) of water were applied every other day with a manual sprinkler. Treatments were replicated 3 times. Percentages of emergence were taken of the number of adults emerging per box. Data were transformed by arc sine procedures before the analysis. Data are presented as actual percent emergence. 194 Results and Discussion Seasonal Occurrence of TPW Eggs in Experimental Plots TPW oviposition peaked early during 1980 (Fig. 23) . Numbers during this year were higher than during 1981. The number of eggs increased sharply during the first 3 months of 1980, and this increase was main- tained until mid-May, 1980, when the tomato leaf area decreased sharply. Average daily temperature in the field increased after the middle of March, 1980, but remained stable around 20-25° C during March through April, 1980, The amount of rainfall was negatively correlated with oviposition. Oviposition increased when rainfall was zero during March and decreased at the beginning of April when rainfall increased. The increase in rainfall seemed to be related to low ovipositional activity during April, and its delayed effects can be seen in Fig. 23. During 1981, increased oviposition was observed as temperature in- creased and rainfall decreased. The temperature increased during January through April, stablizing around 21-24° C during the later months. When rainfall was low the pattern of oviposition increased abruptly. This may suggest a pattern in which moth activity can be related to oviposition patterns. This result agrees with the results found by Simmons and Ellington (1933) who stated that high humidity reduced 13 times the average number of eggs laid by Sitotroga cerealella (Gelechiidae) . Figure 23. Seasonal abundance of TPW eggs in experimental fields, related to temperature and rainfall regimes during A) 1980 and B) 1981, in Homestead, Florida. 196 C (0 ■5. \ M O) « 10. 5_ 20 10 E 80 " £ — 60 . TJ . •^ 40 C ■ ■M M 20 . 0 J 1 « r -I 1 1 r Sampling Week juB|d/se6a Mdl (lutu) ije^ujey 198 Seasonal Occurrence of TPW Larvae in Southern Florida TPW larval density decreased from January 3 through the 2nd week of February, then peaked at the beginning of March, 1980 (Fig. 24) . The population declined through March and the 1st weeks of April , but steadily increased during the 3rd week of this month, to reach the highest density in May. Most growers disced and mowed the tomato fields during the 4th week of May, causing populations to decrease through June-October. Despite this, larval injuries (Fig. 24) were found on volunteer plants during these months. I consider that field infestation is dependent on temperature and rainfall regimes and crop management. It is observed in Fig. 30 that despite an increase in temperature during June-August, rainfall also increased during these months, in- dicating a possible negative effect on pest abundance. During 1981, TPW larval density in commercial tomato fields was almost negligible. Population densities were low as late as Februairy 25, and population growth was not evident until early April. Early, low temperatures seemed to be the factor suppressing the pest this year; however, the high temperatures and low rainfall coincided with the pest peaks. Effect of Plant Water on Oviposition Effect of oviposition response to leaf water content of the whole plant has been demonstrated for Pieris rapae (L.) by Wolfson (1980). Highest water content of each plant was found in those plants planted Figure 24. Seasonal abundance of TPW larvae in tomato fields, related to temperature and rainfall regimes during 1980-81, in Homestead, Florida. 200 3 io_ U 30. 15. 0. 500. se 300 . 100 . — 13— 1 — Tr-r ^C^ -1 1 T I. I 1 1-"^^ Sampling Week 201 2Q2 latest (4 weeks old) . Greater mean nimibers of eggs obtained (Table 42) corresponded again to the youngest crop followed by those crops 3, 4, 5, and 6 months old. The niombers of eggs oviposited among the oldest crops were not significantly different. When plantings were 3, 4, 5, and 6 months old, the water content of plants was 1.46, 2.69, 9.79, and 8.72 times less, respectively, than that found on 2-month old plants. These results may indicate the possible relationship between water content and oviposition; however, when these parameters were regressed, the co- efficient of determination was as low as 0.27, despite an F value of 7.06. Perhaps the high (71.37%) CV obtained tended to minimize the effect of this relationship. Oviposition preference is not only related to water content, but possibly to the amo;ints of other substances in the plant leaf. Effect of Simulated Rainfall on Larvae The data in Table 43 show the effects of rainfall on the number of injuries on each plant. The numbers of larval injuries per plant were always significantly lower when water was sprayed on the foliage. Thus, compared to the soil irrigated treatment, the numbers of injuries were reduced 53% and 48% when water was applied to the foliage. The dif- ferences during 10 days after treatment are shown in Fig. 25. It was found that when water was applied continuously to the foliage, the larvae stopped mining the leaves and started feeding externally on the leaf, or constructed a silk tent to protect against the excessive amount of water. Leaf consumption was also reduced 1.46 and 3.57 times when the double and continuous water spray was used. Since there was no 20.3 Table 42. Plant water content in five tomato plantings related to oviposition by the tomato pinworm. Planting No. Leaf Water Content (Fresh Mean Eggs/Plant Weight - Dry Weight) (g) 1 11.05 d* 0.56 d 2 9.05 e 0.46 e 3 35.83 c 0.63 c 4 65.99 b 1.51 b 5 96.43 a 1.84 a * Numbers followed by different letters were significantly dif- ferent by Duncan Multiple Range Test (P=0.05). 204 Table 43. Effect of simulated rainfall on foliar larval injuries caused by the tomato pinworm Keiferia lycopersicella on plants grown under greenhouse conditions . Treatment Average Number of Leaf Consumption injuries/plant (mg) * 100 cc water applied daily as a mist 45.78 b** 353.2 b 200 cc water applied as spray twice a day 42.071 b 853.2 ab 200 cc water applied to the soil every other day 86.667 a 1261.0 a * Leaf consumption measured days after treatment on larvae. **Values followed by different letters were significantly different by Duncan Multiple Range Test (P=0.05) . 0) M 3 -P M-l « n U Qi m Ck >i h 1 .-1 rri •w C (0 •H T3 u 3 Tl tji > -P (t3 C (d •> tH tn U c o i^ •H a) ■u a, -H T3 en G 0 u !-l (U (I) -a G ^ 3 C 1-1 H c (0 n1 4-1 - > ' 1 mm 19 ' 1 ••■ 0 a 5 1 u> U) 1 V __ j: 0 o u o CN 0) -^ w 3 rH 0 fH x: ^ to CO u U .-( a; (T3 ■73 -H c u 3 -H n-t E r^ Ul CN C o ■ -H (1) -P u •rH 3 nn tTi c •H <) h u 216 T3 T3 (J 3 W1 o •r- Z£ C O (O (C OJ (O 3 a; c in (1) 0] M v 3 0^ ^ ■H «J Cm JS 119 ui/saunfui /v\dl o in o ,XU/S»UBU 0|BU10J_ 220 Table 45. Effect of crop age of post-harvested tomato plants on volun- teer plants and number of tomato pinworm lajrval injuries. Crop Months After Mean Plants/ Mean TPW Injuries/ Planted Harvest m m Oct. 30, 1980 3 0.1222 b* 0.1389 b Nov. 25, 1980 2 0.41 b 0.641 b Dec. 30, 1980 1 2.53 a 4.006 a * Values followed by different letters were significantly different according Duncan Multiple Range Test {P=0.05). 221 of herbicide used in the new crop. Field 2 had a slower increase of plant and larval levels after entering a period of abandonment- Experiment — 1981 Results of the second experiment are expressed in Table 45. Tomato crops planted earlier (Oct. -Nov. , 1980) has a significantly lower mean number of volunteer plants than the younger one (Dec, 1980). The mean number of TPW injuries was significantly larger in the yoxinger planting than in the others. Abandoned fields had generally lower n\imbers of plants than mowed and disced fields. The numbers of injuries were similar for any treatment. Effects of planting time on the treatments were obvious (Tables 46-47) . The older planting had a greater number of injuries per plant when abandoned than when the planting was mowed or 2 disced. The second planting (Nov. , 1980) again had more injuries per m if abandoned compared to the other treatments. In contrast, the younger planting had a higher number of injuries when disced and mowed than when abandoned. The effects are explicable. Older fields have fewer viable seeds that will germinate than those from younger fields in which seeds are immediately incorporated into the soil. Effects from secondary host plants as oversummering sites have been suggested for the TPW. The availability of off-season tomatoes helps to maintain the TPW population but also helps in build up of natural enemies. Use of abandoned or "U-pick" fields for pest management of TPW by constant release of parasitoids would be a practice to reduce pinworm populations for the next season, without interfering with the fairmer interests and environmental concerns. Table 46. General effect of cultural practices on volunteer tomatoes and infestation by tomato pinworm. 2 2 Treatment Mean Plants/m Mean TPVJ Injuries/m Disced 1.377 a* 1.46 Mowed 1.15 a 1.70 Abandoned 0.54 b 1.67 * Values followed by the same letter are not significantly different according to Duncan's Multiple Range Test. [■able 47. Effect of planting age and cultural practices on volunteer tomato plants and number of TPW injuries. Planting 2 2 Treatment Number plants/m Number injuries/m 1) Oct. 30, 1980 disced mowed abandoned 2) Nov. 25, 1980 disced mowed abandoned 3) Dec. 30, 1980 disced mowed abandoned 0.00 0.00 0.075 b* 0.13 0.316 a 0.28 0.017 c 0,16 b 0.283 b 0.16 b 1.066 a 1.51 2.333 b 4.233 2.166 a 4.55 1.333 a 3.23 * Values followed by the same letter are not significantly different according to Duncan's Multiple Range Test (P=0.05). 224 Consequently, effects of cultural practices to help control the TPW in tomatoes can be considered in two ways. First, these practices may be considered in relation to the pest density or, second, in relation to the impact on natural control agents. For the most part, cultural measures usually modify the environment to the disadvantage of the pest (Anonymous 1979) . However, cultural practices can destroy the host and the parasites or natural enemies of the host. In southern Florida, the widespread cultural practices could account for part of the insect reduction during the past 10 years. The impact of some of the post- harvest practices related to pest density has been demonstrated in this study . GENERAL DISCUSSION AND CONCLUSIONS In this research useful data were gathered about tomato plant phenology, spatial distribution of eggs and larval populations of Keiferia lycopersicella (Walsingham) , and effects of tomato pinworm (TPW) larvae on tomato yield. Studies of the effects of natural enemies, rainfall, edgerows and cultural practices on the population dynamics of this insect provided important data for better management of TPW. A different IPM approach is suggested based on data about different stages of development of tomatoes. 'Flora-Dade' tomato phenology was described based on 3 major stages: vegetative, reproductive and senescent. Vegetative stages (TV -TV ) were determined by presence of primary leaves and secondary vegetative growth on plants 1-35 days old. Repro- ductive stages TR -TR were based on nimbers of flowers and fruits on plants 40-110 days old. The results indicate that reproductive stages (TR -T?v ) should be subdivided so more precise pest management decisions can be made. This classification can be refined by dividing the TR_ stage into 2 substages. One stage is during fruit formation and the other stage is during harvest maturity. The rationale behind this is to improve economic thresholds during fruit formation and just before hairvest. Research on sampling TPW eggs indicated that high cost ($8.4-84) of egg sampling combined with high SE/x ratio (20-100) reduce the practicality 225 226 of sampling this immature stage in the field. Data on TPW distribution indicated that eggs were found mainly (44-68%) in the middle-upper canopy of the plant. Therefore, proportional sampling can be allocated for the upper external stratum (nh=6) followed by the middle internal stratum (nh=5) , and 3,4,4 and 1 samples for lower external, upper, middle and lower internal canopies, respectively. Apparently, the research of Burton and Schuster (1981) provided similar data that allow the hypothesis of female attraction for oviposition in the upper plant- part. Future research must be carried out on the relationship between female trapping and egg presence in the field. The research reported here indicates the need for more evidence on TPW oviposition on dif- ferent tomato plant stages . Research on sampling TPW injuries per plant indicated that the percent of the SE corresponding to the mean was 11-31% for 20-25 plants sampled when the population was low (0.2-2.12 injuries per plant) and was 21-29% for 15-20 plants when the population was high (11.05-17.2). By sampling two leaves from the upper and lower canopies I account for 32-34% of the total lairval injury per plant. Results indicated a measure of efficiency (RNP) fluctuating between 0.49-1.20 for 50 and 5 plants inspected when the population was small and 0.49-3.18 for 50 and 5 plants whan the population was large. Results generally agree with those of Wellik et al. (1979) indicating that larvae are mainly (50-75%) located in the middle-lower plant canopy. Therefore, more samples should be allocated to the middle and lower strata. The average niomber (n=20) was 2, 3, and 5 samples for upper, middle and lower external canopies, and 0, 5, and 3 from upper, middle and lower internal canopies. 227 respectively. I recommend more research on the relationship between oviposition and the population index (niimber of injuries per plant) . These data are necessary to establish the prediction of economic injury levels. Economic injury level studies provided useful data to determine yield losses from high levels of larvae (1-14) per plant. The largest yield reduction was from 12-14 TPW larvae per plant. Since the TPW larva attacks both leaves and fruits, the results suggest that sampling from the lower canopy will be more useful than sampling from the upper canopy. These results disagree with those obtained by Wolfenbarger et al. (1975). Conflicting views about sampling to detect an economic injury level will be fewer if EIL is determined for every stage of tomatoes. I suggest further research to develop economic injury levels for the 4 main stages of development. In this case, it would help to avoid relying on the stages of the plant close to harvest (TR ) , when it is too late to apply control measures . The role of several factors (parasitoids, field edgerows , rainfall, and horticultural practices) influenced TPW population dynamics in southern Florida. For instance, the effect of natural enemies was an important TPW mortality factor. The role of larval and egg parasitoids increased after the main crop harvest. Levels of TPW larval para- sitization fluctuated between 39.3-42.3% during 1980-81. I consider the larval parasitoid Apan teles spp. and the egg parasitoid Trichogramma pretiosum Riley the most promising biological control agents for TPW in southern Florida. Apanteles spp. appeared earlier during the winter season, increasing in density during the months of April-June. T. 228 pretiosiim was foxind to be a TPW egg parasitoid with an intermediate level (33-73%) of parasitism. A conclusion from these data is that studies should be encouraged which focus on the effect of parasitization of larvae and eggs following releases of these beneficials in post- harvested fields. I consider post-harvest agroecosystem management to be a good strategy to assure a low pest density. I recommend further study of TPW parasitoids to find pesticide resistant strains. Results describing the patterns of colonization of TPW in the field provide answers to the acciomulation of TPW in several areas of the field. Data suggested that this microplepidopteran tends to aggregate near field borders, especially near windfalls. These results may en- courage further research in dispersal of gelechiids. To understand the dispersion of the different TPW generations, it is important to estab- lish which generation migrates over long distances and which disperses over nearby fields. From a practical standpoint, such knowledge of TPW aggregation can be used for control and monitoring of TPW populations. Research on effects of abiotic factors such as rainfall indicated reasons for TPW population reduction during 1980-81. The use of arti- ficial rainfall on TPW larvae and pupae demonstrated that when plant foliage was irrigated there was a behavioral change in larval foliar consumption which resulted in 50% reduction of injuries by larvae compared to injuries on soil-irrigated plants. Adult emergence was reduced 93% when high levels of water were applied to the soil. Never- theless, these experiments require a more elaborate microclimatic study of the pest. I consider it useful to link the similarities of popula- tion dynamics of this pest and the potato tuberworm because of their parallel activities related to temperature and rainfall regimes. 229 The results from evaluation of cultural practices on populations of TPW indicate that post-harvested tomato crops planted earlier (Oct.- Nov. , 1980) had a significantly lower mean number of volunteer plants than did the crop planted later (Dec, 1980). Mean numbers of injuries 2 per m were higher in crops planted later (Dec, 1980) than in crops planted earlier (Oct. -Nov., 1980). Despite the complexity and dif- ficulty of proving which cultural practices are most adequate for a sound TPW management program, two different approaches could be taken. First, practices such as mowing, burning, and discing may eliminate tomatoes as a source of TPW infestation. This implies more supervision from farmers and agricultural agents of post-harvested fields. Secondly, if habitat management of tomatoes is desired the use of "U-pick," abandoned fields or fields where tomatoes grow voluntarily, should be used as a source for gathering parasites and predators. Natural enemies augmentation could be used to reduce TPW infestations for the next tomato growing season. This may be a large step toward improved management of the TPW. Several additional studies are needed so that TPW population assessment can be conducted most efficiently. At present, entomologists must continue research on TPW monitoring as well as studying the pos- sible relationship between egg numbers and adults caught by pheromone trapping. In general, a better IPM program in tomatoes will develop when sampling techniques as well as multiple economic injury levels are determined for TPW and other direct pests of tomatoes. REFERENCES Abdel-Salam, A.M., A.M. Assem, and G.A. Abdel-Shaheed. 1971. Experi- mental studies on tomato pests. Z. Angew Entomol. 69:55-59. Alvarez- Rodriguez, J. A. 1977. Crop life tables for appraisal of pest injury to tomatoes. Ph.D. Dissertation. Univ. of Florida, Gaines- ville, Fla. 105 p. Ambrust, E.J., and G.G. Gyrisco. 1975. Forage crops insect pest management. In R.L. Metcalf and W.H. Luckman (eds.). Introduction to insect pest management. John Wiley and Sons, New York, pp. 445-469. Andrewartha, H.G. , and L.C. Birch. 1974. The distribution and abun- dance of animals, 6th ed. Univ. Chicago Press, Chicago. 781 p. Anonymous. 1971. Crop loss assessment methods. FAO manual on the evaluation and prevention of losses by pests, diseases and weeds, L. Chiarappa, ed. Food and Agric. Organ. United Nations, Rome. 136 p. Anonymous. 1979. The basic principles of insect population suppression and management. U.S.D.A. Agric. Handb. No. 512. Washington, D.C. 659 p. Anonymous. 1981. Planted acreage. Florida agriculture state vegetable preliminary survey, 1980-1981. Fla. Crop Livest. Rep. Serv. 16 p. Anonymous. 1982. Annual report, 1980-1981. Fla. Tomato Comm. Orlando, Fla. 21 p. Anscombe, F.J. 1950. Sampling theory of the nagative binomial and log- arithmic series distributions. Biometrika 37:358-382. Antonio, A.Q. 1977. The behavior and biological assay of the tomato pinworm moth, Keiferia lycopersicella (Walsingham) Gelechiidae: Lepidoptera, and preliminary field studies of its sex pheromone . M.S. Thesis. Univ. of Florida, Gainesville, Fla. 65 p. Barfield, C.S. 1981. Understanding and implementing pest management strategies in agricultural systems. Burgess Publ . Co., Minne- apolis. 165 p. 230 231 Batiste, W.C., J. Joos, and R.C. King. 1970a. Studies on sources of the tomato pinworm attacking tomatoes in northern California. J. Econ. Entomol. 63:1484-1486. Batiste, W.C., R.C. King, and J. Joos. 1970b. Field laboratory evalu- ations of insecticides for control of the tomato pinworm. J. Econ, Entomol. 63:1479-1484. Batiste, VI. C, and W.H. Olson. 1973. Laboratory evaluations of some solanaceous plants as possible hosts for the tomato pinworm. J. Econ. Entomol. 66:109-111. Beck, S. , and L.M. Schoonhoven. 1980. Insect behavior and plant re- sistance. In E.G. Maxwell and P.R. Jennings (eds.). Breeding plants resistant to insects. John Wiley and Sons, New York, pp. 115-135. Bliss, C.I., and R.A. Fisher. 1953. Fitting the negative binomial distribution to biological data. Biometrics 9:176-200. Borror, D. J. , D.M. Delong, and C.A. Triplehom. 1976. An introduction to the study of insects, 4th ed. Holt, Rinehart and Winston, New York. 825 p. Bottrell, D.R. 1979. Integrated pest management. U.S. Gov. Print. Off., VJashington, D.C. 120 p. Brazzell, J.R., and D.F. Martin. 1957. Oviposition sites of the pink bollworm on cotton plants. J. Econ. Entomol. 50:122-124. Bryan, K.H., J.W. Strobel, and J.D. Dalton. 1967. Effects of plant populations, fertilizer rates on tomato yields on Rockdale soil. Proc. Fla. State Hort. Soc. 80:149-156. Burnet, T. 1958, Effect of host distribution on the Encarsia formosa (Hymenoptera: Chalcidoidea) , Can. Entomol. 90:179-191. Burton, R.L., and D.J. Schuster. 1981. Oviposition stimulant for tomato pinworms from surfaces of tomato plants. Ann. Entomol. Soc. Am. 74:512-515, Busch, A. 1928. Phthorimaea lycopersicella new species (Family Gele- chiidae) , a leaf feeder on tomato (Lep.). Hawaii Entomol. Soc, Proc, 7:171-176. Capps, H.W, 1946. Description of the larva of Keiferia peniculo Heinrich, with a key to the larvae of related species attacking eggplant, pepper, potato and tomato in the United States. Ann. Entomol, Soc. Am. 39:561-563. 232 Cardona, C. , and E.R. Oatman. 1971. Biology of Apanteles dignus (Hymenoptera: Braconidae) a primary parasite of the tomato pin- worm. Ann. Entomol. Soc. Am. 64:996-1007. Castineiras, A., and L.R. Hernandez. 1980. New Hosts of Spilochalcis hirti femora (Ashmead) (Hymenoptera: Chalcididae) from Cuba. Lab. Cent. Lucha Biol. Dir. Gral. Sanidad Veg. Minist. Agric. Havana, Cuba, No. 209. 9 p. Chapman, A.J. , L.W. Noble, O.T. Robertson, and L.C. Fife. 1960. Sur- vival of the pink bollworm under various cultural and climatic conditions. U.S.D.A. Texas Agric. Sta. Res. Rep. No. 34. 21 p. Clayton, T.E., and T.J. Henneberry. 1982. Pink bollworm: effect of soil temperature on moth emergence in field and laboratory studies. Environ. Entomol. 11:147-149. Cloudsley-Thompson, J.L. 1962. Microclimates and the distribution of terrestrial arthropods. Ajinu. Rev. Entomol. 7:199-219. Cochran, W.G. 1977. Sampling techniques, 3rd ed. John Wiley and Sons, New York. 428 p. Condrashoff, S.F. 1964. Bionomics of the aspen leafminer, Phyl- locnistis populiella Cham. (Lep. : Gracillaridae) . Can. Entomol. 96:857-874. Cooper, A.J. 1964. Relations between growth of leaves, fruit and shoot of glasshouse tom.ato plants. J. Hort. Sci. 39:173-181. Cosenza, G.W. , and H.B. Green. 1979. Behavior of the tomato fruitworm, Heliothis zea (Boddie) on susceptible and resistant lines of pro- cessing tomatoes. Hort. Sci. 14:171-172. Cottam, G. , and J.T. Curtis. 1956. The use of distance measures in phytosociological sampling. Ecology 37:451-460. Davis, D.R. , R.R. Kincaid, and F.tM. Rhodes. 1970. Mulches reduce soil temperature under tomato and tobacco plants in Florida. Proc. Fla. State Kort. Soc. 83:117-119. Dethier, V.G. 1959a. Food-plant distribution and density and larval dispersal as factors affecting insect populations. Can. Entomol. 91:531-596. Dethier, V.G. 1959b. Egg-laying habits of Lepidoptera in relation to available focd. Can. Entomol. 91:554-561. Djamin, A. 1970. Biologies of the tomato pinworm Keiferia lycoper- sicella (Wals.) Lepidoptera: Gelechiidae and its parasite Apanteles scutellaris Muesebeck (Hymenoptera: Braconidae) with special refer- ence to the influence of temperature on population increase. Ph.D. Dissertation. Univ. of California, Berkeley. 126 p. 233 Doreste, S., and Nieves, M. 1968. Laboratory studies on the life-cycle of the tobacco, patato and tomato leafminer, Phthorimaea operculella. Agron. Trop. 18:461-474. Elliott, J.M. 1979. Some methods for statistical analysis of samples of benthic invertebrates. Freshwater Biol. Assoc. Publ. No, 25. The Ferry House, Ambleside, pp. 14-36. Elmore, J.C. 1937. The tomato pinworm. U.S.D.A. Circ. No. 440. 8 p. Elmore, J.C, and A.F. Rowland. 1943. Life history and control of the tomato pinworm. U.S.D.A. Tech. Bull. No. 841. 30 p. Fehr, W.R., C.E. Caviness, D.T. Burmood, and J.S. Pennington. 1971. Stage of development for soybeans. Glycine max (L.) Merrill. Crop Sci. 11:929-931. Garcia, F. , C. Cardona, A. Saldarriaga, and R. Cardenas. 1974. Scrobi- palpula absoluta (Meyrick) a new pest of tomatoes in Colombia. Mem. 2nd Congr. Soc. Entomol. Colomb. Cali, Colombia, pp. 49-57. Geraldson, CM. 1962. Growing tomatoes and cucxambers with high anal- ysis fertilizer and plastic mulch. Proc. Fla. Stat Hort. Soc. 75: 253-260. Geraldson, CM. 1975. Evaluation of tomato production efficiency with relevance to contributing components. Proc. Fla. State Hort. Soc. 88:152-155. Click, P. A., and L.W. Noble. 1961. Airborne movement of the pink boll- worm and other arthropods. Agric. Res. Serv. U.S.D.A. Tech. Bull. No. 1255. 19 p. Gomez, K.A. , and R.C Bernardo. 1974. Estimation of stem borer damage in rice fields. J. Econ. Entomol. 67:509-513. Gossard, T.W., and R.E. Jones. 1977. The effect of age and weather on egg laying in Pieris rapae L. J. Appl. Ecol. 14:65-71, Greene, R. , K. Mathis, L. Polopolus , and J. Holt. 1980. Economic data for Florida agriculture, 1975-1980. Univ. of Fla. Inst. Food and Agric. Sci., Gainesville, Fla. 152 p. Grooves, G.R. 1974. Report of the year. Dep. Agric. and Fish. Ber- muda . 9 p . Hall, D.G., and G.L. Teetes. 1982. Yield loss-density relationships of four species of panicle-feeding bugs in sorgh\am. Environ. Entomol. 11:738-741. 234 Harcourt, D.G. 1962. Design of a sampling plan for studies on the population dynamics of the imported cabbage worm, Fieri s rapae (L.) (Lepidoptera: Pieridae) . Can. Entomol. 94:849-859. Harding, J. A. 1971. Field comparisons of insecticidal sprays for con- trol of four tomato insects in south Texas. J. Econ. Entomol. 64:1302-1304. Hassell, M.P. 1966. Evaluation of parasite and predator responses. J. Anim. Ecol. 35:65-75. Henson, W.R., and R.W. Stark. 1959. The description of insect numbers. J. Econ. Entomol. 52:847-850. Hillhouse, T.L., and H.N. Pitre. 1974. Comparison of sampling techni- ques to obtain measurement of insect populations on soybeans. J. Econ. Entomol. 67:411-414. Hinton, H.E. 1981. Biology of Insect eggs. Pergamon Press, Oxford. pp. 51-85. Hofmaster, R.M. 1949, Biology and control of the tomato txoberworm with special reference to eastern Virginia. Va. Truck Exp. Sta. Bull. 111:1828-1382. Hopkins, A.R., R. Moore, and W. James. 1982. Economic injury level for Heliothis sp. larvae on cotton plants in the four-true-ieaf to pinhead square stage. J. Econ. Entomol. 75:328-332. Hurd, R.G., A. P. Gay, and A.C. Mountifield. 1979. The effect of partial flower removal on the relation betv/een root, shoot, and fruit growth in the indeterminate tomato. Ann. Appl. Biol. 93:77-89. Iwao, S. 1971. Dynamics of nvimbers of a phytophagous lady beetle, Spilachna vlgintioctomaculata , living in patchily distributed habitats, In_ P.J. den Boer and G.R. Gradwell (eds.), Proc. Adv. Study Inst. Dynamics Number Populations, Oosterbeek, 1970. Oosterbeek. pp. 129-147. Johnson, A.W. 1979. Tobacco budworm damage to flue-cured lobacco at different plant growth stages. J. Econ. Entomol. 72:602-605. Johnson, F.A. , P.. I. Sailer, J.E. Brogdon, J.E. Arnold, and C. Musgrave. 1978. Current policies for utilization of Trichogramma. In Trich.ogramma research and development conference: status and use strategies, December 14-15, Atlanta, Ga. pp. 13-14. Kaae, R.S., H.H. Shorey, L.K. Gaston, and D. Sellers. 1977. Sex phero- mones of Lepidoptera: seasonal distribution of male Pectinophora qossypiella in cotton-growing area. Environ. Entomol. 6:284-286. 235 Katti, S.K. 1966. Interrelations among generalized distributions and their components. Biometrics 22:44-52. Kemp, W.P., and G.A. Simmons. 1978. The influence of stand factors on parasitism of spruce budworm eggs by Trichogramma mi nu turn. Envi- ron. Entomol. 7:685-688. Kennedy, G.G., and W.R. Henderson. 1978. A laboratory assay for resis- tance to the tobacco homworm in Lycopersicon and Solanum spp, J. Am. Soc. Hort. Sci. 103:334-336. Kennedy, G.G., and R.T. Yamamoto. 1979. A toxic factor causing resis- tance in wild tomato to the tobacco homworm and some other insects . Entomol. Exp. et Appl. 26:121-126. Keularts, J.L.W. 1980. Effect of the vegetable leafminer Liriomyza sativae Blanchard, and the associated plant pathogens on yield and quality of the tomato, Lycopersicon esculeutum Mill, cv. Walter. Ph.D. Dissertation. Univ. of Florida, Gainesville, Fla. 154 p. Krebs, C.J. 1978. Ecology: the experimental analysis of distribution and abundance, 2nd ed. Harper and Row, New York. 678 p. Krombreii, K.V. , P.P. Hurd, Jr., and D.R. Smith. 1979. Catalog of Hy- menoptera in America north of Mexico, Vol. 3. Smithson. Inst. Press, Washington, D.C. 2735 p. Lange, H.W. , and L. Bronson. 1981. Insect pests of tomatoes. Ann. Rev. Entomol. 26:345-371. Lenghty, H.W., M.H. Gallatin, J.L. Malcom, and F.B. Smith. 1954. Soil associations of Dade County, Florida. Univ. of Fla. Agric. Exp. Sta. Circ. No. 5-77. 2 p Lewis, T. 1969. The distribtuion of flying insects near a low hedge- row. J. Appl. Ecol. 6:443-452. Lindquist, R.K. 1975. Insecticides and insecticide combinations for control of tom.ato pinworm larvae on greenhouse tomatoes: a progress report. Ohio Agric. Res. Dev. Cent. Wooster, Ohio. No. 82. pp. 37-39. Lloyd, M. 1967. Mean crowding. J. Anim. Ecol. 36:1-30. Loucks, O.L. 1970. Evolution of diversity, efficiency and community stability. Am. Zool, 10:17-25. Luckman, W.H. , and R.L. Metcalf. 1975. The pest management concept. In_ R.L. Metcalf and W.H. Luckman (eds.) , Introduction to pest manage- ment. John Wiley and Sons, New York. pp. 3-36. 235 Lyons, L.A. 1964. The spatial distribution of two sawflies and methods of sampling for the study of population dynamics. Can. Entomol. 96:1375-1407. Mallea, A.R., G. Macola, S. Garcia, J.G. Bahamondes , and L.A. Suarez. 1972. Nicotiana tabacum L. var. virginica, a new food-plant of Scrobipalpula absoluta (Meyrrick) Povolny (Gelechiidae - Lepi- doptera) . Rev. Fac. Cienc. Agric. Univ. Nal. Cuyo Argentina, pp. 13-15. Marsh, P.M. 1975. A new species of Apan teles from South America being introduced into California (Hymenoptera: Braconidae) . Pan-Pac. Entomol. 51:143-146. McLaughlin, J.R. , A.O. Antonio, S.L. Pee, and D.R. Minnick. 1979. Sex pheromone biology of the adult tomato pinworm, Keiferia lycoper- sicella (Wals.) . Fla . Entomol. 62:35-41. Meisner, J., K.R.S. Aschner, and D. Lavie. 1974. Factors influencing the attraction to oviposition of the potato tubermoth, Gno r imo s chema operculella Zell. Z. Angew. Entomol. 77:179-189. Messenger, P.S. 1959. Bioclimatic studies with insects. Annu. Rev. Entomol. 4:183-205. Messenger, P.S., E. Biliotti, and R. van den Bosch. 1976. The impor- tance of natural enemies in integrated control. In_. C.B. Huf faker and P.S. Messenger (eds.). Theory and practice of biological control. Academic Press, New York. pp. 543-563. Middlekauff, W.W., C.Q. Gonzalez, and R.C. King. 1963. Effect of vari- ous insecticides in control of caterpillars attacking tomato in California. J. Econ. Entomol. 56:155-158. Milne, A. 1957. The natural control of insect populations. Can. Entomol. 89:193-213. Mitchell, A. J. , and M.R. Ensign. 1928. The climate of Florida. Univ. of Fla. Agric. Exp. Sta. Bull. No. 200. 300 p. Morrill, A.W. 1925. Commercial entomology on the west coast of Mexico. J. Econ. Entomol. 18:707-716. Morrisson, G. , and D. Strong. 1980. Spatial variations in host den- sity and the intensity of parasitism: some empirical examples. Environ. Entomol. 9:149-152. Mumeek, A.E. 1924. The effects of fruit on vegetative growth in plants. Proc. Am. Soc. Hort. Sci. 21:274-276. Nicholson, A.J. 1958. Dynamics of insect populations. Annu. Rev. Entomol. 3:107-135. 237 Nishijima, Y. 1960. Host plant preference of soybean podborer, Grapholita glicinivorella Matsumura (Lep. , Eucosmidae) - Entomol. Exp. et Appl. 3:38-47. Oatman, E.R. 1970. Ecological studies of the tomato pinworm on tomato in southern California. J. Econ. Entomol. 63:1531-1534. Oatman, E.R., J. A. Wyman, and G.R. Platner. 1979. Seasonal occurrence and parasitization of the tomato pinworm on fresh market tomatoes in southern California. Environ. Entomol. 8:661-664. Pianka, E.R. 1978. Evolutionary Ecology, 2nd ed. Harper and Row, New York. 397 p. Pimentel, D. 1961a. The influence of plant spatial patterns on insect populations. Ann. Entomol. Soc. Am. 54:61-69. Pimentel, D. 1961b. Species diversity and insect population outbreaks. Ann. Entom.ol. Soc. Am. 54:76-86. Peterson, A. 1964. Entomological techniques: how to work with insects, 10th ed. Entomol. Reprint Specialists, Los Angeles. 435 pp. Poe, S.L. 1972. Populations of insect pests on commodities grown on the west coast of Florida. Univ. of Fla. , Inst. Food and Agric. Sci., Bradenton A.R.E.C. Res. Rep. GC72-10. 9 p. Poe, S.L. 1973. The tomato pinworm in Florida. Univ. of Fla., Inst. Food and Agric. Sci., Bradenton A.R.E.C. Res. Rep. GC73-2. 4 p. Poe, S.L. 1974a. Populations of arthropod pests on commodities grown on the west coast of Florida in 1973. Univ. of Fla., Inst. Food and Agric. Sci., Bradenton A.R.E.C. Res. Rep. GC74-1. 7 p. Poe. S.L. 1974b. Emergence of Keiferia lycopersicella (Lepidoptera: Gelechiidae) and Apanteles sp. (Hymenoptera: Braconidae) from pupae and soil treated with insect growth regulators. Entomophaga 19: 205-211. Poe, S.L., J. P. Crill, and P.H. Everett. 1975. Ecology of tomato pin- worm relative to production management in semi-tropical agriculture. Proc. Fla. State Hort. Soc. 8:160-165. Poe, S.L., and P.H. Everett. 1974. Comparison of single and combined insecticides for control of tomato pinworm in Florida. J. Econ. Entomol. 67:671-674. Pohronezny, K. , V.H. Waddill, W.M. Stall, and W. Dankers. 1978. Inte- grated control of the vegetable leafminer (Liriomyza sativae Blan- chard) during the 1977-78 tomato season in Dade County, Florida. Proc. Fla. State Hort. Soc. 91:322-327. 238 Poole, R.W. 1974. 7^ introduction to quantitative ecology. McGraw- Hill, New York. 532 p. Povolny, D. 1973. Keiferia brunea sp. n. , taxonomic status of the Neo- tropical genera Keiferia Busck and Tildenia Povolny and their economical importance (Lepidoptera: Gelechiidae) . Acta Univ. Agric. Fac. Agron. Brno. 21:603-615. Povolny, D. 1977. On the neotropical species of Gnorimoschemini (Lepi- doptera: Gelechiidae) mining solanaceae. Acta Univ. Agric. Fac. Agron. Brno. 23:379-393. Prada, M. , and P.J. Gutierrez. 1974. Preliminary contribution to the microbiological control of Scrobipalpula absoluta (Meyrick) with Neoplectana carpocapsae Weisser and Bacillus thuringiensis Berl. in tomato Lycopersicon esculentum Mill. Acta Agron. 24:116-137. Price, J.F., and S.L. Poe. 1977. Influence of stake and mulch on lepi- dopterous pests of tomato. Fla. Entomol. 60:173-176. Price, P., and G. Waldbauer. 1975. Ecological aspects of pest manage- ment. In R.L. Metcalf and W.H. Luckman (eds.) , Introduction to insect pest management. John Wiley and Sons, New York. pp. 37-73. Price, P.W. 1976. Colonization of crops by arthropods: non-equilibrium communities in soybean fields. Environ. Entomol. 5:605-611. Purseglove, J.W. 1968. Tropical crops: Dicotyledons. John Wiley and Sons, New York. 2:530-538. Quiroz , E.G. 1976. New information on the biology of the tomato moth, Scrobipalpula absoluta (Meyr.). Agric. Tech. Santiago, Chile 36: 82-86. Ramsany, M. 1980. Developing a sampling plan for estimating the abso- lute population of Stomoxys nigra Macquart (Diptera, Muscidae) in Mauritius. Insect Sci. Appl. 1:133-137. Reynolds, H.T., P.K. Adkinson, and R.F. Smith. 1975. Cotton insect pest management. In_ R.L. Metcalf and W.H. Luckman (eds.). Introduction to insect pest management. John Wiley and Sons, New York. pp. 379-443. Richards, O.W. 1961. The theoretical and practical study of natural in- sect populations. Annu. Rev. Entomol. 6:147-162. Romshe, F.A. 1942. Correlation between fresh weight and area of tomato leaves. Proc. Am. Soc. Hort. Sci. 40:482. Ruesink, W.G. 1980. Introduction to sampling theory. In_ M. Kogan and D.C. Herzog (eds.). Sampling methods in soybean entomology. Spring- er-Verlag, New York. pp. 61-104. 239 Ruesink, W.G,, and M. Kogan. 1975. The quantitative basis of pest man- agement: sampling and measuring. In_ R.L. Metcalf and W.H. Luckman (eds.), Introduction to insect pest management. John Wiley and Sons, Nev7 York. pp. 309-351. Schalk, J.M., and A.K. Stoner. 1976. A bioassay differentiates resis- tance to the Colorado potato beetle on tomatoes. J. Am. Soc. Hort. Sci. 101:74-76. Schoonhoven, L.M. 1972. Plant recognition by lepidopterous larva. In H.F. van Emden (ed.). Insect plant relationships. Symp. R. Ento- mol. Soc. London 6:87-99. Schuster, D.J. 1977a. Resistance in tomato accessions to the tomato pinworm. J. Econ. Entomol. 70:434-436. Schuster, D.J. 1977b. Effect of tomato cultivars on insect damage and chemical control. Fla. Entomol. 60:227-232. Schuster, D.J. 1980. Tomato pinworm: larval survival, development and damage on tomato treated with organotin compounds. J. Econ. Entomol. 73:310-312. Schuster, D.J. 1982. Dipel plus coax for insect control on tomato, 1981. ln_A.C. York (ed.). Insecticide and acaricide tests. Entomol. Soc. Am., College Park, Md. 7:121. Schuster, D. J. , and P.H. Everett. 1982. Insect control on tomatoes in Florida, In^ A.C. York (ed.). Insecticide and acaricide tests. Entomol. Soc. Am., College Park, Md. 7:123. Schuster, D.J., V.H. Waddill, J.J. Augustine, and R.B. Volin. 1979. Field comparisons of Lycopersicon accessions for resistance to the tomato pinworm and vegetable leafminer. J. Am. Soc. Hort. Sci. 104:170-172. Seabrok, W.D. 1977. Insect chemosensory responses to other insects. Iri H.H. Soley and J.J. McKevey, Jr. (eds.). Chemical control of insect behavior. John Wiley and Sons, New York. pp. 15-43. Shelton, A.M., and J. A. Wyman. 1979. Seasonal patterns of potato tiiber- worm moth abundance as determined by pheromone trapping. Environ. Entomol. 8:541-543. Shepard, K. , and G.R. earner. 1976. Distribution of insects in soybean fields. Can. Entomol. 108:767-771. Shiyomi, M. 1976. Generalized spatial patterns in animal populations using a form of the Poisson distribution. Bull. Nat. Inst. Agric . Sci. Ser. A, Tokyo 23:69-83. 240 Simmons, P., and G.W. Ellington. 1933. Life history of the Angoimiois grain moth in Maryland. U.S.D.A. Tech. Bull. No. 351, Washington, B.C. pp. 28-30- Smith, R.F. 1970. Pesticides: their use and limitations. In R.L. Rabb and F.F. Buthrie (eds.). Concept of pest management. N.C. State Univ. Press, Raleigh, pp. 103-118. Snedecor, G.W. , and W.G. Cochran. 1967, Statistical Methods. Iowa State Univ. Press, Ames. 507 p. Sokal, R.R. , and F.J. Rohlf. 1969. Biometry: the principles and prac- tice of statistics in biological research. Freeman and Co., San Francisco. 776 p. Southwood, T.R.E. 1978. Ecological Methods, 2nd ed. John Wiley and Sons, New York. 524 p. Sparks, M.R., and J.S. Cheatham. 1970. Responses of a laboratory strain of the tobacco homworm, Manduca sexta to artificial ovi position sites. J. Econ. Entomol. 63:428-431. Stern, V.M. 1973. Economic thresholds. Annu. Rev. Entomol. 18:259-280. Stem, V.M. 1979. Long and short range dispersal of the pink boll- worm Pectinophora gossypiella over southern California. Environ. Entomol. 8:524-527. Stern, V.M. , P.L. Adkinson, 0. Beingolea G. , and G.A. Viktorov. 1976. Cultural controls. In C.B. Huffaker and P.S. Messenger (eds.). Theory and practice of biological control. Academic Press, New York. pp. 593-613. Stinner, R.E., R.L. Rabb, and J.R. Bradley, Jr. 1977. Natural factors operating in the population dynamics of Heliothis zea in North Carolina. Proc. XV Int. Congr. Entomol. London, pp. 622-642. Stone, J.D., and L.P. Pedigo. 1972. Development and economic injury level of the green cloverworm on soybean in Iowa. J. Econ. Entomol. 55:197-201. Sugimoto, T. 1976. On distribution of egg population of leaf mining fly, Phytomiza ranunculi Schrank (Diptera: Agronyzidae) among leaves and in a leaf. Mem. Fac. Agric. Kinki Univ. 9:11-18. Swank, G.R. 1937. Tomato pinworm (Gnorimoschema lycopersicella Busck) in Florida. Fla. Entomol. 20:33-42. Swesey, O. 1928. Notes on the tomato leafminer, Phthorimaea lycoper- sicella Busck, in Hawaii (Lep.). Proc. Hawaii Entomol. Soc. 7:177. 241 Taylor, L.R. 1961. Aggregation, variance and the mean. Nature 189: 732-735. Taylor, L.R. 1965. A natioral law for the spatial disposition of insects. Proc. XII Int. Congr. Entomol. London, pp. 396-397. Thomas, C.A. 1933. Observations on the tomato pinworm (Gnorimoschema lycopersicella Busck) and the eggplant leaf miner (G. glochinella Zeller) in Pennsylvania. J. Econ. Entomol. 26:137-143. Thompson, H.C. , and W.C. Kelly. 1957. Vegetable crops, 5th ed. McGraw- Hill Co., New York. pp. 471-500. Traynier, R.M. 1975. Field and laboratory experiments on the site of oviposition by the potato moth Phthorimaea operculella (Zell.) (Lepidoptera: Gelechiidae) . Bull. Entomol. Res. 65:395-398. Tsai-Shia, C. 1979. The relations of population dynamics of the army- worm Leucanea separata Walker to relative humidity and rainfall. Acta Entomol. Sin. 22:404-412. van den Bosch, R. , O. Beingolea G. , M. Hafez, and L.A. Falcon. 1976. Biological control of insect pests of row crops. In_ C.B. Huf faker and P.S. Messenger (eds.) , Theory and practice of biological control. Academic Press Inc., New York. pp. 443-456. van Emden, H.F. 1965. The effect of uncultivated land on the distri- bution of cabbage aphid (Brevicome brassica) on an adjacent crop. J. Appl. Ecol. 2:171-176. van Emden, H.F., and G.F. Williams. 1974. Insect stability and di- versity in agroecosystems . Annu. Rev. Entomol. 19:455-475. Van Steenwick, R.A. , G. Ballmer, A.L. Page, T.J. Gauje, and H.T. Reynolds. 1978. Dispersal of Rubidium-marked pink bollworm. Environ. Entomol. 7:608-613, Volin, R.B., and H.H. Bryan, 1976. Flora-Dade : a fresh market tomato with resistance to verticilli\im wilt. Univ. of Fla. Agric. Exp. Sta. Giro. S-246. 9 p. Waddill, V.H. 1975. Tomato pinworm control. Insecticide and acaricide tests. Entomol. Soc . Am. 2:97. Waddill, V.H. 1980. Evaluation of insecticides used on demand for the tomato pinworm, • 1979. Agric. Res. Ed. Cent. Univ. of Fla. 2 p. Ward, C.R., E.R. Mitchell, A.N. Sparks, H. Serrate, and D. Villarroel. 1980. Response of the fall armyworm and other lepidopterous pests of Boliva to synthetic pheromones. Fla. Entomol. 63:151-153. 242 Warren, L.O. 1956. Behavior of Angoijmois grain moth on several strains of com at 2 moisture levels. J. Econ. Entomol. 49:316-319. Watson, J.R., and W.L. Thompson. 1932. Pinworm on tomatoes. Fla. Entomol. 16:14. Weinberg, H.L., and W.H. Lange. 1980. Developmental rate and lower temperature threshold of the tomato pinworm. Environ. Entomol. 9:245-246. Wellik, M.J., J.E. Slosser, and R.D. Kirby. 1979. Evaluation of pro- cedures for sampling Heliothis zea and Keiferia lycopersicella on tomatoes. J. Econ. Entomol. 72:777-780. Wolfenbarger, D.O., J. A. Cornell, S.D. Walker, and D.A. Wolfenberger. 1975. Control and sequential sampling for damage by the tomato pinworm. J. Econ. Entomol. 68:458-460. Wolfenbarger, D.O., and S.L. Poe. 1973. Tomato pinworm control. Proc. Fla. State Hort. Soc. 86:139-143. Wo If son, J.L. 1980. Oviposition response of Pieris rapae to environ- mentally induced variation in Brassica nigra. Entomol. Exp. et Appl. 27:223-232. Wyman, J, A. 1979. Effect of trap design attractant release rate on tomato pinworm catches. J. Econ. Entomol. 72:865-868. Yamamoto, R.T. , R.Y. Jenkins, and R.K. McClusky. 1969. Factors de- termining the selection of plants for oviposition by the tobacco homworm xManduca sexta. Entomol. Exp. et Appl. 12:504-508. APPENDIX EXPLANATORY TABLES FOR CHAPTERS II AND III 244 g d 0 0 c c •H •H £1 ^ 0) c c > n •H 0 0 ■H CO ■P w tn +J M (d tn tn (fl •H cn ■H •H tr O 0) 0 0 (1) ft z eu ft 2 a 245 g g g S s g a Q 0 Q 0 O 0 0 c c c c a c c •H -H •H -H •H •r< •H X} ja ja XI ^ Xt XJ 0) 0) 0) 0) (D o OJ c > a > c > C > fi > fi c > c > 0 ■H 0 •H 0 •H 0 ■H 0 ■H 0 0 •rH o ■H m 4-1 to 4-1 tn +J m J-) M 4J 01 01 4-1 01 4-1 w ■^^ w •H tn lO en -H 01 fd Ul 01 (0 01 (C ■i-{ 0) -H 01 •H en •H 03 -H Di •H •rH en •H cn 0 0 0 0 0 Q) 0 0 0 S ■0 0 0) 0 0) C > C > c > c > c > c > o •H 0 ■H 0 •H 0 -rH o •H 0 ■f^ o ■r^ tn 4J w -P tn -U tn ■P en ■P tn ■P w +J tn td m fl ';i ITS U3 IB en (0 tn (0 tn m H D^ •H tj> •H Cn •H CP ■r-i CP •H en ■H w n (D 0 0) 0 a; 0 0) 0 OJ o (P 0 (1) p. z cu 2 eu 2 cu z ft z ft z ft z 3 .H C •H 4-J iH G •H n S-l c; a 1 < •^\ 247 3 00 C •H 4-1 rH c •H 0 U ( ) a, 1 < 248 g S ^ S e Q 0 o 0 Ci c c c a c •H •rl •H ■H ■H XI XI ^ ,Q ^ 0) m 0) 0) d) c c > c > C > a > G > a 0 •H 0 •H o •H 0 ■H O •H m 01 +J m +J m -P in 4J W 4J it) m (0 m (C m (3 to nj U) Ifl ■H •H &> ■rH CI •H Oi -H CP •W m 0 0 0) 0 0) o 0) 0 0) 0 0) ft cu 2 eu 2 ft 2 ft 2 ft z T3 a,> 3 CO ■H +J rH C •H O U U a, 1 <: fi nJ .H Ui m a 0 u o 0 +J it3 ^ (J -P iH X C) (T\ -p i-H Tl ^ C 0 JJ u* G () •r^ O U) -p in en in w H tji ■r4 0 OJ n tu 2 ft 250 P. S g ri 0 0 0 o c a G G •H •H •H ■H ^ ^ ^ ^ 0) 0) 0) 0) G > c > G > C > 0 •iH 0 •H 0 •H 0 -H Ul 4-) tn -P w +J tn +J m <3 en (d en nj en (rt ■H Cr> ■H 01 •H iji •H m 0 0) 0 OJ 0 G > G > (> ■H 0 ■r^ 0 -rH tn -P tn -P tn -P in cfl tn (tl tn tn -H en •H Oi •H ni n OJ 0 (P 0 (11 tl4 2 A 2 ft 2 251 0 0 c a ■H -H -a ^ 0) (1) fi > C > n •H 0 •H fO ■U tn +J en nj CQ la •H CP ■H tr o (D 0 (1) Ch S a- a 252 o 0 o c c a •H •H -H ^ J2 ^ fi > C > 0 ■W 0 -H 0 ■H tn •4-1 M •P tn -P rn nj M m en Ifl •H Cn ■H cn •H m o i (1 Oh (C u << 2 253 s d R 0 0 o c c a -rH ■H ■H £ ^ XI fi > c: > n o •H 0 -H 0 ■H o 1/1 ■P en +J M +J If) M 03 Cfl (C Ul n3 C(l H cr> ■H Cn •H CP ■H 0 0) 0 (U 0 01 n Pj 2 cu 3 cu 2 ft 0 0 0 c c c •H •iH ■iH ja ^ X! 0) c > > C 0 -H 0 ■H •H o m 4J m -P +J U) U) n3 M (3 05 tn •H cn •H Di 01 ■H n q; 0 lU (D n ft 2 ft a 2 ft 254 S M y ^ 3 H •^ 0 0 o o 0 0 n a a c c c rt a ■H ■H ■H •iH -H -H •H Xi £: X! X! ^ ja X! 0) (U a > G > c; > c > C > c > C > o ■H 0 ■H 0 ■W 0 •H 0 -H 0 ■H 0 -iH U) ■P tn +J w +J m +J Ul +J w -P CQ -P Ui ta tn 10 CO ITJ CO ns ca lO M (C m (TJ -H Di ■H en •H Ol •H Oi •iH en -W Cn ■H tjl 0 (1) 0 (1) 0 0) 0 0) 0 (U 0 0) 0 0) n- 2 eu 2 eu 2 o. s ft z ft 2 ft 2 P CO c ■rH +J c >1 0 rd u s 255 cn o H iH TJ (U 3 iH CO C ■H +J c >i o m u S o LO 0) Lfl tH X) ca Eh 256 p "* c CM ■H +J c u n a, u < M ^ 257 e g S g g 0 0 0 c a a c •.M ■H •rt ■H ^ X! ^ ^ (1) OJ (U 0) a > c > a > c > c 0 •H 0 ■1-1 0 •r4 0 •H f) [0 +J M ■p en +J en -P rn en d Ul fO tn (0 en ra tn ■H CP ■H tjl ■H D> ■H CTi •rH 0 OJ o 1 O nJ u S Cl. • () ■H ^^ CO C) m .H o -M "fc m • e Xi () o H o •-i -P fH C (Ti (1) rri 04 tP 0 + 1 259 g (rt w c n K •H H ■p a) (1) Tl ;5 ft EH 0 260 + 1 + 1 G r O 05 •r-l H +J m -P o CI 0 Qj r-i R 1-1 fl 0 T-l m a) fri > CIJ (]) ■a cu 14-i H 0 + 1 II 2 ^ C a -p r; ^^ W ca ^ en s w + 1 -' 1 X 261 + 1 u o IH (I) rH r On « fc H ffl ai ■P C c (]> o h — 1 r\ 4J 0 (t1 rH tj 0} 0 > r-l m i-H Ti ■rt m rri n m (U a; •^ n3 tij JJ El W 262 3 0 ^ -C CO Wl ai x: ^ a 2 7. +j. w + 1 " 1 X 253 a) £15 r-{ H fi ^ (d +J m fi (1) t: ^, ^j 0 ■r-i r-l 264 IJ « iH H U4 R ^ n3 +J m C fl) >i g ^^ a 3 0 ■r~i iH C 0) •rH > (1) i-l TJ (rt > >w u C (0 H 0) m ;s (« u< -p H CO BIOGRAPHICAL SKETCH Jorge E. -Pena was born on April 8, 1948, in Call, Colombia. He received his high school certificate in 1967 from Colegio Benjamin Herrera in Call, Colombia. He began his undergraduate studies in 1968 at the Universidad Nacional de Colombia, Facultad de Agronomia, and re- ceived the Bachelor of Science degree with a major in agronomy in January 1973. In March 1973, he started working for CIAT (Centre Internacional de Agricultura Tropical) as a research assistant in a cassava entomology program. In 1977 he was awarded a scholarship to pursue a Master of Science degree in entomology at the University of Florida. He graduated in 1979. He is currently a candidate for the degree of Doctor of Philos- ophy in the Department of Entomology and Hematology at the University of Florida. He is a member of Entomological Society of America, Florida Ento- mological Society, Root Crops Society and Sociedad Cclombiana de Entomoloqia. 265 I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree Doctor of Philosophy. ..^- / V.H. Waddill, Chairman" Associate Professor of Entomology and Nematology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree Doctor of Philosophy. J.L. Stimac Associate Professor of Entomology and Nematology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree Doctor of Philosophy. D.J. Schuster Associate Professor of Entomology and Nematology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ■s^. v4/lWxrv_ S.H. Kerr Professor of Entomology and Nematology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. K.L. Pohronezny Assistant Professor of Plant Pathology This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. April 1983 tack X S^ , /College of Agr^ Dean,/Gollege of Agriculture ( Dean for Graduate Studies and Research UNIVERSITY OF FLORIDA 3 1262 08553 4237