Citation
Relationship between temperature and development of the ectoparasitoid Larra bicolor (Hymenoptera : Sphecidae) and the endoparasitoid Ormia depleta (Diptera : Tachinidae)

Material Information

Title:
Relationship between temperature and development of the ectoparasitoid Larra bicolor (Hymenoptera : Sphecidae) and the endoparasitoid Ormia depleta (Diptera : Tachinidae)
Creator:
Cabrera-Mireles, Hector, 1957-
Publication Date:
Language:
English
Physical Description:
xvi, 176 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Adult insects ( jstor )
Eggs ( jstor )
Female animals ( jstor )
Heat sums ( jstor )
Insects ( jstor )
Parasite hosts ( jstor )
Parasitoids ( jstor )
Pupae ( jstor )
Soil temperature regimes ( jstor )
Temperature ( jstor )
Dissertations, Academic -- Entomology and Nematology -- UF ( lcsh )
Entomology and Nematology thesis, Ph. D ( lcsh )
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2002.
Bibliography:
Includes bibliographical references (leaves 167-174).
General Note:
Printout.
General Note:
Vita.
Statement of Responsibility:
by Hector Cabrera-Mireles.

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:
029216581 ( ALEPH )
50797138 ( OCLC )

Downloads

This item has the following downloads:


Full Text











RELATIONSHIP BETWEEN TEMPERATURE AND DEVELOPMENT OF THE ECTOPARASITOID LARRA BICOLOR (HYMENOPTERA: SPHECIDAE) AND THE
ENDOPARASITOID ORMIA DEPLETA (DIPTERA: TACHINIDAE)










By

HECTOR CABRERA-MIRELES


















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 2002














ACKNOWLEDGMENTS

I would like to extend a special thanks to my committee members who have

greatly helped me improve this dissertation. I probably would not find enough words to thank Dr. J.H. Frank (chair), without his support this dissertation would have been impossible. I would thank Dr. J.L Stimac for his cheerful encouragement, friendship and for serving as a committee member. I would like to thank Dr. J. Jones and Dr. T.J Walker who were kind enough to make suggestions on this dissertation and served as committee members. Sincere thanks are due to Dr. Robert Hemenway who was always available to provide me with help and supplies of all kinds and kindly friendship.

I thank the Instituto Nacional de Investigaciones Forestales, Agricolas y Pecuarias (INIFAP) and the Consejo Nacional de Ciencia y Tecnologia (CONACYT) for having provided me with fmancial support during my doctoral study.





















11













TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS............................... ....................... ii1

LIST OF TABLES...........................................................ii

LIST O F FIG U RES........................................................................................................ xi

ABSTRACT........ .................................................................................................... xv

1. INTRODUCTION............................ ........................... 1

2. REVIEW OF LITERATURE...................................................................... 6

Introduction........................................................................................................ 6
The Pest Mole Crickets, Scapteriscus spp.......................................................... 7
The Digger Wasp Larra bicolor..................................................................... 9
Natural History of Larra bicolor................................................................ 9
Life Cycle of Larra bicolor in the Laboratory...................................... ... 10
Establishment of Larra bicolor in Florida.................................................. 11
The Tachinid Fly Ormia depleta............................................... .......................... 11
Natural History of Ormia depleta................................. ................................. 11
Life History of Ormia depleta in the Laboratory........................ ........ 12
Insect M odeling....................................................................................................... 13
Constant vs Fluctuating Temperature Influence on Insect Development.......... 14
Constant Temperatures................................................. 15
Fluctuating Temperatures...................................... .................................... 17
Applicability of Constant and Fluctuating Temperatures in Insect
Developm ental M odels............................................................................... 18
Temperature-Dependent Models for Studies on Insect Development.................. 19
Linear M odel....................................................................................................... 20
Non-linear Models......................................................... 22
Stochastic Models. ............................................................ 24
M ixed M odels...................................................................................................... 26
Conclusion......................................................................................................... 27

3. DEVELOPMENT OF THE WASP LARRA BICOLOR F.................................... 30

Development and reproduction of the wasp Larra bicolor at constant
Tem peratures........... ........................................... .............................................. 30
Introduction.......................................................................................................... 30


iii









M aterials and M ethods................................ .................... 31
Development..................................................................................................... 31
Life-Fertility Tables..................................................... 32
Results................................................................ ............................................... 33
Development......................................................... ......................................... 33
Reproduction.................................................... 37
Life-FertilityTables ...................................................................................... 38
Discussion............................................................................................................. 39

Development of the wasp Larra bicolor at field-varying temperatures............... 42
Introduction......................................................................................................... 42
M aterials and M ethods...................................................... 43
Results................................................................................................................. 48
Survivorship.................................................................................................... 51
Discussion ................ ........................ ...... ............................................. 52

Degree-day model predicting the development of the wasp Larra bicolor......... 57
Introduction........................... ........................... ............................................. 57
M aterials and M ethods......................................................... 58
Results................................................................................................................. 60
Discussion............................................................................................................ 64

4. DEVELOPMENT OF THE FLY ORMIA DEPLETA............................................. 66

Development of Ormia depleta at constant temperatures............................ 66
Introduction....................................................................................................... 66
M aterials and M ethods............................................................................................ 68
Results........................................ ............................................................. ....... 69
The new strain............................................. ................................................... 69
The old strain........................................................................................................ 74
Discussion................................ .............................. ............................................ 79

Development of the two strains of Ormia depleta under field conditions........... 86
Introduction....................................................................................................... 86
M aterials and M ethods........................................................... 86
The new strain .................................................................................................... 91
The old strain............................................. ................................................... 91
Results....................................... ............................................................. ....... 92
The new strain............................................. .................................................. 92
The old strain............................................. .................................................... 93
Development times............................................................ 95
Discussion................................................................................................................ 97
Development times............................................................ 97
Problems.............................................................................................................. 98
Diapause.............................................................................................................. 99




iv









Degree-day model predicting the development of Ormia depleta ................... 100
Introduction........................ ........................................................................... 100
M aterials and M ethods................................ 101
Results................................................................................................................. 103
The new strain................................... 103
The old Strain................................... 107
Discussion............................................................... ........................................... 111

5. CONCLUSIONS ....................................................................................... ............. 114

The Parasitic W asp Larra bicolor......................................................................... 114
The Parasitic Fly Ormia depleta ................................. 116

APPENDICES

A. DEVELOPMENT OF TWO MOLE CRICKETS, S. VICINUS AND S.
ABBREVIATUS, AT CONSTANT TEMPERATURES......................... 119

Introduction ............................................................................................................. 119
M aterials and M ethods.......................................................................................... 120
Oviposition, fecundity and fertility.............................................. 121
Nymphal development.............................. 122
Results........................................ ............................................................. ....... 122
Scapteriscus vicinus.................................. 122
Oviposition, fecundity and fertility.............................................................. 123
Development.......................................................................... ....................... 126
Scapteriscus abbreviatus.............................. 128
Oviposition, fecundity and fertility................................................................. 128
Development............................................................................................. 131
Discussion...................................................................................................... 133
Scapteriscus vicinus.................................. 133
Scapteriscus abbreviatus.............................. 136
Both species.......................................................................................................... 136

B. PARASITISM OF MOLE CRICKETS BY THE WASP LARRA BICOLOR........ 139

Introduction............................................................................................................. 139
M aterials and methods.................................. 140
Results .................................................................................... ............................ 143
Percent parasitism......................................... ................................................. 143
Parasitoid success.................................. 146
Sex dimorphism ............................... 146
Discussion.................................................... ............................ ........... 148







V









C. SOIL TEMPERATURE ESTIMATE................................ 151

Problem description .............................. 151
M aterials and methods.......................................................................................... 153
Temperature estimates................................ 153
M odel.................................................................................................................. 154
Description ................. ..................... ........ ........................................... 154
Results............ ............................................................................................ 155
Temperature estimate of missed data.............................................................. 159
Discussion........................... ........................ ....................................................... 165

LIST OF REFERENCES ...................... .............................. 167

BIOGRAPHICAL SKETCH ...................................................................................... 175









































VI















LIST OF TABLES

Table Page

Table 1. Insect models cited in the Review of Literature chapter....................................28

Table 2. Development (days) of Larra bicolor life stagesa at constant temperatures...........34

Table 3. Linear regression models (y = developmental rate, x = temperature), lower
developmental thresholds (t), and degree-days required for Larra bicolor life
stages and total to develop at constant temperatures. ....................................36

Table 4. Population reproductive statistics for Larra bicolor from two seasons in
Gainesville, FL (29o40'N) ............................................................................ 38

Table 5. Reproductive parameters for Larra bicolor from two seasons in Gainesville, FL
(29040'N) ...................................................................... 39

Table 6. Codes and depths of sub-chambers and thermocouples where parasitized mole
crickets were confined to study the development of Larra bicolor
underground in the field ...................................... ................ 45

Table 7. Description of the cohorts of the wasp-parasitized hosts by numbers and dates....47 Table 8. Development time ofLarra bicolor at various depths underground in the field at
various dates in Gainesville, FL (29o40'N) ....................................................49

Table 9. Developmental time of four cohorts of Larra bicolor stages and total, averaged
over all depths underground in the field in Gainesville, FL (29040'N). .............49

Table 10. Trimodal total developmental time of Larra bicolor per cohort underground in
the field in Gainesville, FL (29040'N). ..................................... .......50

Table 11. Trimodal life cycle developmental time of Larra bicolor underground in field
conditions of Gainesville, FL (29040'N) ........................................................ 51

Table 12. Number of individuals, corresponding proportion and sex ratio per cohort and
overall of Larra bicolor developing on Scapteriscus borellii underground at
field varying-temperatures...................................... ............ .............51






vii








Table 13. Parameter values for the linear model, lower thresholds and thermal constants
(degree-days) for development of Larra bicolor per stage and total at constant
tem peratures..................................................................................................... 61

Table 14 Degree-days (days) required for Larra bicolor stages and total to develop at
constant and field temperatures .................................................61

Table 15. Statistical significance of slopes in regression analysis of development of Larra
bicolor stages and total, from the field and from incubation chambers ...........64

Table 16. Development of Ormia depleta (new strain) pooled sexes life stages............... 70

Table 17. Development of Ormia depleta (new strain) male stages ................................. 70

Table 18. Development of Ormia depleta (new strain) female stages .............................. 71

Table 19. T-test comparison of developmental time (days) of Ormia depleta (new strain)
between sexes per temperature per stage and total life cycle .......................... 72

Table 20. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta
(new strain) pooled sexes life stages to develop ....................................... 73

Table 21. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta
(new strain) male stages to develop ...............................................73

Table 22. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta
(new strain) female stages to develop. ............................................................73

Table 23. Development of Ormia depleta (old strain) pooled sexes life stages ............... 75

Table 24. Development of Ormia depleta (old strain) male life stages. ........................... 75

Table 25. Development of Ormia depleta (old strain) female life stages ......................... 75

Table 26. T-test comparison of developmental time (days) of Ormia depleta (old strain)
between sexes per temperature per stage and total life cycle ..........................77

Table 27. Linear regression models (Y= developmental rate, x- temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (old
strain) life stages to develop ....................................................78

Table 28. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (old
strain) male life stages to develop................................... ............... 78




viii








Table 29. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (old
strain) female life stages to develop ......................................... ......... 78

Table 30. Comparison of slopes (bl)' and intercepts (a)' between the "new" and "old"
strains of Ormia depleta by t-test and confidence limits at constant
tem peratures..................................................................................................... 84

Table 31. Time periods when experiments were conducted for studying the development
of the two strains of Ormia depleta at field-varying temperatures of
G ainesville, FL .............................................. ............................................ 91

Table 32. Development time in days' and corresponding temperatures during
development per stage life cycle of Ormia depleta, new strain, underground in
sub-chamber A (0-15 cm depth) in field experiments .................................... 92

'Table 33. Development time in days' and corresponding temperatures during
development per stage life cycle of Ormia depleta, new strain, underground in
sub-chamber B (15-30 cm depth) in field experiments......... .................... 93

Table 34. Development time in days', standard errors and corresponding temperatures
during development per stage life cycle of Ormia depleta, old strain,
underground at various depths in field experiments ..................................... 94

Table 35. Parameter values for the linear model, lower thresholds and thermal constants
(degree-days) for development of Ormia depleta (new strain) per stage and
total at constant temperatures. ..................................... 104

Table 36. Degree-days required by Ormia depleta (new strain) to develop in incubation
chambers and at field-varying temperatures ......................... 104

Table 37. Statistical significance of slopes in regression analysis of development of
Ormia depleta (new strain) stages and total, in field and in incubation
cham bers .......................................................................................................... 107

Table 38. Regression slope comparison by 95% confidence limits of degree-days of
developmental stages of Ormia depleta (new strain) developing at fieldvarying temperatures and in incubation chambers at constant temperatures ..... 107

Table 39. Parameter values for the linear model, lower thresholds and thermal constants
(degree-days) for development of Ormia depleta (old strain) per stage and
total at constant temperatures. ..................................... 108

Table 40. Degree-days required by Ormia depleta (old strain) to develop in incubation
chambers and in field-varying temperatures ...................................... 109





ix








Table 41. Statistical significance of slopes in regression analysis of development of
Ormia depleta (old strain) stages and total, in the field and in incubation
cham bers............................................................. ........................................... 111

Table 42. Regression slope test by 95% confident limits of degree-days of larval, pupal
and total development of Ormia depleta (old strain) in field-varying
temperatures and in incubation chambers at constant temperatures ................... 111

Table 43. Fertility and oviposition period of Scapteriscus vicinus reared at 260C and
photoperiod of 14:10, L:D ...................................... 125

Table 44. Egg fertility and development of Scapteriscus vicinus at constant temperatures. 125

Table 45. Nymphal developmental time of Scapteriscus vicinus at two constant
temperatures ............................................... 126

Table 46. Nymphal development of Scapteriscus abbreviatus at constant temperatures..... 131

Table 47. Developmental time of Scapteriscus abbreviatus from egg to adult, grouped by
number of instars to complete development and sex ratio per group .............. 133

Table 48. Numbers of Scapteriscus vicinus mole crickets, and numbers parasitized by the
parasitoid wasp Larra bicolor, trapped in pitfall traps at 2 locations during 2
seasons in Gainesville, Florida. ..................................... 144

Table 49. Success of Larra bicolor parasitizing Scapteriscus vicinus .............................. 146

Table 50. Sex dimorphism in adult and pupal stages of Larra bicolor developed on
naturally field-parasitized host mole crickets ...................................... 147

Table 51. Linear equation models proposed to estimate soil temperatures ....................... 154

Table 52. Coefficient of determination values and statistical significance (p-values) of
slope and intercept for four models estimating soil temperature, using 500,
1000, or 2000 data points. ..................................... 156

Table 53. Coefficient of determination values and statistical significance (p-values) of
intercept and slope of four models estimating soil temperature with soil and
air temperature as independent variables, using 500, 1000, or 2000 data
points. .............................................................................................................. 158










X















LIST OF FIGURES



Figure Page

Figure 1. Theoretical linear and nonlinear model development rate functions. ........... 16

Figure 2. Developmental time of Larra bicolor stages and total at constant
tem peratures...................................................................................... 34

Figure 3. Survivorship of Larra bicolor life stages at constant temperatures ................ 35

Figure 4. Developmental rates of Larra bicolor at constant temperatures. Symbols
show observed data. Line trend fitted to the linear response. ................... 37

Figure 5. Oviposition time of wasps Larra bicolor reared at summer-like and autumnal
seasons in Gainesville, FL; daily oviposition per wasp is represented by
different line pattern per season ........................................ .......... 41

Figure 6. Proportion of individuals from three cohorts entering the life cycle stages of
Larra bicolor developing on Scapteriscus borellii underground at fieldvarying temperatures ...................................... ........................ 52

Figure 7. Degree-days required for pupal development of Larra bicolor underground
at field-varying temperatures. Pupal stage grouped by kind of
developmental times (short, intermediate, and long) ................................ 62

Figure 8. Comparison of degree-days required for Larra bicolor life stages and total
to develop at constant and field varying temperatures. A) Egg, larval and
prepupal development. B) Pupal and total development (short
development only). ....................................................... 63

Figure 9. Development time of Ormia depleta "new" strain, pooled sexes, at constant
tem peratures...................................................................................... 71

Figure 10. Survivorship of Ormia depleta "new" strain, pooled sexes, at constant
tem peratures...................................................................................... 72

Figure 11. Developmental rates of Ormia depleta "new strain", pooled sexes, at
constant temperatures. ........................................................................ 74




xi








Figure 12. Developmental time of Ormia depleta, old strain, life cycle at constant
temperature............................. 76

Figure 13. Survivorship of Ormia depleta old strain, pooled sexes, at constant
temperatures ........................................................... 77

Figure 14. Developmental rates of Ormia depleta old strain, pooled sexes, life cycle at
constant temperatures. ................................... ....................... 79

Figure 15. Developmental times of two strains (new and old) of Ormia depleta,
pooled sexes, at constant temperatures. Means followed by different letter
are statistically different ("t - test", P<0.05). ..................................... 80

Figure 16. Developmental times of two strains (new and old) of Ormia depleta males
at constant temperatures. Means followed by different letter are
statistically different ("t - test", P<0.05) ......................................... 81

Figure 17. Developmental times of two strains (new and old) of Ormia depleta
females at constant temperatures. Means followed by different letter are
statistically different ("t - test", P<0.05) ......................................... 81

Figure 18. Linear development estimates by linear regression equations of the two
strains of Ormia depleta at constant temperatures. Lower development
was prolonged until crossing "y" axis. .................................................... 83

Figure 19. Survival (%) of the two strains of Ormia depleta at constant temperatures.
Means followed by different letter are statistically different ("X2 - test",
P<0.05)............................................................................................... 85

Figure 20. Comparison of larval, pupal and total development of Ormia depleta, new
strain, at two depths at field (F) experiments. Means followed by different
letters are statistically different. Analysis by depth................................ 94

Figure 21. Comparison of larval, pupal and total development of Ormia depleta, old
strain, at two depths in field (F) experiments. Means by experiment are
statistically not different: Duncan's test. ..................................... . 95

Figure 22. Comparison of larval, pupal and total developments of the two strains of
Ormia depleta at two depths underground. Field experiments (F) selected
by similar mean temperatures. NS = new strain, OS = old strain. Means
followed by different letters are statistically different. Mean comparisons
grouped by stage and total. ......................................... ............ 96

Figure 23. Regression lines of development of Ormia depleta (new strain) in
incubation chambers set at constant temperatures (N Chambers - ) and
at field varying-temperatures ( O Field ). A) Larval development
B) Pupal development; C) Total development ...................................... 106



xii








Figure 24. Regression lines of development of Ormia depleta (old strain) in incubation
chambers set at constant temperatures (g Chambers - ) and at field
varying-temperatures ( * Field ). A) Larval development; B) Pupal
development; C) Total development. ..................................... 110

Figure 25. Number of eggs per clutch, time after mating, and accumulated percentage
of eggs of Scapteriscus vicinus at 260C. DAM = Days after
m ating�SEM ......................................................................................... 123

Figure 26. Nymphal developmental time of Scapteriscus vicinus at constant 300C in
an incubation chamber. N=13 for lt instar; N=l for the other instars........ 127

Figure 27. Nymphal developmental time of Scapteriscus vicinus at constant 250C in
an incubation chamber .................................... 127

Figure 28. Number of eggs per clutch (average) and days between clutch and
corresponding proportion (%) of eggs per time of Scapteriscus abbreviatus
at constant 26oC. DAM= days after mating. ..................................... 129

Figure 29. Proportion of eggs of Scapteriscus abbreviatus hatching at constant
tem peratures .......................................................................................... 130

Figure 30. Proportion of eggs from Scapteriscus abbreviatus hatching by clutch
number at constant temperatures. ........................................................... 130

Figure 31. Nymphal development of Scapteriscus abbreviatus at constant 25C ....... 132

Figure 32. Nymphal development of Scapteriscus abbreviatus at constant 30C ....... 132

Figure 33. Number of eggs hatched per clutch of two mole cricket species,
Scapteriscus abbreviatus and S. vicinus at constant 26C ....................... 136

Figure 34. Proportion of eggs hatched of two mole crickets, Scapteriscus abbreviatus
and S. vicinus, at constant temperatures. ..................................... 137

Figure 35. Numbers of parasitized and unparasitized Scapteriscus vicinus mole
crickets trapped in pitfall traps at 2 locations during 2 seasons in
Gainesville, Florida, by date of collection. ..................................... 145

Figure 36. Numbers of nymphs and adults of Scapteriscus vicinus mole crickets, and
proportions of each parasitized by the parasitoid wasp Larra bicolor, trapped in pitfall traps at 2 locations during 2 seasons in Gainesville,
Florida, by pronotal size class ...................................... 145

Figure 37. Pupal length and pupal width comparison between sexes of Larra bicolor.
Circles show the proportion of overlapping sizes between sexes. ........... 148




xlll








Figure 38. Temperatures recorded during 10 days in July from the soil in the
experimental area (Exp.) at two depths (7.5 and 22.5 cm.) and from soil
(15 cm depth) recorded by the Florida Automated Weather Network
(FA W N ). .................................................................................................. 153

Figure 39. Efficiency of soil temperature models to estimate temperature at 7.5 cm
depth. Independent variable: Soil temperature from the experimental area
(Exp. 7.5) at 7.5 cm depth. Dependent variable: Soil temperature (FAWN
soil) from the Florida Automated Weather Network................................. 156

Figure 40. Efficiency of soil temperature models to estimate temperature at 7.5 cm
depth. Independent variable: Soil temperature from the experimental area (Exp. 7.5) at 7.5 cm depth. Dependent variable: Soil temperature (FAWNsoil) and Air temperature (not shown) from the Florida Automated
Weather Network................................. 158

Figure 41. Temperatures recorded at two depths from the experimental area (Exp.) and
from the alternative temperature record source (FAWN). Period from
5 July - 16 Aug. 2000 ........................................................................ 160

Figure 42. Temperatures recorded at 7.5 cm depth from the experimental area and
from the alternative temperature record source (FAWN). Period from
23-30 Aug. 2000. The predicted temperature was estimated with the
equation: Y= -1.96286(Sto) - 0.07562(Ato) + 3.113706(St-) +
0.546506(At.1) -16.8321........................................................................ 161

Figure 43. Temperature for 22.5 cm depth underground before and after the period of
16-23 August 2000. A) Temperatures recorded before the missed
temperature period. B) Temperatures recorded after the missed
temperature period ....................................... 162

Figure 44. Temperatures recorded at two depths from the experimental area (Exp.) and
from the alternative temperature record source (FAWN). Period from
29 Sep. - 11 Oct. 2000.................. ..................................... 163

Figure 45. Temperatures recorded at 7.5 cm depth from the experimental area and from
the alternative temperature record source (FAWN). Period from 29 Sep. 11 Oct. 2000. The predicted temperature was estimated with the equation:
Y= 6.61423(Sto)-0.4854(Ato) -4.4884(St-1) +0.35548(A_-1)-20.061 ............. 163

Figure 46. Temperature model for 22.5 cm depth underground for the period of 29 Sep.
11 Oct. 2000. A) Temperatures recorded before the missed temperature period and used to build the model. B) Temperatures recorded after the
missed temperature ...................................... 164






xiv














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

RELATIONSHIP BETWEEN TEMPERATURE AND DEVELOPMENT OF THE ECTOPARASITOID LARRA BICOLOR (HYMENOPTERA: SPHECIDAE) AND THE
ENDOPARASITOID ORMIA DEPLETA (DIPTERA: TACHINIDAE) By

H6ctor Cabrera-Mireles

August, 2002
Chairman: J. Howard Frank
Major Department: Entomology and Nematology

The rates of development of the second introduced strain (from Bolivia) of the ectoparasitoid wasp Larra bicolor and of the two introduced strains (from subtropical Brazil and from temperate Brazil) of the larviparous endoparasitoid tachinid Ormia depleta were determined when reared at both constant and field temperatures in northern Florida. The data were used to examine the performance of the accumulated degree-day model (DD) to predict field development of both natural enemies parasitizing the mole cricket Scapteriscus borellii. Development and survivorship were studied in incubation chambers at eight constant temperatures, 15.0, 17.5, 20.5, 22.5, 25.0, 27.5, 30.0 and 32.5 + 0. I1C, at 14:10 (L: D) photoperiod. The relationship between temperature and developmental rate was estimated with linear regression analyses. Pooled developmental times for males and females were used to fit the linear regression models. The lower developmental threshold was estimated by the "x-intercept" method. Development of all stages of the wasp occurred at temperatures from 25.0 to 30.00C. No development was


xv








observed at 17.5C, and only one wasp developed at 22.5�C. Development of 0. depleta occurred in the range from 17.0 to 30.00C. The field development was determined by confining newly parasitized hosts into 15 x 5 cm (L x diameter) chambers, filled with autoclaved sand, buried underground at up to 60 cm depth that were operated during all parts of the year except the coldest months. Field experiments with L. bicolor were run during autumn. Temperature underground was recorded every 30 min during field experiments. A degree-day model based upon constant temperature was fitted to data from field experiments for each parasitoid species and strain. A full description is presented of the developmental time (and rates) in day units, for constant and field experiments, together with the regression coefficients from constant temperatures and the validation of both data sets using the accumulated degree-day model. Data about the wasp are summarized in life tables and its field efficacy in parasitizing mole crickets is presented and explained. Data about the fly comparing development per sex and per strain are presented also.























xvi













CHAPTER 1
INTRODUCTION

Three mole cricket species from South America arrived as immigrants to the

southeastern USA in ships' ballast about the year 1900 (Walker and Nickle 1981). They are Scapteriscus abbreviatus Scudder, S. vicinus Scudder and S. borellii Giglio-Tos. The immigrant mole crickets became the most important pests of turf and pasture grasses in Florida. Their feeding and tunneling also destroy seedlings of vegetables, ornamentals, and tobacco. Insecticides commonly applied to control mole crickets in lawns, golf courses, and seedbeds are expensive and not always effective. In pastures there is no control that is economically feasible (Walker 1985a).

Adult S. abbreviatus have very short wings, and are flightless. Its current

distribution is restricted to coastal areas and two small areas inland of Florida. Adult S. vicinus and S. borellii have wings that slightly exceed the length of the body, and can fly. Their current dispersion includes eight states, from North Carolina south to Florida and west to Texas (Frank and Parkman 1999). In 1978, the University of Florida established a mole cricket research program (MC program) because of damage to pasture grasses (Walker 1985a). Damage (including control costs) to turf and pastures in Florida alone may approach $100 million annually (H. Frank personal communication 2001).

Three biological control agents from South America have been successfully

released into Florida for control of immigrant mole crickets, an ectoparasitic wasp Larra bicolor F., an endoparasitic fly Ormia depleta Wiedemann, and a parasitic nematode Steinernema scapterisci Nguyen and Smart (Frank and Parkman 1999).


I





2


The neotropical digger wasp Larra bicolor (Hymenoptera: Sphecidae) has been known for decades as a parasitoid of large nymphs and adults of mole crickets (Castner 1984). A Puerto Rican stock originally from Brazil was introduced into Florida in 1981. L. bicolor wasps were released at five sites in south and central Florida in conjunction with plots of their preferred nectar-source plant Spermacoce verticillata L.

Establishment of Larra bicolor only at Ft. Lauderdale, and not at the four more

northerly sites in Florida, suggested that this wasp, of tropical origin, could not withstand colder or longer winters farther north in Florida (Castner 1988). In the late 1980s a different, possibly more cold-tolerant strain of L. bicolor was imported from Bolivia, released in northern Florida, and a decade later appeared not only to be established but also to be spreading, indicating at least that mole crickets in some unknown proportions were being parasitized (Frank et al. 1995).

Ormia depleta is an endoparasitoid ofScapteriscus spp. adults. Females find hosts by homing on the sounds of calling male crickets (Fowler 1987). 0. depleta was imported from Piracicaba, Brazil (23o30'S), released at numerous sites in Florida, and soon spread throughout south and central Florida. This first strain failed to establish permanent populations in north Florida (>270 N), perhaps due to its poor tolerance to cold and long winters (Walker et al. 1996). In recent years, in late autumn, a few parasitic flies have been captured near Gainesville, and these were considered to be individuals dispersing from south toward north without re-establishing populations. In 1999 a second tachinid strain was imported from southern Brazil (300S); because of its origin farther south (in a cooler climate) it was hoped that this strain might be more cold-tolerant and might establish populations north of 290N in Florida and other states (Frank 1994).





3


The entomopathogenic nematode Steinernema scapterisci Nguyen and Smart

(Rhabditida) was imported from Uruguay and released in Florida in 1985 (Parkman and Frank 1992). This nematode successfully attacks adults and large nymphs of S. borellii and S. vicinus, but is not very effective against S. abbreviatus, and has negligible effect against small and medium nymphs of any of these three pest mole crickets (Frank and Parkman 1999).

There is no information about the effect of temperature on life history or seasonal phenology of any of the insect parasitoids. Limits of geographical expansion of L. bicolor and 0. depleta are poorly understood. The fact that the first introduced strains failed to establish in northern Florida suggests a strong influence of temperature regimes, where either colder winters kill the parasitoids or cause the lack of adequate energy (nectar, etc.) sources for the adults. These arguments led to the introduction of a second strain with the hope that they would be more cold hardy and survive colder winters.

The level of control of mole crickets exerted by these two insect parasitoids in

Florida is questionable (e.g., Potter 1998). The questions arise from lack of adequate data.

Development of plants and invertebrates is temperature-dependent with each organism requiring a specific measure of heat accumulation between lower and upper developmental thresholds to complete development (Hagstrum and Milliken 1988). Application of developmental thresholds and rates in the form of phenological models is often used in agricultural IPM programs to predict and manipulate pest and natural enemies population dynamics in the field (Tauber et al. 1994).

Temperature plays a critical role in determining the rate of development, survival, and reproduction of insect species. Understanding the relationship between temperature





4


and various life-history processes is important in developing and refining computer models for simulating biological control of pest populations. Physiological time measured in degree-days is effective in predicting the seasonality of life history stages of insects. The development of a degree-day model requires the determination of the temperature threshold above which development will occur. Developmental thresholds can be determined directly in the laboratory by measuring the time taken to rear insects through a series of developmental stages at a range of constant temperatures. Then the thresholds can be validated in the field (Wagner et al. 1984a, Campbell et al. 1974).

Phenology of many species of poikilotherms as a function of temperature has

been described by regression equations (Wagner et al. 1984b). Such regression equations can provide a method for calculating developmental times in insect population models or can be used for quantitative comparisons of the effect of temperature on the developmental times of different stages and species of insects (Hagstrum and Milliken 1988).

Understanding the effects of temperature on summary life-history parameters such as the intrinsic rate of increase can be useful for examining the potential effectiveness of a species as a biological control agent, estimating its potential for population growth (Lysyk 2000).

Information describing the life history of pest mole crickets and of their two insect natural enemies released in Florida is inadequate. In the best of the cases, the phenology of the parasitoid is based upon development at a single temperature. Thus, we lack the information necessary to describe development and reproduction of any of the





5


insects involved. No development model has been proposed for any of the imported insect natural enemies of mole crickets.

The objective of this study was to investigate the development of the digger wasp Larra bicolor and of the tachinid fly Ormia depleta parasitising mole crickets at both constant and field temperatures. These data are used to examine the feasibility of using a degree-day accumulation model to predict the development of the two parasitoids in the field.

This dissertation is organized by parasitoid species. The studies on the parasitic wasp present the development of the wasp at constant temperatures followed by the field study and finally a degree-day model simulating the wasp development in the field and its validation.

The parasitic fly studies are composed of three experiments; firstly, studying its development at constant temperatures, then in field experiments and finally presenting the proposal of a degree-day model and its validation in field development of this parasitoid.















CHAPTER 2
REVIEW OF LITERATURE


Introduction

Three species of mole cricket (Scapteriscus spp.) are the most damaging pests of turf and pasture grasses in much of the southeastern United States. They destroy turf and grasses by feeding on roots, leaves and shoots, and by tunneling, which disrupts the soil around roots, preventing uptake of water and nutrients (Hudson 1997, cited in Frank and Parkman 1999, Walker 1985a). Damage to Florida turf was estimated as $30 million annually in Florida alone (Walker 1985a), $45 million (damage and cost of control) in 1986, with a further $33 million in Alabama, Georgia, and South Carolina combined (not including damage in North Carolina, Mississippi, Louisiana, and Texas) ( Frank and Parkman 1999). More recent estimates are of annual expenditure of more than $18 million on chemical insecticides in Florida turf, and of more than $12 million in damage in Georgia turf (Hudson et al. 1988). Damage to pasture grass in Georgia in 1996 was estimated as 0.83 million for control and as $9.8 million in damage (Bunti

n et al. 1997, cited in Frank and Parkman 1999).

The most promising strategy for solving immigrant mole cricket problems is the use of classical biological control. In 1978, in response to increasing concerns of cattleman, turf managers, and homeowners, the University of Florida initiated a research project on mole crickets (Walker 1985a). Three biological control agents from South America have been successfully released into Florida for control of adventive mole


6





7


crickets, Larra bicolor F. (Hymenoptera: Sphecidae), Ormia depleta Wiedemann (Diptera: Tachinidae) and Steinernema scapterisci Nguyen and Smart (Rhabditida: Steinemematidae) (Frank and Parkman 1999).

Information regarding natural history of the Scapteriscus mole crickets and of the insect parasitoids in the continental United States is presented in this chapter.


The Pest Mole Crickets, Scapteriscus spp.


Four mole crickets (Orthoptera: Gryllotalpidae) are listed as occurring in Florida: Neocurtilla hexadactyla Perty, the northern mole cricket; Scapteriscus abbreviatus Scudder, the shortwinged mole cricket; Scapteriscus borellii Giglio-Tos, the southern mole cricket; and Scapteriscus vicinus Scudder, the tawny mole cricket. The northern mole cricket is the only species native to the southeastern USA including Florida. The three other mole cricket species are immigrants from South America that were accidentally brought in by ships' ballast about the year 1900 (Walker and Nickle 1981).

Mole crickets are fossorial insects with greatly enlarged forelegs that are used for digging. Some of the adults and nymphs of a population move about on the soil surface on warm nights with high humidity. Depth of the mole crickets in the soil depends upon temperature and moisture (Frank and Parkman 1999, Lake 2000). High water tables force the mole crickets to the surface. Males are more active than females; maximum average depth reported by Abate (1979) was 50 cm for S. borellii males and 44 cm for S. vicinus males. Scapteriscus mole cricket burrows as deep as 75 cm have been recorded (Frank and Parkman 1999). Subterranean depths of oviposition by S. borellii and S. vicinus are





8


30 cm maximum depth recorded in the field by Hayslip (1943) and a maximum mean depth of 60 cm for S. borellii from laboratory experiments by Lake (2000).

Adults of S. abbreviatus have very short wings. In contrast, wings of S. borellii and S. vicinus slightly exceed the length of the body, which explains their current wide distribution in the southern coastal plains from North Carolina south to Florida and west to Texas. S. abbreviatus cannot fly and its distribution is restricted to some small areas in Florida (Frank and Parkman 1999). S. borellii and S. vicinus sometimes fly in enormous numbers. Their flights apparently serve two functions: (1) local searching for mates and new egg-laying or calling sites and (2) long range dispersal (Walker 1985b).

Adult mole crickets, like many other crickets, produce a calling song that

functions to bring the sexes together. Females sometimes respond to the call by walking or flying to the male (phonotaxis) and mating with him. The calling period is correlated with female activity and flight periods (Walker 1985b). The intensity of the song depends upon male size and soil moisture (Forrest 1985).

The following information was presented by Walker (1985b). Mole crickets spend nearly all their lives underground. Adult S. vicinus fly in large numbers typically in March but as early as February in Florida after warm winters. They mate (although some matting occurred the previous autumn); they begin as eggs laid in underground chambers. Eggs take some three weeks to incubate. Nymphs hatch from eggs as early as April, but continue to hatch from later deposited eggs through June. Nymphs take some five months to develop during summer months, and newly formed adults begin to appear in September. There is a minor peak of flight activity in the autumn, as early as August in the far south and as late as December farther north if the weather remains warm. In most





9


of the southern USA, the spring activities occur in S. borellii about three weeks later than in S. vicinus, but the autumnal activities are at the same time. From central Florida northward, S. vicinus and S. borellii have one generation each year. Most S. vicinus overwinter as adults, and most S. borellii overwinter as large juveniles. In southern Florida S. vicinus maintains its one year life cycle, but S. borellii has two generations annually. In total contrast, all developmental stages of S. abbreviatus occur throughout the year but with two peak ovipositional periods, one in late spring and one in winter.


The Digger Wasp Larra bicolor

Natural History of Larra bicolor

Wasps of the genus Larra (Hymenoptera: Sphecidae) are the only sphecids

known to attack mole crickets, which all of them do. Unlike other digger wasps, which are generally predators, all Larra wasp species are parasitoids (Sailer 1985). Worldwide, 65 species of Larra are known (Bohart and Menke 1976). Only one species, Larra analis F., a parasitoid of the northern mole cricket, is native to the United States. Sixteen species have been recorded from South America, where the diversity of species of mole crickets belonging to the genus Scapteriscus is greatest (Sailer 1985). The following information was presented by Sailer (1985). Apart from preferences for different species of mole crickets, the biology of the various species of Larra is remarkably similar. Females do not construct a nest, and prey paralysis is temporary with the host reviving soon after egg deposition. A single egg is firmly attached in the soft, membranous thoracic tissue between the host's first and second legs. The Larra larva lives ectoparasitically until almost fully grown. It then kills and consumes the rest of its host. Larra females lay about 30 eggs during their life span. Full-grown larvae form underground cocoons of





10


sand grain cemented together with a secretion from the labial glands. The L. bicolor adults emerge from their cells 60 to 80 days after eggs are laid on the hosts. At Fort Lauderdale (south Florida) at least four generations are probably produced annually. Females hunt for mole crickets during early morning and mid afternoon hours. During the middle of the day they visit flowers where they feed on nectar and possibly pollen. L. bicolor is most commonly found on the flowers of two quite unrelated plants, Spermacoce verticillata L. (Rubiaceae) (native to south Florida) and Hyptis atrorubens L. (Lamiaceae). The close association of Larra bicolor with these plants suggests that it may not inhabit areas where these plants are not available. Life Cycle of Larra bicolor in the Laboratory

The following information was presented by Castner (1988). The following data were recorded from parasitoids reared in a laboratory maintained at approximately 26 � 2oC and 65 � 5% relative humidity. The 4.0-4.5 mm Larra egg hatches in 6-7 d. The first instar larva remains fixed to the original point of attachment. The larva passes through five instars, feeding initially on hemolymph, but progressing to body tissues as heavily sclerotized mandibles develop. The larva completes the first four instars 8-9 d after hatching, increasing gradually in size to 12-19 mm. The final larval size is 14-26 mm. By 10-11 d post-hatching (16-18 d post-oviposition), the wasp grub has killed the host and consumed the soft inner tissues. The mature larva spends approximately one day in constructing an ovate cocoon of soil particles or sand grains. The wasp adult emerges 6-8 weeks after the cocoon is completed.








Establishment of Larra bicolor in Florida

The wasp Larra bicolor was the first classical biological control agent of mole

crickets to be introduced and established in Florida. A Puerto Rican stock originally from Brazil was introduced in 1981; wasps were released at five sites in south and central Florida in conjunction with plots of their preferred nectar-source plant Spermacoce verticillata. Establishment of Larra bicolor only at Ft. Lauderdale, and not at the four more northerly sites in Florida, suggested that this wasp, of tropical origin, could not withstand colder or longer winters farther north in Florida (Castner 1988, Frank and Parkman 1999). In 1988-89 a Bolivian stock of L. bicolor, possibly more cold-tolerant, was introduced, and a decade later appeared not only to be established but also to be spreading in northern Florida indicating at least that mole crickets in some unknown proportions were being parasitized (Frank and Parkman 1999).


The Tachinid Fly Ormia depleta

Natural History of Ormia depleta

All species of the family Tachinidae develop as parasites of other arthropods.

Some groups of species are restricted to closely related kinds of host insects; for example the group Ormiini consists of species that parasitize katydids and other night-singing Orthoptera (Sailer 1985). Ormia depleta is a South American tachinid that attacks mole crickets of the genus Scapteriscus (Fowler 1987). Adults are crepuscular and nocturnal parasitoids. Females are phonotactic, locating their hosts by their calling songs, and they deposit larvae on or near the host (Fowler and Kochalka 1985).

The first stock was brought from Piracicaba, Brazil (23oS), and released in Florida in 1988. Since then it has become established in 38 counties in Florida, and was





12


considered as largely responsible for a decline in mole cricket populations (Parkman et al. 1996). No evidence of their extension farther to the north exists despite releases in northwestern Florida, Georgia, Alabama, and North Carolina (Frank et al. 1996, Walker et al. 1996). By 1994, this strain disappeared from the Gainesville area (290 40'N latitude), perhaps due to its poor tolerance to cold and long winters (Walker et al. 1996). In recent years, in late autumn, a few parasitic flies have been captured near Gainesville, and these were considered to be individuals dispersing from south toward north without re-establishing populations.

In 1999 a second tachinid strain was imported. It was hoped that this "new" strain, from 30oS in Brazil, might be more cold-tolerant and could establish populations north of 290N in Florida and other states (Frank 1994, Potter 1998).

Several studies were performed with the first strain of Ormia depleta, focused on establishment and range (Frank et al. 1996), seasonality (Walker et al. 1996), effectiveness in controlling mole crickets (Parkman et al. 1996), techniques for rearing in the laboratory (Wineriter and Walker 1990) and food source of the adults (Welch 2000). This "old" strain was in culture from 1987 (Frank et al. 1996) until 1999, but it is the "new" strain that is now in culture to supply populations. Life History of Ormia depleta in the Laboratory

Wineriter and Walker (1990) determined the life history of the first strain under laboratory conditions at 23-250C. The life cycle of 0O. depleta, from oocyte to adult, is about 36 d. Numbers of planidia/female at time of examination range from 70 to 310, mean = 187 � 45, in laboratory-reared females. Most gravid females contain eggs and immature larvae as well as planidia, suggesting that females deposit larvae over many





13


days. Emergence of adults occurs very early in the morning and spans a 3-4 d period. Emergence of males generally precedes that of females. Sex ratio is usually 1:1.


Insect Modeling

Numerous independent variables influence arthropod development but none is

more important than temperature. In fact, the effects of temperature are so profound that most phenology models consider only this one variable. Other biotic and abiotic variables may influence development times, such as food quantity and quality, population density, humidity, rainfall, pH, and photoperiod (Hagstrum and Milliken 1988, Wagner et al. 1991). Population factors can also influence development times, for example genetic differences (Wagner et al. 1991) and gene expression (Gilbert and Raworth 1996).

The temperature dependence of insect development has been frequently

investigated. Temperature affects both the time of development as well as fecundity; consequently, the appearance and dynamics of insect populations in the field are often dictated by ambient temperatures. For this reason there has been considerable interest for a long time in the temperature relationship of development time and fecundity mainly of filthy, annoying, and disease-transmitting insects in order to apply it in predicting the outbreak-time and the dynamics of pests (Ratte 1985).

Predicting the seasonal occurrence of insects is essential for accurate scheduling of census samples and control tactics. These predictions require an understanding of insect development, and because temperature has a major effect on this process, the thermal requirements of development are often a basic to prediction (Wagner et al. 1984b).





14


Temperature determines life of organisms more than do any other environmental factors. The cause of this temperature sensitivity lies in the fundamental fact that chemical compounds build all organisms and that all processes of life are made by chemical reactions that follow the laws of thermodynamics. Temperature not only acts upon the rate of all chemical reactions according to the Arrhenius equation, but also causes conformational transitions of proteins, phase transition of lipids, change in the structure of water, etc. (Alexandrov 1977, cited by Hoffman 1985). However, many results from field observations and laboratory experiments in which the insects have been subjected to fluctuating temperatures cannot be brought into this concept. A too simplistic physiological concept of the temperature action might, however, lead to erroneous insight into the insect's capabilities of adapting to various thermal environments (Ratte 1985).

Constant vs Fluctuating Temperature Influence on Insect Development

Generally the velocity of insect development increases with elevation of temperature. Many workers have constructed hyperbolae by plotting the constant temperatures at which the insects were reared against the duration of development (Wigglesworth 1972). Janish (1930), cited in Wagner et al. (1984b), stated that the reciprocal of the hyperbolae represents the relative velocity of development, and the line based on the reciprocal shows the rate of acceleration of development with rising temperature. From these observations numerous empirical and physiochemical formulations of insect development have been made. Two such formulations are the daydegree or temperature summation rule and the non-linear inhibition model (Behrens et al. 1983).






15


Wagner et al. (1984b) presents a very clear explanation of what thermal

summation or rate summation means: many empirical and biophysical models describe the time versus temperature or the rate versus temperature curve of insect development. Models of rate versus development relationships are used widely to predict insect development times because mean daily (or hourly) rates can be accumulated under fluctuating temperature environments. These mean rates are summed to equal 1.0, at which time mean cohort development is complete. However models of time versus temperature relationships are primarily descriptive, although the inverse of these regressions also yields development rates. Constant Temperatures

Typically the development time of a larva decreases exponentially with increasing temperature. In almost all insects, at lower temperatures the development time decreases steeply with increasing temperature; at medium temperatures the slope of the curve becomes smaller until it reaches zero at the so-called optimal temperature for development, which lies usually close to the upper limit. The curve of developmental rate (reciprocal developmental time) takes a more or less sigmoidal course including an inflexion point mostly between 200 and 30'C, as well as a maximum typically between 250 and 300C (Ratte 1985). See Figure 1.






16


















Temperature
Figure 1. Theoretical linear and nonlinear model development rate functions.



The developmental rate curve displays a wide, approximately linear, range around inflexion point; a straight line often fits this part of the curve. Consequently, the time of development follows a hyperbolic function of temperature over this range. Similar relationships between developmental rate and temperature have been found frequently and have led to establishment of the "rule of thermal summing" and the classical method of"temperature summation" (Ratte 1985). TC = td (T-To)
Where:
td : time of development TC : thermal constant T : experimental temperature To : threshold temperature T-To: effective temperature above To.


In other words, the time of development multiplied by the effective temperature (T-To) is a constant (thermal constant). "Thermal summing" implies the existence of a distinct threshold temperature (To), at which the developmental rate becomes zero






17


(developmental zero; developmental threshold). The developmental threshold has to be regarded as theoretical because it can be extrapolated only from a linear rate relation and cannot be determined by rearing experiments; the threshold may fall outside the favorable range for development, and also the infinite developmental time at temperatures near the threshold makes its measurement practically impossible. Although the cessation of development below critical temperatures can be observed easily, the concept of real developmental zero has not been unequivocally accepted. When subjected to subthreshold temperature some insects develop faster after subsequent exposure to normal temperature indicating that some development also takes place below the threshold (Andrewartha and Birch 1954, Ratte 1985).

Ratte (1985) points to an important phenomenon of developmental rate curves:

the degree of temperature dependence, expressed as slope of the rate curve, or Qio, shows a wide variation among various insects and some do indeed possess the capability to develop independent of ambient temperature (exemplified with the carpet beetle, Attagenus megatoma).

Fluctuating Temperatures

The developmental time of insects reared under a temperature fluctuating regime has usually been found to differ from that under the corresponding constant temperature. Considerably higher acceleration may occur when the low temperature falls below the developmental threshold. The same has also been observed with nondaily alternating temperatures or single transfers from sub-threshold to favorable temperatures. Consequently, it has been concluded that under thermoperiod conditions also below the threshold "some development may take place": this is assumed to occur because the






18


temperature requirements of some rate-limiting development steps are nevertheless satisfied (Ratte 1985).

In cases where in one species different thermoperiods have been tested within the entire favorable range for development, the following rule has often been observed: temperature fluctuations in the range below the inflexion point of the developmental rate curve may cause an acceleration in development, while the converse is true for thermoperiods alternating about the temperature optimum; there are slight or no deviations in developmental time for constant temperatures with fluctuations around medial temperatures of the favorable range. This rule was propounded quite early and can also be derived from the non-linearity of the developmental rate versus temperature relation established under constant temperature (Ratte 1985). Applicability of Constant and Fluctuating Temperatures in Insect Developmental Models

Most of the models, called physiological models, transform temperature into a physiological scale, such as the classical degree-day summations (Wagner et al. 1984b). To describe field stage frequency data- that is number of individuals in each stage of development- in most cases parameters are fitted in two steps. First, parameters of the physiological scale are estimated from constant temperatures in the laboratory. Second, the parameters of the phenological model are estimated from the proportion of individuals in each development stage on the physiological scale. The applicability of the phenological model requires that thermal constants, determined from constant temperatures in the laboratory, can be extrapolated to natural conditions (Wagner et al.





19


1984b, Candy 1991, Manel and Debouzie 1997a). However, development rates at constant temperatures often differ from those at fluctuating temperatures (Worner 1992).

Manel and Debouzie (1997a) say that deriving the relation from different constant temperatures may be questioned; a model aimed at predicting insect development in natural populations requires other experimental designs - by recording temperature directly in the field or in the laboratory under fluctuating temperatures. Degree-day models extrapolated to natural conditions may or may not be successful for many reasons not related to temperature.

Studies of the effect of fluctuating temperatures on insect development often

show that, given the same mean temperature, insects appear to develop at rates that differ under fluctuating conditions from those at constant temperatures. Such studies may indicate that development is accelerated, decelerated, or the same as that under constant conditions. Such variability in reports appears to arise from studies where a limited temperature range is investigated (Worner 1992). Temperature-Dependent Models for Studies on Insect Development

With the advent of computer simulation to the applied biology sciences, there has been increased interest in mathematical models that describe organism growth and development as a function of temperature (Sharp and DeMichele 1977). Many empirical and biophysical models describe the time versus temperature or the rate versus temperature curve of insect development. Models of rate versus temperature relationships are used widely to predict insect development times because mean daily (or hourly) rates can be accumulated under fluctuating temperature environments (Worner 1992).

Models are important tools for predicting future events for development times in particular and population dynamics in general (Ellsbury 1991). Many pest management





20


systems rely on models of arthropod and host phenology, and there seems to be no end to the assemblage of new approaches contrived for this purpose (Wagner et al. 1991).

Statistical and deterministic simulation modeling rely on a complex process made of trials, errors, and gradual improvement of the simulations (Mespl6 et al. 1996) consider the major problem is to be able to quantify the quality of the simulations in order to know whether a modification of the concepts, the laws simulating the processes or the parameters improve it.

Linear Model

This model assumes a linear relationship between development rate and constant temperature; therefore, the number of degree-days (or thermal units) required to complete development is assumed constant at all favorable temperatures.

The duration of development is calculated by adding up the number of thermal

units (degree-hours or degree-days) contributed at each temperature (Wagner et al. 1984b cite very old reports: Candolle 1855 and Reibisch 1902 among others). The thermal constant or degree-days for completed development are most easily derived from the reciprocal of the slope of the regression line fitted to development rate data, obtained typically at constant temperatures (Worner 1992). This method has been applied to plant phenology (Schroder and Sondgerath 1996)

This degree-day approach is used widely because it requires minimal data for

formulation, is easy to calculate and apply, and often yields approximately correct values. However this approach is valid only over intermediate temperatures. Furthermore, the threshold temperature is often determined empirically by extrapolating the straight line to the temperature axis. For these reasons, the number of degree-days required for complete





21


development is too low at temperatures near the lower threshold and too high at or above the optimum. These limitations are important and should not be overlooked when interpreting the results of this technique (Wagner et al. 1984b, Manel and Debouzie 1997a, b).

The rate summation formula is

D = -r[T(t)llt


where the proportions of development taking place at each time increment are summed so that, when D equals 1, the insect has completed development.

This model measured in physiological units is not purely descriptive or statistical in nature but is biologically motivated. However it is based on population attributes rather than the progression of individuals through stages over time (Candy 1991).

In spite of inherent limitations of the thermal model, to that linear portion of

insects' developmental rate and the obligated definition of a lower threshold, this thermal model cannot be ruled out. As expressed by Liu and Meng (1999) insect development may occur most frequently at some temperature range considered in that degree-day summation and the model may fit well with natural varying temperature development although the logistic model better estimates development below the lower threshold. Logan et al. (1976) considered the degree-day model invalid for mite populations. Tanigoshi et al. (1976) showed that this model did not estimate development of field populations of mites.

Gilbert and Raworth (1996) said that in insects, development rates increase almost linearly with temperature over the normal range of operating temperatures. The published data cover more than 300 species, and in every case the effect of temperature on






22


development rate is effectively linear, but falls off at average daily temperatures higher than those normally experienced in the field. Lamb (1992) working with aphid clones detected some genuine curvature at low temperatures, but for practical purposes the relation was linear.

Belehradek (1935) (cited in Andewartha and Birch 1954) described the response curve with an exponential equation that becomes linear on a log scale. This approach does an inadequate job of describing development rates at the temperature extremes. Non-linear Models

Because of the difficulties of predicting insect development at temperature

extremes, empirical non-linear models have been proposed, attempting to explore this response beyond the intermediate curve temperatures. Non-linearity of development (the Kauffman effect, Worner 1992) or rate summing effect (Ratte 1995) implies that development rates at constant temperature differ from those at fluctuating temperatures. Models have been developed to incorporate nonlinearity and variable temperatures (Hagstrum and Milliken 1991, Manel and Debouzie 1997a, b).

Examples of nonlinear models are presented below (some are described in Table 2-1):

The symmetric and asymmetric catenary model. Proposed by Janish (1930) and cited by Wagner et al. (1984b), this model describes development time as a function of temperature. This equation combines two exponential curves, the reciprocal of which describes the accelerating phase of development rate up to the optimal temperature and the decelerating phase beyond the optimum. This model is criticized for inadequate fits and computational difficulties.





23


Second- and third-degree polynomials. These have been used to regress

development time against temperature although they have not been used for prediction purposes (Tanigoshi et al. 1976). Hagstrum and Milliken (1988) compared the polynomial model: d = t + t2 in which d is developmental time, and t is centigrade temperature, and was less accurate than the Sharp and DeMichele (1977) model in describing developmental time of nine species of stored-product Coleoptera under various temperatures, moisture levels, and diets.

Logistic equation. This describes development rates as a function of temperature (Davidson 1944). This equation is used widely although it is not very descriptive at one or both ends of the response curve. Manel and Debouzie (1997a) concluded that this model gives relatively close estimates of stage duration when applied to a single stage of development and when base temperature is known; but, when base temperature is misestimated, stage duration is either under- or overestimated). Liu and Meng (1999) demonstrated that the logistic model better fit field varying temperature, especially when temperature falls below the lower developmental threshold. This model gives better estimates than the Wang nonlinear model (Wang et al. 1982).

An exponential equation. Similarly, a normal distribution model is proposed by Pradhan (1945 and 1946 as cited in Wagner et al. 1984b). When compared with the logistic model, it yields better results (Siddiqui and Barlow 1972).

A modified sigmoid equation model. This model was proposed by Stinner et al. (1975) to describe the effects of temperature on development rates. This model does not work well at higher temperatures.





24


The two exponential equations model. Logan et al. (1976) described

developmental rates at intermediate and high temperatures. Logan's model was more descriptive than Stinner's modified sigmoid equation and third degree polynomial (Logan et al 1976). Lactin et al. (1995) modified Logan's model; it fitted better at lower temperatures in ten of eleven cases. They considered this model better than Sharp and DeMichele's model. Ryoo et al. (1991) considered this model appropriate in describing development of a larval parasitoid. Gould and Elkinton (1990) compared the linear model, Logan's model, and Sharp and DeMichele's model, in studying the development of a wasp parasitoid, Logan's model better estimated development at temperatures below 15 oC. No differences were observed at range temperatures of 150 and 300C.

A complex biophysical model. This model proposed by Sharpe and DeMichele (1977) describes the non-linear response in development rates at both high and low temperatures as well as the linear response at intermediate temperatures. Schoolfield et al. (1981) improved this model making it suitable to the nonlinear approaches. This model is considered as the most suitable to describing insect development rates as a function of temperature by Wagner et al. (1984b). Shaffer (1983) applies this model to 194 cases from the literature, failing in 20% in explaining variations in developmental rate, making this model not true for all species. Hagstrum and Milliken (1991) say this model provides an alternative to degree-day model accumulation method for prediction of developmental times at fluctuating temperatures in the field. Stochastic Models

Insects reared under identical conditions develop at different rates due to inherent differences among individuals in a population. This stochastic aspect of development is





25


important, although usually overlooked, in population models of poikilotherm

development. It is not surprising, therefore, to find many models, linear and nonlinear,

that describe the effect of temperature on insect development; however, most of these

models only predict development time from mean rate versus constant temperature

relationships. Usually, no consideration is given to the variability in insect development

time among individuals. On the other hand, models that consider the stochastic aspects of

development time use frequency distributions to determine the probabilities of

development time for individuals in a population (Wagner et al. 1984b, Candy 1991,

Manel and Debouzie 1997b).

Wagner et al. (1984 a) say that three general stochastic approaches are

recognized:

Monte Carlo techniques use pseudorandom numbers to estimate the development
time of individuals in a sample population (Menke 1973, Hardman 1976). The
disadvantages of this approach are that it is computationally time-consuming
because random numbers must be generated for each individual in the same
population, and its outputs must be interpreted statistically. Phelps et al. (1993)
consider this method as feasible because it uses a fixed number of individuals
from one generation to the next and simulates the timing of events rather than the
population dynamics of the insects.
Stinner et al. (1975) developed a cumulative function that is fitted to the
developmental times at <1, 25, 50, 100% of measured distributions. This
probability distribution provides a good description of distributions at different constant temperatures, but when used in population models, the function is not
fitted directly to measured distributions.
The final stochastic approach mentioned by Wagner et al. (1984b) applies a
probability function directly to the measured distributions. This approach has a
mechanistic basis proposed by Sharp et al. (1977). These investigators used
normal and quadratic probability density functions to describe the variability in
development rates. Although these functions often provide good results, their
goodness-of-fit capabilities are unsatisfactory when distributions of rate are
asymmetric.

Wang model. Described by Wang et al. (1982), this gives a very good fit to data

over a whole range of constant temperatures that permit complete development. Liu and





26


Meng (1999) compared this model with some other nonlinear models, such as Sharpe and DeMichele (1977), Stinner et al. (1975) and Logan et al. (1976), and found that Wang model offered the highest R2 and the least estimation bias.

In order to try to minimize the effect of asymmetry of distributions of rate, Wagner et al. (1985) propose the Weibull function to describe the distributions of normalized development time with good results. Ellsbury (1991) applies this same function on development of immature stages of clover stem borers at constant temperatures with good results. The predicted distribution of development time can be used to optimize census sampling and control procedures, or it can be used as input into larger population dynamics models (Wagner et al. 1985).

Continuation ratio model. This model deals with probability of individuals being in each stage, as proposed by Candy (1991) and re-evaluated by Manel and Debouzie (1997a, b). This model applied to stage (multistage) frequencies obtained by random daily sampling, gives unbiased estimates only when a common physiological scale applies to each stage. When stages differ in physiological scale it gives nonindependent estimates of duration. The Weibull distribution is integrated to this model. It gives a better fit than the logistic model.

Mixed Models

Recent phenological models incorporate temperature-dependent development and inherent stochastic variation. This approach estimates the proportion of insects in a given development stage as a function of accumulated heat or degree-days (Dennis and Kemp 1988)





27


Stage successive logistic model. Manel and Debouzie (1997a) combined rate summation, logistic equation and a stochastic description of individuals in a binomial probability. This model gives a better estimate than the continuation ratio model (Candy 1991).

Conclusion

As a way to express something that reflects what I perceived through the writing of this paper, I have chosen to quote what Wagner et al. (1985) conclude: 'The introductory paragraph of Uvarov's (1931) classic monograph, Insects and Climate, states "Economic entomologists at the present day are no longer satisfied with merely recording the outbreaks of insect pests and with devising means for their control. They realize more and more that their chief aim and highest ambition must be to foresee and to prevent outbreaks. In order to do this, all conditions accompanying and causing outbreaks must be thoroughly investigated and elucidated; in other words the epidemiology of insect pests must be the central problem of economic entomological research, which should be carried out from the ecological point of view."

There is a greater opportunity today to predict population events, a requisite to prevent outbreaks. Vast amounts of biological and economic data are available on most major pests; and new approaches to population research and modeling, as well as inexpensive high speed computers, provide the means to identify, integrate and apply this knowledge for predictive purposes.

Recent advances in entomology have occurred as a result of interdisciplinary

research. Specialists from outside of entomology have, and will continue, to apply new techniques to our field. These techniques must be integrated into the mainstream of entomological application.





28


Table 1. Insect models cited in the Review of Literature chapter.


Sharpe and DeMichele (1977), modified by Schoolfield et al. (1981):

RH0T)= 298.15 R 298.15 T

Ie HL 1 1)]+ exHH 1 1


Where:
r(T) =mean development rate at temperature T(OK R =the universal gas constant (1.987 cal degree mole-) RH025= development rate at 25 oC (298.15 oK) assuming no enzyme inactivation HA =enthalpy of activation of the reaction that is catalyzed by a rate-controlling enzyme. TL = Kelvin temperature at which the rate-controlling enzyme is half active and half low-temperature inactive.
HL = change in enthalpy associated with low temperature inactivation of the enzyme. TH = Kelvin inactivation at which the rate-controlling enzyme is high active and half high-temperature inactive.
HH = change in enthalpy associated with high-temperature inactivation of the enzyme.


Wagner et al. (1984) say they are aware of no other poikiloterm development rate model
with greater flexibility, goodness-of-fit capability, or of stronger theoretical
foundation. Nevertheless, the use of this model is limited to those with experience
in nonlinear parameter estimation. To extend the Sharp and DeMichele model to
biologists they present a SAS-computer program for its use.

Weibull function:

F(x) = 1- exp(- x - y)/ r )
Where:
F(x) : the probability of complete development at normalized time x y, ri and 0 : are estimated parameters


Logistic equation. Liu and Meng (1999):

V(T)= K + exp(a - fP))
Where:
K, a and 0 : are regression parameters fitted to the data of mean development rates V(T) K means the upper asymptote of this shallow sigmoid curve. Table 1. Continues.






29


Logistic regression equation. Manel and Debouzie (1997a):


1
Pi =
1 + exp(ao +aDD,)

Where:
ao and a1 : are parameters to be estimated DDI : the number of degree-days accumulated at time ti


Continuation ratio model. Candy (1992):

exp(bo, + b,I)
P i+ exp(b01 + b DD,) 1- 1 Pk exp(bo, + b,,DDi) k=1 1+ exp(bo, + b,DD,)

k=1

Where :
bo and bl: are parameters to be estimated. DDI : the number of degree-days accumulated at time ti


Wang model. Wang et al. (1982):


V(T) +exp(Fr(T T)T - T T--ex T
+exp(- - exr(T-To)) T))


Where:
V(T): is the mean development rate at temperature T(oC) H : is the value of the upper asymptote of the curve r is the exponential increase rate To : is the optimum temperature for development TL and TH : are, respectively, the minimum and maximum temperature boundaries
: boundary width at the lower and upper temperatures.














CHAPTER 3
DEVELOPMENT OF THE WASP LARRA BICOLOR F.


Development and reproduction of the wasp Larra bicolor at constant temperatures Introduction

Scapteriscus mole crickets are major pests of turf and pasture grasses in the

southern USA. The three mole cricket species that arrived about 1900 as immigrants have being under classical biological control investigation since 1979 under the University of Florida mole cricket research project (Walker 1985a).

Two introductions of the ectoparasic digger wasp Larra bicolor have been made in Florida. Their different success is reported in chapter 2 (pages 10-11). The role of the environmental temperature has been considered the main limitation for Larra bicolor populations to establish and to disperse in northern Florida and in other southeastern states.

As a way to enhance the biological control of this parasitoid, it is important to

determine its vital statistics that may explain its seasonal phenology. These vital statistics are an important and basic component for understanding the dynamics of a species (Lowry et al. 1992). Fertility life tables are appropriate to study the dynamics of animal populations, especially arthropods, as an intermediate process for estimating parameters related to the population growth potential, also called demographic parameters (Southwood 1978, Maia et al. 2000).






30





31


The general biology of the wasp L. bicolor has been poorly studied. Castner

(1988) reported developmental time of L bicolor under a single temperature regime, but failed to report variance.

Insects develop within a species-specific temperature regime. The inverse

relationship between developmental rate and temperature results in a sigmoid curve with a straight line at the intermediate temperatures. This curve is a fundamental characteristic of an insect life history (Taylor 1981, Wagner et al. 1984b).

The most widely used approach to study the effect of temperature on

developmental rates of poikilothermic organisms is the thermal summation or degree-day

(DD) model. This method is one of the oldest and most commonly used for insect modeling and for predicting biological events controlled by temperature accumulation (Wagner et al. 1984a, b, Higley et al. 1986).

The objective of this study was to investigate the effect of temperature on development, survivorship and reproduction of Larra bicolor. Materials and Methods

Development

Placing a female wasp into a 200-ml plastic vial and then introducing the host

induced parasitism of hosts (S. borellii). Parasitism occurred almost immediately most of the times. Parasitized hosts were placed individually in 200-ml plastic cups filled with wet-sterilized sand.

Development and survivorship were studied in incubation chambers at seven constant temperatures, 17.5, 20.0, 22.5, 25.0, 27.5, 30.0 and 32.5oC, at 14:10 (L:D) photoperiod. Development and survivorship were observed and recorded every six hours





32


until adults emerged. Initial wasp cohorts consisted of 29, 34, 36, 38, 38, 39, 39 individuals respectively per temperature studied.

The relationship between temperature and developmental rate (=1/days) was estimated by linear regression analyses. Pooled developmental times for males and females were used to fit the linear models. The lower developmental threshold (t), the temperature at which development is negligible, was estimated by the "x-intercept" method. Values of t were obtained as follows: t = -a/b, where a and b are parameters from the regression equations (where x = temperature, and y = developmental rate).

The degree-days required for development were calculated by using the equation DD=y(T-t) where 'y' is the developmental time (days), 'T' is the temperature during development, and 't' is the lower developmental threshold. DD were determined for each individual at all temperatures where development was recorded and then averaged by stage.

Life-Fertility Tables

Reproduction of two generations developing during different seasons of the year was studied.

The summer cohort. This consisted of wasps collected during July. Wasps

collected from the field were provided with host mole crickets and their parasitic progeny were kept in the laboratory at 26 + 0.5oC, 65% RH and 14:10 (L:D) photoperiod. Emerging adults were confined in cages, grouped by age. Adult wasps were fed with diluted hummingbird nectar and honey. The cages were additionally provided with a blooming nectar-source plant (Spermacoce verticillata). Summer wasps were given host mole crickets starting a day after emergence, 5 to 7 times a day until the wasp's death.





33


Parasitized hosts were kept under the same laboratory conditions as the adults, until wasp stages completed development.

The autumn cohort. This consisted of wasps from parasitized mole crickets that were kept under field conditions until parasitic stages completed development. Emerged wasp adults during October were subjected to tests. Wasp adults were provided with host mole crickets as for summer wasps. Parasitized hosts were confined underground in screened chambers of 30 cm depth until adult emergence.

Reproductive statistics were summarized in the form of age-specific life fertility tables for both cohorts.

Results

Development

Developmental times for all stages were inversely related to temperatures from 25.0 to 30.00C (Figure 2). The temperature of 32.5oC caused a varied effect on the stages of this parasitoid; the egg stage maintained a similar reduction in development compared with previous temperatures, larva and prepupa stages reached a plateau in development compared with previous temperature (30.00C), and pupa stage slowed down development as almost four times the previous temperature (Table 2). No development was observed at 17.5 and 20.0oC; only one wasp developed at 22.50C. The total development ranged from 142.90 to 40.18 days at 32.5 and 30'C, respectively. Apparently, the upper temperature threshold was reached at 32.50C. Pupal development required from 72.5 to 91% of the total development time. Although the duration of stages varied with temperature, the proportion of total developmental time spent in each stage was consistent regardless of temperature.






34




20 LARVA


15


a 10 EGG


5
PREPUP


20.0 22.5 25.0 27.5 30.0 32.5 35.0

150
0
TOTAL x
o
100 8 x
0 PUPA


50



0
20.0 22.5 25.0 27.5 30.0 32.5 35.0 Temperature OC Figure 2. Developmental time of Larra bicolor stages and total at constant temperatures.






Table 2. Development (days) of Larra bicolor life stagesa at constant temperatures.

Temp. (C) n Egg n Larva n Prepupa n Pupa n Total
32.5 4 3.23 � 0.09 (2.3) 11 7.33 � 020 (5.4) 4 0.47 � 0.12 (0.33) 4 130.02 � 17.38 (91.0) 4 142.90 � 17.60
30.0 14 3.75 � 0.28 (8.6) 7 7.38 � 0.01 (16.6) 8 0.44 � 0.02 (0.01) 8 35.07 � 0.69 (75.3) 8 47.50 + 0.63
27.5 8 4.74 � 0.63 (9.3) 8 9.87 � 0.15 (18.4) 6 0.66 � 0.09 (0.11) 6 41.39 � 0.80 (72.5) 6 57.06 0.95
25.0 11 5.09 � 0.28 (6.8) 4 13.30 � 0,40 (17.2) 9 0.72 � 0.06 (0.01) 9 62.30 � 3.76 (75.9) 9 82.10 3.76
22.5 1 8.63 � 0.00 (7.8) 1 18.13 � 0.00 (18.2) 1 2.00 � 0.00(0.02) 1 81.88 � 0.00 (74.0) 1 110.63 � 0.00
aMean + SEM; percentage of developmental time required for each life stage given in
parenthesis.






35


The survivorship per stage is shown in Figure 3 and described according to Southwood (1978). The maximum survivorship occurred at 300C followed by 25 and 27.50C. Survivorship of cohorts reared at 17.5 and 20.0C followed a type IV curve (100% mortality occurs prior to the adult stage). Cohorts reared at 22.5 and 32.5oC exhibited a type III curve. The remaining cohorts followed a type I curve (mortality acts most heavily on the old individuals).


Egg Larva PP Pupa

100 TEMPERATURE oC

80 -- 32.5 ~ -- 30.0 L e ----- 27.5
o 40 -e- 25.0
-.- 22.5
20 -x- 20.0
0 -+- 17.5
0 100 200 300 400 500 600

Physiological time (Degree-Days > 16.0*C) Figure 3. Survivorship of Larra bicolor life stages at constant temperatures



Linear regression parameters describing the relationship between developmental rate and temperature for all stages are presented in Table 3. Data for 22.5oC were excluded because only one wasp completed development at this temperature. The data at 32.5oC were not used in the regression model for larva, prepupa and pupa because the values were considered beyond the linear portion of growth responsive to temperature (Sharpe et al. 1977), (Figure 4). Coefficients of determination ranged from 0.71 (prepupa) to 0.94 (larva), indicating a good fit in all cases. At the egg stage, data for 32.50C were






36


included, but for all remaining stages data were used for temperatures from 25.0 to 30.0C, because the upper developmental threshold was apparently reached.


Table 3. Linear regression models (y = developmental rate, x = temperature), lower
developmental thresholds (t), and degree-days required for Larra bicolor life
stages and total to develop at constant temperatures.


Degree-days
Stage y intercept�SEM Slope+SEM R2�SE t(C) (DD�SE) Egga -0.4015+0.0348 0.0221�0.0012 0.92�0.0182 18.143 45.446�0.7107 n = 31

Larvab -0.2211�0.0159 0.0119-0.00058 0.94�0.0063 18.620 84.425�0.9331
n = 27

P-pb -3.7870�10.7173 0.2032�0.0263 0.71�0.2838 18.640 5.049�0.1652
n = 27

Pupab -0.038910.0048 0.0024�0.00018 0.88�0.0019 16.01 413.135�5.642 n = 27

Totalb -0.03126�0.00276 0.0018�0.0001 0.93�0.001 16.880 540.84+6.035 n = 27
P-p = prepupa; '25.0, 27.5, 30.0 and 32.50C data only; b25.0, 27.7, and 30.00C data only.



The lower developmental threshold temperature (t) obtained by the "x-intercept" method ranged from 16.01 (pupa) to 18.64 (prepupa). In fact, "t" values for all single stages were close to 18 except for the pupa, and then for total development. The degreedays (M + SEM) above their respective "t", required by the wasp to develop during different life stages are presented in Table 3. The degree-days for the pupal stage to complete development account for almost 76% of the degree-days required for total development.






37





0.40 EGG


__ 0.30

0.20 LARVA 0.10 PUPA S- TOTAL
0.00
22.5 25.0 27.5 30.0 32.5 35.0


4.0

PREPUPA
3.0


2.0


1.0 t


0.0
22.5 25.0 27.5 30.0 32.5 35.0 Temperature oC



Figure 4. Developmental rates of Larra bicolor at constant temperatures. Symbols show observed data. Line trend fitted to the linear response. Reproduction

Total fecundity (eggs laid per female) was recorded from the two cohorts,

summer and autumn wasps (Table 4). Summer wasps oviposited 16 eggs per female, in contrast with 12.5 from autumn wasps. Despite the apparent difference, the means are not statistically different. Egg hatching was statistically the same, with values close to 100% hatching. The female longevity was almost the same in the two cohorts, with a female adult survival of 10 days.





38


Table 4. Population reproductive statistics for Larra bicolor from two seasons in
Gainesville, FL (29040'N).


Season n Fecundity Fertility Female longfemales (eggs per female)ab (% egg hatch) vity (days)a, Summer 7 16.0�2.23a 98.2b 10.14+1.20a Autumnd 4 12.5�2.84a 96.4b 10.50�3.17a aMean+SEM
bMeans followed by different letters are significantly different. CLab-reared wasps at 26 � 0.5oC.
dField-reared wasps, emerged during October. Life-FertilityTables

Life fertility tables were constructed for all cohorts to calculate reproductive statistics (Table 5). The net reproductive rate (Ro), which represents the females produced per female per generation of the two populations, overlaps the confidence limits, which makes them statistically not different. The intrinsic rates of increase (Rm) did not overlap each other, meaning that the two populations increase at different rates, with autumn wasps having the slower population growth and being statistically different. The mean generation times (Tc) did not overlap between the two cohorts.

The differences observed in Rm and Tc are explained by the fact that the autumn generation developed under lower temperatures than did the summer generation. These differences are shown by the oviposition periods in Figure 5, where the effect of contrasting temperatures caused delay of oviposition in the autumnal population, and faster oviposition time in the summer population.





39


Table 5. Reproductive parameters for Larra bicolor from two seasons in Gainesville, FL
(29040'N).


Season Road Rmbd Tc (days)cd Summer 1.76�0.61 0.011+0.006 51.14�5.29 Autumnf 1.37�1.00 0.003�0.007 109.47�18.30 aRo = net reproductive rate
bRm = intrinsic rate of increase
cTc = mean generation time
dMean_+SEM
eLab-reared wasps at 26 � 0.5oC.
fField-reared wasps, emerged during October. Discussion

Development of the digger wasp Larra bicolor was characteristic of

poikilothermic organisms where developmental time decreases (and developmental rate increases) as temperature increases (Tables 4 and 5, and Figure 4) (Sharpe and DeMichele 1977, Wagner et al. 1984a). The linear developmental rates occurred at temperatures from 25.0 to 300C for all stages but the egg, where the developmental rate kept its linearity up to the highest temperature of 32.5oC (Figure 4). The upper temperature threshold was passed at 32.50C for all stages but egg, so that the upper threshold must be at some point between 30 and 32.5oC.

Castner (1988) reared the wasps at 26 �2 oC and 65% relative humidity, showing a broadest developmental variability for this parasitoid compared with the data reported in the present study where the wasps were reared at temperatures with an accuracy of

0.50C which explains the lower variability in developmental times of the wasp.

The lower threshold value observed in the pupal stage along with the higher

number of degree-days required for the pupal stage to develop (almost 76% of the total





40


development) is a strong indicator that this is the resistant stage to inappropriate environmental temperatures at both extremes (Table 3 and Figure 3). It is important to note that the developmental data presented in this study were obtained from wasps developing during summer and exposed to controlled summer-like conditions. The developmental rates under autumn-like conditions at which this parasitoid is able to develop were not studied, and remain unknown.

L. bicolor developed and survived over a broader temperature range (Table 2) than normally occurs underground during spring to fall seasons in Gainesville. The host mole cricket, S. borellii, tolerated extreme temperatures during the study, but the parasitic wasp failed to develop at 17.5 and 20.00C, and only one wasp completed development at 22.5oC. In previous studies (not presented in this dissertation) rearing the three mole cricket species, none survived more than 15 days at temperatures of 15 and 10C nor at 35oC or higher. Those temperature ranges must be outside the limits for these hosts to survive and develop.

Temperature regimes, representing contrasting conditions at which the two

cohorts were reared, did not affect reproductive statistics of the parasitoid, considering fecundity, fertility and adult longevity even though the population reproductive parameters estimated from the life-fertility tables (Table 5) showed differences in the intrinsic rate of increase and in generational time, and then in oviposition time (Figure 5). The temperature regime represented by the season of the year did not alter the reproductive capacity of Larra bicolor. From these results it can be inferred that the temperature regime (say the season of the year) is not an important constraint of the reproductive performance of this parasitoid, at least in Florida. Other factors such as





41


suitable hosts and availability of food source for adult wasps should play the more important role restraining the populational increase of Larra bicolor.

The developmental time was defined per stage and total life cycle. The lower

threshold was determined for all the stages. The upper temperature threshold was reached at 32.5 for all stages but the egg. The degree-day values were estimated for all stages. The reproductive parameters of wasps estimated at summer-like temperature (constant temperatures) were similar to those reared in the field (autumn season).



6
Autumn population
N=5
5 i

4
Summer population
,N=8
0



i II L


0 50 100 150 200 Age (days after emergence)


Figure 5. Oviposition time of wasps Larra bicolor reared at summer-like and autumnal
seasons in Gainesville, FL; daily oviposition per wasp is represented by different
line pattern per season.





42


Development of the wasp Larra bicolor at field-varying temperatures. Introduction.

Studies of the sphecid ectoparasitoid Larra bicolor have been devoted to its behavioral ecology and foraging (Fowler 1989), host searching and mate-searching (Castner and Fowler 1987), and establishment and dispersal (Frank et al. 1995). The only information available about the life cycle and field efficiency was obtained by Castner (1988) with the Puerto Rican biotype that established a population in south Florida but failed to do so in central and north Florida. Practically no information about the Bolivian biotype is available. Extensive data are needed to examine the effect on host mole cricket populations, with geographical, seasonal, and annual variation.

Temperature plays a critical role in determining the rate of development, survival and reproduction of host insect species. The duration of development from birth to first reproduction is one of the most important insect life-history characters influencing numerical change in insect populations (Wagner et al. 1984a). Because the rates of development in insects under natural conditions are largely determined by temperature, a realistic measure of the relationship between temperature and development rate is fundamental to any study of the population dynamics or of the management of any insect (Wagner et al. 1984a, Liu and Meng 1999). The degree-day concept is one of the oldest and most widely used entomological models; it assumes a linear relationship between development rate and constant temperatures, therefore the number of degree-days required to complete development is assumed constant at all favorable temperatures (Wagner et al. 1985). Daily temperature cycles that occur in the field are generally not considered by the degree-day method that uses data collected at constant temperatures in the laboratory to predict developmental times in the field (Higley et al. 1986).





43


Developmental times are known to differ between constant and fluctuating temperatures with the same mean (Hagstrum and Hagstrum 1970). Developmental time data collected at constant temperatures in the laboratory can only be expected to provide a rough estimate of developmental times in the field at fluctuating temperatures (Hagstrum and Milliken 1991).

The relationship of life stages of Larra bicolor to constant temperatures is shown in Table 2, resulting in the proposal of a degree-day model (Table 3). The model, based upon constant temperatures, must be evaluated by its goodness of fit to populations developing in the field.

This work describes the development of egg, larval, prepupal, and pupal stages of Larra bicolor, Bolivian biotype, under field conditions of Gainesville, FL (29040' N). The results of this work will be used, in a subsequent study, to evaluate the efficiency of the degree-day model proposed.

Materials and Methods

Female wasps were captured by net from patches of the food-plant Spermacoce verticillata L. in Gainesville, FL. After capture, the wasps were placed individually in snap-cap plastic vials, and kept alive by cooling them in a chest. The wasps were taken the same day to the laboratory and placed in groups of 10 wasps into a 30x30x12 cm metal-framed screened cage containing a flowering S. verticillata. Water, diluted honey, and diluted artificial hummingbird nectar (@Perky-Pet Products Co.) were supplied also. Unless otherwise indicated the host mole crickets were Scapteriscus borellii adults that were captured in traps baited with synthetic mole crickets song at the Ona REC, FL. The hosts were shipped in buckets partly filled with moist vermiculite.





44


Devices made of PVC pipe were designed to individually confine parasitized hosts underground at a range of depths. These pipe-chambers could be easily buried vertically and are could be easily removed from the ground with little disturbance of mole cricket tunnels inside them. The pipe-chambers were subdivided into four sections with the assumption that each section represented a particular thermal regime and the total individuals confined at each depth can be considered as part of an independent experiment. Thus, it was possible to handle four independent experiments at the same time, covering different possible thermal regimes at which the host mole crickets naturally live and thus the parasitoids develop. The depths of the four chamber sections were: 0-15, 15-30, 30-45, and 45-60 cm. (Table 6).

The pipe-chambers consisted of 5 cm inside diameter x 75 cm long PVC pipes cut lengthwise. Two 2.5 cm wide x 25 cm long windows were cut at the center of each half, covered with aluminum window screen, and sealed with hot melted silicone. The inside of each pipe was subdivided into 15 cm sub-chambers, by placing non-corrosive metallic mesh, fully matching the inner diameter of the pipe once closed, also fixed with hot melted silicone. Starting from the bottom, the pipe-chambers were subdivided into four sub-chambers (15 cm tall x 5 cm diameter), leaving a portion of 15 cm intended to be above the ground level for handling purposes. This chamber design allows the host mole cricket to move freely in a volume of 300 ml of sand. The screened windows allow the sand inside the chambers to be in close contact with the sand outside, allowing the flux of air and moisture from the external soil into the chambers. The temperature was monitored inside and outside the chambers at the different depths, with no difference observed.





45


Moisture was not measured but was estimated by touch; tap water was added when considered necessary.

Each half chamber was filled with autoclaved moist sand, lightly compressed.

Artificial tunnels were made by compressing the moist sand by finger. The halves of the chamber were joined as a sandwich by placing rubber bands at both extremes and at the middle. To check the status of host or parasitoid, or to feed the host, the pipes were withdrawn from the soil, the rubber bands were removed, and each half of the pipechamber was rolled gently to a concave position to avoid disturbance of the contents. Table 6. Codes and depths of sub-chambers and thermocouples where parasitized mole
crickets were confined to study the development ofLarra bicolor underground in
the field.

Sub-chambers Thermocouple Code Depth (cm) Depths (cm)
A 0-15 7.5 B 15 - 30 22.5 C 30 - 45 37.5 D 45-60 52.5

The land where the pipe-chambers were buried was kept free of weeds for the duration of the experiments. It was then assumed that the irradiation, aeration, and moisture (from environmental humidity and precipitation) were evenly distributed without interference from vegetation. The presence of patches of weeds would have modified the flux of these factors. A rectangle of land, Im wide x 6m long, located near the Entomology and Nematology Department building, was tilled to 75 cm depth to allow the manipulation of the pipe-chambers in and out of the ground. Before introducing the parasitized hosts, the pipe-chambers were buried underground for a few days. The





46


chambers were symmetrically distributed, allotting five chambers to the width and 20 chambers to the length of the experimental land. Parasitized hosts were randomly allotted to the sub-chambers.

The soil temperature at which the parasitized host was confined in the subchambers was recorded by placing a thermocouple in the middle of the corresponding sub-chamber (Table 6). Temperatures were recorded as follows using one pipe-chamber buried in the center of the experimental land. Four thermocouples were used, one at the depth of each of the four pipe sub-chambers. The thermocouples were connected to a digital temperature datalogger (@HOBO Mod H8-4, Onset Co.) programmed to record temperatures every 30 min. The temperature recorded was downloaded to a computer for further analysis. Another digital temperature datalogger was used to record air temperature, placing the thermocouple at 1.8 m height above ground under a 20x20 cm shed roof.

The parasitism of hosts was achieved following the procedure described by

Williams (1928) cited by Castner (1988). Placing a female wasp into a 200-ml plastic vial, and then introducing the host S. borellii always resulted in immediate parasitism. Wasp activity was stimulated by adding a heat source with a 500 w incandescent lamp for a few minutes at 300C before offering a host. Maintaining the wasps at a temperature of 27oC was observed to increase their feeding activity; otherwise most of the wasps remained hidden in the plant or in the comers of the cage.

A few minutes after they were parasitized, the hosts were placed in moist sand

inside the pipe-chambers and buried underground. The parasitized hosts were fed twice a





47


week with the same Purina cricket chow food used in the laboratory culture. To prevent hosts dying of dehydration, tap water was added if the sand in the chambers seemed dry.

Four wasp groups captured from the field were used to parasitize host mole crickets. Parasitism of all hosts was achieved in as few days as possible to reduce variability. Each group of parasitized hosts was considered a cohort (Table 7).





Table 7. Description of the cohorts of the wasp-parasitized hosts by numbers and dates.

Cohorts Parasitized hosts1 Parasitism dates
No. No. Initial Last
1 65 22-VIII-2000 31-VIII-2000 2 38 1-IX-2000 11-IX-2000 3 28 30-IX-2000 1 -X-2000
4 20 27-X-2000
'Scapteriscus borellii adults.

Hosts of cohort 1 were randomly allotted to the first three sub-chambers (A, B

and C). Hosts of cohorts 2 to 4 were randomly allotted to the four sub-chambers (A, B, C, and D).

The response variable was the time to develop for egg, larval, prepupal and pupal stages. The observation frequency varied as follows: egg development was recorded every six hours starting 24 hours after oviposition. Larval development was recorded daily starting five days after egg eclosion, continuing until the wasp grub reached a large size, or until the parasitic grub killed the host, then recordings were made every six hours until the beginning of the prepupal stage. Prepupal development was recorded every six hours until pupation was completed. Once pupation was completed, the metallic mesh dividing the sub-chambers was removed to allow the emerging wasps to dig up to the





48


surface. Pupal development was recorded daily, starting 20 days after pupation occurred, until adult emergence. Eventually, tardy pupae were visually inspected to assess whether they were still intact and to ensure the wasp had not escaped. Adult emergence was recorded by the presence of the adult wasp trying to escape from the chambers, but impeded by the fiberglass window-screen at the top of the pipe-chamber; daily inspection of the chambers was made during the experiments. Results

The developmental times for stages and cohorts are presented, sorted by subchambers, stages and total development in Table 8, where the egg stage showed no difference among the sub-chamber depths. The larval stage showed no clear differences in development in cohorts 1 and 2, and no differences in cohort 3. The prepupal stage showed no difference in development related to sub-chamber depth in all the three cohorts evaluated. The pupal stage showed no difference among the sub-chambers where comparisons were possible. The total development showed statistical differences in cohorts 1 and 2. The differences observed in egg, larval and pupal stages were not highly significant. The mean values recorded in pupal and total development were numerically more contrasting among sub-chambers. In order to explain the possible differences in development among cohorts, the developmental times were pooled by cohort (without considering the sub-chamber) and presented in Table 9. The egg development was different among the three cohorts. Larval development in cohort I was highly different from that in cohorts 2 and 3. The prepupal stage showed no difference among cohorts. Pupal and total development showed no statistical difference among cohorts even though the mean values were very distinct, particularly in cohort 1, which was three times less than in cohort 3. The lack of differences in the pupal and total development is highly






49



related to the high variability of the respective estimates, which may mask the possible


differences.





Table 8. Development time ofLarra bicolor at various depths underground in the field at
various dates in Gainesville, FL (29040'N).

Cohort 1 Egga'b Larvaa,b Prepupab Pupaa.b Totalab Depths n Mean � SEM n Mean � SEM n Mean � SEM n Mean � SEM n Mean � SEM
A 11 3.81 �0.12a 10 9.18 � 0.28 a 11 0.78 0.09 a 2 44.23 3.3a 2 58.42 3.4a B 8 3.65 0.11 a 7 8.85 0.27ab 9 0.79 0.09a 2 45.31 �3.6a 2 57.89 3.9a C 10 3.60�0.14a 9 8.17 � 0.19 b 13 0.87 0.09a 4 104.95�62.9a 4 117.32 62.8b

Cohort 2 Egg',b Larvaab Prepupaab Pupaab Totalab Depths n Mean �SEM n Mean �SEM n Mean �SEM n Mean SEM n Mean �SEM
A 10 3.33 � 0.1 a 2 11.02 � 0.69 a 1 0.58 � 0.0 0 0
B 12 3.36 0.1 a 3 11.18 �0.26a 3 0.60 �0.1 a 1 238.00 0.0a 3 246.81 �3.83 C 4 3.33 � 0.1 a 1 9.96 � 0.00 1 0.50 � 0.0 0 1 247.71 � 0.00 D 3 3.56 � 0.2 a 2 11.40 � 1.94a 2 0.67 � 0.2 a 2 160.89 + 69.9 a 1 112.92 � 0.00

Cohort 3 Eggab Larvaab Prepupaa.b Pupaab Totalab Depths n Mean � SEM n Mean SEM n Mean � SEM n Mean � SEM n Mean � SEM
A 6 3.89 � 0.04 a 6 14.50 � 0.43 a 5 0.87 � 0.07 a 0 0 B 2 3.96 � 0.37 a 3 12.90 � 0.49 a 2 0.79 � 0.21 ab 0 0
C 4 4.02 � 0.11 a 4 13.38 � 0.38 a 4 0.55 � 0.03 b 2 210.27 � 10.27 2 227.33 10.46
D 2 3.94 � 0.39 a 2 13.06 � 1.02 a 2 0.54 � 0.08 b 0 0

Cohort 4 Eggab Larvaa,b repupaab Pupaa.b Totala,b Depths n Mean � SEM n Mean � SEM n Mean � SEM n Mean � SEM n Mean � SEM
A 0
B 0
C 2 198.75� 6.00
D 0
aMean days to complete development � Standard error. bDifferent letters among depths indicate statistical difference; Duncan's test when three or

more means; 't' test when two means.




Table 9. Developmental time of four cohorts of Larra bicolor stages and total, averaged
over all depths underground in the field in Gainesville, FL (29040'N).

Eggab Larvaa,b Prepupaab Pupaa,b Totalab
Cohorts n Mean �SEM n Mean � SEM n Mean SEM n Mean � SEM n Mean � SEM
1 28 3.69 0.07b 26 8.74 � 0.17 b 33 0.83 0.05 a 8 74.86�31.28a 8 87.74� 31.17 a 2 29 3.35 0.06c 8 11.04 � 0.43 a 7 0.60 0.06a 3 186.60�45.44a 5 220.21� 26.91 a 3 14 3.95 0.07 a 15 13.10 � 0.74 a 13 0.71 0.06 a 2 210.27 � 10.27 a 2 227.33� 10.46 a 4 ne ne ne ne 2 198.75 � 6.00 a aMean days to complete development + Standard error. bDifferent letters among cohorts indicate statistical difference; Duncan's test when three

or more means.
ne = not evaluated.





50


The mean total developmental values per cohort (from Tables 8 and 9) are

grouped in Table 10, where three developmental time groups are proposed, naming them as short, intermediate and long developmental times. Cohort 1, the summer wasps, produced all of the short total development, 56.59 d (all emerged by 18-24 Oct.), and one wasp that shifted to the long development, 305.8 d (emerged by 27 May). In Cohort 2, one wasp completed development in an intermediate time, 112.91 d (emerged by 30 Dec.), and 4 completed development in the long time, 247 d (all emerged by 8-27 May). Wasps from cohort 3 produced 2 individuals that developed in the long time, 227.3 d (emerged by 7 and 27 May). Cohort 4 produced 2 wasps that developed in the long time, 198.5d (emerged 9 and 21 May). The duration of the total development is presented with 99% confidence limits, with the idea to stress that the proposed groups do not overlap each other. Those limits demonstrate the pertinence of grouping them in those three categories that proves they are statistically different (Table 10). Table 10. Trimodal total developmental time of Larra bicolor per cohort underground in
the field in Gainesville, FL (29040'N).

Shorta'b Intermediatea Longa
Cohorts n Mean t SEM n Mean � SEM n Mean � SEM
1 7 56.59 � 1.43 1 305.79 � 0.00 2 1 112.91 � 0.00 4 247.03 � 2.72 3 2 227.33 �10.46 4 2 198.75 � 6.00
aMean days to complete development + SEM. b99% CL = 4.32 d (52.27-60.91).

Table 11 presents the individuals grouped by their corresponding life cycle stages and total developmental times, noting statistical differences between larval, pupal and total development, among the three life cycle developments.





51


Table 11. Trimodal life cycle developmental time of Larra bicolor underground in field
conditions of Gainesville, FL (29040'N).

Shorfbc Intermediatbc Loncbtc
Stage n Mean � SEM n Mean � SEM n Mean � SEIN
Egg 7 3.49 � 0.39 a 1 3.12 � 0.00 a 7 3.49� 0.32 Larva 7 8.86 � 0.84 a 1 13.33 � 0.00 b 5 11.66 � 3.19 al Prepupa 7 0.83 � 0.31 a 1 0.45 � 0.00 a 5 0.64� 0.25 Pupa 7 43.59 � 3.32 a 1 96.00 � 0.00 b 5 235.61� 43.78 Total 7 56.59 � 3.49 a 1 112.91 � 0.00 b 9 238.45� 25.21
aMean days to complete development. bDifferent letters between cycles indicate statistical difference. Duncan's test when three or more means
cOnly wasps that completed the life cycle were included.

Survivorship

The number of wasp parasitic individuals that survived through the life cycle

stages and their corresponding adult sexes per cohort are presented in Table 12. An

overall 58% of the wasps reached the prepupal stage, which is an indicator that the

developing larva killed the host mole cricket. An overall 47 to 49% of the wasps reached

the pupal stage. An overall 13 to 14% of wasps completed development to the adult

stage, with an overall sex ratio not statistically different from 50% (X2 test).


Table 12. Number of individuals, corresponding proportion and sex ratio per cohort and
overall of Larra bicolor developing on Scapteriscus borellii underground at field
varying-temperatures.

Cohort Eggs Larvae' Prepupael Pupae' Adults' Sexes
1 65 45 (69.2) 40 (61.5) 33 (50.8) 10(15.4) 5M 3F 2 38 34 (89.5) 17 (44.7) 16 (42.1) 5 (13.2) 3M 2F 3 28 19 (67.9) 19 (67.9) 15 (53.6) 2 (7.1) 2F 4 20 ne ne 7 (35.0) 2 (10.0) 2F Overall2 131 98 (74.8) 76 (58.0) 64 (48.8) 17 (13.0) 8M 7F Overall3 151 71 (47.0) 19 (14.5) 8M 9F 'Total number of individuals (percentage inside of parentheses). 2Considering cohorts 1 - 3 only.
3Considering all 4 cohorts.
ne = not evaluated.






52


The proportion of individuals entering the next stage is presented in Figure 6, where the declining populations followed the type III curve explained by Southwood (1978), where mortality occurs gradually by stage, being similar among the three cohorts.


100 80



40
Cohort 2 Cohort 1
20

0
Eggs Larvae Prepupae Pupae Adults Life cycle stages

Figure 6. Proportion of individuals from three cohorts entering the life cycle stages of
Larra bicolor developing on Scapteriscus borellii underground at field-varying
temperatures.



The overall observed sex ratio in this work (which coincides with the observed ratio from constant temperature experiments) is somewhat intriguing, because during the three wasp seasons I scouted for wasps at patches of the nectar-plant S. verticillata (which included various times of the day and various periods of the year) the sex ratio of feeding wasps was always male-biased, most commonly seen as 3 to 10 males per one female, but sometimes reaching the exorbitant ratio of 100:1 M:F, or higher. Discussion

The development of egg, larval, prepupal and pupal stages and thus the total development of the wasp Larra bicolor were evaluated under field conditions underground at three or four depths.





53


The mean developmental times of each cohort at the various depths showed few differences, and in most of the cases there was no difference (Table 8). A comparison of the development of the cohorts showed statistical differences in egg and larval development among cohorts. No differences in prepupal and pupal development were seen (Table 9). Development from egg to prepupa did not vary according to depth in the soil (Table 8).

The pupal stage was the most sensitive to seasonal changes (shown by the four

cohorts that developed at various dates), showing three distinct durations of development time (Tables 10 and 12). The pupal (thus the total) development was classified as short, intermediate and long.

The capability of the pupal stage to diapause confers on this parasitoid the

opportunity to survive the winter when low temperatures may occur and when the food plant, Spermacoce verticillata, may freeze to the ground. Most of the parasitoids that showed long life cycle emerged early and middle May, coinciding with two main factors that favor the parasitoid's survival: a) the food plant starts blooming in early May. This information is based on personal observations made during 1998 to 2001 b) the mole cricket populations are composed of old adults and small and small-medium size mole cricket nymphs. Adult and medium-large nymph host sizes are attacked by this wasp parasitoid during late May, as shown in Appendix B.

The life cycle duration of Larra bicolor was observed to include three distinct development times, where the pupal stage accounted for the variation observed in life cycles as follow: a short duration of 57 d (646 DD), an intermediate duration of 113 d (838 DD), and a long duration of 238 d (1093 DD), (see DD calculation on page 32, and





54


sample size on Table 10). The short cycle occurred in August. A possible seasonal strategy is proposed here to account for ability of the wasp to survive under the environmental condition of Gainesville, FL (29040'):

Overwintered wasps. These wasps spend the winter as pupae from which adults begin to emerge in May. In May, these newly-emerged adults will find the preferred nectar- source plant (Spermacoce verticillata) blooming (in the localities where it occurs), providing energy to the wasps for mating, and for the females to hunt for host mole crickets. In May, these females will encounter overwintered Scapteriscus vicinus and S. borellii mole crickets that will be in the adult stage, providing suitable hosts, although by May they will not be especially abundant because they are undergoing attrition by aging.

First new generation. If the progeny of the overwintered generation of wasps (1. above) complete their life cycle:

i. In a short time (57 d), the adults will begin to emerge in July, a time when very few hosts of suitable size are available. This would be a poor strategy.

ii. In an intermediate time (113 d), they will begin to emerge in September, a time when the new generation of mole crickets exists as medium-sized to large-sized nymphs and is abundant (contrasted with numbers of survivors in May). Undersized hosts are not ideal because they may provide inadequate nutrition for wasp larvae. The longer a female wasp can survive, the greater is her opportunity for encountering large hosts. This appears to be the best strategy.

iii. In a long time (238 d), they will begin to emerge in December, which risks encountering freezing temperatures.





55


The optimal strategy for the progeny of the overwintered wasps (first new generation) is a development time of intermediate length (113 d).

Second new generation. For those wasps (first new generation) whose adults emerged in September, the strategies are to develop:

i. In a short time (57d), to provide a second new generation with adults emerging

in November. This strategy would be highly successful except in years when an early

freeze occurs.

ii. In an intermediate time (113 d), to give rise to adults emerging in JanuaryFebruary. This would be a disastrous strategy, because emerging adults would have

little or no access to nectar.

iii. In a long time (238 d), to give rise to adults emerging in May of the following

year. This would be successful.

That three distinct developmental times of L. bicolor should have been detected is remarkable. All three times (short, intermediate, and long) are needed to optimize its adaptation to the climate of northern Florida. The ability to adapt to the climate of northern Florida must have arisen during the evolution of this species in South America, where the species is widespread. The ability of the wasp to switch between these strategies is a tribute to its phenotypic plasticity. A simpler strategy would have been one with a single annual generation matched to the maximal annual abundance of hosts of suitable size: in other words, an annual cycle with presence of adult wasps only in September-November would have been easier.

Two types of questions are now presented. The basic question is: How is the

plasticity of developmental time triggered and maintained? The applied question is: How





56


far north in the USA can L. bicolor establish a population in the presence of Scapteriscus mole crickets, and will it adopt a life cycle of a single annual generation at its northern limits (as contrasted with the three generations it supports in northern Florida)? It seems obvious that a single annual generation of this wasp would have less effect on mole cricket populations than do the three generations that seem evident in northern Florida.

In the tropics, in areas where Scapteriscus mole crickets breed continuously, L. bicolor might likewise breed continuously, so that the short (57 d) life cycle might maximize its reproductive strategy. In more temperate areas, other kinds of life cycles, adapted to breeding periods of Scapteriscus mole crickets and to availability of nectar, are surely more appropriate. At the extreme northern (and southern!) limit of the range of this wasp, wherever that may be, a single annual generation may be most appropriate to maximize its population size. In the intermediate (subtropical and almost-temperate) areas, other strategies may be better to maximize population size. This wasp, L. bicolor, is exceptional in having the ability to adapt its development time to maximize its population size.

This is the first work describing the life cycle ofLarra bicolor under field conditions. The results present an insight into the performance of this parasitoid in adapting its populations to survive the climate of north Florida (29040'N). The data presented support the hypothesized cold hardiness of this strain and reaffirm the potential of this parasitoid to establish and disperse its populations at this latitude and perhaps beyond toward other states where Scapteriscus mole crickets are already established.





57


Degree-day model predicting the development of the wasp Larra bicolor Introduction

The rate of insect development depends upon the temperature to which the insects are exposed. For a developing insect, the rate at which time passes depends on the combination of its temperature-dependent development rate curve and the temperature at which it lives. The development rate curve may be altered by many factors of the environment, such as humidity, nutrition, or disease, but the curve always remains the same (Taylor 1981, Vinson and Iwantsch 1980). For each species, the temperature below that at which no measurable development occurs is the threshold of development.

The degree-day is a model that has been in use for more than 250 years (Wagner et al. 1984b) describing temperature/development relationship in the species (Honek 1999). The degree-day model is a linear approximation that enables calculating two virtual constants, lower threshold and sum of effective temperatures (an amount of heat needed for completing a development stage). This model fit the development rates within the range of ecologically relevant temperatures (Honek 1999, Campbell et al. 1974). The degree-day model does not work for all species. Hagstrum and Milliken (1991), Tanigoshi et al. (1976) and Logan et al. (1976) indicate this model is particularly invalid for mite populations. Wagner et al. (1984, and 1985) reviewed the efficiency of using constant temperatures to model development at variable temperatures. Hagstrum and Hagstrum (1970), Manel and Debouzie (1997b), showed how variable the results can be. Liu and Meng (1999) indicates that the degree-day model can simulate the development ofMyzuspersicae (Sulzer) very well when temperature does not go below 4oC. The degree-day model also works for Cydia pomonella (L.) (Howell and Neven 2000).





58


Wagner et al. (1984b) report that insects reared under identical conditions develop at different rates due to inherent differences among individuals in a population. Hagstrum and Hagstrum (1970) demonstrate that numerous insects develop at much lower temperatures when fluctuating diurnally than at a comparable mean constant temperature.

Understanding the relationship between temperature and development is

important in constructing computer models for simulating biological control of pest populations (Lysyk 2000).

Scapteriscus mole crickets are pests of pasture- and turf-grasses and vegetables in the southeast and are dispersing westward in the USA. They are soil-inhabiting insects, reaching the ground surface for dispersal and mating purposes. Their life cycles are described in Chapter 2. The parasitoid wasp Larra bicolor is a natural enemy of Scapteriscus mole crickets, and has been released in the USA.

The objective of this work was to evaluate the degree-day model based upon

constant temperatures to predict the field development ofLarra bicolor parasitizing mole crickets in north Florida (29040'N).

Materials and Methods

A set of experiments quantifying developmental time of Larra bicolor per stage and total was performed at seven constant temperatures in incubation chambers. The developmental time ofL. bicolor at field-varying temperatures was determined by studying the development of three wasp cohorts underground during summer, autumn and winter seasons. The parasitic wasp developed on Scapteriscus borellii and temperatures were recorded every 30 min at foir soil depths.






59


Developmental rates of L. bicolor are calculated using the linear equation (Eq. 1) plotted against temperature. The temperatures at the most linear portion of the curve were selected to calculate their linear equations by simple linear regression analysis (Eq. 1).

Y = a + b(x): (Eq.1) Y = Developmental rate (1/days)
a = Intercept
b = Slope
x = Temperature 'C


The lower temperature threshold and thermal constant (degree-days) were calculated from the linear equation, using its regression coefficients where: lower threshold (to) = -(a/b) and thermal constant (k) = 1/b.

The degree-days required per individual and total to complete development at constant and field temperatures are calculated by the following equation:


Y = T (t- t) (Eq. 2)
Y = Degree-days
T = Time (time step = 30 minutes = 1/48 days)
t = Temperature recorded every 30 min
to = Lower temperature threshold (per corresponding stage or total)
n = Number of days for stage to develop

The validation of development of L. bicolor at constant temperatures as a

predictor of development at field temperatures is evaluated by comparing degree-day data sets in the same graph to observe how well the degree-days at constant temperatures match degree-days at field varying temperatures. The statistical comparison of what can be seen from the constant and field temperature linear equations was done by comparison of slopes. The linear equations were calculated by regressing the observed DD with corresponding mean temperatures. Degree-days for development of individuals from the





60


constant temperature experiment used the temperature at which the incubation chamber was set. Degree-days for development of individuals from the field experiments were regressed against the mean temperature recorded during the period of time the development was achieved. All the individuals completing the respective stage were included. The null hypothesis was slopes equal to zero. Results

The wasp Larra bicolor developing in incubation chambers completed its life cycle development at temperatures ranging from 22.5 to 32.50C (only one wasp completed development at 22.50C and was not included in the analysis), although a few eggs developed at 200C and some eggs and larvae developed at 22.50C.

Developmental times were transformed to rates then plotted against their

respective temperature (Figure 2) showing the most linear development from 22.5 to 32.50C for the egg stage, and from 22.5 to 30.00C for all the other stages and total. The linear equation that best estimates development at constant temperatures, lower thresholds and degree-days is shown in Table 13.

The accumulated degree-days required to develop by stage at constant temperatures and at field-varying temperatures are shown in Table 14.

Degree-days required for pupal development of Larra bicolor underground at

field-varying temperatures are presented in Figure 7. The differences in development are clear among the three groups. The pupae showing the short development, considered as the non-diapausing population, were the only ones included in this validation; the other two groups developed at temperatures outside the range studied.





61


Table 13. Parameter values for the linear model, lower thresholds and thermal constants
(degree-days) for development of Larra bicolor per stage and total at constant
temperatures.

Linear equationa Lower threshold oC Thermal constant Stage Y = a + b (x)(Eq. 1) to = -(a / b)b kc = 1 /b Egg Y = -0.401469 0.022128(x) 18.143 45.44�1.45 Larva Y = -0.221118 + 0.011875(x) 18.620 84.42�1.92 Prepupa Y = -3.787811 + 0.203193(x) 18.641 5.05�0.34 Pupa Y = -0.038899 + 0.002429(x) 16.010 413.13�11.59 Total Y = -0.031264 + 0.001852(x) 16.770 541.67�12.41 Y = Developmental rate (Y = 1 / days), a = intercept, b = slope, x = temperature to = lower temperature threshold
k = Degree-days�SEM (days)



Table 14 Degree-days (days) required for Larra bicolor stages and total to develop at
constant and field temperatures.

Temp.oCa N Eggc N Larvac N Prepupac N Pupac N Totalc
25.0 26 43.61 1.62 16 83.17 �3.27 16 4.9 � 0.58 6 400.74 21.17 6 525.8 � 27.36
27.5 23 49.3.7 2.46 12 87.19 8.05 12 5.2�1.59 6 442.13 23.59 6 575.3 � 26.27
30.0 27 44.34 1.82 17 83.39�3.03 17 4.9�1.12 8 401.19 22.78 8 527.9 � 19.72
32.5 25 44.87 1.38
FIELDb 41 40.4 1.17 47 89.7 �2.53 53 6.6 � 0.85 8 481.6 14.39 8 595.1 �16.38 aTemperatures in incubation chambers.
bOverall Mean � SEM of all wasps developing during the summer in underground chambers in the field.
Mean � SEM.

Data for all the individuals sorted by stage and total, whose complete

development was recorded from field and from incubation chambers, are depicted in Figure 8. At most of the degree-days calculated, all the stages fell inside the observed range at constant temperatures. Slopes for constant and field temperature tend to be fairly flat, suggesting the degree-days required by stage are constant along the temperature range where development occurred.






62





1400

1200 0
1000 Long development
1000 i 0
a 5 Intermediate development S800
* U Short development S600

400

200


17.5 20.0 22.5 25.0 27.5 30.0 Temperature OC



Figure 7. Degree-days required for pupal development of Larra bicolor underground at
field-varying temperatures. Pupal stage grouped by kind of developmental times
(short, intermediate, and long).






63



120 i A
LARVA X X
100 DD=84.4 x x x


80

"P FIELD
* 60 CONSTANT 60
I M EGG
DD=45.4
40 1 - -

FIELD

20, PREPUPA
DD=5.1 + + +20.0 22.5 25.0 27.5 30.0 32.5


650 B
TOTAL
600 DD=541.7
-- -

CONSTANT
S500FIELD

1450 PUPA "8 DD=413.1 400 350

300 T
22.5 25.0 27.5 30.0 32.5 Temperature OC

Figure 8. Comparison of degree-days required for Larra bicolor life stages and total to
develop at constant and field varying temperatures. A) Egg, larval and prepupal
development. B) Pupal and total development (short development only).


The slopes per stage and total were evaluated showing that none of the regression

slopes were significantly different between field and constant temperature experiments

(Table 15). These results support the hypothesis that data from constant temperatures can





64


estimate development in field-varying temperatures for Larra bicolor over the range of temperatures in these studies.




Table 15. Statistical significance of slopes in regression analysis of development of Larra
bicolor stages and total, from the field and from incubation chambers.
Development Egg Larva Prepupa Pupa Total
Fielda
pC 0.625 ns 0.283 ns 0.819 ns 0.829 ns 0.741 ns Chambers
pC 0.126 ns 0.286 ns 0.295 ns 0.648 ns 0.574 ns
a. Development in field-varying temperatures b. Development in incubation chambers at constant temperatures C statistic of t-test
ns = statistically not significant

Discussion

Development under a regime of fluctuating temperatures often differs from that under constant temperatures. Higley et al. (1986) mention that the magnitude of the differences between developmental times under constant and fluctuating temperatures depends on the average temperature, amplitude and frequency of the fluctuations.

Larra bicolor development estimated at constant temperatures and evaluated as degree-days was a good estimator of development at field-varying temperatures. The coincidence of the constant temperature degree-day trend with observed development in the field (Figure 29) is the first clue to the efficiency of constant temperature development to predict field development. The lack of significant differences of slopes in regression analysis of all stages under the two conditions studied proves that Larra bicolor possesses a thermal constant that applies at the temperature range used in this study (Table 25). This thermal constant applies to the pupal stage when this develops in a short time, which mostly occurs during summer and early autumn.





65


One of the uses of the results presented in this paper is the possibility to predict the period of time when this parasitoid would be present, thus parasitizing host mole crickets in north Florida. The importance of having this knowledge will allow biologists and extension service personnel to predict when the parasitic wasp would emerge from the soil in spring. By this means the phenological model proposed in this paper can be integrated with the pest management of mole crickets. A direct benefit of this phenological model would be the timing of pesticide applications according to the predicted wasp season.

The use of the model is restricted to the soil temperature range from 25 to 32.50C, which occurs from May to November or early December, which covers the normal wasp season. The soil temperature source can be the Florida Automated Weather Network, whose records proved highly coincident with soil temperature in the experimental area (at 22.5 cm depth, see appendix C) where development of the wasp was studied.

The development of Larra bicolor has been modeled based on constant

temperature studies using the degree-day model approach. The model is a good predictor of development in north Florida soils.

The degree-day model to validate development of Larra bicolor at constant

temperatures proved valuable to predict field-varying temperature development of nondiapausing populations. Extensive work would be required to model diapausing populations.













CHAPTER 4
DEVELOPMENT OF THE FLY ORMIA DEPLETA.


Development of Ormia depleta at constant temperatures Introduction

The tachinid fly Ormia depleta was brought to Florida in 1987 as part of the mole cricket research program and considered potentially the most effective control agent against mole crickets (Fowler 1987, 1989, Frank et al. 1996). This endoparasitic fly, of crepuscular habits, locates host mole crickets by the male host's calling song and deposits larvae (planidia) on or near the host (Fowler 1987, Walker et al. 1996).

The first parasitoid strain ("old" strain, from 23oS in Brazil) released in Florida in 1988 (Frank et al. 1996), was considered as largely responsible for a decline in mole cricket populations (Parkman et al. 1996). By 1994, this "old" strain disappeared from the Gainesville area (290 40'N latitude), perhaps due to its poor tolerance to cold and long winters (Walker et al. 1996). In recent years, in late autumn, a few flies have been captured near Gainesville, but these were considered to be individuals dispersing from south toward north without re-establishing populations.

In 1999 a second tachinid strain ("new" strain) was imported. It was hoped that this "new" strain, from 30oS in Brazil, might be more cold tolerant and could establish populations north of 29oN in Florida and other states (H. Frank personal communication).

Studies performed with the first strain of Ormia depleta focused on establishment and range (Frank et al. 1996), seasonality (Walker et al. 1996), effectiveness in



66





67


controlling mole crickets (Parkman et al. 1996), techniques for rearing it in the laboratory (Wineriter and Walker 1990) and food source (Welch 2000). This "old" strain was in culture from 1987 (Frank et al. 1996) until 1999, but it is the "new" strain that is now in culture to supply populations. Wineriter and Walker (1990) determined the life history of the first strain under laboratory conditions at 23-25oC.

There is no information about the effect of temperature on life history or seasonal phenology. Limits of geographical expansion of 0. depleta are poorly understood. The fact that the "old" introduced strain remained well established in south-central Florida (270N), but disappeared farther north suggests a strong influence of temperature regimes, where either colder winters kill the flies or cause the lack of adequate winter nectar sources (Walker et al. 1996). These arguments led to the introduction of a second strain with the hope that it would be more cold hardy and survive colder winters.

In recent years the efficiency of 0. depleta has been questioned, Potter (1998) stated that this parasitoid does not seen to have a major effect on mole cricket populations, but controversy about the parasitoid role could be due to insufficient information on its biology (Ryoo et al. 1991).

Developmental times of many species of poikilotherms as a function of

temperature have been described by regression equations (Wagner et al. 1984a). Such regression equations can provide a sample of calculating developmental times in insect population models or can be used for quantitative comparisons of the effect of temperature on the developmental times of different stages and species of insects (Hagstrum and Milliken 1988).





68


The work described here was aimed to determine the effect of temperatures on development and survivorship of the "new" strain of Ormia depleta, then compare the developmental times and developmental rates of the two strains. Materials and Methods

The flies used in this research were laboratory-reared flies using the methods described by Wineriter and Walker (1990). The old strain flies were under culture until 1999 when the new strain was brought from South America.

Gravid fly females, recognized by a dark spot seen near the apex of the abdomen due to the presence of fully grown planidia, were dissected to expose planidia on wet paper in a petri dish. Only dark colored (mature first instar) planidia were used to parasitize host mole crickets. Three planidia were placed in the membranous tissue behind and underneath the posterior of a host pronotum, making sure those planidia remained there until the host was placed into its plastic cup (Wineriter and Walker 1990).

Parasitized hosts S. borellii were confined in a 200-ml plastic cup filled with wet sterilized sand. The cups were then transferred to culture chambers at eleven constant temperatures, 10.0, 15.0, 17.0 17.5. 20.0, 22.5, 25.0, 27.5, 30.0, 32.5, 35.0 � 0.20C.

Relative humidity was set at 80%, and the photoperiod at 14: 10 (L: D). 15 newly parasitized hosts (three planidias per host) were placed per temperature-chamber.

Parasitized hosts were fed twice a week following the standard procedure for S. abbreviatus culture. Development was tracked by observing the host through the transparent cup walls every six hours, until parasitic exited the host (by making a hole from inside out of the host's abdomen, killing the host) to pupate in the sand. The pupal development was tracked at the same frequency until adult parasitoid emergence.





69


The response variables recorded per strain were larval and pupal development at each constant temperature. Survivorship and adult sex ratio were calculated at each temperature. Data were sorted by sex and pooled sexes, and comparisons were made between sexes. Developmental times were compared by "t-test" between temperatures. Developmental rates were compared using linear regression models. Degree-days and lower temperature thresholds were calculated from the linear equations. Finally these variables were tested between strains.

Results

The new strain

The host, S. borellii, did not survive more then seven days at temperatures of

10.0, 15.0 and 35.0oC. The parasitoid was able to develop at temperatures ranging from 17.0 to 30.00C, where developmental time was inversely related to temperature, as is common in poikilothermic organisms. Developmental time for larvae ranged from 23.6 d to 6.9 d (pooled sexes), 21.7 to 6.8 (males) and 26.5 to 7.9 d (females). Development time for pupae ranged from 30.4 d to 9.3 d (pooled sexes), 31.0 d to 9.0 d (males) and 29.5 d to 9.35 d (females). Total development time ranged from 54 d to 16.5 d (pooled sexes), 52.7 d to 15.9 d (males), and 56 d to 16.5 d (females). (Tables 16,17 and 18 and Figure 9).

Mean development times were compared per sex by the 't-test' assuming equal variances (Table 19). The null hypothesis was set as males develop in less time than females at either stage or total development. No difference was observed at larval stage. Statistical differences were observed in the pupal stage where males developed in less time at 30.0, 25.0, 22.5, 20.5, and 17.5, thus reflecting the same differences in the total






70


development. Although statistically not different, males also developed faster than

females at 27.50C.





Table 16. Development of Ormia depleta (new strain) pooled sexes life stages.

Temp. Days required to complete developmenta oC n Larva Pupa Total
30.0 37 6.91 � 0.11 (41.9) 9.28 � 0.08 (58.1) 16.48 � 0.13 27.5 27 7.02 � 0.09 (41.5) 9.90 � 0.07 (58.5) 16.92 � 0.08 25.0 47 8.61 � 0.13 (43.4) 11.22 � 0.20 ( 56.6) 19.82 � 0.18 22.5 24 10.22 � 0.19 (41.5) 14.43 � 0.11 (58.5) 24.65 � 0.27 22.0 14 12.43 � 0.47 (44.7) 15.36 � 0.13 (55.3) 27.79 � 0.58 20.0 53 14.25 � 0.19 (41.5) 20.11 � 0.16 (58.5) 34.35 � 0.20 17.5 17 20.98 � 0.38 (41.6) 29.45 � 0.18 (58.4) 50.43 � 0.49 17.0 5 23.60 � 1.25 (43.7) 30.40 � 0.40 (56.3) 54.00 � 0.89 a Mean � SEM; porcentage of developmental time required for each life stage given in parenthesis.






Table 17. Development of Ormia depleta (new strain) male stages.

Temp. Days required to complete developmenta �C n Larva Pupa Total
30.0 18 6.82 � 0.14(43.0) 9.06 � 0.11 (57.0) 15.88 � 0.16 27.5 14 7.10 � 0.13.(42.2) 9.86 � 0.10 (57.8) 16.96 � 0.13 25.0 23 8.64 � 0.20 (41.7) 10.87 � 0.27 (58.3) 19.51 � 0.30 22.5 14 9.90 + 0.19(40.6) 14.14 � 0.10 (59.4) 24.04 � 0.27 20.0 27 14.15 � 0.23 (40.2) 19.64 � 0.21 (59.8) 33.77 � 0.26 17.5 5 20.78 � 0.53 (39.4) 28.68 � 0.28 (60.6) 49.46 � 0.72 17.0 3 21.67 � 0.67(38.8) 31.00 � 0.00 (61.2) 52.67 � 0.67 aMean � SEM; percentage of developmental time required for each life stage given in parenthesis.






71


Table 18. Development of Ormia depleta (new strain) female stages.

Temp. Days required to complete developmenta
0C n Larva Pupa Total
30.0 19 6.99 0.16(43.3) 9.49 � 0.11 (56.7) 16.47� 0.19 27.5 13 6.93� 0.12(41.3) 9.95 � 0.10(58.7) 16.88� 0.11 25.0 24 8.58 � 0.19(40.7) 11.56 � 0.27 (59.3) 20.13 � 0.20 22.5 10 10.66 � 0.33 (40.9) 14.85 � 0.14 (59.1) 25.51 � 0.41 20.0 23 14.36 � 0.36 (40.4) 20.47 � 0.22 (59.6) 34.83 � 0.30 17.5 12 21.06 � 0.51 (38.8) 29.77 � 0.14 (61.2) 50.83 � 0.61 17.0 2 26.5 � 0.5 (38.8) 29.50 � 0.5 (61.2) 56.00 � 0.00
a Mean � SEM; percentage of developmental time required for each life stage given in parenthesis.



60
A
50


40


30

20 TOTAL 10 PUPA LARVA
0 ,
15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 Temperature OC

Figure 9. Development time of Ormia depleta "new" strain, pooled sexes, at constant
temperatures.



Survivorship recorded from each temperature is presented in Figure 10, where maximum survival occurred at 27.50C, followed by 22.5, 20.0, and 25.0C in a group with survivorship greater than 60% and greater than 50 % at 17.5oC. The greater mortality occurred in the larval stage.






72


Table 19. T-test comparison of developmental time (days) of Ormia depleta (new strain)
between sexes per temperature per stage and total life cycle.

Larva Pupa Total Temp. oC Male Female Male Female Male Female 30.0 6.82 a 6.99 a 9.06 a 9.49 b 15.88 a 16.47 b 27.5 7.10 a 6.63 a 9.86 a 9.95 a 16.96 a 16.88 a 25.0 8.64 a 8.58 a 10.87 a 11.56 b 19.51 a 20.13 b 22.5 9.90 a 10.66 a 14.14 a 14.85 b 24.04a 25.51 b 20.0 14.15 a 14.36 a 19.64 a 20.47 b 33.77 a 34.83 b 17.5 20.78 a 21.06 a 28.68 a 29.77 b 49.46 a 50.83 b 17.0 21.67 a 26.50 a 31.00 a 29.50 b 52.67 a 56.00 b
a Means followed by the same letter are not statistically different (one tailed test)



100
90 TEMPERATURE OC 80 -e-15.0 70 ----17.5
60 -- 20.0
- 50 --x 22.5 g 40 ---- 25.0
30 - 27.5 20 --+-- 30.0
- 32.5
10


0 50 100 150 200 250 300 350 Degree-days > 11.2 OC


LARVA PUPA
Figure 10. Survivorship of Ormia depleta "new" strain, pooled sexes, at constant
temperatures.




Regression parameters corresponding to each data set are shown in Tables 20,21

and 22 and in Figure 11. The coefficients of determination ranged from 0.90 to 0.95

representing the good fit of all models. The lower threshold values ranged from 11.1 to

11.6. The accumulated degree-days required to complete development per stage and total

are shown in last column of Tables 20, 21 and 22.






73


Table 20. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (new
strain) pooled sexes life stages to develop.

Degree-days
Stage y Intercept � SEM Slope + SEM R2 � SE t (OC) (Days � SE) Larva -0.092 � 0.0044 0.0082 + 0.0002 0.90 � 0.0109 11.2 123.4 � 0.9 Pupa -0.071 � 0.003 0.0062 + 0.0001 0.92 � 0.0076 11.5 163.6 � 0.9 Total -0.04 � 0.0013 0.0035 � 0.00006 0.95 � 0.0033 11.4 286.6 � 1.3





Table 21. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (new
strain) male stages to develop.
Degree-days
Stage y Intercept � SEM Slope + SEM R2 + SE t (OC) (Days � SE)
Larva -0.090 � 0.0061 0.0082 � 0.0003 0.90 + 0.0106 11.1 124.0 � 1.2 Pupa -0.074 � 0.0044 0.0063 � 0.0002 0.91 � 0.0077 11.6 159.5 � 1.2 Total -0.041� 0.002 0.0036 � 0.00008 0.95 � 0.0034 11.4 283.2 � 1.8





Table 22. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (new
strain) female stages to develop.
Degree-days
Stage y Intercept � SEM Slope + SEM R2 � SE t (OC) (Days � SE)
Larva -0.094 � 0.0063 0.0083 � 0.0003 0.90 � 0.0113 11.4 122.9 � 1.4 Pupa -0.069 � 0.004 0.0061 � 0.0002 0.92 � 0.0073 11.5 167.4 � 1.3 Total -0.040 � 0.0018 0.0035 � 0.00007 0.95 � 0.0032 11.5 289.6 � 1.8






74


0.18

0.16 LARVA

0.14

S0.12
1 PUPA
0.10

0.08

Dfl 0.06 TOTAL

0.04

0.02 0.00
15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 Temperature OC Figure 11. Developmental rates of Ormia depleta "new strain", pooled sexes, at constant
temperatures.



The old strain

This old strain developed at temperatures ranging from 17.5 to 30.00C, showing an inverse relationship between developmental time and temperature. Development time for larvae ranged from 20.39 d to 6.65 d (pooled sexes), 20.52 d to 6.40 d (males) and 20.13 d to 6.78 d (females). Development time for pupae ranged from 32.13 d to 8.72 d (pooled sexes), 31.58 d to 8.48 d (males) and 33.23 d to 8.84 d (females). Total development time ranged from 52.52 to 15.38, 52.10 to 14.88, and 53.35 to 15.63 d (females). (Tables 23, 24 and 25 and Figure 12).






75


Table 23. Development of Ormia depleta (old strain) pooled sexes life stages.

Temp. Days required to complete developmenta OC n Larva Pupa Total
30.0 6 6.40 � 0.27 (43.0) 8.48 � 0.06 (57.0) 14.88 � 0.21 27.5 13 7.15 � 0.31 (42.2) 9.78 � 0.37 (57.8) 16.93 � 0.30 25.0 24 7.95� 0.16(41.7) 11.13� 0.11 (58.3) 19.08� 0.17 22.5 9 9.00 � 0.42(40.6) 13.17 � 0.30 (59.4) 22.17 � 0.26 20.0 15 13.83 � 0.48 (40.2) 20.56 � 0.41 (59.8) 34.39 � 0.29 17.5 6 20.52 � 0.40 (39.4) 31.58 � 0.96 (60.6) 52.10 � 1.01
a Mean � SEM; percentage of developmental time required for each life stage given in parenthesis.






Table 24. Development of Ormia depleta (old strain) male life stages.

Temp. Days required to complete development
OC n Larva Pupa Total
30.0 6 6.78 � 0.11 (43.3) 8.84 � 0.14(56.7) 15.63� 0.21 27.5 13 6.73 � 0.32 (41.3) 9.89 � 0.27 (58.7) 16.62 � 0.52 25.0 24 7.77 � 0.10 (40.7) 12.02 � 0.18 (59.3) 19.78 � 0.20 22.5 9 9.67 � 0.00 (40.9) 13.58 � 0.33 (59.1) 23.35 � 0.33 20.0 15 14.23 � 0.40 (40.4) 20.98 � 0.18 (59.6) 35.21 � 0.40 17.5 6 20.13 � 0.0 (38.8) 33.23 � 0.23 (61.2) 53.35 � 0.23 a Mean t SEM; percentage of developmental time required for each life stage given in parenthesis.




Table 25. Development of Ormia depleta (old strain) female life stages.

Temp. Days required to complete developmenta
OC n Larva Pupa Total
30.0 6 6.65 � 0.13(43.3) 8.72 � 0.12 (56.7) 15.38 � 0.21 27.5 13 6.93 � 0.22 (41.3) 9.84 � 0.21 (58.7) 16.76 � 0.30 25.0 24 7.88� 0.10(40.7) 11.50 � 0.13 (59.3) 19.38 � 0.15 22.5 9 9.22 � 0.29 (40.9) 13.31 � 0.23 (59.1) 22.53 � 0.27 20.0 15 14.10 � 0.30 (40.4) 20.84 � 0.18 (59.6) 34.94 � 0.30 17.5 6 20.39 � 0.26 (38.8) 32.13 � 0.70 (61.2) 52.52 � 0.70
a Mean � SEM; percentage of developmental time required for each life stage given in parenthesis.





76



60

50 A

40

S30

20 TOTAL 10 PUPA LARVA

15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 Temperature OC



Figure 12. Developmental time of Ormia depleta, old strain, life cycle at constant
temperature.



Mean developmental times were compared between sexes per temperature per stage by the 't-test' assuming equal variances (Table 26). The null hypothesis was that males develop in less time than females at either temperature per stage or total development. No difference was observed in the larval stage. In the pupal stage the only difference was observed at 25.0C. In the total life cycle differences were observed at 30.0, 25.0, and 22.5oC. At all cases where differences were observed, males developed in less time than females.

Survival recorded from each temperature is presented in Figure 13, where

maximum survival (87-88%) occurred at 22.5 and 25.0oC, followed by 27.5, 20.0, 30.0 and 17.5oC respectively, in a group with survivorship ranging from 50% to 82%. The greater mortality occurred in the larval stage at all temperatures but 22.50C.






77


Table 26. T-test comparison of developmental time (days) of Ormia depleta (old strain)
between sexes per temperature per stage and total life cycle.


Larvaa Pupaa Totala Temp. *C Male Female Male Female Male Female
30.0 6.82 a 6.99 a 9.06 a 9.49 b 15.88 a 16.47 b 27.5 7.10 a 6.63 a 9.86 a 9.95 a 16.96 a 16.88 a 25.0 8.64 a 8.58 a 10.87 a 11.56 b 19.51 a 20.13 b 22.5 9.90 a 10.66 a 14.14 a 14.85 b 24.04a 25.51 b 20.0 14.15 a 14.36 a 19.64 a 20.47 b 33.77 a 34.83 b 17.5 20.78 a 21.06 a 28.68 a 29.77 b 49.46 a 50.83 b 17.0 21.67 a 26.50 a 31.00 a 29.50 b 52.67 a 56.00 b
a Means followed by the same letter are not statistically different (one tailed test)




100
TEMPERATURE OC
90
-e--15.0
80- 17.5
70 ---20.0
> 60 - 22.5
50 e -- 25.0 S--- 27.5
S40
--+-- 30.0
30 - 32.5
20 10
0 O
0 50 100 150 200 250 300 Degree-days > 11.5 OC

LARVA PUPA
Figure 13. Survivorship of Ormia depleta old strain, pooled sexes, at constant
temperatures.



Regression parameters corresponding to each data set are shown in Tables 27, 28

and 29 and in Figure 14. The coefficients of determination ranged from 0.88 to 0.97

representing a good fit of all models. The lower threshold values ranged from 11.3 to





78


12.7. The accumulated degree-days required to complete development per stage and total

are shown in last column of Tables 27, 28 and 29.




Table 27. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (old
strain) life stages to develop.
Degree-days
Stage y Intercept � SEM Slope � SEM R2 � SE t ("C) (Days + SE) Larva -0.104 � 0.0097 0.009 + 0.0004 0.88 � 0.0121 11.5 112.4 + 1.4 Pupa -0.087 � 0.0048 0.0069 + 0.0002 0.95 � 0.0059 12.6 146.1 � 1.3 Total -0.048 � 0.0026 0.0039 � 0.0001 0.95 � 0.0033 12.2 258.2 � 2.1




Table 28. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (old
strain) male life stages to develop.
Degree-days
Stage y Intercept � SEM Slope � SEM R2 � SE t (oC) (Days � SE) Larva -0.101 � 0.0137 0.009 � 0.0006 0.88 � 0.0117 11.3 113.1 � 2.0 Pupa -0.09 � 0.0071 0.0071+ 0.0003 0.94 � 0.006 12.6 142.1 � 1.8 Total -0.048 � 0.0034 0.004 � 0.0001 0.96 � 0.0029 12.1 254.7 � 3.0





Table 29. Linear regression models (Y= developmental rate, x= temperature), lower
developmental thresholds (t), and degree-days required for Ormia depleta (old
strain) female life stages to develop.

Degree-days
Stage y Intercept � SEM Slope + SEM R2 � SE t (OC) (Days � SE)
Larva -0.106 � 0.014 0.0090 � 0.0006 0.88 � 0.0128 11.7 111.6 � 1.9 Pupa -0.086 �0.0054 0.0068 �0.0002 0.97 � 0.0049 12.7 148.3 � 1.4 Total -0.048 � 0.0034 0.0039 � 0.0002 0.95 � 0.0034 12.3 259.7 � 2.8






79



0.20
0.18 i
LARVA
0.16 0.14 0.12
MPUPA
0.10
S0.08
M t,
c* .TOTAL
0.06

0.04 0.02 0.00
15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 Temperature OC Figure 14. Developmental rates of Ormia depleta old strain, pooled sexes, life cycle at
constant temperatures.



Discussion

Ormia depleta development rates were sensitive to temperatures and showed a linear response at temperatures ranging from 17.5 to 30.0C (Figures 11 and 14). The developmental times reported by Wineriter and Walker (1990) coincided with those reported in this study, for both strains, at temperatures between 22.5 and 25.0oC for each sex and pooled. Larval development was shorter than pupal development at all temperatures, maintaining similar proportions among temperatures (Tables 16-18, 23-25). Wineriter and Walker (1990) observed that emergence pattern was significantly different for the two sexes, with male adults generally emerging first. The difference in development by sex was more evident in the new strain than the old strain as indicated in Tables 19 and 23, respectively.





80


Comparing the development of the two strains (Figures 15,16 and 17), male larval development times were statistically different only at 25.0C. For females and pooled sexes there were differences at 22.5 and 25.0oC. Pupal development for males, females and pooled sexes showed differences at 17.5, 22.5 and 30.0C. Male total development times were different at 25.0C, and females and pooled sexes showed differences at 17.5, 22.5 and 30.0C. The difference in developmental time between sexes was more evident in the new strain than the old, where the larval stage was less variable than pupal and total between the two strains.

The development time per strain at the extreme temperatures was statistically different. The old strain took more time to develop at 17.5C and the new strain took more time at 30.00C (more evident at females, Figure 18, and pooled sexes, Figure 16). In other words, for pupal and total development, the new strain developed faster than the old strain at the lower temperature, but it was the old strain that developed faster than the new strain at the higher temperature.


60
a t a Larva New 50 Larva Old a Pupa New
40 a a a Pupa Old ab
b30 a Total New
a b o Total Old
aa ab a
20 a ab
O b a a aa F a, a a b


17.5 20.0 22.5 25.0 27.5 30.0 Temperature OC
Figure 15. Developmental times of two strains (new and old) of Ormia depleta, pooled
sexes, at constant temperatures. Means followed by different letter are statistically
different ("t - test", P<0.05).






81



60
aa a Larva New
50 m Larva Old o Pupa New
40 a Pupa Old
> 30 -- Ib Total New a a o Total Old as aa ab
20 a aa
Eb a
oll a aalaa aa



17.5 20.0 22.5 25.0 27.5 30.0 Temperature OC


Figure 16. Developmental times of two strains (new and old) of Ormia depleta males at
constant temperatures. Means followed by different letter are statistically different
("t - test", P<0.05).




60
ab a Larva New 50 a Larva Old o Pupa New
40 m Pupa Old a a
ab Total New
b on Total Old
30
ia aa
aaa aa
a ab
f i aJ a a 10



17.5 20.0 22.5 25.0 27.5 30.0 Temperature OC


Figure 17. Developmental times of two strains (new and old) of Ormia depleta females at
constant temperatures. Means followed by different letter are statistically different
("t - test", P<0.05).




Temperatures at which this parasitoid may develop were identified (Figures 9 and

12). The rate of increase decreased at 300C and stopped at 32.50C. the upper temperature





82


threshold must fall between these temperatures. The survivorship observed for the two strains was statistically different only at 250C.

The developmental rates of both strains were compared by means of their

corresponding linear regression models shown in Figures 11 and 14. This comparison was performed to determine whether both lines can be considered to be estimates of a common B (slope) following the procedure of Steel and Torrie (1960: 173), Sokal and Rohlf (1995). The results established that the two strains develop differently at constant temperatures (Table 30). The new strain developed slower than old strain at higher temperatures, but faster than the old strain at low temperatures. The last statement is illustrated in Figure 18, where the estimated developments (by the linear equation) were prolonged until crossing the y axes, so that the lower temperature thresholds (where lines crossed the x axes) are presented per strain. No statistical differences were observed in survivorship of the two strains developing in constant temperatures (Figure 19).






83




0.08
New strain y = 0.0035x - 0.0402
0.06 7
Old strain y = 0.0039x - 0.0476
0.04 .

~ 0.02


0 2 4 6 14T 12 14 16 18 20 22 24 26 28 30 32
-0.02

-0.04 i

-0.06
Temperature (OC)



Figure 18. Linear development estimates by linear regression equations of the two strains
of Ormia depleta at constant temperatures. Lower development was prolonged
until crossing "y" axis.



The lower temperature thresholds for the two strains were compared following the procedures in Infante and Gil (1984) and Sokal and Rohlf (1995), showing the results in Table 30, where the thresholds of the two strains overlapped at 95% confidence making them statistically not different. Nevertheless, it was the new strain that showed the lower temperature at the lower confidence limits of the two strains (Table 20 and 27 and Figure 18).





84


Table 30. Comparison of slopes (bl)1 and intercepts (a)' between the "new" and "old"
strains of Ormia depleta by t-test and confidence limits at constant temperatures.
Slope2 CL1 CL2 Intercept CL1 (LTT) CL2 (LTT) New strain 0.00352 a 0.0034 0.0036 -0.0402 a -0.0380 (10.78) -0.0424 (12.04)
Old strain 0.00391 b 0.0037 0.0041 -0.0476 a -0.0432 (11.05) -0.0520 (13.28)
1 bl and a are the coefficients from the linear regression equations.
2 Means followed by defferent letters are statistically significant. CL1 = 95% upper confidence limits,
CL2 = 95% lower confidence limits.
LTT = lower temperature (OC) threshold


Male, female and total life cycles were compared between the new and old strains of Ormia depleta reared at constant temperatures. In all cases the comparisons showed statistical differences (p< 0.0001) (Figures 15,16 and 17). Based upon these results, it can be said that the two strains develop at different rates, which reinforces the previous results where developmental times were different at pupal and total life cycles.

These findings may support the hypothesis that the new strain is better adapted to develop at extreme temperatures than the old strain that failed to establish populations in north Florida. These results suggest the possible success of the new strain to establish populations north of 270 N latitude in Florida and neighboring states where the pest mole crickets are already established.

The previous information of life history of Ormia depleta has been refined to its relationship to constant temperatures, determining the ranges at which development occurs for the two strains released in Florida. Lower temperature thresholds per sex and pooled sexes were determined, along with a good approximation of what can be considered the upper temperature threshold for the two strains.




Full Text

PAGE 1

RELATIONSHIP BETWEEN TEMPERATURE AND DEVELOPMENT OF THE ECTOPARASITOID LARRA BICOLOR (HYMENOPTERA: SPHECIDAE) AND THE ENDOPARASITOID ORMIA DEPLETA (DIPTERA: TACHINIDAE) By HECTOR CABRERA-MIRELES 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 2002

PAGE 2

ACKNOWLEDGMENTS I would like to extend a special thanks to my committee members who have greatly helped me improve this dissertation. I probably would not find enough words to thank Dr. J.H. Frank (chair), without his support this dissertation would have been impossible. I would thank Dr. J.L Stimac for his cheerful encouragement, friendship and for serving as a committee member. I would like to thank Dr. J. Jones and Dr. T.J Walker who were kind enough to make suggestions on this dissertation and served as committee members. Sincere thanks are due to Dr. Robert Hemenway who was always available to provide me with help and supplies of all kinds and kindly friendship. I thank the Instituto Nacional de Investigaciones Forestales, Agricolas y Pecuarias (INIFAP) and the Consejo Nacional de Ciencia y Tecnologia (CONACYT) for having provided me with financial support during my doctoral study. ii

PAGE 3

TABLE OF CONTENTS Pag ACKNOWLEDGMENTS " LIST OF TABLES vii LIST OF FIGURES xi ABSTRACT xv 1. INTRODUCTION 1 2. REVIEW OF LITERATURE 6 Introduction 6 The Pest Mole Crickets, Scapteriscus spp 7 The Digger Wasp Larra bicolor 9 Natural History of Larra bicolor 9 Life Cycle of Larra bicolor in the Laboratory 10 Establishment of Larra bicolor in Florida 1 1 The Tachinid Fly Ormia depleta 1 1 Natural History of Ormia depleta 1 1 Life History of Ormia depleta in the Laboratory 12 Insect Modeling 13 Constant vs Fluctuating Temperature Influence on Insect Development 14 Constant Temperatures 15 Fluctuating Temperatures 17 Applicability of Constant and Fluctuating Temperatures in Insect Developmental Models 18 Temperature-Dependent Models for Studies on Insect Development 19 Linear Model 20 Non-linear Models 22 Stochastic Models 24 Mixed Models 26 Conclusion 27 3. DEVELOPMENT OF THE WASP LARRA BLCOLOR F 30 Development and reproduction of the wasp Larra bicolor at constant Temperatures 30 Introduction 30 iii

PAGE 4

Materials and Methods 1 Development 31 Life-Fertility Tables 32 Results 33 Development 33 Reproduction 37 Life-FertilityTables 38 Discussion 39 Development of the wasp Larra bicolor at fieldvarying temperatures 42 Introduction 42 Materials and Methods 43 Results 48 Survivorship 51 Discussion 52 Degree-day model predicting the development of the wasp Larra bicolor 57 Introduction 57 Materials and Methods 58 Results 60 Discussion 64 DEVELOPMENT OF THE FLY ORMIA DEPLETA 66 Development of Ormia depleta at constant temperatures 66 Introduction 66 Materials and Methods 68 Results 69 The new strain 69 The old strain 74 Discussion 79 Development of the two strains of Ormia depleta under field conditions 86 Introduction 86 Materials and Methods 86 The new strain 91 The old strain 91 Results 92 The new strain 92 The old strain 93 Development times 95 Discussion 97 Development times 97 Problems 98 Diapause 99 iv

PAGE 5

Degree-day model predicting the development of Ormia depleta 1 00 Introduction 100 Materials and Methods 101 Results 103 The new strain 103 The old Strain 107 Discussion 1 1 1 5. CONCLUSIONS 114 The Parasitic Wasp Larra bicolor 114 The Parasitic Fly Ormia depleta 1 1 6 APPENDICES A. DEVELOPMENT OF TWO MOLE CRICKETS, S. VICINUS AND S. ABBREVIA TUS, AT CONSTANT TEMPERATURES 1 1 9 Introduction 119 Materials and Methods 120 Oviposition, fecundity and fertility 121 Nymphal development 122 Results 122 Scapteriscus vicinus 122 Oviposition, fecundity and fertility 123 Development 126 Scapteriscus abbreviatus 128 Oviposition, fecundity and fertility 128 Development 131 Discussion 133 Scapteriscus vicinus 133 Scapteriscus abbreviatus 136 Both species 136 B. PARASITISM OF MOLE CRICKETS BY THE WASP LARRA BICOLOR 1 39 Introduction 139 Materials and methods 140 Results 143 Percent parasit ism 143 Parasitoid success 146 Sex dimorphism 146 Discussion 148

PAGE 6

C. SOIL TEMPERATURE ESTIMATE 151 Problem description 151 Materials and methods 153 Temperature estimates 153 Model 154 Description 154 Results 155 Temperature estimate of missed data 159 Discussion 165 LIST OF REFERENCES 167 BIOGRAPHICAL SKETCH 175 vi

PAGE 7

LIST OF TABLES Table Eage Table 1. Insect models cited in the Review of Literature chapter 28 Table 2. Development (days) of Larra bicolor life stages 3 at constant temperatures 34 Table 3. Linear regression models (y = developmental rate, x = temperature), lower developmental thresholds (t), and degree-days required for Larra bicolor life stages and total to develop at constant temperatures 36 Table 4. Population reproductive statistics for Larra bicolor from two seasons in Gainesville, FL (29°40'N) 38 Table 5. Reproductive parameters for Larra bicolor from two seasons in Gainesville, FL (29°40'N) 39 Table 6. Codes and depths of sub-chambers and thermocouples where parasitized mole crickets were confined to study the development of Larra bicolor underground in the field 45 Table 7. Description of the cohorts of the wasp-parasitized hosts by numbers and dates. ...47 Table 8. Development time of Larra bicolor at various depths underground in the field at various dates in Gainesville, FL (29°40'N) 49 Table 9. Developmental time of four cohorts of Larra bicolor stages and total, averaged over all depths underground in the field in Gainesville, FL (29°40'N) 49 Table 10. Trimodal total developmental time of Larra bicolor per cohort underground in the field in Gainesville, FL (29°40'N) 50 Table 11. Trimodal life cycle developmental time of Larra bicolor underground in field conditions of Gainesville, FL (29°40'N) 51 Table 1 2. Number of individuals, corresponding proportion and sex ratio per cohort and overall of Larra bicolor developing on Scapteriscus borellii underground at field varying-temperatures 51 vn

PAGE 8

Table 13 Parameter values for the linear model, lower thresholds and thermal constants (degree-days) for development of Larva bicolor per stage and total at constant ^ temperatures Table 14 Degree-days (days) required for Larra bicolor stages and total to develop at constant and field temperatures Table 15 Statistical significance of slopes in regression analysis of development of Larra bicolor stages and total, from the field and from incubation chambers 64 Table 16. Development of Ormia depleta (new strain) pooled sexes life stages 70 Table 17. Development of Ormia depleta (new strain) male stages 70 Table 18. Development of Ormia depleta (new strain) female stages 71 Table 19. T-test comparison of developmental time (days) of Ormia depleta (new strain) between sexes per temperature per stage and total life cycle 72 Table 20. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (new strain) pooled sexes life stages to develop 73 Table 21. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (new strain) male stages to develop 73 Table 22. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (new strain) female stages to develop 73 Table 23. Development of Ormia depleta (old strain) pooled sexes life stages 75 Table 24. Development of Ormia depleta (old strain) male life stages 75 Table 25. Development of Ormia depleta (old strain) female life stages 75 Table 26. T-test comparison of developmental time (days) of Ormia depleta (old strain) between sexes per temperature per stage and total life cycle 77 Table 27. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (old strain) life stages to develop 78 Table 28. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (old strain) male life stages to develop 78 viii

PAGE 9

Table 29 Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (old strain) female life stages to develop 78 Table 30. Comparison of slopes (bl) 1 and intercepts (a) 1 between the "new" and "old" strains of Ormia depleta by t-test and confidence limits at constant 84 temperatures Table 31. Time periods when experiments were conducted for studying the development of the two strains of Ormia depleta at fieldvarying temperatures of Gainesville, FL ^ Table 32. Development time in days 1 and corresponding temperatures during development per stage life cycle of Ormia depleta, new strain, underground in sub-chamber A (0-15 cm depth) in field experiments 92 Table 33. Development time in days 1 and corresponding temperatures during development per stage life cycle of Ormia depleta, new strain, underground in sub-chamber B (15-30 cm depth) in field experiments 93 Table 34. Development time in days 1 , standard errors and corresponding temperatures during development per stage life cycle of Ormia depleta, old strain, underground at various depths in field experiments 94 Table 35. Parameter values for the linear model, lower thresholds and thermal constants (degree-days) for development of Ormia depleta (new strain) per stage and total at constant temperatures 104 Table 36. Degree-days required by Ormia depleta (new strain) to develop in incubation chambers and at fieldvarying temperatures 104 Table 37. Statistical significance of slopes in regression analysis of development of Ormia depleta (new strain) stages and total, in field and in incubation chambers 107 Table 38. Regression slope comparison by 95% confidence limits of degree-days of developmental stages of Ormia depleta (new strain) developing at fieldvarying temperatures and in incubation chambers at constant temperatures 107 Table 39. Parameter values for the linear model, lower thresholds and thermal constants (degree-days) for development of Ormia depleta (old strain) per stage and total at constant temperatures 108 Table 40. Degree-days required by Ormia depleta (old strain) to develop in incubation chambers and in fieldvarying temperatures 109 ix

PAGE 10

Table 41 Statistical significance of slopes in regression analysis of development of Ormia depleta (old strain) stages and total, in the field and in incubation ^ ^ chambers Table 42 Regression slope test by 95% confident limits of degree-days of larval, pupal and total development of Ormia depleta (old strain) in fieldvarying temperatures and in incubation chambers at constant temperatures 1 1 1 Table 43 . Fertility and oviposition period of Scapteriscus vicinus reared at 26°C and photoperiod of 14:10, L:D 125 Table 44. Egg fertility and development of Scapteriscus vicinus at constant temperatures. 125 Table 45. Nymphal developmental time of Scapteriscus vicinus at two constant temperatures Table 46. Nymphal development of Scapteriscus abbreviatus at constant temperatures 131 Table 47. Developmental time of Scapteriscus abbreviatus from egg to adult, grouped by number of instars to complete development and sex ratio per group 133 Table 48. Numbers of Scapteriscus vicinus mole crickets, and numbers parasitized by the parasitoid wasp Larra bicolor, trapped in pitfall traps at 2 locations during 2 seasons in Gainesville, Florida 144 Table 49. Success of Larra bicolor parasitizing Scapteriscus vicinus 146 Table 50. Sex dimorphism in adult and pupal stages of Larra bicolor developed on naturally field-parasitized host mole crickets 147 Table 51. Linear equation models proposed to estimate soil temperatures 154 Table 52. Coefficient of determination values and statistical significance (pvalues) of slope and intercept for four models estimating soil temperature, using 500, 1000, or 2000 data points 156 Table 53. Coefficient of determination values and statistical significance (p-values) of intercept and slope of four models estimating soil temperature with soil and air temperature as independent variables, using 500, 1000, or 2000 data points 158 x

PAGE 11

LIST OF FIGURES Figure Page Figure 1 . Theoretical linear and nonlinear model development rate functions 16 Figure 2. Developmental time of Larra bicolor stages and total at constant temperatures 34 Figure 3. Survivorship of Larra bicolor life stages at constant temperatures 35 Figure 4. Developmental rates of Larra bicolor at constant temperatures. Symbols show observed data. Line trend fitted to the linear response 37 Figure 5. Oviposition time of wasps Larra bicolor reared at summer-like and autumnal seasons in Gainesville, FL; daily oviposition per wasp is represented by different line pattern per season 41 Figure 6. Proportion of individuals from three cohorts entering the life cycle stages of Larra bicolor developing on Scapteriscus borellii underground at fieldvarying temperatures 52 Figure 7. Degree-days required for pupal development of Larra bicolor underground at fieldvarying temperatures. Pupal stage grouped by kind of developmental times (short, intermediate, and long) 62 Figure 8. Comparison of degree-days required for Larra bicolor life stages and total to develop at constant and field varying temperatures. A) Egg, larval and prepupal development. B) Pupal and total development (short development only) 63 Figure 9. Development time of Ormia depleta "new" strain, pooled sexes, at constant temperatures 71 Figure 10. Survivorship of Ormia depleta "new" strain, pooled sexes, at constant temperatures 72 Figure 1 1 . Developmental rates of Ormia depleta "new strain", pooled sexes, at constant temperatures 74 xi

PAGE 12

Figure 12. Developmental time of Ormia depleta, old strain, life cycle at constant temperature Figure 1 3 . Survivorship of Ormia depleta old strain, pooled sexes, at constant ^ temperatures Figure 14. Developmental rates of Ormia depleta old strain, pooled sexes, life cycle at constant temperatures Figure 15. Developmental times of two strains (new and old) of Ormia depleta, pooled sexes, at constant temperatures. Means followed by different letter are statistically different ("t test", P<0.05) 80 Figure 16. Developmental times of two strains (new and old) of Ormia depleta males at constant temperatures. Means followed by different letter are statistically different ("t test", P<0.05) 81 Figure 17. Developmental times of two strains (new and old) of Ormia depleta females at constant temperatures. Means followed by different letter are statistically different ("t test", PO.05) 81 Figure 18. Linear development estimates by linear regression equations of the two strains of Ormia depleta at constant temperatures. Lower development was prolonged until crossing "y" axis 83 Figure 19. Survival (%) of the two strains of Ormia depleta at constant temperatures. Means followed by different letter are statistically different ("X test", PO.05) 85 Figure 20. Comparison of larval, pupal and total development of Ormia depleta, new strain, at two depths at field (F) experiments. Means followed by different letters are statistically different. Analysis by depth 94 Figure 21. Comparison of larval, pupal and total development of Ormia depleta, old strain, at two depths in field (F) experiments. Means by experiment are statistically not different: Duncan's test 95 Figure 22. Comparison of larval, pupal and total developments of the two strains of Ormia depleta at two depths underground. Field experiments (F) selected by similar mean temperatures. NS = new strain, OS = old strain. Means followed by different letters are statistically different. Mean comparisons grouped by stage and total 96 Figure 23. Regression lines of development of Ormia depleta (new strain) in incubation chambers set at constant temperatures ( B Chambers — ) and at field varying-temperatures ( O Field ). A) Larval development; B) Pupal development; C) Total development 106 xii

PAGE 13

Figure 24. Regression lines of development of Ormia depleta (old strain) in incubation chambers set at constant temperatures ( Chambers ) and at field varying-temperatures ( Field ). A) Larval development; B) Pupal development; C) Total development 110 Figure 25. Number of eggs per clutch, time after mating, and accumulated percentage of eggs of Scapteriscus vicinus at 26°C. DAM = Days after mating±SEM 123 Figure 26 Nymphal developmental time of Scapteriscus vicinus at constant 30°C in an incubation chamber. N=13 for 1 st instar; N=l for the other instars 127 Figure 27. Nymphal developmental time of Scapteriscus vicinus at constant 25°C in an incubation chamber ^ 27 Figure 28. Number of eggs per clutch (average) and days between clutch and corresponding proportion (%) of eggs per time of Scapteriscus abbreviatus at constant 26°C. DAM= days after mating 129 Figure 29. Proportion of eggs of Scapteriscus abbreviatus hatching at constant temperatures ^ Figure 30. Proportion of eggs from Scapteriscus abbreviatus hatching by clutch number at constant temperatures 130 Figure 31. Nymphal development of Scapteriscus abbreviatus at constant 25°C 132 Figure 32. Nymphal development of Scapteriscus abbreviatus at constant 30°C 132 Figure 33. Number of eggs hatched per clutch of two mole cricket species, Scapteriscus abbreviatus and S. vicinus at constant 26°C 136 Figure 34. Proportion of eggs hatched of two mole crickets, Scapteriscus abbreviatus and S. vicinus, at constant temperatures 137 Figure 35. Numbers of parasitized and unparasitized Scapteriscus vicinus mole crickets trapped in pitfall traps at 2 locations during 2 seasons in Gainesville, Florida, by date of collection 145 Figure 36. Numbers of nymphs and adults of Scapteriscus vicinus mole crickets, and proportions of each parasitized by the parasitoid wasp Larra bicolor, trapped in pitfall traps at 2 locations during 2 seasons in Gainesville, Florida, by pronotal size class 145 Figure 37. Pupal length and pupal width comparison between sexes of Larra bicolor. Circles show the proportion of overlapping sizes between sexes 148 xiii

PAGE 14

Figure 38. Temperatures recorded during 10 days in July from the soil in the experimental area (Exp.) at two depths (7.5 and 22.5 cm.) and from soil (15 cm depth) recorded by the Florida Automated Weather Network (FAWN) 153 Figure 39. Efficiency of soil temperature models to estimate temperature at 7.5 cm depth. Independent variable: Soil temperature from the experimental area (Exp. 7.5) at 7.5 cm depth. Dependent variable: Soil temperature (FAWN soil) from the Florida Automated Weather Network 156 Figure 40. Efficiency of soil temperature models to estimate temperature at 7.5 cm depth. Independent variable: Soil temperature from the experimental area (Exp. 7.5) at 7.5 cm depth. Dependent variable: Soil temperature (FAWNsoil) and Air temperature (not shown) from the Florida Automated Weather Network 158 Figure 41 . Temperatures recorded at two depths from the experimental area (Exp.) and from the alternative temperature record source (FAWN). Period from 5 July -16 Aug. 2000 160 Figure 42. Temperatures recorded at 7.5 cm depth from the experimental area and from the alternative temperature record source (FAWN). Period from 23-30 Aug. 2000. The predicted temperature was estimated with the equation: Y= -1.96286(S«,) 0.07562(Ato) + 3.1 13706(S,.,) + 0.546506(A,.i) -16.8321 161 Figure 43. Temperature for 22.5 cm depth underground before and after the period of 16-23 August 2000. A) Temperatures recorded before the missed temperature period. B) Temperatures recorded after the missed temperature period 1 62 Figure 44. Temperatures recorded at two depths from the experimental area (Exp.) and from the alternative temperature record source (FAWN). Period from 29 Sep. -11 Oct. 2000 163 Figure 45. Temperatures recorded at 7.5 cm depth from the experimental area and from the alternative temperature record source (FAWN). Period from 29 Sep. 1 1 Oct. 2000. The predicted temperature was estimated with the equation: Y= 6.61423(S«))-0.4854(A t o) -4.4884(S,.,) +0.35548(A,-i)-20.061 163 Figure 46. Temperature model for 22.5 cm depth underground for the period of 29 Sep. 1 1 Oct. 2000. A) Temperatures recorded before the missed temperature period and used to build the model. B) Temperatures recorded after the missed temperature 164 xiv

PAGE 15

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 RELATIONSHIP BETWEEN TEMPERATURE AND DEVELOPMENT OF THE ECTOPARASITOID LARRA BICOLOR (HYMENOPTERA: SPHECIDAE) AND THE ENDOPARASITOID ORMIA DEPLETA (DIPTERA: TACHTNIDAE) By Hector Cabrera-Mireles August, 2002 Chairman: J. Howard Frank Major Department: Entomology and Nematology The rates of development of the second introduced strain (from Bolivia) of the ectoparasitoid wasp Larra bicolor and of the two introduced strains (from subtropical Brazil and from temperate Brazil) of the larviparous endoparasitoid tachinid Ormia depleta were determined when reared at both constant and field temperatures in northern Florida. The data were used to examine the performance of the accumulated degree-day model (DD) to predict field development of both natural enemies parasitizing the mole cricket Scapteriscus borellii. Development and survivorship were studied in incubation chambers at eight constant temperatures, 15.0, 17.5, 20.5, 22.5, 25.0, 27.5, 30.0 and 32.5 ± 0.1 °C, at 14:10 (L: D) photoperiod. The relationship between temperature and developmental rate was estimated with linear regression analyses. Pooled developmental times for males and females were used to fit the linear regression models. The lower developmental threshold was estimated by the "xintercept" method. Development of all stages of the wasp occurred at temperatures from 25.0 to 30.0°C. No development was xv

PAGE 16

observed at 17.5°C, and only one wasp developed at 22.5°C. Development of O. depleta occurred in the range from 17.0 to 30.0°C. The field development was determined by confining newly parasitized hosts into 15 x 5 cm (L x diameter) chambers, filled with autoclaved sand, buried underground at up to 60 cm depth that were operated during all parts of the year except the coldest months. Field experiments with L. bicolor were run during autumn. Temperature underground was recorded every 30 min during field experiments. A degree-day model based upon constant temperature was fitted to data from field experiments for each parasitoid species and strain. A full description is presented of the developmental time (and rates) in day units, for constant and field experiments, together with the regression coefficients from constant temperatures and the validation of both data sets using the accumulated degree-day model. Data about the wasp are summarized in life tables and its field efficacy in parasitizing mole crickets is presented and explained. Data about the fly comparing development per sex and per strain are presented also. xvi

PAGE 17

CHAPTER 1 INTRODUCTION Three mole cricket species from South America arrived as immigrants to the southeastern USA in ships' ballast about the year 1900 (Walker and Nickle 1981). They are Scapteriscus abbreviatus Scudder, S. vicinus Scudder and S. borellii Giglio-Tos. The immigrant mole crickets became the most important pests of turf and pasture grasses in Florida. Their feeding and tunneling also destroy seedlings of vegetables, ornamentals, and tobacco. Insecticides commonly applied to control mole crickets in lawns, golf courses, and seedbeds are expensive and not always effective. In pastures there is no control that is economically feasible (Walker 1985a). Adult S. abbreviatus have very short wings, and are flightless. Its current distribution is restricted to coastal areas and two small areas inland of Florida. Adult S. vicinus and S. borellii have wings that slightly exceed the length of the body, and can fly. Their current dispersion includes eight states, from North Carolina south to Florida and west to Texas (Frank and Parkman 1999). In 1978, the University of Florida established a mole cricket research program (MC program) because of damage to pasture grasses (Walker 1985a). Damage (including control costs) to turf and pastures in Florida alone may approach $100 million annually (H. Frank personal communication 2001). Three biological control agents from South America have been successfully released into Florida for control of immigrant mole crickets, an ectoparasitic wasp Larra bicolor F., an endoparasitic fly Ormia depleta Wiedemann, and a parasitic nematode Steinernema scapterisci Nguyen and Smart (Frank and Parkman 1999). 1

PAGE 18

2 The neotropical digger wasp Larra bicolor (Hymenoptera: Sphecidae) has been known for decades as a parasitoid of large nymphs and adults of mole crickets (Castner 1984). A Puerto Rican stock originally from Brazil was introduced into Florida in 1981 . L. bicolor wasps were released at five sites in south and central Florida in conjunction with plots of their preferred nectar-source plant Spermacoce verticillata L. Establishment of Larra bicolor only at Ft. Lauderdale, and not at the four more northerly sites in Florida, suggested that this wasp, of tropical origin, could not withstand colder or longer winters farther north in Florida (Castner 1988). In the late 1980s a different, possibly more cold-tolerant strain of L. bicolor was imported from Bolivia, released in northern Florida, and a decade later appeared not only to be established but also to be spreading, indicating at least that mole crickets in some unknown proportions were being parasitized (Frank et al. 1995). Ormia depleta is an endoparasitoid of Scapteriscus spp. adults. Females find hosts by homing on the sounds of calling male crickets (Fowler 1987). O. depleta was imported from Piracicaba, Brazil (23°30'S), released at numerous sites in Florida, and soon spread throughout south and central Florida. This first strain failed to establish permanent populations in north Florida (>27° N), perhaps due to its poor tolerance to cold and long winters (Walker et al. 1996). In recent years, in late autumn, a few parasitic flies have been captured near Gainesville, and these were considered to be individuals dispersing from south toward north without re-establishing populations. In 1 999 a second tachinid strain was imported from southern Brazil (30°S); because of its origin farther south (in a cooler climate) it was hoped that this strain might be more cold-tolerant and might establish populations north of 29°N in Florida and other states (Frank 1994).

PAGE 19

3 The entomopathogenic nematode Steinernema scapterisci Nguyen and Smart (Rhabditida) was imported from Uruguay and released in Florida in 1985 (Parkman and Frank 1992). This nematode successfully attacks adults and large nymphs of S. borellii and S. vicinus, but is not very effective against S. abbreviatus, and has negligible effect against small and medium nymphs of any of these three pest mole crickets (Frank and Parkman 1999). There is no information about the effect of temperature on life history or seasonal phenology of any of the insect parasitoids. Limits of geographical expansion of L. bicolor and O. depleta are poorly understood. The fact that the first introduced strains failed to establish in northern Florida suggests a strong influence of temperature regimes, where either colder winters kill the parasitoids or cause the lack of adequate energy (nectar, etc.) sources for the adults. These arguments led to the introduction of a second strain with the hope that they would be more cold hardy and survive colder winters. The level of control of mole crickets exerted by these two insect parasitoids in Florida is questionable (e.g., Potter 1998). The questions arise from lack of adequate data. Development of plants and invertebrates is temperature-dependent with each organism requiring a specific measure of heat accumulation between lower and upper developmental thresholds to complete development (Hagstrum and Milliken 1988). Application of developmental thresholds and rates in the form of phenological models is often used in agricultural IPM programs to predict and manipulate pest and natural enemies population dynamics in the field (Tauber et al. 1994). Temperature plays a critical role in determining the rate of development, survival, and reproduction of insect species. Understanding the relationship between temperature

PAGE 20

4 and various life-history processes is important in developing and refining computer models for simulating biological control of pest populations. Physiological time measured in degree-days is effective in predicting the seasonality of life history stages of insects. The development of a degree-day model requires the determination of the temperature threshold above which development will occur. Developmental thresholds can be determined directly in the laboratory by measuring the time taken to rear insects through a series of developmental stages at a range of constant temperatures. Then the thresholds can be validated in the field (Wagner et al. 1984a, Campbell et al. 1974). Phenology of many species of poikilotherms as a function of temperature has been described by regression equations (Wagner et al. 1984b). Such regression equations can provide a method for calculating developmental times in insect population models or can be used for quantitative comparisons of the effect of temperature on the developmental times of different stages and species of insects (Hagstrum and Milliken 1988). Understanding the effects of temperature on summary life-history parameters such as the intrinsic rate of increase can be useful for examining the potential effectiveness of a species as a biological control agent, estimating its potential for population growth (Lysyk 2000). Information describing the life history of pest mole crickets and of their two insect natural enemies released in Florida is inadequate. In the best of the cases, the phenology of the parasitoid is based upon development at a single temperature. Thus, we lack the information necessary to describe development and reproduction of any of the

PAGE 21

5 insects involved. No development model has been proposed for any of the imported insect natural enemies of mole crickets. The objective of this study was to investigate the development of the digger wasp Larra bicolor and of the tachinid fly Ormia depleta parasitising mole crickets at both constant and field temperatures. These data are used to examine the feasibility of using a degree-day accumulation model to predict the development of the two parasitoids in the field. This dissertation is organized by parasitoid species. The studies on the parasitic wasp present the development of the wasp at constant temperatures followed by the field study and finally a degree-day model simulating the wasp development in the field and its validation. The parasitic fly studies are composed of three experiments; firstly, studying its development at constant temperatures, then in field experiments and finally presenting the proposal of a degree-day model and its validation in field development of this parasitoid.

PAGE 22

CHAPTER 2 REVIEW OF LITERATURE Introduction Three species of mole cricket (Scapteriscus spp.) are the most damaging pests of turf and pasture grasses in much of the southeastern United States. They destroy turf and grasses by feeding on roots, leaves and shoots, and by tunneling, which disrupts the soil around roots, preventing uptake of water and nutrients (Hudson 1997, cited in Frank and Parkman 1999, Walker 1985a). Damage to Florida turf was estimated as $30 million annually in Florida alone (Walker 1985a), $45 million (damage and cost of control) in 1986, with a further $33 million in Alabama, Georgia, and South Carolina combined (not including damage in North Carolina, Mississippi, Louisiana, and Texas) ( Frank and Parkman 1999). More recent estimates are of annual expenditure of more than $18 million on chemical insecticides in Florida turf, and of more than $12 million in damage in Georgia turf (Hudson et al. 1988). Damage to pasture grass in Georgia in 1996 was estimated as 0.83 million for control and as $9.8 million in damage (Bunti n et al. 1997, cited in Frank and Parkman 1999). The most promising strategy for solving immigrant mole cricket problems is the use of classical biological control. In 1978, in response to increasing concerns of cattleman, turf managers, and homeowners, the University of Florida initiated a research project on mole crickets (Walker 1985a). Three biological control agents from South America have been successfully released into Florida for control of adventive mole 6

PAGE 23

7 crickets, Larra bicolor F. (Hymenoptera: Sphecidae), Ormia depleta Wiedemann (Diptera: Tachinidae) and Steinernema scapterisci Nguyen and Smart (Rhabditida: Steinernematidae) (Frank and Parkman 1999). Information regarding natural history of the Scapteriscus mole crickets and of the insect parasitoids in the continental United States is presented in this chapter. The Pest Mole Crickets, Scapteriscus spp. Four mole crickets (Orthoptera: Gryllotalpidae) are listed as occurring in Florida: Neocurtilla hexadactyla Perty, the northern mole cricket; Scapteriscus abbreviates Scudder, the shortwinged mole cricket; Scapteriscus borellii Giglio-Tos, the southern mole cricket; and Scapteriscus vicinus Scudder, the tawny mole cricket. The northern mole cricket is the only species native to the southeastern USA including Florida. The three other mole cricket species are immigrants from South America that were accidentally brought in by ships' ballast about the year 1900 (Walker and Nickle 1981). Mole crickets are fossorial insects with greatly enlarged forelegs that are used for digging. Some of the adults and nymphs of a population move about on the soil surface on warm nights with high humidity. Depth of the mole crickets in the soil depends upon temperature and moisture (Frank and Parkman 1999, Lake 2000). High water tables force the mole crickets to the surface. Males are more active than females; maximum average depth reported by Abate (1979) was 50 cm for S. borellii males and 44 cm for S. vicinus males. Scapteriscus mole cricket burrows as deep as 75 cm have been recorded (Frank and Parkman 1999). Subterranean depths of oviposition by S. borellii and S. vicinus are

PAGE 24

8 30 cm maximum depth recorded in the field by Hayslip (1943) and a maximum mean depth of 60 cm for S. borellii from laboratory experiments by Lake (2000). Adults of S. abbreviatus have very short wings. In contrast, wings of S. borellii and S. vicinus slightly exceed the length of the body, which explains their current wide distribution in the southern coastal plains from North Carolina south to Florida and west to Texas. S. abbreviatus cannot fly and its distribution is restricted to some small areas in Florida (Frank and Parkman 1999). S. borellii and S. vicinus sometimes fly in enormous numbers. Their flights apparently serve two functions: (1) local searching for mates and new egg-laying or calling sites and (2) long range dispersal (Walker 1985b). Adult mole crickets, like many other crickets, produce a calling song that functions to bring the sexes together. Females sometimes respond to the call by walking or flying to the male (phonotaxis) and mating with him. The calling period is correlated with female activity and flight periods (Walker 1985b). The intensity of the song depends upon male size and soil moisture (Forrest 1985). The following information was presented by Walker (1985b). Mole crickets spend nearly all their lives underground. Adult S. vicinus fly in large numbers typically in March but as early as February in Florida after warm winters. They mate (although some matting occurred the previous autumn); they begin as eggs laid in underground chambers. Eggs take some three weeks to incubate. Nymphs hatch from eggs as early as April, but continue to hatch from later deposited eggs through June. Nymphs take some five months to develop during summer months, and newly formed adults begin to appear in September. There is a minor peak of flight activity in the autumn, as early as August in the far south and as late as December farther north if the weather remains warm. In most

PAGE 25

of the southern USA, the spring activities occur in S. borellii about three weeks later than in SI vicinus, but the autumnal activities are at the same time. From central Florida northward, & vicinus and S. borellii have one generation each year. Most S. vicinus overwinter as adults, and most S. borellii overwinter as large juveniles. In southern Florida S. vicinus maintains its one year life cycle, but S. borellii has two generations annually. In total contrast, all developmental stages of S. abbreviatus occur throughout the year but with two peak ovipositional periods, one in late spring and one in winter. The Digger Wasp Larra bicolor Natural History of Larra bicolor Wasps of the genus Larra (Hymenoptera: Sphecidae) are the only sphecids known to attack mole crickets, which all of them do. Unlike other digger wasps, which are generally predators, all Larra wasp species are parasitoids (Sailer 1985). Worldwide, 65 species of Larra are known (Bohart and Menke 1976). Only one species, Larra analis F., a parasitoid of the northern mole cricket, is native to the United States. Sixteen species have been recorded from South America, where the diversity of species of mole crickets belonging to the genus Scapteriscus is greatest (Sailer 1985). The following information was presented by Sailer (1985). Apart from preferences for different species of mole crickets, the biology of the various species of Larra is remarkably similar. Females do not construct a nest, and prey paralysis is temporary with the host reviving soon after egg deposition. A single egg is firmly attached in the soft, membranous thoracic tissue between the host's first and second legs. The Larra larva lives ectoparasitically until almost fully grown. It then kills and consumes the rest of its host. Larra females lay about 30 eggs during their life span. Full-grown larvae form underground cocoons of

PAGE 26

10 sand grain cemented together with a secretion from the labial glands. The L. bicolor adults emerge from their cells 60 to 80 days after eggs are laid on the hosts. At Fort Lauderdale (south Florida) at least four generations are probably produced annually. Females hunt for mole crickets during early morning and mid afternoon hours. During the middle of the day they visit flowers where they feed on nectar and possibly pollen. L, bicolor is most commonly found on the flowers of two quite unrelated plants, Spermacoce verticillata L. (Rubiaceae) (native to south Florida) and Hyptis atrorubens L. (Lamiaceae). The close association of Larra bicolor with these plants suggests that it may not inhabit areas where these plants are not available. Life Cycle of Larra bicolor in the Laboratory The following information was presented by Castner (1988). The following data were recorded from parasitoids reared in a laboratory maintained at approximately 26 ± 2°C and 65 ± 5% relative humidity. The 4.0-4.5 mm Larra egg hatches in 6-7 d. The first instar larva remains fixed to the original point of attachment. The larva passes through five instars, feeding initially on hemolymph, but progressing to body tissues as heavily sclerotized mandibles develop. The larva completes the first four instars 8-9 d after hatching, increasing gradually in size to 12-19 mm. The final larval size is 14-26 mm. By 10-11 d post-hatching (16-18 d post-oviposition), the wasp grub has killed the host and consumed the soft inner tissues. The mature larva spends approximately one day in constructing an ovate cocoon of soil particles or sand grains. The wasp adult emerges 6-8 weeks after the cocoon is completed.

PAGE 27

11 Establishment of Larra bicolor in Florida The wasp Larra bicolor was the first classical biological control agent of mole crickets to be introduced and established in Florida. A Puerto Rican stock originally from Brazil was introduced in 1981; wasps were released at five sites in south and central Florida in conjunction with plots of their preferred nectar-source plant Spermacoce verticillata. Establishment of Larra bicolor only at Ft. Lauderdale, and not at the four more northerly sites in Florida, suggested that this wasp, of tropical origin, could not withstand colder or longer winters farther north in Florida (Castner 1988, Frank and Parkman 1999). In 1988-89 a Bolivian stock of L. bicolor, possibly more cold-tolerant, was introduced, and a decade later appeared not only to be established but also to be spreading in northern Florida indicating at least that mole crickets in some unknown proportions were being parasitized (Frank and Parkman 1999). The Tachinid Fly Ormia depleta Natural History of Ormia depleta All species of the family Tachinidae develop as parasites of other arthropods. Some groups of species are restricted to closely related kinds of host insects; for example the group Ormiini consists of species that parasitize katydids and other night-singing Orthoptera (Sailer 1985). Ormia depleta is a South American tachinid that attacks mole crickets of the genus Scapteriscus (Fowler 1987). Adults are crepuscular and nocturnal parasitoids. Females are phonotactic, locating their hosts by their calling songs, and they deposit larvae on or near the host (Fowler and Kochalka 1985). The first stock was brought from Piracicaba, Brazil (23°S), and released in Florida in 1988. Since then it has become established in 38 counties in Florida, and was

PAGE 28

12 considered as largely responsible for a decline in mole cricket populations (Parkman et al. 1996). No evidence of their extension farther to the north exists despite releases in northwestern Florida, Georgia, Alabama, and North Carolina (Frank et al. 1996, Walker et al. 1996). By 1994, this strain disappeared from the Gainesville area (29° 40'N latitude), perhaps due to its poor tolerance to cold and long winters (Walker et al. 1996). In recent years, in late autumn, a few parasitic flies have been captured near Gainesville, and these were considered to be individuals dispersing from south toward north without re-establishing populations. In 1999 a second tachinid strain was imported. It was hoped that this "new" strain, from 30°S in Brazil, might be more cold-tolerant and could establish populations north of 29°N in Florida and other states (Frank 1994, Potter 1998). Several studies were performed with the first strain of Ormia depleta, focused on establishment and range (Frank et al. 1996), seasonality (Walker et al. 1996), effectiveness in controlling mole crickets (Parkman et al. 1996), techniques for rearing in the laboratory (Wineriter and Walker 1990) and food source of the adults (Welch 2000). This "old" strain was in culture from 1987 (Frank et al. 1996) until 1999, but it is the "new" strain that is now in culture to supply populations. Life History of Ormia depleta in the Laboratory Wineriter and Walker (1990) determined the life history of the first strain under laboratory conditions at 23-25°C. The life cycle of O. depleta, from oocyte to adult, is about 36 d. Numbers of planidia/female at time of examination range from 70 to 310, mean = 187 ± 45, in laboratory-reared females. Most gravid females contain eggs and immature larvae as well as planidia, suggesting that females deposit larvae over many

PAGE 29

13 days. Emergence of adults occurs very early in the morning and spans a 3-4 d period. Emergence of males generally precedes that of females. Sex ratio is usually 1:1. Insect Modeling Numerous independent variables influence arthropod development but none is more important than temperature. In fact, the effects of temperature are so profound that most phenology models consider only this one variable. Other biotic and abiotic variables may influence development times, such as food quantity and quality, population density, humidity, rainfall, pH, and photoperiod (Hagstrum and Milliken 1988, Wagner et al. 1991). Population factors can also influence development times, for example genetic differences (Wagner et al. 1991) and gene expression (Gilbert and Raworth 1996). The temperature dependence of insect development has been frequently investigated. Temperature affects both the time of development as well as fecundity; consequently, the appearance and dynamics of insect populations in the field are often dictated by ambient temperatures. For this reason there has been considerable interest for a long time in the temperature relationship of development time and fecundity mainly of filthy, annoying, and disease-transmitting insects in order to apply it in predicting the outbreak-time and the dynamics of pests (Ratte 1985). Predicting the seasonal occurrence of insects is essential for accurate scheduling of census samples and control tactics. These predictions require an understanding of insect development, and because temperature has a major effect on this process, the thermal requirements of development are often a basic to prediction (Wagner et al. 1984b).

PAGE 30

14 Temperature determines life of organisms more than do any other environmental factors. The cause of this temperature sensitivity lies in the fundamental fact that chemical compounds build all organisms and that all processes of life are made by chemical reactions that follow the laws of thermodynamics. Temperature not only acts upon the rate of all chemical reactions according to the Arrhenius equation, but also causes conformational transitions of proteins, phase transition of lipids, change in the structure of water, etc. (Alexandrov 1977, cited by Hoffman 1985). However, many results from field observations and laboratory experiments in which the insects have been subjected to fluctuating temperatures cannot be brought into this concept. A too simplistic physiological concept of the temperature action might, however, lead to erroneous insight into the insect's capabilities of adapting to various thermal environments (Ratte 1985). Constant vs Fluctuating Temperature Influence on Insect Development Generally the velocity of insect development increases with elevation of temperature. Many workers have constructed hyperbolae by plotting the constant temperatures at which the insects were reared against the duration of development (Wigglesworth 1972). Janish (1930), cited in Wagner et al. (1984b), stated that the reciprocal of the hyperbolae represents the relative velocity of development, and the line based on the reciprocal shows the rate of acceleration of development with rising temperature. From these observations numerous empirical and physiochemical formulations of insect development have been made. Two such formulations are the daydegree or temperature summation rule and the non-linear inhibition model (Behrens et al. 1983).

PAGE 31

15 Wagner et al. (1984b) presents a very clear explanation of what thermal summation or rate summation means: many empirical and biophysical models describe the time versus temperature or the rate versus temperature curve of insect development. Models of rate versus development relationships are used widely to predict insect development times because mean daily (or hourly) rates can be accumulated under fluctuating temperature environments. These mean rates are summed to equal 1 .0, at which time mean cohort development is complete. However models of time versus temperature relationships are primarily descriptive, although the inverse of these regressions also yields development rates. Constant Temperatures Typically the development time of a larva decreases exponentially with increasing temperature. In almost all insects, at lower temperatures the development time decreases steeply with increasing temperature; at medium temperatures the slope of the curve becomes smaller until it reaches zero at the so-called optimal temperature for development, which lies usually close to the upper limit. The curve of developmental rate (reciprocal developmental time) takes a more or less sigmoidal course including an inflexion point mostly between 20° and 30°C, as well as a maximum typically between 25° and 30°C (Ratte 1985). See Figure 1.

PAGE 32

16 The developmental rate curve displays a wide, approximately linear, range around inflexion point; a straight line often fits this part of the curve. Consequently, the time of development follows a hyperbolic function of temperature over this range. Similar relationships between developmental rate and temperature have been found frequently and have led to establishment of the "rule of thermal summing" and the classical method of "temperature summation" (Ratte 1985). TC = td(T-T 0 ) Where: td : time of development TC : thermal constant T : experimental temperature To : threshold temperature T-To: effective temperature above To. In other words, the time of development multiplied by the effective temperature (T-To) is a constant (thermal constant). "Thermal summing" implies the existence of a distinct threshold temperature (To), at which the developmental rate becomes zero

PAGE 33

17 (developmental zero; developmental threshold). The developmental threshold has to be regarded as theoretical because it can be extrapolated only from a linear rate relation and cannot be determined by rearing experiments; the threshold may fall outside the favorable range for development, and also the infinite developmental time at temperatures near the threshold makes its measurement practically impossible. Although the cessation of development below critical temperatures can be observed easily, the concept of real developmental zero has not been unequivocally accepted. When subjected to subthreshold temperature some insects develop faster after subsequent exposure to normal temperature indicating that some development also takes place below the threshold (Andrewartha and Birch 1954, Ratte 1985). Ratte (1985) points to an important phenomenon of developmental rate curves: the degree of temperature dependence, expressed as slope of the rate curve, or Qio, shows a wide variation among various insects and some do indeed possess the capability to develop independent of ambient temperature (exemplified with the carpet beetle, Attagenus megatoma). Fluctuating Temperatures The developmental time of insects reared under a temperature fluctuating regime has usually been found to differ from that under the corresponding constant temperature. Considerably higher acceleration may occur when the low temperature falls below the developmental threshold. The same has also been observed with nondaily alternating temperatures or single transfers from sub-threshold to favorable temperatures. Consequently, it has been concluded that under thermoperiod conditions also below the threshold "some development may take place": this is assumed to occur because the

PAGE 34

18 temperature requirements of some rate-limiting development steps are nevertheless satisfied (Ratte 1985). In cases where in one species different thermoperiods have been tested within the entire favorable range for development, the following rule has often been observed: temperature fluctuations in the range below the inflexion point of the developmental rate curve may cause an acceleration in development, while the converse is true for thermoperiods alternating about the temperature optimum; there are slight or no deviations in developmental time for constant temperatures with fluctuations around medial temperatures of the favorable range. This rule was propounded quite early and can also be derived from the non-linearity of the developmental rate versus temperature relation established under constant temperature (Ratte 1985). Applicability of Constant and Fluctuating Temperatures in Insect Developmental Models Most of the models, called physiological models, transform temperature into a physiological scale, such as the classical degree-day summations (Wagner et al. 1984b). To describe field stage frequency datathat is number of individuals in each stage of developmentin most cases parameters are fitted in two steps. First, parameters of the physiological scale are estimated from constant temperatures in the laboratory. Second, the parameters of the phenological model are estimated from the proportion of individuals in each development stage on the physiological scale. The applicability of the phenological model requires that thermal constants, determined from constant temperatures in the laboratory, can be extrapolated to natural conditions (Wagner et al.

PAGE 35

19 1984b, Candy 1991, Manel and Debouzie 1997a). However, development rates at constant temperatures often differ from those at fluctuating temperatures (Worner 1992). Manel and Debouzie (1997a) say that deriving the relation from different constant temperatures may be questioned; a model aimed at predicting insect development in natural populations requires other experimental designs by recording temperature directly in the field or in the laboratory under fluctuating temperatures. Degree-day models extrapolated to natural conditions may or may not be successful for many reasons not related to temperature. Studies of the effect of fluctuating temperatures on insect development often show that, given the same mean temperature, insects appear to develop at rates that differ under fluctuating conditions from those at constant temperatures. Such studies may indicate that development is accelerated, decelerated, or the same as that under constant conditions. Such variability in reports appears to arise from studies where a limited temperature range is investigated (Worner 1 992). Temperature-Dependent Models for Studies on Insect Development With the advent of computer simulation to the applied biology sciences, there has been increased interest in mathematical models that describe organism growth and development as a function of temperature (Sharp and DeMichele 1 977). Many empirical and biophysical models describe the time versus temperature or the rate versus temperature curve of insect development. Models of rate versus temperature relationships are used widely to predict insect development times because mean daily (or hourly) rates can be accumulated under fluctuating temperature environments (Worner 1 992). Models are important tools for predicting future events for development times in particular and population dynamics in general (Ellsbury 1991). Many pest management

PAGE 36

20 systems rely on models of arthropod and host phenology, and there seems to be no end to the assemblage of new approaches contrived for this purpose (Wagner et al. 1991). Statistical and deterministic simulation modeling rely on a complex process made of trials, errors, and gradual improvement of the simulations (Mesple et al. 1996) consider the major problem is to be able to quantify the quality of the simulations in order to know whether a modification of the concepts, the laws simulating the processes or the parameters improve it. Linear Model This model assumes a linear relationship between development rate and constant temperature; therefore, the number of degree-days (or thermal units) required to complete development is assumed constant at all favorable temperatures. The duration of development is calculated by adding up the number of thermal units (degree-hours or degree-days) contributed at each temperature (Wagner et al. 1984b cite very old reports: Candolle 1855 and Reibisch 1902 among others). The thermal constant or degree-days for completed development are most easily derived from the reciprocal of the slope of the regression line fitted to development rate data, obtained typically at constant temperatures (Worner 1992). This method has been applied to plant phenology (Schroder and Sondgerath 1 996) This degree-day approach is used widely because it requires minimal data for formulation, is easy to calculate and apply, and often yields approximately correct values. However this approach is valid only over intermediate temperatures. Furthermore, the threshold temperature is often determined empirically by extrapolating the straight line to the temperature axis. For these reasons, the number of degree-days required for complete

PAGE 37

21 development is too low at temperatures near the lower threshold and too high at or above the optimum. These limitations are important and should not be overlooked when interpreting the results of this technique (Wagner et al. 1984b, Manel and Debouzie 1997a, b). The rate summation formula is D = £r[T(t)Ji/ where the proportions of development taking place at each time increment are summed so that, when D equals 1, the insect has completed development. This model measured in physiological units is not purely descriptive or statistical in nature but is biologically motivated. However it is based on population attributes rather than the progression of individuals through stages over time (Candy 1991). In spite of inherent limitations of the thermal model, to that linear portion of insects' developmental rate and the obligated definition of a lower threshold, this thermal model cannot be ruled out. As expressed by Liu and Meng (1999) insect development may occur most frequently at some temperature range considered in that degree-day summation and the model may fit well with natural varying temperature development although the logistic model better estimates development below the lower threshold. Logan et al. (1976) considered the degree-day model invalid for mite populations. Tanigoshi et al. (1976) showed that this model did not estimate development of field populations of mites. Gilbert and Raworth (1996) said that in insects, development rates increase almost linearly with temperature over the normal range of operating temperatures. The published data cover more than 300 species, and in every case the effect of temperature on

PAGE 38

22 development rate is effectively linear, but falls off at average daily temperatures higher than those normally experienced in the field. Lamb (1992) working with aphid clones detected some genuine curvature at low temperatures, but for practical purposes the relation was linear. Belehradek (1935) (cited in Andewartha and Birch 1954) described the response curve with an exponential equation that becomes linear on a log scale. This approach does an inadequate job of describing development rates at the temperature extremes. Non-linear Models Because of the difficulties of predicting insect development at temperature extremes, empirical non-linear models have been proposed, attempting to explore this response beyond the intermediate curve temperatures. Non-linearity of development (the Kauffman effect, Worner 1992) or rate summing effect (Ratte 1995) implies that development rates at constant temperature differ from those at fluctuating temperatures. Models have been developed to incorporate nonlinearity and variable temperatures (Hagstrum and Milliken 1991, Manel and Debouzie 1997a, b). Examples of nonlinear models are presented below (some are described in Table 2-1): The symmetric and asymmetric catenary model. Proposed by Janish (1930) and cited by Wagner et al. (1984b), this model describes development time as a function of temperature. This equation combines two exponential curves, the reciprocal of which describes the accelerating phase of development rate up to the optimal temperature and the decelerating phase beyond the optimum. This model is criticized for inadequate fits and computational difficulties.

PAGE 39

23 Secondand third-degree polynomials. These have been used to regress development time against temperature although they have not been used for prediction purposes (Tanigoshi et al. 1976). Hagstrum and Milliken (1988) compared the polynomial model: d = t + 1 2 in which d is developmental time, and t is centigrade temperature, and was less accurate than the Sharp and DeMichele (1977) model in describing developmental time of nine species of stored-product Coleoptera under various temperatures, moisture levels, and diets. Logistic equation. This describes development rates as a function of temperature (Davidson 1944). This equation is used widely although it is not very descriptive at one or both ends of the response curve. Manel and Debouzie (1997a) concluded that this model gives relatively close estimates of stage duration when applied to a single stage of development and when base temperature is known; but, when base temperature is misestimated, stage duration is either underor overestimated). Liu and Meng (1999) demonstrated that the logistic model better fit field varying temperature, especially when temperature falls below the lower developmental threshold. This model gives better estimates than the Wang nonlinear model (Wang et al. 1982). An exponential equation. Similarly, a normal distribution model is proposed by Pradhan (1945 and 1946 as cited in Wagner et al. 1984b). When compared with the logistic model, it yields better results (Siddiqui and Barlow 1972). A modified sigmoid equation model. This model was proposed by Stinner et al. (1975) to describe the effects of temperature on development rates. This model does not work well at higher temperatures.

PAGE 40

24 The two exponential equations model. Logan et al. (1976) described developmental rates at intermediate and high temperatures. Logan's model was more descriptive than Stinner's modified sigmoid equation and third degree polynomial (Logan et al. 1976). Lactin et al. (1995) modified Logan's model; it fitted better at lower temperatures in ten of eleven cases. They considered this model better than Sharp and DeMichele's model. Ryoo et al. (1991) considered this model appropriate in describing development of a larval parasitoid. Gould and Elkinton (1990) compared the linear model, Logan's model, and Sharp and DeMichele's model, in studying the development of a wasp parasitoid, Logan's model better estimated development at temperatures below 15 °C. No differences were observed at range temperatures of 15° and 30°C. A complex biophysical model. This model proposed by Sharpe and DeMichele (1977) describes the non-linear response in development rates at both high and low temperatures as well as the linear response at intermediate temperatures. Schoolfield et al. (1981) improved this model making it suitable to the nonlinear approaches. This model is considered as the most suitable to describing insect development rates as a function of temperature by Wagner et al. (1984b). Shaffer (1983) applies this model to 194 cases from the literature, failing in 20% in explaining variations in developmental rate, making this model not true for all species. Hagstrum and Milliken (1991) say this model provides an alternative to degree-day model accumulation method for prediction of developmental times at fluctuating temperatures in the field. Stochastic Models Insects reared under identical conditions develop at different rates due to inherent differences among individuals in a population. This stochastic aspect of development is

PAGE 41

25 important, although usually overlooked, in population models of poikilotherm development. It is not surprising, therefore, to find many models, linear and nonlinear, that describe the effect of temperature on insect development; however, most of these models only predict development time from mean rate versus constant temperature relationships. Usually, no consideration is given to the variability in insect development time among individuals. On the other hand, models that consider the stochastic aspects of development time use frequency distributions to determine the probabilities of development time for individuals in a population (Wagner et al. 1984b, Candy 1991, Manel and Debouzie 1997b). Wagner et al. (1984 a) say that three general stochastic approaches are recognized: Monte Carlo techniques use pseudorandom numbers to estimate the development time of individuals in a sample population (Menke 1973, Hardman 1976). The disadvantages of this approach are that it is computationally time-consuming because random numbers must be generated for each individual in the same population, and its outputs must be interpreted statistically. Phelps et al. (1993) consider this method as feasible because it uses a fixed number of individuals from one generation to the next and simulates the timing of events rather than the population dynamics of the insects. Stinner et al. (1975) developed a cumulative function that is fitted to the developmental times at <1, 25, 50, 100% of measured distributions. This probability distribution provides a good description of distributions at different constant temperatures, but when used in population models, the function is not fitted directly to measured distributions. The final stochastic approach mentioned by Wagner et al. (1984b) applies a probability function directly to the measured distributions. This approach has a mechanistic basis proposed by Sharp et al. (1977). These investigators used normal and quadratic probability density functions to describe the variability in development rates. Although these functions often provide good results, their goodness-of-fit capabilities are unsatisfactory when distributions of rate are asymmetric. Wang model. Described by Wang et al. (1982), this gives a very good fit to data over a whole range of constant temperatures that permit complete development. Liu and

PAGE 42

26 Meng (1999) compared this model with some other nonlinear models, such as Sharpe and DeMichele (1977), Stinner et al. (1975) and Logan et al. (1976), and found that Wang model offered the highest R 2 and the least estimation bias. In order to try to minimize the effect of asymmetry of distributions of rate, Wagner et al. (1985) propose the Weibull function to describe the distributions of normalized development time with good results. Ellsbury (1991) applies this same function on development of immature stages of clover stem borers at constant temperatures with good results. The predicted distribution of development time can be used to optimize census sampling and control procedures, or it can be used as input into larger population dynamics models (Wagner et al. 1985). Continuation ratio model. This model deals with probability of individuals being in each stage, as proposed by Candy (1991) and re-evaluated by Manel and Debouzie (1997a, b). This model applied to stage (multistage) frequencies obtained by random daily sampling, gives unbiased estimates only when a common physiological scale applies to each stage. When stages differ in physiological scale it gives nonindependent estimates of duration. The Weibull distribution is integrated to this model. It gives a better fit than the logistic model. Mixed Models Recent phenological models incorporate temperature-dependent development and inherent stochastic variation. This approach estimates the proportion of insects in a given development stage as a function of accumulated heat or degree-days (Dennis and Kemp 1988)

PAGE 43

27 Stage successive logistic model. Manel and Debouzie (1997a) combined rate summation, logistic equation and a stochastic description of individuals in a binomial probability. This model gives a better estimate than the continuation ratio model (Candy 1991). Conclusion As a way to express something that reflects what I perceived through the writing of this paper, I have chosen to quote what Wagner et al. (1985) conclude: 'The introductory paragraph of Uvarov's (1931) classic monograph, Insects and Climate, states "Economic entomologists at the present day are no longer satisfied with merely recording the outbreaks of insect pests and with devising means for their control. They realize more and more that their chief aim and highest ambition must be to foresee and to prevent outbreaks. In order to do this, all conditions accompanying and causing outbreaks must be thoroughly investigated and elucidated; in other words the epidemiology of insect pests must be the central problem of economic entomological research, which should be carried out from the ecological point of view." There is a greater opportunity today to predict population events, a requisite to prevent outbreaks. Vast amounts of biological and economic data are available on most major pests; and new approaches to population research and modeling, as well as inexpensive high speed computers, provide the means to identify, integrate and apply this knowledge for predictive purposes. Recent advances in entomology have occurred as a result of interdisciplinary research. Specialists from outside of entomology have, and will continue, to apply new techniques to our field. These techniques must be integrated into the mainstream of entomological application.

PAGE 44

28 Table 1. Insect models cited in the Review of Literature chapter. Sharpe and DeMichele (1977), modified by Schoolfield et al. (1981): RH025 298.15 exp 1 298.15 1 + exp HL ( R TL + exp HH( 1 R [tH T Where: r(T) =niean development rate at temperature T(°K) R =the universal gas constant (1 .987 cal degree" 1 mole" 1 ) RH025= development rate at 25 °C (298.15 °K) assuming no enzyme inactivation HA =enthalpy of activation of the reaction that is catalyzed by a rate-controlling enzyme. TL = Kelvin temperature at which the rate-controlling enzyme is half active and half low-temperature inactive. HL = change in enthalpy associated with low temperature inactivation of the enzyme. TH = Kelvin inactivation at which the rate-controlling enzyme is high active and half high-temperature inactive. HH = change in enthalpy associated with high-temperature inactivation of the enzyme. Wagner et al. (1984) say they are aware of no other poikiloterm development rate model with greater flexibility, goodness-of-fit capability, or of stronger theoretical foundation. Nevertheless, the use of this model is limited to those with experience in nonlinear parameter estimation. To extend the Sharp and DeMichele model to biologists they present a SAS-computer program for its use. Weibull function: F(jc) = l-exp(-[(^-/)/^f) Where: F(x) : the probability of complete development at normalized time x y, T| and P : are estimated parameters Logistic equation. Liu and Meng (1999): v ^ =K /{\ + Q x V {a-pr)) Where: K, a and p : are regression parameters fitted to the data of mean development rates V(T) K means the upper asymptote of this shallow sigmoid curve. Table 1 . Continues.

PAGE 45

29 Logistic regression equation. Manel and Debouzie (1997a): l + exp(ar 0 +a t DD t ) Where: do and cti : are parameters to be estimated DD] : the number of degree-days accumulated at time tj Continuation ratio model. Candy (1992): Pi = exp(b 0l +b ]WD] ) l + exp(* 01 + &,,£>£>,) l f p l cxy(b 0i +b u DDi) h 'Jl + expfc+W) r-l *=1 Where : bo and b\ : are parameters to be estimated. DD] : the number of degree-days accumulated at time t. Wang model. Wang et al. (1982): V{T) = ft l + exp(-r(r-7;)) 1 -exp fj"—T W 1-exp f t u -tX\ j j Where: V(T) : is the mean development rate at temperature T(°C) H : is the value of the upper asymptote of the curve r : is the exponential increase rate To : is the optimum temperature for development Tl and T H : are, respectively, the minimum and maximum temperature boundaries : boundary width at the lower and upper temperatures.

PAGE 46

CHAPTER 3 DEVELOPMENT OF THE WASP LARRA BICOLOR F. Development and reproduction of the wasp Larra bicolor at constant temperatures Introduction Scapteriscus mole crickets are major pests of turf and pasture grasses in the southern USA. The three mole cricket species that arrived about 1 900 as immigrants have being under classical biological control investigation since 1 979 under the University of Florida mole cricket research project (Walker 1985a). Two introductions of the ectoparasic digger wasp Larra bicolor have been made in Florida. Their different success is reported in chapter 2 (pages 10-11). The role of the environmental temperature has been considered the main limitation for Larra bicolor populations to establish and to disperse in northern Florida and in other southeastern states. As a way to enhance the biological control of this parasitoid, it is important to determine its vital statistics that may explain its seasonal phenology. These vital statistics are an important and basic component for understanding the dynamics of a species (Lowry et al. 1992). Fertility life tables are appropriate to study the dynamics of animal populations, especially arthropods, as an intermediate process for estimating parameters related to the population growth potential, also called demographic parameters (Southwood 1978, Maia et al. 2000). 30

PAGE 47

31 The general biology of the wasp L. bicolor has been poorly studied. Castner (1988) reported developmental time of L bicolor under a single temperature regime, but failed to report variance. Insects develop within a species-specific temperature regime. The inverse relationship between developmental rate and temperature results in a sigmoid curve with a straight line at the intermediate temperatures. This curve is a fundamental characteristic of an insect life history (Taylor 1981, Wagner et al. 1984b). The most widely used approach to study the effect of temperature on developmental rates of poikilothermic organisms is the thermal summation or degree-day (DD) model. This method is one of the oldest and most commonly used for insect modeling and for predicting biological events controlled by temperature accumulation (Wagner et al. 1984a, b, Higley et al. 1986). The objective of this study was to investigate the effect of temperature on development, survivorship and reproduction of Larra bicolor. Materials and Methods Development Placing a female wasp into a 200-ml plastic vial and then introducing the host induced parasitism of hosts (S. borellii). Parasitism occurred almost immediately most of the times. Parasitized hosts were placed individually in 200-ml plastic cups filled with wet-sterilized sand. Development and survivorship were studied in incubation chambers at seven constant temperatures, 17.5, 20.0, 22.5, 25.0, 27.5, 30.0 and 32.5°C, at 14:10 (L:D) photoperiod. Development and survivorship were observed and recorded every six hours

PAGE 48

32 until adults emerged. Initial wasp cohorts consisted of 29, 34, 36, 38, 38, 39, 39 individuals respectively per temperature studied. The relationship between temperature and developmental rate (=l/days) was estimated by linear regression analyses. Pooled developmental times for males and females were used to fit the linear models. The lower developmental threshold (t), the temperature at which development is negligible, was estimated by the "x-intercept" method. Values of / were obtained as follows: t = -a/b, where a and b are parameters from the regression equations (where x = temperature, and y = developmental rate). The degree-days required for development were calculated by using the equation DD=y(T-0 where 'y' is the developmental time (days), 'T' is the temperature during development, and 7' is the lower developmental threshold. DD were determined for each individual at all temperatures where development was recorded and then averaged by stage. Life-Fertility Tables Reproduction of two generations developing during different seasons of the year was studied. The summer cohort. This consisted of wasps collected during July. Wasps collected from the field were provided with host mole crickets and their parasitic progeny were kept in the laboratory at 26 ± 0.5°C, 65% RH and 14:10 (L:D) photoperiod. Emerging adults were confined in cages, grouped by age. Adult wasps were fed with diluted hummingbird nectar and honey. The cages were additionally provided with a blooming nectar-source plant (Spermacoce verticillata). Summer wasps were given host mole crickets starting a day after emergence, 5 to 7 times a day until the wasp's death.

PAGE 49

33 Parasitized hosts were kept under the same laboratory conditions as the adults, until wasp stages completed development. The autumn cohort. This consisted of wasps from parasitized mole crickets that were kept under field conditions until parasitic stages completed development. Emerged wasp adults during October were subjected to tests. Wasp adults were provided with host mole crickets as for summer wasps. Parasitized hosts were confined underground in screened chambers of 30 cm depth until adult emergence. Reproductive statistics were summarized in the form of age-specific life fertility tables for both cohorts. Results Development Developmental times for all stages were inversely related to temperatures from 25.0 to 30.0°C (Figure 2). The temperature of 32.5°C caused a varied effect on the stages of this parasitoid; the egg stage maintained a similar reduction in development compared with previous temperatures, larva and prepupa stages reached a plateau in development compared with previous temperature (30.0°C), and pupa stage slowed down development as almost four times the previous temperature (Table 2). No development was observed at 17.5 and 20.0°C; only one wasp developed at 22.5°C. The total development ranged from 142.90 to 40.18 days at 32.5 and 30°C, respectively. Apparently, the upper temperature threshold was reached at 32.5°C. Pupal development required from 72.5 to 91% of the total development time. Although the duration of stages varied with temperature, the proportion of total developmental time spent in each stage was consistent regardless of temperature.

PAGE 50

34 Table 2. Development (days) of Larra bicolor life stages 3 at constant temperatures. Temp. ( °C) n Egg n Larva n Prepupa n Pupa n Total 32 5 4 3.23 ± 0.09 (2.3) 11 7.33 ± 0.20 (5.4 ) 4 0.47 ± 0.12 (0.33) 4 130.02 ±17. 38 (91.0) 4 142.90 ± 17.60 30.0 14 3.75 ±0.28 (8.6) 7 7.38 ±0.01 (16.6) 8 0.44 ± 0.02 (0.01) 8 35.07 ± 0.69 (75.3) 8 47.50 ± 0.63 27.5 8 4.74 ± 0.63 (9.3) 8 9.87 ±0.15 (18.4) 6 0.66 ±0.09 (0.11) 6 41.39 ± 0.80(72 5) 6 57 06± 0.95 25.0 11 5 09 ± 0.28 (6.8) 4 13.30 ±0.40 (17.2) 9 0.72 ±0.06 (0.01) 9 62.30 ± 3.76 (75 9) 9 82.10 ± 3.76 22 5 1 8.63 ± 0.00 (7.8) 1 18.13 ±0.00 (18.2) 1 2.00 ± 0.00 (0.02) 1 81.88 ±0.00 (74.0) 1 110.63 ± 0.00 a Mean ± SEM; percentage of developmental time required for each life stage given in parenthesis.

PAGE 51

35 The survivorship per stage is shown in Figure 3 and described according to Southwood (1978). The maximum survivorship occurred at 30°C followed by 25 and 27.5°C. Survivorship of cohorts reared at 17.5 and 20.0°C followed a type IV curve (100% mortality occurs prior to the adult stage). Cohorts reared at 22.5 and 32.5°C exhibited a type III curve. The remaining cohorts followed a type I curve (mortality acts most heavily on the old individuals). Egg Larva PP Pupa 0 100 200 300 400 500 600 Physiological time (Degree-Days > 16.0°C) Figure 3. Survivorship of Larra bicolor life stages at constant temperatures Linear regression parameters describing the relationship between developmental rate and temperature for all stages are presented in Table 3. Data for 22.5°C were excluded because only one wasp completed development at this temperature. The data at 32.5°C were not used in the regression model for larva, prepupa and pupa because the values were considered beyond the linear portion of growth responsive to temperature (Sharpe et al. 1977), (Figure 4). Coefficients of determination ranged from 0.71 (prepupa) to 0.94 (larva), indicating a good fit in all cases. At the egg stage, data for 32.5°C were

PAGE 52

36 included, but for all remaining stages data were used for temperatures from 25.0 to 30.0°C, because the upper developmental threshold was apparently reached. Table 3. Linear regression models (y = developmental rate, x = temperature), lower developmental thresholds (t), and degree-days required for Larra bicolor life stages and total to develop at constant temperatures. Stage y interceptfcSEM Slope±SEM R 2 ±SE t(°C) Degree-days (DD±SE) Egg a n = 31 -0.4015+0.0348 0.022110.0012 0.92±0.0182 18.143 45.446+0.7107 Larva" n = 27 -0.2211+0.0159 0.0119+0.00058 0.94±0.0063 18.620 84.425±0.9331 P-P b n = 27 -3.7870+0.7173 0.2032±0.0263 0.71+0.2838 18.640 5.049±0.1652 Pupa" n = 27 -0.0389±0.0048 0.0024±0.00018 0.88+0.0019 16.01 413.135±5.642 Total" n = 27 -0.03126+0.00276 0.0018+0.0001 0.93±0.001 16.880 540.84±6.035 P-p = prepupa; a 25.0, 27.5, 30.0 and 32.5°C data only; b 25.0, 27.7, and 30.0°C data only. The lower developmental threshold temperature (/) obtained by the "x-intercept" method ranged from 16.01 (pupa) to 18.64 (prepupa). In fact, "f values for all single stages were close to 18 except for the pupa, and then for total development. The degreedays (M ± SEM) above their respective "f , required by the wasp to develop during different life stages are presented in Table 3. The degree-days for the pupal stage to complete development account for almost 76% of the degree-days required for total development.

PAGE 53

37 0.0 — 22.5 25.0 27.5 30.0 32.5 35.0 Temperature °C Figure 4. Developmental rates of Larra bicolor at constant temperatures. Symbols show observed data. Line trend fitted to the linear response. Reproduction Total fecundity (eggs laid per female) was recorded from the two cohorts, summer and autumn wasps (Table 4). Summer wasps oviposited 16 eggs per female, in contrast with 12.5 from autumn wasps. Despite the apparent difference, the means are not statistically different. Egg hatching was statistically the same, with values close to 100% hatching. The female longevity was almost the same in the two cohorts, with a female adult survival of 1 0 days.

PAGE 54

38 Table 4. Population reproductive statistics for Larra bicolor from two seasons in Gainesville, FL (29°40'N). Season n females Fecundity (eggs per female) ab Fertility (% egg hatch) Female longvity (days) a * Summer 0 Autumn d 7 4 16.0±2.23 a 12.5±2.84 a 98.2 b 96.4 b 10.14±1.20 a 10.50±3.17 a a Mean±SEM b Means followed by different letters are significantly different. c Lab-reared wasps at 26 ± 0.5°C. d Field-reared wasps, emerged during October. Life-FertilityTables Life fertility tables were constructed for all cohorts to calculate reproductive statistics (Table 5). The net reproductive rate (Ro), which represents the females produced per female per generation of the two populations, overlaps the confidence limits, which makes them statistically not different. The intrinsic rates of increase (Rm) did not overlap each other, meaning that the two populations increase at different rates, with autumn wasps having the slower population growth and being statistically different. The mean generation times (Tc) did not overlap between the two cohorts. The differences observed in Rm and Tc are explained by the fact that the autumn generation developed under lower temperatures than did the summer generation. These differences are shown by the oviposition periods in Figure 5, where the effect of contrasting temperatures caused delay of oviposition in the autumnal population, and faster oviposition time in the summer population.

PAGE 55

39 Table 5. Reproductive parameters for Larra bicolor from two seasons in Gainesville, FL (29°40'N). Season Ro ad Rm bd Tc (days) cd Summer 6 1.76±0.61 0.011+0.006 51.14±5.29 Autumn f 1.37±1.00 0.003±0.007 109.47±18.30 a Ro = net reproductive rate b Rm = intrinsic rate of increase c Tc = mean generation time d Mean±SEM e Lab-reared wasps at 26 ± 0.5°C. f Field-reared wasps, emerged during October. Discussion Development of the digger wasp Larra bicolor was characteristic of poikilothermic organisms where developmental time decreases (and developmental rate increases) as temperature increases (Tables 4 and 5, and Figure 4) (Sharpe and DeMichele 1977, Wagner et al. 1984a). The linear developmental rates occurred at temperatures from 25.0 to 30°C for all stages but the egg, where the developmental rate kept its linearity up to the highest temperature of 32.5°C (Figure 4). The upper temperature threshold was passed at 32.5°C for all stages but egg, so that the upper threshold must be at some point between 30 and 32.5°C. Castner (1988) reared the wasps at 26 ±2 °C and 65% relative humidity, showing a broadest developmental variability for this parasitoid compared with the data reported in the present study where the wasps were reared at temperatures with an accuracy of ± 0.5°C which explains the lower variability in developmental times of the wasp. The lower threshold value observed in the pupal stage along with the higher number of degree-days required for the pupal stage to develop (almost 76% of the total

PAGE 56

40 development) is a strong indicator that this is the resistant stage to inappropriate environmental temperatures at both extremes (Table 3 and Figure 3). It is important to note that the developmental data presented in this study were obtained from wasps developing during summer and exposed to controlled summer-like conditions. The developmental rates under autumn-like conditions at which this parasitoid is able to develop were not studied, and remain unknown. L. bicolor developed and survived over a broader temperature range (Table 2) than normally occurs underground during spring to fall seasons in Gainesville. The host mole cricket, S. borellii, tolerated extreme temperatures during the study, but the parasitic wasp failed to develop at 17.5 and 20.0°C, and only one wasp completed development at 22.5°C. In previous studies (not presented in this dissertation) rearing the three mole cricket species, none survived more than 15 days at temperatures of 15 and 10°C nor at 35°C or higher. Those temperature ranges must be outside the limits for these hosts to survive and develop. Temperature regimes, representing contrasting conditions at which the two cohorts were reared, did not affect reproductive statistics of the parasitoid, considering fecundity, fertility and adult longevity even though the population reproductive parameters estimated from the life-fertility tables (Table 5) showed differences in the intrinsic rate of increase and in generational time, and then in oviposition time (Figure 5). The temperature regime represented by the season of the year did not alter the reproductive capacity of Larra bicolor. From these results it can be inferred that the temperature regime (say the season of the year) is not an important constraint of the reproductive performance of this parasitoid, at least in Florida. Other factors such as

PAGE 57

41 suitable hosts and availability of food source for adult wasps should play the more important role restraining the populational increase of Larra bicolor. The developmental time was defined per stage and total life cycle. The lower threshold was determined for all the stages. The upper temperature threshold was reached at 32.5 for all stages but the egg. The degree-day values were estimated for all stages. The reproductive parameters of wasps estimated at summer-like temperature (constant temperatures) were similar to those reared in the field (autumn season). Autumn population N=5 S4 o n E ^2 Summer population N = 8 ! ( '.'Hi 11 'J ill i 50 100 Age (days after emergence) 150 200 Figure 5. Oviposition time of wasps Larra bicolor reared at summer-like and autumnal seasons in Gainesville, FL; daily oviposition per wasp is represented by different line pattern per season.

PAGE 58

42 Development of the wasp Larra bicolor at field-varying temperatures. Introduction. Studies of the sphecid ectoparasitoid Larra bicolor have been devoted to its behavioral ecology and foraging (Fowler 1989), host searching and mate-searching (Castner and Fowler 1987), and establishment and dispersal (Frank et al. 1995). The only information available about the life cycle and field efficiency was obtained by Castner (1988) with the Puerto Rican biotype that established a population in south Florida but failed to do so in central and north Florida. Practically no information about the Bolivian biotype is available. Extensive data are needed to examine the effect on host mole cricket populations, with geographical, seasonal, and annual variation. Temperature plays a critical role in determining the rate of development, survival and reproduction of host insect species. The duration of development from birth to first reproduction is one of the most important insect life-history characters influencing numerical change in insect populations (Wagner et al. 1984a). Because the rates of development in insects under natural conditions are largely determined by temperature, a realistic measure of the relationship between temperature and development rate is fundamental to any study of the population dynamics or of the management of any insect (Wagner et al. 1984a, Liu and Meng 1999). The degree-day concept is one of the oldest and most widely used entomological models; it assumes a linear relationship between development rate and constant temperatures, therefore the number of degree-days required to complete development is assumed constant at all favorable temperatures (Wagner et al. 1985). Daily temperature cycles that occur in the field are generally not considered by the degree-day method that uses data collected at constant temperatures in the laboratory to predict developmental times in the field (Higley et al. 1986).

PAGE 59

43 Developmental times are known to differ between constant and fluctuating temperatures with the same mean (Hagstrum and Hagstrum 1970). Developmental time data collected at constant temperatures in the laboratory can only be expected to provide a rough estimate of developmental times in the field at fluctuating temperatures (Hagstrum and Milliken 1991). The relationship of life stages of Larra bicolor to constant temperatures is shown in Table 2, resulting in the proposal of a degree-day model (Table 3). The model, based upon constant temperatures, must be evaluated by its goodness of fit to populations developing in the field. This work describes the development of egg, larval, prepupal, and pupal stages of Larra bicolor, Bolivian biotype, under field conditions of Gainesville, FL (29°40' N). The results of this work will be used, in a subsequent study, to evaluate the efficiency of the degree-day model proposed. Materials and Methods Female wasps were captured by net from patches of the food-plant Spermacoce verticillata L. in Gainesville, FL. After capture, the wasps were placed individually in snap-cap plastic vials, and kept alive by cooling them in a chest. The wasps were taken the same day to the laboratory and placed in groups of 10 wasps into a 30x30x12 cm metal-framed screened cage containing a flowering S. verticillata. Water, diluted honey, and diluted artificial hummingbird nectar (®Perky-Pet Products Co.) were supplied also. Unless otherwise indicated the host mole crickets were Scapteriscus borellii adults that were captured in traps baited with synthetic mole crickets song at the Ona REC, FL. The hosts were shipped in buckets partly filled with moist vermiculite.

PAGE 60

44 Devices made of PVC pipe were designed to individually confine parasitized hosts underground at a range of depths. These pipe-chambers could be easily buried vertically and are could be easily removed from the ground with little disturbance of mole cricket tunnels inside them. The pipe-chambers were subdivided into four sections with the assumption that each section represented a particular thermal regime and the total individuals confined at each depth can be considered as part of an independent experiment. Thus, it was possible to handle four independent experiments at the same time, covering different possible thermal regimes at which the host mole crickets naturally live and thus the parasitoids develop. The depths of the four chamber sections were: 0-15, 15-30, 30-45, and 45-60 cm. (Table 6). The pipe-chambers consisted of 5 cm inside diameter x 75 cm long PVC pipes cut lengthwise. Two 2.5 cm wide x 25 cm long windows were cut at the center of each half, covered with aluminum window screen, and sealed with hot melted silicone. The inside of each pipe was subdivided into 1 5 cm sub-chambers, by placing non-corrosive metallic mesh, fully matching the inner diameter of the pipe once closed, also fixed with hot melted silicone. Starting from the bottom, the pipe-chambers were subdivided into four sub-chambers (15 cm tall x 5 cm diameter), leaving a portion of 15 cm intended to be above the ground level for handling purposes. This chamber design allows the host mole cricket to move freely in a volume of 300 ml of sand. The screened windows allow the sand inside the chambers to be in close contact with the sand outside, allowing the flux of air and moisture from the external soil into the chambers. The temperature was monitored inside and outside the chambers at the different depths, with no difference observed.

PAGE 61

45 Moisture was not measured but was estimated by touch; tap water was added when considered necessary. Each half chamber was filled with autoclaved moist sand, lightly compressed. Artificial tunnels were made by compressing the moist sand by finger. The halves of the chamber were joined as a sandwich by placing rubber bands at both extremes and at the middle. To check the status of host or parasitoid, or to feed the host, the pipes were withdrawn from the soil, the rubber bands were removed, and each half of the pipechamber was rolled gently to a concave position to avoid disturbance of the contents. Table 6. Codes and depths of sub-chambers and thermocouples where parasitized mole crickets were confined to study the development of Larra bicolor underground in the field. Sub-chambers Thermocouple Code Depth (cm) Depths (cm) A 0-15 7.5 B 15-30 22.5 C 30-45 37.5 D 45 -60 52.5 The land where the pipe-chambers were buried was kept free of weeds for the duration of the experiments. It was then assumed that the irradiation, aeration, and moisture (from environmental humidity and precipitation) were evenly distributed without interference from vegetation. The presence of patches of weeds would have modified the flux of these factors. A rectangle of land, lm wide x 6m long, located near the Entomology and Nematology Department building, was tilled to 75 cm depth to allow the manipulation of the pipe-chambers in and out of the ground. Before introducing the parasitized hosts, the pipe-chambers were buried underground for a few days. The

PAGE 62

46 chambers were symmetrically distributed, allotting five chambers to the width and 20 chambers to the length of the experimental land. Parasitized hosts were randomly allotted to the sub-chambers. The soil temperature at which the parasitized host was confined in the subchambers was recorded by placing a thermocouple in the middle of the corresponding sub-chamber (Table 6). Temperatures were recorded as follows using one pipe-chamber buried in the center of the experimental land. Four thermocouples were used, one at the depth of each of the four pipe sub-chambers. The thermocouples were connected to a digital temperature datalogger (®HOBO Mod H8-4, Onset Co.) programmed to record temperatures every 30 min. The temperature recorded was downloaded to a computer for further analysis. Another digital temperature datalogger was used to record air temperature, placing the thermocouple at 1 .8 m height above ground under a 20x20 cm shed roof. The parasitism of hosts was achieved following the procedure described by Williams (1928) cited by Castner (1988). Placing a female wasp into a 200-ml plastic vial, and then introducing the host S. borellii always resulted in immediate parasitism. Wasp activity was stimulated by adding a heat source with a 500 w incandescent lamp for a few minutes at 30°C before offering a host. Maintaining the wasps at a temperature of 27°C was observed to increase their feeding activity; otherwise most of the wasps remained hidden in the plant or in the corners of the cage. A few minutes after they were parasitized, the hosts were placed in moist sand inside the pipe-chambers and buried underground. The parasitized hosts were fed twice a

PAGE 63

47 week with the same Purina cricket chow food used in the laboratory culture. To prevent hosts dying of dehydration, tap water was added if the sand in the chambers seemed dry. Four wasp groups captured from the field were used to parasitize host mole crickets. Parasitism of all hosts was achieved in as few days as possible to reduce variability. Each group of parasitized hosts was considered a cohort (Table 7). Table 7. Description of the cohorts of the wasp-parasitized hosts by numbers and dates. Cohorts Parasitized hosts 1 Parasitism dates No. No Initial Last 1 65 22-VIII-2000 31-VIII-2000 2 38 1-IX-2000 11-IX-2000 3 28 30-IX-2000 1-X-2000 4 20 27-X-2000 Scapteriscus borellii adults. Hosts of cohort 1 were randomly allotted to the first three sub-chambers (A, B and C). Hosts of cohorts 2 to 4 were randomly allotted to the four sub-chambers (A, B, C, and D). The response variable was the time to develop for egg, larval, prepupal and pupal stages. The observation frequency varied as follows: egg development was recorded every six hours starting 24 hours after oviposition. Larval development was recorded daily starting five days after egg eclosion, continuing until the wasp grub reached a large size, or until the parasitic grub killed the host, then recordings were made every six hours until the beginning of the prepupal stage. Prepupal development was recorded every six hours until pupation was completed. Once pupation was completed, the metallic mesh dividing the sub-chambers was removed to allow the emerging wasps to dig up to the

PAGE 64

48 surface. Pupal development was recorded daily, starting 20 days after pupation occurred, until adult emergence. Eventually, tardy pupae were visually inspected to assess whether they were still intact and to ensure the wasp had not escaped. Adult emergence was recorded by the presence of the adult wasp trying to escape from the chambers, but impeded by the fiberglass window-screen at the top of the pipe-chamber; daily inspection of the chambers was made during the experiments. Results The developmental times for stages and cohorts are presented, sorted by subchambers, stages and total development in Table 8, where the egg stage showed no difference among the sub-chamber depths. The larval stage showed no clear differences in development in cohorts 1 and 2, and no differences in cohort 3. The prepupal stage showed no difference in development related to sub-chamber depth in all the three cohorts evaluated. The pupal stage showed no difference among the sub-chambers where comparisons were possible. The total development showed statistical differences in cohorts 1 and 2. The differences observed in egg, larval and pupal stages were not highly significant. The mean values recorded in pupal and total development were numerically more contrasting among sub-chambers. In order to explain the possible differences in development among cohorts, the developmental times were pooled by cohort (without considering the sub-chamber) and presented in Table 9. The egg development was different among the three cohorts. Larval development in cohort 1 was highly different from that in cohorts 2 and 3. The prepupal stage showed no difference among cohorts. Pupal and total development showed no statistical difference among cohorts even though the mean values were very distinct, particularly in cohort 1, which was three times less than in cohort 3. The lack of differences in the pupal and total development is highly

PAGE 65

49 related to the high variability of the respective estimates, which may mask the possible differences. Table 8. Development time of Larra bicolor at various depths underground in the field at various dates in Gainesville, FL (29°40'N). Cohort 1 Egg ab Larva 3 " Prepupa 3 " Pupa 3 " Total 3 '" Depths n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM A B C 11 8 10 3.81 ±0.12 a 3.65 ±0.11 a 3.60 ±0.14 a 10 7 9 9.18 ±0.28 a 8.85 ± 0.27 ab 8.17 ± 0.19 b 11 9 13 0.78 ± 0.09 a 0.79 ± 0.09 a 0.87 ± 0.09 a 2 2 4 44.23 ± 3.3 a 45.31 ± 3.6 a 104.95 ±62.9 a 2 2 4 58.42 ± 3.4 a 57.89 ± 3.9 a 1 1 7.32 ± 62.8 b Cohort 2 Egg"'" Larva 3 " Prepupa 3 " Pupa ab Total 3 '" Depths n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM A B C D 10 12 4 3 3.33 ±0.1 a 3.36 ±0.1 a 3.33 ± 0.1 a 3.56 ± 0.2 a 2 3 1 2 11.02 ±0.69 a 11.18±0.26a 9.96 ± 0.00 11.40 ± 1.94a 1 3 1 2 0.58 ±0.0 0.60 ± 0.1 a 0.50 ± 0.0 0.67 ± 0.2 a 0 1 0 2 238.00 ± 0.0 a 160.89 ±69.9 a 0 3 1 1 246.81 ± 3.83 247.71 ± 0.00 112.92 ±0.00 Cohort 3 Egg*'" Larva 3 " Prepupa 3 " Pupa 3 " Total 3 " Depths n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM A B C D 6 2 4 2 3.89 ± 0.04 a 3.96 ± 0.37 a 4.02 ±0.11 a 3.94 ± 0.39 a 6 3 4 2 14.50 ±0.43 a 12.90 ±0.49 a 13.38 ±0.38 a 13.06 ±1.02 a 5 2 4 2 0.87 ± 0.07 a 0.79 ±0.21 ab 0.55 ±0.03 b 0.54 ±0.08 b 0 0 2 0 210.27 ±10.27 0 0 2 0 227.33 ± 10.46 Cohort 4 Egg a " Larva 3 " Prepupa 3 " Pupa 3 " Total 3 " Depths n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM A B C D 0 0 2 0 198.75 ± 6.00 a Mean days to complete development ± Standard error. b Different letters among depths indicate statistical difference; Duncan's test when three or more means; 't' test when two means. Table 9. Developmental time of four cohorts of Larra bicolor stages and total, averaged over all depths underground in the field in Gainesville, FL (29°40'N). Egg 3 " Larva 3 " Prepupa 3 " Pupa 3 " Total 3 " Cohorts n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM n Mean ± SEM 1 28 3.69 ± 0.07 b 26 8.74 ± 0.17 b 33 0.83 ±0.05 a 8 74.86 ±31.28 a 8 87.74±31.17a 2 29 3.35 ± 0.06 c 8 11.04 ±0.43 a 7 0.60 ± 0.06 a 3 186.60 ± 45.44 a 5 220.21± 26.91 a 3 14 3.95 ± 0.07 a 15 13.10 ±0.74 a 13 0.71 ± 0.06 a 2 210.27 ± 10.27 a 2 227.33± 10.46 a 4 ne ne ne ne 2 198.75 ± 6.00 a a Mean days to complete development ± Standard error. Different letters among cohorts indicate statistical difference; Duncan's test when three or more means, ne = not evaluated.

PAGE 66

50 The mean total developmental values per cohort (from Tables 8 and 9) are grouped in Table 10, where three developmental time groups are proposed, naming them as short, intermediate and long developmental times. Cohort 1, the summer wasps, produced all of the short total development, 56.59 d (all emerged by 18-24 Oct.), and one wasp that shifted to the long development, 305.8 d (emerged by 27 May). In Cohort 2, one wasp completed development in an intermediate time, 1 12.91 d (emerged by 30 Dec), and 4 completed development in the long time, 247 d (all emerged by 8-27 May). Wasps from cohort 3 produced 2 individuals that developed in the long time, 227.3 d (emerged by 7 and 27 May). Cohort 4 produced 2 wasps that developed in the long time, 198.5d (emerged 9 and 21 May). The duration of the total development is presented with 99% confidence limits, with the idea to stress that the proposed groups do not overlap each other. Those limits demonstrate the pertinence of grouping them in those three categories that proves they are statistically different (Table 10). Table 10. Trimodal total developmental time of Larra bicolor per cohort underground in the field in Gainesville, FL (29°40'N). Short 3 '" Intermediate 3 Long 3 Cohorts n Mean ± SEM n Mean ± SEM n Mean ± SEM 1 7 56.59 ±1.43 1 305.79 ± 0.00 2 1 112.91 ±0.00 4 247.03 ± 2.72 3 2 227.33 ±10.46 4 2 198.75 ± 6.00 a Mean days to complete development ± SEM. b 99%CL = 4.32d (52.27-60.91). Table 1 1 presents the individuals grouped by their corresponding life cycle stages and total developmental times, noting statistical differences between larval, pupal and total development, among the three life cycle developments.

PAGE 67

51 Table 11. Trimodal life cycle developmental time of Larra bicolor underground in field conditions of Gainesville, FL (29°40'N). Shorf bc Intermedial*" c Loncf bc Stage n Mean ± SEM n Mean ± SEM n Mean ± SEN Egg 7 3.49 ± 0.39 a 1 3.1 2 ±0.00 a 7 3.49± 0.32 £ Larva 7 8.86 ± 0.84 a 1 13.33 ± 0.00 b 5 1 1.66 ± 3.19 al Prepupa 7 0.83 ± 0.31 a 1 0.45 ±0.00 a 5 0.64± 0.25 £ Pupa 7 43.59 ± 3.32 a 1 96.00 ± 0.00 b 5 235.61 ±43.78 Total 7 56.59 ± 3.49 a 1 112.91 ± 0.00 b 9 238.45± 25.21 a Mean days to complete development. different letters between cycles indicate statistical difference. Duncan's test when three or more means °Only wasps that completed the life cycle were included. Survivorship The number of wasp parasitic individuals that survived through the life cycle stages and their corresponding adult sexes per cohort are presented in Table 12. An overall 58% of the wasps reached the prepupal stage, which is an indicator that the developing larva killed the host mole cricket. An overall 47 to 49% of the wasps reached the pupal stage. An overall 13 to 14% of wasps completed development to the adult stage, with an overall sex ratio not statistically different from 50% (X test). Table 12. Number of individuals, corresponding proportion and sex ratio per cohort and overall of Larra bicolor developing on Scapteriscus borellii underground at field varying-temperatures . Cohort Eggs Larvae Prepupae 1 Pupae 1 Adults' Sexes 1 65 45 (69.2) 40 (61.5) 33 (50.8) 10(15.4) 5M3F 2 38 34 (89.5) 17(44.7) 16(42.1) 5 (13.2) 3M 2F 3 28 19(67.9) 19(67.9) 15(53.6) 2 (7.1) 2F 4 20 ne ne 7 (35.0) 2 (10.0) 2F Overall 2 131 98 (74.8) 76 (58.0) 64 (48.8) 17(13.0) 8M 7F Overall 3 1^ . . 151 71 (47.0) 19(14.5) 8M 9F Total number of individuals (percentage inside of parentheses). Considering cohorts 1 3 only. 3 Considering all 4 cohorts, ne = not evaluated.

PAGE 68

52 The proportion of individuals entering the next stage is presented in Figure 6, where the declining populations followed the type III curve explained by Southwood (1978), where mortality occurs gradually by stage, being similar among the three cohorts. Eggs Larvae Prepupae Pupae Adults Life cycle stages Figure 6. Proportion of individuals from three cohorts entering the life cycle stages of Larra bicolor developing on Scapteriscus borellii underground at field-varying temperatures. The overall observed sex ratio in this work (which coincides with the observed ratio from constant temperature experiments) is somewhat intriguing, because during the three wasp seasons I scouted for wasps at patches of the nectar-plant S. verticillata (which included various times of the day and various periods of the year) the sex ratio of feeding wasps was always male-biased, most commonly seen as 3 to 10 males per one female, but sometimes reaching the exorbitant ratio of 100:1 M:F, or higher. Discussion The development of egg, larval, prepupal and pupal stages and thus the total development of the wasp Larra bicolor were evaluated under field conditions underground at three or four depths.

PAGE 69

53 The mean developmental times of each cohort at the various depths showed few differences, and in most of the cases there was no difference (Table 8). A comparison of the development of the cohorts showed statistical differences in egg and larval development among cohorts. No differences in prepupal and pupal development were seen (Table 9). Development from egg to prepupa did not vary according to depth in the soil (Table 8). The pupal stage was the most sensitive to seasonal changes (shown by the four cohorts that developed at various dates), showing three distinct durations of development time (Tables 10 and 12). The pupal (thus the total) development was classified as short, intermediate and long. The capability of the pupal stage to diapause confers on this parasitoid the opportunity to survive the winter when low temperatures may occur and when the food plant, Spermacoce verticillata, may freeze to the ground. Most of the parasitoids that showed long life cycle emerged early and middle May, coinciding with two main factors that favor the parasitoid' s survival: a) the food plant starts blooming in early May. This information is based on personal observations made during 1998 to 2001 b) the mole cricket populations are composed of old adults and small and small-medium size mole cricket nymphs. Adult and medium-large nymph host sizes are attacked by this wasp parasitoid during late May, as shown in Appendix B. The life cycle duration of Larra bicolor was observed to include three distinct development times, where the pupal stage accounted for the variation observed in life cycles as follow: a short duration of 57 d (646 DD), an intermediate duration of 1 13 d (838 DD), and a long duration of 238 d (1093 DD), (see DD calculation on page 32, and

PAGE 70

54 sample size on Table 10). The short cycle occurred in August. A possible seasonal strategy is proposed here to account for ability of the wasp to survive under the environmental condition of Gainesville, FL (29°40'): Overwintered wasps. These wasps spend the winter as pupae from which adults begin to emerge in May. In May, these newly-emerged adults will find the preferred nectarsource plant (Spermacoce verticillata) blooming (in the localities where it occurs), providing energy to the wasps for mating, and for the females to hunt for host mole crickets. In May, these females will encounter overwintered Scapteriscus vicinus and S. borellii mole crickets that will be in the adult stage, providing suitable hosts, although by May they will not be especially abundant because they are undergoing attrition by aging. First new generation. If the progeny of the overwintered generation of wasps (1. above) complete their life cycle: i. In a short time (57 d), the adults will begin to emerge in July, a time when very few hosts of suitable size are available. This would be a poor strategy. ii. In an intermediate time (113 d), they will begin to emerge in September, a time when the new generation of mole crickets exists as medium-sized to large-sized nymphs and is abundant (contrasted with numbers of survivors in May). Undersized hosts are not ideal because they may provide inadequate nutrition for wasp larvae. The longer a female wasp can survive, the greater is her opportunity for encountering large hosts. This appears to be the best strategy. iii. In a long time (238 d), they will begin to emerge in December, which risks encountering freezing temperatures.

PAGE 71

55 The optimal strategy for the progeny of the overwintered wasps (first new generation) is a development time of intermediate length (113 d). Second new generation. For those wasps (first new generation) whose adults emerged in September, the strategies are to develop: i. In a short time (57d), to provide a second new generation with adults emerging in November. This strategy would be highly successful except in years when an early freeze occurs. ii. In an intermediate time (1 13 d), to give rise to adults emerging in JanuaryFebruary. This would be a disastrous strategy, because emerging adults would have little or no access to nectar. iii. In a long time (238 d), to give rise to adults emerging in May of the following year. This would be successful. That three distinct developmental times of L. bicolor should have been detected is remarkable. All three times (short, intermediate, and long) are needed to optimize its adaptation to the climate of northern Florida. The ability to adapt to the climate of northern Florida must have arisen during the evolution of this species in South America, where the species is widespread. The ability of the wasp to switch between these strategies is a tribute to its phenotypic plasticity. A simpler strategy would have been one with a single annual generation matched to the maximal annual abundance of hosts of suitable size: in other words, an annual cycle with presence of adult wasps only in September-November would have been easier. Two types of questions are now presented. The basic question is: How is the plasticity of developmental time triggered and maintained? The applied question is: How

PAGE 72

56 far north in the USA can L. bicolor establish a population in the presence of Scapteriscus mole crickets, and will it adopt a life cycle of a single annual generation at its northern limits (as contrasted with the three generations it supports in northern Florida)? It seems obvious that a single annual generation of this wasp would have less effect on mole cricket populations than do the three generations that seem evident in northern Florida. In the tropics, in areas where Scapteriscus mole crickets breed continuously, L. bicolor might likewise breed continuously, so that the short (57 d) life cycle might maximize its reproductive strategy. In more temperate areas, other kinds of life cycles, adapted to breeding periods of Scapteriscus mole crickets and to availability of nectar, are surely more appropriate. At the extreme northern (and southern!) limit of the range of this wasp, wherever that may be, a single annual generation may be most appropriate to maximize its population size. In the intermediate (subtropical and almost-temperate) areas, other strategies may be better to maximize population size. This wasp, L. bicolor, is exceptional in having the ability to adapt its development time to maximize its population size. This is the first work describing the life cycle of Larra bicolor under field conditions. The results present an insight into the performance of this parasitoid in adapting its populations to survive the climate of north Florida (29°40'N). The data presented support the hypothesized cold hardiness of this strain and reaffirm the potential of this parasitoid to establish and disperse its populations at this latitude and perhaps beyond toward other states where Scapteriscus mole crickets are already established.

PAGE 73

57 Degree-day model predicting the development of the wasp Larra bicolor Introduction The rate of insect development depends upon the temperature to which the insects are exposed. For a developing insect, the rate at which time passes depends on the combination of its temperature-dependent development rate curve and the temperature at which it lives. The development rate curve may be altered by many factors of the environment, such as humidity, nutrition, or disease, but the curve always remains the same (Taylor 1981, Vinson and Iwantsch 1980). For each species, the temperature below that at which no measurable development occurs is the threshold of development. The degree-day is a model that has been in use for more than 250 years (Wagner et al. 1984b) describing temperature/development relationship in the species (Honek 1999). The degree-day model is a linear approximation that enables calculating two virtual constants, lower threshold and sum of effective temperatures (an amount of heat needed for completing a development stage). This model fit the development rates within the range of ecologically relevant temperatures (Honek 1999, Campbell et al. 1974). The degree-day model does not work for all species. Hagstrum and Milliken (1991), Tanigoshi et al. (1976) and Logan et al. (1976) indicate this model is particularly invalid for mite populations. Wagner et al. (1984, and 1985) reviewed the efficiency of using constant temperatures to model development at variable temperatures. Hagstrum and Hagstrum (1970), Manel and Debouzie (1997b), showed how variable the results can be. Liu and Meng (1999) indicates that the degree-day model can simulate the development of Myzus persicae (Sulzer) very well when temperature does not go below 4°C. The degree-day model also works for Cydia pomonella (L.) (Howell and Neven 2000).

PAGE 74

58 Wagner et al. (1984b) report that insects reared under identical conditions develop at different rates due to inherent differences among individuals in a population. Hagstrum and Hagstrum (1970) demonstrate that numerous insects develop at much lower temperatures when fluctuating diurnally than at a comparable mean constant temperature. Understanding the relationship between temperature and development is important in constructing computer models for simulating biological control of pest populations (Lysyk 2000). Scapteriscus mole crickets are pests of pastureand turf-grasses and vegetables in the southeast and are dispersing westward in the USA. They are soil-inhabiting insects, reaching the ground surface for dispersal and mating purposes. Their life cycles are described in Chapter 2. The parasitoid wasp Larra bicolor is a natural enemy of Scapteriscus mole crickets, and has been released in the USA. The objective of this work was to evaluate the degree-day model based upon constant temperatures to predict the field development of Larra bicolor parasitizing mole crickets in north Florida (29°40'N). Materials and Methods A set of experiments quantifying developmental time of Larra bicolor per stage and total was performed at seven constant temperatures in incubation chambers. The developmental time of L. bicolor at fieldvarying temperatures was determined by studying the development of three wasp cohorts underground during summer, autumn and winter seasons. The parasitic wasp developed on Scapteriscus borellii and temperatures were recorded every 30 mitt at four soil depths.

PAGE 75

59 Developmental rates of L. bicolor are calculated using the linear equation (Eq. 1) plotted against temperature. The temperatures at the most linear portion of the curve were selected to calculate their linear equations by simple linear regression analysis (Eq. 1). Y = a + b(x): (Eq.l) Y = Developmental rate (1/days) a = Intercept b = Slope x = Temperature °C The lower temperature threshold and thermal constant (degree-days) were calculated from the linear equation, using its regression coefficients where: lower threshold (to) = -(a/b) and thermal constant (k) = 1/b. The degree-days required per individual and total to complete development at constant and field temperatures are calculated by the following equation: Y = tH'-'o) (Eq-2) Y = Degree-days T = Time (time step = 30 minutes = 1/48 days) t = Temperature recorded every 30 min to = Lower temperature threshold (per corresponding stage or total) n = Number of days for stage to develop The validation of development of L, bicolor at constant temperatures as a predictor of development at field temperatures is evaluated by comparing degree-day data sets in the same graph to observe how well the degree-days at constant temperatures match degree-days at field varying temperatures. The statistical comparison of what can be seen from the constant and field temperature linear equations was done by comparison of slopes. The linear equations were calculated by regressing the observed DD with corresponding mean temperatures. Degree-days for development of individuals from the

PAGE 76

60 constant temperature experiment used the temperature at which the incubation chamber was set. Degree-days for development of individuals from the field experiments were regressed against the mean temperature recorded during the period of time the development was achieved. All the individuals completing the respective stage were included. The null hypothesis was slopes equal to zero. Results The wasp Larra bicolor developing in incubation chambers completed its life cycle development at temperatures ranging from 22.5 to 32.5°C (only one wasp completed development at 22.5°C and was not included in the analysis), although a few eggs developed at 20°C and some eggs and larvae developed at 22.5°C. Developmental times were transformed to rates then plotted against their respective temperature (Figure 2) showing the most linear development from 22.5 to 32.5°C for the egg stage, and from 22.5 to 30.0°C for all the other stages and total. The linear equation that best estimates development at constant temperatures, lower thresholds and degree-days is shown in Table 13. The accumulated degree-days required to develop by stage at constant temperatures and at field-varying temperatures are shown in Table 14. Degree-days required for pupal development of Larra bicolor underground at field-varying temperatures are presented in Figure 7. The differences in development are clear among the three groups. The pupae showing the short development, considered as the non-diapausing population, were the only ones included in this validation; the other two groups developed at temperatures outside the range studied.

PAGE 77

61 Table 13. Parameter values for the linear model, lower thresholds and thermal constants (degree-days) for development of Larra bicolor per stage and total at constant temperatures. Stage Linear equation Y = a + b(x)(Eq. 1) / 0 = -(a/b) b Thermal pnrwtant k c = 1 / b Egg Y = -0.401469 0.022 128(x) 18.143 45.44+1.45 Larva Y = -0.221 118 + 0.01 1875(x) 18.620 84.42±1.92 Prepupa Y = -3.787811 +0.203 193(x) 18.641 5.05±0.34 Pupa Y = -0.038899 + 0.002429(x) 16.010 413.13111.59 Total Y = -0.031264 + 0.00 1852(x) 16.770 541.67+12.41 Y = Developmental rate (Y = 1 / days), a = intercept, b = slope, x = temperature = lower temperature threshold k = Degree-days±SEM (days) Table 14 Degree-days (days) required for Larra bicolor stages and total to develop at constant and field temperatures. Temp.°C a N Egg c N Larva c N Prepupa c N Pupa c N Total 0 25.0 26 43.61 ± 1.62 16 83.17 ±3.27 16 4.9 ±0.58 6 400.74 ±21.17 6 525.8 ±27.36 27.5 23 49.3.7 ± 2.46 12 87.19 ±8.05 12 5.2 ±1.59 6 442.13 ±23.59 6 575.3 ± 26.27 30.0 27 44.34 ±1.82 17 83.39 ± 3.03 17 4.9 ± 1.12 8 401.19 ±22.78 8 527.9 ±19.72 32.5 25 44.87 ±1.38 field" 41 40.4 ±1.17 47 89.7 ±2.53 53 6.6 ±0.85 8 481.6 ±14.39 8 595.1 ± 16.38 temperatures in incubation chambers. Overall Mean ± SEM of all wasps developing during the summer in underground chambers in the field. Mean ± SEM. Data for all the individuals sorted by stage and total, whose complete development was recorded from field and from incubation chambers, are depicted in Figure 8. At most of the degree-days calculated, all the stages fell inside the observed range at constant temperatures. Slopes for constant and field temperature tend to be fairly flat, suggesting the degree-days required by stage are constant along the temperature range where development occurred.

PAGE 78

62 o Long development Intermediate development Short development 17.5 20.0 22.5 25.0 Temperature °C 27.5 30.0 Figure 7. Degree-days required for pupal development of Larra bicolor underground at field-varying temperatures. Pupal stage grouped by kind of developmental times (short, intermediate, and long).

PAGE 79

63 20.0 22.5 25.0 27.5 30.0 32.5 B TOTAL DD=541.7 CONSTANT PUPA DD=413.1 22.5 25.0 27.5 Temperature °C 30.0 32.5 Figure 8. Comparison of degree-days required for Larra bicolor life stages and total to develop at constant and field varying temperatures. A) Egg, larval and prepupal development. B) Pupal and total development (short development only). The slopes per stage and total were evaluated showing that none of the regression slopes were significantly different between field and constant temperature experiments (Table 1 5). These results support the hypothesis that data from constant temperatures can

PAGE 80

64 estimate development in fieldvarying temperatures for Larra bicolor over the range of temperatures in these studies. Table 15. Statistical significance of slopes in regression analysis of development of Larra bicolor stages and total, from the field and from incubation chambers. Development Egg Larva Prepupa Pupa Total Field 3 p c 0.625 ns 0.283 ns 0.819 ns 0.829 ns 0.741ns Chambers £ 0.126 ns 0.286 ns 0.295 ns 0.648 ns 0-574 ns a . Development in field-varying temperatures b . Development in incubation chambers at constant temperatures ° statistic of t-test ns = statistically not significant Discussion Development under a regime of fluctuating temperatures often differs from that under constant temperatures. Higley et al. (1986) mention that the magnitude of the differences between developmental times under constant and fluctuating temperatures depends on the average temperature, amplitude and frequency of the fluctuations. Larra bicolor development estimated at constant temperatures and evaluated as degree-days was a good estimator of development at field-varying temperatures. The coincidence of the constant temperature degree-day trend with observed development in the field (Figure 29) is the first clue to the efficiency of constant temperature development to predict field development. The lack of significant differences of slopes in regression analysis of all stages under the two conditions studied proves that Larra bicolor possesses a thermal constant that applies at the temperature range used in this study (Table 25). This thermal constant applies to the pupal stage when this develops in a short time, which mostly occurs during summer and early autumn.

PAGE 81

65 One of the uses of the results presented in this paper is the possibility to predict the period of time when this parasitoid would be present, thus parasitizing host mole crickets in north Florida. The importance of having this knowledge will allow biologists and extension service personnel to predict when the parasitic wasp would emerge from the soil in spring. By this means the phenological model proposed in this paper can be integrated with the pest management of mole crickets. A direct benefit of this phenological model would be the timing of pesticide applications according to the predicted wasp season. The use of the model is restricted to the soil temperature range from 25 to 32.5°C, which occurs from May to November or early December, which covers the normal wasp season. The soil temperature source can be the Florida Automated Weather Network, whose records proved highly coincident with soil temperature in the experimental area (at 22.5 cm depth, see appendix C) where development of the wasp was studied. The development of Larra bicolor has been modeled based on constant temperature studies using the degree-day model approach. The model is a good predictor of development in north Florida soils. The degree-day model to validate development of Larra bicolor at constant temperatures proved valuable to predict field-varying temperature development of nondiapausing populations. Extensive work would be required to model diapausing populations.

PAGE 82

CHAPTER 4 DEVELOPMENT OF THE FLY ORMIA DEPLETA. Development of Ormia depleta at constant temperatures Introduction The tachinid fly Ormia depleta was brought to Florida in 1987 as part of the mole cricket research program and considered potentially the most effective control agent against mole crickets (Fowler 1987, 1989, Frank et al. 1996). This endoparasitic fly, of crepuscular habits, locates host mole crickets by the male host's calling song and deposits larvae (planidia) on or near the host (Fowler 1987, Walker et al. 1996). The first parasitoid strain ("old" strain, from 23°S in Brazil) released in Florida in 1988 (Frank et al. 1996), was considered as largely responsible for a decline in mole cricket populations (Parkman et al. 1996). By 1994, this "old" strain disappeared from the Gainesville area (29° 40 'N latitude), perhaps due to its poor tolerance to cold and long winters (Walker et al. 1996). In recent years, in late autumn, a few flies have been captured near Gainesville, but these were considered to be individuals dispersing from south toward north without re-establishing populations. In 1999 a second tachinid strain ("new" strain) was imported. It was hoped that this "new" strain, from 30°S in Brazil, might be more cold tolerant and could establish populations north of 29°N in Florida and other states (H. Frank personal communication). Studies performed with the first strain of Ormia depleta focused on establishment and range (Frank et al. 1996), seasonality (Walker et al. 1996), effectiveness in 66

PAGE 83

67 controlling mole crickets (Parkman et al. 1996), techniques for rearing it in the laboratory (Wineriter and Walker 1990) and food source (Welch 2000). This "old" strain was in culture from 1987 (Frank et al. 1996) until 1999, but it is the "new" strain that is now in culture to supply populations. Wineriter and Walker (1990) determined the life history of the first strain under laboratory conditions at 23-25°C. There is no information about the effect of temperature on life history or seasonal phenology. Limits of geographical expansion of O. depleta are poorly understood. The fact that the "old" introduced strain remained well established in south-central Florida (27°N), but disappeared farther north suggests a strong influence of temperature regimes, where either colder winters kill the flies or cause the lack of adequate winter nectar sources (Walker et al. 1996). These arguments led to the introduction of a second strain with the hope that it would be more cold hardy and survive colder winters. In recent years the efficiency of O. depleta has been questioned, Potter (1998) stated that this parasitoid does not seen to have a major effect on mole cricket populations, but controversy about the parasitoid role could be due to insufficient information on its biology (Ryoo et al. 1991). Developmental times of many species of poikilotherms as a function of temperature have been described by regression equations (Wagner et al. 1984a). Such regression equations can provide a sample of calculating developmental times in insect population models or can be used for quantitative comparisons of the effect of temperature on the developmental times of different stages and species of insects (Hagstrum and Milliken 1988).

PAGE 84

68 The work described here was aimed to determine the effect of temperatures on development and survivorship of the "new" strain of Ormia depleta, then compare the developmental times and developmental rates of the two strains. Materials and Methods The flies used in this research were laboratory-reared flies using the methods described by Wineriter and Walker (1990). The old strain flies were under culture until 1 999 when the new strain was brought from South America. Gravid fly females, recognized by a dark spot seen near the apex of the abdomen due to the presence of fully grown planidia, were dissected to expose planidia on wet paper in a petri dish. Only dark colored (mature first instar) planidia were used to parasitize host mole crickets. Three planidia were placed in the membranous tissue behind and underneath the posterior of a host pronotum, making sure those planidia remained there until the host was placed into its plastic cup (Wineriter and Walker 1990). Parasitized hosts S. borellii were confined in a 200-ml plastic cup filled with wet sterilized sand. The cups were then transferred to culture chambers at eleven constant temperatures, 10.0, 15.0, 17.0 17.5. 20.0, 22.5, 25.0, 27.5, 30.0, 32.5, 35.0 ± 0.2°C. Relative humidity was set at 80%, and the photoperiod at 14: 10 (L: D). 15 newly parasitized hosts (three planidias per host) were placed per temperature-chamber. Parasitized hosts were fed twice a week following the standard procedure for S. abbreviatus culture. Development was tracked by observing the host through the transparent cup walls every six hours, until parasitic exited the host (by making a hole from inside out of the host's abdomen, killing the host) to pupate in the sand. The pupal development was tracked at the same frequency until adult parasitoid emergence.

PAGE 85

69 The response variables recorded per strain were larval and pupal development at each constant temperature. Survivorship and adult sex ratio were calculated at each temperature. Data were sorted by sex and pooled sexes, and comparisons were made between sexes. Developmental times were compared by "t-test" between temperatures. Developmental rates were compared using linear regression models. Degree-days and lower temperature thresholds were calculated from the linear equations. Finally these variables were tested between strains. Results The new strain The host, S. borellii, did not survive more then seven days at temperatures of 10.0, 15.0 and 35.0°C. The parasitoid was able to develop at temperatures ranging from 17.0 to 30.0°C, where developmental time was inversely related to temperature, as is common in poikilothermic organisms. Developmental time for larvae ranged from 23.6 d to 6.9 d (pooled sexes), 21.7 to 6.8 (males) and 26.5 to 7.9 d (females). Development time for pupae ranged from 30.4 d to 9.3 d (pooled sexes), 3 1 .0 d to 9.0 d (males) and 29.5 d to 9.35 d (females). Total development time ranged from 54 d to 16.5 d (pooled sexes), 52.7 d to 15.9 d (males), and 56 d to 16.5 d (females). (Tables 16,17 and 18 and Figure 9). Mean development times were compared per sex by the 't-test' assuming equal variances (Table 1 9). The null hypothesis was set as males develop in less time than females at either stage or total development. No difference was observed at larval stage. Statistical differences were observed in the pupal stage where males developed in less time at 30.0, 25.0, 22.5, 20.5, and 17.5, thus reflecting the same differences in the total

PAGE 86

70 development. Although statistically not different, males also developed faster than females at 27.5°C. Table 1 6. Development of Ormia depleta (new strain) pooled sexes life stages. Temp. °C Days required to complete development 3 n Larva Pupa Total 30.0 37 6.91 ±0.11 (41.9) 9.28 ±0.08 (58.1) 16.48 ±0.13 27.5 27 7.02 ±0.09 (41.5) 9.90 ± 0.07 (58.5 ) 16.92 ±0.08 25.0 47 8.61 ±0.13(43.4) 11.22 ±0.20 (56.6) 19.82 ±0.18 22.5 24 10.22 ±0.19 (41.5) 14.43 ±0.11 (58.5) 24.65 ± 0.27 22.0 14 12.43 ± 0.47 (44.7) 15.36 ±0.13 (55.3) 27.79 ± 0.58 20.0 53 14.25 ±0.19 (41.5) 20.11 ±0.16(58.5) 34.35 ± 0.20 17.5 17 20.98 ±0.38 (41.6) 29.45 ±0.18 (58.4) 50.43 ± 0.49 17.0 5 23.60 ± 1.25 (43.7) 30.40 ± 0.40 (56.3 ) 54.00 ± 0.89 Mean ± SEM; percentage of developmental time required for each life stage given in parenthesis. Table 1 7. Development of Ormia depleta (new strain) male stages. Temp. Days required to complete development 3 °C n Larva Pupa Total 30.0 18 6.82 ± 0.14(43.0) 9.06 ± 0.11 (57.0) 15.88 ± 0.16 27.5 14 7.10± 0.13(42.2) 9.86 ± 0.10(57.8) 16.96 ± 0.13 25.0 23 8.64 ± 0.20(41.7) 10.87 ± 0.27(58.3) 19.51 ± 0.30 22.5 14 9.90 ± 0.19(40.6) 14.14 ± 0.10(59.4) 24.04 ± 0.27 20.0 27 14.15 ± 0.23 (40.2) 19.64 ± 0.21 (59.8) 33.77 ± 0.26 17.5 5 20.78 ± 0.53 (39.4) 28.68 ± 0.28 (60.6) 49.46 ±0.72 17.0 3 21.67 ± 0.67 (38.8) 31.00 ± 0.00(61.2) 52.67 ± 0.67 Mean ± SEM; percentage of developmental time required for each life stage given in parenthesis.

PAGE 87

71 Table 18. Development of Ormia depleta (new strain) female stages. Temp. Days required to complete development 3 °C n Larva Pupa Total 30.0 19 6.99 ± 0.16(43.3) 9.49 ± 0.11 (56.7) 16.47 ± 0.19 27.5 13 6.93 ± 0.12(41.3) 9.95 ± 0.10(58.7) 16.88 ± 0.11 25.0 24 8.58 ± 0.19(40.7) 11 .56 ± 0.27 (59.3) 20.13 ± 0.20 22.5 10 10.66 ± 0.33(40.9) 14.85 ± 0.14(59.1) 25.51 ± 0.41 20.0 23 14.36 ± 0.36(40.4) 20.47 ± 0.22 (59.6) 34.83 ± 0.30 17.5 12 21.06 ± 0.51 (38.8) 29.77 ± 0.14(61.2) 50.83 ± 0.61 17.0 2 26.5 ± 0.5 (38.8) 29.50 ± 0.5 (61 .2) 56.00 ± 0.00 Mean ± SEM; percentage of developmental time required for each life stage given in parenthesis. 60 50 Survivorship recorded from each temperature is presented in Figure 10, where maximum survival occurred at 27.5°C, followed by 22.5, 20.0, and 25.0°C in a group with survivorship greater than 60% and greater than 50 % at 1 7.5°C. The greater mortality occurred in the larval stage.

PAGE 88

72 Table 19. T-test comparison of developmental time (days) of Ormia depleta (new strain) between sexes per temperature per stage and total life cycle. Larva Pupa Total Temp. °C Male Female Male Female Male Female 30.0 6.82 a 6.99 a 9.06 a 9.49 b 15.88 a 16.47 b 27.5 7.10 a 6.63 a 9.86 a 9.95 a 16.96 a 16.88 a 25.0 8.64 a 8.58 a 10.87 a 11.56 b 19.51 a 20.13 b 22.5 9.90 a 10.66 a 14.14 a 14.85 b 24.04a 25.51 b 20.0 14.15 a 14.36 a 19.64 a 20.47 b 33.77 a 34.83 b 17.5 20.78 a 21.06 a 28.68 a 29.77 b 49.46 a 50.83 b 17.0 21.67 a 26.50 a 31.00 a 29.50 b 52.67 a 56.00 b Means followed by the same letter are not statistically different (one tailed test) Degree-days>11.2°C LARVA PUPA Figure 10. Survivorship of Ormia depleta "new" strain, pooled sexes, at constant temperatures. Regression parameters corresponding to each data set are shown in Tables 20,2 1 and 22 and in Figure 1 1 . The coefficients of determination ranged from 0.90 to 0.95 representing the good fit of all models. The lower threshold values ranged from 11.1 to 1 1 .6. The accumulated degree-days required to complete development per stage and total are shown in last column of Tables 20, 21 and 22.

PAGE 89

73 Table 20. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (new strain) pooled sexes life stages to develop. Stage y Intercept ± SEM Slope ± SEM R 2 ±SE r(°C) Degree-days (Days ± SE) Larva -0.092 ± 0.0044 0.0082 ± 0.0002 0.90 ±0.0109 11.2 123.4 ± 0.9 Pupa -0.071 ± 0.003 0.0062 ± 0.0001 0.92 ± 0.0076 11.5 163.6 ± 0.9 Total -0.04 ±0.0013 0.0035 ± 0.00006 0.95 ± 0.0033 11.4 286.6 ± 1.3 Table 21. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (new strain) male stages to develop. Stage y Intercept ± SEM Slope ± SEM R 2 ±SE t (°C) Degree-days (Days ± SE) Larva -0.090 ± 0.0061 0.0082 ± 0.0003 0.90 ±0.0106 11.1 124.0 ± 1.2 Pupa -0.074 ± 0.0044 0.0063 ± 0.0002 0.91 ± 0.0077 11.6 159.5 ± 1.2 Total -0.041 ±0.002 0.0036 ± 0.00008 0.95 ± 0.0034 11.4 283.2 ± 1.8 Table 22. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (new strain) female stages to develop. Stage y Intercept ± SEM Slope ± SEM R 2 ±SE tec) Degree-days (Days ± SE) Larva -0.094 ± 0.0063 0.0083 ± 0.0003 0.90 ±0.0113 11.4 122.9 ± 1.4 Pupa -0.069 ± 0.004 0.0061 ± 0.0002 0.92 ± 0.0073 11.5 167.4 ±1.3 Total -0.040 ±0.0018 0.0035 ± 0.00007 0.95 ± 0.0032 11.5 289.6 ± 1.8

PAGE 90

74 0.18 0.00 -I— — r15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 Temperature °C Figure 1 1 . Developmental rates of Ormia depleta "new strain", pooled sexes, at constant temperatures. The old strain This old strain developed at temperatures ranging from 17.5 to 30.0°C, showing an inverse relationship between developmental time and temperature. Development time for larvae ranged from 20.39 d to 6.65 d (pooled sexes), 20.52 d to 6.40 d (males) and 20.13 d to 6.78 d (females). Development time for pupae ranged from 32.13 d to 8.72 d (pooled sexes), 31.58 d to 8.48 d (males) and 33.23 d to 8.84 d (females). Total development time ranged from 52.52 to 15.38, 52.10 to 14.88, and 53.35 to 15.63 d (females). (Tables 23, 24 and 25 and Figure 12).

PAGE 91

75 Table 23. Development of Ormia depleta (old strain) pooled sexes life stages. Temp. Days required to complete development 3 °C n Larva Pupa Total 30.0 6 6.40 ± 0.27 (43.0) 8.48 ± 0.06 (57.0) 14.88 ± 0.21 27.5 13 7.15 ± 0.31 (42.2) 9.78 ± 0.37 (57.8) 16.93 ± 0.30 25.0 24 7.95 ± 0.16(41.7) 11.13 ± 0.11 (58.3) 19.08 ± 0.17 22.5 9 9.00 ± 0.42 (40.6) 13.17 ± 0.30(59.4) 22.17 ± 0.26 20.0 15 13.83 ± 0.48 (40.2) 20.56 ± 0.41 (59.8) 34.39 ± 0.29 17.5 6 20.52 ± 0.40 (39.4) 31.58 ± 0.96 (60.6) 52.10 ± 1.01 a Mean ± SEM; percentage of developmental time required for each life stage given in parenthesis. Table 24. Development of Ormia depleta (old strain) male life stages. Temp. Days required to complete development 3 °C n Larva Pupa Total 30.0 6 6.78 ± 0.11 (43.3) 8.84 ± 0.14(56.7) 15.63 ± 0.21 27.5 13 6.73 ± 0.32(41.3) 9.89 ± 0.27 (58.7) 16.62 ± 0.52 25.0 24 7.77 ± 0.10(40.7) 12.02 ± 0.18(59.3) 19.78 ± 0.20 22.5 9 9.67 ± 0.00(40.9) 13.58 ± 0.33(59.1) 23.35 ± 0.33 20.0 15 14.23 ± 0.40(40.4) 20.98 ± 0.18 (59.6) 35.21 ± 0.40 17.5 6 20.13 ± 0.0(38.8) 33.23 ± 0.23(61.2) 53.35 ± 0.23 a Mean ± SEM; percentage of developmental time required for each life stage given in parenthesis. Table 25. Development of Ormia depleta (old strain) female life stages. Temp. Days required to complete development 3 °C n Larva Pupa Total 30.0 6 6.65 ± 0.13(43.3) 8.72 ± 0.12 (56.7) 15.38 ± 0.21 27.5 13 6.93 ± 0.22(41.3) 9.84 ± 0.21 (58.7) 16.76 ± 0.30 25.0 24 7.88 ± 0.10(40.7) 11.50 ± 0.13 (59.3) 19.38 ± 0.15 22.5 9 9.22 ± 0.29(40.9) 13.31 ± 0.23 (59.1) 22.53 ± 0.27 20.0 15 14.10 ± 0.30(40.4) 20.84 ± 0.18(59.6) 34.94 ± 0.30 17.5 6 20.39 ± 0.26(38.8) 32.13 ± 0.70(61.2) 52.52 ±0.70 a Mean ± SEM; percentage of developmental time required for each life stage given in parenthesis.

PAGE 92

76 60 Q LARVA TOTAL PUPA 0 15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 Temperature °C Figure 12. Developmental time of Ormia depleta, old strain, life cycle at constant temperature. Mean developmental times were compared between sexes per temperature per stage by the 't-test' assuming equal variances (Table 26). The null hypothesis was that males develop in less time than females at either temperature per stage or total development. No difference was observed in the larval stage. In the pupal stage the only difference was observed at 25.0°C. In the total life cycle differences were observed at 30.0, 25.0, and 22.5°C. At all cases where differences were observed, males developed in less time than females. Survival recorded from each temperature is presented in Figure 13, where maximum survival (87-88%) occurred at 22.5 and 25.0°C, followed by 27.5, 20.0, 30.0 and 17.5°C respectively, in a group with survivorship ranging from 50% to 82%. The greater mortality occurred in the larval stage at all temperatures but 22.5°C.

PAGE 93

77 Table 26. T-test comparison of developmental time (days) of Ormia depleta (old strain) between sexes per temperature per stage and total life cycle. Larva 3 Pupa 3 -y_ i Iota 3 Temp. °C Male Female Male Female Male Female 30.0 6.82 a 6.99 a 9.06 a 9.49 b 15.88 a 16.47 b 27.5 7.10a 6.63 a 9.86 a 9.95 a 16.96 a 16.88 a 25.0 8.64 a 8.58 a 10.87 a 11.56 b 19.51 a 20.13 b 22.5 9.90 a 10.66 a 14.14 a 14.85 b 24.04a 25.51 b 20.0 14.15a 14.36 a 19.64 a 20.47 b 33.77 a 34.83 b 17.5 20.78 a 21.06 a 28.68 a 29.77 b 49.46 a 50.83 b 17.0 21.67 a 26.50 a 31.00 a 29.50 b 52.67 a 56.00 b Means followed by the same letter are not statistically different (one tailed test) Degree-days > 1 1 .5 °C LARVA PUPA Figure 13. Survivorship of Ormia depleta old strain, pooled sexes, at constant temperatures. Regression parameters corresponding to each data set are shown in Tables 27, 28 and 29 and in Figure 14. The coefficients of determination ranged from 0.88 to 0.97 representing a good fit of all models. The lower threshold values ranged from 1 1 .3 to

PAGE 94

78 12.7. The accumulated degree-days required to complete development per stage and total are shown in last column of Tables 27, 28 and 29. Table 27. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (old strain) life stages to develop. Degree-days Stage y Intercept ± SEM Slope ± SEM R 2 ±SE fro (Days ± SE) Larva -0.104 ±0.0097 0.009 ± 0.0004 0.88 ±0.0121 11.5 112.4 ± 1.4 Pupa -0.087 ± 0.0048 0.0069 ± 0.0002 0.95 ± 0.0059 12.6 146.1 ± 1.3 Total -0.048 ± 0.0026 0.0039 ± 0.0001 0.95 ± 0.0033 12.2 258.2 ±2.1 Table 28. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (old strain) male life stages to develop. Degree-days Stage y Intercept ± SEM Slope ± SEM R2±SE M°C) (Days ± SE) Larva -0.101 ± 0.0137 0.009 ±0.0006 0.88 ± 0.0117 11.3 113.1 ± 2.0 Pupa -0.09 ± 0.0071 0.0071± 0.0003 0.94 ± 0.006 12.6 142.1 + 1.8 Total -0.048 ± 0.0034 0.004 ±0.0001 0.96 ± 0.0029 12.1 254.7 ± 3.0 Table 29. Linear regression models (Y= developmental rate, x= temperature), lower developmental thresholds (t), and degree-days required for Ormia depleta (old strain) female life stages to develop. Stage y Intercept ± SEM Slope ± SEM R 2 ±SE t(°C) Degree-days (Days ± SE) Larva -0.106 ±0.014 0.0090 ± 0.0006 0.88 ±0.0128 11.7 111.6 ± 1.9 Pupa -0.086 ± 0.0054 0.0068 ± 0.0002 0.97 ± 0.0049 12.7 148.3 ± 1.4 Total -0.048 ± 0.0034 0.0039 ± 0.0002 0.95 ± 0.0034 12.3 259.7 ± 2.8

PAGE 95

79 0.20 0.18 0.00 -I— — ,— — i— ! 15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 Temperature °C Figure 14. Developmental rates of Ormia depleta old strain, pooled sexes, life cycle at constant temperatures. Discussion Ormia depleta development rates were sensitive to temperatures and showed a linear response at temperatures ranging from 17.5 to 30.0°C (Figures 1 1 and 14). The developmental times reported by Wineriter and Walker (1990) coincided with those reported in this study, for both strains, at temperatures between 22.5 and 25.0°C for each sex and pooled. Larval development was shorter than pupal development at all temperatures, maintaining similar proportions among temperatures (Tables 16-18, 23-25). Wineriter and Walker (1990) observed that emergence pattern was significantly different for the two sexes, with male adults generally emerging first. The difference in development by sex was more evident in the new strain than the old strain as indicated in Tables 19 and 23, respectively.

PAGE 96

80 Comparing the development of the two strains (Figures 15,16 and 17), male larval development times were statistically different only at 25.0°C. For females and pooled sexes there were differences at 22.5 and 25.0°C. Pupal development for males, females and pooled sexes showed differences at 17.5, 22.5 and 30.0°C. Male total development times were different at 25.0°C, and females and pooled sexes showed differences at 17.5, 22.5 and 30.0°C. The difference in developmental time between sexes was more evident in the new strain than the old, where the larval stage was less variable than pupal and total between the two strains. The development time per strain at the extreme temperatures was statistically different. The old strain took more time to develop at 1 7.5°C and the new strain took more time at 30.0°C (more evident at females, Figure 18, and pooled sexes, Figure 16). In other words, for pupal and total development, the new strain developed faster than the old strain at the lower temperature, but it was the old strain that developed faster than the new strain at the higher temperature. a t E3 Larva New Larva Old O Pupa New Pupa Old 17.5 20.0 22.5 25.0 Temperature °C 27.5 30.0 Figure 15. Developmental times of two strains (new and old) of Ormia depleta, pooled sexes, at constant temperatures. Means followed by different letter are statistically different ("t test", P<0.05).

PAGE 97

81 17.5 20.0 22.5 25.0 Temperature °C 27.5 30.0 Figure 16. Developmental times of two strains (new and old) of Ormia depleta males at constant temperatures. Means followed by different letter are statistically different ("t-test", PO.05). 17.5 20.0 22.5 25.0 Temperature °C 27.5 30.0 Figure 1 7. Developmental times of two strains (new and old) of Ormia depleta females at constant temperatures. Means followed by different letter are statistically different ("t-test", PO.05). Temperatures at which this parasitoid may develop were identified (Figures 9 and 12). The rate of increase decreased at 30°C and stopped at 32.5°C. the upper temperature

PAGE 98

82 threshold must fall between these temperatures. The survivorship observed for the two strains was statistically different only at 25°C. The developmental rates of both strains were compared by means of their corresponding linear regression models shown in Figures 1 1 and 14. This comparison was performed to determine whether both lines can be considered to be estimates of a common 6 (slope) following the procedure of Steel and Torrie (1960: 173), Sokal and Rohlf (1995). The results established that the two strains develop differently at constant temperatures (Table 30). The new strain developed slower than old strain at higher temperatures, but faster than the old strain at low temperatures. The last statement is illustrated in Figure 1 8, where the estimated developments (by the linear equation) were prolonged until crossing the y axes, so that the lower temperature thresholds (where lines crossed the x axes) are presented per strain. No statistical differences were observed in survivorship of the two strains developing in constant temperatures (Figure 19).

PAGE 99

83 0.08 D -0.06 Temperature (°C) Figure 1 8. Linear development estimates by linear regression equations of the two strains of Ormia depleta at constant temperatures. Lower development was prolonged until crossing "y" axis. The lower temperature thresholds for the two strains were compared following the procedures in Infante and Gil (1984) and Sokal and Rohlf (1995), showing the results in Table 30, where the thresholds of the two strains overlapped at 95% confidence making them statistically not different. Nevertheless, it was the new strain that showed the lower temperature at the lower confidence limits of the two strains (Table 20 and 27 and Figure 18).

PAGE 100

84 Table 30. Comparison of slopes (bl) 1 and intercepts (a) 1 between the "new" and "old" strains of Ormia depleta by t-test and confidence limits at constant temperatures. Slope 2 CL1 CL2 Intercept CL1 (LTT) CL2 (LTT) New strain 0.00352 a 0.0034 0.0036 -0.0402 a -0.0380 (10.78) -0.0424(12.04) Old strain 0.00391 b 0.0037 0.0041 -0.0476 a -0.0432(11.05) -0.0520 (13.28) 1 b1 and a are the coefficients from the linear regression equations. 2 Means followed by defferent letters are statistically significant. CL1 = 95% upper confidence limits, CL2 = 95% lower confidence limits. LTT = lower temperature (°C) threshold Male, female and total life cycles were compared between the new and old strains of Ormia depleta reared at constant temperatures. In all cases the comparisons showed statistical differences (p< 0.0001) (Figures 15,16 and 17). Based upon these results, it can be said that the two strains develop at different rates, which reinforces the previous results where developmental times were different at pupal and total life cycles. These findings may support the hypothesis that the new strain is better adapted to develop at extreme temperatures than the old strain that failed to establish populations in north Florida. These results suggest the possible success of the new strain to establish populations north of 27° N latitude in Florida and neighboring states where the pest mole crickets are already established. The previous information of life history of Ormia depleta has been refined to its relationship to constant temperatures, determining the ranges at which development occurs for the two strains released in Florida. Lower temperature thresholds per sex and pooled sexes were determined, along with a good approximation of what can be considered the upper temperature threshold for the two strains.

PAGE 101

85 Figure 19. Survival (%) of the two strains of Ormia depleta at constant temperatures. Means followed by different letter are statistically different ("X 2 test", P<0.05).

PAGE 102

86 Development of the two strains of Ormia depleta under field conditions Introduction Studies of the tachinid endoparasitoid Ormia depleta have been devoted to its laboratory rearing techniques and seasonality of the adult stage. The only information available about the life cycle is under laboratory conditions. No study of this kind has ever made under field conditions. Efficacy of this fly as biological control agent has been evaluated by direct and indirect means in Florida, but data so far obtained are sparse. Extensive data are needed to examine effect on host mole cricket populations, with geographical, seasonal, and annual variation. This section describes the development of larval and pupal stages of Ormia depleta under field conditions at Gainesville, Florida (29°40'N). The two strains of this parasitoid were compared. Materials and Methods Gravid females of Ormia depleta new strain were obtained from the culture at the UF Entomology and Nematology Department, Gainesville, FL. Gravid females of O. depleta old strain were brought from the GCREC, Bradenton, FL, where they were captured in a trap using synthetic mole cricket song as bait. After capture at night, the flies were preserved alive by cooling them in a chest. The flies were still alive when they arrived in Gainesville. Unless otherwise indicated, the host mole crickets were Scapteriscus borellii adults that were captured in traps baited with synthetic mole cricket song at the Ona REC, FL. The hosts were shipped in buckets partly filled with moist vermiculite.

PAGE 103

87 Devices made of PVC pipe were designed to individually confine parasitized hosts underground at a range of depths. These pipe-chambers could easily be buried vertically and could be easily removed from the ground with little disturbance of mole cricket tunnels inside them. The pipe-chambers were subdivided into four sections with the assumption that each section represented a particular thermal regime and the total individuals confined at each depth can be considered as part of an independent experiment. Thus, it was possible to handle four independent experiments at the same time, covering different possible thermal regimes at which the host mole crickets naturally live and thus the parasitoids develop. The depths of the four chamber sections were: 0-15, 15-30, 30-45, and 45-60 cm (Table 6.) The pipe-chambers consisted of 5 cm x 75 cm long PVC pipes that were cut lengthwise. A 2.5 cm wide x 25 cm long window was cut at the center of each half and sealed with aluminum window screen fixed with hot melted silicone. The inside of each pipe was subdivided into 15 cm sub-chambers, by placing non-corrosive metallic mesh, fully matching the inner diameter of the pipe once closed, also fixed with hot melted silicone. Starting from the bottom, the pipe-chambers were subdivided into four subchambers (15 cm tall x 5 cm diameter), leaving a portion of 15 cm intended to be above the ground level for handling purposes. This chamber design allows the host mole cricket to move freely in a volume of 300 ml of sand. The screened windows allow the sand inside the chambers to be in close contact with the sand outside, allowing the flux of air and moisture from the external soil into the chambers. The temperature was monitored inside and outside the chambers at the different depths, with no difference observed.

PAGE 104

88 Moisture was not measured but was estimated by touch; tap water was added when considered necessary. Each half chamber was filled with autoclaved moist sand, lightly compressed. Compressing the moist sand by finger made artificial tunnels. The halves of the chamber were joined as a sandwich by placing rubber bands at both extremes and at the middle. To check the status of host or parasitoid, or to feed the host, the pipes were withdrawn from the soil, the rubber bands were removed, and each half of the pipe-chamber was rolled gently to a concave position to avoid disturbance of the contents. The land where the pipe-chambers were buried was kept free of weeds for the duration of the experiments. It was then assumed that the irradiation, aeration, and moisture (from environmental humidity and precipitation) were evenly distributed without interference from vegetation. The presence of patches of weeds would have modified the flux of these factors. A rectangle of land, 1 m wide x 6 m long, located near the Entomology and Nematology Department building, was tilled to 75 cm depth to allow the manipulation of the pipe-chambers in and out of the ground. Before placing the parasitized hosts into the chambers, these were buried underground for a few days. The chambers were symmetrically distributed, allotting five chambers to the width and 20 chambers to the length of the experimental land. Parasitized hosts were randomly allotted to the sub-chambers. The parasitism of hosts was achieved following the procedure described by Wineriter and Walker (1990). Planidia were removed from gravid flies and 3 planidia were placed on the soft connective tissue behind and underneath the posterior of a host pronotum. Only mature planidia were used, recognized by their dark color. Immature

PAGE 105

89 (light colored) planidia were discarded. A few minutes after they were parasitized, the hosts were placed in moist sand inside the pipe-chambers and buried underground. The parasitized hosts were fed twice a week with the same Purina cricket chow food in use in the laboratory culture. Tap water was added every time the sand in the chambers was noted as dry to prevent hosts dying of dehydration. The newly formed pupae were kept unmanaged the first five days, then they were transferred to a small container made of 2.5 cm diameter and 5 cm long transparent plastic tube with ends capped with fiberglass window screen tightened with small rubber bands. The small containers were half-filled with the same sand where the pupation occurred. The pupae were lightly covered with sand. The empty space in the small container allowed the newly emerged flies to crawl once emerged, and sometimes even to expand their wings. After the transfer to the small containers was made, the pupae were observed every six hours until adults emerged. The unopened pupae were kept under observation up to 15 days in the same place after the last adult emergence observed. Then the remaining pupae were taken to the laboratory to determine whether they were still alive, by dissecting them. Newly emerged adults were sexed. Surviving adult hosts were kept in the same sub-chamber for as long as 30 days more or until their death. The soil temperature, at which the parasitized host was confined in the subchambers, was recorded by placing a thermocouple in the middle of the corresponding sub-chamber. Temperatures were recorded as follows using one pipe-chamber buried in the center of the experimental land. Four thermocouples were used, one at the depth of each of the four pipe sub-chambers. The thermocouples were connected to a digital temperature datalogger (HOBO® Mod H8-4, Onset Co.) programmed to record

PAGE 106

90 temperatures every 30 min. The temperature recorded was downloaded to a computer for further analysis. Another digital temperature datalogger was used to record air temperature, with the thermocouple at 2 m height above ground under a small shed roof. The response variables were the time to develop for larval and pupal stages. The observation frequency was every six hours, starting the fifth day after parasitism until the last cricket died because of the parasitic larval development. Pupal development was monitored every six hours starting five days after the fully formed pupa was observed, ending when the fly adult emerged. The development of the two strains of the parasitic fly Ormia depleta parasitizing Scapteriscus mole cricket adults was evaluated at field varying temperatures through several experiments. The experiments were coded by the letter 'F' followed by the corresponding experiment number to track the period of time when each experiment was performed (Table 3 1 ). Then, a coded letter was added, corresponding to the sub-chamber depth and midpoint where the thermocouple was placed following the same nomenclature shown in Table 6 (p. 45). (i.e., Fl-A = field experiment number one, which corresponds to the most superficial chamber of 0-15 cm depth, with temperature recorded at its midpoint at 7.5 cm). The data are presented by strain.

PAGE 107

91 Table 3 1 . Time periods when experiments were conducted for studying the development of the two strains of Ormia depleta at field-varying temperatures of Gainesville, FL. Period from narasitism to adult emergence (days) llvTT 911 ulll F1 Mav-Tnne ?000 iviay Jtiii^. Z-\J\J\J F2 Julv 2000 %J LI 1 * . — \J \J\J F3 A nonet ^IpntPTTiHft* "7000 F4 September October, 2000 F5 October November, 2000 F6 March -April, 2001 F7 April -May, 2001 Old strain Fl April -May, 2001 F2 April -May, 2001 The new strain The development of the new strain was studied through seven experiments during 2000 and 2001, using the first two sub-chambers, A and B. Scapteriscus borellii was the only mole cricket used. The experimental design was random for all the experiments The old strain Two field experiments were run during April-May 2001. The first experiment used all four sub-chambers and two mole cricket species, S. borellii and S. vicinus; the experimental design was a factorial 2x4 with 10 repetitions. The second experiment used the two more superficial sub-chambers and one host species, S. borellii.

PAGE 108

92 Results The new strain The results are shown in Tables 32 and 33, where it can be seen that the pupal stage takes longer to develop than larval stage. The effect of the seasonal temperatures is better noted in Figures 22 and 23, particularly for the influence of the maximum temperatures recorded during the development of the parasitoid by date. In the superficial sub-chambers (0-15 cm depth) larval development ranged from 6.65 d (summer) to 10.12 d (fall); pupal development ranged from 9.0 d (summer) to 13.09 d (fall); total development ranged from 15.92 d (summer) to 28.07d (spring). In the second chambers (15-30 cm depth) larval development ranged from 6.92 d (summer) to 9.97 d (spring); pupal development ranged from 9.25 d (summer) to 12.54 d (fall); total development ranged from 16.17 d (summer) to 28.76 d (spring). The F-test showed statistical significance among stages and total development (Tables 32 and 33). Table 32. Development time in days and corresponding temperatures during development per stage life cycle of Ormia depleta, new strain, underground in sub-chamber A (0-15 cm depth) in field experiments. Experiment Stages Temperatures °C Code Larva Pupa Total Mean Minimum Maximum F1-A 8.12 ± 0.15 c 10.17 ± 0.24 d 18.29 ± 0.28 f 28.1 24.3 36.6 F2-A 8.15 + 0.10 c 9.00 ± 1.08 e 15.92+1 .08 g 29.1 23.6 41.1 F3-A 8.30 ±0.41 c 10.77 ± 0.59 c 19.07 ± 0.76 de 29.1 22.6 39.7 F4-A 6.65 ± 0.39 d 12.50 ± 0.32 b 19.14 ± 0.56 d 26.3 16.0 37.8 F5-A 10.12 ±0.37 a 13.09 ±0.43 a 23.21 ± 0.51 b 22.7 16.0 28.9 F6-A LD LD 28.07 ± 2.56 a 22.9 9.8 35.7 F7-A 9.77 ± 0.15 b 10.97 ± 0.26 c 20.74 ± 0.28 c 24.7 15.2 36.6 'Means (±SEM), followed by different letters are statistically different. Analysis by stage. LD = Lost data.

PAGE 109

93 Table 33. Development time in days 1 and corresponding temperatures during development per stage life cycle of Ormia depleta, new strain, underground in sub-chamber B (15-30 cm depth) in field experiments Experiment Code Stages Temperatures °C Larva Pupa Total Mean Minimum Maximum F1-B 7.57 ± 0.22 c 9.44 ± 0.33 e 17.01 ± 0.33 ef 28.5 19.1 39.1 F2-B 7.46 ±0.37 c 9.25 ± 0.00 ef 16.17 ± 0.62 g 29.0 25.6 34.4 F3-B 7.66 ±0.09 c 10.06 ± 0.12 d 17.73 ±0.18d 29.2 24.8 34.0 F4-B 6.34 ± 0.22 d 1 1.26 ± 0.30 b 17.60 ± 0.45 de 25.6 19.6 32.4 F5-B 9.97 ± 0.23 a 12.54 ±1.40 a 22.51 ± 1.53 b 22.8 16.0 28.9 F6-B LD LD 28.76 ± 1 .49 a 22.5 13.7 29.1 F7-B 9.79 ±0.36 ab 11.23 ±0.37 be 21.01 ± 0.49 c 24.5 19.0 30.7 1 Means (±SEM), followed by different letters are statistically different. Analysis by stage. LD = Lost data. Although the mean temperatures were very similar during the experiments F2 and F3 at the two depths, the two stages and total development were statistically different. Something similar occurred with experiments F5 and F6, where the total development of the parasitoid was statistically different (Figure 20). Studies of the effect of fluctuating temperatures on insect development often show that development is more rapid at greater diurnal fluctuations (Worner 1992, Hagstrum and Milliken 1991). The old strain The Ftest showed no statistical difference between hosts based upon which it was decided to consider all units as a single host (no distinction of host species). The mean developmental times and corresponding mean temperatures recorded during their development per depth are shown in Table 34.

PAGE 110

94 F7 F1 F2 F3 F4 F5 F7 F1 F2 F3 F4 F5 F7 F1 F2 F3 F4 F5 F6 LARVA PUPA TOTAL Experiments per life stage Figure 20. Comparison of larval, pupal and total development of Ormia depleta, new strain, at two depths at field (F) experiments. Means followed by different letters are statistically different. Analysis by depth. Table 34. Development time in days', standard errors and corresponding temperatures during development per stage life cycle of Ormia depleta, old strain, underground at various depths in field experiments. Experiment Stages Temperatures °C Code Larva Pupa Total Mean Minimum Maximum F1-A 8.55 ±0.1 a 13.45 ±0.1 a 22.00 ± 0.0 a 24.5 15.2 33.6 F1B 8.44 ± 0.2 a 12.92 ±0.1 a 21.36 ±0.3 a 24.6 20.2 28.3 F1-C 8.90 ± 0.2 a 12.69 ±0.1 a 21.60 ±0.3 a 23.9 21.7 26.3 F1-D 7.92 ± 0.3 ab 13.55 ±0.3 a 21.47 ±0.1 a 23.5 22.1 24.8 F2-A 8.00 ± 0.0 c 12.90 ±0.1 a 20.90 ±0.1 b 23.7 15.2 33.6 F2B 8.12±0.1 be 13.13±0.2 a 21.25 ±0.2 ab 23.8 19.0 27.5 'Means (±SEM), followed by different letters are statistically different. Analysis by stage. The mean development times at the several depths by field experiments were not different, according to Duncan's multiple range test, Figure 21. This is not very surprising considering that during the spring, the temperatures were less variable, remaining rather cool and without big differences in mean soil temperatures (maximum

PAGE 111

95 difference was 1 .1°C). In the Fl experiment can be observed how variability in temperature decreases with depth in the soil, Table 34. 25 F1 F2 F1 F2 F1 F2 LARVA PUPA TOTAL Field experiments per life stage and total. Figure 21. Comparison of larval, pupal and total development of Ormia depleta, old strain, at two depths in field (F) experiments. Means by experiment are statistically not different: Duncan's test. Development times The development time fluctuated according to the temperatures, and temperatures varied according to the time of the year. This relationship makes it difficult to compare the development times between strains and between experiments. However, temperatures in spring (experiments F6 and F7 with the new strain, and Fl and F2 with the old strain) were the least variable. The mean temperature recorded during experiment F5 (Oct.Nov.) with the new strain was very similar to those aforementioned. All these experiments (four data sets for larval and pupal development and five data sets for total development) were grouped to run F-tests and mean comparisons between experiments and fly strains at the same depth (Figure 22).

PAGE 112

96 30.0 F5-NS F7-NS Fl-OS F2-OS F5-NS F7-NS Fl-OS F2-OS F5-NS F6-NS F7-NS FI-OS F2-OS LARVA PUPA TOTAL Q 30 25 20 15 10 5 0 30.0 SUB-CHAMBER A b b b b a A c d U I J F5-NS F7-NS Fl-OS F2-OS F5-NS F7-NS Fl-OS F2-OS F5-NS F6-NS F7-NS Fl-OS F2-OS LARVA PUPA TOTAL Figure 22. Comparison of larval, pupal and total developments of the two strains of Ormia depleta at two depths underground. Field experiments (F) selected by similar mean temperatures. NS = new strain, OS = old strain. Means followed by different letters are statistically different. Mean comparisons grouped by stage and total. Larval development time was evaluated from two experiments per strain (thus 4 data sets), where all four sets were statistically different in each of the two. Pupal development times were in general not different from each other, but experiment F7 with the new strain, where the highest temperatures occurred (relative to the temperature range of all the experiments considered in this comparison), resulted in the shortest

PAGE 113

97 development time, statistically different. Total development time in F6 (new strain) was longer than in all other experiments, coinciding with the lowest temperatures recorded. The differences were not very clear between total developments of the two strains in chambers A and B. Discussion Development times The development of larval and pupal stages and thus the total development of the two strains of Ormia depleta were evaluated under field conditions. Seven field experiments at two depths were performed with the new strain, and two field experiments, at four depths in one experiment and two depths at the other, with the old strain. The whole set of experiments was accomplished by use of 22 parasitoid cohorts (= 22 experiments). A comparison of the development of the two strains showed no clear differences between them when the temperature regimes were most similar. The mean developmental times of the new strain recorded at the two depths show little significant difference, and in most of the cases there was no difference between the two depths evaluated (Figure 20). The mean developmental times of the old strain recorded at the four depths show few significant differences. The differences observed in development times of the new strain occurred during the highest temperatures recorded in the year, with the direct effect of stronger irradiation that where absorbed and transmitted through the bare soil in the experimental land, primarily affects the most superficial sub-chamber. Apparently, the depth of soil around the deepest sub-chambers, perhaps aided by moisture content of the soil, reduced the effect of that stronger solar irradiation during most of the year, which might partly explain the lack of differences between the deepest sub-chambers. Presence of vegetation

PAGE 114

98 might add further insulation, so the development of this parasitoid might be expected to fluctuate less under vegetated soil where temperature and moisture would be more stable. Problems It is important to mention that the most important problem faced during the experiments with the new strain was caused by two fire ant species, Solenopsis invicta Buren and Diplorhopium sp. The huge mobilization capability and aggression of S. invicta inflicted severe losses on the host mole crickets, thus affecting the parasitic larval development. The small fire ant Diplorhopium sp. was more frequently seen devouring the newly emerging fly adults, causing the loss of data on sex of the newly emerged flies. Most of the cases when the small fire ant attacked a still living mole cricket, which was weakened due to the parasitic effect, the parasitic larvae were able to pupate inside the body of the dying mole cricket. In very few cases the same behavior was observed when S. invicta attacked. Very often the parasitoid pupation happened very close to the mole cricket carcass, which can be an important source of moisture for the developing pupae. F6 was the experiment most severely affected by a nocturnal S. invicta invasion, forcing me to remove the chambers to another side of the experimental land, but not before losing most of the hosts and developing larvae. To avoid problems with ants in future experiments, it would be advisable to try covering the lateral windows with a fine metallic mesh able to prevent the small fire ants from entering the chambers but allowing the temperature, moisture and aeration flux into the chambers. The other major problem was with the temperature-recording datalogger. It stopped several times during experiments Fl, F3, F4 and F5, losing temperature data for some period of time in each of these experiments. The lost temperatures were estimated

PAGE 115

99 by means of modeling portions of my temperature records using the Florida Automated Weather Network (FAWN) data from a station in Gainesville at a very short distance from the experimental land. A stepwise linear regression method was applied. It used 500 of my data points as dependent variables, and the air and soil temperature records from FAWN as independent variables. The procedural details of the temperature model are presented in Appendix C. Diapause Wineriter and Walker (1990) speculated that Ormia depleta may overwinter as a pupa, and may diapause. However, the experiments reported here gave no indication of diapause, and some pupae developed during November when on some days, for short periods of time, the temperature in the more superficial chambers dropped to 14°C. The experiments reported here were not intended to investigate the possibility of diapause and they do not rule it out. First, some of the mole crickets overwinter as adults, after having mated late in the year, resuming reproduction next year, thus suspending reproduction during the coldest months. Suspended development of Ormia depleta planidia might conceivably occur inside such mole cricket hosts due to factors other than temperature. Second, experiments were not conducted during the coldest months, the very time when diapause would be likely to occur.

PAGE 116

100 Degree-day model predicting the development of Ormia depleta. Introduction Because temperature affects the development rate, fecundity, and mortality (and therefore the dynamics) of insect populations, considerable research on insect development in relation to temperature has been done. The purpose of many studies was to determine an appropriate development rate function for a phenological or general population model that can be used to predict important events in insect life cycles, or insect abundance, for control purposes (Candy 1991, Worner 1992, Lactin et aL 1995), which has been extended to estimating the effect of parasitoids on host populations (Gould and Elkinton 1990). The rate of insect development depends upon the temperature to which the insects are exposed. For a developing insect, the rate at which time passes depends on the combination of its temperature-dependent development rate curve and the temperature at which it lives. The development rate curve may be altered by many factors of the environment, such as humidity, nutrition, or disease, but the curve always remains (Taylor 1981, Vinson and Iwantsch 1980). For each species, the temperature below which no measurable development occurs is its threshold of development. The amount of heat required over time for an insect to complete some aspect of development is considered to be a thermal constant (Campbell et al. 1974). Scapteriscus mole crickets are pests of pasture and turf grasses and vegetables in the southeast and are spreading westward in the USA. They are soil-inhabiting insects, reaching the ground surface for dispersal and mate finding. Their life cycles are described in Chapter 2.

PAGE 117

101 Ormia depleta is a South American tachinid that attacks mole crickets of the genus Scapteriscus (Fowler 1987). Adults are crepuscular and nocturnal parasitoids. Females are phonotactic, locating their hosts by their calling songs, and they deposit larvae on or near the host (Fowler and Kochalka 1985). Its life cycle is described in Chapter 2. Two strains of the larviparous endoparasitic Ormia depleta have been released in the USA. The first strain (old strain) released in Florida established populations in central and south Florida, failing to survive in the northern part of this state, perhaps in some way due to low winter temperatures (Frank and Parkman 1999). A second strain (new strain) has been released in Georgia and North Carolina, and it may be released in north Florida, with the hope that it will tolerate winter temperatures and establish populations. There would, however, be no point in making further releases of the new strain unless evidence can be obtained that it is somehow more tolerant of colder winter temperatures. The life cycle duration and some basic statistics of both fly strains at constant temperatures and all the data necessary to build a degree-day model of this parasitoid are presented in Tables 19-21. The life cycle of both strains developing underground in the field during all but the coldest months of the year is presented in Tables 26-28. The objective of this work is to evaluate the degree-day model based upon constant temperatures to fit the field development of Ormia depleta parasitizing mole crickets in north Florida (29°40'N). Materials and Methods Ormiaparasitized mole crickets were held at several constant temperatures. The developmental time per Ormia individual is converted to developmental rate by the reciprocal of time. The rates are then regressed against their corresponding temperatures

PAGE 118

102 per life cycle stage and total development. A degree-day model is built based upon regression parameters (Eq. 1). The total degree-days required to develop are calculated by the reciprocal of the slope (bi), and the lower threshold is calculated by dividing the intercept by the slope and multiplying it by -1. Y = a + b(x): (Eq.l) Y = Developmental rate (1/days) a = Intercept b = Slope x = Temperature °C Once having calculated the linear equation describing the development of the parasitoid at constant temperatures, the corresponding lower threshold per stage and total are estimated. The degree-days required to complete development at field-varying temperatures are estimated by the addition of degree-day per unit of time at which the temperature is recorded per stage and total at various depths in each field experiment. The estimation of the degree-days is performed by equation 2: r=jrr(f-0 Eq . 2 Y = Degree-days T = Time (time step = 30 minutes = 1/48 days) / = Temperature recorded every 30 min to = Lower temperature threshold (per corresponding stage or total) n = Number of days for stage to develop i = Number of individuals The equation is applied to every temperature recorded during the development (recorded every 30 min) of every parasitic individual, and then the degree-days are averaged per cohort and plotted on the same graph together with the estimated degree-

PAGE 119

103 days at constant temperatures. All available mean degree-days of individuals that developed at every depth were included, grouping the data per larval, pupal and total development per fly strain. Results The two strains of Ormia depleta completed development at constant temperatures ranging from 17 to 30°C. None of the larvae completed development under 17°C or above 30°C. Developmental times (days) were transformed to rates, then plotted against their respective temperatures (Figure 8, and Figure 15,), showing a linear development covering the full temperature range (17 to 30°C) for larval, pupal and total development. The development of O. depleta at field-varying temperatures was recorded from 12 cohorts of the new strain, and from six cohorts of the old strain. All experiments included two soil depths; the old strain included four soil depths in one data set (see Table 6, p 57). The new strain field experiments covered all the seasons of the year but winter; the old strain experiments were run during spring only. The new strain. The linear equation that best estimates development of O. depleta (new strain) at constant temperatures, lower thresholds and degree-days are presented as a linear equation model in Table 35. These parameters were calculated from studies at constant temperatures (Table 2 p. 48).

PAGE 120

104 Table 35. Parameter values for the linear model, lower thresholds and thermal constants (degree-days) for development of Ormia depleta (new strain) per stage and total at constant temperatures. Stage Linear equation 3 Y = a + b (x) Lower threshold (°C) /« = -(a/b) Thermal constant 0 k= 1 /a Larva Y = -0.092 + 0.0082(x) 11.22 123.41+0.90 Pupa Y = -0.071 +0.0062(x) 11.54 163.6±0.92 Total Y = -0.040 + 0.0035(x) 11.40 286.60±1.3 Y = Developmental rate (Y = 1 / days), a = intercept, b = slope, x = temperature to = lower temperature threshold; k = Degree-days±SEM (days) Table 36. Degree-days required by Ormia depleta (new strain) to develop in incubation chambers and at field-varying temperatures. Larva Pupa Total Condition Temp.T h DD+SEM Temp.T DD+SEM Temp.T h DD+SEM Constant 17.0 136.34±7.2 17.0 166.59+2.2 17.0 302.93+5.3 17.5 131.69+2.3 17.5 176.09+1.1 17.5 301.31+3.4 20.0 124.68+1.5 20.0 170.52+1.4 20.0 295.31+1.6 22.0 131.87+4.3 22.0 160.94+1.4 22.0 292.81+5.4 22.5 115.59+2.1 22.5 158.48+1.2 22.5 273.73+3.1 25.0 118.47+1.8 25.0 159.78+1.6 25.0 277.35+1.8 27.5 114.24+1.4 27.5 158.27+1.1 27.5 279.29+3.1 30.0 128.74+1.5 30.0 171.44+1.6 30.0 300.18+2.3 Field 3 Fl A 28.8 140.68+1.3 27.8 166.18+1.9 28.33 305.59+2.3 Fl B 29.3 136.7+1.9 27.8 154.41+2.5 28.51 290.61+2.6 F2 A 29.3 124.7+0.0 28.9 156.37+0.0 29.05 279.10+0.6 F2B 29.2 124.1+0.0 28.9 160.87+0.0 29.00 288.10+0.0 F3 A 29.4 148.7+2.6 28.8 184.86+2.7 29.01 332.43+3.0 F3 B 29.1 137.1+0.8 29.3 178.4+0.6 29.23 316.03+1.4 F4 A 29.4 115.4+1.1 24.7 159.2+1.9 26.40 267.07+2.8 F4B 27.6 122.2+2.7 25.6 163.9+1.6 26.34 284.84+2.9 F5 A 23.3 115.6+0.8 22.7 145.7+2.0 22.75 253.83+1.7 F5B 24.3 116.3+1.1 23.5 154.6+6.5 23.65 269.91+6.9 F7 A 23.3 110.5+0.5 26.1 159.9+2.2 24.74 275.32+1.8 F7B 23.3 119.5+2.7 26.6 157.8+2.0 24.54 271.58+2.5 a Codes per experiment (see Tables 6, p. 45 and 31, p. 88). Mean temperature during development; b Average degree-days calculated with Eq. 2, all observations included.

PAGE 121

105 The lower threshold values were used in Eq. 2 to calculate the accumulated degree-days required for O. depleta stages to develop at the two rearing conditions. Table 36 shows the mean degree-days per fly cohort and corresponding mean temperatures at which the development occurred during all experiments. Mean temperatures ranged from 22.7 to 29.4°C, which fell inside the optimal temperature range, estimated through the constant temperature experiments. The complete data set, comprising all observed developed O. depleta new strain, along with individual temperature at which each individual developed are depicted in Figure 24. The dispersion of the data appears to show a positive trend with relation with temperature increases. The data regressed against corresponding temperature, including those constant temperatures comprising field development only (Table 37), show the slope values and statistical significance from regression analysis, demonstrating statistical significance of the slopes for all stages under both rearing conditions. The next step was to evaluate whether the slopes of both temperature conditions matched the estimate of the degree-days for development at field-varying temperature. The slope from constant temperatures was compared with the slope from field-varying temperatures for each stage; calculating the 95% confidence limits made the comparison (Table 38). All confidence limits overlapped between constant and field-varying temperatures, meaning that regression slopes from constant temperature experiments are statistically not different from slopes from field temperatures. The constant temperature data set can be used to predict development of this parasitoid at any period of time of the year but winter.

PAGE 122

106 200 0 — —r-i— — r— r15.0 17.5 20.0 22.5 25.0 27.5 30.0 400 . 300 >. (0 ? © 200 £ a 0) a 100 O TOTAL 15.0 17.5 20.0 22.5 25.0 Temperature °C 27.5 30.0 Figure 23. Regression lines of development of Ormia depleta (new strain) in incubation chambers set at constant temperatures ( Chambers ~ ~ ) and at field varyingtemperatures ( ^ Field ). A) Larval development; B) Pupal development; C) Total development.

PAGE 123

107 Table 37. Statistical significance of slopes in regression analysis of development of Ormia depleta (new strain) stages and total, in field and in incubation chambers. Condition Larva Pupa Total Field a 2.97 (p= 0.001) 3.31 (p= 0.001) 8.77 (p= 0.01) Chambers b 2.35 (p= 0.01) 3.01 (p=0.01) 8.25 (p= 0.001) a Development at field-varying temperatures (n = 156). b Development in incubation chambers at constant temperatures (n = 135). c Slope from regression pooling all observations (statistical "p" value inside parenthesis) Table 38. Regression slope comparison by 95% confidence limits of degree-days of developmental stages of Ormia depleta (new strain) developing at field-varying temperatures and in incubation chambers at constant temperatures. Larva Pupa Total UCL a LCL" UCL a LCL b UCL a LCL" Field" 2.28 3.65 2.58 4.04 7.65 9.90 Chambers' 1 1.85 2.86 2.37 3.66 7.31 9.19 "Upper confidence limits b Lower confidence limits. °A11 observations included (n = 156) included: only constant temperatures in the range of field development (22.5 to 30°C) (n = 135). The old Strain. The linear equation that best estimates development of O. depleta (old strain) at constant temperatures, lower thresholds and degree-days are presented in Table 39. The equation coefficients were calculated from studies at constant temperatures (Tables 24 to 26).

PAGE 124

108 Table 39. Parameter values for the linear model, lower thresholds and thermal constants (degree-days) for development of Ormia depleta (old strain) per stage and total at constant temperatures. Linear equation" Lower treshold (°C) D Thermal constant 0 Stages Y = a+b(x) (Eq. 1) t 0 = a / b k=l/b Larva Y = -0.104 + 0.0090(x)) 11.52 1 12.4+1.36 Pupa Y = -0.087 + 0.0069(x) 12.61 146.1±1.26 Total (All stages) Y = -0.048 + 0.003 9(x) 12.20 258.2+2.14 Y = Developmental rate (Y = 1 / days), a = intercept, b = slope, x = temperature. to = lower temperature threshold k = Degree-days±SEM. The lower threshold values were used in Eq. 2 to calculate the accumulated degree-days required for O. depleta stages to develop under the two rearing conditions. Table 30 shows the mean degree-days per fly cohort and corresponding mean temperatures at which the development occurred during all experiments. Mean field temperatures ranged from 22.5 to 24.5°C, which fell inside the optimal temperature range, estimated through the constant temperature experiments. The narrow temperature range was because experiments with the old strain were performed during April and May, when soil temperatures remain rather low when compared with summer and fall months. The complete data set, comprising all observed developed O. depleta new strain, along with the individual temperature at which each individual developed are depicted in Figure 29; although development at constant temperatures ranged from 17 to 30°C, Figure 29 shows those temperatures that cover field development temperatures only. The dispersion of the data appears to show a positive trend with relation with temperature increases. The data were regressed against corresponding temperature. Table 3 1 shows the slope values and statistical significance from regression analysis, demonstrating statistical significance of the slopes for all stages at both rearing conditions. The next step

PAGE 125

109 was to evaluate whether the slopes of both temperature conditions matched the estimate of the degree-days to develop at field-varying temperature. The slope from constant temperatures was compared versus the slope from field-varying temperatures estimating development at the same stage; calculating the 95% confidence limits made the comparison (Table 32). All confidence limits overlapped between constant and fieldvarying temperatures, meaning that regression slopes from constant temperature experiments are statistically not different of slopes from field temperatures. The constant temperature data set can be used to predict development of this parasitoid at any period of time of the year but winter. Table 40. Degree-days required by Ormia depleta (old strain) to develop in incubation chambers and in field-varying temperatures. Larva Pupa Total Condition Temp.°C DD±SEM D Temp.T DD±SEM b Temp.T DD±SEM b Constant 17.5 115.03±3.3 17.5 149.94±1.9 17.5 249.33±2.4 20.0 113.28±3.4 20.0 145.67±1.9 20.0 244.79±2.5 22.5 1 04.3 7± 1.4 22.5 140.16±1.8 22.5 244.60±1.9 25.0 106.01±1.9 25.0 142.29±1.4 25.0 244.58±1.7 27.5 115.74±2.9 27.5 151.69±1.1 27.5 247.94±2.3 30.0 113.70±3.8 30.0 145.28±1.9 30.0 253.98±2.1 Field 3 Fl A 24.3 1 1 1 .3=1=3. 1 23.0 144.87±1.6 23.7 253.46±2.4 Fl B 24.5 110.5±1.9 23.8 135.05±2.5 23.7 249.83±1.3 Fl C 23.9 109.1±0.0 22.9 138.76±3.0 23.3 240.62±1.1 Fl D 23.5 95.2±0.0 28.9 136.87±0.0 23.1 234.41±3.2 F2 A 24.4 103.5±0.2 23.2 184.86±0.9 23.6 239.81±1.3 F2B 24.5 105.1±0.8 23.4 142.4±2.3 23.8 247.44±2.2 a Codes per experiment (see Table 6, p. 59 and Table 31, p. 92). Mean temperature during development completion. b Average degree-days calculated with Eq. 2, all observations included.

PAGE 126

110 150 w 100 >» < (0 ® 100 £ 3 50 15.0 o o o 0 0 -©A> LARVA 15.0 17.5 20.0 22.5 25.0 27.5 30.0 B* PUPA 17.5 20.0 22.5 25.0 27.5 30.0 300 250 w 200 (0 | 150 E O) S 100 50 • * 1 * O TOTAL 15.0 17.5 20.0 22.5 25.0 Temperature °C 27.5 30.0 Figure 24. Regression lines of development of Ormia depleta (old strain) in incubation chambers set at constant temperatures ( B Chambers — ) and at field varyingtemperatures ( Field ). A) Larval development; B) Pupal development; C) Total development.

PAGE 127

Ill Table 41 . Statistical significance of slopes in regression analysis of development of Ormia depleta (old strain) stages and total, in the field and in incubation chambers. Condition Larva Pupa Total Field 3 4.09 (p= 0.001) 6.23 (p= 0.001) 14.99 (p= 0.001) Chambers b 4.36 (p= 0.001) 5.33 (p= 0.001) 14.66 (p= 0.001) "Development at field-varying temperatures (n = 85) b Development in incubation chambers at constant temperatures (n = 33 ) c Slope from regression pooling all observations (statistical "p" value inside parentheses). ns = not significant Table 42. Regression slope test by 95% confident limits of degree-days of larval, pupal and total development of Ormia depleta (old strain) in field-varying temperatures and in incubation chambers at constant temperatures. Larva Pupa Total UCL a LCL b UCL a LCL" UCL a LCL Field 0 0.254 7.94 3.489 10.816 9.653 20.342 Chambers' 1 2.147 6.57 2.147 6.574 10.48 18.929 "Upper confidence limits ''Lower confidence limits. C A11 observations included (n = 85) included constant temperatures ranging field development (22.5 to 30°C) (n= 33). Discussion Development of eighteen cohorts of Ormia depleta (two strains) parasitizing mole crickets was studied at fieldvarying temperatures. The larval, pupal and total development degree-day model based upon experiments at constant temperatures proved good predictors of development at field conditions. Ormia depleta development estimated at constant temperatures and evaluated as degree-days was a good estimator of development at fieldvarying temperatures. The coincidence of the constant temperature degree-day trend with observed development in

PAGE 128

112 the field (Figures 24 and 25), is the first clue to the efficiency of constant temperature development to predict field development The significance of slopes in regression analysis under the two conditions studied suggests that the degree-days required for O. depleta increase with temperature increases, although the thermal constant applies at the temperature range established in this study (Table 28). Greater variability was observed among data obtained using the new strain than the old strain. This is explained by the fact that experiments with the new strain were performed at a wide range of naturally occurring temperatures in Gainesville, FL (29°40'N). The data on the old strain were less variable, due to the narrow temperature range experienced in the field. Although the degree-days development from constant temperatures fitted development in the field, much more work will be necessary to determine the developmental response of the old strain during the year. The results with the new strain are more relevant, not only because they are more complete, but also because this strain has been released in several states north of Florida. The phenological model developed in this work may provide the opportunity to understand and predict development of this parasitoid during the months when mole crickets are active. One of the uses of the results presented in this paper is the ability to predict the period of time when this parasitoid would be active in relation to the host mole cricket calling seasons (flying and mating seasons) in north Florida. This knowledge will allow biologists and extension service personnel to use soil temperatures to when adult flies

PAGE 129

113 will emerge from the soil and become active. Thus, the phenological model proposed in this paper can be integrated with the pest management of mole crickets. An example of a direct benefit of this phenological model would be the timing of pesticide applications according to the predicted fly season. The use of the model is restricted to soil temperatures ranging from 17.0 to 30°C, which occurs from April to November or early December. This is the normal fly season. The soil temperature source can be the Florida Automated Weather Network, whose records proved highly coincident with soil temperature in the experimental area (at 22.5 cm depth, see appendix C) where development of the fly was studied.

PAGE 130

CHAPTER 5 CONCLUSIONS The objective of this dissertation was to describe the development of two insect parasitoids of mole crickets, the wasp Larra bicolor and the fly Ormia depleta, under a range of constant temperatures in the laboratory and at field-varying temperatures. The stages of work for each species were (a) experimental design, (b) conduct of the experiments and collection of data, (c) analysis of data, (d) fitting of the data to an accumulated degree-day model, and (e) validation of the models. In addition, although this was not included in the original plan, levels of parasitism of mole crickets by Larra bicolor were measured in two field populations in the autumn of each of two years. The Parasitic Wasp Larra bicolor 1 . The population reproductive statistics, such as fecundity, fertility and female longevity, of the progeny of wasps collected from the field during summer and autumn were statistically not different. Reproductive parameters such as net reproductive rate, intrinsic rate of increase, and generational time, were statistically different. These results indicate that the season of the year, defined by lower temperatures and shorter photo-phase during autumn, causes the wasp to modify the rates of development and reproduction. The statistical difference between seasons is explained by the fact that this parasitoid is adapted to modify its metabolism to overwinter as a pupa, which is reflected by doubling of the generational time in the progeny of wasps collected in the autumn. 114

PAGE 131

115 2. Determination of development times of Larra bicolor at seven constant temperatures permitted an estimation of the lower developmental thresholds for egg, larva, prepupa, pupa, and total development. A precise upper developmental threshold could not be established with these data. However, the results indicate that the upper developmental threshold is near 32.5°C. The maximum survivorship of total development (from egg to adult) observed at constant temperatures was observed at 30°C, the minimum at 32.5°C, which supports the proposed upper threshold. 3. The development of the wasp Larra bicolor was studied under field conditions of Gainesville, FL. No statistical differences in development were observed when tracking the wasp developing at several underground depths. Statistical differences were observed among wasp cohorts that started developing in summer and autumn. Instead of the data showing a continuum from short to long developmental times, there was a trimodal pattern, with short (56.6 d), intermediate (1 12.9 d), and long developmental time (238.5 d), respectively. The long development that may occur from summer populations, more commonly from autumnal wasps, is considered to be the diapausing generation that provides the wasp populations the opportunity to survive cool winters and adult food shortage. This long development appears to be synchronized with the blooming of a plant that provides an important nectar source. This plant starts blooming in early May at this latitude (29°40'N). The overall survivorship of the wasp, from egg to adult, at fieldvarying temperatures was 1314%; 58% of the wasps reached the prepupal stage, which indicates that the host mole cricket was devoured by the wasp grub. The overall sex ratio observed was 1:1.

PAGE 132

116 4. Calculation was made of the degree-days required by stage and total development of Larra bicolor at constant temperatures. The linear models thus derived were applied to calculate the development of the wasp at field-varying temperatures. The accumulated degree-days required for wasps developing in the field were the same or very similar to those recorded at constant temperatures for the eggs, larvae, and prepupae. However, the pupal stage at constant temperature had only a short developmental time. Thus, the development times of pupae in the field, and the total development times (all successive immature stages) were underestimated for those individuals having pupae that took an intermediate time or a long time to develop. 5. Adult wasps of the strain that was released in north Florida, were observed parasitizing mole crickets. On average, 23% of the mole crickets captured in pitfall traps throughout the autumn of each of two years at two sites near Gainesville, FL (29° 40' N), were found to be parasitized by wasp eggs or larvae. Generational mortality inflicted on mole crickets by the wasp is obviously much higher. There was no statistical difference between parasitized mole crickets by sex, size, and developmental stages. Nymphs of small-medium size were observed being parasitized and some of these succeeded in producing adult wasps. Based upon these findings, the parasitic wasp can be considered to play an important role as a biological agent, reducing mole cricket populations in north Florida. The Parasitic Fly Ormia depleta. 1 . The development of the two strains (Piracicaba and Osorio) of the tachinid fly Ormia depleta was studied per sex, stage, and total development at constant temperatures. Both strains develop at temperatures between 17 to 30°C; developmental time was

PAGE 133

117 inversely related to temperature. Statistical differences were observed at pupal (thus at total development) stage between sexes, where males developed faster than females. Larvae developed faster than pupae at all temperatures. The survivorship of the strains differed, the Osorio (new) strain showed maximum survivorship at 27°C followed by 30°C; the Piracicaba (old) strain showed maximum survivorship at 22.5° followed by 25°C. Lower developmental thresholds were estimated followed by the degree-days required per stage and total by strain, to complete development. A precise upper developmental threshold could not be established with these data, however, the results indicate the upper developmental threshold to be near 32.5°C, because no development of larvae and pupae occurred at this temperature. The new strain thus has a lower temperature threshold and it develops faster at lower temperatures than does the old strain. The old strain develops faster at the highest temperature (30.0°C). These findings suggest that the new strain is better adapted to succeed at lower temperature regimes, and might be expected to establish a population farther north in the northern hemisphere. 2. Field evaluations were made, during April-November for the new strain, and during April-May for the old strain, at 29°40'N. The development time fluctuated according to the temperatures, and temperatures varied according to the season of the year. The differences were not very clear between total development times of the two strains developing during April-May 2001, thus exposed to very similar temperature regimes. The degree-days required for total development fluctuated from 250 to 330°C-day, for all field experiments.

PAGE 134

118 3. A degree-day model was constructed using the results from the study of Ormia depleta development at constant temperatures. The model was then evaluated for its goodness of fit to the field data. Field data fell into the predicted range. Future Research 1 . Suggestions for future research include evaluation of development times and survivorship for wasps and flies outdoors during the coldest months (DecemberMarch), and further evaluation of diapause. It is not known whether the flies can diapause. The observed trimodal pattern of development time of the wasp pupa is at least curious; it may be an artifact of too few observations, or it may be real. If it is real, then its control mechanism requires investigation.

PAGE 135

APPENDIX A DEVELOPMENT OF TWO MOLE CRICKETS, S. VICINUS AND S. ABBREVIATUS, AT CONSTANT TEMPERATURES. Introduction Three species of mole crickets (Scapteriscus spp.) are the most damaging turf and pasture grasses pests in much of the southeastern United States. They destroy turf and grasses by feeding on roots, leaves and shoots, and by tunneling, which disrupts the soil around roots, preventing uptake of water and nutrients (Walker 1985a, Hudson 1987). The pest mole cricket populations are spreading in the southern USA, damaging turf and pastures grasses. Several authors (Hayslip 1943, Forrest 1985, Hudson 1987, and Braman 1993) studied the general biology, oviposition and development of mole cricket species. Those studies did not consider the effect of temperatures on the life cycle development of the mole crickets. Understanding the relationship between mole cricket life cycle and temperature would help interpretation of the seasonality of the pests and their interaction with the imported parasitoids in adapting to the pests' life cycle in Florida (see Chapter 2). Having the description of the effect of temperatures on life cycle of mole crickets is the primary step toward modeling the mole cricket development that would enable the construction of models to understand the interrelationship of the organisms involved in the biological control of mole crickets. This work was aimed to determine the life cycle 119

PAGE 136

120 development of two mole cricket species, Scapteriscus vicinus and S. abbreviatus at constant temperatures in incubation chambers. Materials and Methods Host mole cricket Scapteriscus vicinus adults were captured in traps baited with species-specific synthetic mole cricket song at Gainesville, FL (29°40'N). The hosts were transferred to the laboratory in buckets partly filled with moist sand. The S. abbreviatus mole crickets used in this research were obtained from the culture in the UF-Gainesville Entomology and Nematology Department. The study of development of mole crickets requires keeping them in individual containers to prevent cannibalism, so this procedure was applied to S. abbreviatus in culture. Each mole cricket was confined in 300 ml containers, filled with autoclaved sand. The containers consisted of 7.5 cm diameter * 7.5 cm tall PVC-pipe with both ends capped with plastic petri dishes kept in place by rubber bands. The bottom of the containers was covered with black plastic to stimulate (by exclusion of light) the mole cricket female to oviposit, which facilitates the location and extraction of the eggs. To start the experiment, it was necessary to be sure that all females were fertilized to produce progeny, starting ovogenesis at same time. The S. abbreviatus were all virgins, because in culture they developed in individual containers. Some of the S. vicinus (all captured from the field), may have already mated by the time they were collected (Walker and Nation 1982), but it was assumed that most of the females were virgins, or just recently mated; to promote uniformity, all the female crickets were collected during two days of the same week in early March, confining them immediately. Females were provided with males for a period of seven days, to ensure mating, then the males were

PAGE 137

121 removed. The mid-point of the week was considered the date of mating, followed by ovogenesis. Thirty two S. vicinus and 28 S. abbreviatus females were included in this work. The study was initiated the second week of March 1999, coinciding with the beginning of the S. vicinus activity season in Gainesville, FL; S. abbreviatus populations are not established at this northern latitude (29°40' N) (see Chapter 2). The reproductive females were kept at a constant 26°C and photoperiod of 14:10 (D:L); the sand was kept wet by adding tap water when needed. The females were fed twice a week with the same mole cricket chow in use in the culture. Oviposition, fecundity and fertility The number of egg clutches per female and the periods of time between clutches were quantified. Thus, the oviposition period per mole cricket species was determined. To determine when a clutch of eggs was laid, it was necessary to open the container and look for the eggs or for the oviposition chamber, causing minimal disturbance. The observations were made twice daily because, in previous experiments, S. vicinus was observed eating its own eggs even the same day they were oviposited. Although this was never observed to happen with S. abbreviatus, the observations were made at the same frequency, as a standard method. Sometimes a female may oviposit the ^ total egg clutch in two steps. Usually spaced within a few hours, oviposition occasionally was spaced over to two or three days; when this happened, all eggs were considered to belong to a single egg clutch, but were labeled with distinct dates. The same day the eggs were laid, they were confined in 2.5 cm plastic petri dishes containing a wet sponge wrapped with filter paper. The petri dishes were labeled with a

PAGE 138

122 code for identification, date when oviposited, consecutive number of egg clutch, and number of eggs. The egg clutches were randomly allotted to six environmental chambers, set at temperatures of 10, 15, 17, 20, 25, and 30°C, respectively. The proportion of eggs hatching was quantified per egg clutch Then, the period of time required for the eggs to develop at each temperature was quantified. The observations of egg hatching were made daily, by looking at the petri dish containing the eggs, without opening it. Nymphal development As soon as hatchlings were observed, they were confined individually in 50 ml snap-capped plastic vials half filled with wet autoclaved sand; the vials were labeled, identifying them by code, date when where oviposited, hatching date, and consecutive number of nymph. Because nymphs usually eat their exuviae soon after molting, the hatchlings were marked with acrylic paint by placing a dot on their pronotum, allowing identification when they molted. Once molted, a new paint dot was placed, repeating it after every molt. When the adult stage was reached, it was sexed and marked for followup purposes. By this means, the number of nymphs and the period between nymphs per temperature were determined. Results Scapteriscus vicinus Out of the 32 females put to mate, only 27 were selected, because two died and three devoured the male during the first two days, making it questionable whether they had mated.

PAGE 139

123 Oviposition, fecundity and fertility The ovipositions recorded by female (Table 43) showed that 57 egg clutches were oviposited by 27 S. vicinus, comprising 941 eggs that gave an average of 34.9 eggs/female during her lifetime. The average oviposition period estimated was as 61 days. The time estimated between clutches was 28.6 d. The number of eggs per clutch was reduced with the number of clutches per female (Figure 25), producing the maximum number of eggs at the first egg clutch, then declining at the second clutch. The second, third and fourth egg clutches contained a similar number of eggs, but the fifth (last) contained fewer. DAM= 35+2.1 58±5.9 74±7.2 100±12.0 89±0.0 1 2 3 4 5 Clutch number Figure 25. Number of eggs per clutch, time after mating, and accumulated percentage of eggs of Scapteriscus vicinus at 26°C. DAM = Days after mating±SEM The proportion of eggs that hatched per temperature is presented in Table 44, where is clear that the maximum egg survivorship occurred at 25°C (22.7%) followed by 30°C (16.5%) and 20°C (8.6%). No eggs hatched at temperatures below 20°C. In a

PAGE 140

124 separate experiment I exposed some eggs and adults of S. vicinus to 35°C with no survivorship observed of either eggs or adults. The time spent for egg development varied from 14.5 to 44.3 d, varying according to the temperature at which they were confined (Table 44). The unhatched eggs were kept two weeks at the same corresponding temperature, then transferred to a 26°C room and kept for two more weeks, and finally they were observed under the microscope to determine whether some development could be detected.

PAGE 141

125 Table 43. Fertility and oviposition period of Scapteriscus vicinus reared at 26°C and MC a 1 2 3 4 5 Eggs No. Ovipos n CODE F1 b fct I tz TO d e M 1 \J E3 b TO d ATO 6 E4 b TO d e ATO E5 b TO d e ATO Total Clutches Period v1 v1 1 1 1 U I O I D 9*% Zu I o 38 37 3 38 v2 v2 32 1 0 32 1 10 v3 v3 Zl ZD 9 1 Z I Do R1 0 I 9? Z O 1 2 93 1 4 27 120 79 4 120 v4 v4 ZD 9fi ZO I 0 1 1 T7 74 51 3 74 v5 v5 1 /I 14 ^7 14 1 37 v6 v6 99 z.Z OO 1 2 24 62 9 1 4 76 1 1 3 79 7 10 89 61 5 89 v7 v7 14 43 14 1 43 v8 v8 11 49 1 2 30 79 23 2 79 v21 v9 30 1 0 9 33 43 39 2 43 V22 v10 26 1 0 21 1 8 28 20 1 1 39 67 3 39 V23 V11 29 12 29 1 12 v24 v12 28 1 2 5 92 104 33 2 104 v25 v13 22 1 3 1 3 7 24 35 2 24 v26 v14 27 1 7 1 1 14 31 38 2 31 v27 v15 23 17 5 35 72 3 14 86 31 3 86 v28 v16 3 1 9 3 1 19 v29 v17 9 34 1 1 48 82 20 2 82 v30 v18 25 OO 9fi ZD 9fi ZD D*t 9 1 Z I 9^ ZO O/ 1 3 1 7 1 00 85 4 100 v31 v19 26 Zo Z4 -i 7 9 4Z 50 2 42 v32 v20 16 1 0 4 9 1 Z 06 -1 Q 9A ZU f o 50 3 73 v33 v21 28 54 6 31 84 4 1 2 96 38 3 96 V34 v22 28 57 12 9 66 40 2 66 V35 v23 32 59 32 59 v36 v24 6 66 6 66 v37 v25 5 68 5 68 v38 v26 9 73 9 73 v39 V27 20 74 20 74 SUM 941 57 MEAN 20 1 13.6 13.6 12.7 7 MEAN 34.7 28.1 57.5 17 3 73.6 15.7 99.7 10 89 16.5 2.1 61.0 a. Female mole cricket code b. Number of eggs per clutch. c. Time in days after mating. d. Time in days after previous clutch. e. Accumulated time in days after mating. Table 44. Egg fertility and development of Scapteriscus vicinus at constant temperatures. Temperature Fertility Development °C % Hatching Days 10 0 0 15 0 0 17 0 0 20 8.6 ±5.82 44.25 ±2.01 25 22.70 ± 7.09 19.31 ±0.69 30 12.50 ±4.89 14.50 ±0.65

PAGE 142

126 Development The hatchlings obtained at each temperature were transferred into a small vial filled with wet sand and kept at the same temperature. Their development was tracked per individual. No egg development occurred at temperatures of 17°C or below; although some eggs hatched at 20°C, all hatchlings died at first instar. Therefore, nymphal development was studied at 25 and 30°C only. Table 45 presents the developmental time of nymphs recorded from chambers at temperatures of 25 and 30°C. The hatchlings showed high susceptibility to 30°C, most dying after the first instar; only one individual developed until the ninth instar and then died without reaching the adult stage. At 25°C a few nymphs developed until the ninth instar and died. Vestigial wings were apparent from the 7 th and 8 th instars. The maximum development time recorded was 165 and 230 days at 30 and 25°C respectively. Table 45. Nymphal developmental time of Scapteriscus vicinus at two constant temperatures. 30°C 25 U C Instars n Per instar Accumulated n Per instar Accumulated 1 13 17.43±0.29 17.43±0.29 28 20.21 ±0.74 20.21 ±0.74 2 1 30 47 17 16.88±2.11 38.73±2.21 3 1 LD LD 11 18.27±2.37 54.27±5.11 4 1 19 66 9 23.33±3.18 77.77±7.38 5 1 19 85 8 22.38±0.94 104.25±7.49 6 1 28 113 7 20.57±0.72 121.29±7.91 7 1 18 131 5 22.40±2.77 140.83±10.76 8 1 25 156 4 40.50±4.97 177.40±17.68 9 1 8 164 2 40.50±20.5 230.50±27.5 Developmental times between instars ranged from 8 to 30 d at 30°C and from 18 to 40 d at 25°C. The development of the only nymph that developed at 30°C (Figure 26)

PAGE 143

127 showed no consistent trend. At 25°C duration of the first seven instars was similar, then the last two instars, 7 th and 8 th , required a longer time to complete development (Figure 27). 123456789 Instars Figure 26. Nymphal developmental time of Scapteriscus vicinus at constant 30°C in an incubation chamber. N=13 for 1 st instar; N=l for the other instars. 123456789 Instars Figure 27. Nymphal developmental time of Scapteriscus vicinus at constant 25°C in an incubation chamber.

PAGE 144

128 Scapteriscus abbreviatus After the mating period, 25 females survived and all were included in the experiments. The males were taken out of the oviposition chambers, keeping the females in their respective chamber. The starting time of the oviposition time was the mid-point of the week when females mated. Oviposition, fecundity and fertility The mean oviposition period was as 101.2 d, ranging from 28 to 190 d with 21.7 days between clutches. The total number of clutches oviposited by the 25 females was 1 19, averaging 4.76 clutches per female. The total number of eggs recorded from the 25 females was 3918, giving a mean of 156.7 eggs per female during its lifetime (Table 46). The average number of eggs produced per female per clutch is shown in Figure 28, where the mean number of eggs of the three first clutches comprised 50% of the eggs produced during each female's lifetime lifetime. The 50% were oviposited during the first 60 days after mating. The number of eggs per clutch declined with the clutch number (Figure 29).

PAGE 145

129 DAM=23 42 58 79 97 113 129 171 167 123456789 Clutch number Figure 28. Number of eggs per clutch (average) and days between clutch and corresponding proportion (%) of eggs per time of Scapteriscus abbreviates at constant 26°C. DAM= days after mating. The proportion of eggs hatching was strongly related to the temperature at which they developed (Figure 30). The maximum egg survivorship was recorded at 25°C (51.26%), followed by 30°C (48.72%), and finally at 20°C (20.88%). No development occurred at 17°C or below. The unhatched eggs were treated as for S. vicinus. The relationship between egg fertility and egg clutch number is not very clear (Figure 30), presenting strong variation from one egg clutch to the other with no significance among them. The only trend that was seen is that fertility of the first three clutches was higher than in the others.

PAGE 146

130 Figure 29. Proportion of eggs of Scapteriscus abbreviates hatching at constant temperatures. 120 Clutch number Figure 30. Proportion of eggs from Scapteriscus abbreviatus hatching by clutch number at constant temperatures.

PAGE 147

131 Development The hatchlings at each temperature were transferred into small vials filled with wet sand, confining them individually at the same temperature. No hatchlings from 20°C survived the first instar. Therefore, larval development reported is from two temperatures, 25 and 30°C (Table 46). Table 46. Nymphal development of Scapteriscus abbreviatus at constant temperatures 25°C 30°C Stadia Per instar 1 Accumulative 1 1 Per instar Accumulative 1 1 32.38 ± 0.60 32.39 ± 0.60 28.13 ± 1.03 28.13 ± 1.03 2 29.64 ± 2.46 50.41 ± 2.94 6.83 ± 1.72 39.56 ± 6.54 3 22.39 ±1.33 63.88 ± 3.18 13.00 ±3.59 76.26 ± 5.29 4 19.37 ± 1.31 79.04 ± 2.04 18.60 ± 1.69 102.20 ± 7.42 5 23.64 ± 1.24 102.50 ± 2.30 22.75 ± 2.35 121.00 ± 9.46 6 26.44 ±2.10 136.17 ± 5.18 18.30 ±1.93 140.60 ± 12.21 7 34.91 ±4.33 172.83 ± 8.46 28.78 ± 4.93 175.20 ± 15.03 8 30.29 ±5.19 164.17 ± 11.48 37.75 ±6.52 171.50 ±21.06 9 35.00 ± 0.00 189.00 ± 0.00 48.00 ± 0.00 191.50 ±35.50 l.Mean±SEM Nine instars were recorded during S. abbreviatus development. Time between instars ranged from 19 to 35 d at 25°C and from 6 to 48 at 30°C. The accumulated mean time required to pass through the nine instars at 25°C was 189 d, and 191.5 d at 30°C. The developmental times varied among the instars, the intermediate instars tending to develop in a shorter time than those at the extremes. This tendency was observed at the two temperatures evaluated (Figures 3 1 and 32). Fifty percent of the nymphal development was reached by the 5 th instar at 25°C and by the 3 rd to 4 th instars at 30°C.

PAGE 148

132 123456789 In stars Figure 32. Nymphal development of Scapteriscus abbreviates at constant 30°C. The survivorship observed at the two temperatures was very low considering the total development from egg to adult. Thirteen individuals (out of 898 eggs; 498 hatchlings) reached adult stage at 25°C and 9 (out of 737 eggs; 384 hatchlings) at 30°C. The low adult proportion may have been due to the intense handling by applying the

PAGE 149

133 paint dot and constant shaking of the nymphs when recording the molting time (observing the paint dot present or absent). Data from the few survivors (Table 47) show the time to develop from egg to adult, and sex ratio. The adult stage was reached from three instars, the 7 th instar being the most common. The overall development at 25°C was 188.9 d and at 30°C was 202.7 d. The overall sex ratio was female-biased at 25°C and male-biased at 30°C Table 47. Developmental time of Scapteriscus abbreviates from egg to adult, grouped by number of instars to complete development and sex ratio per group. 25°C 30°C Instar No. Time (d)' Sex ratio 2 No. Time (d) 1 Sex ratio 10 184.0±16.9 6:4 5 208.4±5.3 2:3 8 th 2 213.5±30.5 2:0 2 199.5±6.5 0:2 9 th 1 189.0±0.00 0:1 2 191.5±35.5 1:1 Overall 13 188.9±13.7 8:5 9 202.7±7.06 3:6 1 . Mean days±SEM 2. Female: male Discussion Scapteriscus vie in us S. vicinus females were able to oviposit up to 5 clutches with an overall average of 2.1 clutches per female with a mean number of 34.9 eggs per female. The most evident reduction in number of eggs per clutch occurred at the fifth clutch The overall mean oviposition period was as 61 days. Forrest (1985) reported up to 10 egg clutches per female (without specifying mole cricket species, either S.vicinus or S. borellii) Braman (1993) reported up to 8 egg clutches for S. vicinus females and up to seven clutches for

PAGE 150

134 S. borellii. Braman (1993) reported an average of 3.1 clutches per S. vicinus female with 45.8 eggs total per female and an average of 2.4 clutches with 58.2 eggs for S. borellii females. Meanwhile, Worsham and Reed (1912) cited in Braman (1993) reported a range of 29 to 60 eggs total per female S vicinus. Hayslip (1943) reported an oviposition period of29to38days. There are several disagreements among the reports regarding the oviposition description of S. vicinus. Part of the cause of this variability might be the temperatures at which the studies were run. The results of the work presented here considered the oviposition process at constant temperatures in incubation chambers. Other authors studied the oviposition process at field temperatures, which may have been subject to the naturally fluctuating temperature regimes at varying latitudes. The eggs of S. vicinus developed at temperatures ranged from 20 to 30°C, failing to develop at lower temperature than 20°C. Fecundity and fertility were strongly affected by the temperatures at which they were reared, with the maximum values at 25°C, which is an indicator that this temperature is close to the optimum temperature for egg development. Braman (1993) reported egg development required 2-3 weeks at 27°C, that is close to that observed in this work between 25 and 30°C (19.3-14.5 d). Forrest (1985) reported that fertility decreases with egg clutch number; the later clutches a female produced had lower percent hatch which agrees with the observations of this work (Figures 3 1 and 32). The mean number of eggs per female was as 16.1, ranging from 3 to 80 during its lifetime. The maximum fertility observed was 22.7%

PAGE 151

135 Nymphal development was strongly affected by the ambient temperature. Development of more than one nymph was recorded only at 25°C, showing that the immatures of this species are very susceptible to the temperatures at which they live. Even though several nymphs reached the ninth instar, none of them reached the adult stage. Most of the 8 th instars developed vestigial wings but fail to fully develop, which I attribute to the experimental conditions at which they were reared, because all the older nymphs died during molting. The possibility that the ninth instar nymphs were ready to transform to the adult stage is reinforced by Braman (1993) who observed individuals reaching adulthood at 8 th and 9 th instars. Scapteriscus borellii (=acletus) was reported to reach adulthood at 7 l to 10 th instars by Hudson (1987). S. abbreviatus also reached adulthood at 7 th to 9 th instars (see S. abbreviatus results in this account), indicating a standard range of instars for all three mole crickets to complete development to adulthood. The total development of Scapteriscus vicinus from egg to the 9 th nymphal instar was estimated as 250 days, considering egg development (19.4 d) at 26°C plus nymphal development (238.5d) at 25°C. Nevertheless, development time varies at temperatures between 20 to 30°C. Hayslip (1943) reported (what he called "fall adult") nymphal development of 135 d plus 18 d of egg development (= 154 d total). Braman (1993) reported the nymphal development as 23 32.5 weeks (= 161-227 d). Hudson (1987) reported 223 d for total development (from egg to adult) for S. borellii (=acletus). Advances in the study of the life cycle of Scapteriscus vicinus were made, establishing possible temperature boundaries for development, and more specific information was obtained for S. vicinus developing at 25°C.

PAGE 152

136 Scapteriscus abbreviatus Results with this mole cricket species were more consistent, with several individuals reaching adulthood at the two temperatures where development occurred. There is no published information about general biology or life cycle development of this mole cricket species to compare. The lack of this information is explained by the fact that S. abbreviatus populations are restricted to a small area, close to the original areas where they arrived in the early 1 900s. The lack of long wings prevents this species from dispersing readily. Both species S. abbreviatus is highly fecund, ovipositing more than twice the number of clutches as does S. vicinus (Tables 43 and 44) and the proportion of eggs hatched per clutch and per temperature was greater in S. abbreviatus than S. vicinus (Figures 33 and 34). The time to develop from egg to adult was shorter than in S vicinus (Tables 45 and 46). 90 -I 1 23456789 Clutch number Figure 33. Number of eggs hatched per clutch of two mole cricket species, Scapteriscus abbreviatus and S. vicinus at constant 26°C.

PAGE 153

137 Temperature °C Figure 34. Proportion of eggs hatched of two mole crickets, Scapteriscus abbreviatus and S. vicinus, at constant temperatures. S. abbreviatus developed as did S. vicinus, but with greater success in reaching the adult stage, perhaps due to this species being already in culture, and thus laboratoryadapted (at 26±1°C), or due to its higher fitness and reproductive capacity. That fewer individuals reached adulthood at 30°C than 25°C, which coincides with the higher overall survivorship at the same temperature reinforces the possibility that 25°C is about optimal for S. abbreviatus development. Another clue supporting it is the lesser time to develop at 25°C than 30°C, which indicates the latter temperature surpasses the optimum, slowing down the development (requiring more time to complete development) (Wagner et al. 1984a). That development of the two mole cricket species occurred at only a narrow range of temperatures makes modeling difficult, but it established temperature constraints. Future works should include intermediate temperatures between 20° and 30°C, and possibly some higher than 30° but less than 35°C.

PAGE 154

138 Nymphal development of S. abbreviates was well described and might give some clues to make culture more efficient. For example, the clutch number could be used to eliminate females once they have oviposited 75% of the expected oviposition total, saving time and money. Modeling the egg development would enhance the probability of predicting successive nymphal events. For example, the time spent at 20°C was 44 d which, coupled with the estimated 50% of flying females in the spring, helps to predict when nymphal stages will be present (something very hard to monitor in the field).

PAGE 155

APPENDIX B PARASITISM OF MOLE CRICKETS BY THE WASP LARRA BICOLOR Introduction Larra spp. wasps have been identified as specialist natural enemies of mole crickets in the tropics and subtropics (Adamson 1942, Castner 1983, 1988, Frank et al. 1995, Frank and Parkman 1999, Menke 1992, Williams 1928); they use particular species of mole crickets as hosts for individual ectoparasitic larvae; after 10-12 days mature larvae devour their weakened hosts (Castner 1988). Species of Larra were used in the earliest attempts at classical biological control of mole crickets, in Hawaii in the 1920s and Puerto Rico in the 1930s (Frank et al. 1995). No analyses of their effect on mole cricket populations on these islands appear to have been published (Castner 1983, 1988, Frank 1994, Frank et al. 1995). Adults of L. bicolor are specialist parasitoids of large nymphs and adults of pest mole crickets of the genus Scapteriscus (Castner 1984, 1988, Frank and Parkman 1999, Frank et al. 1995). In the early 1980s L. bicolor wasps were released at five Florida sites in conjunction with plots of their preferred nectar-source plant, Spermacoce verticillata L. (Frank et al. 1995). At that time only wasps released at one of the sites (in Fort Lauderdale) established viable populations (Frank et al. 1995). Castner (1988) reported 0% parasitism of Scapteriscus vicinus Scudder mole crickets (and <1% of S. abbreviatus Scudder) trapped at Fort Lauderdale during the period April 1985 through December 1987, leading him to question the value of the wasp in a biological control program 139

PAGE 156

140 against mole crickets. In the late 1980s a different, perhaps more cold-tolerant strain of L. bicolor was imported from Bolivia, released in northern Florida (approximately 29°40'N), and a decade later appeared not only to be established but also to be spreading (Frank and Parkman 1999), indicating at least that mole crickets in some unknown proportions were being parasitized. Again, no follow-up evaluation has ever been made, either of parasitism efficiency or of geographical spread of the wasp (Frank and Parkman 1999). Noting the lack of data on percent parasitism, at the same time that the wasp is being touted as a major player in integrated pest management of mole crickets in the southeast (Frank and Parkman 1999), an objective was set to document percent parasitism of pest mole crickets by Larra bicolor in the field. A second objective was to analyze for an effect of mole cricket size on percent parasitism. Materials and methods In the summer of 1999, 3 pitfall traps were established at each of 2 sites: the University of Florida Horse Teaching Unit (HTU), and the Department of Fisheries and Aquatic Sciences (FAS) field laboratory situated «10 km almost directly north of the HTU, in northwest Gainesville. A row of 25 S. verticillata plants had earlier been established at each of the sites. Each trap, a modification of the design of Lawrence (1982), consisted of a 20-liter plastic bucket in which four 7.5-cm holes had been drilled every 90° around the top edge; four 7.5-cm dia. PVC pipes, each 3 meters long and with a 2.5-cm slit cut from their entire length, were fitted into the holes. Traps were checked weekly from 2 September to 30 December 1999 and from 14 September to 7 December 2000, periods of 18 and 12 weeks respectively, and the

PAGE 157

141 samples processed as described above. For the 2000 samples, only mole crickets trapped through 30 November were considered in analysis because an early hard freeze (air temperature = 4°C) marked the end of the wasp season; Spermacoce plants were killed at that time and wasps were observed dead on the ground. Pronotal measurements were grouped into size classes of 0.5and 1 .0-mm intervals for statistical analysis. Proportions of mole crickets parasitized were compared for significant differences between the 2 sites, among total mole crickets trapped per location per year, among size classes of nymphs and between nymphs and adults, and between the 2 years of the study using chi-square tests (Snedecor 1 946). Only those samples that included at least 1 parasitized mole cricket, regardless of stage, were used for comparison of total mole crickets against percent parasitized, and only size classes with more than 10 individuals were included in analysis of that parameter, to conform to chi-square requirements. Analysis of the significance of mole cricket stage considered only samples that included both nymphs and adults; samples lacking one or the other life stage were excluded. The success of the wasp in parasitizing the different mole cricket sizes explained before, was evaluated by tracking the developmental success of the parasitic stages collected from pitfall trap captures, and compared with success of parasitic stages in laboratory trials. The success of the parasitic stages was defined as the parasitic larva reaching the pupal stage, which implies the wasp grub kills the host by devouring it and then building a cocoon. Another way to evaluate the wasp's success was by comparing the proportion of wasps reaching the adult stage (adult emergence). Other parameters

PAGE 158

142 describing the wasp's success were the development time, sex ratio, and the possible interaction between molting nymphs and parasitoid success in killing the host. The proportions of each the three cohorts were compared by the chi-square test and development times were compared by the t-test (Snedecor 1946). The three cohorts evaluated are described as follow: a) Field-parasitized hosts. 40 (nymphs and adults) mole crickets bearing a wasp parasitic stage. These hosts were parasitized by wasps in the field. The parasitized hosts were captured in pitfall traps (described above). The age of the stages (eggs and larvae) attached to mole crickets was unknown, b) Laboratoryinduced parasitism (Lab.l). A cohort of 32 nymphal mole cricket hosts (in which molting had occurred sometime in the five days before the test) was selected. The hosts were exposed to laboratory-reared wasps. Parasitized hosts were confined as in the field cohort, c) Laboratory-induced parasitism (Lab. 2) A cohort of 37 (close to molt) nymphal mole cricket hosts was used. Their last molt was at least 1 5 days before the test. The reason for use of the two laboratory cohorts was based upon previous experiences rearing mole crickets, where molting of medium to large nymphs occurred every 15 to 30 days. In this way the first laboratory cohort did not face the risk that the parasitoid would be shed along with the exuviae. The developing parasitoids in the second laboratory cohort had a high chance of encountering molting by the host. This second laboratory cohort would provide a good insight of the wasp's success while developing on molting nymphal host stages. The three cohorts were kept under the same laboratory conditions, 26 ±0.5°C, 14:10 (D:L) photoperiod, 60% RH. The containers were filled with autoclaved sand, kept wet by adding tap water once a week. While alive, the hosts were fed twice a week with

PAGE 159

143 Purina cricket chow. The transparent walls of the containers allowed assessment of parasitoid development, pupation, and wasp emergence. The differences in sizes between sexes of Larra bicolor were evaluated from the pupae and adults obtained from those field-parasitized mole crickets captured from pitfall traps. Pupae were measured alive, or their exuviae were measured. Adults were marked with paint at emergence, to record association with the pupae from which they had emerged, and were measured after death. The variables considered were adult wasp: body length, left forewing span, and thoracic width; pupa: length, and maximum width (of the thorax). All variables were measured in millimeters. The variables were sorted by sex and analyzed independently, and by coupling corresponding pupal with adult dimensions. The statistical analyses were linear regression, t-test comparison, and 99% confidence intervals of the mean per variable between sexes. Results Percent parasitism In total, 195 mole crickets were trapped at HTU during the 2 seasons of the study, with a total parasitism of 24.1%. At FAS, 204 mole crickets were trapped during the 2 seasons; 18.1% of these were parasitized (Table 48, Figure 35). There was a significant difference (p <0.10) in the percent parasitism of mole crickets trapped between the 2 sites in 1999, with 22.3% (n = 31 of 139 individuals parasitized) at HTU versus 12.6% (n = 1 1 of 87 parasitized) at FAS, but in 2000 percentages of 28.6 (n = 16 of 56) at HTU and 22.2 (n = 26 of 1 17) at FAS were statistically the same.

PAGE 160

144 Table 48. Numbers of Scapteriscus vicinus mole crickets, and numbers parasitized by the parasitoid wasp Larra bicolor, trapped in pitfall traps at 2 locations during 2 seasons in Gainesville, Florida. Horse Teaching Unit (HTU) Fisheries & Aquatic Sciences (FAS) No. mole Number % No. mole Number % Year crickets parasitized parasitized crickets parasitized parasitized 1999 139 31 22.3 87 11 12.6 2000 56 16 28.6 117 26 22.2 Totals 195 47 24J 204 37 18.1 The parasitism proportion at FAS recorded during the two years was statistically different (p <0.10), because it was almost twice in 2000 that of 1999. No difference was observed between years at HTU. Data for the two years and two locations were pooled before testing for a difference of mole cricket sex on percent parasitism, but no difference was found. Over the 2 years of the study, pronotal lengths of trapped mole crickets ranged from 5.4 to 9.15 mm (mean = 7.63 mm, n = 21 1) for nymphs and 6.9 to 9.55 mm (mean = 8.32 mm, n = 187) for adults. The average pronotal length of adult males (mean = 8.42 mm, n = 124) was significantly greater than that of females (mean = 8.10, n = 63; twosample t-test, p «0.01). The adult data are compatible with ranges of 7.0 to 10.0 (mean = 8.3) mm for adult males and 6.6 to 8.8 (mean = 7.9) mm for adult females reported by Forrest (1987). There was no evidence that larger nymphs (>7.5 mm) were more heavily parasitized at either site in either year, and there was no evidence that adults were parasitized in higher proportions than were nymphs (Fig. 36). Mole crickets were thus grouped together regardless of size or stage for further analysis. Overall parasitized nymphs were compared with overall parasitized adults. There was a statistical difference during 1999 at both places, but 2000. Pooled data from the two places and the two years showed no difference.

PAGE 161

145 Fisheriesand Aquatic Sciences, 1999 Fisheries a nd Aquatic Sci 9/2 9/16 9/30 10/14 10/28 1 1/11 11/25 12/9 12/23 9/14 9/28 10/12 10/26 11/9 11/23 12/7 collection date collection date Horse Teaching Unit. 1999 Horse Teaching Unit, 2000 z 25 1 20 o E 15 o « 10 III... [/ paras itized Bunparasibzed ll.ll.. 10/14 10/28 11/11 collection date 12/9 12/23 25 20 15 ^parasitized fjunparasitized l.-i 9/14 9/28 10/12 10/26 11/9 11/23 12/7 collection date Figure 35. Numbers of parasitized and unparasitized Scapteriscus vicinus mole crickets trapped in pitfall traps at 2 locations during 2 seasons in Gainesville, Florida, by date of collection. Fisheries and Aquatic Sciences, 1999 Fisheriesand Aquatic Sciences, 2000 pa ra sitized Jul.. 9/2 9/1 6 9/30 10/14 10/28 11/11 11 /2 5 12/9 12/23 collection dale 9/14 9/28 10/12 1 0/26 1 1/9 11/23 12/7 collection date Horse Teaching Unit, 1999 R para sitized I unp a tasitize d liL.Ui. 9/2 9/1 6 9/30 10/14 10/28 11/11 11/25 12/9 12/23 c ol le ctlon d a te Horse Teaching Unit. 200 0 pqp a ta sitized pju npa ra sitize d 9/14 9/2 6 10/1 2 10/26 11/9 11/23 12/7 c o Me ctio n d a te Figure 36. Numbers of nymphs and adults of Scapteriscus vicinus mole crickets, and proportions of each parasitized by the parasitoid wasp Larra bicolor, trapped in pitfall traps at 2 locations during 2 seasons in Gainesville, Florida, by pronotal size class.

PAGE 162

146 Parasitoid success There were no differences among the three cohorts in the proportions of wasp success in killing the host, wasp emergence, wasp sex ratio, unrelated host death, and development time overall and by sex (Table 49). In the Lab. 2 cohort, 19% of the developing parasitoids were shed along with the molt of nymphal hosts. All the mole cricket nymphs that molted survived to the adult stage. Table 49. Success of Larra bicolor parasitizing Scapteriscus vicinus. COHORTS SUBJECT FIELD LAB. 1 LAB. 2 OVERALL Parasitic success 1 93.3 a 100.0 a 100.0 a 97.6 Parasitoid died 2 6.6 0.0 0.0 2.4 Unrelated host death 3 25.0 a 15.6 a 13.9 a 18.5 Wasp adult emergence 4 50.0 a 59.3 a 48.0 a 52.4 Wasp Sex ratio 0.53 : 0.46 0.37 : 0.63 0.5 : 0.5 0.47 : 0.53 Parasitoid shed by host molt 5 0.0 0.0 19.3 7.3 Parasitoid development (egg to adult) in days. Overall 130.0 155.1 a 148.3 a 144.5 Females 132.0 146.1 a 151.8 a 142.2 Males 128.0 162.1 a 144.8 a 146.5 1 . Wasps that reached the pupal stage. 2. Proportion not including unrelated host death. 3. Hosts that died for unknown reason, not attributable to development of the parasitoid. 4. Proportion considering all wasp pupae. Not considering parasitoid death nor unrelated host death. 5. Eggs and larvae that were shed with the host exuviae when molting. Means followed by same letters are not statistically different (Duncan's test). Comparison by subject. Sex dimorphism Thirty-four adult wasps emerged out of the total field-parasitized mole crickets captured from pitfall traps, and were 16 males and 18 females. They, and the pupae from which they developed, were measured. The t-test comparing all five variables between wasp sexes gave highly significant differences between all comparisons, indicating a good separation of mean measurements (Table 50). Regressions of pupal dimensions with

PAGE 163

147 corresponding (coupled) adult dimensions indicated no significant linear relationship between any combination per wasp sex nor overall. Table 50. Sex dimorphism in adult and pupal stages of Larra bicolor developed on naturally field-parasitized host mole crickets. Adult wasp Pupal wasp Wasp Thorax 3 Wing" Body c Length" Width 6 Female* 3.89±0.2a 12.88 ± 0.6 a 18.55 ± 0.9 a 20.74 ± 1.1 a 8.38±0.4a Male' 3.31±0.1b 11.62 ± 0.4 b 1402 ± 0.6 b 1 1801 ± 0.7 b 7.51±0.3b a. Wider segment of the pronotum (mm). b. Left forewing span (mm). c. Body length from the head to the tip of the abdomen (mm). d. Maximum length of the unhatched pupa (mm). e. Maximum equatorial diameter of the pupa (mm). f. Mean±99%CL. Means followed by different letters between sexes are statistically different; 't' test.. Despite the clear separation of mean measurements, there was overlap in size ranges of the two sexes. The pupal lengths of 37 % of the males overlapped with those of females, and the pupal widths of 68% of the males likewise overlapped (Fig. 37).

PAGE 164

148 25 20 FEMALE 15 E E 68% MALE / MALE 10 FEMALE PUPAL LENGTH 5 PUPAL WIDTH 0 5 10 15 20 25 mm Figure 37. Pupal length and pupal width comparison between sexes of Larra bicolor. Circles show the proportion of overlapping sizes between sexes. Because of overlap, pupal dimensions did not provide a highly reliable means of differentiating sex of adults. When a wasp emerges, a portion of the abdominal apex of the pupal exuviae is lost, making it impossible subsequently to measure the length. The true mortality of mole crickets subjected to parasitism by L. bicolor can be shown to be much higher than the average 24% reported for the 13-week period ending 7 December 2000 at both sites. Development time of L. bicolor, from egg through larval stages, is three weeks (Castner 1983), and there is a continuous overlapping of parasitoid generations. However, Scapteriscus vicinus is univoltine (Walker 1985b). Consider the following analogy: if a regiment of soldiers is subject to 24% casualties during any 3week period of a war, what will be the total percentage attrition of the regiment after 1 3 weeks assuming no reinforcements? The answer is about 70% for the soldiers as for S. Discussion

PAGE 165

149 vicinus, calculated as 100 (l-(l-0.24) 4 33 ) where 4.33 is 13/3. This high level of mortality nevertheless excludes mortality occurring in other months. Juvenile stages of S. vicinus occur in north Florida from early May until December, reaching their maximum numbers during July and August. Data on the number of nymphal stadia are inconclusive at present; Walker (1985b) reported 6 to 7, Braman (1993) reported 10. Adults start appearing in late August, reaching their maximum numbers during late November and December. This species overwinters as large nymphs (15%) and adults (85%) (Hayslip 1943, Walker 1985b). Limiting parasitism only to adult mole crickets would provide the wasp a narrow window of opportunity for success; wasps would have largely died out at the time of adult population peak. Parasitizing nymphs as well as adults enable the wasp to increase generation turnover and population numbers, but with risks for wasp life cycle success. Although 8 1 % of eggs and larvae were observed to reach the pupal stage (so killing the host) while developing on ready-to-molt hosts (Lab. 2 cohort), 19% did not succeed. In at least four cases the parasitic larvae were able to remain on the host after it molted, and in one occasion a medium-large wasp larva was shed along with the host exuviae. The high wasp success, 93 to 100% (97% overall), in reaching the pupal stage indicates the very high efficiency of this parasitoid in killing the host, and the 48 to 59.3 (52.4% overall) of wasp emergence supports it. Increased percent parasitism during the second year at one of the study sites (FAS) suggests that L. bicolor increased its population there. It was believed that small or medium host nymphs might be limited in their capacity to provide enough nutrients for wasp grub development; the nymphs' own development is further limited by the presence of the grub (Castner 1988). Also, small

PAGE 166

150 nymphs might be more vulnerable to predation, causing the loss of any attached eggs or grubs. However, this work shows not only that nymphs as small as 6.5 mm pronotal length were parasitized in the field, but that development of wasp grubs occurred successfully on such small hosts. No difference was seen between development time of males and of females (Table 24). This does not match Castner's (1988) observations; he considered that males develop in a shorter period of time than do the females. Half («50%) of the wasps that reached the pupal stage emerged in a period of 152.2 d. The remaining half did not emerge and died, which can be attributed to inadequate laboratory conditions to keep them alive, or to the lack of a necessary environmental stimulus to emerge. One of the most difficult problems in studying Larra bicolor development is the prolonged time for the pupa to develop, causing delay and inaccuracy in knowing the wasp's sex, one of the most important items in evaluating the parasitoid's performance as a biological control agent. In order to solve this problem, the wasp sexes were characterized by their dimensions. The pupal length and width may be used as a method of determining the adult's sex, but this is not entirely reliable because of overlap in size ranges. The body length of the adult correlates well with its sex, but is less accurate than dissection of the genitalia. The problem to wait such a long time for the adult to emerge and then know its sex can be solved by taking the pupal dimensions (length or width) that can predict the wasp sex.

PAGE 167

APPENDIX C SOIL TEMPERATURE ESTIMATE Problem description During developmental studies, the two parasitoids of mole crickets, the tachinid fly Ormia depleta and the digger wasp Larra bicolor, were kept underground at field temperatures in the natural habitat of their host. The soil temperatures were constantly recorded every 12 min at a range of soil depths. During 1998, 1 used two dataloggers that stopped working very often. In June 2000 a digital datalogger was purchased, which has four thermocouples to record the same number of temperatures at the same time. Unfortunately one thermocouple did not work and the datalogger recorded temperature from three points, two underground and one from air. The soil depths where temperature was recorded were 7.5 and 22.5 cm, and the air temperature was recorded at 1.8 m. Recordings were made every 30 min. In October of the same year, a fourth thermocouple was purchased along with another datalogger with the capability to record temperature from one point only. The four-point recording temperature datalogger was set to record temperature at four soil depths, 7.5, 22.5, 37.5, and 52.5 cm respectively. The single thermocouple datalogger was set to record air temperature at 1.8 m. Both dataloggers were set to record temperature every 30 min. The only problem with the four thermocouple datalogger was that it stopped working twice due to a loose battery connection. The dates when temperatures were not recorded are from 1 6-23 August, from 29 September to 1 1 October, and 2-4 November. 151

PAGE 168

152 Thus this problem caused the lack of those temperature recordings during the development of the parasitoids underground. Therefore, it was necessary to estimate soil temperatures from recordings from the meteorological station nearest to the location of the experiments. Fortunately, the Florida Automated Weather Network (FAWN) has a station less than a mile from the experimental plot. The FAWN records soil temperature at 15 cm depth underground and air temperature at 1.8 m, among other conventional recordings. The FAWN records are taken every 1 5 minutes. The FAWN soil temperature is recorded at an intermediate depth of my experimental plot, where thermocouples were set at 7.5 and 22.5 cm, which means FAWN was set at the middle depth of my experiment. I compared the FAWN soil records with my own that were recorded at the same time. This provided an intermediate set of records (15 cm) between my uppermost depths (7.5 and 22.5 cm), which I took to be the expected result. The FAWN records followed the trend of the temperature recorded from the experiments, but failed to match the precise values of my data (Figure 38).

PAGE 169

153 0123456789 10 Days Figure 38. Temperatures recorded during 10 days in July from the soil in the experimental area (Exp.) at two depths (7.5 and 22.5 cm.) and from soil (15 cm depth) recorded by the Florida Automated Weather Network (FAWN). I built a mathematical model to obtain a better estimate of the temperatures at the study site during the periods when the experimental data were lacking. Materials and methods Temperature estimates. The temperature estimate was performed by applying stepwise linear regression analysis. The independent variable was the temperatures recorded by the Hobo datalogger in the experimental area; the dependent variables were soil and air temperatures from FW. The models evaluated, adding the coefficients in delayed times, are shown in Table 51.

PAGE 170

154 Table 5 1 . Linear equation models proposed to estimate soil temperatures. Model Description AM J 1 1 Model l Y 7.5 _ i u, 1 /c \ a + bl(Sto) Model 2 Y 7 .5 = a + bl(S, 0 ) + b2(S,.,) Model 3 Y7.5 = a + bl (S t0 ) + b2 (S t -i) + b3 (S t-2) Model 4 Y7.5 = a + bl (S t0 ) + b2 (S ,.,) + b3 (S t2 ) + b4(S t-3) Model 5 Y7.5 = a+ bl (S t0 ) + b2 (A t0 ) Model 6 Y7.5 = a+ bl (S, 0 ) + b2 (A, 0 )+ b3 (S t ,)+b4 (A t l) Model 7 Y7.5 = a+bl (S t0 ) + b2 (A t0 )+ b3 (S,. 0+b4(A t . i)+b5 (S, . 2 )+ b6 (A t 2 ) Model 8 Y7.5 = a+ bl (S, 0 ) + b2 (A t0 )+ b3 (S ,. i)+b4(A,. .)+b5 (S, -2)+ b6 (A ,. 2 )+b7 (S t . 3 )+ b8 (A t 3 ) Where: Y 7.5 = Temperature recorded at the experiment at 7.5 cm depth underground. S = Soil temperature from Florida Automated Weather Network (FAWN) database A = Air temperature from Florida Automated Weather Network (FAWN) database to = Temperature recorded at time zero t-1 = Temperature delayed one time step ( 30 min). t-2 = Temperature delayed two times step ( 30 min). t-3 = Temperature delayed three times step ( 30 min). Those models were run including either 500, 1000, or 2000 data entries; each data entry was the temperature recorded each 30 min underground in the experimental area. The linear equation obtained by running the first data set was then run to another data set, a group of data recorded at various dates, not included in the calculation of the equation, so that the equation should mimic the new temperature data set. The predicted data were plotted with the new temperature data, thus evaluated by two ways, first the predicted values should follow the experiment's data trend, and second the predicted data should match the values of the experiment's new data. Their R 2 values and the significance of their coefficients evaluated the proposed models. The equation that better fit those criteria would be selected to estimate the missed data. The data set that was used to build and evaluate the models consisted of the data next to the periods where the experimental data were not recorded. Therefore, the equation was built using the data previous to the empty period, and then the validation

PAGE 171

155 was run with data after the same empty period. It was assumed that one equation would be required per soil depth to estimate the missed data of the two different periods. Results Before starting to build the temperature model to estimate the missed data, the estimation procedure was first applied to all the known and contiguous series of data. This was intended to build and to validate the model working with known data, which are contiguous in time. To accomplish this trial, the July data set was chosen. The proposed models calculated include 500, 1000 and 2000 data points. They were plotted along with independent and dependent variables, so that their trends and values were compared. The models including soil temperature as the only independent variable are depicted in Figure 40, where the model 1 (Figure 39, A) appeared to be the most variable, particularly when including 1000 data points; the other three models (Figure 39, B, C, and D) showed very similar predicted outputs when using any of the input (500, 1000, 2000) levels. The comparison of the regression coefficients of the models considering the soil temperature as the only independent variable is shown in Table 52. The R 2 value of model 1 is clearly lower than that of the other three models. The R 2 values for models 2 to 4 are very similar. The regression coefficients of model 1 are all statistically significant. The coefficients of model 2 were all significant when using 500 and 1000 data points, but one coefficient was not significant with 2000 data points. Models 3 and 4 presented at least one coefficient not significant with the three (500, 1000, 2000) levels of data points, becoming less significant as it included more coefficients.

PAGE 172

156 Figure 39. Efficiency of soil temperature models to estimate temperature at 7.5 cm depth. Independent variable: Soil temperature from the experimental area (Exp. 7.5) at 7.5 cm depth. Dependent variable: Soil temperature (FAWN -soil) from the Florida Automated Weather Network. Table 52. Coefficient of determination values and statistical significance (p-values) of slope and intercept for four models estimating soil temperature, using 500, 1000, or 2000 data points. Model 1 Model 2 Model 3 Model 4 500 1000 2000 500 1000 2000 500 1000 2000 500 1000 2000 R 2 0.66 0.77 0.81 0.70 0.80 0.81 0.71 0.81 0.81 0.73 0.81 0.81 a 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 b1 0.001 0.001 0.001 0.001 0.001 0.001 >0.05 0.001 0.05 > 0.05 0.001 0.05 b2 0.001 0.001 > 0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05 b3 0.001 0.001 > 0.05 >0.05 >0.05 >0.05 b4 0.001 0.001 0.05

PAGE 173

157 The temperature estimates by the proposed models including soil an air temperature as the independent variables are depicted in Figure 40. A 4-day output is shown because the lines become difficult to discern visually with longer outputs. The outputs in Figure 40 show how similar the predicted temperatures are among the three levels of input. Model 1 (Figure 40, A) appeared to be the most variable, particularly when including 1 000 data points. The other three models (Figure 40, B, C, and D) showed very similar predicted outputs at the three levels of input. Comparison of the regression coefficients of the models considering soil and temperature as the independent variables is shown in Table 53. The R 2 value of model 1 is clearly lower than for the other three models. The R values for models 2 to 4 are very similar. The regression coefficients of models 1 and 2 are all statistically significant. Most of the coefficients of models 3 and 4 are not significant for the three data input levels, becoming getting less significant as they include more coefficients.

PAGE 174

158 42 38 34 a 30 E 0) H 26 22 Model 7 Exp . 7.5 cm Model 6 E xp. 7.5 cm 42 38 o o Q) 34 IS i01 Q. 30 E 0) l26 22 Model 8 Exp. 7.5 cm 2000 2 Days Figure 40. Efficiency of soil temperature models to estimate temperature at 7.5 cm depth. Independent variable: Soil temperature from the experimental area (Exp. 7.5) at 7.5 cm depth. Dependent variable: Soil temperature (FAWN-soil) and Air temperature (not shown) from the Florida Automated Weather Network. Table 53. Coefficient of determination values and statistical significance (pvalues) of intercept and slope of four models estimating soil temperature with soil and air temperature as independent variables, using 500, 1 000, or 2000 data points. Model 5 Model 6 Model 7 Model 8 500 1000 2000 500 1000 2000 500 1000 2000 500 1000 2000 R 2 0.71 0.82 0.84 0.72 0.83 0.86 0.73 0.83 0.87 0.74 0.84 0.88 a 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 b1 0.001 0.001 0.001 0.001 0.001 0.001 > 0.05 >0.05 > 0.05 > 0.05 >0.05 > 0.05 b2 0.001 0.001 0.001 0.001 0.05 0.001 > 0.05 >0.05 > 0.05 > 0.05 >0.05 > 0.05 b3 0.05 0.05 0.001 > 0.05 >0.05 > 0.05 > 0.05 >0.05 > 0.05 b4 0.05 0.001 0.001 > 0.05 >0.05 > 0.05 > 0.05 >0.05 > 0.05 b5 > 0.05 >0.05 > 0.05 > 0.05 >0.05 > 0.05 b6 > 0.05 >0.05 > 0.05 > 0.05 >0.05 > 0.05 b7 > 0.05 >0.05 > 0.05 > 0.05 >0.05 > 0.05 b8 > 0.05 > 0.05 > 0.05 > 0.05 >0.05 > 0.05

PAGE 175

159 Temperature estimate of missed data The differences between models 1 and 2 were not quite evident, considering soil alone or with air as independent variables. Regardless of input level (500, 1 000, or 2000 data points), the predicted temperatures were very similar (except model 1 with 1 00 data with soil and air variables). The more apparent differences were the R values that were higher in models 2 and 6 than in models 1 and 5, with all coefficients statistically significant. Models 3 and 4 models 7 and 8, although producing similar predicted outputs, showed no significance among their coefficients, with mostly similar R 2 values compared with models 2 and 6. The choice of the best model to estimate soil temperature lay between models 2 and 6, with 500 or 1000 data points. Considering the higher R values, the statistical significance of all coefficients and the not very complicated data management required to achieve a good temperature prediction, model 2 was chosen, with 1 000 data points. Then this model was applied to estimate the temperatures during those periods when data from the experimental area were lacking. The model building was performed using the temperatures recorded in the experimental area at 7.5 cm depth from 5 July to 16 August 2000. The data are shown in Figure 42, where is evident the disparity between the temperature in the experimental area at 7.5 cm depth and the FW recordings. The similarity of temperatures at 22.5 cm depth and those from FW are evident in the same Figure 41 .

PAGE 176

160 40 Figure 41. Temperatures recorded at two depths from the experimental area (Exp.) and from the alternative temperature record source (FAWN). Period from 5 July 16 Aug. 2000. By applying model 6 with 1000 data points, the proposed model was: Y= -1.96286(S t0 ) -0.07562(A, 0 ) + 3.1 13706(S,.,) + 0.546506(A t .i)-16.8321. Thus, applying this model to the next data series, including dates from 23-30 August 2000 the predicted temperatures, shown in Figure 42, fitted the observed temperatures in trend and values. The similarity of the FW-soil data to the temperature records of the experimental area was evaluated with the data after and before the missed data, and shown in Figure 43 (A and B respectively).

PAGE 177

161 40 o o 20 0 2 3 4 5 6 7 8 Days Experiment Predicted FW Figure 42. Temperatures recorded at 7.5 cm depth from the experimental area and from the alternative temperature record source (FAWN). Period from 23-30 Aug. 2000. The predicted temperature was estimated with the equation: Y= -1.96286(S t o) 0.07562(A t0 ) + 3.1 13706(S t .i) + 0.546506(A t .,)-16.8321. The same equation applied to the data of August (Figure 43) was applied to the data next to missed data of September. The output showed the same trend but failed to match the values, so that it was decided to build a model using the more contiguous data to the missed ones, following the same procedure explained before. The data before the missed period are shown in Figure 44, where it is evident the lower seasonal temperatures, noting fewer differences between the experimental and FAWN temperatures. Thus the model building procedure was applied, producing the equation: Y= 6.61423(S,o)-0.4854(A t0 ) -4.4884(S,.,) +0.35548(A t .i)-20.061. This equation was then run with the observed temperatures after the missed period, with results in Figure 45. The

PAGE 178

162 predicted temperatures followed the trend and matched the values of the observed temperatures in the experimental area. Similarity of the FAWN soil temperatures to those observed in the experimental area at 22.5 cm depth was proved by plotting both temperatures using data before and after the missed period (Figure 46, A and B respectively). 24 34 32 O 2 30 3 M k_ fl> o. 28 E
PAGE 179

163 Exp. 22.5 cm FW 3 Days Figure 44. Temperatures recorded at two depths from the experimental area (Exp.) and from the alternative temperature record source (FAWN). Period from 29 Sep. 1 1 Oct. 2000. 28 Days Figure 45. Temperatures recorded at 7.5 cm depth from the experimental area and from the alternative temperature record source (FAWN). Period from 29 Sep. 1 1 Oct. 2000. The predicted temperature was estimated with the equation: Y= 6.61423(S t0 )-0.4854(A,o) -4.4884(S t .,) +0.35548(A,.i)-20.061.

PAGE 180

164 Figure 46. Temperature model for 22.5 cm depth underground for the period of 29 Sep. 1 1 Oct. 2000. A) Temperatures recorded before the missed temperature period and used to build the model. B) Temperatures recorded after the missed temperature.

PAGE 181

165 Discussion The datalogger set to record soil temperatures in the experimental area stopped working twice, missing the temperatures of several days. This problem coincided with phenological studies of two parasitoids that develop mostly underground. An alternative temperature data set was considered to supply the missing temperatures. The most proximate climate station was less than a mile from the experimental area; the Florida Automated Weather Network (FAWN) has an automated climate datalogger that includes soil temperature recordings at 15 cm depth. The datalogger in the experimental area, however, was set to record temperatures at 7.5 and 22.5 cm depth. A comparison of the records was made, and it showed disparity between data from 7.5 cm depth in the experimental area and those from FW (soil). Differences between data from 22.5 cm depth in the experimental area and those from FW (soil) were negligible. Due to lack of fit between the upper sub-chamber temperatures and those from FAWN, a model was developed to estimate the missed temperatures at the experimental area. The linear equation model developed to estimate temperature from August was applied to estimate temperatures from October, noting a poor fit of the predicted data. The same result occurred when using the model based upon October data to predict August temperatures. The lack of fit of the models built based upon data from different seasons of the year led to building and use of models based upon data of the same season. The resultant predicted data matched the temperatures used for validation (data after periods when temperatures were missed).

PAGE 182

166 The following speculation seems appropriate: i. Perhaps mole crickets tried to dig deeper away from soil surface to prevent desiccation. The upper chamber (sub-chamber A= 0-1 5cm depth) lost moisture at higher rate than deeper sub-chambers, particularly the first 5 cm. ii. 95% of the hosts died at the bottom of the sub-chambers -during 1999 the chambers consisted of PVC pipes up to 60 cm length kept underground without divisions, where parasitized hosts died at the bottom. iii. It is very likely the development of the parasitoids occurs deep underground (based upon i and ii) -deeper than the upper sub-chamberwhere temperature (and moisture) is more stable, escaping the higher temperatures, particularly during the summer months (see Tables 46 and 47). iv. That soil temperatures at depths of 22.5 cm or deeper were so similar in the experimental area and at FAWN, suggests that low variation can be expected in Alachua County, and probably elsewhere in north Florida. v. If the assumption made (in iv) is true, then the proposed models in chapters 3 and 4 can be applicable to a wider area to predict development of the wasp Larra bicolor (short development during spring and fall), and the fly Ormia depleta.

PAGE 183

LIST OF REFERENCES Abate, T. 1979. Plant material consumption and subterranean movements of mole crickets (Orthoptera: Gryllotalpidae: Scapteriscus) as determined by radioisotope techniques, with notes on materials for laboratory feeding. M.S. thesis, Univ. Florida. Adamson, A.M. 1942. Mole cricket parasites of the genus Larra in Trinidad. Trop. Agric. 9: 43^15. Andrewartha, H.G., and L.C. Birch. 1954. The distribution and abundance of animals. University of Chicago Press; Chicago, IL. Behrens, W., K.H. Hoffman, S. Kampa, S. Gabler, and G. Menke-Wallner. 1983. Effects of diurnal thermoperiods and quickly oscillating temperatures on the development and reproduction of crickets, Gryllus bimaculatus. Oecologia 59: 279-287. Bohart, R.M. and A.S. Menke. 1976. Sphecid wasps of the world -A generic revision. Univ. of California Press; Berkeley. Braman, S.K. 1993. Progeny production, number of instars and duration of development of tawny mole cricket (Orthoptera: Gryllotalpidae). J. Entomol. Sci. 28: 327-330. Campbell, A., B.D. Frazer, N. Gilbert, A.P. Gutierrez, and M. Mackauer. 1974. Temperature requirements of some aphids and their parasites. J. Appl. Ecol. 1 1 : 431-438. Candy, S.G. 1991. Modeling insect phenology using ordinal regression and continuation ratio models. Environ. Entomol. 20: 190-195. Castner, J.L. 1983. Biology and ecology of the mole cricket parasitoid Larra bicolor. M.S. thesis, Univ. Florida. Castner, J.L. 1984. Suitability of Scapteriscus spp. mole crickets (Orthoptera: Gryllotalpidae) as hosts of Larra bicolor (Hymenoptera: Sphecidae). Entomophaga 29: 323-329. 167

PAGE 184

168 Castner, J.L. 1988. Evaluation of Larra bicolor as a biological control agent of mole crickets. Ph.D. diss., Univ. Florida. Castner, J.L. and H.G. Fowler. 1987. Diel patterns of Larra bicolor (Hymenoptera: Sphecidae) in Puerto Rico. J. Entomol. Sci. 22: 77-83. Davidson, J. 1944. On the relationship between temperature and rate of development of insects at constant temperatures. J. Anim. Ecol. 13: 2638. Dennis, B., and W.P. Kemp. 1988. Further statistical inference methods for a stochastic model of insect phenology. Environ. Entomol. 17: 887-893. Ellsbury, MM. 1991. Development of immature stages of the clover stem borer (Coleoptera: Languriidae) at constant temperatures. Environ. Entomol. 20: 1 1 Sill 86. Forrest, T. G. 1985. Reproductive behavior. In: Mole crickets in Florida. Walker, T. J. (ed.). Univ. Florida Agric. Exp. Sta. Bull. 846. p. 10-15. Forrest, T. G. 1987. Insect size tactics and developmental strategies. Oecologia 75: 178184. Fowler, H.G. 1987. Field confirmation of the phonotaxis of Euphasiopteryx depleta (Diptera: Tachinidae) to calling males of Scapteriscus vicinus (Orthoptera: Grillotalpidae). Fla. Entomol. 70: 409-410. Fowler, H. G. 1989. Optimization of nectar foraging in a solitary wasp (Hymenoptera: Sphecidae: Larra sp.). Naturalia 14: 13-17. Fowler, H.G. and Kochalka, J.N. 1985. New record of Euphasiopteryx depleta (Diptera: Tachinidae) from Paraguay: attraction to broadcast calls of Scapteriscus acletus (Orthoptera: Grillotalpidae). Fla. Entomol. 68: 225-226. Frank, J. H. 1994. Inoculative biological control of mole crickets. In: Leslie, A. R. [ed.], Handbook of Integrated pest management for turf and ornamentals. CRC Press; Boca Raton, FL. p. 467-475. Frank, J. H., and J. P. Parkman. 1999. Integrated pest management of pest mole crickets with emphasis on the southeastern U.S. Integr. Pest Manage. Rev. 4: 39-52. Frank, J. H., J. P. Parkman, and F. D. Bennett. 1995. Larra bicolor (Hymenoptera: Sphecidae), a biological control agent of Scapteriscus mole crickets (Orthoptera: Gryllotalpidae), established in northern Florida. Fla. Entomol. 78: 619-623.

PAGE 185

169 Frank, J.H., Walker, T.J., and J.P Parkman. 1996. The introduction, establishment and spread of Ormia depleta in Florida. Biol. Control 6: 368-377. Gilbert, N., and D.A. Raworth. 1996. Insects and temperature-a general theory. Can. Entomol. 128: 1-13. Gould, J.R., and J.S. Elkinton. 1990. Temperature-dependent growth of Cotesia melacosela (Hymenoptera: Braconidae), a parasitoid of the gypsy moth (Lepidoptera: Lymantriidae). Environ. Entomol. 19: 859-865. Hagstrum, D.W., and G.A. Milliken. 1988. Quantitative analysis of temperature, moisture and diet factors affecting insect development. Ann. Entomol. Soc. Am. 81: 539546. Hagstrum, D.W., and G.A. Milliken. 1991. Modeling differences in insect developmental times between constant and fluctuating temperatures. Ann. Entomol. Soc. Am. 84: 369-379. Hagstrum, D.W., and W.R. Hagstrum. 1970. A simple device for producing fluctuating temperatures, with an evaluation of the ecological significance of fluctuating temperatures. Ann. Entomol. Soc. Am. 63: 1385-1389. Hardman, J.M. 1976. Life table data for use in deterministic and stochastic simulation models predicting the growth of insect populations under Malthusian conditions. Can. Entomol. 108: 897-906. Hayslip, N.C. 1943. Notes on biological studies of mole crickets at Plant City, Florida. Fla. Entomol. 26: 32-46. Higley, L.G., L.P. Pedigo, and K.R. Ostlie. 1986. Degday: a program for calculating degree-days, and assumptions behind degree-day approach. Environ. Entomol. 15: 999-1016. Hoffman, K.H. 1985. Metabolic and enzyme adaptation to temperature. In: KH Hoffman [ed.] Environmental physiology and biochemistry of insects. Springer; New York. Honek, A. 1999. Constraints on thermal requirements for insect development. Entomol. Science 2: 615-621. Howell, J,H., and L.G. Neven. 2000. Physiological development time and zero development temperature of the codling moth (Lepidoptera: Tortricidae). Environ. Entomol. 29: 766-772.

PAGE 186

170 Hudson, W,G. 1987. Variability in development of Scapteriscus acletus (Orthoptera: Gryllotalpidae). Fla. Entomol. 70: 403-404. Hudson, W,G., J.H. Frank, and J.L. Castner. 1988. Biological control of Scapteriscus spp. mole crickets (Orthoptera: Gryllotalpidae) in Florida. Bull. Entomol. Soc. Am. 34: 192-198. Infante, G.S., and L.G.P. Zarate. 1984. Metodos estadisticos. 2d ed. Editorial Trillas; Mexico City. Lactin, D.J., N.J. Holliday, D.L. Johnson, and R. Craigen. 1995. Improved rate model of temperature-dependent development by arthropods. Environ. Entomol. 24: 68-7i. Lake P 2000 Behaviors of Pheropsophus aequinoctialis (Coleoptera: Carabidae) affecting its ability to locate its larval food, eggs of Scapteriscus spp. (Orthoptera: Gryllotalpidae); and the effect of moisture on oviposition depth in Scapteriscus abbreviatus. M.S. thesis, Univ. Florida. Lamb R J. 1992. Developmental rate of Acyrthosiphon pisum (Homoptera: Aphididae) at ' low temperatures; implications to estimating rate parameters for insects. Environ. Entomol. 21: 10-19. Lawrence, K.O. 1982. A linear pitfall trap for mole crickets and other soil arthropods. Fla. Entomol. 65: 376-377. Lowry V K., J.H. Smith, Jr. and F.L. Mitchell. 1992. Life-fertility tables for ' Frankliniellafusca (Hinds) and F. occidentalis (Pergande) (Thysanoptera: Thripidae) on peanut. Ann. Entomol. Soc. Am. 85: 744-754. Liu, S.-S. and X.-D. Meng. 1999. Modelling development time of Myzus persicae (Hemiptera: Aphididae) at constant and natural temperatures. Bull. Entomol. 89: 53-63. Logan, J. A., D.J. Wollkind, S.C. Hoyt, and L.K. Tanigoshi. 1976. An analytic model for description of temperature dependent rate phenomena in arthropods. Environ. Entomol. 5: 1113-1140. Lysyk, T.J. 2000. Relationship between temperature and life history parameters of ' Muscidifurax raptor (Hymenoptera: Pteromalidae). Environ. Entomol. 29: 596605.

PAGE 187

171 Maia A H.N. AJ.B. Luiz, and C. Campanhola. 2000. Statistical inference on associated fertility life table parameters using jackknife technique: computational aspects. J. Econ. Entomol. 93:511-578. Manel, S., and D. Debouzie. 1997a. Logistic regression and continuation ratio models to ' estimate insect development under variable temperatures. Ecol. Model. 98: 237243. Manel, S., and D. Debouzie. 1997b. Modeling insect development time of two or more ' larval stages in the field under variable temperatures. Environ. Entomol. 26: 163169. Mesple, F., M. Troussellier, C. Casellas, and P. Legendre. 1996. Evaluation of simple statistical criteria to qualify a simulation. Ecol. Model. 99: 9-18. Menke, W.W. 1973. A computer simulation model: the velvetbean caterpillar in the soybean agrosystem. Fla. Entomol. 56: 92-102. Menke, A. S. 1992. Mole cricket hunters of the genus Larra in the New World (Hymenoptera: Sphecidae, Larrinae). J. Hymenoptera Res. 1: 175-234. Parkman, J.P., J.H. Frank., T.J. Walker, and D.J. Schuster. 1996. Classical biocontrol of Scapteriscus spp. (Orthoptera: Grillotalpidae) in Florida. Environ. Entomol. 25: 1415-1420. Parkman, J.P., and J.H. Frank. 1992. Infection of sound-trapped mole crickets, Scapteriscus spp. by Steinernema scapterisci. Environ. Entomol. 75: 163-165. Phelps, K., R.H. Collier, R.J. Reader, and S. Finch. 1993. Monte Carlo simulation method for forecasting the timing of pest insect attacks. Crop Protection 12: 335342. Potter, D.A. 1998. Destructive turfgrass insects: biology, diagnosis and control. Ann Arbor Press; MI. Ratte, H.T. 1985. Temperature and insect development. In: KH Hoffman [ed.] Environmental physiology and biochemistry of insects. Springer; New York. Ryoo, M.I., Y.S. Hong, and C.K. Yoo. 1991. Relationship between temperature and development of Lariophagus distinguendus (Hymenoptera: Pteromalidae, an ectoparasitoid of Sitophilus orizae (Coleoptera: Curculionidae). J. Econ. Entomol. 84: 825-829.

PAGE 188

172 Sailer, R.I. 1985. Biological control of mole crickets: natural enemies. In: Mole crickets in Florida. Walker, T. J. (ed.). Univ. Florida Agric. Exp. Sta. Bull. p. 23-32. Schoolfield, R.M., P.J.H. Sharpe, and C.E. Magnuson. 1981. Nonlinear regression of biological temperature-dependant model based on absolute reaction-rate theory. J Theor. Biol. 88: 719-731. Schroder, U., and D. Sondgerath. 1996. The concept of biological time for computing the switching points of a growth model for winter wheat. Ecol. Model. 88: 1-8. Shaffer, P. 1983. Prediction of variation in development period of insects and mites reared at constant temperatures. Environ. Entomol. 12: 1012-1019. Sharpe, P.J.H., and D.W. DeMichele. 1977. Reaction kinetics of poikilotherm development. J. Theor. Biol. 64: 649-670. Sharpe, P.J.H., G.L. Curry, D.W. DeMichele, and C.L. Cole. 1977. Distribution model of organism development times. J. Theor. Biol. 66: 21-38. Siddiqui, W.H., and C.A. Barlow. 1973. Population growth of Drosophila malanogaster (Diptera: Drosophilidae) at constant and alternating temperatures. Ann. Entomol. Soc. Am. 65:993-1001. Snedecor, G. W. 1946. Statistical methods applied to experiments in agriculture and biology, 4th ed. Iowa State College Press; Ames. Sokal, R.R., and F.J. Rohlf. 1995. Biometry: the principles and practice of statistics in biological research. 3 rd ed.W.H. Freeman and Company; San Francisco. Southwood, T.R.E. 1978. Ecological methods. 2nd ed. Wiley; New York. Steel, R.G.D. and J.H. Torrie. 1960. Principles and procedures of statistics; McGrawHill; New York. Stinner, R.E., G.D. Butler Jr., J.S. Becheler, and C. Turtle. 1975. Simulation of temperature development in population dynamics models. Can. Entomol. 107: 1167-1174. Tanigoshi, L.K., R.W. Browne, S.C. Hoyt, and R.F. Lagier. 1976. Empirical analysis of variable temperature regimes on life stage development and population growth of Tetranychus mcdanieli (Acarina: Tetranychidae). Ann. Entomol. Soc. Am. 69: 712-716.

PAGE 189

173 Tauber, M.J., C.A.Tauber, and J.P. Nyrop. 1994. Soil moisture and postdormancy emergence of Colorado potato beetles (Coleoptera: Chrysomelidae): descriptive model and field emergence patterns. Environ. Entomol. 23: 1485-1496. Taylor, F. 1981. Ecology and evolution of physiological time in insects. Am. Nat. 117: 123. Vinson, S.B. and G.F. Iwantsch. 1980. Host regulation by insect parasitoids. Quarterly Rev. Biol. 55: 143-165. Wagner, T.L., H.-I. Wu, P.J.H. Sharpe, and R.N. Coulson. 1984a. Modeling distributions of insect development time: a literature review and application of the Weibull function. Ann. Entomol. Soc. Am. 77: 475-487. Wagner, T.L., H. Wu, P.J.H. Sharpe, R.M. Schoolfield, and R.N. Coulson. 1984b. Modeling insect development rates: a literature review and application of a biophysical model. Ann. Entomol. Soc. Am. 77: 208-225. Wagner, T.L., H.-I. Wu, R.M. Feldman, P.J.H. Sharpe, and R.N. Coulson. 1985. Multiple-cohort approach for simulating development of insect populations under variable temperatures. Ann. Entomol. Soc. Am. 78: 691-704. Wagner, T.L., R.L. Olson, and J.L. Willers. 1991. Modeling arthropod development time. J. Agric. Entomol. 8: 251-270. Walker, T. J. (ed.). 1985a. Mole crickets in Florida. Univ. Florida Agric. Exp. Sta. Bull. 846. Walker, T. J. 1985b. Biology of pest mole crickets: systematics and life cycles. In: Mole crickets in Florida. Walker, T. J. (ed.). Univ. Florida Agric. Exp. Sta Bull. p. 310. Walker, T.J. and J.L. Nation. 1982. Sperm storage in mole crickets; fall matings fertilize spring eggs in Scapteriscus acletus. Fla. Entomol. 65: 283-285. Walker, T.J. and D.A. Nickle. 1981. Introduction and spread of pest mole crickets: Scapteriscus vicinus and S. acletus reexamined. Ann. Entomol. Soc. Am. 74: 158163. Walker, T.J., Parkman, J.P., Frank, J.H. and D.J. Schuster. 1996. Seasonality of Ormia depleta and limits to its spread. Biol. Control 6: 378-383.

PAGE 190

174 Wang, R.S., Z.X. Lan, and Y.C. Ting. 1982. Studies on mathematical models of the relationships between insect development and temperature. Acta Ecol. Sinica 2: 47-57. Welch, C.H. 2000. Gas chromatographic analyses of crop sugars of Ormia depleta established in Florida. M.S. thesis, Univ. Florida. Wigglesworth, V.B. 1972. The principles of insect physiology. 7 th ed. Chapman and Hall; London. Williams, F. X. 1928. Studies in tropical wasps-their hosts and associates. Bull. Exp. Sta. Hawaiian Sugar Planters' Assoc., Entomol. Ser. no. 19: 30-60. Wineriter, S.A. and T.J. Walker. 1990. Rearing phonotactic parasitoid flies (Diptera: Tachinidae, Ormiini, Ormia depleta). Entomophaga 35: 621-632. Worner, S.P. 1992. Performance of pheno logical models under variable temperature regimes: consequences of the Kaufmann or rate summation effect. Environ. Entomol. 21:689-699.

PAGE 191

BIOGRAPHICAL SKETCH Hector Cabrera-Mireles was born on July 25, 1957, in Monterrey, Nuevo Leon, Mexico. He grew up in Monterrey, where he attended school and graduated from high school in 1973. He then went to earn his Bachelor of Science degree in parasitology from the agronomy school at the Universidad Autonoma de Nuevo Leon (UANL). After graduation, he went to work for INIFAP, in Tabasco State, southern Mexico. He worked there on coconut palm pests. After five years, he was accepted into the graduate program of the Institute Tecnologico y de Estudios Superiores de Monterrey (ITESM), to study plant health. After finishing his master's degree he continued working for INI A (today INIFAP) in the research program on fruit flies research, focusing on mango fruits. In 1997 he was accepted into the graduate program of the University of Florida Entomology and Nematology Department to study biocontrol. After finishing his doctoral degree, he will continue working for INIFAP in Veracruz state, Mexico. 175

PAGE 192

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. J£jLk J. Howard Frank, Chair Professor of Entomology and Nematology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Jeyw L./Stimac Professor of Entomology and Nematology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy . ' A Thomas J. Walker Professor of Entomology and Nematology 1 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. [James W. Jones distinguished Professor of Agricultural and Biological Engineering This dissertation was submitted to the Graduate Faculty of the College of Agricultural and Life Sciences and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August 2002 Dean, College of Agric^turaPand Life" Sciences J Dean, Graduate School


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EZOZJQIPR_BHWWNF INGEST_TIME 2015-03-05T21:07:07Z PACKAGE AA00028847_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES