Citation
Sampling methods to estimate absolute and relative density of adults and larvae of Liriomyza trifolii (Burgess) on fresh market tomatoes and biology of Liriomyza trifolii under laboratory conditions

Material Information

Title:
Sampling methods to estimate absolute and relative density of adults and larvae of Liriomyza trifolii (Burgess) on fresh market tomatoes and biology of Liriomyza trifolii under laboratory conditions
Creator:
Zoebisch, Tomas Gunter, 1958-
Publication Date:
Language:
English
Physical Description:
xvi, 163 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Eggs ( jstor )
Female animals ( jstor )
Insecticides ( jstor )
Larvae ( jstor )
Leafminers ( jstor )
Leaves ( jstor )
Oviposition ( jstor )
Population estimates ( jstor )
Seasons ( jstor )
Tomatoes ( jstor )
Dissertations Academic -- Entomology and Nematology -- UF
Entomology and Nematology thesis Ph. D
Liriomyza -- Density ( lcsh )
Population density -- Measurement -- Mathematical models ( lcsh )
Tomatoes -- Diseases and pests -- Density ( lcsh )
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1988.
Bibliography:
Includes bibliographical references (leaves 147-162).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Tomas Gunter Zoebisch.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
030416437 ( ALEPH )
20809301 ( OCLC )

Downloads

This item has the following downloads:


Full Text










SAMPLING METHODS TO ESTIMATE ABSOLUTE AND RELATIVE DENSITY OF ADULTS AND LARVAE OF Liriomyza trifolii (Burgess) ON FRESH MARKET TOMATOES AND BIOLOGY OF
Liriomyza trifolii UNDER LABORATORY CONDITIONS













By

TOMAS GUNTER ZOEBISCH


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1988
















ACKNOWLEDGEMENTS


I am very thankful to Dr. D. J. Schuster, chairman of my supervisory committee, for his great friendship, guidance, and financial and invaluable moral support.

I would like to express my gratitute to Dr. J. L.

Stimac for his skillful help in statistical analyses, and his lively humor while he served on my supervisory committee.

My sincerest thanks go to Dr. G. H. Smerage for

providing me with concepts in mathematical models and also with moral support while he served on my supervisory committee.

I would like to thank Dr. S. H. Kerr, for his academic advice and administrative support, and Dr. J. R. Strayer for his moral support and professional guidance, both serving on my supervisory committee.

The technicians at the Gulf Coast Research and

Education Center helped to make my field and laboratory experiments possible. I am particularly thankful to Ken Kiger for his help in establishing and maintaining my experimental plots.

The faculty members at the Gulf Coast Research and Education Center were tremendously supportive to me ii









emotionally. I would like to thank Dr. S. S. Woltz for allowing me to use his personal computer to learn SAS and Dr. J. B. Dr. Kring for motivating me and giving me a good time in the laboratory.

Dr. J. Allen gave me valuable advice on data analyses to whom I grateful express my thanks.

I also appreciate very much the friendship of Dr. K. J. Patel and his wife, Falguni, who helped me emotionally to finish my degree.

Mr. G. Zoebisch supported me financially, making my coming to the U.S. possible. I am grateful for his financial support.

Mr. and Mrs. Hof also supported me morally and financially during my hardest times. Their help is gratefully acknowledged.

I thank my parents for their great love, moral support, encouragement, and financial support under all circumstances.

Church members from the Church of Jesus Christ and Latter Day Saints in Bradenton and Gainesville always supported me and my family with great blessings. I appreciate very much their help.

Most of all, I would like to express my thanks and love to my beautiful wife and daughter, whose moral and emotional support enabled me to finish my dissertation.


iii















TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS...............................** ii

LIST OF TABLES ..................******* ****** **** * viii

LIST OF FIGURES............. *.* * *............... xii

ABSTRACT...................*..................... xiv

CHAPTER

I INTRODUCTION...............*........*....... 1

II LITERATURE REVIEW ........................... 6

Systematics and Distinguishing
Characteristics of Liriomyza
trifolii.......... . .. ...... .......... 6
Life Cycle and Biology of L.
trifolii..... .... ..................... 11
Biology and Behavior of
Adults............. ............ ........ . 11
Oviposition and Egg
Developmental Time....................... 14
Larval Development........................ 16
Pupal Developmental Time................ 17
Geographical Distribution of
L. trifolii ... ..... ................. .... 17
Alternative Hosts of L. trifolii............ 20
Crop Damage and Effects on Yield............ 21
Reduction in Crop Yield................... 21
Entry of Pathogens and Physiological
Changes due to Larval Mining.............. 23
Reduction of Photosynthetic
Rates....... .................. . 23
Chemical Control of L. trifolii
on Tomato........... .................... .. 24
Resistance of Leafminers to
Insecticides Used in Tomatoes......... . 27
Effects of Insecticides on
Parasitoids of Liriomyza
Leafminers.................. .......... 29









Pae

Biological Control of Liriomyza
trifolii................................. 32
Parasitoids of Liriomyza
trifolii on Tomatoes...................... 32
Predators of Liriomyza
Leafminers................................. 33
Effects of Cultural Methods on
Populations of Liriomyza
Leafminers................................. 35
Sampling Methods Used to Monitor
Populations of Liriomyza
Leafminers............................... 37
Methods to Estimate Larval
Densities................................. 37
Methods to Estimate Adult
Densities................................ 40

III PLANT PRODUCTION AND MAINTENANCE
OF L. trifolii COLONIES...................... 45

Introduction........................ ...... 45
Tomato Plant Production..................... 46
Screenhouse Maintenance .................... 47
Establishment and Maintenance
of the L. trifolii Colony................. 47

IV PREPARATION OF EXPERIMENTAL PLOTS
AND CONSTRUCTION OF FIELD CAGES
TO SAMPLE L. trifolii ADULT
POPULATIONS................................. 50

Introduction................................ 50
Preparation of Study Plots.................. 50
Transplanting Procedures ................. 51
Staking and Tying Procedures.............. 52
Pest, Disease and Weed Control............ 52
Construction of Field Cages to
Sample Adult L. trifolii.................. 53

V ESTIMATION OF ADULT DENSITIES AND
CALIBRATION OF A RELATIVE SAMPLING
METHOD FOR ESTIMATION OF ADULT
POPULATION DENSITY OF L. trifolii........... 56

Introduction........... .................... 56
Materials and Methods....................... 57
Sampling Method on the Unknown
Field Population ........................... 57
Sampling Method on the Known
Leafminer Population..................... 59








Pae


Influence of Sticky Card
Position on Leafminer Catches............ 61
Determination fo Insecticidal
Effects on Leafminers Treated
Inside Field Cages...................... 61
Results and Discussion ...................... 62
Generation of the Calibration
Equations on the Known Population........ 62
Generation of the Calibration
Equations on the Field Population........ 63
Generation of Calibration Equations
to Estimate Absolute Densities Based
on Relative Sampling Data............... 64
Validation of Predicted Adult
Leafminer Densities....................... 65
Proportion and Number of Females and
Males Collected on the Sticky Cards
During Relative and Absolute Sampling
Procedures................................ 69
Sampling at Three Different
Heights................................... 82
Response of Flies Exposed to Pyrethrum
Insecticide in the Laboratory............. 85

VI ESTIMATES OF FIRST-SECOND AND
THIRD STAGE LARVAL DENSITIES OF
Liriomyza trifolii ON TOMATOES.............. 87

Introduction................................ 87
Materials and Methods....................... 90
Results and Discussion....................... 92

VII FEMALE SURVIVAL, OVIPOSITION, EGG AND
LARVAL DEVELOPMENT OF Liriomyza trifolii
ON TOMATO FOLIAGE................ ......... 104

Introduction................................ 104
Materials and Methods....................... 106
Planting of Host plants................... 106
Determination of Adult Female
Longevity, Adult Female Survival, Oviposition Rate, Fertility, and
Egg Development Rate................... 106
Determination of Larval
Development .............. .... ......... 107
Equations Used to Describe the
Biological Processes Studied in
This Chapter.......................... 108









Page

Results and Discussion ...................... 109
Adult Female Longevity and
Survival, Fertility, and Egg
Developmental Rate...................... 109
Larval Development........................ 121
Pupal Development......................... . 132
Comparison of the Regression
Equations................................. 133
Proposal of a Conceptual
Model on Population Dynamics
of L. trifolii.......................... 134

VIII CONCLUSIONS........................................ 142

LITERATURE CITED......................... ............. 147

BIOGRAPHICAL SKETCH.................................. 163


vii














LIST OF TABLES


Pae


Table 2-1. Table 2-2. Table 3-1. Table 5-1. Table 5-2. Table 5-3.


Larval development time of L. trifolii at similar temperatures in foliage of different hosts.............

Pupal development time of L. trifolii at similar temperatures from larvae that developed in foliage of different hosts................

Insecticides and fungicides applied to control insect pests and fungal diseases on experimental plots during the fall 1986 and spring spring 1987 tomato growing seasons at the Gulf Coast Research and Education Center ......................*

Proportion of females/males collected on the sticky cards on the unknown population (inside and outside cages) and on the known population inside the cages during the fall 1986 season......... . ............................

Proportion of females/males collected on the sticky cards on the unknown population (inside and outside cages) and on the known population inside the cages during the spring 1986 season....................... ...........

Mean number of flies of the unknown population outside and inside cages and known population inside cages collected in the south (sunny) and north (shady) side of the tomato field rows sampled during the fall 1986 season ......................................


viii










Table 5-4. Table 5-5. Table 5-6. Table 5-7. Table 5-8. Table 5-9. Table 5-10. Table 5-11.


Mean number of flies of the unknown population outside and inside cages and known population inside cages collected in the south and north side of the tomato field rows sampled during the spring 1987 season .......................****.............

Mean number of females and males of unknown population outside and inside cages collected in the south (sunny) and north (shady) side of the tomato field rows sampled during the fall 1986 season ......................

Mean number of females and males of known population inside cages collected in the south (sunny) and north (shady) side of the tomato field rows sampled during the fall 1986 season ...................................

Mean number of females and males of unknown population outside and inside cages collected in the south and north side of the tomato field rows sampled during the spring 1987 season....................................

Mean number of females and males of known population inside cages collected in the south and north side of the tomato field rows sampled during the spring 1987 season . . .....................................

Average laboratory-reared females collected on yellow sticky cards when feeding and mating status varied with a total of 20 females per cage.................................

Mean proportion of female L. trifolii (female/female+male) collected on sticky cards at three heights during the spring 1987 season ..............................

Mean number of females collected on yellow sticky cards at three different heights during the spring 1987 season ... .. ...............................


Page










Table 5-12. Table 5-13. Table 5-14. Table 5-15. Table 6-1. Table 6-2. Table 6-3. Table 6-4. Table 6-5.


Mean number of males collected on yellow sticky cards at three different heights during the spring 1987 season.................... *..*

Mean number of total flies collected on yellow sticky cards at three different heights during the spring 1987 season ...............................

Response of laboratory-reared L. trifolii PrentoxR-treated lumite screen after 24 h ..................**

Response of laboratory-reared L. trifolii Prentox -treated lumite screen after 48 h .........................

Regression analysis results with an R2 >0.50 on relative and absolute sampling procedures on large larvae sampled on the fourth, seventh and tenth nodes on lateral stems on six randomly selected plants..................

Regression analysis results with an R2 >0.50 on relative and absolute sampling procedures on small larvae sampled on the fourth, seventh and tenth nodes on lateral stems on six randomly selected plants..................

Regression analysis results with an R2 >0.50 on relative and absolute sampling procedures on large and small larvae sampled on the fourth, seventh and tenth nodes on four lateral stems and main stem of six randomly selected plants.........................

Percentage of sampling occasions in which the regression analysis yielded an R2 > 0.50 out of twelve sampling dates on large (L) and small (S) larvae on main stems............

Percentage of sampling occasions in which the regression analysis yielded an R2 > 0.50 out of twelve sampling dates on large (L) and small
(S) larvae on four lateral stems..........


Pace











Table 6-6.






Table 7-1. Table 7-2. Table 7-3. Table 7-4. Table 7-5. Table 7-6. Table 7-7. Table 7-8. Table 7-9.


Percentage of sampling occasions in which the regression analysis yielded an R2 > 0.50 out of twelve sampling dates on large (L) and small (S) larvae on four lateral stems and main stem ......................

Mean adult female longevity (days) at four constant temperatures.............

Mean total eggs laid per female throughout its adult lifetime at four constant temperatures................

Magnitude and time of peak oviposition rate of L. trifolii at four constant temperatures under laboratory conditions...............

Percent of L. trifolii larvae that hatched at four constant temperatures.....

Mean egg developmental time (days) at four constant temperatures.............

R2 values obtained on linear, quadratic and exponential regression equations from the biological processes studied at four constant temperatures (13.9, 20, 25 and 32 *C) ..................

Linear equations for temperaturedependent-development of Liriomyza trifolii pupae ............................

Slope (1/K) and threshold temperature estimates (To) obtained from linear regression equations that describe some of the biological processes of L. trifolii studied in this chapter.........

Mean and variance of the time of biological processes at four constant temperatures (13.9, 20, 25, and 32 C .....


Pacie


99 109



116




117 118 118





122 133





140 141















LIST OF FIGURES


Pa-9e


Fig. 5-1. Fig. 5-2.


Fig. 6-1.





Fig. 7-1. Fig. 7-2. Fig. 7-3.



Fig. 7-4. Fig. 7-5. Fig. 7-6.


Mean � 95% C.I. adult flies of L. trifolii collected and predicted on a per two plant basis during the fall 1986 season............. ......................

Mean � 95% C.I. adult flies of L. trifolii collected and predicted on a per two plant basis during the spring 1987 season ...........................................

Mean total � 95% C.I. L. trifolii larvae (small and large) collected on the seventh node on four lateral stems on six plants per sampling date ......................................

Adult female longevity (in days) at four constant temperatures 13.9, 20, 25 and 32 *C)...................

Adult female survival rate (1/day) at four constant temperatures 13.9, 20, 25 and 32 'C)...............

Mean number of eggs laid per day per female at four constant temperatures .............................

Mean number of eggs laid per day per female at (13.9, 20, 25 and 32 *C ) throughout its lifetime...........

Total eggs laid per female at four constant temperatures (13.9, 20, 25 and 32 *C ) ...............................

Mean egg developmental time at four constant temperatures (13.9, 20, 25 and 30 *C).........................


xii


101 110 111



112 114 115 119











Fig. 7-7.


Mean egg developmental rate at four constant temperatures (13.9, 20, 25 and 32 *C).........................


Fig. 7-8. Larval developmental time at four
constant temperatures (13.9, 20,
25 and 32 �C) .............................

Fig. 7-9. Small larvae developmental time at
four constant temperatures (13.9,
20, 25, and 32 �C) ........................


Fig.


7-10.


Fig. 7-11. Fig. 7-12. Fig. 7-13.


Large larvae developmental time at four constant temperatures (13.9, 20, 25, and 32 *C) ........................

Larval developmental rate (1/day) at four constant temperatures (13.9, 20, 25, and 32 �C).......................

Larval developmental rate (1/day) transformed to Napierian logarithms) at four constant temperatures (13.9, 20, 25, and 32 *C).......................

Developmental rate (1/day) of small larvae at four constant temperatures (13.9, 20, 25, and 32 *C) .................


Fig. 7-14. Developmental rate (1/day) of small
larvae (transformed to Napierian
logarithms) at four constant
temperatures (13.9, 20, 25, and 32 *C)....

Fig. 7-15. Developmental rate (1/hour) of large
larvae at four constant temperatures
(13.9, 20, 25, and 32 �C) .................

Fig. 7-16. Developmental rate (1/day) of large
larvae (transformed to Napierian
logarithms) at four constant
temperatures (13.9, 20, 25, and 32 �C)....

Fig. 7-17. Diagram of a conceptual model of
Liriomyza trifolii on tomatoes............

Fig. 7-18. Representative probability density
function of development time in
a life stage..............................


120 123 124



125 126 127 128 129



130 131 136 138


xiii














Abstract of Dissertation Presented to the Graduate
School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SAMPLING METHODS TO DETERMINE ABSOLUTE AND RELATIVE DENSITY OF ADULTS AND LARVAE OF Liriomyza trifolii
(Burgess) ON FRESH MARKET TOMATOES AND BIOLOGY OF
Liriomyza trifolii UNDER LABORATORY CONDITIONS By

Tomas Gunter Zoebisch

December 1988

Chairman : D. J. Schuster
Major Department: Entomology and Nematology

A sampling method for estimating absolute adult

leafminer populations on a per-two-plant basis using yellow sticky cards as a relative sampling method was developed under field conditions. The following equations for estimating adult leafminer densities were developed: Fd=

8.352+0.182Fac and Sd=9.765+0.485Sc where Fd and Sd are the absolute average densities estimated for the fall and spring seasons, respectively, and Fa and Sac are the mean number of flies collected on sticky cards during the fall and spring seasons, respectively.

Relative sampling methods applied to L. trifolii larvae were not reliable through time. Regression equations that relate small larvae (<1.5 mm long) sampled on six plants with the total larvae/six plants yielded an R2 value >


xiv








0.50 in 9 out of 12 sampling dates when data sampling the 4th, 7th, and 10th nodes on four randomly selected lateral stems per plant (relative sample) were related to the total number of larvae per plant (absolute sample).

Several biological processes of L. trifolii, female adult longevity and survival rate, oviposition, egg developmental rate, and small (< 1.5 mm) and large ( 1.5 mm) larval developmental rates were numerically described using linear, quadratic and exponential regression equations. Based on regression coefficients (R2), most of the biological processes were described well by linear equations. Only total number of eggs laid per female and large larval developmental rate were better described by quadratic and exponential equations, respectively. Large larval developmental rate was well described using two linear equations over predefined temperature ranges.

A conceptual population dynamics model was developed for estimating subsequent larval populations from observed adult populations and in response to temperature in the absence of mortality and migration. The information generated using the sampling method developed herein for estimating absolute adult leafminer densities would be an input to the model. Descriptions of biological processes in the model are based on parameters for rates of oviposition and development of immatures established in this research. The objective in developing a population dynamics model of








L. trifolii is to predict the dynamics of larval populations from adult populations for timely treatment to prevent economic losses.


xvi














CHAPTER I
INTRODUCTION


The leafminer fly, Liriomyza trifolii (Burgess)

(Diptera: Agromyzidae), has become an important pest in Florida horticultural and vegetable crops, including tomato, celery and chrysanthemum (Stegmaier 1966). Due to the frequent use of insecticides, L. trifolii has developed resistance to most organochlorine and organophosphate insecticides used in vegetable crops. Leafminers have been recurring pests in Florida since the late 1940s due to the wide and general use of DDT (Wolfenbarger 1958). Although leafminer populations can be controlled by natural enemies (Pohronezny and Waddill 1978), the use of broad spectrum insecticides to control primary pests on tomatoes has allowed leafminer populations to increase to damaging density levels. As a consequence, the use of broad spectrum insecticides had to be reduced as much as possible, and implementation of other, mainly biological, control strategies became necessary. Therefore, integrated pest management (IPM) programs for tomatoes have been developed in Florida since 1978 (Pohronezny and Waddill 1978). The






2

main objective of these programs has been to develop economically, technically and ecologically sound systems of integrated pest management.

Leafminers (Liriomyza spp.) attack the foliage of toamtoes (a non-marketable part of the plant) and are, therefore, classified as indirect pests (Ruesink and Kogan 1975). Unfortunately, the high value of tomato crops reduces the options of pest management tactics (Bottrell 1979), and chemical control is predominant to prevent fruit damage by other insects (Lange and Bronson 1981). Implementation of IPM programs for tomatoes in Florida has allowed growers to maintain competitive prices (e.q., in comparison to West Mexican producers during the winter season) by reducing production costs. This reduction has been achieved by using hybrid varieties such as 'FTE-12', 'Duke', and 'Sunny', and fewer pesticide applications (Van Sickle and Belibasis 1985). In 1984/85, pesticide costs ranged from $453.43 to $501.96 per acre (Van Sickle and Belibasis 1985). In 1986/87, they ranged from $472.15 to $511.72 per acre (Taylor and Smith 1987).

Advanced research and applications in IPM increasingly employ systems analyses. Three major objectives of system analyses in IPM are: 1) to establish comprehensive information about agroecosystem composition, which includes the crop, pest and beneficial organisms, cultural practices, weather, and management; 2) to predict agroecosystem responses to specific environmantal and man-imposed inputs






3

as well as behavioral features of these systems; and 3) to select optimal management strategies for crop development and control of pest populations for economical production of crops with high yield and quality (Smerage et al. (1980). Models are essential for expressing system compositions, and mathematical models are used for analyzing system behavior. Mathematical models permit prediction of pest population densities under various environmental and management conditions.

Descriptions of biological processes determining the dynamics of pest populations must be known to generate appropriate mathematical models. Data obtained under field and laboratory conditions are used to develop descritpions of those processes such as development, oviposition, parasitism, predation, movement, and insecticide mortality caused by insecticides (Smerage et al. 1980).

Sampling techniques have to be developed for generating information on field populations in order to formulate process descriptions and to provide input to population dynamics models. Absolute population densities (i.e. on a unit area or volume basis) are required for process descriptions and input to models. Present sampling techniques (which yield relative population density estimates) for L. trifolii are inadequate to estimate those absolute population densities. Current IPM programs utilize relative sampling methods (i.e. number of insects collected on a per trap basis, or per swing net collection, etc.) to






4

monitor a pest for treatment decisions at the lowest cost possible. Since absolute sampling methods under practical conditions are expensive and time consuming, numerical relationships between absolute and relative sampling methods are needed for generating absolute densities from relative estimates for population dynamics models (Luna et al. 1982). The hope is that a reliable numerical relationship can be established between relative sampling and absolute densities for development of valid mathematical models.

Adult leafminers are attracted to yellow surfaces

(Affeldt et al. 1983), and yellow sticky cards have been used to sample adult leafminers. However, no relationships (relative or absolute) have been established between yellow sticky card catches and adult leafminer population density. Estimated absolute adult populations could be used to predict and control subsequent larval populations in time to prevent economic damage.

The objectives of my studies were as follows:

1) To develop a sampling technique for estimating absolute

adult leafminer density in fields of staked tomatoes. 2) To establish a numerical relationship between yellow

sticky card catches (relative method of for sampling

leafminer adult population) and the estimated absolute

density.

3) To establish similar procedures on small (first and

second instar) and large (third instar) leafminer

larvae.






5

4) To determine parameters of basic population processes

such as oviposition, egg development, larval

development, and adult survival under laboratory

conditions, to generate useful information for

developing a population dynamics model.













CHAPTER II
LITERATURE REVIEW

Systematics and Distinguishing Characteristics of Liriomyza trifolii

The genus Liriomyza (from Greek "lirion" (lily) and "myza" (suck)) was established by Mik in 1894, when he described a leafminer from lilies (Lilium martagon L.) as L. urophorina. The basal segment of the female ovipositor of this species is nearly as long as the abdomen (Oatman and Michelbacher 1958). Liriomyza is a very homogenous group, differing little in coloration and structure from the type specimens (Frick 1952) which makes the species in this genus difficult to identify. Males in this genus possess a stridulating organ which consists of a chitinized ridge on the hind femora and a line of scales on the sides of the abdomen (Spencer 1984).

Liriomyza trifolii was first described as Oscinis trifolii (in the family Oscinidae, now Agromyzidae) by Burgess in 1880 from foliage of white clover (Trifolium repens L.) and has since been confused and synonymized with several agromyzid fly species. The first misidentification was made by Coquillett in 1898 when he studied specimens reared from potato at Foristell, Montana and specimens reared from white clover in Washington, D.C. He identified






7

his specimens as Agromyza diminuta Walker and stated that these were identical to those described by Burgess (1880) (as Q. trifolii) and by Riley (1884) (as Q. brassicae). Thus, 0. trifolii was placed in the genus Aqromyza as A. trifolii. Aldrich (1905) stated that the synonymy created by Coquillett (1898) was unrecognizable.

In 1913, Malloch synonymized 0. trifolii and 0.

brassicae with the European species Aqromyza pusilla Meigen and doubtfully included Coquillett's (1898) A. diminuta in his synonymy (Spencer 1981b). In the same year Melander stated that A. diminuta was synonymized by Coquillett (1898) with 0. trifolii (which Melander (1913) considered as A. scutellata) and that Coquillett's description was too brief to allow for proper identification. He placed A. diminuta as a synonym of A. scutellata (which belongs to the genus Metopomyza (Spencer 1981b)).

0. trifolii and 0. brassicae were still treated as
synonyms of A. pusilla by Frost (1924) who stated that this latter species was a "decidedly variable species" (p. 51).

De Meijere (1925) described the larva of L.

lequminosarum and stated that this species might have been identical to Burgess' (1880) Q. trifolii (=A. trifolii). He rightly pointed out that the name trifolii was not applicable because "Kaltenbach (1874) had earlier cited A. trifolii". Hendel (1938) also noticed this homonymy but stated that this homonymy "was very probable, but not very certain" (p. 214). Spencer (1981b) considered A. trifolii a






8
synonym of A. nana (Meigen). Frick (1953) placed A. trifolii as a synonym of L. congesta and stated that A. trifolii was not a homonym of Kaltenbach's (1874) A. trifolii, creating a secondary homonym and therefore a new synonym. He also created a new synonym by placing L. trifolii as a synonym of L. congesta. In addition, he mentioned that, since the type of A. trifolii did not exist in the U.S. National Museum collection and that it seemed probable that it was the same species that Burgess described, his synonymy would remain valid until the status of A. trifolii would be clarified. A name once placed in homonymy must be rejected, however (article 35 of the International Commision on Zoological Nomenclature (Schenk and McMasters 1936)). Therefore, a new name is needed for the economically important leafminer fly, Liriomyza trifolii (Zoebisch 1984).

Frick (1955) described a new species of Liriomyza as L. alliovora (named the Iowa onion miner) which he synonymized with Liriomyza allia (Frost). Frost (1962) described L. archboldi collected at the Archbold Biological Station in Highlands County, Florida and mentioned that this species was similar to L. trifolii (Burgess). Spencer (1965), after studying the male genitalia of Frost's L. archboldi concluded that it was the same species as L. trifolii. He also discussed the homonymy of Kaltenbach's A. trifolii and the Q. trifolii described by Burgess (1880), and made a wrong conclusion stating that these were not homonyms.






9

Presently much of the confusion in the correct

identification of Liriomyza leafminers has been eliminated by the detailed studies of the male genitalia (Spencer 1981a) and by the use of starch gel electrophoresis techniques to correctly diagnose closely related species (Zehnder et al. 1983). The same technique has proven useful to separate larvae of Liriomyza species (Menken and Ulenberg 1983).

Spencer (1981b) listed two characters that made L.

trifolii adults easily distinguishable from a very closely related species, L. sativae Blanchard. First, the color of the mesonotum of L. trifolii is grayish-black and matte; second, the upper orbits and most of the hind-margin of the eye is yellow, with both vertical bristles on yellow ground. The first characteristic, however, may become altered depending on the method of preservation. K. A. Spencer (Dept. Biol. Sci., Univ. Exeter, England) observed that the mesonotum of specimens of L. trifolii preserved in alcohol may become shiny, acquiring a similar appearance to the mesonota of L. sativae. The shiny mesonotum of L. sativae was considered by Spencer (1981b) the most conspicuous difference to separate L. trifolii from L. sativae. The hind-margin of the eye of L. sativae is entirely black and the vertical bristle is on black background. Even the male genitalia are similar in L. trifolii and L. sativae, the curvature of the genitalia of L. sativae being less pronounced than in L. trifolii (Spencer 1981a). Knodel-






10

Montz and Poe (1982) studied the female genitalia of L. sativae and L. trifolii and concluded that the egg guide was V-shaped in L. trifolii whereas the egg guide of L. sativae was acutely angled. The denticles in the former species are angular while in the latter they are elongate.

The description of the adult is based on a neotype

designated by Spencer (1965) from a male reared from alfalfa (Medicago sativa L.) in Indiana and is stored in the U.S. National Museum. Spencer (1965) described it as follows:

Orbits entirely yellow, both vertical bristles on
yellow ground; black of occiput reaching margin beyond
outer vertical bristle; all antennal segments bright
yellow,third only finely pubescent; mesonotum blackish grey, distinctly polinose; acrostichals irregularly in
3 or 4 rows in from, reduced to two rows behind, yellow
patch at each corner adjoining scutellum; mesopleuron
with black patch normally extending along lower margin,
sternopleura largely black, upper margin yellow;
abdomen with tergites variably yellow laterally and on hind margins; coxae yellow, femora largely so but with
slight, variable brownish striation; tibiae and tarsi
darker, brown. (p. 37-38).

Genitalia are illustrated in Spencer (1973).

Larvae of L. trifolii as well as those from the family Agromyzidae can be distinguished by the presence of a pair of anterior spiracles that is located dorsally in contrast to the lateral position that these have in other cyclorrhaphous Diptera (Peterson 1979). Larvae and pupae of L. trifolii and L. sativae are difficult to separate by conventional techniques. Menken and Ulenberg (1986) characterized the larvae and pupae of these species using starch gel electrophoresis.






11

The eggs of L. trifolii are microscopic, oval, creamy white, and measure approximately 0.2 x 0.1 mm (Bartlett and Powell 1981). Eggs of L. trifolii probably increase in size similar to those of L. congesta (as L. trifolii) (Dimetry 1971) and L. sativae (as L. pusilla) (Tilden (1950) after being deposited.

Life Cycle and Biology of L. trifolii Biology and Behavior of Adults

Adult L. trifolii are small flies with a body length of ca. 1.30 mm and a wing length of 1.5 mm (Burgess (1880)). They emerge with the aid of their ptilinum through the dorsal anterior end of the puparium (Parrella 1987). Adult flies usually emerge during early morning hours primarily between 0930 h and 1230 h showing a pronounced diurnal periodicity (Charlton and Allen 1981, Vercambre 1980). Nevertheless, L. trifolii adults have been collected in light traps (Spencer 1965, Stegmaier 1966).

Adult emergence depends in part on relative humidity, the type of substrate, and depth of burial in which the larva pupates. Charlton and Allen (1981) studied adult emergence under various humidity conditions. Very few adults (6%) emerged from newly transformed pupae completely exposed at a low relative humidity (11%), while at 100% RH, 88% adults emerged. In another experiment they buried pupae about one cm in sand and peat varying the amounts of water added by soil weight. They observed that maximum emergence occurred at 11.4% and 27% added water by weight in sand and






12

peat, respectively. They also determined that at 25 *C up to 96% of the puparia would survive after being submerged in water for 4 hours whereas none would survive after being submerged for 75 h. Taking additionally the factors of depth burial and soil grain size into account, Oetting (1983) determined that the pupal survival was reduced with increasing gravel size and depth of burial in gravel.

Adult flies begin activity at sunrise and are most

active during midmorning hours (Parrella et al. 1981). At 26.7�0.5 *C Parrella et al. (1983b) observed that females oviposited in chrysanthemum leaves within 24 h after being placed in association with males, which indicates that mating also occurs shortly after adult emergence. Mating and oviposition may be influenced by several factors one of which is temperature (chapter VII).

Males of L. trifolii are considered polygamous (Bodri and Oetting 1985). Oatman and Michelbacher (1958) observed multiple matings in L. Dictella (now synonymized with L. sativae (Spencer 1968).

Feeding and oviposition behavior of adult female

leafminers has been studied quantitatively by Bethke and Parrella (1985). They described the sequence and the probability of events that occurred once the female flies started puncturing the leaf of a host (chrysanthemum or tomato) with their ovipositor to feed on the exudates of the ruptured cells. Their observations can be summarized as follows: upon initiation of a puncture or 'stipple' the






13

ovipositor is positioned perpendicularly and touching the leaf. It is first thrust rapidly and then more slowly, the change in thrust speed initiating the penetration of the leaf surface. In the majority of punctures, a female moves her abdomen from side to side to create during the slower ovipositor thrusting larger, fan-shaped leaf punctures. When the abdomen is not twisted from side to side, the ovipositor penetrates the leaf surface in a straight back direction to create a smaller tubular-shaped stipple. Females lay eggs in tubular punctures. Occasionally (16% and 21% of the times in chrysanthemum and tomato, respectively) an egg is deposited, almost always in the tubular punctures. After a puncture is completed, the female backs over the puncture to feed on the exudates of the macerated leaf cells. Both puncture types are used for feeding.

Musgrave et al. (1978) reported that males of L.

sativae do not possess structures to puncture host leaves and they stated that males may feed at wounds produced by the females. Bodri and Oetting (1985) stated that males also fed on plant exudates on leaf punctures made by female flies but did not present any data to support this. Zoebisch and Schuster (1987a) did not find increased survival of males provided access to these exudates compared to males not provided access to exudates. Thus, males apparently do not feed on puncture exudates.






14

Oviposition and Eqq Developmental Time

Eggs are laid singly and parallel to the surface of a leaf in the leaf mesophyll in the smaller, tubular punctures. On chrysanthemums lower leaves are preferred for oviposition while younger leaves are preferred for feeding (Knodel-Montz et al. 1983).

Oviposition of L. trifolii varies from host to host and is greatly influenced by the presence of additional carbohydrate sources. Chandler and Gilstrap (1986a) determined in the laboratory that mated L. trifolii oviposited an average of 17.9�28.8 (Mean�SD) on bell peppers during their whole life span at a constant temperature of 24
*C. Leibee (1984) studied the oviposition rate of L. trifolii in celery and concluded that oviposition was highest at temperatures that ranged from 25 *C (total of 159 eggs per female) to 35 *C ( total of 118 eggs per female). He provided the flies with honey, however. Parrella (1984) studied the oviposition rate of L. trifolii on chrysanthemum at five constant temperatures and determined that from 21.1
*C to 32.2 *C the total number of eggs during the whole life span of the flies was highest (188.53-233.91 eggs/female) and that these numbers did not differ statistically. He also provided the flies with honey. Care should be taken when comparing oviposition rates on different hosts since in many studies the flies were provided with honey or other carbohydrate sources. Charlton and Allen (1981) found that honey significantly increased fecundity of L. trifolii.






15

Numerous weeds have been found to be suitable hosts for L. trifolii. Smith and Hardman (1986) concluded that L. trifolii successfully oviposited on 16 species of weeds common in or around greenhouses in Nova Scotia, Canada. The number of eggs/100lcm of foliage ranged from 1 on Glecoma hederacea L. (Labiatae) (least suitable host) to 32 on Solanum dulcamara L. (Solanaceae). Zoebisch and Schuster (1987a) determined the fecundity and ovipositional preference of L. trifolii on tomato and three weed species and found that tomato and nightshade (Solanum americanum Mill.) where overall the most suitable hosts for oviposition. Tomato-reared flies oviposited an average of 34.6 and 31.8 eggs during their whole life span at 22-26 �C on tomato and nightshade foliage, respectively.

Natural sources of carbohydrates such as aphid honeydew could be utilized by adult L. trifolii. Zoebisch and Schuster (1987a) determined in the laboratory that aphid honeydew significantly increased oviposition of L. trifolii on tomato foliage even when mated females were exposed to this carbohydrate source for only 24 h.

The egg developmental time of L. trifolii is influenced primarily by temperature and not host plant species. Charlton and Allen (1981) reported that the egg developmental time on pink beans ranged from 2 days (at 32.5

*C) to 11.2 days (at 13.8 *C). In celery, egg developmental times range from 1.99 days (35 *C) to 9.97 days (15 *C) (Leibee 1984). In bell peppers the mean egg developmental






16

time is 4.2�0.1 (Mean�SD) days at a constant temperature of 24 *C (Chandler and Gilstrap 1986a). In tomato cv. 'Hayslip' foliage the mean egg developmental times ranged from 1.99 (32 *C) to 11.5 (13.9 *C) days (chapter VII). Larval Development

Upon maturity the first stage larva punctures the

chorion of the anterior pole of the egg with its mouthhooks to exit the egg (Dimetry 1971). The larva also exerts pressure to the eggshell to split it at the anterior end and completes development undergoing four molts (Parrella 1987). Larvae of L. trifolii feed on the leaf's mesophyll moving in one direction creating a linear or ophionome mine (Hering 1951) leaving a trail of fecal material that alternates from one side to another.

Larval development, unlike egg development, is strongly influenced by both temperature and host plant (Table 2-1).

Upon completing their development, larvae of L.

trifolii exit their mines by cutting a crescent-shaped slit close to or at the apex of the mine and dropping to the ground to seek a dark place to pupate (Leibee 1986). Larvae exit their mines during early daylight hours (Charlton and Allen 1981). Pupation seems to be delayed for only a limited amount of time regardless of lighting conditions (Leibee 1986).








Pupal Developmental Time

In contrast to the differences in developmental times

observed in larvae of L. trifolii, pupal developmental times are similar regardless of the host in which the larvae developed (Table 2-2).

Equations to calculate developmental rates of pupae

from larvae that developed in different hosts are listed in Parrella (1987). Developmental threshold temperatures for puparia are slightly higher than those of the larvae (Table 2-2).

In Italy Suss et al. (1984) observed puparia diapausing under laboratory conditions at temperatures below 16 *C.

Geographical Distribution of L. trifolii

L. trifolii is one of several species of agromyzid leafminers that has been spread by man (Frick 1952). Spencer (1968) stated that the genus Liriomyza is of Holarctic origin. L. trifolii is of Nearctic origin (Spencer 1981b). Due to the spread by man, L. trifolii is now considered a cosmopolitan species (Minkenberg and van Lenteren 1986) and it has been recorded in North and South America, Europe, Africa, and Asia. It is able to survive in areas with severe periods of sub-zero temperatures, but it survives best in subtropical and tropical regions (Spencer 1973). In America the northern limit of L. trifolii is Ontario (Spencer and Stegmaier 1973) and its distribution ranges as far south as Colombia (Poe and Montz 1981). Patel (1987) summarized the geographical distribution of L.











Table 2-1. Larval development time of L. trifolii at similar
temperatures in foliage of different hosts.

Temperature Host plant Development Citation
(�C) time (days)


celery


5.36 6.77 7.97 11.98
threshold#


chrysanthemum


Leibee (1984)


Bodri and Oetting (1985)


7.00 7.60 11.57 13.67


pink beans


32.5 30.0 25.0 20.0
N/A@


tomato


32.2 26.7 21.1 7.8


Charlton and Allen (1981)

4.50 5.10 4.70 8.00 threshold

Schuster and Patel (1985)

3.50 4.40 7.10
threshold


35.0 30.0 25.0
20.0 8.4


35.0 30.0 25.0
20.0


#Estimated threshold temperature for development (*C). @Datum not available.








Table 2-2. Pupal developmental time of L. trifolii at similar
temperatures from larvae that developed in foliage of
different hosts.

Temperature Host plant Development Citation
(C) time (days)

celery Leibee (1984) 35.0 6.69 30.0 6.76 25.0 8.37
20.0 13.45 10.3 threshold@

chrysanthemum Parrella et al. (1981) 32.2 7.25 26.7 9.15
21.1 14.15
9.0 threshold

pink beans Charlton and Allen (1981)

32.5 6.90 30.0 6.80 25.0 8.50 11.0 threshold

@Estimated threshold tempertaure for development (*C) using linear regression equations. #Supposed development threshold temperature.






20

trifolii. In addition to the countries included in his list, L. trifolii has possibly been reported in Puerto Rico (Pdrez 1974).

The rapid spread of L. trifolii worldwide is primarily due to failure of quarantine procedures (Parrella and Keil 1984) because of the difficulty in detecting the small oviposition punctures.

Alternative Hosts of L. trifolii

Spencer (1964) and Stegmaier (1966) consider L.

trifolii a polyphagous species. It attacks several crops and has been reported in 148 host plant species in 31 families (Patel 1987).

Schuster et al. (1982) surveyed weeds associated with tomato fields in Hillsborough County, Florida. They found that the predominant flora on the perimeters of these fields was different and changed throughout the season. In the spring season of 1982 almost 50% of the leafminer larvae were observed in foliage of ground cherry (Physalis sp.) during the month of February, while most of the larvae (ca. 90%) were collected from black nightshade (Solanum nigrum L.) until the end of that season. Throughout the growing season 80-95% of Liriomyza larvae were observed in foliage of black nightshade, ground cherry, and Spanish needle (Bidens bipinnata L.). These authors concluded that weeds might serve as sources of low numbers of Liriomyza spp. in tomatoes. Wolfenbarger (1961) concluded the same when he surveyed weeds associated with tomato and potato fields in






21

Florida. Zoebisch and Schuster (1987a) determined in the laboratory that downy ground cherry (Physalis pubescens L.) was the least suitable host when compared to tomato, American black nightshade (Solanum americanum Mill.) and common beggar-tick (Bidens alba L. (DC)).

Stegmaier (1966) and Wolfenbarger (1961) believed that weeds in Florida acted as reservoirs of L. trifolii throughout the year, thus maintaining continuous populations while crops were not in cultivation.

Crop Damage and Effects on Yield

Parrella (1987) summarized six ways in which a crop might be damaged by Liriomyza spp.: 1) reduction in the aesthetic value of ornamental plants, 2) destruction of young seedlings, 3) reduction of crop yields, 4) transmission of diseases, 5) acceleration of leaf abcission causing scalding of fruit, and 6) causing some plant species to be quarantined.

In tomatoes reductions in crop yield, entry of leaf

pathogens through mines, defoliation, and reduction in leaf photosynthetic rates have been reported. Reduction in Crop Yield

Wolfenbarger (1948) reported in an insecticidal trial that a 'serpentine leaf miner' (possibly L. trifolii or L. sativae) was the most serious pest among other pests like banded cucumber beetles (Diabrotica balteata) Lec., southern armyworm (Spodoptera eridania (Cram.), and tomato hornworms






22
(Manduca sexta (Haw.)). Plots sprayed with parathion yielded an average of 60.2 kg compared to the control plots (13.7 kg/plot).
Jones and Kelsheimer (1963) reported that insecticides applied for control of lepidopterous larvae and leafminers on tomatoes in Florida affected yields. Plots treated with dimethoate yielded more than plots treated with parathion or azinphos-methyl. They reported that leafminer and larval insecticides affected yields although they did not report any apparent insect infestation on the plants.

Poe (1974) indicated that the southern armyworm (SpodoDtera exiqua (Hiibner)) and the tomato fruitworm (Heliothis zea (Boddie)) were consistently the more damaging pests of staked tomatoes in Florida. He observed L. sativae in his experimental plots but did not ascribe any yield losses to this insect. Levins et al. (1975) studied the effect of diazinon on populations of L. sativae and effects on yield in tomatoes and concluded that leafminers did not reduce yields at the densities present.

Effects on yield in tomatoes due to leafminer damage

were studied during four seasons in Florida by Schuster and Jones (1976). Yield (measured as weight of fruit) was significantly increased over the check plots of 'Walter' tomatoes only in one season when leafminer populations were treated with oxamyl.






23

Entry of Pathogens and Physiological Changes due to Larval Mining

Keularts (1980) studied the effects of leafmining by L. trifolii on tomato as a source of entry for pathogenic organisms. He successfully recovered the pathogens Alternaria alternata (Fries) Keisler and Xanthomonas vesicatoria (Doidge) Dows) from leafmines. These pathogens can cause major defoliation on tomatoes in Florida.

Schuster (1978) reported that up to 90% of the foliage on tomato may be lost, if leafminers are not controlled. Fruit thus exposed to the sun may suffer scalding and moisture loss (Musgrave et al. 1975b). Reduction of Photosynthetic Rates

Johnson et al. (1983) determined that photosynthetic rates in mined tissues of tomatoes (hybrid 6718 VF) were reduced 62% by the feeding damage exerted by L. sativae larvae. They concluded that mining injury primarily affected gas exchange thus reducing stomatal conductance in unmined leaf tissues close to the mine. They further suggested that more studies were needed to understand the relationship between leafminer injury, photosynthetic rates, and fruit production physiology in tomatoes.

On lima beans Martens and Trumble (1987) found that mature leaves mined by L. trifolii larvae virtually recovered from injury. Damaged cells were replaced by photosynthetically active cells.






24

Chemical Control of L. trifolii on Tomato

Until 1982 there has been confusion regarding the identity of Liriomyza leafminers studied on tomatoes. Leibee (1981) provided an excellent review of the insecticidal control of Liriomyza leafminers on vegetables until 1981. He concluded that the short effective life of insecticides and the adverse effects that these compounds had on the leafminer parasite species complex had contributed to several outbreaks of leafminers in vegetable crops.

Powell (1981) treated L. trifolii populations in

England in greenhouse tomatoes with heptenophos, oxamyl, permethrin, and resmethrin in an effort to eradicate this pest. Tomatoes grown outside greenhouses were treated with DDT and trichlorfon. No tests of insecticide efficacy were performed at that time. Bartlett and Powell (1981) concluded that greenhouse populations of L. trifolii in England could be eradicated but that field populations provided a constant threat of reintroduction. Schuster and Everett (1983a) reported on the effectiveness of abamectin, SD 52618 (5,6-dihydro-2-(aci-nitromethyl)-4H-l,3-thiazine, calcium salt (2:1)), cypermethrin, fenvalerate, encapsulated methyl parathion, chlorpyriphos, methamidophos, and the pyrethroid AC 222,705 against L. trifolii in the field and in the laboratory. They concluded that abamectin and cyromazine were effective in controlling larvae in the field and in the laboratory. Low adult mortality was observed






25

although abamectin reduced oviposition and feeding. In the laboratory they observed that shortly before or upon emergence, larvae died in foliage treated with abamectin. Of the other compounds none provided satisfactory control and encapsulated methyl parathion at 119.8 g AI/liter resulted in more leafmines relative to the water check.

In another field experiment Schuster and Everett

(1983a) collected significantly fewer puparia on leaves treated with cyromazine and abamectin (except in plots treated with 2.27 g ai/378.5 1). As many puparia as those collected from the control plots sprayed with water were collected where insecticides such as methamidophos, azadirachtin, fenvalerate, and the pyrethroid Pay-OffR were used.

In California, Trumble (1983) reported that the insecticides permethrin, fenvalerate, BactospeineR, abamectin, and cypermethrin significantly reduced leafminer (L. trifolii and L. sativae) densities when compared to the control plots. L. sativae may have predominated over L. trifolii and may have been controlled controlled by a wider range of insecticides, since L. trifolii was not satisfactorily controlled on tomatoes with fenvalerate in Florida (Schuster and Everett 1983b). Parrella and Keil (1985) compared the toxicity to methamidophos in four species of leafminers and concluded that L. trifolii reared






26

from Florida celery was the most tolerant species when compared to the other leafminer species including L. sativae.

Woets (1985) mentioned that L. trifolii was difficult to control in British greenhouses due to its tolerance to many common insecticides. In addition, the disruption of biological control of the greenhouse whitefly by insecticides applied for control of L. trifolii made management of leafminers more difficult. Aldicarb, oxamyl, dimethoate, and heptenophos are among the insecticides listed by this author for the control of L. trifolii.

In France, Martin (1984) and Martin and Filliol (1985) reported that fenthion, methomyl, methidathion, dichlorvos and isathrine were used on a prophylactic basis to control L. trifolii by growers in that country.

Of the insecticides currently under development against many pests, abamectin, azadirachtin, and cyromazine are the most promising for the satisfactory control of L. trifolii. The active ingredients of abamectin consist of a mixture of avermectin B1a (minimum 80%) and avermectin Bb (maximum 20%) (Brown and Dybas 1982), a macrocyclic lactone isolated from the soil microorganism Streptomyces avermitilis (Burg et al. 1979). It inhibits gamma aminobutyric acid-mediated neuromuscular transmission. The low number of feeding stipples and eggs observed by Schuster and Everett (1983a) in females treated with abamectin may be due to the inhibition of the oviposition functions.






27

Azadirachtin, natural product obtained from seeds of

the neem tree (Azadirachta indica Juss.), acts as a feeding inhibitor and/or as an insect growth regulator against many insect species (Warthen 1979). Although Schuster and Everett (1983a) reported that a water extract of neem seed was not as effective as abamectin against L. trifolii on tomatoes. Webb et al. (1983) determined that significantly fewer eggs were laid by L. trifolii on leaves of 'Henderson Bush' lima beans treated with a water extraction of neem seed. They also reported that larval mortality reached 100% shortly after the eggs hatched. Systemic effects of neem seed extract on L. trifolii larvae were reported in chrysanthemum by Larew et al. (1985).

Cyromazine (N-cyclopropyl-l,3,5-triazine-2,4,6triamine) has a mode of action similar to that of an insect growth regulator and possibly acts as a chitin inhibitor (Trumble 1985a). Larvae treated with cyromazine fail to pupate (Schuster and Everett 1983b). Resistance of Leafminers to Insecticides Used in Tomatoes

In 1957 Genung reported the ineffectiveness of

toxaphene for controlling the serpentine leafminer on tomatoes cv. 'Hayslip'. Wolfenbarger (1958) concluded that chlordane, lindane, toxaphene, and aldrin had become ineffective for leafminer control on tomatoes and potatoes in south Florida. Brogdon (1961) stated that leafminers were a more severe and constant problem in southern Florida than in the central and northern parts of the state. At






28
that time growers were seeking label approval for the insecticide naled to control leafminers on tomatoes.

The insecticides azinphos methyl, permethrin, and Lorsban did not significantly reduce the number of leafmines on tomatoes in experimental plots in Bradenton in 1980 (Schuster and Everett 1981).

Schuster and Everett (1983b) determined that under

field conditions the insecticides cypermethrin, fenvalerate, microencapsulated parathion, and chlorpyrifos were ineffective against L. trifolii in the Bradenton, Florida area.

Using standard probit analysis methods Parrella and

Keil (1985) determined that L. trifolii adults from Florida celery were the most tolerant to repeated applications of methamidophos (LD, = 10.8 mg/ml, slope = 3.09) when compared to L. trifolii adults collected on chrysanthemum in California (LDW = 1.93 mg/ml, slope = 2.17) where methamidophos is not used.

Resistance of L. trifolii to insecticides has been

monitored using yellow sticky cards. Haynes et al. (1986) used these cards with permethrin and chlorpyrifos incorporated at desired concentrations in the sticky material (TangletrapR) in chrysanthemum greenhouses in California. They determined that this method was reliable, simple, and accurate to evaluate insecticide resistance of adult leafminers.






29

Effects of Insecticides on Parasitoids of Liriomyza Leafminers

The impact that insecticides have on leafminer

parasitoids varies among crops and geographical areas. Since the use of organochlorine insecticides, increases in leafminer populations have been recorded due to the apparent ability of the leafminers to develop tolerance and/or resistance to synthetic organic compounds. Leafminers have been classified as secondary pests (Pohronezny and Waddill 1978) which means that whenever their natural enemies are eliminated through the use of broad spectrum insecticides their population densities may increase to economically damaging levels. Insecticides can interact with parasites in several ways: 1) have no effect, which means that parasitoids are not affected and would still control the treated pest population; 2) have broad spectrum toxicity which means that pest and natural enemy populations are affected, and 3) have selective toxicity which means that certain natural enemies are effected.

Getzin (1960) suggested that leafminer chemical control methods should be combined with biological control methods in which only leafminers are controlled. Chemicals that would allow such pest control strategies have not been available until recently. Parrella et al. (1983a) determined under laboratory conditions that the insect growth regulators cyromazine and Ro 13-5223 had no effect on survival of Chrysonotomyia parksi, and still 80% of L.








trifolii was controlled. The use of broad spectrum insecticides has been the most common in tomato crops due to the high value of the crop (Bottrell 1979).

In relation to the use of broad spectrum insecticides

Shorey and Hall (1963) and Getzin (1960) reported increasing leafminer populations in areas treated with DDT and methoxychlor, dieldrin, and lindane. With the wide use of organophosphate insecticides, leafminer parasitoids responded similarly as to the organochlorine compounds. Oatman and Kennedy (1976) and Johnson et al. (1980a) realized the deleterious effects that methomyl had on leafminer parasitoid populations. They observed significantly higher leafminer densities on tomato plots treated with methomyl. Oxamyl was observed to have similar effects when applied weekly on tomatoes for leafminer control (Schuster et al. 1979). The problems arising with these kinds of compounds rested in their broad spectral activity.

Regarding selective toxicities, under certain

circumstances, insecticides have been shown to affect some leafminer parasitoid species more than others. Poe et al. (1978) determined that most leafminer parasitoids were reared from tomato foliage treated with a mixture of leptophos and endosulfan in contrast to foliage treated with oxamyl, which yielded the lowest number of parasitoids. They also reported D. intermedius as the most abundant parasitoid in Florida tomatoes.






31

A biotic larvicide (Bacillus thuringiensis Berliner

var. kurstaki) and an ovicide (chlordimeform) were found to be more specific for the control of lepidopterous pests on tomatoes, although the leafminer parasite Chrysonotomyia punctiventris was also adversely affected by chlordimeform (Johnson et al. 1980a). Zehnder and Trumble (1985a) determined that a higher percentage of Chrysonotomyia punctiventris emerged from organophosphate-treated celery leaf samples while Diqlyphus spp. were more abundant in foliage treated with the pyrethroid permethrin. These trends may vary from one season to another, however. Trumble (1985a) observed that the insecticide abamectin altered the composition of parasitoid species in celery, reducing populations of D. intermedius in 1982, but not in 1983. Schuster and Price (1985) reported that C. punctiventris was more abundant in sprayed tomato plots than ODius sp. and D. intermedius in nonsprayed plots. Schuster (1985) found that Chrysonotomyia was more abundant in tomatoes sprayed with permethrin and methamidophos while Ovius sp. and D. intermedius were found primarily in nontreated tomatoes. This indicates that more detailed studies are needed to determine the effects of insecticides on leafminer parasitoids in several crops under various environmental conditions such as greenhouses and open fields.






32

Biological Control of Liriomyza trifolii Parasitoids of Liriomyza trifolii on Tomatoes

Johnson and Hara (1987) presented a comprehensive list of hymenopterous parasitoids of Liriomyza leafminers in North America including Hawaii. Of the 37 parasitoid species listed, 7 were reported attacking Liriomyza on tomato: DiqlvDyhus begini (Ashmead) (Eulophidae), Chrysocharis Darksi Crawford (Eulophidae), and Chrysonotomyia punctiventris (Crawford) (Eulophidae) in California; D. pulchripes and Ogius dimidiatus Ashmead (Braconidae) in Ohio; Chrysonotomyia formosa and D. intermedius (Girault) in Florida; and C. punctiventris and D. begini in Hawaii.

Minkenberg and van Lenteren (1986) provided a

comprehensive list of the identified parasitoids of L. trifolii reported worldwide. They listed a total of 28 species and mentioned that the biology of only a few was known to some extent.

The predominant parasitoids of L. trifolii may change
from one season to another or one crop to another. Schuster (1985) reported that in West Central Florida, ODius sp. and Chrysonotomyia sp. predominated in 1980 and D. intermedius predominated in 1981. Chrysonotomyia sp. again predominated in 1983 and 1984. Zehnder and Trumble (1984a) found that significantly fewer Chrysocharis parksi Crawford were reared from L. trifolii in a celery field adjacent to a tomato






33

field where L. sativae predominated. D. intermedius was more abundant in the celery field where L. trifolii predominated.

Classical biological control of L. trifolii provides a fertile field of study. Due to the polyphagous nature of this leafminer species, a thorough knowledge of the leafminer's alternative hosts/crops and the ability of its parasitoids to search for larvae under varied conditions is needed. Before any kind of releases will be effective, preferences of the parasitoids for L. trifolii in different crop habitats must be understood. Johnson and Hara (1987) found that the major parasitoid species of major Liriomyza spp. in North America and Hawaii have been found consistently within a major crop. Also parasitoid diversity may vary when the major leafminer species differ between adjacent crops. Zehnder and Trumble (1984a) found that significantly fewer Chrysocharis parksi Crawford were reared from L. trifolii in a celery field adjacent to a tomato field where L. sativae predominated. P. intermedius was more abundant in the celery field where L. trifolii predominated.

Predators of Liriomyza Leafminers

Predators of Liriomyza spp. have been given less emphasis in their role as agents regulating leafminer populations although they have been reported as early as 1913 by Webster and Parks. They listed a mite species in the genus Erythraeus and a hemipteran (Triphleps sp.) which






34
preyed on Agromyza pusilla Meigen. Webster and Parks (1913) were working with L. sativae, L. trifolii, L. huidobrensis, and other undetermined species and not A. pusilla, however (Spencer 1981b). Hemiptera in the genus Nabis and probably also in the genus Geocoris have been reported to prey on larvae of L. sativae on celery (Genung et al. 1978). Johnson et al. (1980b) observed that green lacewing larvae (Neuroptera: Chrysopidae) preyed on L. sativae larvae collected in pupal trays in fresh market tomatoes in Irvine, California. Chrysopid larvae have also been observed preying on Liriomyza spp. larvae in California (Trumble and Nakakihara 1983). In Colombia, Prieto and Chacon de Ulloa (1980) observed two predators of adult L. trifolii on chrysanthemum (a Diptera: Dolichopodidae and an arachnid in the family Oxyopidae). They observed a third predator (a small ant in the subfamily Ponerinae) which attacked larvae that had just exited the mines. Freidberg and Gijswijt (1983) observed adult Drapetis subaenescens (Collin) (Diptera: Empididae) feeding in and around greenhouses on adult L. trifolii in Israel. Under laboratory conditions they observed that this empidid lived one month and consumed between 16 and 20 adult L. trifolii per individual predator. Another empidid (Tachydromia annulata Fallen) was observed to prey on L. trifolii adults and it preferred leafminers over leafminer parasitoids. These authors suggested that






35

these predators might suppress leafminer populations satisfactorily but their biology was still unknown to mass rear them.

Effects of Cultural Methods on Populations of Liriomyza Leafminers

Webster and Parks (1913) correctly pointed out that the 'serpentine leafminer' was controlled by natural enemies (parasitoids) in the U.S., and that these had prevented this leafminer from becoming destructively abundant. The first cultural control method suggested by them was cutting an alfalfa crop for hay once damage was observed in the plants' foliage. They also proposed deep fall plowing in the East arid regions to bury the leafminer pupae deeply enough to prevent adult emergence. For the western area they suggested controlling the weeds along ditch banks and uncultivated fields to diminish the pupal population which would subsequently hibernate. Musgrave et al. (1976) determined that stripping and trimming celery plants reduced total mine numbers of L. sativae from 90-100% in experimental and commercial plots in Florida.

The destruction of crop residues as a means of

controlling populations of Liriomyza spp. in Florida to prevent population buildups in fields newly planted with vegetable crops was concurrently suggested by Adlerz (1961), Brogdon (1961), Kelsheimer (1961), and Wolfenbarger (1961). Wolfenbarger (1961) stated that this was only a partial solution to the leafminer problem in Florida, however.






36

Broad-leaved weeds were believed to serve as reservoirs for pests, including L. sativae, initially invading fields planted to vegetable and ornamental crops in Florida (Musgrave et al. 1975b). Therefore, proper sanitation and destruction of any crop residues and broad-leaved weeds would prevent or delay leafminer migration to newly cultivated fields. Vercambre (1980) suggested similar cultural practices in Reunion for preventing L. trifolii from attacking newly planted fields.

Wolfenbarger and Moore (1968) found significantly fewer mines of Liriomyza sp. in foliage of tomato plants protected by vertical strips of aluminum than in plants protected by mulches made of plasticized, and paper-backed aluminum in South Florida. They also found significantly fewer mines in cotyledons of squash plants protected with aluminum foil and aluminum scrap when compared to the unprotected (check) plants. Chalfant et al. (1977) also reported fewer L. sativae on brown-paper-mulched yellow summer squash compared to non-mulched squash in Georgia. Contrary to these results, Webb and Smith (1973) found most adults of L. sativae in foliage of snap beans mulched with aluminum foil. Their explanation of these results was based on the demonstration of Oatman and Michelbacher (1958) that leafminers exhibited positive phototaxis and would therefore be attracted to reflective mulches. Price and Poe (1976) observed that polyethylene plastic paper mulched tomatoes






37

and staked tomatoes supported the largest leafminer populations when compared to only mulched or staked or nonmulched/non-staked tomatoes. A parasite (Opius sp.) was collected in significantly lower numbers from leafminer pupae collected from staked plants than from non-staked plants which may explain in part the lower densities of leafminers observed in the non-staked plots.

Fertilizer levels also influence leafminer damage. Harbaugh et al. (1983) found that damage by L. trifolii increased linearly as leaf nitrogen in chrysanthemum cv. 'Manatee Yellow Iceberg' increased from 2.2% to 4.0%. They also indicated that N was the most critical factor correlated with leafminer damage although caution should be taken interpreting the N effect, because the controlledrelease fertilizer formulations result in changes in P and K with any change in N.

Sampling Methods Used to Monitor Populations of Liriomyza Leafminers

Methods to Estimate Larval Densities

Larval densities have been estimated by different

methods to determine the effectiveness of insecticides or to study the population dynamics of Liriomyza leafminers in several crops. Oatman and Michelbacher (1958) estimated larval populations of the melon leafminer (L. sativae; misidentified as L. pictella (Thomson) (Spencer 1981a)) by taking random samples consisting of 20 to 50 leaves from the center of melon plants. They concluded that leafminer populations varied considerably from host to host, field to






38

field, and even within areas in the same field. Highest leafminer larval densities were found along the edges of the fields during the months of July and August in the San Joaquin Valley in California.

Wolfenbarger and Wolfenbarger (1966) established a

sequential sampling program to determine the effectiveness of a leafminer control procedure on tomatoes in South Florida. They proposed decision lines to spray or not to spray when 40% and 10% of the leaves sampled at random from a plant averaged one or more mines per leaflet. They considered that the leafminer had been adequately controlled if 10% of the leaves sampled averaged one or had one or more mines. A level of 40% or more of the leaves bearing mines was considered as inadequate control.

In celery, Musgrave et al. (1979) determined that random samples of 100 mature petioles per 4.86 ha were statistically precise enough to plot population trends of L. sativae larvae. Larval densities were estimated counting the maggots in 10 mature trifoliolates per sample per plot.

Schuster and Beck (1983) developed a rating system to estimate densities of total leafmines on tomatoes. This rating system correlated closely with actual counts and reduced the time required by field scouts to assess densities of leafmines later during the tomato growing seasons in west central Florida.

Populations of L. trifolii larvae have been monitored in celery using leaflet samples and counting puparia after






39

3, 7, and 14 days (Foster 1986). Based on indices of dispersion he determined that the leafminer population had an aggregated distribution in commercial celery fields. Jones and Parrella (1986a) developed binomial sampling plans on chrysanthemum in two greenhouses and determined that a 100-leaf random sample per 2000 m2 was appropriate to estimate larval densities of L. trifolii. Johnson et al. (1980b) developed a sampling method based on a linear regression relating the number of live medium and large larvae of L. sativae collected on a per leaflet basis (independent variable) and pupae collected on styrofoam trays placed underneath double-row plotted tomato plants (dependent variable). The coefficient of determination obtained in this numerical relationship was 0.767. This sampling method has not been found useable under Florida field conditions (Schuster, D.J. pers. comm.).

In snap beans cv. 'Nemasnap' and 'Eagle' Hanna et al. (1987) developed a sequential sampling plan counting leafmines present on 10 randomly selected trifoliate leaves in each of four plots. To obtain desired sampling precision levels they related the number of leaves required to sample (dependent variable), and the number of leafmines of L. sativae per leaf (independent variable). They determined that this sampling plan was robust for type of cultivar, nitrogen fertilizer, and a range of pesticide application frequencies.






40

Chandler and Gilstrap (1986b) determined with a

stratified sampling method that L. trifolii larvae on bell peppers after plants had reached 75 mm in height, mature leaves should be sampled to detect the greatest proportion of the larval populations.

Methods to Estimate Adult Densities

Some of the first sampling methods of adult leafminers were developed by Musgrave et al. (1979). They determined that 24 sweep samples taken at random in celery per 4.86 ha provided precise estimates of population trends of L. sativae adults. Adult samples consisted of 10 sweeps with a 38.1 cm diameter sweep net over 6.1 m - 7.6 m of celery. They correctly indicated that the development of sampling procedures for pests offered the key to the judicious use of pest management tactics. Price (1982) reported that some chrysanthemum growers in Colombia sampled adult L. trifolii using sweep nets and other growers used D-Vac suction machines to monitor adult L. trifolii populations. A D-Vac suction machine was also used by Trumble and Nakakihara (1983) to monitor adult leafminers in celery.

Adult leafminer flies are attracted to wavelengths in

the green region (500-540 nm) and the yellow region (540-600 nm) (Affeldt et al. 1983). These intervals correspond to the maximum reflectance of green leaf plants (Shull 1929). This information has permitted the development of traps consisting of a surface of the proper reflectance coated with sticky substances.






41

Tryon et al. (1980) collected significantly more adult flies on yellow cards than on yellow-green, orange, green and blue cards. Results obtained by Yudin et al. (1987) on lettuce in Hawaiian farms agree with these results.

Musgrave et al. (1975b) stated that adult L. sativae could be detected by trapping them on 7.6 cm X 12.7 cm bright yellow cards coated with some sticky material. They suggested that the number of adults captured on several cards after 24 h would indicate their relative abundance in the field and that samples should be taken at least weekly. Early detection and monitoring of adult leafminers could therefore lead to improved population management through the precise application of control measures. Musgrave et al. (1975a) sampled L. sativae in plots with 17 vegetable garden varieties in north Florida. Yellow sticky card counts indicated that adult leafminers were distributed randomly in the field. Weather conditions tended to influence sticky card counts, particularly windy or rainy periods which reduced the number of individuals collected. No relationships were established between adult counts on yellow cards and subsequent larval population densities.

Yellow cards covered with polybutanate have been used in England to monitor adult L. trifolii populations in some nurseries. Powell (1979) indicated that the frequent presence of both Phytomyza sp. and L. trifolii leafminers in greenhouses precluded the confirmation of L. trifolii by examining the foliage alone. He also mentioned the






42

difficulty in detecting low densities of L. trifolii using yellow sticky cards. Despite this, these cards proved to be useful when infestations were slight or the foliage of the greenhouse crops was dense and leafminer larvae were therefore difficult to detect (Powell 1981).

Dispersal of adult L. sativae has been studied using yellow traps. Tryon et al. (1980) collected significantly more flies on cards located on the periphery of a transplant production range nearest the prevailing wind and within 34 m of a commercial tomato farm. Their results indicated that L. sativae moved in the direction of prevailing winds and for relatively short distances. Numbers of trapped flies declined when wind gusts during daylight hours were greater than 32 km/h. Jones and Parrella (1986b) determined that in a chrysanthemum greenhouse the average distance flown by female L. trifolii from a known point of release was greater (21.5 m) than that flown by males (18.0 m).

The shape of the sticky cards (square, rectangular,

triangular, and circular) had no influence on the number of leafminers (L. trifolii and L. sativae) collected in different cultivars such as alfalfa, bell pepper, cantaloup, and greenhouse grown chrysanthemums (Chandler 1981). He collected more males on all trap shapes, and yellow opaque traps attracted significantly more flies than yellow fluorescent translucent cards. Zehnder and Trumble (1984b) also reported that a greater proportion of male L. trifolii and L. sativae were caught on yellow sticky traps in fresh






43

market tomatoes. They suggested that these biased catches were due to female flies spending more time on leaves during oviposition and that males tended to visit more leaves in search of food and females. Jones and Parrella (1986b) captured significantly more male than female L. trifolii on yellow sticky cards in a chrysanthemum greenhouse. Webb et al. (1985) conducted several experiments under greenhouse conditions to determine the response of L. trifolii and L. sativae to yellow sticky cards. When no plants were present, there were no signs of a sex bias for capture of either species. Nevertheless, they collected a larger proportion of L. trifolii females during 4 of 10 trapping periods in a commercial chrysanthemum greenhouse in Maryland. Chandler (1985) concluded that both sexes of L. trifolii appeared equally responsive to yellow cards placed at different heights in bell peppers. With the exception of 13 of 130 instances, no significant differences between males and females collected on the yellow cards were noted. Where differences were found, more males than females were captured, particularly when trap catches peaked.

Sticky cards placed horizontally at different heights in staked tomatoes in Florida have not revealed a spatial preference of Liriomyza spp. leafminers for a certain plant stratum (Schuster and Beck 1981). With vertically placed traps, L. trifolii preferred lower plant heights in fresh market tomatoes in California (Zehnder and Trumble 1984b) and bell peppers in Texas (Chandler 1985).






44

Sequential sampling plans for monitoring Liriomyza spp. adults have been developed on greenhouse chrysanthemums (Parrella and Jones 1985) and fresh market tomatoes (Zehnder and Trumble 1985b).













CHAPTER III
PLANT PRODUCTION AND MAINTENANCE OF COLONIES
OF Liriomyza trifolii (Burgess) Introduction

Methods for rearing Liriomyza spp. leafminers have been described by several authors. The techniques used to maintain L. trifolii colonies to obtain adults for field and laboratory experiments were similar to those used by Ketzler and Price (1982) and Patel (1987). Due to the polyphagous nature of L. trifolii, several host plants in the families Compositae, Leguminosae and Solanaceae can be used to rear them in high numbers. Due to the suitabilty of tomato for feeding and oviposition of L. trifolii (Zoebisch and Schuster 1987b), colonies were maintained using tomatoes cv. 'Hayslip'. Although other hosts like American nightshade (Solanum americanum Mill.) (Zoebisch and Schuster 1987b) or pink beans (Phaseolus vulgaris L.) (Charlton and Allen 1981) could have been used to rear high numbers of L. trifolii, tomatoes were used as host plants because field and laboratory experiments were done on tomato plants.

Approximately 1000-1500 L. trifolii adults were

produced daily using the rearing procedures described in this chapter.






46

Tomato Plant Production

Seeds obtained from tomatoes cv. "Hayslip' at the Gulf Coast Research and Education Center were planted in SpeedlingR (Speedling Inc., Sun City, Florida) inverted pyramid cellular styrofoam trays. Each cell was filled with SpeedlingR Peat-Lite Mix and 3 to 6 seeds were placed about

0.5 cm deep in each cell. Two trays were planted in a screenhouse each Wednesday or Thursday to obtain enough plants to maintain leafminer colonies. Depending on ambient temperature seed germination took 3 to 5 days. During cold days in winter, warm air was blown in the screenhouse with a liquid propane heater to prevent drops in temperature that significantly may have affected plant growth. During hot days in late spring and summer, a fan was used to circulate the air inside the screenhouse to prevent the temperature from becoming too high ( 37.78 *C). Temperature as well as relative humidity fluctuated considerably in the screenhouse (up to � 15.56 *C and �65% RH, respectively).

Seedling trays were watered at least once a day. In addition, plants were fertilized using a HyponexR every Tuesday and Friday with 379 g of 20-20-20 (N, P, K) NutriLeafR (Miller Chemical & Fertilizer Co., Hannover, Pennsylvania) fertilizer in 10 1 of water to give a 500 ppm N solution. Every Friday 9.29 g/l of hydrated magnesium sulfate was added to the fertilizer mixture. Once the seedlings had the first two true leaves they were removed from the styrofoam trays and carefully separated and






47

repotted in white 15.2 cm plastic pots. Four seedlings were placed per pot if the plants were used for colony maintenance and only one seedling per pot if they were used in leafminer development studies. Potted plants were maintained similarly as seedlings.

Screenhouse Maintenance

Weeds growing underneath benches were removed every two weeks to keep plants free of pests such as armyworms (Spodoptera spp.), and russett mites (Aculops lycopersici (Massee)). Armyworms and leafminers appeared at such low densities that infested tomato leaflets were removed manually. Plants infested with russet mites (primarily during the spring 1987 season) were discarded and the benches holding these plants were drenched with the acaricide dicofol. Treated benches were not used for two weeks.

Plant diseases such as leafmold became a problem by the end of the fall 1986 season and 3.875 1 of the fungicide Bravo 500 were applied weekly at 3.87 cc/l until runoff to control this disease.

Establishment and Maintenance of the L. trifolii Colony

Adult flies from a colony previously established for

ca. 4 years at the Gulf Coast Research and Education Center (GCREC) were used to initiate a new colony. Infested tomato foliage in experimental fields at the GCREC was collected in August 1986 and placed in Tupperware Superseal 30 cm x 30 cm x 12 cm plastic storage containers to collect puparia.






48

These containers had one 6.5 cm diameter organdy clothcovered ventilation hole in each sidewall and two similar holes on the lid. A 0.8 cm mesh hardware cloth elevated 2 cm above the bottom of the plastic containers was used to obtain newly emerged larvae. The bottoms of the containers were coated with a thin layer of Teflon (Fluon AD-1; Northeast Chemical Co., Inc.) to prevent puparia from sticking to the surface. Once adults emerged from these puparia they were added to the colony. This procedure was done when the leafminer colony was increased at the beginning and six months later to maintain a genetically diversified population.

Four 7-8 week-old plants were placed in each of two 61 cm x 61 cm x 61 cm oviposition cages (BioquipR cat. no. 1452D) equipped with a stockinette sleeve to exchange the plants (every day) and introduce newly emerged adult flies (ca. 200 every other day). During the fall 1986 season new plants and flies were placed in cages between 0800 h and 0900 h while during the spring 1987 season these procedures were effected between 1600 h and 1800 h. Oviposition cages as well as plants infested with eggs and larvae were kept in a rearing facility maintained between 18 �C and 22 *C. Relative humidity fluctuated between 40% and 100% despite the constant use of an electric humidifier. A 12L:12D photoperiod was maintained with white fluorescent lights.

After ca. 6 days (when the larvae were about to exit their mines) the foliage of each tomato plant was clipped






49

and placed in a TupperwareR plastic storage container to obtain puparia. Containers used for the collection of puparia were kept under the same conditions as above. Puparia were collected with a no. 2 camel hair brush 48 to 72 h later to allow the pupal skin to harden and therefore prevent damaging puparia during recollection. About 100-200 puparia were placed in 30 g clear plastic cups. A filter paper strip streaked with honey was placed through the lid of these cups to provide adult leafminers with a carbohydrate source. Adults used for the colony were transferred directly into the oviposition cages located in the rearing room. Care was taken not to release too many adults into the oviposition cages to avoid intraspecific competition among larvae (Parrella 1983).













CHAPTER IV
PREPARATION OF EXPERIMENTAL PLOTS AND CONSTRUCTION OF FIELD CAGES TO SAMPLE
L. trifolii ADULT POPULATIONS Introduction

Tomatoes in the Palmetto-Ruskin area are grown on sandy soils using staked culture and sometimes subsurface irrigation. The tomato varieties most commonly grown in this area are cv. 'Hayslip' and cv. 'Sunny'. 'Hayslip' is a late, jointless, moderately large-vined, determinate, openpollinated cultivar developed by the Institute of Food and Agricultural Sciences at the University of Florida. 'Sunny' is a midseason, jointed, determinate, hybrid developed by Asgrow. Both varieties are resistant to Verticillium wilt, Fusarium wilt (race 1 and race 2), and gray leaf spot (Maynard 1987).

Agronomic practices followed during two seasons (fall 1986 and spring 1987) at the Bradenton Gulf Coast Research and Education Center were similar to those outlined by Hochmuth et al. (1988).

Preparation of Study Plots

Fields were plowed and disced to bury old crop refuse prior to planting. Two weeks prior to transplanting, three lands (each land consisting of a field of 0.2 ha with an irrigation ditch on both sides) were rototilled and prepared






51

for cultivation. Beds were formed and a 18-0-25 fertilizer was spread on the beds in double shoulder bands. Superphosphate, with 36.32 k fritted trace elements per ton was broadcast over the entire bed width.

Following these procedures the beds were fumigated with methyl bromide and mulched. Tomato plants were transplanted two weeks later.

During the fall 1986 season white polyethylene mulch

and during the spring 1987 season black mulch was used as a means of improving moisture and fertilizer conservation, and weed control. Row spacing was 2.74 m and plant spacing 0.46 m. Water was supplied by seep irrigation through ditches located between every four planted rows. The water level was maintained at 38.1 cm to 45.7 cm below the bed surface to properly irrigate the plants. Transplanting Procedures

Transplants produced in a multi-cell styrofoam tray purchased from Speedling, Inc. were kept one to two weeks prior to transplanting in the screenhouse. They were watered as few times as possible which made them more resistant to field conditions. Before transplanting they were treated with a mixture of the fungicides Demosan 65 WP at 0.908 k a.i./378.5 1 and Truban 40 WP at 499.4 g a.i/378.5 1.

After punching holes every 45.7 cm on the beds

transplants were set manually on the rows on September 22 in






52

1986 and on February 24 in 1987. The transplants had a well developed root ball with the growing medium attached to their roots.

Staking and Tying Procedures

Tomato plants in the Palmetto-Ruskin area are staked to provide fruits higher in quality, and easier to harvest than ground tomatoes. One hundred and twenty cm long by 2.5 cm diameter wooden stakes were placed between each plant in the center of the bed about 2 to 3 weeks after transplanting. The stakes were driven into the ground with pneumatic hammers. Plants were tied with a string about one month after transplanting. The string was wrapped around each stake and past both sides of the tomato plants to provide vertical support. They were tied 3 times during both seasons.

Pest, Disease and Weed Control

During the fall 1986 season southern armyworm

(Spodoptera eridania (Cramer)) was the most prevalent insect pest. Only few tomato pinworms (Keiferia lycopersicella (Walsingham)) and tomato fruitworms (Heliothis zea (Boddie)) were observed. Early blight (Alternaria solani) was the most damaging disease affecting the foliage during the fall 1986 season despite the weekly application of fungicides. This resulted in ca. 50% defoliation by the end of the season. A low incidence of bacterial spot (Xanthomonas campestris pv. vesicatoria) and target spot (Corvnespora cassicola) was also observed on lower leaves during the last






53

four weeks of the fall 1986 season. Weeds growing next to the planted beds and in the irrigation ditches were controlled with paraquat on October 13. A rototiller was used to control weeds on the beds between planted beds on September 23 and October 31.

In the spring season of 1987 southern armyworm was

again the prevalent pest and few tomato pinworms and cabbage loopers (Trichoplusia ni (Hubner)) were observed. Low densities of potato aphids (Macrosiphum euphorbiae (Thomas)) were observed 6 weeks after transplanting but did not increase to damaging levels. There was no significant damage incurred by any bacterial or fungal disease during the spring 1987 season. Weeds were controlled similarly as in the fall 1986 season.

Pesticides and fungicides applied during the fall 1986 and spring 1987 season were applied with a high clearance tractor and are listed in table 3-1.

Construction of Field Cages to Sample Adult L. trifolii

Six 1.8 mx 1.8 mx 1.8 m outdoor cage frames of 1.3 cm diameter galvanized steel electrical conduit were constructed. Sponge strips (5.08 cm x 5.08 cm x 1.8 m) were supported with duct tape at the base of the frames to seal the cage bottoms. The cage frames were covered with fine (0.05 cm x 0.05 cm mesh) lumite screens (Bioquip ProductsR cat. no. 1406D) with bottom edges reinforced with saran tape provided with heavy duty brass grommets every 30.5 cm. In the 1986 fall season the bottom of the screens was secured






54

with two strips of duct tape to the inner side of the cage frames to prevent any insects from escaping. In the 1987 spring season the strips of duct tape were substituted by strips of 0.05 cm x 0.05 cm mesh lumite strips sewn to the
R
screens and glued to the frames with Liquid Nails.

Based on the dimensions of these cages and the plant spacing in the field plots two plants could be enclosed within one cage. Since the plants were grown on raised beds, cages had to be placed on fiberwood boards that supported the cages. Twelve 2.44 m x 1.22 m x 1.27 cm fiberwood boards were painted with gray paint (Scotty'sR no. 52) to protect the boards and to closely match the the color of the sandy soil in the field. A thin masonite board painted gray was nailed on to one side of each fiberwood board with tacks. This board provided a smooth surface which was needed to find collected field specimens. Based on the plant and stake spacing in the field, slots were cut on one long side of the boards so that two plants and five stakes would fit between two boards. Two hinges (with the supporting pins removed) were placed at 0.46 m from the outer margins of the boards on each side to serve as coupling points between two boards. Once the boards were placed on the beds the hinges were held together with nails that kept the boards together and cages were bolted to the boards.








Table 3-1. Insecticides and fungicides applied to control insect
pests and fungal diseases on experimental plots during the
fall 1986 and spring 1987 tomato growing seasons at the
at the Gulf Coast Research and Education Center.

Season Pesticide Rateu Category Date applied


Fall 1986
Dithane M 22 Special Dithane M 22 Special Dithane M 22 Special
Lannate L
Dithane M 22 Special
Tribasic CuSO4
Lannate L Lannate L
Dithane M 22 Special
Tribasic CuSO4
Bravo 500 Lannate L
Dithane M 22 Special
Tribasic CuSO4
Bravo 500
Benlate
Lannate L Bravo 500

Spring 1987
Dithane M 22 Special Dithane M 22 Special Dithane M 22 Special
Bravo 500
Dithane M-45
Bravo 500
Lannate 2L
Bravo 500 Bravo 500
Dipel
Bravo 500
Dipel
Bravo 500
Dipel
Bravo 500
Lannate 2L


681 681 681 0.9 681 1.8 0.9 0.9 681 1.8
1.4 0.9 681 1.8
1.4 454 0.9 1.4


1.4 681 681 1.4 681 1.4 2.2 1.4 1.4 454 1.4 454 1.4 454 1.4 2.2


g/0.4 g/0.4
g/0.4 1/0.4 g/0. 4 k/0.4
1/0.4 1/0.4
g/0.4 k/0.4
1/0.4 1/0.4 g/0. 4 k/0.4 1/0.4 g/0.4 1/0.4 1/0.4


1/0.4 g/0. 4 g/0. 4 1/0.4 g/0.4
1/0.4 1/0.4 1/0.4 1/0.4 g/0.4
1/0.4 g/0. 4 1/0.4
g/0. 4 1/0.4 1/0.4


Fungicide Fungicide Fungicide Insecticide Fungicide Bactericide Insecticide Insecticide Fungicide Bactericide Fungicide Insecticide Fungicide Bactericide Fungicide Fungicide Insecticide Fungicide


Fungicide Fungicide Fungicide Fungicide Fungicide Fungicide Insecticide Fungicide Fungicide Insecticide Fungicide Insecticide Fungicide Insecticide Fungicide Insecticide


@ 189.25-378.5 1/0.4 ha of formulation were the size of the plants.


applied depending on


Oct. Oct. Oct. Oct. Oct. Oct. Oct. Nov. Nov. Nov. Nov. Nov. Nov. Nov. Dec. Dec. Dec. Dec.


Mar. Mar. Apr. Apr. Apr. Apr. May May May May May May May May Jun. Jun.













CHAPTER V
ESTIMATION OF ABSOLUTE ADULT DENSITIES AND
CALIBRATION OF A RELATIVE SAMPLING METHOD
FOR ESTIMATION OF ADULT POPULATION DENSITY OF L. trifolii

Introduction

Integrated pest management (IPM) programs, particularly those utilizing population dynamics models, require accurate, reliable estimates of absolute pest population densities (Marston et al. 1976). Field scouts, however, must be able to produce accurate and consistent relative estimates of population density in a quick and easy manner. Scouts must also be able to take samples at a reasonable cost to growers (Linker et al. 1984). Calibration of the relative-density estimates to absolute-density estimates is essential if scouting data are to be used in predictive management models (Luna et al. 1982).

Due to the large size of insect field populations, samples must be taken to estimate their densities. L. trifolii adult populations have been monitored on several occasions with yellow cards covered with some sticky material. No attempts to relate sticky card catches to actual adult densities have been made, however. Therefore, no predictions of larval densities exceeding action thresholds on tomatoes can be made using sticky card catches






57

of adult leafminers. Under the conditions prevalent in staked tomato cultivars in central Florida, estimating adult Liriomyza populations to predict larval populations using yellow sticky cards is more practical and feasible than taking samples with a D-VacR or a sweep net.

Zehnder and Trumble (1984b) studied the spatial and diel activity of L. trifolii and L. sativae adults using yellow sticky cards in fresh market tomatoes in California. They established a relationship between puparia collected in styrofoam pupal trays and adults trapped two weeks later and concluded that the pupal counts provided a suitable tool for forecasting adult leafminer population sizes. They concluded that most leafminers were collected on sticky traps at lower or middle plant heights.

Field cages were used to develop calibration equations to determine absolute densities of L. trifolii adults in staked and mulched tomatoes using sticky card counts from the field population and counts on known caged populations. To study the influence of sticky card position in relation to plant height, samples at three plant heights were taken during the spring 1987 season.

Materials and Methods

Sampling Method on the Unknown Field Population

Simple random samples of adult L. trifolii were taken with 12.7 cm x 7.6 cm yellow sticky cards (Sticky StripsR Olson Products, Medina, Ohio) during the fall 1986 and spring 1987 seasons. During the fall 1986 season sampling






58

sites were selected at random using a random number table and during the spring 1987 season random numbers were generated using a computer program. The first sample during both seasons was taken ca. 6 weeks after transplanting.

The fiberwood boards were placed underneath two plants during the afternoon, removing the adjacent plants, to provide a basis for the field cages. Three wooden 30.5 cm garden stakes were stapled together longitudinally. Two small binder clips were glued to the tips of the garden stakes to hold the sticky cards on each side of the row. These stakes were stapled at the middle height of the stake between the two tomato plants enclosed in a field cage. The total length of the stapled garden stakes was 91.5 cm. At the same time, another set of similar stakes was stapled at the middle plant height between two plants located about 2 m from the fiberwood boards. The sticky cards were always placed on the east side of the cages to prevent shading effects of the cage in the the morning. Twenty four hours later the cages were carefully placed on the boards and the yellow sticky cards were fixed to the binder clips. At the same time a pair of sticky cards were similarly affixed outside the cages. Prior to placing the sticky cards inside or outside the cages they were sprayed with a pressurized insect trapping adhesive (Tangle-Trap , Tanglefoot Co., Grand Rapids, Michigan) to make sure that their entire surface was sticky. Although the sticky cards had a






59

thin layer of adhesive from the factory, some parts, particularly the edges, were lacking sufficient adhesive material.

On the next day, the sticky cards were recovered and the cages were sprayed with PrentoxR EC (a mixture of 1.2% pyrethrins, 9.6% piperonyl butoxide, 81.2% petroleum distillates, and 8% inert ingredients) at a rate of 8.6 ml/l with a SoloR model 423 back pack mist blower to kill the flies inside the cages that were not trapped on the sticky cards. The mist blower was operated at full thrust while spraying the field cages. Cards next to each cage were also collected on that day. After ca. 30 minutes, flies killed by the insecticide were collected from the fiberwood base with an aspirator made of a flexible rubber tube and a glass eye-dropper with a peace of organdy cloth as a filter. This aspirating technique was performed to make absolute estimates of the population in case a poor relationship between sticky card counts inside and outside the cages was found. These data would also be used to adjust the equations generated for the absolute density estimation based on sticky card counts.

Sampling Method on the Known Leafminer Population

Twenty four hours after sampling the unknown

population, an equal number of one-day-old, unfed male and unfed/unmated female L. trifolii adults was introduced into each field cage. Flies from the laboratory colony were introduced into 30 g transparent plastic cups for their






60

subsequent release in the field cages. The plastic cups were placed between the plants on the boards inside the cages. Two new sticky cards were clipped on to the stake before the flies were released inside each cage. A day later, the sticky cards were collected and each cage was sprayed again with PrentoxR. Dead flies were collected as before.

During the fall 1986 season, 5 pairs of L. trifolii

were introduced into each cage in the first week of sampling when the average plant height was 85.5 cm. Every week thereafter an additional 5 pairs were released per cage. It was possible to take samples for 6 weeks. At the end of the growing season only two pairs were introduced per cage to determine if at low densities the adults could be still collected with the sticky cards. Since a very low proportion was trapped at this density these data were not taken into consideration for statistical analyses.

During the spring 1987 season the number of flies

introduced into the cages was increased as plant height was measured. Thus, 5 pairs were introduced when plants measured an average height of 51.2 cm, 13 pairs when plants measured an average height of 72.9 cm, 24 pairs when plants measured an average height of 96.8 cm, 33 pairs when plants measured an average height of 120.4 cm, 40 pairs when plants measured an average height of 126.0 cm, and 45 pairs when plants measured an average height of 127.6 cm.






61

Influence of Sticky Card Position on Leafminer Catches

During the spring 1987 season samples of the field

population were taken using yellow sticky cards placed at three heights. The positions of the cards were defined as low, medium, and high in relation to plant height. Low cards were placed at a height equivalent to one quarter plant height, medium cards were placed at a height equivalent to one half of the plant height, and the high cards were placed just above the plant canopy. Six pairs of sticky cards for each height category were placed at random in the field. Samples were taken twice a week for a total of 6 weeks.

Determination of Insecticidal Effects on Leafminers Treated Inside Field Cages

To insure that the short residual pyrethrin insecticide had no adverse effects on the flies introduced into the cages, two laboratory experiments were completed. In the first experiment, two pairs of one-day-old, unfed flies were introduced in 30 g plastic cups with a small 2.5 cm x 2.5 cm piece of lumite screen that had been treated with PrentoxR and had been exposed to the sun for 24 h. A total of 7 cups was prepared, using another 7 cups with an untreated piece of lumite as a control. The cups were placed in a rearing room with fluorescent lamps. The second experiment was similar to the first except that the flies (2 one-day-old unfed pairs) were introduced in clip cages (Zoebisch 1984) holding a small piece of lumite inbetween the lids. This increased the probability of contact with the treated






62

surface. A total of 7 clip cages with treated lumite pieces (as in the previous experiment) and another 7 clip cages with untreated lumite screens were used. Mortality was recorded 24 h and 48 h later.

Results and Discussion

Generation of the Calibration Equations on the Known Population

Calibration equations were developed to establish a

numerical relationship between the number of flies collected on yellow sticky cards inside the cages and the known number of introduced flies into a field cage on a per two plant basis. To estimate unknown populations on a per two plant basis these equations would be used to calibrate absolute density estimates. These equations were generated regressing the known number of flies introduced into the cages against the number of flies trapped on the sticky cards inside the cages. For the fall 1986 season the following equation was obtained: FIk = 7.109 + 1.765 FIc (R = 0.87) (1) where FIk equals the known number of flies introduced per cage and FIC equals the number of flies collected on the sticky cards.

During the spring 1987 season the following equation was obtained:

SIk = 5.886 + 1.896 SIC (R2 = 0.93) (2) The terminology for the variables is the same as that of equation (1).






63

Generation of the Calibration Equations on the Field Population

Relating the number of flies trapped outside cages and the number of flies trapped from the unknown population inside the cages was done to establish a relationhip between flies trapped on a per two plant basis and an unknown area. This numerical relationship would be combined with the relationship obtained on the known population to estimate an absolute density of flies collected on yellow sticky cards on a per two plant basis of an unknown field population.

Linear regression analyses were therefore performed on counts of flies from the unknown population trapped on the sticky cards outside and inside cages. Linear equations relating number of flies collected outside the cages as the independent variable and flies collected inside as the dependent variable were developed. The linear regression equation to relate catches of adult flies inside and outside cages for the fall 1986 season is: FIu = 0.704 + 0.103 FOu (R2 = 0.66) (3) where FIu is the number of flies trapped inside the cages, and FOu is the number of flies trapped outside the cages. For the spring 1987 season the following equation was determined:

SIu = 2.046 + 0.256 SOu (R2 = 0.78) (4) Dependent and independent variables have the same meaning as those in the equation for the fall 1986 season.

A Student t-test was used to search for differences in slope coefficients between equations (1) and (2) and (3) and






64

(4) to determine if data from both seasons could be pooled. The t-value to compare the slope coefficients was computed using the following formula (Snedecor and Cochran 1967):

t = (bi - b2)/!(sl/SS(FOu) + (s2z/SS(SOu)) where b, and b2 are the slope coefficients, s,2 and s22 are the variances obtained from the standard error estimates of the slope coefficients and SS(FOu) and SS(SOu) are the sum of squares of the model term obtained in the General Linear Models Procedure using SAS (SAS Institute 1986). The number of degrees of freedom (df) used to search the t-value in a table was computed by the following formula (Snedecor and Cochran 1967):

df = (n, - 2) + (n2 -2)

where n, and n2 are the number of samples taken throughout the fall and spring seasons (36 during each season). Significant differences of the t-values at a P<0.05 level were found when comparing the slope coefficients of equations (1) and (2) and (3) and (4); therefore they must be used separately according to the season (fall or spring) to determine absolute densities. Generation of Calibration Equations to Estimate Absolute Densities Based on Relative Sampling Data

Assuming that FIu=FI (or SIu=SIc), the estimation of absolute adult leafminer density on a per two plant basis involves the creation of an equation for each season substituting the independent variable from the equation of the known population (equations (1) or (2)) by the equation of the unknown population (equations (3) or (4)). In






65

general terms this procedure involves the following steps, using equations obtained for the fall season:

FIu = b0 + bIFOu and FIk = bo' + b 1'FIC

Using the above equations and assuming that FIu= FIe the following equation is obtained:

FIk = bo' + bi ' b +b1QI

where bO' + bI' is the adjustment from trap catches per two plants to calibrate between the actual number of adult flies per two plants versus the number of flies collected on yellow sticky cards, and [b0 + bFpl adjusts yellow sticky card catches from an unknown area to catches on a per two plant basis.

For the fall, FIu (solved from equation (3)) would be susbstituted into equation (1) and for the spring, SIu (solved from equation (4)) would be susbstituted into equation (2). For estimating the average number of adult leafminers per two plants the following equations were obtained:

Fe = 8.352 + 0.182 Fac (5) So = 9.765 + 0.485 Sac (6) where Fd and So are the absolute average densities estimated for the fall and spring seasons, respectively, and Fa and Sac are the mean number of flies collected on sticky cards during the fall and spring season, respectively. Validation of Predicted Adult Leafminer Densities

To determine how well the equations would predict an unknown population of adults based on sticky card catches,






66

predicted values were compared to estimated field density for the fall (Fig. 5-1) and spring (Fig. 5-2) seasons. To do this the following procedure was developed: a) from the data obtained from sampling the known population an average proportion of the aspirated flies was computed to determine what proportion of the total number of flies inside a cage was not collected on the sticky cards (i.e. (females + males) divided by the total known number introduced per cage), and b) the number of flies collected by aspiration was adjusted using the value of the mean proportion obtained from the known population (i.e. observed value from the aspirated unknown population divided by the mean proportion of the known population) to compute the means and confidence intervals (95%) for the aspirated unknown population.

For the fall season all values obtained to determine an average density of flies on a per two plant basis using equation (5) were within the 95% confidence intervals of estimated field densities of mean number of flies per two plants (Fig. 5-1). For the spring season the first two mean number of flies per two plants estimates computed with equation (6) were above the upper limit of the corresponding 95% confidence limits (Fig. 5-2). The two points overestimated density at population levels below 5 flies/2 plants.

The mathematical model developed for the absolute adult fly density estimation is appropriate for use because 10 out of 12 estimates were obtained within the 95% confidence























40


35


30



115






10
15





0'
0 2 4 8 SAMPUNG DATES a CoLLEcrm MEAN A PREDICTED MEAN

Fig. 5-1. Mean �95% C.I. adult flies of L.
trifoli collected and predicted on a per
two plant basis during the fall 1986 season.







68

















70


s0o


80go


40


30


2D


10 *A


0~ I II
0 2 4 8 SAMPLING DATES o COLLECI~ED MEAN A PREDICHD MEAN

Fig. 5-2. Mean �95% C.I. adult flies of L.
trifoli collected and predicted on a per
two plant basis during the spring 1987 season.






69

intervals. The two misses are most likely at densities below treatment levels. To make a decision, as an example, suppose that an average of 20 flies were collected on a per sticky card basis similar to that used to develop the model during the tenth week of the fall season. Substituting the value obtained from the field collection into equation (5), the following estimate of flies per two plants would be obtained: F, = 8.352 + 0.182 * 20, which equals 11.992 (=12 rounded to a whole number) flies per two plants. Proportion and Number of Females and Males Collected on the Sticky Cards During Relative Sampling and Absolute Sampling Procedures

Since female flies cause part of the damage to the crop by stippling and ovipositing inside the foliage, numerical analyses should be done only for females. To separate females from males while counting adults trapped on sticky cards in the field is quite time consuming and under most cases a stereoscope is needed. Therefore, the average proportion of females to males per sampling date was computed to determine if this proportion remained constant through time (Tables 5-1 and 5-2). During the 1986 fall season the mean proportion female/(female + male) of the unknown field population outside and inside the cages, and the known population inside the cages varied up to 2.83, 1.69, and 4.9 times among sampling dates throughout the season, respectively. This variation has no apparent pattern through time (Table 5-1). During the spring season of 1987 the mean proportion from the unknown field






70

population outside and inside the cages, and the known population inside the cages varied up to 2.33, 4.43, and

1.12 times among sampling dates throughout the season with no apparent pattern through time (Table 5-2). Therefore, the data of males and females obtained during the sampling procedure were pooled to generate equations that represent the estimation of the total number of flies collected on a per plant basis.

Although proportions of females and males collected on yellow sticky cards are quite variable and not statistically different, the population sex ratio is approximately 1:1 (Zehnder and Trumble 1984a) who determined the sex ratio populations of L. trifolii based on larvae collected from foliage of celery and tomato. They collected a larger number of males in fresh market tomatoes on sticky cards, however (Zehnder and Trumble 1984b), which coincides with the results obtained in this study. In contrast, Webb et al. (1985) collected an equal number of males and females on yellow sticky cards in a greenhouse with no plants. In a commercial chrysanthemum greenhouse in Baltimore they collected significantly more females than males during 4 out of 10 trapping periods. During the rest of the sampling periods the sex ratio was close to 1:1.

Numerical analyses were done to compare number of flies collected on north and south sides. During the fall 1986 season, sticky cards on the south row side were exposed to the sun while those on the north side were not exposed.








Table 5-1. Proportions of females/(females+males) collected on
the sticky cards on the unknown population (inside and
outside cages) and on the known population inside the
cages during the fall 1986 season.


Sampling date


Mean proportion (9/(9+a))


Unknown population outside cages


Unknown population
inside cages


0.167 � 0.066 � 0.134 � 0.059 � 0.124 � 0.114 � 0.111 �


0.418 a@
0.021 a 0.039 a 0.031 a 0.016 a 0.040 a 0.017 #


0.292 0.292
0.403 0.292
0.492 0.417 0.365


0.213 a 0.100 a 0.141 a 0.187 a 0.067 a 0.138 a 0.035 #


Known population inside cages


0.617 � 0.126 � 0.302 � 0.188 � 0.286 � 0.297 � 0.303 �


0.173 a 0.046 a 0.056 a 0.024 a 0.026 a 0.030 a 0.069 #


@Data transformed to arcsine x; (x=proportion (9/9+0)) prior to analysis but presented in the original scale. Means followed by the same letter vertically are not significantly different 4P<0.05, ANOVA). Data are presented as mean � S.E. Overall mean � S.E.


Nov Nov Nov Dec Dec Dec


Nov Nov Nov Dec Dec Dec








Table 5-2. Proportions of females/(females+males) collected on
the sticky cards on the unknown population (inside and
outside cages) and on the known population inside the cages
during the spring 1987 season.


Mean proportion (9/(9+d))


Sampling
date


Unknown population outside cages


0.056
0.054 0.060
0.126 0.078 0.099
0.079


0.056 0.035
0.021 0.037 0.037
0.012 0.012


Unknown population
inside cages


0.167 0.117 0.114 0.155
0.040 0.177 0.128


� 0.114 � 0.053 � 0.030 + 0.065
� 0.036 + 0.027 + 0.021


Known population
inside cages


0.192 � 0.197 � 0.178 � 0.194 � 0.199 �
0.181 � 0.190 �


0.045 a 0.039 a 0.012 a 0.027 a 0.020 a 0.019 a
0.009 #


@Data transformed to arcsine jx; (x=proportion 9/9+d) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different JP<0.05, ANOVA). Data are presented as mean � standard error. Overall mean � standard error.


Apr May May May May Jun


May May May May May Jun






73

Although proportions of females and males collected on yellow sticky cards are quite variable and not statistically different, the population sex ratio is approximately 1:1 (Zehnder and Trumble 1984a) who determined the sex ratio populations of L. trifolii based on larvae collected from foliage of celery and tomato. They collected a larger number of males in fresh market tomatoes on sticky cards, however (Zehnder and Trumble 1984b), which coincides with the results obtained in this study. In contrast, Webb et al. (1985) collected an equal number of males and females on yellow sticky cards in a greenhouse with no plants. In a commercial chrysanthemum greenhouse in Baltimore they collected significantly more females than males during 4 out of 10 trapping periods. During the rest of the sampling periods the sex ratio was close to 1:1.

Numerical analyses were done to compare number of flies collected on north and south sides. During the fall 1986 season, sticky cards on the south row side were exposed to the sun while those on the north side were not exposed. During the spring 1987 season sticky cards on both row sides were exposed to the sun. During the fall 1986 season significantly more flies were collected on the south (sunny row side) during the last 4 sampling dates on the unknown population outside and inside cages and during all sampling dates, except the first one, inside cages of the known population (Table 5-3). During the spring 1987 season, significantly more flies from the unknown population inside






74

cages were collected only on the last sampling day (June 4) on the north side (Table 5-4). From the known population during the spring 1987 season significantly more flies were collected on the south side on the first sampling day (May 2), and on the second sampling date (May 9) significantly more flies were collected in the north side.

In most cases (primarily when the population densities started increasing) significantly more males were collected on the sunny (south) side (Tables 5-5 and 5-6). Females seemed to have responded more uniformly to any light intensity (Tables 5-5 and 5-6). Few significant differences between the north and south sides were obtained throughout the spring season (Tables 5-7 and 5-8).

Some of the factors that influence the skewed sex ratio obtained on yellow sticky cards may be that males spend more time searching for females and therefore encounter yellow cards more often or they are more attracted to these cards than are females.

Males responded more uniformly to cards exposed to the sun which indicates that they are more sensitive to the intensity of light reflected on the cards than females. This response may function as a primary orientation cue for males to search for females. In addition, males seem to be able to discern between light intensities under varying conditions.

During most sampling dates in the fall 1986 season, significantly more males were collected on the sunny side








Table 5-3. Mean number of flies of unknown population outside
and inside cages and known population inside cages collected
in the south (sunny) and north (shady) side of the tomato
field rows sampled during the fall 1986 season.

Sampling date South side North side

Unknown population collected outside cages

Nov 13 0.667 � 0.333 a@ 0.333 � 0.211 a Nov 20 17.167 � 3.799 a 9.833 � 2.496 a Nov 27 6.833 � 3.825 a 9.500 � 1.088 b Dec 4 13.667 � 2.629 a 3.167 � 1.545 b Dec 11 16.167 � 7.039 a 0.667 � 6.380 b Dec 16 18.333 � 3.084 a 4.500 � 1.708 b
22.889 � 7.299 # 7.250 � 2.333 #

Unknown population collected inside cages

Nov 16 1.000 � 0.447 a@ 0.500 � 0.224 a Nov 23 2.000 � 0.449 a 1.000 � 0.365 a Nov 30 4.167 � 0.792 a 1.333 � 0.333 b Dec 7 1.500 � 0.428 a 0.333 � 0.211 b Dec 14 5.667 � 0.989 a 2.000 � 0.577 b Dec 18 2.667 � 0.715 a 0.833 � 0.307 b
2.834 � 1.279 # 0.999 � 0.247 #

Known population collected inside cages

Nov 16 2.500 � 0.671 a@ 1.000 � 0.517 a Nov 23 5.167 � 0.401 a 2.500 � 0.500 b Nov 30 9.167 � 0.946 a 3.333 � 0.882 b Dec 7 15.833 � 0.946 a 5.500 � 0.882 b Dec 14 13.667 � 1.202 a 8.333 � 1.542 b Dec 18 18.333 � 1.085 a 9.500 � 1.648 b
10.778 � 1.018 # 5.028 � 1.284 #

@Data transformed to (./x+-.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test). Data are presented as mean � S.E. Overall mean � S.E.








Table 5-4. Mean number of flies of unknown population outside
and inside cages and known population inside cages
collected in the north and south side of the tomato
field rows sampled during the spring 1987 season.

Sampling date South side North side

Unknown population collected outside cages

Apr 30 0.667 � 0.333 a@ 2.000 � 0.632 a May 7 6.667 � 1.022 a 5.833 � 1.740 a May 14 10.000 � 4.251 a 11.000 � 3.425 a May 21 9.833 � 1.580 a 9.333 � 2.076 a May 28 2.667 � 1.116 a 3.167 � 0.946 a Jun 4 38.333 � 4.958 a 31.167 � 7.463 a 11.361 � 1.513 # 10.417 � 4.097 #

Unknown population collected inside cages

Apr 30 0.833 � 0.307 a@ 0.833 � 0.447 a May 7 1.667 � 0.494 a 2.333 � 0.667 a May 14 4.500 � 0.671 a 4.167 � 1.195 a May 21 4.167 � 0.872 a 7.000 � 1.525 a May 28 0.833 � 0.477 a 1.500 � 0.719 a Jun 4 7.500 � 0.764 a 11.500 � 1.335 b 3.258 � 1.071 # 4.556 � 1.658 #

May 2 4.000 � 0.517 b@ 0.833 � 0.163 a May 9 1.667 � 0.615 a 9.617 � 0.401 b May 16 12.167 � 1.778 a 10.833 � 1.600 a May 21 13.667 � 1.977 a 14.500 � 2.337 a May 28 17.333 � 1.430 a 21.667 � 1.109 a Jun 4 22.833 � 2.762 a 21.333 � 2.104 a 11.945 � 3.263 # 13.056 � 3.236 #

@Data transformed to (/~TE.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test). Data are presented as mean � S.E. #Overall mean � S.E.








Table 5-5. Mean number of females and males of unknown
population outside and inside cages collected in the north and south side of the tomato field rows sampled during the
fall 1986 season.

Sample date South side North side

Unknown population collected outside cages (Females)


Nov Nov Nov Dec Dec Dec


0.000
1.167 4.500 1.000 4.833 1.833
2.194


0.000 0.447 0.619 0.872 0.872
1.447 0.818


0.167 0.833 1.333 0.167 3.167 1.167 1.139


0.167 2.069
0.422 0.167 0.307
0.477 0.452


Unknown population collected outside cages (Males)


Nov Nov Nov Dec Dec Dec


0.667
16.170 32.333
12.667 45.833
16.500 20.695


0.333 a 3.719 a 2.249 b 2.268 b 6.321 b 1.918 b
6.509


0.167 9.000
8.167 3.000 13.000 3.333 6.111


0.167 2.066 1.327 0.683 2.581 1.520 1.939


Unknown population collected inside cages (Females)


0.333
0.667 2.167 0.667 2.667 1.500 1.334


0.333 0.333 1.880
0.422 0.667
0.500 0.383


0.167 0.500 0.667 0.167 1.167
0.500 0.528


0.516 a 0.224 a 0.211 a 0.167 a 0.307 a 0.224 a 0.152 *


Unknown population collected inside cages (Males)


0.667 � 1.333 � 2.000 � 0.833 � 3.000 � 1.167 � 1.500 �


0.211 a 0.211 a 0.577 a 0.307 a 0.632 b 0.305 b 0.355 #


0.333
0.500 0.667 0.167 0.833 0.333
0.472


0.211 a 0.224 a 0.211 a 0.167 a 0.478 a 0.211 b 0.010 #


@Data transformed to (jx+0.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test). Data are presented as mean � S.E. #Overall mean � S.E.


Nov Nov Nov Dec Dec Dec


Nov
Nov Nov Dec Dec Dec








Table 5-6. Mean number of females and males of known population
inside cages collected in the sunny and shady side of the
tomato field rows sampled during the fall 1986 season.

Sample date South side North side


Known population collected inside cages (Females)

Nov 16 1.167 � 0.307 a@ 0.167 � 0.167 a Nov 23 0.833 � 0.307 a 0.167 � 0.167 a Nov 30 2.833 � 0.872 a 1.000 � 0.258 b Dec 7 2.000 � 0.516 a 2.167 � 0.792 a Dec 14 4.667 � 0.558 a 1.667 � 0.211 b Dec 18 4.500 � 0.619 a 3.833 � 1.327 a
2.667 � 0.669 1.500 � 5.569 #

Known population collected inside cages (Males)

Nov 16 1.333 � 0.307 a 0.833 � 0.543 a Nov 23 4.333 � 0.211 a 2.333 � 0.422 b Nov 30 6.333 � 0.558 a 2.333 � 0.843 b Dec 7 13.833 � 0.871 a 3.333 � 0.803 b Dec 14 9.000 � 0.857 a 6.667 � 1.520 a Dec. 18 13.833 � 0.980 a 5.667 � 1.358 b
8.111 � 2.079 # 3.528 � 0.905 #

@Data transformed to (!x+0.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test).
Overall means as mean � S.E.








Table 5-7. Mean number of females and males of unknown
population outside and inside cages collected in the north
and south side of tomato field rows sampled during the
spring 1987 season.

Sample date South side North side

Unknown population collected outside cages (Females)


0.167 � 0.167 � 0.667 � 1.167 0.167 � 3.333 �
0.945 �


0.167 a@
0.167 a 0.494 a 0.543 a 0.167 a 0.615 a 0.505 #


0.000 0.167 0.667 1.500
0.500 3.667 1.084


0.000 a 0.167 a 0.422 a 0.619 a 0.342 a 0.211 a 0.559 #


Unknown population collected outside cages (Males)


0.500 � 6.500 � 9.333 � 8.667 � 2.500 � 35.000 �
10.417 �


0.224 1.148 3.827 1.358 0.992
4.817 5.098


0.167 5.667 10.333 7.833
2.667 27.500
9.028


0.632
1.801
2.449 1.621 0.667 7.451 3.997


Unknown population collected inside cages (Females)


0.667 0.333
0.500 0.500 0.000 1.500 0.500


0.167 0.211 0.224 0.224 0.000
0.342 0.553


1.167 0.157
0.500 1.167 0.167
1.667 0.806


0.167 0.167 0.224 0.509 0.167 0.211 0.649


Unknown population collected inside cages (Males)


0.667 � 1.333 � 4.000 � 3.667 � 0.833 � 6.000 � 2.750 �


0.333 0.558
0.816 0.954 0.477 0.577 0.818


0.667
2.167 3.667 5.833 1.333 9.833
3.917


0.498
0.703 1.145 0.938 0.558
1.424 1.311


@Data transformed to (jx+0.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test). Data are presented as mean � S.E. #Overall mean � S.E.


Apr May May May May Jun


Apr May May May May Jun


Apr May May May May Jun


Apr. May May May May Jun








Table 5-8. Mean number of females and males of known population
inside cages collected in the south and north side of the
tomato field rows sampled during the spring 1987 season.

Sampling date South side North side

Known population collected inside cages (Females)

May 2 0.167 � 0.167 b 0.833 � 0.167 a May 9 1.000 � 0.449 a 1.167 � 0.601 a May 16 2.333 � 0.760 a 1.833 � 0.307 a May 23 2.667 � 0.760 a 3.000 � 0.578 a May 30 3.833 � 1.302 a 4.000 � 0.843 a Jun 7 3.667 � 0.715 a 4.333 � 0.843 a
2.278 � 0.555 # 2.528 � 0.562 #

Known population collected inside cages (Males)

May 2 0.667 � 0.211 a 3.167 � 0.477 b May 9 8.167 � 0.401 a 0.500 � 0.342 b May 16 8.500 � 0.847 a 10.333 � 1.874 a May 23 11.833 � 2.182 a 10.667 � 1.606 a May 30 17.833 � 2.212 a 13.333 � 1.686 a Jun 7 17.667 � 1.382 a 18.500 � 2.565 a
10.778 � 2.488 # 9.417 � 2.358 #

@Data transformed to (/x+0.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test). Data are presented as mean � S.E. Overall mean � S.E.






81

inside or outside the cages, regardless of the origin of the flies. Temperature differences between sunny and shady sides may also influence the number of flies collected on sticky cards at both sides although light intensity may be more important.

Females, in contrast, may be attracted more intensively in staked tomatoes by factors other than reflectance to search for suitable oviposition and feeding sites. Affeldt et al. (1983) observed that more L. sativae were collected on surfaces with more light, particularly those facing the morning sun. Although leafminers were not observed on an hourly basis, the possibility exists that most males were trapped during morning hours on the cards facing the sun. Zehnder and Trumble (1984b) reported that most L. trifolii flight activity in fresh market tomatoes peaked from 0700 h to 1100 h due primarily to movement of males.

Female catches were more variable than male catches in both seasons. At the end of the 1987 spring season two samples on yellow sticky cards inside cages on known numbers of females were taken. In the first sample 20 one-day-old unfed females were introduced into each of three cages while 20 one-day-old females fed with honey for 24 h were introduced into the other 3 field cages. No significant differences between sticky card catches were observed (Table 5-9). In a second experiment 20 mated (3 cages) versus 20 unmated (3 cages) one-day-old unfed females were compared and no significant differences in trap catches were observed






82

(Table 5-9). More experiments would be needed to determine

which factors influence the trapability of females versus

that of males.


Table 5-9. Average laboratory-reared females collected on yellow
sticky cards inside cages when feeding and mating status
varied with a total of 20 females were introduced per cage. Age, feeding and Mean � S.E. number of females mating status collected on sticky cards

1 day, unfed,
unmated 3.667 � 0.439 a
1 day, fed,
unmated 3.000 � 0.577 a
1 day, unfed,
unmated 3.000 � 0.577 a
1 day, unfed,
mated 4.000 � 0.577 a @Data transformed to (!x+0.5) prior to analysis but presented in the original scale. Means followed by the same letter vertically are not significantly different (P<0.05, Student's t-test).


Sampling at Three Different Heights

The proportion (female/female+male)) of L. trifolii

collected at three different heights was not significantly

different during 8 out of 12 sampling dates (Table 5-10).

Most females and males were collected during the last three weeks and most of them were collected at the lower or middle

height (Tables 5-11, 5-12, and 5-13) which agrees with

results obtained by Zehnder and Trumble (1984b).

Densities during the first 3 sampling weeks were

probably too low to allow for a distinction in plant strata.

Due to the high number of males collected throughout the








Table 5-10. Mean proportion of female L. trifolii (female/female
+ male) collected on sticky cards at three heights during
the spring 1987 season.

Sampling date Sticky cards' position High Medium Low

May 7 0.125 � 0.085 a' 0.075 � 0.048 a 0.039 � 0.023 a May 11 0.026 � 0.017 a 0.070 � 0.034 b 0.222 � 0.070 b May 14 0.065 � 0.021 a 0.065 � 0.022 a 0.095 � 0.026 a May 17 0.104 � 0.029 a 0.134 � 0.028 a 0.171 � 0.047 a May 21 0.088 � 0.052 a 0.126 � 0.037 a 0.207 � 0.121 a May 25 0.092 � 0.050 b 0.240 � 0.034 a 0.196 � 0.051ab May 28 0.072 � 0.019 a 0.078 � 0.037 a 0.239 � 0.010 a May 31 0.025 � 0.011 c 0.149 � 0.046 b 0.294 � 0.047 a Jun 4 0.119 � 0.031 a 0.110 � 0.023 a 0.139 � 0.017 a Jun 7 0.062 � 0.031 a 0.015 � 0.002 a 0.098 � 0.040 a Jun 11 0.014 � 0.023 b 0.104 � 0.064 a 0.173 � 0.144 a Jun 14 0.091 � 0.048 a 0.085 � 0.038 a 0.147 � 0.103 a
0.074 � 0.015 # 0.108 � 0.021 # 0.168 � 0.029 #

@Data transformed to arcsine x (x=proportion 9/9+c) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different ;P<0.05, ANOVA). Data are presented as mean � S.E. Overall mean � S.E.

Table 5-11. Mean number of females collected on yellow sticky cards at three different heights during the spring 1987 season.

Sampling date Sticky cards' position High Medium Low

May 7 0.333 � 0.211 a@ 0.333 � 0.211 a 0.667 � 0.211 a May 11 0.333 � 0.211 a 1.167 � 0.307 a 1.167 � 0.401 a May 14 1.000 � 0.365 a 1.500 � 0.619 a 1.333 � 0.422 a May 17 1.333 � 0.422 b 2.667 � 0.494 ab 4.333 � 1.085 a May 21 0.667 � 0.333 a 2.667 � 1.022 a 1.833 � 0.477 a May 25 1.167 � 0.600 a 2.833 � 0.600 a 2.333 � 0.667 a May 28 0.333 � 0.211 b 0.667 � 0.336 a 1.000 � 0.000ab May 31 1.000 � 0.773 b 2.167 � 0.477 ab 2.667 � 0.494 a Jun 4 4.500 � 0.922 b 7.167 � 0.401 a 8.333 � 1.256 a Jun 7 1.000 � 0.447 b 3.833 � 0.703 a 2.833 � 1.046ab Jun 11 0.333 � 0.211 b 2.333 � 0.494 a 2.333 � 0.558 a Jun 14 2.667 � 0.422 b 4.833 � 0.980 ab 6.333 � 1.564 a

@Data transformed to (jx+0.5) prior to analysis but presented in the original scale. Means (� S.E.) followed by the same letter horizontally are not significantly different (P<0.05, ANOVA).







Table 5-12. Mean number of males collected on yellow sticky
cards at three different heights during the spring 1987
season.

Sampling date Sticky cards' position
High Medium Low

May 7 5.667 � 2.028 bw 9.667 � 3.499 ab 18.333 � 3.685 a May 11 8.167 � 1.815 a 8.167 � 1.352 a 4.333 � 1.022 a May 14 15.000 � 2.309 a 19.500 � 3.871 a 13.000 � 1.437 a May 17 11.500 � 2.141 a 19.500 � 2.907 a 18.667 � 4.088 a May 21 8.000 � 1.366 b 17.000 � 2.098 a 14.500 � 3.334ab May 25 10.167 � 1.250 a 9.000 � 1.571 a 10.833 � 2.429 a May 28 3.333 � 1.202 a 5.167 � 1.424 a 6.333 � 2.290 a May 31 1.000 � 0.365 a 15.500 � 3.149 b 7.667 � 2.319 b Jun 4 26.333 � 4.248 b 48.000 � 9.533 ab 49.833 � 4.512 a Jun 7 21.000 � 7.510 b 53.833 � 6.416 a 26.000 � 2.633 b Jun 11 17.667 � 4.920 a 21.667 � 3.041 a 19.667 � 6.136 a Jun 14 30.667 � 6.092 a 58.333 � 13.258 a 47.167 � 10.316 a

@Data transformed to (/x0.5) prior to analysis but presented in the original scale. Means (� S.E.) followed by the same letter horizontally are not significantly different (P<0.05, ANOVA).

Table 5-13. Mean number of total flies collected on yellow
sticky cards at three different heights during the spring
1987 season.

Sampling date Sticky cards' position
High Medium Low

May 7 6.000 � 1.915 b@ 10.000 � 3.386 ab 19.000 � 3.751 a May 11 8.500 � 1.979 a 9.333 � 3.502 a 5.500 � 3.332 a May 14 16.000 � 2.408 a 21.000 � 10.583 a 14.333 � 1.520 a May 17 12.833 � 2.272 a 22.167 � 2.960 a 23.000 � 4.496 a May 21 8.667 � 1.308 b 19.667 � 2.789 a 16.333 � 2.916 a May 25 11.333 � 1.358 a 11.833 � 1.939 a 13.167 � 2.574 a May 28 3.667 � 1.333 a 5.833 � 1.720 a 7.333 � 2.290 a May 31 37.333 � 8.550 a 17.667 � 2.871 b 10.333 � 2.704 b Jun 4 30.833 � 3.995 b 55.167 � 9.676 a 58.167 � 5.300 a Jun 7 22.000 � 7.920 b 57.667 � 7.070 a 28.333 � 2.949 b Jun 11 18.000 � 2.449 a 24.000 � 3.033 a 22.000 � 6.022 a Jun 14 33.333 � 6.184 a 63.167 � 13.751 a 53.500 � 10.760 a

@Data transformed to (-x+0.5) prior to analysis but presented in the original scale. Means (� S.E.) followed by the same letter horizontally are not significantly different (P<0.05, ANOVA).




Full Text

PAGE 1

SAMPLING METHODS TO ESTIMATE ABSOLUTE AND RELATIVE DENSITY OF ADULTS AND LARVAE OF Liriomvza trifolii (Burgess) ON FRESH MARKET TOMATOES AND BIOLOGY OF Liriomvza trifolii UNDER LABORATORY CONDITIONS By TOMAS GUNTER ZOEBISCH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1988

PAGE 2

ACKNOWLEDGEMENTS I am very thankful to Dr. D. J. Schuster, chairman of my supervisory committee, for his great friendship, guidance, and financial and invaluable moral support. I would like to express my gratitute to Dr. J. L. Stimac for his skillful help in statistical analyses, and his lively humor while he served on my supervisory committee. My sincerest thanks go to Dr. G. H. Smerage for providing me with concepts in mathematical models and also with moral support while he served on my supervisory committee. I would like to thank Dr. S. H. Kerr, for his academic advice and administrative support, and Dr. J. R. Strayer for his moral support and professional guidance, both serving on my supervisory committee. The technicians at the Gulf Coast Research and Education Center helped to make my field and laboratory experiments possible. I am particularly thankful to Ken Kiger for his help in establishing and maintaining my experimental plots. The faculty members at the Gulf Coast Research and Education Center were tremendously supportive to me

PAGE 3

emotionally. I would like to thank Dr. S. S. Woltz for allowing me to use his personal computer to learn SAS and Dr. J. B. Dr. Kring for motivating me and giving me a good time in the laboratory. Dr. J. Allen gave me valuable advice on data analyses to whom I grateful express my thanks. I also appreciate very much the friendship of Dr. K. J. Patel and his wife, Falguni, who helped me emotionally to finish my degree. Mr. G. Zoebisch supported me financially, making my coming to the U.S. possible. I am grateful for his financial support. Mr. and Mrs. Hof also supported me morally and financially during my hardest times. Their help is gratefully acknowledged. I thank my parents for their great love, moral support, encouragement, and financial support under all circumstances . Church members from the Church of Jesus Christ and Latter Day Saints in Bradenton and Gainesville always supported me and my family with great blessings. I appreciate very much their help. Most of all, I would like to express my thanks and love to my beautiful wife and daughter, whose moral and emotional support enabled me to finish my dissertation. iii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS LIST OF TABLES viii LIST OF FIGURES xii ABSTRACT CHAPTER I INTRODUCTION * 1 II LITERATURE REVIEW 6 Systematics and Distinguishing Characteristics of Liriomvza trifolii 6 Life Cycle and Biology of L. trifolii 11 Biology and Behavior of Adults 11 Oviposition and Egg Developmental Time 14 Larval Development 16 Pupal Developmental Time 17 Geographical Distribution of L. trifolii IV Alternative Hosts of L. trifolii 20 Crop Damage and Effects on Yield 21 Reduction in Crop Yield 21 Entry of Pathogens and Physiological Changes due to Larval Mining 23 Reduction of Photosynthetic Rates 23 Chemical Control of L. trifolii on Tomato 24 Resistance of Leafminers to Insecticides Used in Tomatoes 27 Effects of Insecticides on Parasitoids of Liriomvza Leafminers 29 iv

PAGE 5

Page Biological Control of Liriomvza trifolii ^2 Parasitoids of Liriomvza trifolii on Tomatoes 32 Predators of Liriomvza Leaf miners 33 Effects of Cultural Methods on Populations of Liriomvza Leaf miners 35 Sampling Methods Used to Monitor Populations of Liriomvza Leaf miners 37 Methods to Estimate Larval Densities 37 Methods to Estimate Adult Densities 4° III PLANT PRODUCTION AND MAINTENANCE OF L. trifolii COLONIES 45 Introduction 45 Tomato Plant Production 46 Screenhouse Maintenance 47 Establishment and Maintenance of the L. trifolii Colony 47 IV PREPARATION OF EXPERIMENTAL PLOTS AND CONSTRUCTION OF FIELD CAGES TO SAMPLE L. trifolii ADULT POPULATIONS 50 Introduction 50 Preparation of Study Plots 50 Transplanting Procedures 51 Staking and Tying Procedures 52 Pest, Disease and Weed Control 52 Construction of Field Cages to Sample Adult L. trifolii 53 V ESTIMATION OF ADULT DENSITIES AND CALIBRATION OF A RELATIVE SAMPLING METHOD FOR ESTIMATION OF ADULT POPULATION DENSITY OF L. trifolii 56 Introduction 56 Materials and Methods 57 Sampling Method on the Unknown Field Population 57 Sampling Method on the Known Leaf miner Population 59 V

PAGE 6

Page Influence of Sticky Card Position on Leafminer Catches 61 Determination fo Insecticidal Effects on Leafminers Treated Inside Field Cages 61 Results and Discussion 62 Generation of the Calibration Equations on the Known Population 62 Generation of the Calibration Equations on the Field Population 63 Generation of Calibration Equations to Estimate Absolute Densities Based on Relative Sampling Data 64 Validation of Predicted Adult Leafminer Densities 65 Proportion and Number of Females and Males Collected on the Sticky Cards During Relative and Absolute Sampling Procedures 69 Sampling at Three Different Heights 82 Response of Flies Exposed to Pyrethrum Insecticide in the Laboratory 85 VI ESTIMATES OF FIRST-SECOND AND THIRD STAGE LARVAL DENSITIES OF Liriomvza trifolii ON TOMATOES 87 Introduction 87 Materials and Methods 90 Results and Discussion 92 VII FEMALE SURVIVAL, OVI POSITION, EGG AND LARVAL DEVELOPMENT OF Liriomvza trifolii ON TOMATO FOLIAGE 104 Introduction 104 Materials and Methods 106 Planting of Host plants 106 Determination of Adult Female Longevity, Adult Female Survival, Oviposition Rate, Fertility, and Egg Development Rate 106 Determination of Larval Deve 1 opment 107 Equations Used to Describe the Biological Processes Studied in This Chapter 108 vi

PAGE 7

Page Results and Discussion 109 Adult Female Longevity and Survival, Fertility, and Egg Developmental Rate 109 Larval Development 121 Pupal Development 132 Comparison of the Regression Equations 133 Proposal of a Conceptual Model on Population Dynamics of L. trifolii 134 VIII CONCLUSIONS 1*2 LITERATURE CITED 1^7 BIOGRAPHICAL SKETCH 163 6vii

PAGE 8

LIST OF TABLES Page Table 2-1. Table 2-2. Table 3-1. Table 5-1, Table 5-2. Larval development time of L. trifolii at similar temperatures in foliage of different hosts.... Pupal development time of L. trifolii at similar temperatures from larvae that developed in foliage of different hosts Insecticides and fungicides applied to control insect pests and fungal diseases on experimental plots during the fall 1986 and spring spring 1987 tomato growing seasons at the Gulf Coast Research and Education Center Proportion of females/males collected on the sticky cards on the unknown population (inside and outside cages) and on the known population inside the cages during the fall 1986 season Proportion of females/males collected on the sticky cards on the unknown population (inside and outside cages) and on the known population inside the cages during the spring 1986 season 18 19 55 71 72 Table 5-3. Mean number of flies of the unknown population outside and inside cages and known population inside cages collected in the south (sunny) and north (shady) side of the tomato field rows sampled during the fall 1986 season 75 viii

PAGE 9

Page Table 5-4. Mean number of flies of the unknown population outside and inside cages and known population inside cages collected in the south and north side of the tomato field rows sampled during the spring 1987 season 76 Table 5-5. Mean number of females and males of unknown population outside and inside cages collected in the south (sunny) and north (shady) side of the tomato field rows sampled during the fall 1986 season 77 Table 5-6. Mean number of females and males of known population inside cages collected in the south (sunny) and north (shady) side of the tomato field rows sampled during the fall 1986 season 78 Table 5-7. Mean number of females and males of unknown population outside and inside cages collected in the south and north side of the tomato field rows sampled during the spring 1987 season 79 Table 5-8. Mean number of females and males of known population inside cages collected in the south and north side of the tomato field rows sampled during the spring 1987 season 80 Table 5-9. Average laboratory-reared females collected on yellow sticky cards when feeding and mating status varied with a total of 20 females per cage 82 Table 5-10. Mean proportion of female L. trif olii (female/female+male) collected on sticky cards at three heights during the spring 1987 season 83 Table 5-11. Mean number of females collected on yellow sticky cards at three different heights during the spring 1987 season 83 ix

PAGE 10

Page Table 5-12 Mean number of males collected on yellow sticky cards at three different heights during the spring 1987 season 84 Table 5-13. Mean number of total flies collected on yellow sticky cards at three different heights during the spring 1987 season 84 Table 5-14. Response of laboratory-reared L. trifolii Prentox -treated lumite screen after 24 h 86 Table 5-15. Response of laboratory-reared L. trifolii Prentox -treated lumite screen after 48 h 86 Table 6-1. Regression analysis results with an R^ >0.50 on relative and absolute sampling procedures on large larvae sampled on the fourth, seventh and tenth nodes on lateral stems on six randomly selected plants Table 6-2 Table 6-3. Table 6-4 Regression analysis results with an R^ >0.50 on relative and absolute sampling procedures on small larvae sampled on the fourth, seventh and tenth nodes on lateral stems on six randomly selected plants Regression analysis results with an R^ >0.50 on relative and absolute sampling procedures on large and small larvae sampled on the fourth, seventh and tenth nodes on four lateral stems and main stem of six randomly selected plants 93 95 97 Percentage of sampling occasions in which the regression analysis yielded an R^ > 0.50 out of twelve sampling dates on large (L) and small (S) larvae on main stems.... 98 Table 6-5. Percentage of sampling occasions in which the regression analysis yielded an R^ > 0.50 out of twelve sampling dates on large (L) and small (S) larvae on four lateral stems 99 X

PAGE 11

Page Table 6-6. Percentage of sampling occasions in which the regression analysis yielded an > 0.50 out of twelve sampling dates on large (L) and small (S) larvae on four lateral stems and main stem 99 Table 7-1. Mean adult female longevity (days) at four constant temperatures 109 Table 7-2. Mean total eggs laid per female throughout its adult lifetime at four constant temperatures 116 Table 7-3. Magnitude and time of peak oviposition rate of L. trif olii at four constant temperatures under laboratory conditions 117 Table 7-4. Percent of L. trifolii larvae that hatched at four constant temperatures 118 Table 7-5. Mean egg developmental time (days) at four constant temperatures 118 Table 7-6. values obtained on linear, quadratic and exponential regression equations from the biological processes studied at four constant temperatures (13.9, 20, 25 and 32 'C) 122 Table 7-7. Linear equations for temperaturedependent-development of Liriomyza trifolii pupae 133 Table 7-8. Slope (1/K) and threshold temperature estimates (Tq) obtained from linear regression equations that describe some of the biological processes of L. trifolii studied in this chapter 140 Table 7-9. Mean and variance of the time of biological processes at four constant temperatures (13.9, 20, 25, and 32 "C 141 xi

PAGE 12

LIST OF FIGURES Page Fig. 5-1. Mean ± 95% C.I. adult flies of L. trifolii collected and predicted on a per two plant basis during the fall 1986 season Fig. 5-2. Mean ± 95% C.I. adult flies of L. trifolii collected and predicted on a per two plant basis during the spring 1987 season Fig. 6-1. Mean total ± 95% C.I. L. trifolii larvae (small and large) collected on the seventh node on four lateral steins on six plants per sampling date 101 Fig. 7-1. Adult female longevity (in days) at four constant temperatures 13.9, 20, 25 and 32 °C) 110 Fig. 7-2. Adult female survival rate (1/day) at four constant temperatures 13.9, 20, 25 and 32 'C) Ill Fig. 7-3. Mean number of eggs laid per day per female at four constant temperatures 112 Fig. 7-4. Mean number of eggs laid per day per female at (13.9, 20, 25 and 32 *C ) throughout its lifetime 114 Fig. 7-5. Total eggs laid per female at four constant temperatures (13.9, 20, 25 and 32 'C ) 115 Fig. 7-6. Mean egg developmental time at four constant temperatures (13.9, 20, 25 and 30 'C) 119 Xll

PAGE 13

Page Fig. 7-7. Mean egg developmental rate at four constant temperatures (13.9, 20, 25 and 32 °C) 120 Fig. 7-8. Larval developmental time at four constant temperatures (13.9, 20, 25 and 32 'C) 123 Fig. 7-9. Small larvae developmental time at four constant temperatures (13.9, 20, 25, and 32 'C) 124 Fig. 7-10. Large larvae developmental time at four constant temperatures (13.9, 20, 25, and 32 'C) 125 Fig. 7-11. Larval developmental rate (1/day) at four constant temperatures (13.9, 20, 25, and 32 'C) 126 Fig. 7-12. Larval developmental rate (1/day) transformed to Napierian logarithms) at four constant temperatures (13.9, 20, 25, and 32 °C) 127 Fig. 7-13. Developmental rate (1/day) of small larvae at four constant temperatures (13.9, 20, 25, and 32 'C) 128 Fig. 7-14. Developmental rate (1/day) of small larvae (transformed to Napierian logarithms) at four constant temperatures (13.9, 20, 25, and 32 'C) 129 Fig. 7-15. Developmental rate (1/hour) of large larvae at four constant temperatures (13.9, 20, 25, and 32 'C) 130 Fig. 7-16. Developmental rate (1/day) of large larvae (transformed to Napierian logarithms) at four constant temperatures (13.9, 20, 25, and 32 "C) 131 Fig. 7-17. Diagram of a conceptual model of Liriomyza trifolii on tomatoes 136 Fig. 7-18. Representative probability density function of development time in a life stage 138 • t • Xlll

PAGE 14

Abstract of Dissertation Presented to the Graduate school of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SAMPLING METHODS TO DETERMINE ABSOLUTE AND RELATIVE DENSITY OF ADULTS AND LARVAE OF Liriomvza trifolii (Burgess) ON FRESH MARKET TOMATOES AND BIOLOGY OF Liriomvza trifolii UNDER LABORATORY CONDITIONS By Tomas Giinter Zoebisch December 1988 Chairman : D. J. Schuster Major Department: Entomology and Nematology A sampling method for estimating absolute adult leafminer populations on a per-two-plant basis using yellow sticky cards as a relative sampling method was developed under field conditions. The following equations for estimating adult leafminer densities were developed: F^^= 8.352+0.182Fg^ and S^^=9. 765+0. 4855^^ where F^ and S^ are the absolute average densities estimated for the fall and spring seasons, respectively, and F,,^, and S^^ are the mean number of flies collected on sticky cards during the fall and spring seasons, respectively. Relative sampling methods applied to L. trifolii larvae were not reliable through time. Regression equations that relate small larvae (<1.5 mm long) sampled on six plants with the total larvae/six plants yielded an R^ value > xiv

PAGE 15

0.50 in 9 out of 12 sampling dates when data sampling the 4th, 7th, and 10th nodes on four randomly selected lateral stems per plant (relative sample) were related to the total number of larvae per plant (absolute sample) . Several biological processes of L. trifolii . female adult longevity and survival rate, oviposition, egg developmental rate, and small (< 1.5 mm) and large (> 1.5 mm) larval developmental rates were numerically described using linear, quadratic and exponential regression equations. Based on regression coefficients (R ) , most of the biological processes were described well by linear equations. Only total number of eggs laid per female and large larval developmental rate were better described by quadratic and exponential equations, respectively. Large larval developmental rate was well described using two linear equations over predefined temperature ranges. A conceptual population dynamics model was developed for estimating subsequent larval populations from observed adult populations and in response to temperature in the absence of mortality and migration. The information generated using the sampling method developed herein for estimating absolute adult leafminer densities would be an input to the model. Descriptions of biological processes in the model are based on parameters for rates of oviposition and development of immatures established in this research. The objective in developing a population dynamics model of XV

PAGE 16

L. trifolii is to predict the dynamics of larval populations from adult populations for timely treatment to prevent economic losses. xvi

PAGE 17

CHAPTER I INTRODUCTION The leafminer fly, Liriomvza trifolii (Burgess) (Diptera: Agromyzidae) , has become an important pest in Florida horticultural and vegetable crops, including tomato, celery and chrysanthemum (Stegmaier 1966) . Due to the frequent use of insecticides, L. trifolii has developed resistance to most organochlorine and organophosphate insecticides used in vegetable crops. Leaf miners have been recurring pests in Florida since the late 1940s due to the wide and general use of DDT (Wolf enbarger 1958) . Although leafminer populations can be controlled by natural enemies (Pohronezny and Waddill 1978) , the use of broad spectrum insecticides to control primary pests on tomatoes has allowed leafminer populations to increase to damaging density levels. As a consequence, the use of broad spectrum insecticides had to be reduced as much as possible, and implementation of other, mainly biological, control strategies became necessary. Therefore, integrated pest management (IPM) programs for tomatoes have been developed in Florida since 1978 (Pohronezny and Waddill 1978) . The 1

PAGE 18

main objective of these programs has been to develop economically, technically and ecologically sound systems of integrated pest management. Leafminers f Liriomvza spp.) attack the foliage of toamtoes (a non-marketable part of the plant) and are, therefore, classified as indirect pests (Ruesink and Kogan 1975) . Unfortunately, the high value of tomato crops reduces the options of pest management tactics (Bottrell 1979) , and chemical control is predominant to prevent fruit damage by other insects (Lange and Bronson 1981) . Implementation of IPM programs for tomatoes in Florida has allowed growers to maintain competitive prices (e.g., in comparison to West Mexican producers during the winter season) by reducing production costs. This reduction has been achieved by using hybrid varieties such as 'FTE-12', 'Duke', and 'Sunny', and fewer pesticide applications (Van Sickle and Belibasis 1985). In 1984/85, pesticide costs ranged from $453.43 to $501.96 per acre (Van Sickle and Belibasis 1985). In 1986/87, they ranged from $472.15 to $511.72 per acre (Taylor and Smith 1987). Advanced research and applications in IPM increasingly employ systems analyses. Three major objectives of system analyses in IPM are: 1) to establish comprehensive information about agroecosystem composition, which includes the crop, pest and beneficial organisms, cultural practices, weather, and management; 2) to predict agroecosystem responses to specific environmantal and man-imposed inputs

PAGE 19

3 as well as behavioral features of these systems; and 3) to select optimal management strategies for crop development and control of pest populations for economical production of crops with high yield and quality (Smerage et al. (1980). Models are essential for expressing system compositions, and mathematical models are used for analyzing system behavior. Mathematical models permit prediction of pest population densities under various environmental and management conditions . Descriptions of biological processes determining the dynamics of pest populations must be known to generate appropriate mathematical models. Data obtained under field and laboratory conditions are used to develop descritpions of those processes such as development, oviposition, parasitism, predation, movement, and insecticide mortality caused by insecticides (Smerage et al. 1980). Sampling techniques have to be developed for generating information on field populations in order to formulate process descriptions and to provide input to population dynamics models. Absolute population densities (i.e. on a unit area or volume basis) are required for process descriptions and input to models. Present sampling techniques (which yield relative population density estimates) for L. trifolii are inadequate to estimate those absolute population densities. Current IPM programs utilize relative sampling methods (i.e. number of insects collected on a per trap basis, or per swing net collection, etc.) to

PAGE 20

monitor a pest for treatment decisions at the lowest cost possible. Since absolute sampling methods under practical conditions are expensive and time consuming, numerical relationships between absolute and relative sampling methods are needed for generating absolute densities from relative estimates for population dynamics models (Luna et al. 1982). The hope is that a reliable numerical relationship can be established between relative sampling and absolute densities for development of valid mathematical models. Adult leafminers are attracted to yellow surfaces (Affeldt et al. 1983), and yellow sticky cards have been used to sample adult leafminers. However, no relationships (relative or absolute) have been established between yellow sticky card catches and adult leaf miner population density. Estimated absolute adult populations could be used to predict and control subsequent larval populations in time to prevent economic damage. The objectives of my studies were as follows: 1) To develop a sampling technique for estimating absolute adult leaf miner density in fields of staked tomatoes. 2) To establish a numerical relationship between yellow sticky card catches (relative method of for sampling leafminer adult population) and the estimated absolute density. 3) To establish similar procedures on small (first and second instar) and large (third instar) leafminer larvae.

PAGE 21

4) To determine parameters of basic population processes such as oviposition, egg development, larval development, and adult survival under laboratory conditions, to generate useful information for developing a population dynamics model.

PAGE 22

CHAPTER II LITERATURE REVIEW Systematics and Distinguishing Characteristics of Liriomyza trifolii The genus Liriomyza (from Greek "lirion" (lily) and "myza" (suck)) was established by Mik in 1894, when he described a leafminer from lilies ( Lilium martagon L.) as L. urophorina . The basal segment of the female ovipositor of this species is nearly as long as the abdomen (Oatman and Michelbacher 1958) . Liriomyza is a very homogenous group, differing little in coloration and structure from the type specimens (Frick 1952) which makes the species in this genus difficult to identify. Males in this genus possess a stridulating organ which consists of a chitinized ridge on the hind femora and a line of scales on the sides of the abdomen (Spencer 1984) . Liriomyza trifolii was first described as Oscinis trifolii (in the family Oscinidae, now Agromyzidae) by Burgess in 1880 from foliage of white clover ( Tri folium repens L.) and has since been confused and synonymized with several agromyzid fly species. The first misidentif ication was made by Coguillett in 1898 when he studied specimens reared from potato at Foristell, Montana and specimens reared from white clover in Washington, D.C. He identified 6

PAGE 23

7 his specimens as Aaromyza diminuta Walker and stated that these were identical to those described by Burgess (1880) (as O. trifolii) and by Riley (1884) (as 0. brassicae) . Thus, 0. trifolii was placed in the genus Aaromyza as A. trifolii . Aldrich (1905) stated that the synonymy created by Coquillett (1898) was unrecognizable. In 1913, Malloch synonymized O. trifolii and O. brassicae with the European species Agromyza pusilla Meigen and doubtfully included Coquillett 's (1898) A. diminuta in his synonymy (Spencer 1981b) . In the same year Melander stated that A. diminuta was synonymized by Coquillett (1898) with 0. trifolii (which Melander (1913) considered as A. scutellata ) and that Coquillett 's description was too brief to allow for proper identification. He placed A. diminuta as a synonym of A. scutellata (which belongs to the genus Metopomy za (Spencer 1981b) ) . O. trifolii and O. brassicae were still treated as synonyms of A. pusilla by Frost (1924) who stated that this latter species was a "decidedly variable species" (p. 51) . De Meijere (1925) described the larva of L. leauminosarum and stated that this species might have been identical to Burgess' (1880) 0. trifolii (=A. trifolii). He rightly pointed out that the name trifolii was not applicable because "Kaltenbach (1874) had earlier cited A. trifolii". Hendel (1938) also noticed this homonymy but stated that this homonymy "was very probable, but not very certain" (p. 214). Spencer (1981b) considered A. trifolii a

PAGE 24

8 synonym of A. nana (Meigen) . Frick (1953) placed A. trifolii as a synonym of L. conaesta and stated that A. trif olii was not a homonym of Kaltenbach's (1874) A. trifolii . creating a secondary homonym and therefore a new synonym. He also created a new synonym by placing L. trifolii as a synonym of L. congesta. In addition, he mentioned that, since the type of A. trifolii did not exist in the U.S. National Museum collection and that it seemed probable that it was the same species that Burgess described, his synonymy would remain valid until the status of A. trifolii would be clarified. A name once placed in homonymy must be rejected, however (article 35 of the International Commision on Zoological Nomenclature (Schenk and McMasters 1936)). Therefore, a new name is needed for the economically important leafminer fly, Liriomvza trifolii (Zoebisch 1984) . Frick (1955) described a new species of Liriomyza as i. alliovora (named the Iowa onion miner) which he synonymized with Liriomvza allia (Frost). Frost (1962) described L. archboldi collected at the Archbold Biological Station in Highlands County, Florida and mentioned that this species was similar to L. trifolii (Burgess). Spencer (1965), after studying the male genitalia of Frost's L. archboldi concluded that it was the same species as L. trifolii. He also discussed the homonymy of Kaltenbach's A. trifolii and the O. trifolii described by Burgess (1880) , and made a wrong conclusion stating that these were not homonyms.

PAGE 25

9 Presently much of the confusion in the correct identification of Liriomyza leafminers has been eliminated by the detailed studies of the male genitalia (Spencer 1981a) and by the use of starch gel electrophoresis techniques to correctly diagnose closely related species (Zehnder et al. 1983). The same technique has proven useful to separate larvae of Liriomyza species (Menken and Ulenberg 1983) . Spencer (1981b) listed two characters that made L. trif olii adults easily distinguishable from a very closely related species, L. sativae Blanchard. First, the color of the mesonotum of L. trifolii is grayish-black and matte; second, the upper orbits and most of the hind-margin of the eye is yellow, with both vertical bristles on yellow ground. The first characteristic, however, may become altered depending on the method of preservation. K. A. Spencer (Dept. Biol. Sci., Univ. Exeter, England) observed that the mesonotum of specimens of L. trifolii preserved in alcohol may become shiny, acquiring a similar appearance to the mesonota of L. sativae . The shiny mesonotum of L. sativae was considered by Spencer (1981b) the most conspicuous difference to separate L. trifolii from L. sativae . The hind-margin of the eye of L. sativae is entirely black and the vertical bristle is on black background. Even the male genitalia are similar in L. trifolii and L. sativae, the curvature of the genitalia of L. sativae being less pronounced than in L. trifolii (Spencer 1981a) . Knodel-

PAGE 26

10 Montz and Poe (1982) studied the female genitalia of L. sativae and L. trifolii and concluded that the egg guide was V-shaped in L. trifolii whereas the egg guide of L. sativae was acutely angled. The denticles in the former species are angular while in the latter they are elongate. The description of the adult is based on a neotype designated by Spencer (1965) from a male reared from alfalfa ( Medicago sativa L.) in Indiana and is stored in the U.S. National Museum. Spencer (1965) described it as follows: Orbits entirely yellow, both vertical bristles on yellow ground; black of occiput reaching margin beyond outer vertical bristle; all antennal segments bright yellow, third only finely pubescent; mesonotum blackish grey, distinctly polinose; acrostichals irregularly in 3 or 4 rows in from, reduced to two rows behind, yellow patch at each corner adjoining scutellum; mesopleuron with black patch normally extending along lower margin, sternopleura largely black, upper margin yellow; abdomen with tergites variably yellow laterally and on hind margins; coxae yellow, femora largely so but with slight, variable brownish striation; tibiae and tarsi darker, brown, (p. 37-38) . Genitalia are illustrated in Spencer (1973) . Larvae of L. trifolii as well as those from the family Agromyzidae can be distinguished by the presence of a pair of anterior spiracles that is located dorsally in contrast to the lateral position that these have in other cyclorrhaphous Diptera (Peterson 1979) . Larvae and pupae of L. trifolii and L. sativae are difficult to separate by conventional techniques. Menken and Ulenberg (1986) characterized the larvae and pupae of these species using starch gel electrophoresis.

PAGE 27

11 The eggs of L. trifolii are microscopic, oval, creamy white, and measure approximately 0.2 x 0.1 mm (Bartlett and Powell 1981) . Eggs of L. trifolii probably increase in size similar to those of L. conqesta (as L. trifolii ) (Dimetry 1971) and L. sativae (as L. pusilla ) (Tilden (1950) after being deposited. Life Cycle and Biology of L. trifolii Biology and Behavior of Adults Adult L. trifolii are small flies with a body length of ca. 1.30 mm and a wing length of 1.5 mm (Burgess (1880)). They emerge with the aid of their ptilinum through the dorsal anterior end of the puparium (Parrella 1987) . Adult flies usually emerge during early morning hours primarily between 0930 h and 12 30 h showing a pronounced diurnal periodicity (Charlton and Allen 1981, Vercambre 1980) . Nevertheless, L. trifolii adults have been collected in light traps (Spencer 1965, Stegmaier 1966). Adult emergence depends in part on relative humidity, the type of substrate, and depth of burial in which the larva pupates. Charlton and Allen (1981) studied adult emergence under various humidity conditions. Very few adults (6%) emerged from newly transformed pupae completely exposed at a low relative humidity (11%) , while at 100% RH, 88% adults emerged. In another experiment they buried pupae about one cm in sand and peat varying the amounts of water added by soil weight. They observed that maximum emergence occurred at 11.4% and 27% added water by weight in sand and

PAGE 28

12 peat, respectively. They also determined that at 25 °C up to 96% of the puparia would survive after being submerged in water for 4 hours whereas none would survive after being submerged for 75 h. Taking additionally the factors of depth burial and soil grain size into account, Getting (1983) determined that the pupal survival was reduced with increasing gravel size and depth of burial in gravel. Adult flies begin activity at sunrise and are most active during midmorning hours (Parrella et al. 1981) . At 26.710.5 "C Parrella et al. (1983b) observed that females oviposited in chrysanthemum leaves within 24 h after being placed in association with males, which indicates that mating also occurs shortly after adult emergence. Mating and oviposition may be influenced by several factors one of which is temperature (chapter VII) . Males of L. trifolii are considered polygamous (Bodri and Getting 1985) . Gatman and Michelbacher (1958) observed multiple matings in L. pictella (now synonymized with L. sativae (Spencer 1968) . Feeding and oviposition behavior of adult female leafminers has been studied quantitatively by Bethke and Parrella (1985). They described the sequence and the probability of events that occurred once the female flies started puncturing the leaf of a host (chrysanthemum or tomato) with their ovipositor to feed on the exudates of the ruptured cells. Their observations can be summarized as follows: upon initiation of a puncture or 'stipple' the

PAGE 29

13 ovipositor is positioned perpendicularly and touching the leaf. It is first thrust rapidly and then more slowly, the change in thrust speed initiating the penetration of the leaf surface. In the majority of punctures, a female moves her abdomen from side to side to create during the slower ovipositor thrusting larger, fan-shaped leaf punctures. When the abdomen is not twisted from side to side, the ovipositor penetrates the leaf surface in a straight back direction to create a smaller tubular-shaped stipple. Females lay eggs in tubular punctures. Occasionally (16% and 21% of the times in chrysanthemum and tomato, respectively) an egg is deposited, almost always in the tubular punctures. After a puncture is completed, the female backs over the puncture to feed on the exudates of the macerated leaf cells. Both puncture types are used for feeding. Musgrave et al. (1978) reported that males of L. sativae do not possess structures to puncture host leaves and they stated that males may feed at wounds produced by the females. Bodri and Getting (1985) stated that males also fed on plant exudates on leaf punctures made by female flies but did not present any data to support this. Zoebisch and Schuster (1987a) did not find increased survival of males provided access to these exudates compared to males not provided access to exudates. Thus, males apparently do not feed on puncture exudates.

PAGE 30

14 Oviposition and Egg Developmental Time Eggs are laid singly and parallel to the surface of a leaf in the leaf mesophyll in the smaller, tubular punctures. On chrysanthemums lower leaves are preferred for oviposition while younger leaves are preferred for feeding (Knodel-Montz et al. 1983). Oviposition of L. trifolii varies from host to host and is greatly influenced by the presence of additional carbohydrate sources. Chandler and Gilstrap (1986a) determined in the laboratory that mated L. trifolii oviposited an average of 17.9±28.8 (MeantSD) on bell peppers during their whole life span at a constant temperature of 24 •C. Leibee (1984) studied the oviposition rate of L. trifolii in celery and concluded that oviposition was highest at temperatures that ranged from 25 "C (total of 159 eggs per female) to 35 "C ( total of 118 eggs per female). He provided the flies with honey, however. Parrella (1984) studied the oviposition rate of L. trifolii on chrysanthemum at five constant temperatures and determined that from 21.1 •C to 32.2 °C the total number of eggs during the whole life span of the flies was highest (188.53-233.91 eggs/female) and that these numbers did not differ statistically. He also provided the flies with honey. Care should be taken when comparing oviposition rates on different hosts since in many studies the flies were provided with honey or other carbohydrate sources. Charlton and Allen (1981) found that honey significantly increased fecundity of L. trifolii .

PAGE 31

15 Numerous weeds have been found to be suitable hosts for L. trifolii. Smith and Hardman (1986) concluded that L. trif olii successfully oviposited on 16 species of weeds common in or around greenhouses in Nova Scotia, Canada. The number of eggs/lOOcm^ of foliage ranged from 1 on Glecoma hederacea L. (Labiatae) (least suitable host) to 32 on Solanum dulcamara L. (Solanaceae) . Zoebisch and Schuster (1987a) determined the fecundity and ovipositional preference of L. trifolii on tomato and three weed species and found that tomato and nightshade (Solanum americanum Mill.) where overall the most suitable hosts for oviposit ion. Tomato-reared flies oviposited an average of 34.6 and 31.8 eggs during their whole life span at 22-26 'C on tomato and nightshade foliage, respectively. Natural sources of carbohydrates such as aphid honeydew could be utilized by adult Xi« trifolii. Zoebisch and Schuster (1987a) determined in the laboratory that aphid honeydew significantly increased oviposition of L. trifolii on tomato foliage even when mated females were exposed to this carbohydrate source for only 24 h. The egg developmental time of L. trifolii is influenced primarily by temperature and not host plant species. Charlton and Allen (1981) reported that the egg developmental time on pink beans ranged from 2 days (at 32.5 •C) to 11.2 days (at 13.8 °C) . In celery, egg developmental times range from 1.99 days (35 'C) to 9.97 days (15 "C) (Leibee 1984). In bell peppers the mean egg developmental

PAGE 32

16 time is 4.2±0.1 (MeanlSD) days at a constant temperature of 24 °C (Chandler and Gilstrap 1986a) . In tomato cv. 'Hayslip' foliage the mean egg developmental times ranged from 1.99 (32 'C) to 11.5 (13.9 'C) days (chapter VII). Larval Development Upon maturity the first stage larva punctures the chorion of the anterior pole of the egg with its mouthhooks to exit the egg (Dimetry 1971) . The larva also exerts pressure to the eggshell to split it at the anterior end and completes development undergoing four molts (Parrella 1987) . Larvae of L. trifolii feed on the leaf's mesophyll moving in one direction creating a linear or ophionome mine (Hering 1951) leaving a trail of fecal material that alternates from one side to another. Larval development, unlike egg development, is strongly influenced by both temperature and host plant (Table 2-1) . Upon completing their development, larvae of L. trifolii exit their mines by cutting a crescent-shaped slit close to or at the apex of the mine and dropping to the ground to seek a dark place to pupate (Leibee 1986) . Larvae exit their mines during early daylight hours (Charlton and Allen 1981) . Pupation seems to be delayed for only a limited amount of time regardless of lighting conditions (Leibee 1986) .

PAGE 33

17 Pupal Developmental Time In contrast to the differences in developmental times observed in larvae of L. trifolii . pupal developmental times are similar regardless of the host in which the larvae developed (Table 2-2) . Equations to calculate developmental rates of pupae from larvae that developed in different hosts are listed in Parrella (1987). Developmental threshold temperatures for puparia are slightly higher than those of the larvae (Table 2-2) . In Italy Siiss et al. (1984) observed puparia diapausing under laboratory conditions at temperatures below 16 'C. Geographical Distribution of L. trifolii Ij. trifolii is one of several species of agromyzid leaf miners that has been spread by man (Frick 1952) . Spencer (1968) stated that the genus Liriomyza is of Holarctic origin. L. trifolii is of Nearctic origin (Spencer 1981b) . Due to the spread by man, L. trifolii is now considered a cosmopolitan species (Minkenberg and van Lenteren 1986) and it has been recorded in North and South America, Europe, Africa, and Asia. It is able to survive in areas with severe periods of sub-zero temperatures, but it survives best in subtropical and tropical regions (Spencer 1973). In America the northern limit of L. trifolii is Ontario (Spencer and Stegmaier 1973) and its distribution ranges as far south as Colombia (Foe and Montz 1981) . Patel (1987) summarized the geographical distribution of L.

PAGE 34

18 Table 2-1. Larval development time of L. trifolii at similar temperatures in foliage of different hosts. Temperature Host plant Development time fdavs) Citation 35.0 30.0 25.0 20.0 8.4 celery 5.36 6.77 7.97 11.98 threshold* Leibee (1984) chrysanthemum Bodri and Getting (1985) 35.0 30.0 25.0 20.0 7.00 7.60 11.57 13.67 pink beans Charlton and Allen (1981) 32.5 30.0 25.0 20.0 N/A® 4.50 5.10 4.70 8.00 threshold tomato Schuster and Patel (1985) 32.2 26.7 21.1 7.8 3.50 4.40 7.10 threshold ^Estimated threshold temperature for development ('C). ®Datum not available.

PAGE 35

19 Table 2-2. Pupal developmental time of L. trifolii at similar temperatures from larvae that developed in foliage of different hosts. Temperature (°C) Host plant Development time (days) Citation 35.0 30.0 25. 0 20.0 10 . 3 celery 6.69 6.76 8 . 37 13.45 threshold® Leibee (1984) 32.2 26.7 21.1 9.0 chrysanthemum 7.25 9.15 14.15 threshold Parrella et al. (1981) pink beans Charlton and Allen (1981) 32.5 30.0 25.0 11.0* 6.90 6.80 8.50 threshold ^Estimated threshold tempertaure for development ('C) using linear regression equations. ^Supposed development threshold temperature.

PAGE 36

20 trif olii . In addition to the countries included in his list, L. trif olii has possibly been reported in Puerto Rico (Perez 1974) . The rapid spread of L. trifolii worldwide is primarily due to failure of quarantine procedures (Parrella and Keil 1984) because of the difficulty in detecting the small oviposition punctures. Alternative Hosts of L. trifolii Spencer (1964) and Stegmaier (1966) consider L. trifolii a polyphagous species. It attacks several crops and has been reported in 148 host plant species in 31 families (Patel 1987). Schuster et al. (1982) surveyed weeds associated with tomato fields in Hillsborough County, Florida. They found that the predominant flora on the perimeters of these fields was different and changed throughout the season. In the spring season of 1982 almost 50% of the leafminer larvae were observed in foliage of ground cherry f Phvsalis sp.) during the month of February, while most of the larvae (ca. 90%) were collected from black nightshade (Solanum nigrum L.) until the end of that season. Throughout the growing season 80-95% of Liriomyza larvae were observed in foliage of black nightshade, ground cherry, and Spanish needle (Bidens bipinnata L.). These authors concluded that weeds might serve as sources of low numbers of Liriomyza spp. in tomatoes. Wolfenbarger (1961) concluded the same when he surveyed weeds associated with tomato and potato fields in

PAGE 37

21 Florida. Zoebisch and Schuster (1987a) determined in the laboratory that downy ground cherry ( Physalis pubescens L. ) was the least suitable host when compared to tomato, American black nightshade ( Solanum americanum Mill.) and common beggar-tick (Bidens alba L. (DC) ) . Stegmaier (1966) and Wolfenbarger (1961) believed that weeds in Florida acted as reservoirs of L. trif olii throughout the year, thus maintaining continuous populations while crops were not in cultivation. Crop Damage and Effects on Yield Parrella (1987) summarized six ways in which a crop might be damaged by Liriomy za spp.: 1) reduction in the aesthetic value of ornamental plants, 2) destruction of young seedlings, 3) reduction of crop yields, 4) transmission of diseases, 5) acceleration of leaf abcission causing scalding of fruit, and 6) causing some plant species to be quarantined. In tomatoes reductions in crop yield, entry of leaf pathogens through mines, defoliation, and reduction in leaf photosynthetic rates have been reported. Reduction in Crop Yield Wolfenbarger (1948) reported in an insecticidal trial that a 'serpentine leaf miner' (possibly L. trifolii or L. sativae) was the most serious pest among other pests like banded cucumber beetles ( Diabrotica balteata) Lee, southern armyworm ( Spodoptera eridania (Cram.), and tomato hornworms

PAGE 38

22 (Manduca sexta (Haw.))* Plots sprayed with parathion yielded an average of 60.2 kg compared to the control plots (13.7 kg/plot) . Jones and Kelsheimer (1963) reported that insecticides applied for control of lepidopterous larvae and leaf miners on tomatoes in Florida affected yields. Plots treated with dimethoate yielded more than plots treated with parathion or azinphos-methyl. They reported that leafminer and larval insecticides affected yields although they did not report any apparent insect infestation on the plants. Poe (1974) indicated that the southern armyworm f Spodoptera exiqua (Hiibner) ) and the tomato fruitworm (Heliothis zea (Boddie) ) were consistently the more damaging pests of staked tomatoes in Florida. He observed L. sativae in his experimental plots but did not ascribe any yield losses to this insect. Levins et al. (1975) studied the effect of diazinon on populations of L. sativae and effects on yield in tomatoes and concluded that leafminers did not reduce yields at the densities present. Effects on yield in tomatoes due to leafminer damage were studied during four seasons in Florida by Schuster and Jones (1976) . Yield (measured as weight of fruit) was significantly increased over the check plots of Walter' tomatoes only in one season when leafminer populations were treated with oxamyl.

PAGE 39

23 Entry of Pathogens and PhYsioloaical Changes due to Larval Mining Keularts (1980) studied the effects of leafmining by L trifolii on tomato as a source of entry for pathogenic organisms. He successfully recovered the pathogens Alternaria alternata (Fries) Keisler and Xanthomonas vesicatoria (Doidge) Dows) from leafmines. These pathogens can cause major defoliation on tomatoes in Florida. Schuster (1978) reported that up to 90% of the foliage on tomato may be lost, if leaf miners are not controlled. Fruit thus exposed to the sun may suffer scalding and moisture loss (Musgrave et al. 1975b). Reduction of Photosynthetic Rates Johnson et al. (1983) determined that photosynthetic rates in mined tissues of tomatoes (hybrid 6718 VF) were reduced 62% by the feeding damage exerted by L. sativae larvae. They concluded that mining injury primarily affected gas exchange thus reducing stomatal conductance in unmined leaf tissues close to the mine. They further suggested that more studies were needed to understand the relationship between leaf miner injury, photosynthetic rates and fruit production physiology in tomatoes. On lima beans Martens and Trumble (1987) found that mature leaves mined by L. trifolii larvae virtually recovered from injury. Damaged cells were replaced by photosynthetically active cells.

PAGE 40

24 Chemical Control of L. trifolii on Tomato Until 1982 there has been confusion regarding the identity of Liriomyza leaf miners studied on tomatoes. Leibee (1981) provided an excellent review of the insecticidal control of Liriomyza leafminers on vegetables until 1981. He concluded that the short effective life of insecticides and the adverse effects that these compounds had on the leaf miner parasite species complex had contributed to several outbreaks of leafminers in vegetable crops. Powell (1981) treated L. trifolii populations in England in greenhouse tomatoes with heptenophos, oxamyl, permethrin, and resmethrin in an effort to eradicate this pest. Tomatoes grown outside greenhouses were treated with DDT and trichlorfon. No tests of insecticide efficacy were performed at that time. Bartlett and Powell (1981) concluded that greenhouse populations of L. trifolii in England could be eradicated but that field populations provided a constant threat of reintroduction. Schuster and Everett (1983a) reported on the effectiveness of abamectin, SD 52618 (5,6-dihydro-2-(aci-nitromethyl)-4H-l,3-thiazine, calcium salt (2:1)), cypermethrin, fenvalerate, encapsulated methyl parathion, chlorpyriphos, methamidophos , and the pyrethroid AC 222,705 against L. trifolii in the field and in the laboratory. They concluded that abamectin and cyromazine were effective in controlling larvae in the field and in the laboratory. Low adult mortality was observed

PAGE 41

25 although abamectin reduced oviposition and feeding. In the laboratory they observed that shortly before or upon emergence, larvae died in foliage treated with abamectin. Of the other compounds none provided satisfactory control and encapsulated methyl parathion at 119.8 g Al/liter resulted in more leaf mines relative to the water check. In another field experiment Schuster and Everett (1983a) collected significantly fewer puparia on leaves treated with cyromazine and abamectin (except in plots treated with 2.27 g ai/378.5 1). As many puparia as those collected from the control plots sprayed with water were collected where insecticides such as methamidophos , azadirachtin, fenvalerate, and the pyrethroid Pay-Off'^ were used. In California, Trumble (1983) reported that the insecticides permethrin, fenvalerate, Bactospeine'', abamectin, and cypermethrin significantly reduced leafminer (L. trifolii and L. sativae) densities when compared to the control plots. L. sativae may have predominated over L. trifolii and may have been controlled controlled by a wider range of insecticides, since L. trifolii was not satisfactorily controlled on tomatoes with fenvalerate in Florida (Schuster and Everett 1983b) . Parrella and Keil (1985) compared the toxicity to methamidophos in four species of leafminers and concluded that i. trifolii reared

PAGE 42

26 from Florida celery was the most tolerant species when compared to the other leaf miner species including L. sativae . Woets (1985) mentioned that L. trifolii was difficult to control in British greenhouses due to its tolerance to many common insecticides. In addition, the disruption of biological control of the greenhouse whitefly by insecticides applied for control of t. trifolii made management of leafminers more difficult. Aldicarb, oxamyl, dimethoate, and heptenophos are among the insecticides listed by this author for the control of L. trifolii. In France, Martin (1984) and Martin and Filliol (1985) reported that fenthion, methomyl, methidathion, dichlorvos and isathrine were used on a prophylactic basis to control L. trifolii by growers in that country. Of the insecticides currently under development against many pests, abamectin, azadirachtin, and cyromazine are the most promising for the satisfactory control of L. trifolii. The active ingredients of abamectin consist of a mixture of avermectin B,a (minimum 80%) and avermectin B,b (maximum 20%) (Brown and Dybas 1982), a macrocyclic lactone isolated from the soil microorganism Streptomyces avermitilis (Burg et al. 1979). It inhibits gamma aminobutyric acid-mediated neuromuscular transmission. The low number of feeding stipples and eggs observed by Schuster and Everett (1983a) in females treated with abamectin may be due to the inhibition of the oviposition functions.

PAGE 43

27 Azadirachtin, natural product obtained from seeds of the neem tree ( Azadirachta indica Juss.)# acts as a feeding inhibitor and/or as an insect growth regulator against many insect species (Warthen 1979) . Although Schuster and Everett (1983a) reported that a water extract of neem seed was not as effective as abamectin against L. trifolii on tomatoes. Webb et al. (1983) determined that significantly fewer eggs were laid by L. trifolii on leaves of 'Henderson Bush' lima beans treated with a water extraction of neem seed. They also reported that larval mortality reached 100% shortly after the eggs hatched. Systemic effects of neem seed extract on L. trifolii larvae were reported in chrysanthemum by Larew et al. (1985). ^ Cyromazine (N-cyclopropyl-l,3,5-triazine-2,4,6triamine) has a mode of action similar to that of an insect growth regulator and possibly acts as a chitin inhibitor (Trumble 1985a) . Larvae treated with cyromazine fail to pupate (Schuster and Everett 1983b) . Resistance of Leafminers to Insecticides Used in Tomatoes In 1957 Genung reported the ineffectiveness of toxaphene for controlling the serpentine leafminer on tomatoes cv. 'Hayslip'. Wolfenbarger (1958) concluded that chlordane, lindane, toxaphene, and aldrin had become ineffective for leafminer control on tomatoes and potatoes in south Florida. Brogdon (1961) stated that leafminers were a more severe and constant problem in southern Florida than in the central and northern parts of the state. At

PAGE 44

28 that time growers were seeking label approval for the insecticide naled to control leafminers on tomatoes. The insecticides azinphos methyl, permethrin, and Lorsban'^ did not significantly reduce the number of leafmines on tomatoes in experimental plots in Bradenton in 1980 (Schuster and Everett 1981) . Schuster and Everett (1983b) determined that under field conditions the insecticides cypermethrin, fenvalerate, microencapsulated parathion, and chlorpyrifos were ineffective against L. trifolii in the Bradenton, Florida area. Using standard probit analysis methods Parrella and Keil (1985) determined that L. trifolii adults from Florida celery were the most tolerant to repeated applications of methamidophos (LDg^ = 10.8 mg/ml, slope = 3.09) when compared to L. trifolii adults collected on chrysanthemum in California {LD^ = 1.93 mg/ml, slope = 2.17) where methamidophos is not used. Resistance of L. trifolii to insecticides has been monitored using yellow sticky cards. Haynes et al. (1986) used these cards with permethrin and chlorpyrifos incorporated at desired concentrations in the sticky material (Tangletrap") in chirysanthemum greenhouses in California. They determined that this method was reliable, simple, and accurate to evaluate insecticide resistance of adult leafminers.

PAGE 45

29 Effects of Insecticides on Parasitoids of Liriomyza Leafminers The impact that insecticides have on leafminer parasitoids varies among crops and geographical areas. Since the use of organochlorine insecticides, increases in leafminer populations have been recorded due to the apparent ability of the leafminers to develop tolerance and/or resistance to synthetic organic compounds. Leafminers have been classified as secondary pests (Pohronezny and Waddill 1978) which means that whenever their natural enemies are eliminated through the use of broad spectrum insecticides their population densities may increase to economically damaging levels. Insecticides can interact with parasites in several ways: 1) have no effect, which means that parasitoids are not affected and would still control the treated pest population; 2) have broad spectrum toxicity which means that pest and natural enemy populations are affected, and 3) have selective toxicity which means that certain natural enemies are effected. Getzin (1960) suggested that leafminer chemical control methods should be combined with biological control methods in which only leafminers are controlled. Chemicals that would allow such pest control strategies have not been available until recently. Parrella et al. (1983a) determined under laboratory conditions that the insect growth regulators cyromazine and Ro 13-5223 had no effect on survival of Chrvsonotomyia parks i , and still 80% of L.

PAGE 46

30 trifolii was controlled. The use of broad spectrum insecticides has been the most common in tomato crops due to the high value of the crop (Bottrell 1979) . In relation to the use of broad spectrum insecticides Shorey and Hall (1963) and Getzin (1960) reported increasing leafminer populations in areas treated with DDT and methoxychlor, dieldrin, and lindane. With the wide use of organophosphate insecticides, leafminer parasitoids responded similarly as to the organochlorine compounds. Oatman and Kennedy (1976) and Johnson et al. (1980a) realized the deleterious effects that methomyl had on leafminer parasitoid populations. They observed significantly higher leafminer densities on tomato plots treated with methomyl. Oxamyl was observed to have similar effects when applied weekly on tomatoes for leafminer control (Schuster et al. 1979). The problems arising with these kinds of compounds rested in their broad spectral activity. Regarding selective toxicities, under certain circumstances, insecticides have been shown to affect some leafminer parasitoid species more than others. Poe et al. (1978) determined that most leafminer parasitoids were reared from tomato foliage treated with a mixture of leptophos and endosulfan in contrast to foliage treated with oxamyl, which yielded the lowest number of parasitoids. They also reported D. intermedius as the most abundant parasitoid in Florida tomatoes.

PAGE 47

31 A biotic larvicide ( Bacillus thurinaiensis Berliner var. kurstaki ) and an ovicide (chlordimeform) were found to be more specific for the control of lepidopterous pests on tomatoes, although the leaf miner parasite Chrysonotomyia punctiventris was also adversely affected by chlordimeform (Johnson et al. 1980a). Zehnder and Trumble (1985a) determined that a higher percentage of Chrysonotomyia punctiventris emerged from organophosphate-treated celery leaf samples while Diqlyphus spp. were more abundant in foliage treated with the pyrethroid permethrin. These trends may vary from one season to another, however. Trumble (1985a) observed that the insecticide abamectin altered the composition of parasitoid species in celery, reducing populations of D. intermedius in 1982, but not in 1983. Schuster and Price (1985) reported that C. punctiventris was more abundant in sprayed tomato plots than Opius sp. and D. intermedius in nonsprayed plots. Schuster (1985) found that Chrysonotomyia was more abundant in tomatoes sprayed with permethrin and methamidophos while Opius sp. and D. intermedius were found primarily in nontreated tomatoes. This indicates that more detailed studies are needed to determine the effects of insecticides on leafminer parasitoids in several crops under various environmental conditions such as greenhouses and open fields.

PAGE 48

32 Biological Control of Liriomyza trifolii Parasitoids of Liriomyza trifolii on Tomatoes Johnson and Kara (1987) presented a comprehensive list of hymenopterous parasitoids of Liriomyza leafminers in North America including Hawaii. Of the 37 parasitoid species listed, 7 were reported attacking Liriomyza on tomato: Diqlyphus beqini (Ashmead) (Eulophidae) , Chrysocharis parksi Crawford (Eulophidae) , and Chrysonotomyia punctiventris (Crawford) (Eulophidae) in California; D. pulchripes and Opius dimidiatus Ashmead (Braconidae) in Ohio; Chrysonotomyia formosa and D. intermedius (Girault) in Florida; and C. punctiventris and D. beqini in Hawaii. Minkenberg and van Lenteren (1986) provided a comprehensive list of the identified parasitoids of L. trifolii reported worldwide. They listed a total of 28 species and mentioned that the biology of only a few was known to some extent. The predominant parasitoids of L. trifolii may change from one season to another or one crop to another. Schuster / (1985) reported that in West Central Florida, Opius sp. and Chrysonotomyia sp. predominated in 1980 and D. intermedius predominated in 1981. Chrysonotomyia sp. again predominated in 1983 and 1984. Zehnder and Trumble (1984a) found that significantly fewer Chrysocharis parksi Crawford were reared from L. trifolii in a celery field adjacent to a tomato

PAGE 49

33 field where L. sativae predominated. D. intermedius was more abundant in the celery field where L. trifolii predominated . Classical biological control of Jj. trifolii provides a fertile field of study. Due to the polyphagous nature of this leafminer species, a thorough knowledge of the leafminer's alternative hosts/crops and the ability of its parasitoids to search for larvae under varied conditions is needed. Before any kind of releases will be effective, preferences of the parasitoids for L. trifolii in different crop habitats must be understood. Johnson and Kara (1987) found that the major parasitoid species of major Liriomyza spp. in North America and Hawaii have been found consistently within a major crop. Also parasitoid diversity may vary when the major leafminer species differ between adjacent crops. Zehnder and Trumble (1984a) found that significantly fewer Chrvsocharis parksi Crawford were reared from L. trifolii in a celery field adjacent to a tomato field where L. sativae predominated. D. intermedius was more abundant in the celery field where L. trifolii predominated. Predators of Liriomyza Leafminers Predators of Liriomyza spp. have been given less emphasis in their role as agents regulating leafminer populations although they have been reported as early as 1913 by Webster and Parks. They listed a mite species in the genus Ervthraeus and a hemipteran (Triphleps sp.) which

PAGE 50

34 preyed on Agromyza pusilla Meigen. Webster and Parks (1913) were working with L. sativae . L. trifolii . L. huidobrensis . and other undetermined species and not A. pusilla . however (Spencer 1981b) . Hemiptera in the genus Nabis and probably also in the genus Geocoris have been reported to prey on larvae of L. sativae on celery (Genung et al. 1978) . Johnson et al. (1980b) observed that green lacewing larvae (Neuroptera: Chrysopidae) preyed on L. sativae larvae collected in pupal trays in fresh market tomatoes in Irvine, California. Chrysopid larvae have also been observed preying on Liriomyza spp. larvae in California (Trumble and Nakakihara 1983). In Colombia, Prieto and Chacon de Ulloa (1980) observed two predators of adult L. trifolii on chrysanthemum (a Diptera: Dolichopodidae and an arachnid in the family Oxyopidae) . They observed a third predator (a small ant in the subfamily Ponerinae) which attacked larvae that had just exited the mines. Freidberg and Gijswijt (1983) observed adult Drapetis subaenescens (Collin) (Diptera: Empididae) feeding in and around greenhouses on adult L. trifolii in Israel. Under laboratory conditions they observed that this empidid lived one month and consumed between 16 and 20 adult L. trifolii per individual predator. Another empidid ( Tachvdromia annulata Fallen) was observed to prey on L. trifolii adults and it preferred leafminers over leafminer parasitoids. These authors suggested that

PAGE 51

35 these predators might suppress leafminer populations satisfactorily but their biology was still unknovm to mass rear them. Effects of Cultural Methods on Populations of Liriomyza Leafminers Webster and Parks (1913) correctly pointed out that the 'serpentine leafminer' was controlled by natural enemies (parasitoids) in the U.S., and that these had prevented this leafminer from becoming destructively abundant. The first cultural control method suggested by them was cutting an alfalfa crop for hay once damage was observed in the plants' foliage. They also proposed deep fall plowing in the East arid regions to bury the leafminer pupae deeply enough to prevent adult emergence. For the western area they suggested controlling the weeds along ditch banks and uncultivated fields to diminish the pupal population which would subsequently hibernate. Musgrave et al. (1976) determined that stripping and trimming celery plants reduced total mine numbers of L. sativae from 90-100% in experimental and commercial plots in Florida. The destruction of crop residues as a means of controlling populations of Liriomyza spp. in Florida to prevent population buildups in fields newly planted with vegetable crops was concurrently suggested by Adlerz (1961) , Brogdon (1961), Kelsheimer (1961), and Wolf enbarger (1961). Wolfenbarger (1961) stated that this was only a partial solution to the leafminer problem in Florida, however.

PAGE 52

Broad-leaved weeds were believed to serve as reservoirs for pests, including L. sativae, initially invading fields planted to vegetable and ornamental crops in Florida (Musgrave et al. 1975b). Therefore, proper sanitation and destruction of any crop residues and broad-leaved weeds would prevent or delay leafminer migration to newly cultivated fields. Vercambre (1980) suggested similar cultural practices in Reunion for preventing L. trifolii from attacking newly planted fields. Wolfenbarger and Moore (1968) found significantly fewer mines of Liriomyza sp. in foliage of tomato plants protected by vertical strips of aluminum than in plants protected by mulches made of plasticized, and paper-backed aluminum in South Florida. They also found significantly fewer mines in cotyledons of squash plants protected with aluminum foil and aluminum scrap when compared to the unprotected (check) plants. Chalfant et al. (1977) also reported fewer L. sativae on brown-paper-mulched yellow summer squash compared to non-mulched squash in Georgia. Contrary to these results, Webb and Smith (1973) found most adults of L. sativae in foliage of snap beans mulched with aluminum foil. Their explanation of these results was based on the demonstration of Oatman and Michelbacher (1958) that leafminers exhibited positive phototaxis and would therefore be attracted to reflective mulches. Price and Poe (1976) observed that polyethylene plastic paper mulched tomatoes

PAGE 53

37 and staked tomatoes supported the largest leafminer populations when compared to only mulched or staked or nonmulched/non-staked tomatoes. A parasite ( Opius sp.) was collected in significantly lower numbers from leafminer pupae collected from staked plants than from non-staked plants which may explain in part the lower densities of leafminers observed in the non-staked plots. Fertilizer levels also influence leafminer damage. Harbaugh et al. (1983) found that damage by L. trifolii increased linearly as leaf nitrogen in chrysanthemum cv. •Manatee Yellow Iceberg' increased from 2.2% to 4.0%. They also indicated that N was the most critical factor correlated with leafminer damage although caution should be taken interpreting the N effect, because the controlledrelease fertilizer formulations result in changes in P and K with any change in N. Sampling Methods Used to Monitor Populations of Liriomyza Leafminers Methods to Estimate Larval Densities Larval densities have been estimated by different methods to determine the effectiveness of insecticides or to study the population dynamics of Liriomyza leafminers in several crops. Oatman and Michelbacher (1958) estimated larval populations of the melon leafminer (L. sativae ; misidentif ied as L. pictella (Thomson) (Spencer 1981a) ) by taking random samples consisting of 20 to 50 leaves from the center of melon plants. They concluded that leafminer populations varied considerably from host to host, field to

PAGE 54

38 field, and even within areas in the same field. Highest leafminer larval densities were found along the edges of the fields during the months of July and August in the San Joaquin Valley in California. Wolfenbarger and Wolfenbarger (1966) established a sequential sampling program to determine the effectiveness of a leafminer control procedure on tomatoes in South Florida. They proposed decision lines to spray or not to spray when 40% and 10% of the leaves sampled at random from a plant averaged one or more mines per leaflet. They considered that the leafminer had been adequately controlled if 10% of the leaves sampled averaged one or had one or more mines. A level of 40% or more of the leaves bearing mines was considered as inadequate control. In celery, Musgrave et al. (1979) determined that random samples of 100 mature petioles per 4.86 ha were statistically precise enough to plot population trends of X*. sativae larvae. Larval densities were estimated counting the maggots in 10 mature trifoliolates per sample per plot. Schuster and Beck (1983) developed a rating system to estimate densities of total leaf mines on tomatoes. This rating system correlated closely with actual counts and reduced the time required by field scouts to assess densities of leafmines later during the tomato growing seasons in west central Florida. Populations of L. trifolii larvae have been monitored in celery using leaflet samples and counting puparia after

PAGE 55

39 3, 7, and 14 days (Foster 1986). Based on indices of dispersion he determined that the leafminer population had an aggregated distribution in coimnercial celery fields. Jones and Parrella {1986a) developed binomial sampling plans on chrysanthemum in two greenhouses and determined that a 100-leaf random sample per 2000 m^ was appropriate to estimate larval densities of L. trifolii. Johnson et al. (1980b) developed a sampling method based on a linear regression relating the number of live medium and large larvae of L. sativae collected on a per leaflet basis (independent variable) and pupae collected on styrofoam trays placed underneath double-row plotted tomato plants (dependent variable) . The coefficient of determination obtained in this numerical relationship was 0.767. This sampling method has not been found useable under Florida field conditions (Schuster, D.J. pers. comm.). In snap beans cv. 'Nemasnap' and 'Eagle' Hanna et al. (1987) developed a sequential sampling plan counting leafmines present on 10 randomly selected trifoliate leaves in each of four plots. To obtain desired sampling precision levels they related the number of leaves required to sample (dependent variable) , and the number of leafmines of i. sativae per leaf (independent variable) . They determined that this sampling plan was robust for type of cultivar, nitrogen fertilizer, and a range of pesticide application frequencies.

PAGE 56

40 Chandler and Gilstrap (1986b) determined with a stratified sampling method that L. trifolii larvae on bell peppers after plants had reached 75 mm in height, mature leaves should be sampled to detect the greatest proportion of the larval populations. Methods to Estimate Adult Densities Some of the first sampling methods of adult leaf miners were developed by Musgrave et al. (1979). They determined that 24 sweep samples taken at random in celery per 4.86 ha provided precise estimates of population trends of L. sativae adults. Adult samples consisted of 10 sweeps with a 38.1 cm diameter sweep net over 6.1 m 7.6 m of celery. They correctly indicated that the development of sampling procedures for pests offered the key to the judicious use of pest management tactics. Price (1982) reported that some chrysanthemum growers in Colombia sampled adult L. trifolii using sweep nets and other growers used D-Vac suction machines to monitor adult L. trifolii populations. A D-Vac suction machine was also used by Trumble and Nakakihara (1983) to monitor adult leafminers in celery. Adult leafminer flies are attracted to wavelengths in the green region (500-540 nm) and the yellow region (540-600 nm) (Affeldt et al. 1983). These intervals correspond to the maximum reflectance of green leaf plants (Shull 1929) . This information has permitted the development of traps consisting of a surface of the proper reflectance coated with sticky substances.

PAGE 57

41 Tryon et al. (1980) collected significantly more adult flies on yellow cards than on yellow-green, orange, green and blue cards. Results obtained by Yudin et al. (1987) on lettuce in Hawaiian farms agree with these results. Musgrave et al. (1975b) stated that adult L. sativae could be detected by trapping them on 7.6 cm X 12.7 cm bright yellow cards coated with some sticky material. They suggested that the number of adults captured on several cards after 24 h would indicate their relative abundance in the field and that samples should be taken at least weekly. Early detection and monitoring of adult leafminers could therefore lead to improved population management through the precise application of control measures. Musgrave et al. (1975a) sampled L. sativae in plots with 17 vegetable garden varieties in north Florida. Yellow sticky card counts indicated that adult leafminers were distributed randomly in the field. Weather conditions tended to influence sticky card counts, particularly windy or rainy periods which reduced the number of individuals collected. No relationships were established between adult counts on yellow cards and subsequent larval population densities. Yellow cards covered with polybutanate have been used in England to monitor adult L. trif olii populations in some nurseries. Powell (1979) indicated that the frequent presence of both Phytomyza sp. and L. trifolii leafminers in greenhouses precluded the confirmation of trifolii by examining the foliage alone. He also mentioned the

PAGE 58

42 difficulty in detecting low densities of L. trifolii using yellow sticky cards. Despite this, these cards proved to be useful when infestations were slight or the foliage of the greenhouse crops was dense and leafminer larvae were therefore difficult to detect (Powell 1981) . Dispersal of adult L. sativae has been studied using yellow traps. Tryon et al. (1980) collected significantly more flies on cards located on the periphery of a transplant production range nearest the prevailing wind and within 34 m of a coimercial tomato farm. Their results indicated that Ij. sativae moved in the direction of prevailing winds and for relatively short distances. Numbers of trapped flies declined when wind gusts during daylight hours were greater than 32 km/h. Jones and Parrella (1986b) determined that in a chrysanthemum greenhouse the average distance flown by female L. trifolii from a known point of release was greater (21.5 m) than that flown by males (18.0 m) . The shape of the sticky cards (square, rectangular, triangular, and circular) had no influence on the number of leafminers (L. trifolii and L. sativae ) collected in different cultivars such as alfalfa, bell pepper, cantaloup, and greenhouse grown chrysanthemums (Chandler 1981) . He collected more males on all trap shapes, and yellow opaque traps attracted significantly more flies than yellow fluorescent translucent cards. Zehnder and Trumble (1984b) also reported that a greater proportion of male L. trifolii and L. sativae were caught on yellow sticky traps in fresh

PAGE 59

43 market tomatoes. They suggested that these biased catches were due to female flies spending more time on leaves during oviposition and that males tended to visit more leaves in search of food and females. Jones and Parrella (1986b) captured significantly more male than female L. trifolii on yellow sticky cards in a chrysanthemum greenhouse. Webb et al. (1985) conducted several experiments under greenhouse conditions to determine the response of L. trifolii and L. sativae to yellow sticky cards. When no plants were present, there were no signs of a sex bias for capture of either species. Nevertheless, they collected a larger proportion of L. trifolii females during 4 of 10 trapping periods in a commercial chrysanthemum greenhouse in Maryland. Chandler (1985) concluded that both sexes of L. trifolii appeared equally responsive to yellow cards placed at different heights in bell peppers. With the exception of 13 of 130 instances, no significant differences between males and females collected on the yellow cards were noted. Where differences were found, more males than females were captured, particularly when trap catches peaked. Sticky cards placed horizontally at different heights in staked tomatoes in Florida have not revealed a spatial preference of Liriomyza spp. leaf miners for a certain plant stratum (Schuster and Beck 1981) . With vertically placed traps, L. trifolii preferred lower plant heights in fresh market tomatoes in California (Zehnder and Trumble 1984b) and bell peppers in Texas (Chandler 1985) .

PAGE 60

44 Sequential sampling plans for monitoring Liriomyza spp. adults have been developed on greenhouse chrysanthemums (Parrella and Jones 1985) and fresh market tomatoes (Zehnder and Trumble 1985b) .

PAGE 61

CHAPTER III PLANT PRODUCTION AND MAINTENANCE OF COLONIES OF Liriomyza trifolii (Burgess) Introduction Methods for rearing Liriomyza spp. leafminers have been described by several authors. The techniques used to maintain L. trifolii colonies to obtain adults for field and laboratory experiments were similar to those used by Ketzler and Price (1982) and Patel (1987) . Due to the polyphagous nature of L. trifolii . several host plants in the families Compositae, Leguminosae and Solanaceae can be used to rear them in high numbers. Due to the suitabilty of tomato for feeding and oviposition of L. trifolii (Zoebisch and Schuster 1987b) , colonies were maintained using tomatoes cv. •Hayslip'. Although other hosts like American nightshade (Solanum americanum Mill.) (Zoebisch and Schuster 1987b) or pink beans (Phaseolus vulgaris L. ) (Charlton and Allen 1981) could have been used to rear high numbers of L. trifolii . tomatoes were used as host plants because field and laboratory experiments were done on tomato plants. Approximately 1000-1500 L. trifolii adults were produced daily using the rearing procedures described in this chapter. 45

PAGE 62

46 Tomato Plant Production Seeds obtained from tomatoes cv. Hayslip' at the Gulf Coast Research and Education Center were planted in Speedling" (Speedling Inc., Sun City, Florida) inverted pyramid cellular styrofoam trays. Each cell was filled with Speedling" Peat-Lite Mix and 3 to 6 seeds were placed about 0.5 cm deep in each cell. Two trays were planted in a screenhouse each Wednesday or Thursday to obtain enough plants to maintain leaf miner colonies. Depending on ambient temperature seed germination took 3 to 5 days. During cold days in winter, warm air was blown in the screenhouse with a liquid propane heater to prevent drops in temperature that significantly may have affected plant growth. During hot days in late spring and summer, a fan was used to circulate the air inside the screenhouse to prevent the temperature from becoming too high (>37.78 °C) . Temperature as well as relative humidity fluctuated considerably in the screenhouse (up to ± 15.56 'C and ±65% RH, respectively). Seedling trays were watered at least once a day. In addition, plants were fertilized using a Hyponex" every Tuesday and Friday with 379 g of 20-20-20 (N, P, K) NutriLeaf'^ (Miller Chemical & Fertilizer Co. , Hannover, Pennsylvania) fertilizer in 10 1 of water to give a 500 ppm N solution. Every Friday 9.29 g/1 of hydrated magnesium sulfate was added to the fertilizer mixture. Once the seedlings had the first two true leaves they were removed from the styrofoam trays and carefully separated and

PAGE 63

47 repotted in white 15.2 cm plastic pots. Four seedlings were placed per pot if the plants were used for colony maintenance and only one seedling per pot if they were used in leafminer development studies. Potted plants were maintained similarly as seedlings. Screenhouse Maintenance Weeds growing underneath benches were removed every two weeks to keep plants free of pests such as armyworms ( Spodoptera spp.), and russett mites ( Aculops lycopersici (Massee) ) . Armyworms and leafminers appeared at such low densities that infested tomato leaflets were removed manually. Plants infested with russet mites (primarily during the spring 1987 season) were discarded and the benches holding these plants were drenched with the acaricide dicofol. Treated benches were not used for two weeks . Plant diseases such as leafmold became a problem by the end of the fall 1986 season and 3.875 1 of the fungicide Bravo'' 500 were applied weekly at 3.87 cc/1 until runoff to control this disease. Establishment and Maintenance of the L. trifolii Colony Adult flies from a colony previously established for ca. 4 years at the Gulf Coast Research and Education Center (GCREC) were used to initiate a new colony. Infested tomato foliage in experimental fields at the GCREC was collected in August 1986 and placed in Tupperware'' Superseal 30 cm x 30 cm X 12 cm plastic storage containers to collect puparia.

PAGE 64

48 These containers had one 6.5 cm diameter organdy clothcovered ventilation hole in each sidewall and two similar holes on the lid. A 0.8 cm mesh hardware cloth elevated 2 cm above the bottom of the plastic containers was used to obtain newly emerged larvae. The bottoms of the containers were coated with a thin layer of Teflon (Fluon AD-1; Northeast Chemical Co., Inc.) to prevent puparia from sticking to the surface. Once adults emerged from these puparia they were added to the colony. This procedure was done when the leafminer colony was increased at the beginning and six months later to maintain a genetically diversified population. Four 7-8 week-old plants were placed in each of two 61 cm X 61 cm X 61 cm oviposition cages (Bioquip'' cat. no. 1452D) equipped with a stockinette sleeve to exchange the plants (every day) and introduce newly emerged adult flies (ca. 200 every other day) . During the fall 1986 season new plants and flies were placed in cages between 0800 h and 0900 h while during the spring 1987 season these procedures were effected between 1600 h and 1800 h. Oviposition cages as well as plants infested with eggs and larvae were kept in a rearing facility maintained between 18 "C and 22 'C. Relative humidity fluctuated between 40% and 100% despite the constant use of an electric humidifier. A 12L:12D photoperiod was maintained with white fluorescent lights. After ca. 6 days (when the larvae were about to exit their mines) the foliage of each tomato plant was clipped

PAGE 65

49 and placed in a Tupperware*^ plastic storage container to obtain puparia. Containers used for the collection of puparia were kept under the same conditions as above. Puparia were collected with a no. 2 camel hair brush 48 to 72 h later to allow the pupal skin to harden and therefore prevent damaging puparia during recollection. About 100-200 puparia were placed in 30 g clear plastic cups. A filter paper strip streaked with honey was placed through the lid of these cups to provide adult leafminers with a carbohydrate source. Adults used for the colony were transferred directly into the oviposition cages located in the rearing room. Care was taken not to release too many adults into the oviposition cages to avoid intraspecif ic competition among larvae (Parrella 1983) .

PAGE 66

CHAPTER IV PREPARATION OF EXPERIMENTAL PLOTS AND CONSTRUCTION OF FIELD CAGES TO SAMPLE L. trifolii ADULT POPULATIONS Introduction Tomatoes in the Palmetto-Ruskin area are grown on sandy soils using staked culture and sometimes subsurface irrigation. The tomato varieties most commonly grown in this area are cv. 'Hayslip' and cv. 'Sunny'. 'Hayslip' is a late, jointless, moderately large-vined, determinate, openpollinated cultivar developed by the Institute of Food and Agricultural Sciences at the University of Florida. 'Sunny* is a midseason, jointed, determinate, hybrid developed by Asgrow. Both varieties are resistant to Verticillium wilt, Fusarium wilt (race 1 and race 2) , and gray leaf spot (Maynard 1987) . Agronomic practices followed during two seasons (fall 1986 and spring 1987) at the Bradenton Gulf Coast Research and Education Center were similar to those outlined by Hochmuth et al. (1988). Preparation of Study Plots Fields were plowed and disced to bury old crop refuse prior to planting. Two weeks prior to transplanting, three lands (each land consisting of a field of 0.2 ha with an irrigation ditch on both sides) were rototilled and prepared 50

PAGE 67

51 for cultivation. Beds were formed and a 18-0-25 fertilizer was spread on the beds in double shoulder bands. Superphosphate, with 36.32 k fritted trace elements per ton was broadcast over the entire bed width. Following these procedures the beds were fumigated with methyl bromide and mulched. Tomato plants were transplanted two weeks later. During the fall 1986 season white polyethylene mulch and during the spring 1987 season black mulch was used as a means of improving moisture and fertilizer conservation, and weed control. Row spacing was 2.74 m and plant spacing 0.46 m. Water was supplied by seep irrigation through ditches located between every four planted rows. The water level was maintained at 38.1 cm to 45.7 cm below the bed surface to properly irrigate the plants. Transplanting Procedures Transplants produced in a multi-cell styrofoam tray purchased from Speedling, Inc. were kept one to two weeks prior to transplanting in the screenhouse. They were watered as few times as possible which made them more resistant to field conditions. Before transplanting they were treated with a mixture of the fungicides Demosan'' 65 WP at 0.908 k a.i./378.5 1 and Truban" 40 WP at 499.4 g a.i/378.5 1. After punching holes every 45.7 cm on the beds transplants were set manually on the rows on September 22 in

PAGE 68

52 1986 and on February 24 in 1987. The transplants had a well developed root ball with the growing medium attached to their roots. Staking and Tying Procedures Tomato plants in the Palmetto-Ruskin area are staked to provide fruits higher in quality, and easier to harvest than ground tomatoes. One hundred and twenty cm long by 2.5 cm diameter wooden stakes were placed between each plant in the center of the bed about 2 to 3 weeks after transplanting. The stakes were driven into the ground with pneumatic hammers. Plants were tied with a string about one month after transplanting. The string was wrapped around each stake and past both sides of the tomato plants to provide vertical support. They were tied 3 times during both seasons. Pest. Disease and Weed Control During the fall 1986 season southern armyworm ( Spodoptera eridania (Cramer) ) was the most prevalent insect pest. Only few tomato pinworms ( Keiferia lycopersicella (Walsingham) ) and tomato fruitworms (Heliothis zea (Boddie) ) were observed. Early blight (Alternaria solani) was the most damaging disease affecting the foliage during the fall 1986 season despite the weekly application of fungicides. This resulted in ca. 50% defoliation by the end of the season. A low incidence of bacterial spot f Xanthomonas campestris pv. vesicatoria ) and target spot f Corvnespora cassicola ) was also observed on lower leaves during the last

PAGE 69

53 four weeks of the fall 1986 season. Weeds growing next to the planted beds and in the irrigation ditches were controlled with paraquat on October 13. A rototiller was used to control weeds on the beds between planted beds on September 23 and October 31. In the spring season of 1987 southern armywonn was again the prevalent pest and few tomato pinworms and cabbage loopers f Trichoplusia ni (Hiibner) ) were observed. Low densities of potato aphids f Macrosiphum euphorbiae (Thomas)) were observed 6 weeks after transplanting but did not increase to damaging levels. There was no significant damage incurred by any bacterial or fungal disease during the spring 1987 season. Weeds were controlled similarly as in the fall 1986 season. Pesticides and fungicides applied during the fall 1986 and spring 1987 season were applied with a high clearance tractor and are listed in table 3-1. Construction of Field Cages to Sample Adult L. trifolii Six 1.8mxl.8mxl.8m outdoor cage frames of 1.3 cm diameter galvanized steel electrical conduit were constructed. Sponge strips (5.08 cm x 5.08 cm x 1.8 m) were supported with duct tape at the base of the frames to seal the cage bottoms. The cage frames were covered with fine (0.05 cm X 0.05 cm mesh) lumite screens (Bioquip Products'' cat. no. 1406D) with bottom edges reinforced with saran tape provided with heavy duty brass grommets every 30.5 cm. In the 1986 fall season the bottom of the screens was secured

PAGE 70

54 with two strips of duct tape to the inner side of the cage frames to prevent any insects from escaping. In the 1987 spring season the strips of duct tape were substituted by strips of 0.05 cm x 0.05 cm mesh lumite strips sewn to the screens and glued to the frames with Liquid Nails . Based on the dimensions of these cages and the plant spacing in the field plots two plants could be enclosed within one cage. Since the plants were grown on raised beds, cages had to be placed on fiberwood boards that supported the cages. Twelve 2.44 m x 1.22 m x 1.27 cm fiberwood boards were painted with gray paint (Scotty's'' no. 52) to protect the boards and to closely match the the color of the sandy soil in the field. A thin masonite board painted gray was nailed on to one side of each fiberwood board with tacks. This board provided a smooth surface which was needed to find collected field specimens. Based on the plant and stake spacing in the field, slots were cut on one long side of the boards so that two plants and five stakes would fit between two boards. Two hinges (with the supporting pins removed) were placed at 0.46 m from the outer margins of the boards on each side to serve as coupling points between two boards. Once the boards were placed on the beds the hinges were held together with nails that kept the boards together and cages were bolted to the boards.

PAGE 71

55 Table 3-1. Insecticides and fungicides applied to control insect pests and fungal diseases on experimental plots during the fall 1986 and spring 1987 tomato growing seasons at the at the Gulf Coast Research and Education Center. Season Pesticide Rate®" Category Date applied Fall 1986 uiunane n opeciaj. ool g/0. A 4 Via na r ungiciae vJCT. . •7 uxunane n opecxax ool g/0 . A H Via na f ungiciae T A Diunane n 22 special ool g/0. 4 Vk ^ na r ungicicie OCu . 21 T.anna^p T. 1/0 4 ha Tri«!Pr't 1 r* 1 rfp Oct Dithane M 22 Special 681 g/0. 4 ha Fungicide Oct. 28 Tribasic CuSO^ 1.8 k/0. 4 ha Bactericide Oct. 28 Lannate L 0.9 1/0. 4 ha Insecticide Oct. 28 Lannate L 0.9 1/0. 4 ha Insecticide Nov. 4 Dithane M 22 Special 681 g/0. 4 ha Fungicide Nov. 4 Tribasic CuSO^ 1.8 k/0. 4 ha Bactericide Nov. 4 Bravo 500 1.4 1/0. 4 ha Fungicide Nov. 10 Lannate L 0.9 1/0. 4 ha Insecticide Nov. 21 Dithane M 22 Special 681 g/0. 4 ha Fungicide Nov. 21 Tribasic CuSO^ 1.8 k/0. 4 ha Bactericide Nov. 21 Bravo 500 1.4 1/0. 4 ha Fungicide Dec. 2 Benlate 454 g/0. 4 ha Fungicide Dec. 2 Lannate L 0.9 1/0. 4 ha Insecticide Dec. 9 Bravo 500 1.4 1/0. 4 ha Fungicide Dec. 9 Spring 1987 Dithane M 22 Special 1.4 1/0. 4 ha Fungicide Mar. 24 Dithane M 22 Special 681 g/0. 4 ha Fungicide Mar. 31 Dithane M 22 Special 681 g/0. 4 ha Fungicide Apr. 6 Bravo 500 1.4 1/0. 4 ha Fungicide Apr. 13 Dithane M-45 681 g/0. 4 ha Fungicide Apr. 21 Bravo 500 1.4 1/0. 4 ha Fungicide Apr. 28 Lannate 2L 2.2 1/0. 4 ha Insecticide May 4 Bravo 500 1.4 1/0. 4 ha Fungicide May 5 Bravo 500 1.4 1/0. 4 ha Fungicide May 12 Dipel 454 g/0. 4 ha Insecticide May 12 Bravo 500 1.4 1/0. 4 ha Fungicide May 19 Dipel 454 g/0. 4 ha Insecticide May 19 Bravo 500 1.4 1/0. 4 ha Fungicide May 26 Dipel 454 g/0. 4 ha Insecticide May 26 Bravo 500 1.4 1/0. 4 ha Fungicide Jun. 2 Lannate 2L 2.2 1/0. 4 ha Insecticide Jun. 2 189.25-378.5 1/0.4 ha of formulation were applied depending on the size of the plants.

PAGE 72

CHAPTER V ESTIMATION OF ABSOLUTE ADULT DENSITIES AND CALIBRATION OF A RELATIVE SAMPLING METHOD FOR ESTIMATION OF ADULT POPULATION DENSITY OF L. trifolii Introduction Integrated pest management (IPM) programs, particularly those utilizing population dynamics models, require accurate, reliable estimates of absolute pest population densities (Marston et al. 1976). Field scouts, however, must be able to produce accurate and consistent relative estimates of population density in a quick and easy manner. Scouts must also be able to take samples at a reasonable cost to growers (Linker et al. 1984). Calibration of the relative-density estimates to absolute-density estimates is essential if scouting data are to be used in predictive management models (Luna et al. 1982). Due to the large size of insect field populations, samples must be taken to estimate their densities. Xi« trifolii adult populations have been monitored on several occasions with yellow cards covered with some sticky material. No attempts to relate sticky card catches to actual adult densities have been made, however. Therefore, no predictions of larval densities exceeding action thresholds on tomatoes can be made using sticky card catches 56

PAGE 73

57 of adult leaf miners. Under the conditions prevalent in staked tomato cultivars in central Florida, estimating adult Liriomyza populations to predict larval populations using yellow sticky cards is more practical and feasible than taking samples with a D-Vac'^ or a sweep net. Zehnder and Trumble (1984b) studied the spatial and diel activity of L. trifolii and L. sativae adults using yellow sticky cards in fresh market tomatoes in California. They established a relationship between puparia collected in styrofoam pupal trays and adults trapped two weeks later and concluded that the pupal counts provided a suitable tool for forecasting adult leaf miner population sizes. They concluded that most leafminers were collected on sticky traps at lower or middle plant heights. Field cages were used to develop calibration equations to determine absolute densities of L. trifolii adults in staked and mulched tomatoes using sticky card counts from the field population and counts on known caged populations. To study the influence of sticky card position in relation to plant height, samples at three plant heights were taken during the spring 1987 season. Materials and Methods Sampling Method on the Unknown Field Population Simple random samples of adult L. trifolii were taken with 12.7 cm X 7.6 cm yellow sticky cards (Sticky Strips'^; Olson Products, Medina, Ohio) during the fall 1986 and spring 1987 seasons. During the fall 1986 season sampling

PAGE 74

58 sites were selected at random using a random number table and during the spring 1987 season random numbers were generated using a computer program. The first sample during both seasons was taken ca. 6 weeks after transplanting. The fiberwood boards were placed underneath two plants during the afternoon, removing the adjacent plants, to provide a basis for the field cages. Three wooden 30.5 cm garden stakes were stapled together longitudinally. Two small binder clips were glued to the tips of the garden stakes to hold the sticky cards on each side of the row. These stakes were stapled at the middle height of the stake between the two tomato plants enclosed in a field cage. The total length of the stapled garden stakes was 91.5 cm. At the same time, another set of similar stakes was stapled at the middle plant height between two plants located about 2 m from the fiberwood boards. The sticky cards were always placed on the east side of the cages to prevent shading effects of the cage in the the morning. Twenty four hours later the cages were carefully placed on the boards and the yellow sticky cards were fixed to the binder clips. At the same time a pair of sticky cards were similarly affixed outside the cages. Prior to placing the sticky cards inside or outside the cages they were sprayed with a pressurized insect trapping adhesive (Tangle-Trap", Tanglefoot Co., Grand Rapids, Michigan) to make sure that their entire surface was sticky. Although the sticky cards had a

PAGE 75

59 thin layer of adhesive from the factory, some parts, particularly the edges, were lacking sufficient adhesive material . On the next day, the sticky cards were recovered and the cages were sprayed with Prentox'^ EC (a mixture of 1.2% pyrethrins, 9.6% piperonyl butoxide, 81.2% petroleum distillates, and 8% inert ingredients) at a rate of 8.6 ml/1 with a Solo*^ model 423 back pack mist blower to kill the flies inside the cages that were not trapped on the sticky cards. The mist blower was operated at full thrust while spraying the field cages. Cards next to each cage were also collected on that day. After ca. 30 minutes, flies killed by the insecticide were collected from the fiberwood base with an aspirator made of a flexible rubber tube and a glass eye-dropper with a peace of organdy cloth as a filter. This aspirating technique was performed to make absolute estimates of the population in case a poor relationship between sticky card counts inside and outside the cages was found. These data would also be used to adjust the equations generated for the absolute density estimation based on sticky card counts. Sampling Method on the Known Leaf miner Population Twenty four hours after sampling the unknown population, an equal number of one-day-old, unfed male and unfed/unmated female L. trifolii adults was introduced into each field cage. Flies from the laboratory colony were introduced into 30 g transparent plastic cups for their

PAGE 76

60 subsequent release in the field cages. The plastic cups were placed between the plants on the boards inside the cages. Two new sticky cards were clipped on to the stake before the flies were released inside each cage. A day later, the sticky cards were collected and each cage was sprayed again with Prentox''. Dead flies were collected as before. During the fall 1986 season, 5 pairs of L. trifolii were introduced into each cage in the first week of sampling when the average plant height was 85.5 cm. Every week thereafter an additional 5 pairs were released per cage. It was possible to take samples for 6 weeks. At the end of the growing season only two pairs were introduced per cage to determine if at low densities the adults could be still collected with the sticky cards. Since a very low proportion was trapped at this density these data were not taken into consideration for statistical analyses. During the spring 1987 season the number of flies introduced into the cages was increased as plant height was measured. Thus, 5 pairs were introduced when plants measured an average height of 51.2 cm, 13 pairs when plants measured an average height of 72.9 cm, 24 pairs when plants measured an average height of 96.8 cm, 33 pairs when plants measured an average height of 120.4 cm, 40 pairs when plants measured an average height of 126.0 cm, and 45 pairs when plants measured an average height of 127.6 cm.

PAGE 77

61 Influence of Sticky Card Position on Leafminer Catches During the spring 1987 season samples of the field population were taken using yellow sticky cards placed at three heights. The positions of the cards were defined as low, medium, and high in relation to plant height. Low cards were placed at a height equivalent to one quarter plant height, medium cards were placed at a height equivalent to one half of the plant height, and the high cards were placed just above the plant canopy. Six pairs of sticky cards for each height category were placed at random in the field. Samples were taken twice a week for a total of 6 weeks. Determination of Insecticidal Effects on Leafminers Treated Inside Field Cages To insure that the short residual pyrethrin insecticide had no adverse effects on the flies introduced into the cages, two laboratory experiments were completed. In the first experiment, two pairs of one-day-old, unfed flies were introduced in 30 g plastic cups with a small 2.5 cm x 2.5 cm piece of lumite screen that had been treated with Prentox'' and had been exposed to the sun for 24 h. A total of 7 cups was prepared, using another 7 cups with an untreated piece of lumite as a control. The cups were placed in a rearing room with fluorescent lamps. The second experiment was similar to the first except that the flies (2 one-day-old unfed pairs) were introduced in clip cages (Zoebisch 1984) holding a small piece of lumite inbetween the lids. This increased the probability of contact with the treated

PAGE 78

62 surface. A total of 7 clip cages with treated lumite pieces (as in the previous experiment) and another 7 clip cages with untreated lumite screens were used. Mortality was recorded 24 h and 48 h later. Results and Discussion Generation of the Calibration Ecfuations on the Known Po pulation Calibration equations were developed to establish a numerical relationship between the number of flies collected on yellow sticky cards inside the cages and the known number of introduced flies into a field cage on a per two plant basis. To estimate unlcnown populations on a per two plant basis these equations would be used to calibrate absolute density estimates. These equations were generated regressing the known number of flies introduced into the cages against the number of flies trapped on the sticky cards inside the cages. For the fall 1986 season the following equation was obtained: FI^ = 7.109 + 1.765 FIp (R^ = 0.87) (1) where FI^ equals the known number of flies introduced per cage and FI^ equals the number of flies collected on the sticky cards. During the spring 1987 season the following equation was obtained: SI^ = 5.886 + 1.896 SI^ (R^ = 0.93) (2) The terminology for the variables is the same as that of equation (1) .

PAGE 79

63 Generation of the Calibration Ecmations on the Field Population Relating the number of flies trapped outside cages and the number of flies trapped from the unknown population inside the cages was done to establish a relationhip between flies trapped on a per two plant basis and an unknown area. This numerical relationship would be combined with the relationship obtained on the known population to estimate an absolute density of flies collected on yellow sticky cards on a per two plant basis of an unknown field population. Linear regression analyses were therefore performed on counts of flies from the unknown population trapped on the sticky cards outside and inside cages. Linear equations relating number of flies collected outside the cages as the independent variable and flies collected inside as the dependent variable were developed. The linear regression equation to relate catches of adult flies inside and outside cages for the fall 1986 season is: Fly = 0.704 + 0.103 F0„ (R^ = 0.66) (3) where FI^ is the number of flies trapped inside the cages, and FOy is the number of flies trapped outside the cages. For the spring 1987 season the following equation was determined: Sly = 2.046 + 0.256 SO^ (R^ = 0.78) (4) Dependent and independent variables have the same meaning as those in the equation for the fall 1986 season. A Student t-test was used to search for differences in slope coefficients between equations (1) and (2) and (3) and

PAGE 80

64 (4) to determine if data from both seasons could be pooled. The t-value to compare the slope coefficients was computed using the following formula (Snedecor and Cochran 1967) : t = (b, b2)/y(s,VsS(F0J + (S2VSS(S0J) where b, and bg are the slope coefficients, s^^ and are the variances obtained from the standard error estimates of the slope coefficients and SS(FOy) and SS(SOy) are the sum of squares of the model term obtained in the General Linear Models Procedure using SAS (SAS Institute 1986) . The number of degrees of freedom (df) used to search the t-value in a table was computed by the following formula (Snedecor and Cochran 1967) : df = (n, 2) + (ng -2) where n, and n2 are the number of samples taken throughout the fall and spring seasons (36 during each season) . Significant differences of the t-values at a P<0.05 level were found when comparing the slope coefficients of equations (1) and (2) and (3) and (4) ; therefore they must be used separately according to the season (fall or spring) to determine absolute densities. Generation of Calibration Equations to Estimate Absolute Densities Based on Relative Sampling Data Assuming that FIy=FIj. (or 81^=81^.) , the estimation of absolute adult leafminer density on a per two plant basis involves the creation of an equation for each season substituting the independent variable from the equation of the known population (equations (1) or (2)) by the equation of the unknown population (equations (3) or (4)). In

PAGE 81

65 general terms this procedure involves the following steps, using equations obtained for the fall season: FI, = bo + b,FO„ and FI^ = bo' + b/FI, Using the above equations and assuming that FI^ = FI^ the following equation is obtained: FI, = bo' + b/Ibo_jL^iFQ^ where bo' + b, ' is the adjustment from trap catches per two plants to calibrate between the actual number of adult flies per two plants versus the number of flies collected on yellow sticky cards, and Cbp + b ^ FO ^I adjusts yellow sticky card catches from an unknown area to catches on a per two plant basis. For the fall, FI^^ (solved from equation (3)) would be susbstituted into equation (1) and for the spring, SI, (solved from equation (4)) would be susbstituted into equation (2) . For estimating the average number of adult leafminers per two plants the following equations were obtained: F,^ = 8.352 + 0.182 F^ (5) S,^ = 9.765 + 0.485 S^^ (6) where and S^^ are the absolute average densities estimated for the fall and spring seasons, respectively, and Fgg and S^^ are the mean number of flies collected on sticky cards during the fall and spring season, respectively. Validation of Predicted Adult Leafminer Densities To determine how well the equations would predict an unknown population of adults based on sticky card catches.

PAGE 82

66 predicted values were compared to estimated field density for the fall (Fig. 5-1) and spring (Fig. 5-2) seasons. To do this the following procedure was developed: a) from the data obtained from sampling the known population an average proportion of the aspirated flies was computed to determine what proportion of the total number of flies inside a cage was not collected on the sticky cards (i.e. (females + males) divided by the total known number introduced per cage) , and b) the number of flies collected by aspiration was adjusted using the value of the mean proportion obtained from the known population (i.e. observed value from the aspirated unknown population divided by the mean proportion of the known population) to compute the means and confidence intervals (95%) for the aspirated unknown population. For the fall season all values obtained to determine an average density of flies on a per two plant basis using equation (5) were within the 95% confidence intervals of estimated field densities of mean number of flies per two plants (Fig. 5-1) . For the spring season the first two mean number of flies per two plants estimates computed with equation (6) were above the upper limit of the corresponding 95% confidence limits (Fig. 5-2) . The two points overestimated density at population levels below 5 flies/2 plants. The mathematical model developed for the absolute adult fly density estimation is appropriate for use because 10 out of 12 estimates were obtained within the 95% confidence

PAGE 83

67 40 SB 90 Ei I! IB 10 s r S O COLLECTED MEAN 4 SAMPLING DAIES A PREDICTED MEAN Fig. 5-1. Mean ±95% C.I. adult flies of trifolii collected and predicted on a per two plant basis during the fall 1986 season.

PAGE 84

68 0 2 O COLLECTED MEAN SAMPLING DATES A PREDICTED MEAN Fig. 5-2. Mean ±95% C.I. adult flies of L* trifolii collected and predicted on a per two plant basis during the spring 1987 season.

PAGE 85

69 intervals. The two misses are most likely at densities below treatment levels. To make a decision, as an example, suppose that an average of 20 flies were collected on a per sticky card basis similar to that used to develop the model during the tenth week of the fall season. Substituting the value obtained from the field collection into equation (5) , the following estimate of flies per two plants would be obtained: F^^ = 8.352 + 0.182 * 20, which equals 11.992 (=12 rounded to a whole number) flies per two plants. Proportion and Number of Females and Males Collected on the Sticky Cards During Relative Sampling and Absolute Sampling Procedures Since female flies cause part of the damage to the crop by stippling and ovipositing inside the foliage, numerical analyses should be done only for females. To separate females from males while counting adults trapped on sticky cards in the field is quite time consuming and under most cases a stereoscope is needed. Therefore, the average proportion of females to males per sampling date was computed to determine if this proportion remained constant through time (Tables 5-1 and 5-2) . During the 1986 fall season the mean proportion female/ (female + male) of the unknown field population outside and inside the cages, and the known population inside the cages varied up to 2.83, 1.69, and 4.9 times among sampling dates throughout the season, respectively. This variation has no apparent pattern through time (Table 5-1) . During the spring season of 1987 the mean proportion from the unknown field

PAGE 86

population outside and inside the cages, and the known population inside the cages varied up to 2.33, 4.43, and 1.12 times among sampling dates throughout the season with no apparent pattern through time (Table 5-2) . Therefore, the data of males and females obtained during the sampling procedure were pooled to generate equations that represent the estimation of the total number of flies collected on a per plant basis. Although proportions of females and males collected on yellow sticky cards are quite variable and not statistically different, the population sex ratio is approximately 1:1 (Zehnder and Trumble 1984a) who determined the sex ratio populations of L. trifolii based on larvae collected from foliage of celery and tomato. They collected a larger number of males in fresh market tomatoes on sticky cards, however (Zehnder and Trumble 1984b) , which coincides with the results obtained in this study. In contrast, Webb et al. (1985) collected an equal number of males and females on yellow sticky cards in a greenhouse with no plants. In a commercial chrysanthemum greenhouse in Baltimore they collected significantly more females than males during 4 out of 10 trapping periods. During the rest of the sampling periods the sex ratio was close to 1:1. Numerical analyses were done to compare number of flies collected on north and south sides. During the fall 1986 season, sticky cards on the south row side were exposed to the sun while those on the north side were not exposed.

PAGE 87

71 Table 5-1. Proportions of females/ (females+males) collected on the sticky cards on the unknown population (inside and outside cages) and on the known population inside the cages during the fall 1986 season. Sampling Mean proportion (9/(9+cf)) date Unknown population Unknown population outside cages inside cages Nov 16 0.167 ± 0.418 a® 0.292 ± 0.213 a Nov 22 0.066 ± 0.021 a 0.292 ± 0.100 a Nov 28 0.134 ± 0.039 a 0.403 ± 0.141 a Dec 4 0.059 ± 0.031 a 0.292 ± 0.187 a Dec 11 0.124 ± 0.016 a 0.492 ± 0.067 a Dec 16 0.114 ± 0.040 a 0.417 ± 0.138 a U . Ill X 0 . 017 0 . 365 X 0 . 035 Known population inside cages Nov 18 0.617 ± 0.173 a Nov 24 0.126 ± 0.046 a Nov 30 0.302 ± 0.056 a Dec 6 0.188 ± 0.024 a Dec 13 0.286 ± 0.026 a Dec 18 0.297 ± 0.030 a 0.303 ± 0.069 * ®Data transformed to arcsine /k; (x=proportion (9/9+cf) ) prior to analysis but presented in the original scale. Means followed by the same letter vertically are not significantly different (P<0.05, ANOVA) . Data are presented as mean ± S.E. Overall mean ± S.E.

PAGE 88

72 Table 5-2. Proportions of females/ (females+males) collected on the sticky cards on the unknown population (inside and outside cages) and on the known population inside the cages during the spring 1987 season. Sampling date Mean proportion (9/(9+cf)) Apr 30 May 7 May 14 May 21 May 28 Jun 4 Unknown population outside cages 0.056 0.054 0.060 0.126 0.078 0.099 0.079 + + + + + + + 0.056 0.035 0.021 0.037 0.037 0.012 0.012 @ Unknown population inside cages 0. 167 0.117 0.114 0. 155 0.040 0.177 0.128 + + + + + + + 0.114 0.053 0.030 0.065 0.036 0.027 0.021 a a a a a a # Known population inside cages May May 2 9 May 16 May 23 May 30 Jun 6 0.192 0. 197 0.178 0.194 0.199 0.181 0.190 0.045 0.039 0.012 0.027 0.020 0.019 0.009 a a a a a a # ®Data transformed to arcsine Jy.; (x=proportion 9/9+cf) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, ANOVA) . Data are presented as mean ± standard error. Overall mean ± standard error.

PAGE 89

73 Although proportions of females and males collected on yellow sticky cards are quite variable and not statistically different, the population sex ratio is approximately 1:1 (Zehnder and Trumble 1984a) who determined the sex ratio populations of L. trifolii based on larvae collected from foliage of celery and tomato. They collected a larger number of males in fresh market tomatoes on sticky cards, however (Zehnder and Trumble 1984b) , which coincides with the results obtained in this study. In contrast, Webb et al. (1985) collected an equal number of males and females on yellow sticky cards in a greenhouse with no plants. In a commercial chrysanthemum greenhouse in Baltimore they collected significantly more females than males during 4 out of 10 trapping periods. During the rest of the sampling periods the sex ratio was close to 1:1. Numerical analyses were done to compare number of flies collected on north and south sides. During the fall 1986 season, sticky cards on the south row side were exposed to the sun while those on the north side were not exposed. During the spring 1987 season sticky cards on both row sides were exposed to the sun. During the fall 1986 season significantly more flies were collected on the south (sunny row side) during the last 4 sampling dates on the unknown population outside and inside cages and during all sampling dates, except the first one, inside cages of the known population (Table 5-3). During the spring 1987 season, significantly more flies from the unknown population inside

PAGE 90

74 cages were collected only on the last sampling day (June 4) on the north side (Table 5-4) . From the known population during the spring 1987 season significantly more flies were collected on the south side on the first sampling day (May 2) , and on the second sampling date (May 9) significantly more flies were collected in the north side. In most cases (primarily when the population densities started increasing) significantly more males were collected on the sunny (south) side (Tables 5-5 and 5-6) . Females seemed to have responded more uniformly to any light intensity (Tables 5-5 and 5-6) . Few significant differences between the north and south sides were obtained throughout the spring season (Tables 5-7 and 5-8) . Some of the factors that influence the skewed sex ratio obtained on yellow sticky cards may be that males spend more time searching for females and therefore encounter yellow cards more often or they are more attracted to these cards than are females. Males responded more uniformly to cards exposed to the sun which indicates that they are more sensitive to the intensity of light reflected on the cards than females. This response may function as a primary orientation cue for males to search for females. In addition, males seem to be able to discern between light intensities under varying conditions. During most sampling dates in the fall 1986 season, significantly more males were collected on the sunny side

PAGE 91

75 Table 5-3. Mean number of flies of unknown population outside and inside cages and known population inside cages collected in the south (sunny) and north (shady) side of the tomato field rows sampled during the fall 1986 season. Sampling date South side North side Unknown population collected outside cages Nov 13 0.667 ± 0.333 a*^ ^ 0.333 ± 0.211 a Nov 20 17.167 ± 3.799 a 9.833 ± 2.496 a Nov 27 6.833 ± 3.825 a 9.500 ± 1.088 b Dec 4 13.667 ± 2.629 a 3.167 ± 1.545 b Dec 11 16.167 ± 7.039 a 0.667 ± 6.380 b Dec 16 18.333 ± 3.084 a 4.500 ± 1.708 b 22.889 ± 7.299 * 7.250 ± 2.333 * Unknown population collected inside cages Nov 16 1.000 ± 0.447 a® 0.500 ± 0.224 a Nov 23 2.000 ± 0.449 a 1.000 ± 0.365 a Nov 30 4.167 ± 0.792 a 1.333 ± 0.333 b Dec 7 1.500 ± 0.428 a 0.333 ± 0.211 b Dec 14 5.667 ± 0.989 a 2.000 ± 0.577 b Dec 18 2.667 ± 0.715 a 0.833 ± 0.307 b 2.834 ± 1.279 * 0.999 ± 0.247 * Known population collected inside cages Nov 16 2.500 ± 0.671 a® 1.000 ± 0.517 a Nov 23 5.167 ± 0.401 a 2.500 ± 0.500 b Nov 30 9.167 ± 0.946 a 3.333 ± 0.882 b Dec 7 15.833 ± 0.946 a 5.500 ± 0.882 b Dec 14 13.667 ± 1.202 a 8.333 ± 1.542 b Dec 18 18.333 ± 1.085 a 9.500 ± 1.648 b 10.778 ± 1.018 * 5.028 ± 1.284 * ®Data transformed to (yx+0.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test) . Data are presented as mean ± S.E. Overall mean ± S.E.

PAGE 92

76 Table 5-4. Mean number of flies of unknown population outside and inside cages and known population inside cages collected in the north and south side of the tomato field rows sampled during the spring 1987 season. Sampling date South side North side Unknown population collected outside cages Apr 30 0. 667 + 0.333 a® 2.000 ± 0.632 a May 7 6. 667 + 1.022 a 5.833 + 1.740 a May 14 10. 000 + 4.251 a 11.000 + 3.425 a May 21 9.833 + 1.580 a 9.333 + 2.076 a May 28 2 . 667 + 1.116 a 3.167 + 0.946 a Jun 4 38.333 + 4.958 a 31.167 + 7.463 a 11.361 + 1.513 * 10.417 + 4.097 * Unknown population collected inside cages Apr 30 0.833 + 0.307 a® 0.833 + 0.447 a May 7 1. 667 + 0.494 a 2.333 + 0.667 a May 14 4 . 500 + 0.671 a 4.167 + 1.195 a May 21 4.167 + 0.872 a 7.000 + 1.525 a nay Z o 0.833 + 0.477 a 1.500 + 0.719 a Jun 4 7.500 + 0.764 a 11.500 + 1.335 b 3.258 + 1.071 * 4.556 + 1.658 * May 2 4.000 + 0.517 b® 0.833 + 0.163 a May 9 1.667 + 0.615 a 9.617 + 0.401 b May 16 12.167 + 1.778 a 10.833 + 1.600 a May 21 13.667 + 1.977 a 14.500 + 2.337 a May 28 17.333 + 1.430 a 21.667 + 1.109 a Jun 4 22.833 + 2.762 a 21.333 + 2.104 a 11.945 + 3.263 * 13.056 + 3.236 * ®Data transformed to (yx+0.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test) . Data are presented as mean ± S.E. 'overall mean ± S.E.

PAGE 93

77 Table 5-5. Mean number of females and males of unknown population outside and inside cages collected in the north and south side of the tomato field rows sampled during the fall 1986 season. Sample date South side North side Unknown population collected outside cages (Female Nov 13 0.000 ± 0.000 a® 0. 167 + 0.167 a Nov 20 1.167 ± 0.447 a 0.833 + 2.069 a Nov 27 4.500 ± 0.619 b 1.333 + 0.422 a Dec 4 1.000 ± 0.872 a 0. 167 + 0.167 a Dec 11 4.833 ± 0.872 a 3.167 ± 0.307 a Dec 16 1.833 ± 1.447 a 1. 167 ± 0.477 a 2.194 ± 0.818 * 1.139 ± 0.452 * Unknown population collected outside cages (Males) Nov 13 0.667 ± 0.333 a 0.167 + 0.167 a Nov 20 16.170 ± 3.719 a 9.000 + 2.066 a Nov 27 32.333 ± 2.249 b 8.167 + 1.327 a Dec 4 12.667 ± 2.268 b 3.000 + 0.683 a Dec 11 45.833 ± 6.321 b 13.000 + 2.581 a Dec 16 16.500 ± 1.918 b 3.333 ± 1.520 a 20.695 ± 6.509 * 6.111 + 1.939 * Unknown population collected inside cages (Females Nov 13 0.333 ± 0.333 a 0.167 + 0.516 a Nov 20 0.667 ± 0.333 a 0.500 + 0.224 a Nov 27 2.167 ± 1.880 a 0.667 + 0.211 a Dec 4 0.667 ± 0.422 a 0. 167 ± 0.167 a Dec 11 2.667 ± 0.667 a 1.167 ± 0.307 a Dec 16 1.500 ± 0.500 a 0.500 ± 0.224 a 1.334 ± 0.383 * 0.528 ± 0.152 * Unknown population collected inside cages (Males) Nov 16 0.667 ± 0.211 a 0.333 + 0.211 a Nov 23 1.333 ± 0.211 a 0.500 + 0.224 a Nov 30 2.000 ± 0.577 a 0.667 + 0.211 a Dec 7 0.833 ± 0.307 a 0.167 + 0.167 a Dec 14 3.000 ± 0.632 b 0.833 + 0.478 a Dec 18 1.167 ± 0.305 b 0.333 ± 0.211 b 1.500 ± 0.355 * 0.472 + 0.010 * ®Data transformed to (yx+0. 5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test) . Data are presented as mean ± S.E. Overall mean ± S.E.

PAGE 94

78 Table 5-6. Mean number of females and males of known population inside cages collected in the sunny and shady side of the tomato field rows sampled during the fall 1986 season. Sample date South side North side Known population collected inside cages (Females) Nov 16 1.167 ± 0.307 a® 0.167 ± 0.167 a Nov 23 0.833 ± 0.307 a 0.167 ± 0.167 a Nov 30 2.833 ± 0.872 a 1.000 ± 0.258 b Dec 7 2.000 ± 0.516 a 2.167 ± 0.792 a Dec 14 4.667 ± 0.558 a 1.667 ± 0.211 b Dec 18 4.500 ± 0.619 a 3.833 ± 1.327 a 2.667 ± 0.669 * 1.500 ± 5.569 * Known population collected inside cages (Males) Nov 16 1.333 ± 0.307 a 0.833 ± 0.543 a Nov 23 4.333 ± 0.211 a 2.333 ± 0.422 b Nov 30 6.333 ± 0.558 a 2.333 ± 0.843 b Dec 7 13.833 ± 0.871 a 3.333 ± 0.803 b Dec 14 9.000 ± 0.857 a 6.667 ± 1.520 a Dec. 18 13.833 ± 0.980 a 5.667 ± 1.358 b 8.111 ± 2.079 * 3.528 ± 0.905 * ®Data transformed to (Jx+oTsj prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test) . 'overall means as mean ± S.E.

PAGE 95

79 Table 5-7. Mean number of females and males of unknown population outside and inside cages collected in the north and south side of tomato field rows sampled during the spring 1987 season. Sample date South side North side Unknown population collected outside cages (Female Apr 30 0.167 ± 0.167 a® 0.000 + 0.000 a May 7 0.167 ± 0.167 a 0.167 + 0.167 a May 14 0.667 ± 0.494 a 0.667 + 0.422 a May 21 1.167 ± 0.543 a 1.500 + 0.619 a May 28 0.167 ± 0.167 a 0.500 + 0.342 a Jun 4 3.333 ± 0.615 a 3.667 + 0.211 a 0.945 ± 0.505 * 1.084 + 0.559 * Unknown population collected outside cages (Males) Apr 30 0.500 ± 0.224 a 0.167 + 0.632 a May 7 6.500 ± 1.148 a 5.667 + 1.801 a May 14 9.333 ± 3.827 a 10.333 + 2.449 a May 21 8.667 ± 1.358 a 7.833 + 1.621 a May 28 2.500 ± 0.992 a 2.667 + 0.667 a Jun 4 35.000 ± 4.817 a 27.500 + 7.451 a 10.417 ± 5.098 * 9.028 + 3.997 * Unknown population collected inside cages (Females) Apr 30 0.667 ± 0.167 a 1. 167 + 0.167 a May 7 0.333 ± 0.211 a 0.157 + 0.167 a May 14 0.500 ± 0.224 a 0.500 + 0.224 a May 21 0.500 ± 0.224 a 1.167 + 0.509 a May 28 0.000 ± 0.000 a 0.167 + 0.167 a Jun 4 1.500 ± 0.342 a 1.667 + 0.211 a 0.500 ± 0.553 * 0.806 + 0.649 * Unknown population collected inside cages (Mai Apr. 30 0.667 ± 0.333 a 0.667 + 0.498 a May 7 1.333 ± 0.558 a 2. 167 + 0.703 a May 14 4.000 ± 0.816 a 3.667 + 1.145 a May 21 3.667 ± 0.954 a 5.833 + 0.938 a May 28 0.833 ± 0.477 a 1.333 + 0.558 a Jun 4 6.000 ± 0.577 a 9.833 + 1.424 b 2.750 ± 0.818 * 3.917 + 1.311 * Data transformed to (Jx+oTsj prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test) . Data are presented as mean ± S.E. Overall mean ± S.E.

PAGE 96

80 Table 5-8. Mean number of females and males of known population inside cages collected in the south and north side of the tomato field rows sampled during the spring 1987 season. Sampling date South side North side Known population collected inside cages (Females) May 2 0.167 ± 0.167 b® 0.833 + 0.167 a May 9 1.000 ± 0.449 a 1.167 + 0.601 a May 16 2.333 ± 0.760 a 1.833 + 0.307 a May 23 2.667 ± 0.760 a 3.000 + 0.578 a May 30 3.833 ± 1.302 a 4.000 + 0.843 a Jun 7 3.667 ± 0.715 a 4.333 + 0.843 a 2.278 ± 0.555 * 2.528 + 0.562 # Known population collected inside cages May 2 0.667 ± 0.211 a 3.167 + 0.477 b May 9 8.167 ± 0.401 a 0.500 + 0.342 b May 16 8.500 ± 0.847 a 10.333 + 1.874 a May 23 11.833 ± 2.182 a 10.667 + 1.606 a May 30 17.833 ± 2.212 a 13.333 + 1.686 a Jun 7 17.667 ± 1.382 a 18.500 + 2.565 a 10.778 ± 2.488 * 9.417 + 2.358 * '-Data transformed to (yx+0.5) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different (P<0.05, Student's t-test) . Data are presented as mean ± S.E. 'overall mean ± S.E.

PAGE 97

81 inside or outside the cages, regardless of the origin of the flies. Temperature differences between sunny and shady sides may also influence the number of flies collected on sticky cards at both sides although light intensity may be more important. Females, in contrast, may be attracted more intensively in staked tomatoes by factors other than reflectance to search for suitable oviposition and feeding sites. Affeldt et al. (1983) observed that more L. sativae were collected on surfaces with more light, particularly those facing the morning sun. Although leaf miners were not observed on an hourly basis, the possibility exists that most males were trapped during morning hours on the cards facing the sun. Zehnder and Trumble (1984b) reported that most L. trifolii flight activity in fresh market tomatoes peaked from 0700 h to 1100 h due primarily to movement of males. Female catches were more variable than male catches in both seasons. At the end of the 1987 spring season two samples on yellow sticky cards inside cages on known numbers of females were taken. In the first sample 20 one-day-old unfed females were introduced into each of three cages while 20 one-day-old females fed with honey for 24 h were introduced into the other 3 field cages. No significant differences between sticky card catches were observed (Table 5-9) . In a second experiment 20 mated (3 cages) versus 20 unmated (3 cages) one-day-old unfed females were compared and no significant differences in trap catches were observed

PAGE 98

82 (Table 5-9) . More experiments would be needed to determine which factors influence the trapability of females versus that of males. Table 5-9. Average laboratory-reared females collected on yellow sticky cards inside cages when feeding and mating status varied with a total of 20 females were introduced per cage. Age, feeding and Mean ± S.E. number of females mating status collected on sticky cards 1 day, unfed. a® unmated 3.667 + 0.439 1 day, fed, unmated 3.000 + 0.577 a 1 day, unfed, a® unmated 3.000 + 0.577 1 day, unfed. mated 4.000 + 0.577 a ®Data transformed to {Jx+oTbj prior to analysis but presented in the original scale. Means followed by the same letter vertically are not significantly different (P<0.05, Student's t-test) . Sampling at Three Different Heights The proportion ( female/ f emale+male) ) of L. trifolii collected at three different heights was not significantly different during 8 out of 12 sampling dates (Table 5-10) . Most females and males were collected during the last three weeks and most of them were collected at the lower or middle height (Tables 5-11, 5-12, and 5-13) which agrees with results obtained by Zehnder and Trumble (1984b) . Densities during the first 3 sampling weeks were probably too low to allow for a distinction in plant strata. Due to the high number of males collected throughout the

PAGE 99

83 Table 5-10. Mean proportion of female L. trifolii (female/female + male) collected on sticky cards at three heights during the spring 1987 season. Sampling date Sticky cards' position High Medium Low May 1 0 • 12d 0 . 085 a 0 . 075 0 • 048 a 0 • 039 1 0 • 023 a May 11 0 . 026 0 . 017 a 0 . rt ^7 rt 070 1 0 . 034 D 0 . 222 1 0 . 070 b May 14 0. 065 + 0. 021 a 0. 065 + 0. 022 a 0. 095 + 0. 026 a May 17 0. 104 + 0. 029 a 0. 134 + 0. 028 a 0. 171 + 0. 047 a May 21 0. 088 + 0. 052 a 0. 126 + 0. 037 a 0. 207 + 0. 121 a May 25 0. 092 + 0. 050 b 0. 240 + 0. 034 a 0. 196 + 0. 051ab May 28 0. 072 + 0. 019 a 0. 078 + 0. 037 a 0. 239 + 0. 010 a May 31 0. 025 + 0. Oil c 0. 149 + 0. 046 b 0. 294 + 0. 047 a Jun 4 0. 119 + 0. 031 a 0. 110 + 0. 023 a 0. 139 + 0. 017 a Jun 7 0. 062 + 0. 031 a 0. 015 + 0. 002 a 0. 098 + 0. 040 a Jun 11 0. 014 + 0. 023 b 0. 104 + 0. 064 a 0. 173 + 0. 144 a Jun 14 0. 091 + 0. 048 a 0. 085 + 0. 038 a 0. 147 + 0. 103 a 0. 074 + 0. 015 # 0. 108 + 0. 021 # 0. 168 + 0. 029 # ®Data transformed to arcsine Jy. (x=proportion 9/9+cf) prior to analysis but presented in the original scale. Means followed by the same letter horizontally are not significantly different jP<0.05, ANOVA) . Data are presented as mean ± S.E. Overall mean ± S.E. Table 5-11. Mean number of females collected on yellow sticky cards at three different heights during the spring 1987 season. Sampling date Sticky cards' position High Medium Low May 7 0. 333 + 0. 211 0. 333 + 0. 211 a 0. 667 + 0. 211 a May 11 0. 333 + 0. 211 a 1. 167 + 0. 307 a 1. 167 + 0. 401 a May 14 1. 000 + 0. 365 a 1. 500 + 0. 619 a 1. 333 + 0. 422 a May 17 1. 333 + 0. 422 b 2. 667 + 0. 494 ab 4. 333 + 1. 085 a May 21 0. 667 + 0. 333 a 2. 667 + 1. 022 a 1. 833 + 0. 477 a May 25 1. 167 + 0. 600 a 2. 833 + 0. 600 a 2. 333 + 0. 667 a May 28 0. 333 + 0. 211 b 0. 667 + 0. 336 a 1. 000 + 0. OOOab May 31 1. 000 + 0. 773 b 2. 167 + 0. 477 ab 2. 667 + 0. 494 a Jun 4 4. 500 + 0. 922 b 7. 167 + 0. 401 a 8. 333 + 1. 256 a Jun 7 1. 000 + 0. 447 b 3. 833 + 0. 703 a 2. 833 + 1. 046ab Jun 11 0. 333 + 0. 211 b 2. 333 + 0. 494 a 2. 333 + 0. 558 a Jun 14 2. 667 + 0. 422 b 4. 833 + 0. 980 ab 6. 333 + 1. 564 a '^Data transformed to (yx+0.5) prior to analysis but presented in the original scale. Means (± S.E.) followed by the same letter horizontally are not significantly different (P<0.05, ANOVA).

PAGE 100

84 Table 5-12. Mean number of males collected on yellow sticky cards at three different heights during the spring 1987 season. Sampling date Sticky cards' position High Medium Low May 7 5. 667 + 2. 028 b"^ 9. 667 + 3. 499 ab 18. 333 + 3. 685 a May 11 8 . 167 + 1 . 815 8 . 167 + 1 . 352 •J '.J 4Li 4 333
PAGE 101

85 spring season the results obtained when total number of flies was analyzed were similar to those obtained when only number of males was analyzed. No L. sativae were collected on the yellow sticky cards although a few specimens were recovered from larval samples. Zehnder and Trumble {1984b) indicated that most L. trifolii were collected on cards placed at the low plant height. In contrast, L. sativae was collected more abundantly on cards in the middle plant height. In this study, L. trifolii was equally abundant in low and middle plant strata perhaps due to the lack of competition with L. sativae . Response of Flies Exposed to Pyrethrum Insecticide in the Laboratory Adult flies exposed to the treated and untreated lumite screens in the laboratory lived the same (Tables 5-14 and 515) . Therefore, Prentox'' did not have adverse effects on the known leafminer populations in the cages. No phytotoxicity was observed in the field after the pyrethrum insecticide had been applied up to three times per week on the same plants.

PAGE 102

86 Table 5-14. Response of laboratory-reared L. trifolii to Prentox'^-treated lumite screen after 24 h. Treatment Avg. dead females (24h) Avg. dead males (24h) Inside clip cages with 0.000 ± 0.000 a® 0.143 ± 0.232 a® Prentox Inside clip cages without 0.143 ± 0.232 a 0.000 ± 0.000 a Prentox'^ Inside cups 0.000 ± 0.000 a® 0.143 ± 0.232 a® with Prentox" Inside cups without 0.000 ± 0.000 a 0.000 ± 0.000 a Prentox" ®Means (±S.E.) followed vertically by the same letter between two treatments are not significantly different (P<0.05, Student's t-test) . A total of 14 females were treated. Table 5-15. Response of laboratory-reared L. trifolii to Prentox -treated lumite screen after 48 h. Treatment Avg. dead females (24h) Avg. dead males (48h) Inside clip cages with 2.000 ± 0.000 a® 0.429 ± 0.276 a® Prentox Inside clip cages without 1.714 ± 0.264 a 1.857 ± 0.232 a Prentox Inside cups 2.000 ± 0.000 a® 1.714 ± 0.364 a® with Prentox Inside cups without 1.857 ± 0.232 a 2.000 ± 0.000 a Prentox ®Means (±S.E.) followed vertically by the same letter between two treatments are not significantly different (P<0.05, Student's t-test) . A total of 14 females were treated.

PAGE 103

CHAPTER VI ESTIMATES OF FIRST-SECOND AND THIRD STAGE LARVAL DENSITIES OF Liriomvza trifolii ON TOMATOES Introduction Since the development of organic insecticides Liriomyza sativae and Liriomyza trifolii have become important pests in vegetable crops. To develop integrated pest management (IPM) programs several researchers have tried to develop larval sampling techniques to establish timing of pesticide applications and to develop IPM programs. Oatman and Michelbacher (1958) estimated larval populations of L. sativae (misidentif ied as L. pictella (Thomson) (Spencer (1981a)) on melon taking random samples and concluded that leaf miner populations varied considerably from host to host, field to field, and even within areas in the same field. Wolfenbarger and Wolfenbarger (1966) recorded leafminer (identifed as L. sativae ) infestation levels as number of mines per leaf. Sample size and category of mines (i.e. whether live or dead leafminer larvae are present) were not given, however. Musgrave et al. (1975a) sampled L. sativae larvae in several vegetable crops and classified 'active mines' as those being occupied by live larvae. 'Active mines' are those in which larvae are alive and are light 87

PAGE 104

88 yellow in contrast to larvae that are dead and are dark yellow or brown. They concluded that a more precise estimation of fly and parasite populations was obtained from mine counts and reared samples than from adult counts on yellow sticky cards. However, mine counts were more time consuming. Foster (1986) also determined that L. trifolii larvae in celery had an aggregated distribution. Musgrave et al. (1976) determined that sampling celery trifoliates was not a reliable method for determining seasonal accumulation of mines. To determine trends in populations of L. trifolii larvae in celery Musgrave et al. (1979) determined that samples of 100 mature petioles/4.86 ha block provided estimates with a 20% precision level ((S.E./x,) *100; X|=mean number larvae collected per block). Schuster and Beck (1981) developed a visual rating index system to determine total leaf miner ( Liriomyza spp., most likely L. trifolii and L. sativae) injury to tomato foliage. This system was based on a scale from 1 to 8. To make this system more sensitive to the number of mines just below the damage threshold of 4 mines/terminal 3 leaflets (Pohronezny and Waddill 1978), increases of 1 in the scale from 1 to 4 corresponded to increments of 1 in the number of leaf mines per terminal 3 leaflets. Increments of 1 in the scale at or above a rating of 5 corresponded to increments of 6 leaf mines. A rating of 8 indicated the complete destruction of the leaf. In contrast, counts of active mines were less precise. In preliminary observations

PAGE 105

89 Schuster and Beck (1981) determined that the largest percent of active larvae could be sampled by counting them from the top to the bottom of the plants on the terminal trifoliate at the seventh node. The accuracy and precision of counting active mines on the terminal three leaflets of this node from the top was not determined, however. Larvae of L. trifolii have been monitored in California in chrysanthemums and the highest larval density has been reported from the center of the plants (Parrella and Jones 1984). Chandler and Gilstrap (1986) determined that mature leaves in the middle third of bell pepper plants taller than 75 mm supported significantly more larvae of L. trifolii per leaf than leaves positioned in other parts of the plants. Wardlow (1984) considered that leaf samples on tomatoes should be taken randomly from the middle leaves and that the assessment of 'new' mines played an essential role to determine the number of parasites that had to be introduced in a greenhouse to successfully control populations of X,. brvoniae Kalt. in southern England. Larval populations have also been assessed using styrofoam trays to collect larvae that exit the leaves to pupate. Johnson et al. (1980b) successfully used these trays and concluded that changes in populations of L. sativae were quickly detected, permitting control measures to be taken before more serious damage occurred. The objective of the present study was to develop a relative sampling method to estimate the number of active

PAGE 106

90 small (< 1.5 mm) and large (> 1.5 mm) L. trifolii larvae on a per plant basis in the least labor demanding way. The objective of sampling on a per plant basis is based on the high variability of larval distribution within a field and on the sampling technique used by field scouts. The reliability of the sampling method was analyzed to determine if it is applicable for making control decisions. Although dead larvae were also recorded, only active larvae were taken into account because of the continuous foliar damage they produce until they exit their mines to pupate on the ground. Materials and Methods During the spring 1987 season six tomato plants were selected at random in the same fields at the Gulf Coast Research and Education Center where L. trifolii adult sampling methods were developed. This procedure was followed twice a week, every Tuesday and Friday for 6 weeks starting when tomato plants measured an average of 73 cm (May 5) . Prior to sampling, plants were numbered in each row and their position was selected according to a random number generator program. Plants were isolated from the rest of the plants by cutting the strings that supported them. The relative sample on parts of the plant was taken by cutting 4 lateral stems and the main stem and counting the larvae on the 4th, 7th and 10th terminal trifoliate originating from their respective nodes, counting them from the top to the bottom. Absolute density estimates on a per

PAGE 107

91 plant basis were taken by counting every larva and empty mine on each of the selected plants. Larvae were categorized by size as large live, small live, large dead, and small dead. Empty mines were also recorded. A larva was categorized as large when its length equaled or exceeded about 1.5 mm and small if it was smaller than 1.5 mm. Dead larvae were brown or dark yellow compared to live larvae which were light yellow. Large larvae were collected in Tupperware" plastic containers with a hardware cloth divider similar to those used in the leaf miner colonies. A day or two later these larvae were placed in one-pint Fonda" ice cream containers to rear adult leafminers and parasites. Simple and stepwise linear regressions were performed relating total larval counts of large live and small live larvae per plant with larval counts (large and small larvae) made on the terminal trifoliate of the main stem at each sampled node. A second set of similar equations with the same relationship of variables was developed taking into account relative larval samples of lateral stems only. A third set of equations (using simple linear regression) was developed relating total counts of large and small larvae found per plant to total counts of larvae found in the three nodes on main stems and laterals. Regression analyses that yielded an value greater than, or equal to 0.50, were considered appropriate for

PAGE 108

92 being used in the development of a reliable sampling method. Therefore, only the equations with an value > 0.50 are listed and discussed. Results and Discussion Regression equations were generated to calibrate trifoliate samples to a per plant basis to use these equations in decision making in the scouting programs. Simple and stepwise regression equation results and terms, where an R^ value > 0.50 was obtained, are listed in Tables 6-1, 6-2, and 6-3. The regression analyses were summarized by calculating the percentage of occasions where the R^ value was > 0.50 (Tables 6-4, 6-5, and 6-6). When large or small larvae were counted on the main stem only, results were inconsistent through time taking into account simple regression analyses. An R^ > 0.50 was obtained in only 25% of the sampling dates on small larvae (on node 7) and 25% on large larvae on node 10 (Table 6-4) . Using stepwise regression analyses, no improvements were obtained by combining large larvae data from the different nodes. With small larvae up to 50% of the times an R^ > 0.50 was obtained combining data on the 7th and 10th nodes (Table 6-4). This indicates that sampling the main stem will not provide reliable information to predict absolute density. When large larvae data on the three nodes of four randomly selected lateral stems were regressed separately against the total number of larvae collected on a per plant

PAGE 109

93 (0 t7> g C (U rH •H 4J rH to Pu e -H nS (C U M Q) 0) 4-) P eO 3 -p •H C P (1) n) -P iH (0 •H CM O 0) rH — a g in as tfi CM Id :) 0 in •rl fi ^ 0 rH \ (0 T3 in p a) in n E -P rH 0) m c t-i (0 m in rH (0 0) H (0 CM m > rH >i 0) H (0 -P in (0 iH O C (U (0 H C U VI 00 0 RJ •H iH >i in Hi iH (0 C g WOO in tT> 0) C •-^ 0) 0) n) in 3 ffi • (1) -H C rH O 0) O 1 0 •rt VO M M-l (0 a 0 CO 0) 0) rH in ^ rH in "S* in (N O 00 CTi 1* in in VC VD 00 r~ 00 • • • o CM a> (N r00 00 m a^ CN {NJ o ^ m fO rH in in CN "* 00 cTi 00 O CN rCO in o o o o O rH r~ rH O O CN 00 o m CN rH 'S' in o in m 0^ in CN CM O rH ^ o p • a o Q} U u u 0) o H CO Hi N n r« o +J • +) ^ rH a 0 Oa O 0^1^ ^ 0) u o> u 0) o 0 0 • • o O • • )H 0) U 0) O 0 rH H 0 0 Q) O O 0) u u 4J 0 o +J 0 + +> + e rH rH C rH C • • c • • iJ H W ^A H to W H CO to

PAGE 110

O VO CM in 00 00 (Ti • • • • o If) in VD CN riH 00 o^ in in v£> in o i-H o in in o CN in vo ^ t~» n in CM rM in o o ^ o CM in CO iH in ^ 'a' i-i in o^ ^ o^ • • • • O 'JT i-t CM iH I 00 O O O O o o o in o n r-i in CM n CM tH I I vD o m m O «X) iH • t • • o a> CM t-t CM I O vo CO in in iH o CO "fl" d o in >^ in o^ in CM I 1-H CN 00 o^ rin rH ro o m o in CM CM r~ ^ CM cn r-H O O O (J Q) O O O + +J ^ c • • • 1-3 H w to w

PAGE 111

95 Ml W _J _1_J 1 fft B iQ 10 III H Al O 4^ in 4-* iD p " t O C CO 0 ft in (0 in *^ Q) c o (1) .c > -M •H C ID +J w > 0) 'd IN H C UJ (0 ni ift u 1 Ci 4-> ft iH lU nH VD &r) (D U IN • O Q) A\ " f-H ft lO +J CO (N C M lO P ft (J iT) ^ It 1 "H C W > *> ft (0 'D 1 ft U 1 1 1 /i\ iH iH • ^ Cu 0) Ift U ) M E -P i lo 01 r— 1 w C E 0) 0 0 M 13 lO 0^ 03 c (U (U (0 IT) rr I 1 1 1 ? CO • d) tH C4 U 01 0 1 0 •H V£> ^ >4-t CQ Cu 0 CO O O VO r~ O (N -l 0) >-t 0) MOO rH MOO (1) a 0) 0^ 0) 04 0) o o CO (U o o +) 0 +J 0 o P 0 + +J + C .H r~ C tH tH c • • c • • H w W H W (0 H CO CO H CO CO CO M CO CO

PAGE 112

^ 00 00 o^ r~ro o^ CT^ r~ o ^ in 'i" 00 00 00 0\ 00 00 iH O fO 00 00 i-i 00 CM • • • • o o in 00 vc CO CTi i-H iH 00 t-i r-i I I CM o Oi o csi in \a n • • • • o 1 CO o iH 00 m 1 in n in in CO P» in o in O CO rH o CM ^ m V£) in ON in ^ in 00 o a> • • • • o 00 n o in rH rH CM rH rH 1 CM a\ o o o O in o in in in o 00 CM rH 1 rH • •0 in fO ^ vo O 00 CM vo c 00 O rH O 00 00 rH o n • • • • +J o in VO n o V ^ rH X rH rH rH 0) 1 O o CM (Xi +J r» rH an +J ^ r~ rH rH OiCO CO CO vo 0) a) o o • • + o • • • 0) rH U 0 0 H 0 0 0 rH Ui (1) u u Q) O U CJ + P + 10 C • • c * * * EH (0 H to w CO M CO CO CO

PAGE 113

97 c rH •H \ •H C vo Oi 0 § n CM "Sin 0^ to 0) cri in in 00 O 00 \ t a • • • • 0) 0 vo o CM o CM I-•P c n CM 3 0 -P o o en to C in vo o ^ ^ (U • • • m +J o m vo CM c c Id lO 0) A > 4J •H C •P 0) rO > iH Q) o> CM 'a* 0) CO CM 00 00 CM r~ to • • • • • • 0) in o ^ rH o rH CO G Xi +J CO fH (N rH 0 -P 10 o 3 vo 'arO in 0 CM vo . vw c • • • o •H in o CO 00 A\ C rH rH rH ^\ 0 • a M to CM 00 ^ OS +J 10 in rCM 0) c CM • • • C tH (0 o in r~ 1 6 Ou rH xi (0 j-> to Ta ^ (D 4J rH 10 CJ to > 0) in P M rH rH (0 Q) 3 rH to in to rH 0) rH >i »H rH rH in (0 £ CO g 0 •rl to T) to c; rH >i'd (0 rH C ^1 in (0 (0 lO Q) -rl vo 00 tj^ to 00 00 C M • • 0 (0 in o in •H rH 0 rH CO CO c to 0) 0 g in vo 00 o U (1) ' — . CM CTN Cy> to 4J in • • • 0) 0) to o rH ^ « u rH 3 rH 0) 10 to 10 n • (1) M c > Oi Pi n O 0) 0 u -P • -P • 1 O -P •H 10 CU 0 Ou 0 vo Vh (0 to rH 0) u 0) CJ arH to V u 0) 0) U 0) U 0) rH 0) a. (U Ch ^ u P 0 P 0 (0 10 C rH C rH EH M C/1 to H W

PAGE 114

98 Table 6-4. Percentage of sampling occasions in which the regression analyses yielded an > 0.5 out of 12 sampling dates on large (L) and small (S) larvae on main stems. Larval size (L or S) and regression used® Percent in which r2 > 0.5 \ 8.33 0.00 25.00 16.67 8.33 16.67 33.33 S4 8.33 S7 25.00 S10 16.67 41.67 S4+S10 25.00 S7+S10 50.00 S4+S7+S10 33.33 '-The numerical subscripts indicate the the plant node from which the larvae were collected. The dependent variables for these equations are the mean total large or total small larvae collected on a per plant basis.

PAGE 115

99 Table 6-5. Percentage of sampling occasions in which the regression analyses yielded an > 0.5 out of 12 sampling dates on large (L) and small (S) larvae on four lateral stems. i-iarvai size (L or S) ana regression used® Percent in which > 0.5 T 0 T 16 . 67 ^10 16. 67 8 . 33 25. 00 41.67 T. 4-T -4-T 50 . 00 S4 16.67 S7 25.00 S10 25.00 S4+S7 50.00 S4+S10 50.00 S7+S10 50.00 S,+S,+S,0 75.00 ®^The numerical subscripts indicate the the plant node from which the larvae were collected. The dependent variables for these equations are the mean total large or total small larvae collected on a per plant basis. Table 6-6. Percentage of sampling occasions in which the regression analyses yielded an R^ > 0.5 out of 12 sampling dates on large and small larvae on four lateral stems and main stem. Larval size Percent in which R^ > 0.5 Large Small 33.33 41.67

PAGE 116

100 basis (Table 6-1), an > 0.50 was obtained only in 16.67% of the sampling dates (Table 6-5) . Stepwise regression analyses yielded an > 0.50 on 50% of the sampling dates when data were combined over all three nodes. Analyzing data for small larvae separately for each node yielded simple regression equations (Table 6-2) with an R^ > 0.50 in no more than 25% of the sampling dates (Table 6-5) . Combining data for small larvae on the 4th, 7th and 10th nodes resulted in 75% of the sampling dates yielding an R^ > 0.50. Combining data from the main stem and the lateral stems yielded less consistent results (Table 6-6) , which indicates that including the main stem does not improve the predictions. Taking into account the total larvae on the terminal trifoliate of the seventh node, an action threshold of a mean of 0.7 active larvae per terminal trifoliate (Pohronezny et al. 1986) was not exceeded season-long (Fig. 6-1) . Although most of the confidence intervals are wide, the upper limits never exceeded the action threshold. If they had, the sample mean value could have fallen below the action threshold while the actual larval density might have been in the upper part of the confidence interval, thus exceeding the action threshold. No action thresholds have been determined taking into account larval sizes. The development of a descriptive model on L. trifolii larval population dynamics requires equations that generate

PAGE 117

101 140 JULIAN DATE Fig. 6-1. Mean ± 95% C.I. total trifolii larvae (small and large) collected on the seventh node on four lateral steins on six plants per sampling date.

PAGE 118

102 precise and consistent results. Experiments under controlled conditions should be done to determine at which larval density yield is affected significantly to justify a pesticide application. Also, sampling methods which yield reliable and consistent larval population estimates need to be developed. Based on the results obtained during the spring 1987 season, very intensive sampling methods must be developed to generate consistent estimation equations. An important factor that may influence the consistency in sampling leafminer larvae is the spatial distribution of larvae within the plant. A clumped distribution may increase the variability to such a degree that no significant numerical relationships can be obtained by sampling on randomly selected lateral stems. If more samples per plant are taken to compare the relations, the sampling costs might be too high. Even though the highest proportion of larvae have been collected on the seventh node throughout a season (Schuster and Beck 1981) , the within plant distribution seems to play an important role to establish accurate and precise larval population density estimates . Another important factor that may influence the within plant larval distribution is parasitism. Unfortunately, no studies have been done on this subject. Leafminer parasitoids are able to control leafminer larval populations if no broad spectrum insecticides are used. These parasitoids may have a spatial preference to feed and

PAGE 119

103 oviposit in/on leafminer larvae. Different parasitoid species may parasitize larvae at different strata. In general, I conclude that practical experience in establishing action thresholds is an important factor that should be taken into account. Based on the results obtained during the spring 1987 season, I suggest the following matters: 1) to sample terminal trifoliates on only lateral stems in as many positions as possible within the plant to obtain less variable results, and 2) to determine action threshold levels under controlled situations with known larval densities. Unfortunately, these procedures would be quite time consuming and expensive.

PAGE 120

CHAPTER VII FEMALE SURVIVAL, OVIPOSITION, EGG AND LARVAL DEVELOPMENT OF L. trifolii ON TOMATO FOLIAGE Introduction Detailed knowledge of the basic biological characteristics of leafminers is necessary for development of population dynamics models. To develop process-based mathematical models, one needs to study and analyze processes like adult survival rate, oviposition, and and stage-specific development rates and mortalities. Parrella (1987) provided a summary of biological processes, like temperature-dependent egg, larval, and pupal development, of L. trifolii on different hosts. These differences were determined at different constant temperatures under laboratory conditions. Unfortunately, in most cases, adult leafminers under laboratory conditions have been provided with a carbohydrate source, consequently laboratory oviposition rates most likely have been greater than field rates. Zoebisch and Schuster (1987a) determined under laboratory conditions that L. trifolii laid significantly more eggs when females had access to aphid honeydew, which is a source of carbohydrates. Also, larval development has not been studied with consideration of different developmental stages . 104

PAGE 121

105 Some of the biological processes described in this chapter have been studied on celery by Leibee (1984) , on beans by Charlton and Allen (1981) , on chrysanthemum by Parrella et al. (1981), and on Dendranthema sp. by Miller and Isger (1985) (chapter II). The numerical descriptions of biological processes of L. trifolii provided by these researchers was based only on linear regression equations. In this research, linear, quadratic and exponential regression equations were investigated to determine which best described the biological processes of L. trifolii on tomato. Although an exponential equation can not be obtained by linear regression with SAS software, the dependent variable can be transformed into Napierian logarithms and regressed linearly against an independent variable using the general linear model procedure (GLM) of the SAS software. The objectives of this study were to describe numerically at four constant temperatures (13.9, 20, 25, and 32 'C) oviposition rate, egg developmental rate, female adult longevity, female adult survival, and larval developmental rate for first-second (small) and third (large) larval stages, with the expectation that the information obtained could be used in future population dynamics models. As a first step a conceptual model was proposed. In the original plan, data were to be collected at 10 *C, but practically no oviposition or larval development were observed at that temperature.

PAGE 122

106 Materials and Methods Planting of Host Plants Tomato plants cv. 'Hayslip' were planted and grown in , the screenhouse used for maintaining leaf miner colonies. Young seedlings were transplanted singly into 2 inch x 2 inch plastic pots and grown for four weeks prior to beginning experiments. Determination of Adult Female Longevity and Survival. Oviposition rate. Fertility . and Egg Development Rate Experiments determining adult longevity, adult survival, oviposition rate, fecundity and egg developmental rate were replicated three times. Each chamber was operated at an assigned constant temperature in all replications that it was used. Each replication consisted of eight pairs of leafminer adults inside a temperature incubator until the females died. The adults were kept at constant temperature and a photoperiod of LD 12:12 inside each incubator. Unmated flies were obtained from isolated puparia stored in plastic gel capsules. One newly emerged leafminer pair was placed on a leaflet of each four-week-old plant inside a clip cage similar to those used by Zoebisch (1984). These cages were placed on terminal leaflets of the second or third node, and, if leafminer survival exceeded the utility of second and third node terminal leaflets of a plant, they were transferred to another plant on terminal leaflets of

PAGE 123

107 the same nodes. Dead male leaf miners were replaced with newly emerged males until the females died. Flies were not provided any carbohydrate. Adult survival was observed daily, and dead females were recorded every morning between 0900 and 1030 h when eggs were counted. Oviposition was determined daily between 0900 and 1030 h, and the eggs were marked with a blue felt tip marker to aid detection of hatching. Egg hatch was determined by inspecting the leaflets twice a day for newly hatched larvae. At 32 °C, eggs were observed for 2 days after oviposition, and at 25 'C, 20 'C and 13.9 'C they were observed for 3, 5, and 7 days, respectively. Newly emerged larvae were killed with an insect pin to prevent them damaging the leaflets. Determination of Larval Development Four leaflets per two tomato plants, similar to those used for oviposition, were infested with two pair of oneday-old adult leafminers for two hours in a rearing room at about 25 'C. The plants were infested from 0900 to 1100 h to provide adult female leafminers enough time to deposit at least 3 eggs per leaflet. This experiment also was replicated three times. The infested plants were kept inside this rearing room, and hatching eggs were observed after two days. To prevent larval intraspecif ic competition, only three larvae per leaflet that hatched at the same time were allowed to survive. As soon as larvae

PAGE 124

108 had hatched, they were transferred to the constant temperature chambers. A total of 24 larvae per temperature chamber per replication were observed. Development of larvae at 25 "C and 32 "C was observed twice a day, when their size was close to the third instar. Larvae were categorized as small when they measured between 0.15 mm and 1.5 mm and large when they were longer than 1.5 mm. These sizes were taken from Johnson et al. (1980b). Equations Used to Describe the Biological Processes Studied in This Chapter Data were analyzed by linear, quadratic and exponential regression equations for describing the biological processes studied in this chapter using the GLM procedure of the PC SAS statistical package (SAS 1987) . Linear equations provided information for comparison of some of the biological processes of L. trif olii on tomato studied with other hosts like celery, and chtysanthemum. A threshold temperature for a particular biological process (primarily developmental rate) can be computed with linear regression. Quadratic equations also can be used in computing a threshold temperature. An exponential equation can be used to describe a biological process within a limited temperature range, but a threshold temperature is not defined for an exponential equation.

PAGE 125

109 Results and Discussion Adult Female Longevity and Survival. Fertility, and Egg Developmental Rate Longevity of adult leafminer females was quite variable. Linear, quadratic and exponential equations regressing number of days lived against temperature, in °C, yielded R^ values of 0.658, 0.659 and 0.681, respectively (Fig. 7-1) . Mean adult female longevity is listed in Table 7-1. Table 7-1. Mean adult female longevity (days) at four constant temperatures. Temperature ('C) N Mean ± S. E . days 32 22 3.45 + 0.13 25 22 4.86 + 0.21 20 22 7.55 + 0.32 13.9 24 8.88 + 0.46 When adult female longevity was transformed into survival rate, regression analyses similar to those done on female longevity yielded similar results. The highest value (0.694) was obtained for the quadratic regression equation (Fig. 7-2) . The mean number of eggs laid per female per day is influenced by temperature (Fig. 7-3) . Linear, quadratic, and exponential (transforming the dependent variable into Napierian logarithms) equations regressing mean number of eggs laid per female per day (oviposition rate) against

PAGE 126

110 12 11 10 Quadratic aguation F," 0.002T* 0.393T + 14.169 R*0.659 F,Adult feaale longevity (in days) T " Temperature in *C Linear equation F,-0.316T + 13.354 R^0.658 X 9 8 ' 3 ' 9 S 5 •• I I 20 25 TEMPERATDR£ (*C) 10 15 30 35 Fig. 7-1. Adult female longevity (in days) at four constant temperatures (13.9, 20, 25 and 32 'C) .

PAGE 127

Ill Fig. 7-2. Adult female survival rate (1/day) at four constant temperatures (13.9, 20, 25 and 32 'C) .

PAGE 128

112

PAGE 129

113 temperature (in °C) were developed. values of 0.735, 0.737, and 0.724 were obtained, respectively. Based on the quadratic equation (Fig. 7-4) , which had the highest R^, the estimated oviposition threshold temperature was 12.02 °C. This estimate compares favorably with observations of Leibee (1984) for L. trifolii on celery. He concluded that at 15 •C oviposition was nearly arrested, although he did not attempt to estimate the threshold temperature. Analyzing total eggs laid per female throughout its lifetime at the four constant temperatures yielded results with the highest R^ value (0.511) when a quadratic equation (instead of a linear or exponential equation) was fitted to the total number of eggs laid per female at each temperature (Fig. 7-5) . The threshold temperature of oviposition was estimated from the quadratic regression curve to be 11.9 'C, which was very similar to the estimate by Patel (1981) for L. sativae on tomato (11.5 'C) . It was also close to the estimate obtained when oviposition rate was computed from the quadratic equation in Fig. 7-4. The estimated temperature at maximum oviposition rate, based on the equation in Fig. 7-5, equaled 27.3 'C, which was close to that obtained by Parrella (1985) on L. trifolii on chrysanthemum . From 16 females observed in two replications at 10 'C, only two females deposited a total of 3 eggs; this probably was because 10 'C is close to the threshold temperature for

PAGE 130

114 Quadratic Aquation E^-0.0077T* + 1.050T 12.127 R*0.737 E^Mean nuaber of eggs laid per day per female T" Teaperature in 'C Linear equation E^0.696T 8.417 R*0.735 * * TEKPERATDSE (*C) Fig. 7-4. Mean number of eggs laid per day per female at 13.9, 20, 25, and 32 'C throughout its lifetime.

PAGE 131

115 Fig. 7-5. Total eggs laid per female at four constant temperatures (13.9, 20, 25, and 32 *C .

PAGE 132

116 oviposition which is below the estimated threshold temperature (11.9 °C) . These females did not feed on the leaflets and consequently produced no eggs. During the oviposition experiments, 2 females at each of three temperatures, 32 "C, 25 "C, and 20 'C, were lost on the second experimental day when transferred from one leaflet to another. Only three flies (15%) at 13.9 "C did not lay any eggs during their lifetimes. These results differ from those of Patel (1981) who observed 8 (40%) , 4 (20%), 2(10%), and 0 females of L. sativae out of 20 females not laying any eggs at 15.5 'C, 21.1 'C, 26.7 'C, and 32.2 'C, respectively. The average total number of eggs laid per adult female at four constant temperatures are listed in Table 7-2. Oviposition rate is also related to a female's age. At higher temperature, adult females lay eggs at higher rate, but their life span is shorter. The peak of oviposition rate occurs at higher temperature due to an increased accumulation of heat units (degree-days; Table 7-3). Table 7-2. Mean total eggs laid per female throughout its adult lifetime at four constant temperatures. Temperature (°C) Mean ± S.E. eggs laid 32 53.68 ± 3.68 25 64.23 ± 4.07 20 47.95 ± 5.10 13.9 15.63 ± 2.34

PAGE 133

117 In the species closely related to L. trifolii . peaks of average daily oviposition rate of L. sativae on tomato at 32.2 "C, 26.7 'C, 21.1 'C, and 15.5 'C are on days 3, 3, 6, and 4, respectively (Patel 1981). The maximum oviposition rate of L. trifolii on celery ranged from 35 to 39 eggs per day per female on days 1, 2, and 4 at 35 'C, 30 'C, and 25 "C, respectively (Leibee 1984) . Unfortunately, these results can not be compared with those obtained for L. trifolii on tomato because Leibee (1984) provided the adult flies with honey. Table 7-3. Magnitude and time of peak oviposition rate of L. trifolii at four constant temperatures under laboratory conditions. Temperature rc) Day of peak oviposition Proportion of mean total eggs/ female Degree-days 32 2 0.408 64.0 25 3 0.304 75.0 20 4 0.260 80.0 13.9 8 0.170 111.2 Based on the results obtained for total number eggs laid per female, oviposition is quite variable. Possible factors influencing oviposition other than temperature are female size, photoperiod, host plant and fertilizer. Fertility was high at all temperatures (Table 7-4) and similar to that observed on tomato-reared and weed-reared L. trifolii on tomato (Zoebisch 1984). Only at 13.9 *C was fertility significantly lower but still high (81.48%; Table 7-4).

PAGE 134

118 Mean egg developmental time ranged from 1.99 days at 32 •C to 11.08 days at 13.9 °C (Table 7-5; Fig. 7-6). Among the linear, quadratic and exponential descriptions of mean egg developmental rate, the quadratic equation generated Table 7-4. Percent of L. trifolii larvae that hatched at four constant temperatures. Temperature (°C) Mean hatching percent 32 91.43 a' 25 91.89 a 20 89.46 a 13.9 81.48 b Data transformed to arcsine y%/100 prior to analysis but presented in the original scale. Means followed by the same letter vertically are not significantly different (P<0.05, ANOVA) . Table 7-5. Mean egg developmental time (days) at four constant temperatures. Temperature ( ° C) Mean ± S .E. egg dev. time 32 1.99 + 0.02 25 2.47 + 0.15 20 4.08 + 0.24 13.9 11.08 + 0.21 the highest value (0.975; Fig. 7-7). The linear equation also had a high value (0.957; Fig. 7-7). Since its R^ value was only 1.9% smaller than that for the quadratic equation, the linear equation was used to describe egg developmental rate. Based on the linear regression equation the estimated threshold temperature of egg development was 9.52 "C. This estimate is close to those determined by

PAGE 135

119 TEMPERATURE 'C Fig. 7-6. Mean egg developmental time at four constant temperatures (13.9, 20, 25 and 32 'C) .

PAGE 136

120 Quadratic •
PAGE 137

121 by Charlton and Allen (1981) on beans (10.0 °C) and on chrysanthemum (13.4 °C) and by Leibee (1984) on celery (12.8 °C) . Except for total eggs laid per female, which was better described by a quadratic equation, the biological processes studied in this section were well described by linear equations (Table 7-6) . Larval Development Data for larval developmental time at four constant temperatures is illustrated in Figs. 7-8, 7-9, and 7-10. Linear, quadratic, and exponential regression of development rate (1/day) as a function of temperature, without consideration of larval size, yielded high values (0.939, 0.940 and 0.948, respectively; Figs. 7-11 and 7-12). When size was considered (Figs. 7-13 to 7-16) , the highest values were obtained with exponential regression equations (0.900 for small larvae and 0.896 for large larvae). The largest differences in R^ values for the three regression equations were obtained for large larvae 0.579 (linear equation), 0.612 (quadratic equation) and 0.896 (exponential equation) . To get higher R^ values in linear regression of large larvae developmental rate, only data between 13.9 'C and 25 'C were regressed against temperature to generate equation I in Fig. 7-15; a second equation (II in Fig. 7-15) was obtained regressing data obtained at 25 °C and 32 °C. Comparison, in terms of R^, of regression equations describing the processes studied is presented in Table 7-6.

PAGE 138

122 Table 7-6. values obtained in linear, quadratic and exponential regression equations from the biological processes studied at four constant temperatures (13.9, 20 , 25 and 32 °C) . Biological process Linear equation Quadratic equation Exponential equation 9 longevity (days) 0.658 0.659 0.681 9 survival rate (1/day) 0.662 0.694 0.681 X eggs/day/ 9* 0.735 0.737 0.724 total eggs/9 0.320 0.511 0.292 X egg devel. rate (1/day) 0.957 0.975 0.884 Larval devel. rate (1/day) 0.939 0.940 0.948 Large larvae devel. rate (1/hour) 0.579 0.612 0.896 Large larvae devel . rate (1/hour) Equation I 0.764 Large larvae devel . rate (1/hour) Equation II 0.579 Small larvae devel . rate (1/day) 0.884 0.885 0.900 x equals mean Since large larvae developed quickly (less than one day) at high temperature, longevity was expressed in hours instead of days. The lower value of equation II in Fig. 7-15 could be improved by checking the development of large

PAGE 139

123 .00 • .75 • .50 .25 .00 .75 -f .50 .25 .00 .75 .50 .25 .00 .75 .50 .25 .00 .75 .50 .25 .00 .75 .50 .25 .00 .75 .50 1.25 1.00 .75 .50 r.25 '.00 ,75 .50 .25 .00 .75 .50 .25 .00 1.75 1.50 1.25 .00 .75 .50 .25 .00 t10 — I— 15 20 25 TENPERATORE ('C) 30 35 Fig. 7-8. Larval developmental time at four constant temperatures (13.9, 20, 25 and 32 'C).

PAGE 140

124 1 1 nn 1 fl •7«> in 9 K in nn O 7*t m o i^n >• Q 9 R Q Q 2V • uu O • / 9 s fi i\n o • Dv fi 9i\ O • «9 n o • uu 7.75 7.50 7.25 7.00 a 6.75 6.50 6.25 Q 6.00 H 5.75 5.50 i 5.25 5.00 4.75 4.50 4.25 4.00 3.75 3.50 3.25 3.00 2.75 2.50 2.25 10 15 20 25 TEMPERATURE ('C) 30 35 Fig. 7-9. Small larvae developmental time at four constant temperatures (13.9, 20, 25 and 32 'C) .

PAGE 141

125 10 15 20 25 TEMPERATURE (*C) 30 35 Fig. 7-10. Large larvae developmental tine at four constant temperatures (13.9, 20, 25 and 32 'C) .

PAGE 142

126 0.35 0.30 0.25 0.20 • a >3 0.15 0.10 0.05 Quadratic aquation I^0.00009T* + 0.008T 0.060 R*0.940 I^Larval davelop»ental rate (1/day) T« Temperature in 'C Linear equation 1^0.012T 0.104 R*0.939 TEHPERATORE (*C) Fig. 7-11. Larval developmental rate (1/day) at four constant temperatures (13.9, 20, 25 and 32 'C).

PAGE 143

127 Fig. 7-12. Larval developmental rate (1/day) (transformed to Napierian logarithms) at four constant temperatures (13.9, 20, 25 and 32 'C) .

PAGE 144

128 0.45 4 0.40 0.35 Quadratic equation I^--0. 00097* + 0.019T 0.151 R*« 0.885 I^" Saall larvae developmental rate T» Temperature in *C Linear equation I^0.015T 0.106 0.884 K 0.30 0.25 H 5 0.20 0.15 0.10 25 TEMPERATOBE (*C) Fig. 7-13. Developmental rate (1/day) of small larvae at four constant temperatures (13.9, 20, 25 and 32 *C) .

PAGE 145

129 I^0.072T 3.204 R*0.900 I^Small larvae developmental rate transformed to Napierian logarithms T" Temperature in 'C — I15 10 20 25 TEKPERATOKZ (*C) 30 35 Fig. 7-14. Developmental rate (1/day) of small larvae (transformed to Napierian logarithms) at four constant temperatures (13.9, 20, 25 and 32 'C)

PAGE 146

130 0.18 0.16 Quadratic equation " I,0.00013T* -'0.003T + 0.026 R*0.612 I^o Large larvae developmental rate T" Temperature in *C 0.14 s h' 0.12 S 0.10 Linear equation I,0.003T 0.0383 R*0.579 Equation 1: I,0.0022T 0.022 R*0.764 Large larvae developmental rate TTemperature (in 'C) S 25 'C 0.08 Equation il: I,0.603T 0.038 — — — — R*0.579 I^' Large larvae developmental rate TTemperature (in 'C) > 25 'C S 32 'C N 0.06 So. 04 0.02 0.00 TEMPERATDRE ('C) Fig* 7-15. Developmental rate (1/hour) of large larvae at four constant temperatures (13.9, 20, 25 and 32 'C)

PAGE 147

131 I^O.IOOT 2.758 R*0.896 L,Devalopmental rat* of large larvae transformed to Napierian logarithms 10 15 20 25 30 35 TEMPERATURE ('C) Fig. 7-16. Developmental rate of large (1/day) larvae (transformed to Napierian logarithms) at four constant temperatures (13.9, 20, 25 and 32 'C) .

PAGE 148

132 larvae at a shorter time interval, particularly during morning hours when large larvae typically exit their mines (Charlton and Allen 1981) . Large larvae occasionally exit their mines during the afternoon, but if they do not exit in the morning, they usually remain in the leaves until the next morning. This prolongs by one day the time recorded as developmental period. The estimated threshold temperature of larval development, when size was not taken into account, was 8.67 "C, which is similar to the estimates by Leibee (1984) in celery (8.4 °C) , Charlton and Allen (1981) in beans (8.5 •C), and Schuster and Patel (1985) in tomato (7.8 'C) . For small larvae, the estimated threshold temperature was 7.07 °C, and for large larvae it was 9.9 'C (using equation I). For large larvae, the estimated threshold temperature of development (9.9 'C) was 40% greater than that estimated for small larvae (7.07 'C) . This may indicate that large larvae are more sensitive to low temperatures. Pupal Development Studies of temperature-dependent development of the pupal stage of L. trifolii have been done by Leibee (1984) on celery, Charlton and Allen (1981) on beans, Miller and Isger (1985) on Dendranthema sp. and Parrella et al. (1981) on chrysanthemum. Linear equations regressing pupal developmental rate on temperature obtained by these researchers are similar. values obtained by Leibee (1984) and Parrella et al. (1981) were 0.96 and 0.997,

PAGE 149

133 respectively. Linear regression equations describing pupal developmental rate were obtained by Parrella (1987) using published data by Charlton and Allen (1981) and Miller and Isger (1985) on pupal developmental time. These equations are listed in Table 7-7. The threshold temperature of development for those equations ranged between 8.0 and 10.0 "C. Pupal developmental rate was not ddetermoned in the research reported herein because results similar to those listed in Table 7-7 were obtained previously by Dr. D. J. Schuster (pers. comm.) on tomato. Table 7-7. Linear equations for temperature-dependentdevelopment of Liriomvza trifolii pupae. Host Development rate Estimated Reference regressed on threshold temperature temperature for (°C) development Celery y=0. 00760X-0. 0779' 10. 3 Leibee (1984) Bean y=0. 00662X-0. 0529 8. 0 Charlton and Allen (1981) Chrysanthemum y=0. 00600X-0. 0539 9. 0 Parrella et al. (1981) Dendranthema y=0. 00691X-0. 0771 10. 4 Miller and sp. Isger (1985) Equations were obtained from Parrella (1987) . Comparison of the Regression Equations Linear, quadratic, and exponential equations were studied for describing some of the biological processes of L. trifolii. Comparison of those equations was based on R^ values. Regression equations presented in this chapter indicate that most of the biological processes of Ij.

PAGE 150

134 trifolii are well described by linear equations. Only two biological processes were better described by a quadratic equation and by an exponential equation (total eggs laid per female and large larvae developmental time, respectively) . The development processes studied in this chapter were quite variable. Therefore, values for linear, quadratic, or exponential equations describing a biological process were similar. Most of the linear equations presented in this chapter are adequate for describing the oviposition and development processes of L. trifolii on tomatoes. Proposal of a Conceptual Model on Population Dynamics of L. trifolii Models are effective implements for organizing, expressing, and analyzing knowledge about systems. The philosophy of integrated pest management has led to widespread interest in the use of systems analysis. It has been focused primarily on 1) analyses of agroecosystem components including crop, pest and beneficial organisms, cultural practices, weather, and management factors; 2) predictions of agroecosystem responses to specific environmental and man-imposed inputs and the general behavioral features of these systems and 3) selection of optimal management strategies for crop development and control of pest populations for economical production of crops with high yield and quality (Smerage et al. 1980). A conceptual model is proposed below for estimating subsequent field populations of larvae from observed adult populations. The proposed model incorporates oviposition

PAGE 151

135 and development, as described in this chapter, which are some of the fundamental processes determining the dynamics of a population of L. trifolii. This model is viewed as an initial step toward eventual development of a more complex population dynamics model which would include mortality, such as parasitism and migration. It is left to other scientists to develop a mathematical model utilizing the conceptual model and data presented in this chapter. The proposed conceptual model for estimating larval population dynamics from observed adult populations is illustrated in Fig. 7-17. An informational component estimates absolute adult population density, n^, from field samples (sticky cards) using the sampling method discussed in chapter V. Estimate n^ is the input (mean number of females) as a function of time to the biological process model which includes oviposition and development of eggs and small and large larvae development rate. Large larvae at the conclusion of their development transfer to the pupal sink. The other input to the model is field temperature. From temperature and the estimated adult population, oviposition on the sampling day is estimated. Since female age can not be determined under field conditions, the adult female population is assumed to be uniformly distributed in age and oviposition is estimated using the linear or quadratic equation in Fig. 7-4 for the average number of eggs deposited per day per female for the temperature extant. Based on these assumptions, flow f^, of eggs into the population model is defined as

PAGE 152

136 (0 X a c H a «> o (0 o c o •H rH o •H u 4J •H U c a> 0) (0 (0 (0 N > 4J 0) o 0) O 3 •O R) 0) •P 3 iH o to c >1 •H n c (U •a 0) (0 c <0 (0 3 P a c 0) o 0-H «3 C 4J E O (0 CO OH II 3 II H « a H + O C -H o H C to Cl) -H ^ T3 O 6 u H i 3 II II < O O •H

PAGE 153

137 ^0,1 " ^0 '^AE where Gq = mean number of eggs laid per day per female and n^ = adult female density. Within a population, there are several age classes, which present the problem of adequately accounting for the delay between input and output flows of each age-class due to ageing or development (Smerage 1985) . With tp denoting the development time (residence time, delay, or transit time) of a life stage, 1/tp is the rate of development (passage) at the given temperature. Development time, t^, is a random variable, with a probability density function, mean and variance (Fig. 7-18) . The challenge or task in developing a population model, is to account for tp in a representation of development within a life stage. Erlang (exponential delay) and Bessell delay networks are means for accomplishing that representation (Smerage 1985) . The first order Erlang equations illustrate representation of development in the egg and small and large larvae stages in Fig. 7-17. The functional expression = nj refers to the flow of individuals from stage i to the next stage and n-^ equals the population density of stage i. Parameter Gj of stage i (egg, small or large larvae) is defined as G, = E/t„

PAGE 154

138 Fig. 7-18. Representative probability density function of development time in a life stage.

PAGE 155

139 where tjj, equals the develpment time (day or hour) of stage i at current temperature T, and E equals the area defined for the leaf miner population. If developmental rate 1/tpj of a stage is linearly related to temperature, then l/tpi = (T-To)/Kj and Gi = E(T-Toj)/Ki where Tq; and 1/Kj equal developmental threshold temperature and the slope of development rate as a function of temperature for the stage. Parameter K is equivalent to the degree-day longevity of the stage, i.e., the heat units accumulated through time to complete the stage. Slopes and threshold temperatures of stages obtained from linear regressions of data in this chapter are listed in Table 7-8. With linear temperature development in the absence of mortality and assuming that the egg and small and large larvae populations are uniformly distributed over age in their respective stages, the instantaneous flow of individuals from stage i to stage (i+1) (i.e. eggs to first instar larvae, first instar larvae to large larvae, and large larvae to pupae) is described by the following equation = nj where G. is given above. The differential equation describing the rate of change of population Nj of the ith

PAGE 156

140 stage due to flows from one stage to another and based on the first-order Erlang function is as follows dN/dt = f,,_i fii,, where Nj = Enj. This equation is used for first order representations of age classes in population system models. Higher order Erlang functions or Bessell networks (Smerage 1985) should be used for more accurate representation of development time. The order of Erlang or Bessell delay used Table 7-8. Slope (1/K) and threshold temperature estimates (Tgj) obtained from linear regression equations that describe some of the biological processes of L. trifolii derived from data analyzed in this chapter. Biological Slope Threshold temperature process coefficient estimate (1/K) To, X eggs/day/ 9* 0.696 12.09 X egg devel. 0.023 9.52 rate (1/day) Larval devel. 0.012 8.67 rate (1/day) Large larvae devel. rate 0.003 12.67 (1/hour) Large larvae devel. rate 0.002 10.00 (1/hour) Equation I Large larvae devel. rate 0.003 12.67 (1/hour) Equation II Small larvae 0.015 7.07 devel . rate (1/day) 'x equals mean

PAGE 157

141 is determined by the variance of development time, t^ (Smerage 1989) . Therefore, the variances of development time data in this chapter are given in Table 7-9. Table 7-9. Mean and variance of the time of biological processes at four constant temperatures (13.9, 20, 25 and 32 'C) included in the conceptual model. Biological process X eggs/day/9* 13.9 'C 20 'C 25 'C 32 'C X egg devel . time (days) 13.9 'C 20 'C 25 'C 32 'C Larval devel. time (days) 13.9 'C 20 'C 25 'C 32 'C Small larvae devel . time (days) 13.9 'C 20 'C 25 'C 32 'C Large larvae devel. time (days) 13.9 'C 20 'C 25 °C 32 'C N Mean Variance 24 22 22 22 @ 72 72 72 72 72 72 72 72 72 72 72 72 1.422 4.149 10.835 13.173 11.083 4.083 2.473 1.990 14.347 7.371 5.024 3.510 10.392 5.087 3.771 2.809 3.955 2.285 1.253 0.701 0.972 2.053 12.096 12.390 0.1332 0.1692 0.0697 0.0007 0.2298 0.2182 0.0746 0.2348 0.1757 0.0143 0.1210 0.2315 0.0411 0. 1220 0.1153 0.0222 X equals mean N refers to the number of replications

PAGE 158

CHAPTER VIII CONCLUSIONS Sampling methods for adult Lirioinvza spp. on vegetable crops employ sweep nets, D-Vac suction machines and yellow sticky cards. Larval sampling is based on foliar counts of mines, and pupae are collected on styrofoam trays placed underneath the plants. Although yellow sticky cards have been used in Florida for about 10 years for monitoring Liriomyza spp. , methodology has not been developed for estimating absolute adult densities from card counts. In this research, equations for estimating absolute adult density from yellow card count were generated. Adult population density is estimated on a per two plant basis, using one model with two sets of parameters (fall and spring seasons, respectively) . Although this sampling method seems to be a relative sampling method, it was considered absolute because cages of constant volume were used through time. For estimating the average number of adult leafminers per two plants the following equations were obtained: F^ = 8.352 + 0.182 F^^^ S^^ = 9.765 + 0.485 S^^^. where F^^ and S^^ are the absolute average densities estimated for the fall and spring seasons, respectively, and 142

PAGE 159

143 Fjjg and Sjjj. are the mean number of flies collected on sticky cards during the fall and spring season, respectively. These mathemathical models for fall and spring seasons were validated using data obtained during both seasons. The field sampling techniques and absolute population estimate developed in this research are considered reliable, because validation data fell within 95% confidence intervals. The adult population density estimator developed in this study could be used to evaluate the effectiveness of insecticides on adult L. trifolii in experimental plots. Data obtained in experiments on insecticide effectiveness could then be integrated into a population dynamics model to improve control procedures. The absolute density estimator would manipulate field data to generate the required input to a population dynamics model. An important factor that should be studied is the influence of light intensity on the efficiency of yellow sticky cards in the field. Furthermore, sticky cards in rows oriented N-S instead of E-W, as they were in these studies, might attract a higher number of leaf miners and thus, may overestimate absolute adult density. Since the adult population density estimator was developed under experimental conditions (i.e. in small plots and planting tomatoes with a non-planted row inbetween each row) it should be tested under commercial conditions to determine if its estimates still are reliable. An important factor to be determined is the action threshold for adult

PAGE 160

144 leaf miners. Other important factors that should be studied are adult immigration/ emigration from adjacent fields due to wind currents, harvests of other Liriomvza spp. suitable crops close to tomato fields, and weeds surrounding the tomato fields. The decisions to spray a tomato field in an IPM program in Florida is based on the average number of larvae collected at the terminal three leaflets of the seventh node of one stem per tomato plant (Pohronezny et al. 1986) . Data obtained in this research during the spring 1987 season indicate that there is no consistency in absolute density estimates at this node sampling four (instead of only one) lateral stems per plant. Based on this fact, larval density might be above the action threshold (0.7 larvae per terminal trifoliate on the seventh node of a lateral stem) , and consequently economic losses might occur. Estimates of larval densities were improved by combining data obtained on the fourth, seventh and tenth nodes, taking into account only lateral stems. Since Liriomyza spp. larvae are the most damaging stage, most sampling methods have focused on this life stage. More information is needed to develop a reliable sampling method for estimating L. trifolii larval densities. A factor that should be determined is withinplant distribution of larvae in order to improve sampling methods. Once these methods have been developed, valid information could be transferred to a population dynamics model that includes larval stages.

PAGE 161

145 Parasitoids seem to control L. trifolii as long as no broad spectrum insecticides are used. Sampling methods for estimating parasitoid population density need to be developed, and studies of parasitoid biology should be done in order to incorporate the effects of parasitoids in a population dynamics model of L. trifolii. Equations were formulated to describe oviposition and immature development of L. trifolii at four constant temperatures (13.9, 20, 25 and 32 *C) . Number of eggs laid per day per female plotted against four constant temperatures (Fig. 7-3) yielded a pattern similar to that presented by Patel (1981) on L. sativae on tomato. The oviposition threshold temperature estimated in chapter VII (11.9 "C) was close to that obtained by Patel (1981) on tomato (11.5 'C) . The 27.3 *C temperature of maximum oviposition based on the quadratic equation in Fig. 7-5, compared favorably with the results obtained by Parrella (1984) on L. trifolii on chrysanthemum. Lowering temperature progressively reduced the maximum number of eggs laid during any day, but the total number of eggs is strongly influenced by factors other than temperature, like female size, host plant, and fertilizer. The longevity of adult female L. trifolii increased with reduced temperature. Although this increased the oviposition period, it did not result in greater total egg production because oviposition rate is reduced at lower -

PAGE 162

146 temperature. In general, adult female longevity and oviposition rate are inversely related to temperature. The equation describing larval developmental rate, not taking into account larval size, was similar to that obtained by Schuster and Patel (1985). Taking larval size into account, a higher frequency of observations is required at high temperature (> 25 'C) for third stage larvae, in order to obtain a linear equation with an acceptable R^. This equation could then be used more reliably in a population dynamics model. A conceptual population dynamics model of L. trifolii was developed for predicting larval population based on an adult population sampled in a field. Equations obtained in chapter VII to describing the biological processes of L. trifolii would be the basis of the mathematical model. Equations developed in chapter V to estimate absolute adult field densities, would be used to generate the input to the model. An advantage of a population dynamics model could be timely treatment of a leafminer larval population based on adult catches. The impact of parasitism, plant resistance, and cultural methods could also be evaluated more effectively if this model were expanded to include those factors. Also scouts could save time counting leafminer adults on sticky cards instead of larvae, since up to 50% of a scout's time at any sampling site may be spent assessing only Liriomvza spp. (Pohronezny et al. 1986).

PAGE 163

LITERATURE CITED Adlerz, W. C. 1961. Control of leaf miner on watermelon in central and south Florida. Proc. Fla. State Hortic. Soc. 81: 176-180. Affeldt, H. A., R. W. Thimijan, F. F. Smith, and R. E. Webb. 1983. Response of the greenhouse whitefly (Homoptera: Aleyrodidae) and the vegetable leaf miner (Diptera: Agromyzidae) to photospectra . J. Econ. Entomol. 76(6): 1405-1409. Aldrich, J. M. 1905. A catalogue of North American Diptera. Smithsonian Institute. Washington, D. C. Bartlett, P. W. , and D. F. Powell. 1981. Introduction of American serpentine leaf miner, Liriomyza trifolii, into England and Wales and its eradication from commercial nurseries, 1977-81. Plant Pathol. 30: 185193. Bethke, J. A., and M. P. Parrella. 1985. Leaf puncturing, feeding and oviposition behavior of Liriomyza trifolii . Entomol. Exp. Appl. 39: 149-154. Bodri, M. S., and R. D. Getting. 1985. Assimilation of radioactive phosphorus by Liriomyza trifolii (Burgess) (Diptera: Agromyzidae) from feeding at different temperatures on different chrysanthemum cultivars. Proc. Entomol. Soc. Wash. 87(4): 770-776. Brogdon, J. E. 1961. Summary of leafminer control in Florida. Proc. Fla. State Hortic. Soc. 74: 143. Bottrell, D. R. 1979. Integrated pest management. U.S. Government Printing Office. Washington, DC. 120 p. Brown, R. D. , and R. A. Dybas. 1982. MK-936: a novel miticide/insecticide for control of Liriomyza leaf miners. Pages 59-61 in S. L. Poe, ed. Proceedings of the Third Annual Industry Conference on the Leafminer. The Center for Commercial Floriculture, Growers Division, Alexandria, VA. 147

PAGE 164

148 Burg, R. W. , B. M. Miller, E. D. Baker, J. Birnbaum, S. A. Currie, R. Hartman, Y.-L. Kong, R. L. Monaghan, G. Olson, I. Putter, J. B. Tunac, H. Wallick, E. O. Stapley, R. Oiwa, and S. Omura. 1979. Avennectins, new family of potent anthelmintic agents: producing organisms and fermentation. Antimicrob. Agents Chemother. 15: 361-367. Burgess, E. 1880. The clover Oscinis ( Oscinis trifolii. Burgess [n.sp.].) order Diptera; family Oscinidae. Report of the Commisioner of Agriculture, 1879, Washington, DC. Chalfant, R. B. , C. A. Jaworski, A. W. Johnson, and D. R. Sumner. 1977. Reflective film mulches, millet barriers, and pesticides: effects on watermelon mosaic virus, insects, nematodes, soil-borne fungi, and yield of yellow summer squash. J. Amer. Soc. Hortic. Sci. 102(1): 11-15. Chandler, L. D. 1981. Evaluation of different shapes and color intensities of yellow traps for use in monitoring of dipterous leaf miners. Southwestern Entomol. 6(1): 23-27. Chandler, L. D. 1985. Flight activity of Liriomyza trifolii (Diptera: Agromyzidae) in relationship to placement of yellow traps in bell pepper. J. Econ. Entomol. 78(4): 825-828. Chandler, L. D. , and F. E. Gilstrap. 1986a. Biology of Liriomyza trifolii (Burgess) on bell peppers under constant temperature conditions. Southwestern Entomol. 11(4): 269-276. Chandler, L. D. , and F. E. Gilstrap. 1986b. Within-plant larva distribution of Liriomyza trifolii (Burgess) (Diptera: Agromyzidae) on bell peppers. Environ. Entomol. 15(1): 96-99. Charlton, C. A., and W. W. Allen. 1981. The biology of Liriomyza trifolii on beans and chrysanthemums. Pages 42-49 in D.J. Schuster, ed. Proceedings of IFAS Industry Conference on Biology an Control of Liriomyza Leafminers. Institute of Food and Agricultural Sciences. University of Florida. Gainesville, FL. Cochran, W. G. 1963. Sampling techniques. Second edition. John Wiley and Sons, Inc. New York. Coquillett, D. W. 1898. On the habits of the Oscinidae and Agromyzidae, reared at the United States Department of Agriculture. Bulletin 10, new series. Division of Entomolology . Department of Agriculture. Washington, DC.

PAGE 165

149 Dimetry, N. Z. 1971. Biological studies on a leaf mining Diptera, Liriomvza trifolii Burgess attacking beans in Egypt. Bull. Soc. Entomol. Egypte 55: 55-59. De Meijere, J. C. H. 1925. Die Larven der Agromyzinen. Tijdschr. Entomol. 68: 195-293. Duncan, D. B. 1955. Multiple range and multiple F tests. Biometrics 11: 1-42. Foster, R. E. 1986. Monitoring populations of Liriomyza trifolii (Diptera: Agromyzidae) in celery with pupal counts. Fla. Entomol. 69(2): 292-298. Freidberg, A., and M. J. Gijswijt. 1983. A list and preliminary observations on natural enemies of the leafminer, Liriomyza trifolii (Burgess) (Diptera: Agromyzidae) in Israel. Israel J. Entomol. 17: 115116. Frick, K. E. 1952. A generic revision of the family Agromyzidae (Diptera) with a catalogue of New World species. Univ. Calif. Publ. Entomol. 8(8): 339-452. Frick, K. E. 1953. Some additions and corrections to the species list of North American Agromyzidae (Diptera) . Can. Entomol. 85: 68-76. Frick, K. E. 1955. Nearctic species in the Liriomyza pusilla complex. No. 3 L. alliovora, a new name for the Iowa onion miner (Diptera: Agromyzidae) . J. Kansas Entomol. Soc. 28(3): 88-92. Frost, S. W. 1924. A study of the leaf-mining Diptera of North America. Mem. Cornell Univ. Agric. Exp. Stat. 78: 1-228. Frost, S. W. 1962. Liriomyza archboldi, a new species (Dipt., Agromyzidae). Entomol. News 73(1): 51-53. Genung, W. G. 1957. Some possible cases of insect resistance to insecticides in Florida. Proc. Fla. State Hortic. Soc. 70: 148-152. Genung, W. G., V. L. Guzman, M. J. Janes, and T. A. Zitter. 1978. The first four years of integrated pest management in Everglades celery: part I. Proc. Fla. State Hortic. Soc. 91: 275-284. Getzin, L. W. 1960. Selective insecticides for vegetable leafminer control and parasite survival. J. Econ. Entomol. 53(5): 872-875.

PAGE 166

150 Hanna, H. Y., R. N. Story, and A. J. Adams. 1987. Influence of cultivar, nitrogen, and frequency of insecticide application on vegetable leafminer (Diptera: Agromyzidae) population density and dispersion on snap beans. J. Econ. Entomol. 80(1): 107-110. Harbaugh, B. K. , J. F. Price, and C. D. Stanley. 1983. Influence of leaf nitrogen on leafminer damage and yield of spray chrysanthemum. HortScience 18(6): 880881. Haynes, K. F. , M. P. Parrella, J. T. Trumble, and T. A. Miller. 1986. Monitoring insecticide resistance with yellow sticky cards. Calif. Agric. 40: 11-12. Hendel, F. 1938. Agromyzidae. Pages 213-215 in Lindner, E., ed. Die Fliegen der palaearktischen Region. VIg. Stuttgart . Bering, E. M. 1951. Biology of the leaf miners. Uitgeverij Dr. W. Junk' s-Gravenhage. N. V. Drukkerij Hooiberg, EPE. Netherlands. Hochmuth, G. J., D. N. Maynard, and S. P. Kovach. 1988. Selection of an irrigation system for vegetable production in Florida. Circular 530-C. Florida Cooperative Extension Service. Institute of Food and Agricultural Sciences. University of Florida. Gainesville, FL. Johnson, M. W. , and A. H. Kara. 1987. Influence of host crop on parasitoids (Hymenoptera) of Liriomyza spp. (Diptera: Agromyzidae). Environ. Entomol. 16(2): 339344. Johnson, M. W. , E. R. Oatman, and J. A. Wyman. 1980a. Effects of insecticides on populations of the vegetable leafminer and associated parasites on summer pole tomatoes. J. Econ. Entomol. 73: 61-66. Johnson, M. W. , E. R. Oatman, J. A. Wyman, and R. A. van Steenwyk. 1980b. A technique for monitoring Liriomyza sativae in fresh market tomatoes. J. Econ. Entomol. 73(4): 552-555. Johnson, M. W. , S. C. Welter, N. C. Toscano, I. P. Ting, and J. T. Trumble. 1983. Reduction of tomato leaflet photosynthesis rates by mining activity of Liriomyza sativae (Diptera: Agromyzidae). J. Econ. Entomol. 76(5): 1061-1063.

PAGE 167

151 Jones, J. P., and E. G. Kelsheimer. 1963. The compatibility of several fungicides and insecticides used for the control of diseases, lepidopterous larvae, and leaf miners on tomato. Proc. Fla. State Hortic. Soc. 76: 126-130. Jones, J. P., and M. P. Parrella. 1986a. Development of sampling strategies for larvae of Liriomyza trifolii (Diptera: Agromyzidae) in chrysanthemums. Environ. Entomol. 15(2): 269-273. Jones, J. P., and M. P. Parrella. 1986b. The movement and dispersal of Liriomyza trifolii (Diptera: Agromyzidae) in a chrysanthemum greenhouse. Ann. Appl. Biol. 109: 33-39. Kaltenbach, J. H. 1874. Die Pf lanzenfeinde aus der Klasse Insekten. Verhaeltniss der Naturfreunde Vereinigung. Preussen, Rheinlande, Bonn. Kelsheimer, E. G. 1961. Problems associated with insect control on tomatoes. Proc. Fla. State Hortic. Soc. 74: 156-158. Ketzler, L. D. , and J. F. Price. 1982. Methods for growers to evaluate effects of their cultural practices on Liriomyza trifolii leaf miners in a simple laboratory. Proc. Fla. State Hortic. Soc. 95: 162-164. Keularts, J. L. W. 1980. Effect of the vegetable leaf miner, Liriomyza sativae Blanchard, and the associated plant pathogens on yield and quality of the tomato, Lvcopersicon esculentum Mill. cv. Walter. Ph. D. Disseration. University of Florida. Gainesville, FL. Knodel-Montz, J. J., and S. L. Poe. 1982. Ovipositor morphology of three economically important Liriomyza species (Diptera: Agromyzidae). Pages 186-195 in S. L. Poe, ed. Proceedings of the Third Annual Industry Conference on the Leaf miner. The Center for Commercial Floriculture, Growers Division, Alexandria, VA. Knodel-Montz, J. J., S. L. Poe, and R. E. Lyons. 1983. Effects of plant growth hormones on oviposition site selection of Liriomyza trifolii (Burgess) (Diptera: Agromyzidae). Va. J. Sci. 34(3): 102. Lange, H.W., and L. Bronson. 1981. Insect pests of tomatoes. Annu. Rev. Entomol. 26: 345-371.

PAGE 168

152 Larew, H. G. , J. J. Knodel-Montz , R. E. Webb, and J. D. Warthen. 1985. Liriomvza trifolii (Burgess) (Diptera: Agromyzidae) control on chrysanthemum by neem seed extract applied to soil. J. Econ. Entomol. 78((1): 80-84. Leibee, G. L. 1981. Insecticidal control of Liriomyza spp. on vegetables. Pages 216-220 in D. J. Schuster, ed. Proceedings of the IFAS-Industry Conference on Biology and Control of Liriomyza Leaf miners. Institute of Food and Agricutural Sciences. University of Florida. Gainesville, FL. Leibee, G. L. 1984. Influence of temperature on development and fecundity of Liriomvza trifolii (Burgess) (Diptera: Agromyzidae) on celery. Environ. Entomol. 13(2): 497-501. Leibee, G. L. 1986. Effect of light on the pupariation of Liriomyza trifolii (Diptera: Agromyzidae) . Fla. Entomol. 69(4): 758-759. Levins, R. A., S. L. Poe, R. C. Littell, and J. P. Jones. 1975. Effectiveness of a leafminer control program for Florida tomato production. J. Econ. Entomol. 68(6): 772-774. Linker, H. M. , F. A. Johnson, J. L. Stimac, and S. L. Poe. 1984. Analysis of sampling procedures for corn earworm and fall armyworm (Lepidoptera: Noctuidae) in peanuts. Environ. Entomol. 13(1): 75-78. Luna, J. M. , H. M. Linker, J. L. Stimac, and S. L. Rutherford. 1982. Estimation of absolute larval densities and calibration of relative sampling methods for velvetbean caterpillar, Anticarsia qemmatalis (Hubner) in soybean. Environ. Entomol. 11(2): 497-502. Malloch, J. R. 1913. A revision of the species in Agromyza Fallen, and Cerodontha Rondani. (Diptera) . Ann. Entomol. Soc. Amer. 6(3) 269-336. Marston, N. L. , C. E. Morgan, G. D. Thomas, and C. M. Ignoffo. 1976. Evaluation of four techniques for sampling soybean insects. J. Kans. Entomol. Soc. 49: 389-400. Martens, B. , and J. T. Trumble. 1987. Structural and photosynthetic compensation for leafminer (Diptera: Agromyzidae) injury in lima beans. Environ. Entomol. 16: 374-378. Martin, C. 1984. La mineuse americaine: Liriomyza trifolii . Premier bilan en Rousillon. Elements de prophylaxie. Rev. Hortic. 244: 39-43.

PAGE 169

153 Martin, C. , and Filliol, I. 1985. Liriomyza trifolii. Mouche mineuse des cultures maraicheres. Resultats de differents types de piegeage par panneaux jaunes englues. Def. Veg. 234: 33-36. Maynard, D. N. 1987. Commercial vegetable Cultivars for Florida. Circular 530-C. Florida Cooperative Extension Service. Institute of Food and Agricultural Sciences. University of Florida. Gainesville, FL. Melander, A. L. 1913. A synopsis of the dipterous groups Agromyzinae, Milichiinae, Ochthiphilinae and Geomyzinae. J. N.Y. Entomol. Soc. 21: 219-300. Menken, B. J., and S. A. Ulenberg. 1983. Diagnosis of the leafminers Liriomyza bryoniae and L. trifolii by means of starch gel electrophoresis. Entomol. Exp. Appl. 34: 205-208. Menken, S. B. , and S. A. Ulenberg. 1986. Allozymatic diagnosis of four economically important Liriomyza species (Diptera, Agromyzidae) . Ann. Appl. Biol. 109: 41-47. Mik, J. 1894. Ueber eine neue Aqromyza deren Larven in den Bluethenknospen von Lilium martagon leben. Wien. Entomol. Z. 13: 284-290. Miller, G. W. , and M. B. Isger. 1985. Effects of temperature on the development of Liriomyza trifolii (Burgess) (Diptera: Agromyzidae). Bull. Entomol. Res. 75: 321-328. Minkenberg, 0. P. J. M. , and J. C. van Lenteren. 1986. The leafminers Liriomyza bryoniae and L. trifolii (Diptera: Agromyzidae), their parasites and host plants: a review. Agric. Univ. Waageningen Papers 86(2): 50 p. Musgrave, C. A, H. W. Beck, S. L. Poe, W. H. Denton, J. O. Strandberg, J. M. White, W. G. Genung, and V. L. Guzman. 1979. Dispersion analysis and sampling plans for insect pests in Florida celery. Proc. Fla. State Hortic. Soc. 92: 106-108. Musgrave, C. A., D. R. Bennett, S. L. Poe, and J. M. White. 1976. Pattern of vegetable leaf miner infestations in Florida celery. Proc. Fla. State Hortic. Soc. 89: 150154. Musgrave, C. A., S. L. Poe, and D. R. Bennett. 1975a. Leaf miner population estimation in polycultured vegetables. Proc. Fla. State Hortic. Soc. 88: 156-160.

PAGE 170

154 Musgrave, C. A. , S. L. Poe, G. H. Smerage, and W. D. Eshleman. 1978. Analysis of the dynamics of insect populations in celery production. Proc. Fla. State Hortic. Soc. 91: 271-275. Musgrave, C. A., S. L. Poe, and H. V. Weems, Jr. 1975b. The vegetable leafminer, Liriomyza sativae Blanchard (Diptera: Agromyzidae) in Florida. Florida Department of Agricuture and Consumer Services. Division of Plant Industry. Entomology Circular 162. Oatman, E. R. , and G. G. Kennedy. 1976. Methomyl induced outbreak of Liriomyza sativae on tomato. J. Econ. Entomol. 69(5): 667-668. Oatman, E. R. , and A. E. Michelbacher . 1958. The melon leaf miner, Liriomyza pictella (Thomson) (Diptera: Agromyzidae). Ann. Entomol. Soc. Amer. 51: 557-566. Getting, R. D. 1983. The influence of selected substrates on Liriomyza trif olii emergence. J. Ga. Entomol. Soc. 18(1): 120-124. Parrella, M. P. 1983. Intraspecif ic competition among larvae of Liriomyza trifolii (Diptera: Agromyzidae) : effects on colony production. Environ. Entomol. 12(5) 1412-1414. Parrella, M. P. 1984. Effect of temperature on oviposition, feeding, and longevity of Liriomyza trifolii (Diptera: Agromyzidae). Can. Entomol. 116: 85-92. Parrella, M. P. 1987. Biology of Liriomyza. Annu. Rev. Entomol. 32: 201-224. Parrella, M. P., W. W. Allen, and P. Morishita. 1981. Leafminer species causes California mum growers new problems. Calif. Agric. 35(9-10): 28-30. Parrella, M. P., G. D. Christie, and K. L. Robb. 1983a. Compatibility of insect growth regulators and Chrvsocharis parksi (Hymenoptera: Eulophidae) for the control of Liriomyza trifolii (Diptera: Agromyzidae) . J. Econ. Entomol. 76(4): 949-951. Parrella, M. P., and V. P. Jones. 1984. Coping with the •leafminer crisis'. Calif. Agric. 38: 17-18. Parrella, M. P., and V. P. Jones. 1985. Yellow traps as monitoring tools for Liriomyza trifolii (Diptera: Agromyzidae) in chrysanthemum greenhouses. J. Econ. Entomol. 78(1): 53-56.

PAGE 171

155 Parrella, M. P., and C. B. Keil. 1984. Insect pest management: the lesson of Liriomyza . Bull. Entomol. Soc. Amer. 30(2): 22-25. Parrella, M. P., and C. B. Keil. 1985. Toxicity of methamidophos to four species of Agromyzidae. J. Agric. Entomol. 2(3): 34-237. Parrella, M. P., K. L. Robb, and J. Bethke. 1983b. Influence of selected host plants on the biology of Liriomyza trif olii (Diptera: Agromyzidae) . Ann. Entomol. Soc. Amer. 76: 112-115. Parrella, M. P., K. L. Robb, and J. Bethke. 1981. Oviposition and pupation of Liriomyza trifolii (Burgess) . Pages 50-55 in D. J. Schuster, ed. Proceedings of IFAS-Industry Conference on Biology and Control of Liriomyza Leafminers. Institute of Food and Agricultural Sciences. University of Florida. Gainesville, FL. Patel, K. J. 1981. Influence of temperature on the fecundity of Liriomyza sativae Blanchard (Diptera: Agromyzidae) and rate of development of L. sativae and Diqlyphus intermedius (Girault) (Hymenoptera: Eulphidae) . M.S. Thesis. University of Florida. Gainesville, FL. Patel, K. J. 1987. Parasitization of Liriomyza trifolii (Burgess) by Diqlyphus intermedius (Girault). Ph.D. Dissertation. University of Florida. Gainesville, FL. Perez, R. P. 1974. Liriomyza munda Frick (Diptera: Agromyzidae) attacking beans and cucumbers in Puerto Rico. J. Agric. Univ. P.R. 57(4): 350. Peterson, A. 1979. Larvae of insects. Part II. Edwards Brothers, Inc. Ann Arbor, Michigan. Poe, S. L. 1974. Selective application of insecticides for sustained fruit yield of tomato. Proc. Fla. State Hortic. Soc. 87: 165-169. Poe, S. L. , P. H. Everett, D. J. Schuster, and C. A. Musgrave. 1978. Insecticidal effects on Liriomyza sativae larvae and their parasites on tomato. J. Ga. Entomol. Soc. 13(4): 322-327. Poe, S. L. , and J. J. Montz. 1981. Preliminary results of a leafminer species survey. Pages 24-34 in D. J. Schuster, ed. Proceedings of IFAS Industry Conference on Biology and Control of Liriomyza Leafminers. Institute of Food and Agricultural Sciences. University of Florida. Gainesville, FL.

PAGE 172

156 Pohronezny, K. L. , and V. H. Waddill. 1978. Integrated pest management development of an alternative approach to control of tomato pests in Florida. University of Florida Extension Plant Pathololgy Report no. 22. Pohronezny, K. L. , V. H. Waddill, D. J. Schuster, and R. M. Sonoda. 1986. Integrated pest management for Florida tomatoes. Amer. Phytopathol. Soc. 70(2): 96-102. Powell, D. F. 1979. Eradication of alien pests of glasshouse crops in the United Kingdom. Pages 259-267 in D. L. Ebbels and J. E. King eds. Plant Health. Ministry of Agriculture, Fisheries and Food. Harpenden, England. Powell, D. F. 1981. The eradication campaign against American serpentine leaf miner, Liriomyza trifolii, at Efford Experimental Horticulture Station. Plant Pathol. 30: 195-204. Price, J. F. 1982. An assessment of leafminer management methods on export chrysanthemum enterprises in Coloir±)ia, South America. Proc. Fla. State Hortic. Soc. 95: 146-148. Price, J. F., and S. L. Poe. 1976. Response of Liriomyza (Diptera: Agromyzidae) and its parasites to stake and mulch culture of tomatoes. Fla. Entomol. 59(1): 85-88. Prieto, A. J., and Chac6n de Ulloa. 1980. Biologia y ecologia de Liriomyza trifolii (Burgess) (Diptera: Agromyzidae) minador de crisantemo en el Departamento del Valle del Cauca. Rev. Colombiana Entomol. 6(3/4): 77-84. Riley, H. 1884. Annual Report of the U.S. Department of Agriculture. Washington, DC. Ruesink, W. G., and M. Kogan. 1975. The quantitative basis of pest management: sampling and measuring. Pages 309351 in R. L. Metcalf, and W. H. Luckman, eds. Introduction to Insect Pest Management. John Wiley and Sons, NY. SAS Institute, Inc. 1985. SAS procedure guide for personal computer. Version 6th ed. Gary, NC. Schenk, E. T., and J. H. McMasters. 1936. Procedure in Taxonomy. Stanford Univ. Press. Palo Alto, CA. Schuster, D. J. 1978. Vegetable leafminer control on tomato, 1977. Insectic. Acaricide Tests 3: 108.

PAGE 173

157 Schuster, D. J. 1985. Seasonal abundance of Liriomyza leafminers and their parasitoids in fresh market tomatoes grown on the west coast of Florida. Pages 1920 in J. J. Knodel-Montz , ed. An Informal Conference on Liriomyza Leafminers. USDA/ARS Technical Information Bulletin. Schuster, D. J., and H. W. Beck. 1981. Sampling and distribution of Liriomyza on tomatoes. Pages 106-128 in D. J. Schuster, ed. Proceedings of IFAS-Industry Conference on Biology and Control of Liriomyza Leafminers. Institute of Food and Agricultural Sciences. University of Florida. Gainesville, FL. Schuster, D. J., and H. W. Beck. 1983. Visual rating system for assessing Liriomyza spp. (Diptera: Agromyzidae) leaf mining on tomato. J. Econ. Entomol. 76(6): 1465-1466. Schuster, D. J., and P. H. Everett. 1981. Insect control on tomato, 1980. Insectic. Acaricide Tests 6: 98. Schuster, D. J., and P. H. Everett. 1983a. Insect control on tomatoes in southwest Florida, spring 1982. Insectic. Acaricide Tests 8: 149. Schuster, D. J., and P. H. Everett. 1983b. Response of Liriomyza trifolii (Diptera: Agromyzidae) to insecticides on tomato. J. Econ. Entomol. 76(5): 11701174. Schuster, D. J., and J. P. Jones. 1976. Effect of leaf miner control on tomato yield. Proc. Fla. State Hortic. Soc. 89: 154-156. Schuster, D. J., c. A. Musgrave, and J. P. Jones. 1979. Vegetable leafminer and parasite emergence from tomato foliage sprayed with oxamyl. J. Econ. Entomol. 72(2): 208-210. Schuster, D. J., and K. J. Patel. 1985. Development of Liriomyza trifolii (Diptera: Agromyzidae) larvae on tomato at constant temperatures. Fla. Entomol. 68(1): 158-161. Schuster, D. J., and J. F. Price. 1985. Impact of insecticides on lepidopterous larval control and leafminer parasite emergence on tomato. Proc. Fla. State Hortic. Soc. 98: 248-251.

PAGE 174

158 Schuster, D. J., T. G. Zoebisch, and J. P. Gilreath. 1982. Ovipositional preference and larval development of Liriomyza trifolii on selected weeds. Pages 137-145 in S. L. Poe, ed. Proceedings of the Third Annual Industry Conference on the Leaf miner. The Center for Commercial Floriculture, Growers Division, Alexandria, VA. Shorey, H. H. , and I. M. Hall. 1963. Toxicicty of chemical and microbial insecticides to pest and beneficial insects on poled tomatoes. J. Econ. Entomol. 56(5): 813-817. Shull, C. A. 1929. A spectrophotometric study of reflection of light from leaf surfaces. Bot. Gaz. 87: 583-607. Smerage, G. H. 1985. Bessell delay networks for population models. Trans. ASAE (28) : 1269-1278) . Smerage, G. H. 1989. Models of development in insect populations. In Estimation and Analysis of Insect Populations. McDonald, L. L. , F. J. Manley, J. A. Lockwood, and J. A. Logan, eds. Springer Verlag. New York, NY. (in press). Smerage, G. H., C. A. Musgrave, S. L. Poe, and W. D. Eshleman. 1980. Plant protection through integrated management: Systems analysis of insect population dynamics. Institute of Food and Agricultural Sciences. Gainesville, FL. Smith, R. F., and J. M. Hardman. 1986. Rates of feeding, oviposition, development, and survival of Liriomyza trifolii (Burgess) (Diptera: Agromyzidae) on several weeds. Can. Entomol. 118: 753-759. Snedecor, G. W. , and W. G. Cochran. 1967. Statistical Methods. Iowa State University Press. Ames, lA. Spencer, K. A. 1963. A synopsis of the neotropical Agromyzidae. Trans. R. Entomol. Soc. Lond. 115: 352372. Spencer, K. A. 1964. The species-host relationship in the Agromyzidae (Diptera) as an aid to taxonomy. Xllth Int. Congr. Entomol. London, p. 101-102. Spencer, K. A. 1965. A clarification of the status of Liriomyza trifolii in Florida (Diptera: Agromyzidae). Proc. Entomol. Soc. Wash. 67(1): 32-40. Spencer, K. A. 1968. Relationships between the Nearctic and Palaearctic Agromyzidae (Diptera) . Xlllth Int. Congr. Entomol. Moscow, p. 1-6.

PAGE 175

159 Spencer, K. A. 1973. Agromyzidae (Diptera) of economic importance. Pages 37-59 in Dr. W. Junk, ed. The Hague. Netherlands. Spencer, K. A. 1981a. A revisionary study of the leafmining flies (Agromyzidae) of California. University of California. Division of Agricultural Sciences. Special Publications no. 3273. Spencer, K. A. 1981b. Morphological characteristics and brief taxonomic history of Liriomyza . Pages 12-23 in D.J. Schuster, ed. Proceedings of IFAS-Industry Conference on Biology and Control of Liriomyza Leaf miners. Institute of Food and Agricultural Sciences. University of Florida. Gainesville, FL. Spencer, K. A. 1984. The Agromyzidae (Diptera) of Colombia, including a new species attacking potato in Bolivia. Rev. Colombiana Entomol. 10(1/2): 3-33. Spencer, K. A., and C. E. Stegmaier, Jr. 1973. Agromyzidae of Florida with a supplement of species from the Caribbean. Arthropods of Florida and Neighboring Areas. Vol. 7. Florida Department of Agriculture and Consumer Services. Division of Plant Industry. Gainesville, FL. Stegmaier, C. E., Jr. 1966. Host plants and parasites of Liriomyza trifolii in Florida (Diptera; Agromyzidae) . Fla. Entomol. 49(2): 75-80. Siiss, L. , G. Agosti, and M. Costanzi. 1984. Liriomyza trifolii, note di biologia. Inf. Fitopatol. 2: 8-12. Taylor, G. T., and S. A. Smith. 1987. Production costs for selected Florida vegetables, 1986-87. Economy Information Report no. 234. Food and Resource Economics Department. Agricultural Experiment Stations. Institute of Food and Agricultural Sciences. Univ. of Florida. Gainesville, FL. Tilden, J. W. 1950. Oviposition and behavior of Liriomyza pusilla (Meigen) . Pan-Pac. Entomol. 26(3): 119-121. Trumble, J. T. 1983. Suppression of lepidopterous larvae and leaf miners on tomatoes, 1982. Insectic. Acaricide Tests 8: 150. Trumble, J. T. 1985a. Integrated pest management of Liriomyza trifolii : influence of avermectin, cyromazine, and methomyl on leaf miner ecology in celery. Agric. Ecosyst. Environ. 12: 181-188. Trumble, J. T. 1985b. Planning ahead for leaf miner control. Calif. Agric. 35(9-10): 8-9. j

PAGE 176

160 Trumble, J. T. , and H. Nakakihara. 1983. Occurrence, parasitization, and sampling of Liriomyza species (Diptera: Agromyzidae) infesting celery in California. Environ. Entomol. 12(3): 810-814. Tryon, E. H., Jr., S. L. Poe, and H. L. Cromroy. 1980. Dispersal of vegetable leafminer onto a transplant production range. Fla. Entomol. 63(3): 292-295. Van Sickle, J. J., and Belibasis, E. 1985. Update on Florida West Mexico competition in the fresh market tomato industry. Pages 79-97 in Florida Tomato Institute. University Florida Vegetable Crops Department (D. N. Maynard, Program Coordinator) . Vegetable Crops Extension Report VEC 85-2. Vercambre, B. 1980. Etudes realis^es a la Reunion sur la mouche maraichere: Liriomyza trifolii Burgess. Rev. Agric. Sucriere lie Maurice 59: 147-157. Wardlow, L. R. 1984. Monitoring the activity of tomato leaf miner ( Liriomyza bryoniae Kalt.) and its parasites in commercial glasshouses in southern England. Med. Fac. Landbouww. Rijksuniv. Gent 49 (3a): 781-791. Warthen, J. D. 1979. Azadirachta indica: a source of insect feeding inhibitors and growth regulators. Science and Education Administration. U.S.D.A., Agriculture Revenue Manual, Northeastern, Series no. 4. Webb, R. E., and F. F. Smith. 1973. Influence of reflective mulches on infestations of Liriomyza munda in snap bean foliage. J. Econ. Entomol. 66(2): 539540. Webb, R. E., M. A. Hinebaugh, R. K. Lindquist, and M. Jacobson. 1983. Evaluation of aqueous solution of neem seed extract against Liriomyza sativae and L. trifolii (Diptera: Agromyzidae). J. Econ. Entomol. 76(2): 357-362. Webb, R. E., F. F. Smith, A. M. Wieber, H. Larew II, and J. J. Knodel-Montz . 1985. The comparative responses of a vegetable leafminer, Liriomyza sativae (Blanchard) (Diptera: Agromyzidae) and the greenhouse whitefly, Trialeurodes vaporariorum (Westwood) (Homoptera: Aleyrodidae) , to visual stimuli. Pages 51-72 in J. J. Knodel-Montz, ed. An Informal Conference on Liriomyza Leaf miners. USDA/ARS Technical Information Bulletin. Webster, F. M. , and T. H. Parks. 1913. The serpentine leaf-miner. J. Agric. Res. 1(1): 59-87.

PAGE 177

161 Woets, J. 1985. Integrated programmes for specific crops. Tomatoes. Pages 166-174 in N. W. Hussey and N. Scopes eds. Biological Pest Control: The Glasshouse Experience. Blandford, Poole Dorset, England. Wolfenbarger, D. A., and D. O. Wolfenbarger . 1966. Tomato yields and leaf miner infestations and a sequential sampling plan for determining need for control treatments. J. Econ. Entomol. 59(2): 279-283. Wolfenbarger, D. 0. 1948. Control studies on the serpentine leaf miner on potato and tomato. Proc. State Hortic. Soc. 62: 181-186. Wolfenbarger, D. 0. 1958. Serpentine leaf miner: brief history and summary of a decade of control measures in south Florida. J. Econ. Entomol. 51: 357-359. Wolfenbarger, D. O. 1961. Leaf mining insects, especially the serpentine miners on vegetable crop plants and their control. Proc. Fla. State Hortic. Soc. 74: 131133. Wolfenbarger, D. 0., and W. D. Moore. 1968. Insect abundances on tomatoes and squash mulched with aluminum and plastic sheetings. J. Econ. Entomol. 61(1): 34-36. Yudin, L. S., W. C. Mitchell, and J. J. Cho. 1987. Color preference of thrips (Thysanoptera: Thripidae) with reference to aphids (Homoptera: Aphididae) and leaf miners in Hawaiian lettuce farms. J. Econ. Entomol. 80(1): 51-55. Zehnder, G. W. , and J. T. Trumble. 1984a. Host selection of Liriomyza species (Diptera: Agromyzidae) and associated parasites in adjacent plantings of tomato and celery. Environ. Entomol. 13: 492-496. Zehnder, G. W. , and J. T. Trumble. 1984b. Spatial and diel activity of Liriomvza species (Diptera: Agromyzidae) in fresh market tomatoes. Environ. Entomol. 13(5): 14111416. Zehnder, G. W. , and J. T. Trumble. 1985a. Impact of currently registered insecticides on the Liriomyza/parasite complex in celery, 1984. Pages 2127 in An Informal Conference on Liriomyza Leaf miners. USDA/ARS Technical Information Bulletin. Zehnder, G. W. , and J. T. Trumble. 1985b. Sequential sampling plans with fixed levels of precision for Liriomyza species (Diptera: Agromyzidae) in fresh market tomatoes. J. Econ. Entomol. 78(1): 138-142.

PAGE 178

162 Zehnder, G. W. , J. T. Trumble, and W. R. White. 1983. Discrimination of Liriomyza species (Diptera: Agromyzidae) using electrophoresis and scanning electron microscopy. Proc. Entomol. Soc. Wash. 85: 564-574. Zoebisch, T. G. 1984. Oviposition and development of Liriomyza trif olii (Burgess) (Diptera: Agromyzidae) on foliage of tomato and selected weeds. M. S. thesis. University of Florida. Gainesville, FL. Zoebisch, T. G., and D. J. Schuster. 1987a. Longevity and fecundity of Liriomyza trifolii (Diptera: Agromyzidae) exposed to tomato foliage and honeydew in the laboratory. Environ. Entomol. 16(4): 100-1003. Zoebisch, T. G. , and D. J. Schuster. 1987b. Suitability of foliage of tomatoes and three weed hosts for oviposition and development of Liriomyza trifolii (Diptera: Agromyzidae). J. Econ. Entomol. 80(4): 758762.

PAGE 179

BIOGRAPHICAL SKETCH Tomds Giinter Zoebisch was born in Mexico City, in 1958, and attended the German Alexander von Humboldt School in Mexico City. In 1980, he enrolled at the Universidad Autonoma Metropolitana and received the Bachelor of Science degree in biology in 1980. In the same year, he worked at the Institute de Ecologia in Mexico City under the supervision of Dr. Rapoport in a project on the ecology of weeds. At the same time, he was a biology teacher at the Institute Cumbres Highschool in Mexico City. In 1981, he moved to Cuernavaca, Morelos, to study cattle tick ecology at the Centre Nacional de Parasitologia Animal, where he was chairman of the Department of Ecology. In 1982, he enrolled in the graduate program in the Entomology and Nematology Department at the University of Florida and has been employed by the department as a research and teaching assistant. After his M.S. program, he continued his studies, pursuing the Ph.D. degree under the supervision of Dr. D.J. Schuster. His primary objectives are to develop integrated pest management programs in vegetable and/or field crops at an agriculture research center. 163

PAGE 180

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. David J^. S/5huster, Chairman Professor of Entomology and Hematology 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. Jejjfiry l/ Stimac Associate Professor of Entomology and Hematology 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. Stratton H. Kerr Professor of Entomology and Hematology 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. Strayer sor of Entomology and Hematology 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. Len H. Smerage ''Associate Professor of Agricu: Engineering ural

PAGE 181

This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1988 Dean, Graduate School