<%BANNER%>

Monitoring and Managing Blueberry Gall Midge (Diptera

Permanent Link: http://ufdc.ufl.edu/UFE0024872/00001

Material Information

Title: Monitoring and Managing Blueberry Gall Midge (Diptera Cecidomyiidae) in Rabbiteye Blueberries
Physical Description: 1 online resource (117 p.)
Language: english
Creator: Roubos, Craig
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: aprostocetus, blueberries, cecidomyiidae, dasineura, ipm, midge, oxycoccana, vaccinium
Entomology and Nematology -- Dissertations, Academic -- UF
Genre: Entomology and Nematology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Blueberry gall midge, Dasineura oxycoccana (Johnson) is a key pest of blueberries in the southeastern United States. In Florida, it emerges in January and can cause considerable damage to developing blueberry flower buds. The objectives of this study were to develop new monitoring tools and improve management strategies for D. oxycoccana. Experiments were conducted in rabbiteye blueberry, Vaccinium ashei Reade, plantings in north-central Florida. Emergence traps were evaluated for their effectiveness in capturing adult midges emerging from pupation sites and at predicting subsequent larval infestation. The Petri dish and bucket traps caught the highest number of adult midges in 2007 and 2008, respectively. Data from both traps showed peak adult emergence preceding peak larval infestation by one week. Laboratory experiments were conducted to determine the minimum temperature for midge pupation and the number of degree days (DD) required. The most reliable estimates of these parameters were calculated using a linear model. The minimum temperature estimate for pupation was 8.9 degrees C and the degree-day estimate was 134 DD. These estimates can be used to develop models for predicting midge activity in the field. I conducted efficacy trials comparing reduced-risk insecticides to a conventional insecticide (malathion) and a control. Malathion and acetamiprid significantly reduced the number of midge larvae per flower bud compared to the control, but the acetamiprid data were inconclusive because the post-treatment larval infestation was not significantly different from the pre-treatment infestation. I conducted a survey of blueberry gall midge parasitoids in north-central Florida. Adult wasps were reared from the larvae collected from blueberry flower and leaf buds, or larvae were mounted on slides and immature parasitoids were counted in vivo. The two most common genera reared from midge larvae were Platygaster (Platygastridae) and Aprostocetus (Eulophidae). Aprostocetus sp. was the primary parasitoid of midge larvae in leaf buds. The highest percentage of larvae parasitized was found in the outer layers of the leaf buds. Third-instar midge larvae were parasitized most often. Dasineura oxycoccana was the most commonly collected midge pest in blueberry leaf buds. Prodiplosis vaccinii was collected but did not occur in high numbers.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Craig Roubos.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Liburd, Oscar E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024872:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024872/00001

Material Information

Title: Monitoring and Managing Blueberry Gall Midge (Diptera Cecidomyiidae) in Rabbiteye Blueberries
Physical Description: 1 online resource (117 p.)
Language: english
Creator: Roubos, Craig
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: aprostocetus, blueberries, cecidomyiidae, dasineura, ipm, midge, oxycoccana, vaccinium
Entomology and Nematology -- Dissertations, Academic -- UF
Genre: Entomology and Nematology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Blueberry gall midge, Dasineura oxycoccana (Johnson) is a key pest of blueberries in the southeastern United States. In Florida, it emerges in January and can cause considerable damage to developing blueberry flower buds. The objectives of this study were to develop new monitoring tools and improve management strategies for D. oxycoccana. Experiments were conducted in rabbiteye blueberry, Vaccinium ashei Reade, plantings in north-central Florida. Emergence traps were evaluated for their effectiveness in capturing adult midges emerging from pupation sites and at predicting subsequent larval infestation. The Petri dish and bucket traps caught the highest number of adult midges in 2007 and 2008, respectively. Data from both traps showed peak adult emergence preceding peak larval infestation by one week. Laboratory experiments were conducted to determine the minimum temperature for midge pupation and the number of degree days (DD) required. The most reliable estimates of these parameters were calculated using a linear model. The minimum temperature estimate for pupation was 8.9 degrees C and the degree-day estimate was 134 DD. These estimates can be used to develop models for predicting midge activity in the field. I conducted efficacy trials comparing reduced-risk insecticides to a conventional insecticide (malathion) and a control. Malathion and acetamiprid significantly reduced the number of midge larvae per flower bud compared to the control, but the acetamiprid data were inconclusive because the post-treatment larval infestation was not significantly different from the pre-treatment infestation. I conducted a survey of blueberry gall midge parasitoids in north-central Florida. Adult wasps were reared from the larvae collected from blueberry flower and leaf buds, or larvae were mounted on slides and immature parasitoids were counted in vivo. The two most common genera reared from midge larvae were Platygaster (Platygastridae) and Aprostocetus (Eulophidae). Aprostocetus sp. was the primary parasitoid of midge larvae in leaf buds. The highest percentage of larvae parasitized was found in the outer layers of the leaf buds. Third-instar midge larvae were parasitized most often. Dasineura oxycoccana was the most commonly collected midge pest in blueberry leaf buds. Prodiplosis vaccinii was collected but did not occur in high numbers.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Craig Roubos.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Liburd, Oscar E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024872:00001


This item has the following downloads:


Full Text

PAGE 1

1 MONITORING AND MANAGING BLUEBERRY GALL MIDGE (DIPTERA: CECIDOMYIIDAE) IN RABBITEYE BLUEBERRIES By CRAIG RICHARD ROUBOS 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 2009

PAGE 2

2 2009 Craig Richard Roubos

PAGE 3

3 To my family for their unfailing support

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank everyone who helped in preparing this dissertation. I thank my major professor, Dr. Oscar E. Liburd, and the members of my graduate committee, Drs. Rob ert McSorley, Paul M. Lyrene, Frank Slansk y Jr., and Dakshina Seal, for their guidance. Thanks to Dr. Blair Sampson at the U nited S tates D epartment of A griculture, Agriculture Research Service, Small Fruit Res earch Station in Poplarville, Mississippi for use of his laboratory space and his helpf ul input. I would like to thank the members of the Small Fruit and Ve getable Integrated Pest Management Laboratory for their help in collecting and processing data. Special thanks go to the blueberry growers who allowed me to use their farms and collect data This research was funded by the Florida Blueberry Growers Association USDA Pest Management Alternatives (P MAP) program, and the Interregional Research Project No. 4 (IR 4) Finally, I thank my parents for their continued support and my wife, Gayl e, for her encouragement and patience through this whole process.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 8 LIST OF FIGURES .............................................................................................................................. 9 ABSTRACT ........................................................................................................................................ 11 CHAPTER 1 INTRODUCTION ....................................................................................................................... 13 2 LITERATURE REVIEW ........................................................................................................... 16 Blueberry Pr oduction Around the World .................................................................................. 16 Blueberries in Florida .................................................................................................................. 17 Arthropod Pests of Blueberry ..................................................................................................... 19 Dise ases of Blueberry ................................................................................................................. 22 Blueberry Gall Midge ................................................................................................................. 22 Taxonomy ............................................................................................................................. 22 Life Cycle ............................................................................................................................. 23 Pest Status ............................................................................................................................ 26 Emergence ............................................................................................................................ 27 Mating Behavior and Pheromones ...................................................................................... 28 Monitoring ............................................................................................................................ 30 Management ......................................................................................................................... 31 Insecticides ................................................................................................................... 31 Biological control ......................................................................................................... 31 Cultural control ............................................................................................................. 32 3 EVALUATION OF EMERGE NCE TRAPS FOR MONITORING BLUEBERRY GALL MIDGE, DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE) ................................................................................................................... 33 Materials and Methods ................................................................................................................ 34 2006 Pilot Test ..................................................................................................................... 35 2007 Experiment .................................................................................................................. 36 2008 Experiment .................................................................................................................. 36 Results .......................................................................................................................................... 37 2006 Pilot Test ..................................................................................................................... 37 2007 Experiment .................................................................................................................. 38 2008 Experiment .................................................................................................................. 38 Discussion .................................................................................................................................... 39

PAGE 6

6 4 WITHIN FIELD DISTRIB UTION OF BLUEBERRY G ALL MIDGE, DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE) ........................................... 49 Materials and Methods ................................................................................................................ 50 2006: Experiment to Investigate Dasineura oxycoccana Distribution Patterns with Variation in Field Borders ............................................................................................... 51 2007: Experiment to Investigate Dasineura oxycoccana Distribution Patterns in Isolated Plots with Potential Edge Effects ...................................................................... 52 2008: Experiment to Investigate Dasineura oxycoccana Distribution Patterns in Isol ated Plots with Potential Edge Effects: Second Year .............................................. 52 Results .......................................................................................................................................... 53 2006: Experiment to Investigate Dasineura oxycoccana Distribution Patterns with Variation in Field Borders ............................................................................................... 53 2007: Experiment to Investigate Dasineura oxycoccana Distribution Patterns in Isolated Plots with Potential Edge Effects ...................................................................... 54 2008: Experiment to Investigate Dasineura oxycoccana Distribution Patterns in Isolated Plots .................................................................................................................... 54 Discussion .................................................................................................................................... 55 5 BLUEBERRY GALL MIDGE DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE), PUPATION AND E MERGENCE UNDER VARYING TEMPERATURE CONDITIONS ......................................................................... 64 Materials and Methods ................................................................................................................ 66 Development of Midges in the Laboratory ........................................................................ 66 Temperature Models ............................................................................................................ 67 Trapping Midge Adults in the Field ................................................................................... 68 Results .......................................................................................................................................... 69 Linear Model ........................................................................................................................ 70 Nonlinear Model .................................................................................................................. 70 Field Data ............................................................................................................................. 70 Discussion .................................................................................................................................... 71 6 EVALUATION OF REDUCE D -RISK INSECTICIDES FO R MANAGEMENT OF BLUEBERRY GALL MIDGE DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE) .............................................................................................. 7 8 Materials and Methods ................................................................................................................ 79 2006 Efficacy Trial .............................................................................................................. 80 2007 Efficacy Trial .............................................................................................................. 80 2009 Efficacy Trial .............................................................................................................. 81 Results .......................................................................................................................................... 82 2006 Efficacy Trial .............................................................................................................. 82 2007 Efficacy Trial .............................................................................................................. 82 2009 Efficacy Trial .............................................................................................................. 82 Discussion .................................................................................................................................... 83

PAGE 7

7 7 PARASITOIDS OF BLUEB ERRY GALL MIDGE, DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE) IN NO RTH CENTRAL FLORIDA ........... 88 Materials and Methods ................................................................................................................ 90 Identification of Parasitoids ................................................................................................ 90 Patter ns of Parasitoid Activity ............................................................................................ 91 Results .......................................................................................................................................... 92 Identification of Parasitoids ................................................................................................ 92 Pa tterns of Parasitoid Activity ............................................................................................ 92 Discussion .................................................................................................................................... 94 8 GALL MIDGE SPECIES C OMPLEX IN FLORIDA RA BBITEYE BLUEBERRIES ........ 98 Materials and Methods ................................................................................................................ 99 Results an d Discussion ............................................................................................................. 100 9 GENERAL CONCLUSIONS .................................................................................................. 103 Emergence Trap for Adult Midge Monitoring ........................................................................ 103 Development Rate Model for Midge Pupae ............................................................................ 104 Managing Blueberry Gall Midge with Reduced-Risk Insecticides ........................................ 105 Identifying Natural Enemies of Blueberry Gall Midge in Florida ......................................... 105 LIST OF REFERENCES ................................................................................................................. 107 BIOGRAPHICAL SKETCH ........................................................................................................... 117

PAGE 8

8 LIST OF TABLES Table page 3 1 Emergence trap evaluation, pi lot experiment 2006. ............................................................ 42 3 2 Emergence trap evaluation for 2007. .................................................................................... 42 3 3 Mean number of Dasineura oxycoccana females and males per trap ( SEM) and sex ratio of captured adults in 2007. ............................................................................................ 42 3 4 Emergence trap evaluation for 2008. .................................................................................... 43 3 5 Mean number of Dasineura oxycoccana females and males per trap ( SEM) and sex ratio of captured adults in 2008. ............................................................................................ 43 4 1 Weekly mean Dasineura oxycoccana infestation (larvae per flower bud) ( SEM) in different rows of plot B in 2007. ........................................................................................... 58 4 2 Mean Dasineura oxycocc ana infestation (larvae per flower bud) ( SEM) in different rows of plot B by sample date in 2008. ................................................................................ 58 5 1 Time to emergence of Dasineura oxycoccana adults reared in 5 -ml vial pupation containers. ............................................................................................................................... 75 5 2 Time to emergence of Dasineura oxycoccana adults reared in 947 ml deli cup pupation containers. ............................................................................................................... 75 5 3 Estimates of the thermal constant for Dasineura oxycoccana pupation under consta nt t emperatures. ........................................................................................................................... 75 5 4 Parameter estimates (mean SEM) and statistics from the nonlinear Lactin regression. ............................................................................................................................... 76 6 1 Insecticides used in efficacy trials for blueberry gall midge control in rabbiteye flower buds. ............................................................................................................................ 85

PAGE 9

9 LIST OF FIGURES Figure page 3 1 Emergence traps evaluated for effectiveness in catching Dasineura oxycoccana adults in 2 007 and 2008. ........................................................................................................ 44 3 2 Mean number of Dasineura oxycoccana adults captured in emergence traps in plots A and B 2007. ......................................................................................................................... 45 3 3 Mean number of Dasineura oxycoccana adults per emergence trap and mean number of larvae per blueberry flower bud collected at an organic blueberry farm i n Gainesville, FL in 2007. ........................................................................................................ 46 3 4 Mean number of Dasineura oxycoccana adults captured in emergence traps in plots A and B in 2008. .................................................................................................................... 47 3 5 Mean number of Dasineura oxycoccana adults per emergence trap and mean number of larvae per blueberry flower bud collected at an organic blueberry farm in G ainesville, FL in 2008. ........................................................................................................ 48 4 1 Map of Plot A from a bluebe rry farm in Gainesville, FL. ................................................... 59 4 2 Map of Plot B from a blueberry farm in Gainesville, FL. ................................................... 60 4 3 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot A over the 2006 sampling period. ............................................................................................................ 61 4 4 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot A at each sampling date in 2006. ........................................................................................................... 61 4 5 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot B over the 2007 sampling peri od. ............................................................................................................ 62 4 6 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot B at each sampling date in 2007. ........................................................................................................... 62 4 7 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot B over the 2008 sampling peri od. ............................................................................................................ 63 5 1 Development rates of Dasineura oxycoccana pupae in 5 ml vials at four constant temperatures ........................................................................................................................... 77 5 2 Development rates of Dasineura oxycoccana pupae in deli cups at four constant temperatures. ........................................................................................................................... 77 6 1 Mean ( SEM) Dasineura oxycoccana larvae per flower bud from 2006 insecticide evaluation (North Florida Research and E ducation Center, Live Oak, FL). ...................... 86

PAGE 10

10 6 2 Mean ( SEM) Dasineura oxycoccana larvae per flower bud from 2007 insecticide evaluation (North Florida Research and Education Center, Live Oak, FL). ...................... 86 6 3 Mean ( SEM) Dasineura oxycoccana larvae per flower bud from 2009 insecticide evaluation (North Florida Research and Education Center, Live Oak, FL). ...................... 87 7 1 Mean num ber of Dasineura oxycoccana larvae per leaf bud collected in 2009 at a blueberry farm in Gainesville, FL. ........................................................................................ 97 7 2 Mean percent paras itism of Dasineura oxycoccana larvae by Aprostocetus sp. ............... 97 8 1 Wing margins of gall midges collecte d at blueberry farms in North -Central Florida. .... 102

PAGE 11

11 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 MONITORING AND MANAGING BLUEBERRY GALL M IDGE (DIPTERA: CECIDOMYIIDAE IN RABBITEYE BLUEBERRIES By Craig Richard Roubos August 2009 Chair: Oscar E. Liburd Major: Entomology and Nematology Blueberry gall midge, Dasineura oxycoccana (Johnson) is a key pest of blueberries in the southeastern United States. In Florida, it emerges in January and can cause considerable damage to developing blueberry flower buds. The objectives of this study were to develop new monitoring tools and impr ove management strategies for D. oxycoccana Experiments were conducted in rabbiteye blueberry, Vaccinium ashei Reade, plantings in north-central Florida. Emergence traps were evaluated for their effectiveness in capturing adult midges emerging from pupation sites and at predicting subsequent larval infestation. The Petri dish and bucket traps caught the high est number of adult midges in 2007 and 2008, respectively. Data from both traps showed peak adult emergence preceding peak larval infestation by one week. Laboratory experiments were conducted to determine the minimum temperature for midge pupation and t he number of degree days (DD) required. The most reliable estimates of these parameters were calculated using a linear model. The minimum temperatu re estimate for pupation was 8.9 degrees C and the degree -day estimate was 134 DD. These estimates can be used to develop models for predicting midge activity in the field. I conducted efficacy trials comparing reduced risk insecticides to a conventional insecticide (malathion) and a control. Malathion and acetamiprid significantly reduced the

PAGE 12

12 number of midge larvae per flower bud compared to the control, but the acetamiprid data were inconclusive because the post -treatment larval infestation was not significantly different from the pre treatment infestation. I conducted a survey of blueberry gall midge para sitoids in north -central Florida. Adult wasps were reared from the larvae collected from blueberry flower and leaf bud s or larvae were mounted on slides and immature parasitoids were counted in vivo The two most common genera reared from midge larvae w ere Platygaster (Platygastridae) and Aprostocetus (Eulophidae). Aprostocetus sp. was the primary parasitoid of midge larvae in leaf buds. The highest percentage of larvae parasitized was found in the outer layers of the leaf buds. Thirdinstar midge lar vae were parasitized most often. Dasineura oxycoccana was the most commonly collected midge pest in blueberry leaf buds. Prodiplosis vaccinii was collected but did not occur in high numbers.

PAGE 13

13 CHAPTER 1 INTRODUC TION Blueberries are one of the few cultivated fruits native to North America ( Williamson and Lyrene 2004a ). The industry has expanded to Europe, South America, and Australasia, but North America r emains the major producer (Stri k and Yarborough 2005). Bl ueberries belong to the genus Vaccinium which also includes cranberries and huckleberries. Vaccinium species were first introduced to cultivation in the late 1700s but the industry did not begin to expand until the turn of the 20th century (T rehane 2004) Many species of Vaccinium are also grown for ornamental purposes (Trehane 2004). The blueberry industry has been expanding rapidly over the past two decades. From 1982 to 1992, the area planted to blueberries in the United Sta tes increased by 19% (Moor e 1994). From 1992 to 2003, highbush acreage in the United States increased by 24% and lowbush acreage increased by 33% (Strik and Yarborough 2005). Over the past five years, the area planted to blueberries in Florida has increased 13% ( NASS -USDA 2008). From April to May, Florida is one of the major producer s of fresh blueberries in the United States (Williamson et al. 2000). Mild winters and warm spring temperatures allow for early production, but the chilling required for blueberry dormancy ma y be inadequate. Chilling requirements vary among different blueberry cultivars (Lyrene 1989). When chilling requirements are not met, plants are unproductive. Southern highbush blueberries are hybrids based largely on crosses between Vaccinium corymbos um and V. darrowi They have low chilling requirements, which have made it possible to grow blueberries throughout Florida (Lyrene 1989). Expansion of blueberry production has been influenced by the publics interest in the health benefits of Vaccinium fr uits as well as finding alternatives to citrus as the main crop in Florida. Vaccinium fruits are rich in phenolic compounds and essential nutrients known to have

PAGE 14

14 antioxidant and anticarcinogenic properties (Zheng and Wang 2003). Prior et al. (1998) deter mined that on a fresh weight basis, blueberries had the highest antioxidant capacity compared to other fruits and vegetables. Unfortunately, blueberries are threatened by a number of insect pests. Two of the most wide -spread pests are the blueberry maggo t, Rhagoletis mendax Curran, and cranberry fruit worm, Acrobasis vaccinii Riley (Turner and Liburd 2007). Both of these pests damage the fruit directly. In the southeastern United States, the blueberry gall midge, Dasineura oxycoccana (Johnson), has emer ged as a significant pest. Blueberry gall midge damages developing flower and leaf buds, limiting fruit production. The importance of blueberry gall midge was only recognized since the early 1990s (Lyrene and Payne 1995). In most blueberry growing regi ons of North America it is not considered a pest of any significance. Barnes (1948) listed it as a pest on cranberries but not blueberries. Although it is a key pest, little has been published on blueberry gall midge biology and management. Only Sampson et al. (2002 & 2006) and Sarzynski and Liburd (2003) have discussed certain aspects with regards to biology and management of blueberry gall midge. The only management tools currently available are insecticides. Malathion, spinosad and diazinon have been used for blueberry gall midge control (Liburd 2004). The midge spends most of its lifecycle within the buds (larva) or the soil (pupa), which shelters it from insecticides (Lyrene and Payne 1995). Systemic insecticides may be more efficacious for midge control (Bosio et al. 1998). A number of new reducedrisk insecticides have systemic activity. Due to blueberry gall midges short life cycle and multiple generations per year, the development of insecti cide resistance is possible (Sampson et al. 2002). Monitoring blueberry gall midge is c omplicated by its small size (2 3 mm ) and short adult lifespan (1 2 days) (Gagn 1989). Sticky traps hung in the canopy of blueberry bushes

PAGE 15

15 have proven ineffective (Sarzynski and Liburd 2003). Adults can be damage or key identifying characteristics obscured by the adhesive on the sticky trap. Midges infesting buds can be detected through an emergence technique or by dissecting buds, but these techniques are time consuming and not practical for most growers (Sarzynski and Liburd 2003). An effective monitoring technique is essential for directing the timing of insecticide applications. Biological control has been explored as a control strategy in other blueberry growing regions. Several species of wasps have been i dentified as parasitoids of blueberry gall midge. In Mississippi, 30 40% of midge larvae were found parasitized during peak parasitoid activity (Sampson et al. 2002). The midge population was subsequently reduced 75% (Sampson et al. 2002). The most c ommonly encountered parasitoids belonged to the families Platygastridae and Eulophidae (Hymenoptera). A new species of Aprostocetus (Hymenoptera: Eulophidae: Tetrastichinae) was discovered which attacked blueberry gall midge early in the season (Sampson e t al. 2006). Sampson et al. (2002) cautioned that insecticide applications should be carefully timed to avoid disrupting natural enemies. There is a need for research covering all aspects of blueberry gall midge biology, behavior, and management. The main objectives of this work were to improve monitoring and management of blueberry gall midge adults in rabbiteye blueberries. My goals were to design an effective emergence trap capable of detecting blueberry gall midge adults early in the season, devel op a temperature model for estimating the rate of midge development, evaluate new reduced -risk insecticides for midge control, and identify hymenopteran parasitoids of blueberry gall midge in Florida. Knowledge gained will help develop a comprehensive pes t management program for this pest in Florida.

PAGE 16

16 CHAPTER 2 LITERATURE REVIEW Blueberry Production Around t he World North America produces the majority of the worlds blueberries (Strick and Yar borough 2005). Thirty-six states in the United States and six Canadian provinces have notable blueberry production regions (Trinka 1997). Blueberries are also cultivated in Europe (Poland, Germany, Belgium, France, Italy, Spain, and the Netherlands), South America (Chile and Argentina ), New Zealand and Austr al ia (P liszka 1993, Naumann 1993, Schick 1993, Patel 1993). China is also emerging as a commercial blueberry producer (Li et al. 2006). Blueberry ( Vaccinium spp.) is one of the few cultivated fruits native to North Amer ica (Williamson and Lyrene 2004a ). The three main species of blueberry cultivated in North America are Vaccinium corymbosum L. (highbush), V. ashei Reade (rabbiteye), and V. angustifolium Aiton (lowbush). Lowbush production is limited to the Northeast, particularly Maine ( Trehane 2004). T he top five blueberry producing states (excluding lowbush blueberries) in terms of acreage are Michigan, New Jersey, Georgia, North Carolina and Oregon ( NASS USDA 2008). In 2007 the United States produced over 1 41 thousand tons of blueberries. The total value of utilized production was over $ 519 million ( NASS -USD A 2008). Highbush blueberries were introduced to Europe in the 1920s and 1930s (Naumann 1993). The majority of blueberries grown in Europe are produced for the fresh market (Naumann 1993). Nativ e Vaccinium species in Europe include bilberries ( V. myrtillis L.), whortleberry ( V. uliginosum L.), cowberries ( V. vitis -idaea L.) and small cranberries ( V. oxycoccus L.) (Pliszka 1993). Blueberries have a limited market in South America so most are expor ted t o the Northern hemisphere (Schick 1993). Chile is the main blueberry producing country in South America with

PAGE 17

17 approximately 10,700 ha planted as of 2007 (Baados 2009). Blueberries were first introduced into Chile for experimental purposes in 1979 (Sc hick 1993). Highbush, southern highbush, and rabbiteye blueberries are all grown commercially (Lyrene and Muoz 1997). Blueberries were introduced to Argentina in the early 1990s, and by 2005 there were approximately 1600 ha planted (Taquini 2006). Sout hern highbush varieties are the only commercially grown blueberries in Argentina (Taquini 2006). The blueberry industries in Australia and New Zealand began in the early 1980s. In Australia, the main types of blueberries grown are southern highbush and ra bbiteye, while in New Zealand the main types are northern highbush and rabbiteye (Patel 1993). New Zealand exports up to 75% of its blueberry crop to North America, Europe and Asia (Patel 1993). Blueberries in Florida In Florida, cultivated blueberries are grown on approximately 1011 acres. Total blueberry production for the state in 2007 was 3900 tons with a value of over $39 million ( NASS -USDA 2008). Although Florida ranks eighth in the nation in terms of total blueberry production, its blueberry in dustry is very important due to the states mild winters, which allow for earlier production. In most blueberry growing areas in the northern hemisphere crops are harvested from June through August (Lyrene 1989). Southern highbush blueberries (interspec ific hybrids of Vaccinium ashei V. corymbosum and V. darrowi ) grown in central Florida, however, can be harvested in April and May (Lyrene 1989). For many years Florida was the only supplier of fresh blueberries in commercial quantities during these mon ths (Crocker and Willis 1989) but Georgia and California have since surpassed Florida as the major early -season producers (Blizzard 2006, NeSmith 2006) Blueberries grown in Californias Central San Joaquin Valley

PAGE 18

18 can be harvested in early may (Blizzard 2006). In Georgia, where both rabbiteye and southern highbush are grown, the production season lasts from mid -April into July (NeSmith 2006). In Florida, the blueberry production area increased almost 220% from 1992 to 2007 (NASS -USDA 2008). The blueberr y production trend in Florida has been toward more southern highbush and less rabbiteye. Between 1989 and 2000, for instance, southern highbush acreage in Florida increased by 23% while rabbiteye acreage decreased by 56% (Williamson et al. 2000). Souther n highbush was developed at the University of Florida and released for cultivation in 1975 (Strik and Yarborough 2005). Early flowering makes southern highbush profitable, but it also makes it susceptible to late winter or early spring free zes (Williamson and Lyrene 2004b ). Florida growers suffer substantial losses from such freezes. The most common protective measure is overhead irrigation (Lyrene 1989). There is an increased market for early -season blueberries such as southern highbush with a higher ma rket value than for later ripening berries (Sarzynski and Liburd 2003). Florida blueberries sold at $5 00 per pound in 2007 w hile the national average was $2 11 per pound ( NASS -USDA 2008). In Georgia, rabbiteye remains the most widely planted blueberry ty pe (Trinka 1997). Rabbiteye blueberry plants are large and vigorous and are capable of carrying large fruit loads. They maintain high fruit firmness and flavor when berries are ripening, often during periods of hot, rainy weather, and some cultivars are well suited to mechanical harvest (Lyrene and Payne 1995). Rabbiteye blueberry ripens about a month later than southern highbush (Lyrene and Payne 1995). While southern highbush blueberries are grown for the high -priced export market, rabbiteye blueberri es are grown for u -pick and local mark ets (Williamson and Lyrene 2004a ). Rabbiteye and southern highbush grow well on well -drained acid soils (pH 4.0 to 5 .5) (Williamson and Lyrene 2004b ). Rabbiteye is native to north Florida and is better adapted to the

PAGE 19

19 growing c onditions in this area (Crocker et al. 1994). Florida soil is generally not high in organic matter. Rabbiteye can produce satisfactorily on soil with only 1% organic matter, but southern highbush is not recommended on soils with less than 3% organic mat ter (Williamson and Lyrene 2004b ). Bees are essential pollinators in cultivated blueberries. For good commercial yield about 80% fruit set is required (Shutak and Marucci 1966). The most important blueberry pollinators in the southeast are the European honeybee, Apis mellifera L., bumblebees, Bombus spp. and Habropoda laboriosa (F.), a native bee (Williamson and Lyrene 2004c Sampson and Cane 2000). The mason bee Osmia ribifloris Cockerell was also found to be an efficient pollinator of rabbit eye blueberries although it does not naturally occur with this type of blueberry (Sampson et al. 2004). The most efficient pollinators sonicate flowers (vibrate wing muscles at high frequency causing pollen to pour out of flowe rs) (Williamson and Lyrene 2 004c ). Rabbiteye blueberry cultivars are largely self incompatible ( Williamson and Lyrene 2004c ). Without cross pollination only about 4% of flowers will set fruit (Sampson and Cane 2000). Pollination experiments from Japan showed that cross pollination of rabbiteye blueberries resulted in larger and earlier ripening berries tha n self -pollination (Tamada et al. 1977). Self -pollinated southern highbush and northern highbush cultivars will set fruit, but cross -pollination accelerates fruit development, in creases seed content and fruit size (Shutak and Marucci 1966, Williamson and Lyrene 2004c ). Arthropod Pests of Blueberry Blueberries suffer attack from a number of arthropod pests and diseases. Key pest status varies depending on region. Major arthropod pests include blueberry gall midge ( Dasineura oxycoccana (Johnson) ) blueberry maggot ( Rhagoletis mendax Curran), cranberry fruitworm

PAGE 20

20 (Acrobasis vaccinii Riley), blueberry bud mite ( Acalitus vaccinii Keifer), and flower thrips (Frankliniella spp.) (Liburd and Arevalo 2006). Blueberry Maggot: Rhagoletis mendax Curran (Diptera: Tephritidae). The blueberry maggot is one of the most serious pest s of blueberries in North America east of the Rocky Mountains (Liburd and Arevalo 200 6). It was first reported in Ma ine and New Hampshire in 1914 and has since spread throughout the eastern United States and Canada (Steck and Payne 1998). Cultivated hosts include lowbush and highbush, blueberries ( Vaccinium angustifolium and V. corymbosum respectively) in the northeas tern and north -central blueberry growing areas, but it can also infest blueberries in the southeast (Steck and Payne 1998). In Florida and Georgia, its native host is deerberry, V. stamineum (Steck and Payne 1998). Blueberry maggot overwinters as a pupa in the soil. Females insert eggs beneath the skin of developing berries seven to 10 days after emergence (Marucci 1966). A single female can produce 25 to 100 eggs (Steck and Payne 1998). Eggs hatch in two to seven days, and larvae feed in the fruit pulp for about 20 days (Marucci 1966). This long feeding period means that blueberry maggot often remains in the berry until after harvest (Marucci 1966). There is zero tolerance for maggot in fested blueberries (Liburd et al. 2000). Pherocon AM yellow sticky boards and sphere traps have been used extensively for blueberry maggot adult monitoring (Liburd et al. 2000). Blueberry maggot management strategies include reduced risk insecticides, i nsecticide treated spheres, and biological control ( Diachasma alloeum (Muesebeck) (Hymenoptera: Braconidae)) (Liburd and Arevalo 2006). Cranberry Fruitworm : Acrobasis vaccinii Riley (Lepidoptera: Pyralidae). The cranberry fruitworm is a pest in all bluebe rry growing regions in the eastern United St ates (Turner and Liburd 2007). In addition to blueberries, it feeds on huckleberries, apples, and cranberries

PAGE 21

21 (Beckwith 1941). Eggs are laid at the blossom end of unripe fruit and hatch within five days (Turner and Liburd 2007). Newly hatched larvae enter blueberries at the stem end. They tunnel to the center, consuming pulp and the seeds (Beckwith 1941). Cranberry fruitworm can feed on multiple berries (up to eight) during larval development (Turner and Libu rd 2007). Damaged blueberry fruit clusters are covered with a silken web. Cranberry fruitworm can be monitored effectively with winged traps baited with cranberry fruitworm pheromone lures (Sarzynski and Liburd 2003). These traps should be placed on fiel d borders and in the center of the field. Blueberry Bud Mite : Acalitus vaccinii (Keifer) (Arachnida: Acari: Eriophyidae). The blueberry bud mite was first recognized as an economic pest of blueberries in 1940 (Fulton 1940). Highbush blueberries were obs erved to be heavily infested with bud mites while lowbush blueberries were moderately infested (Neunzig and Galletta 1977). This mite is found mainly under the large outer bud scales of blueberry bushes (Fulton 1940). Development from egg to adult takes 15 days (Baker and Neunzig 1970). The highest concentration of mites is found on terminal buds. Monitoring can be difficult due its small size (200 microns). Cells fed on by blueberry bud mite become enlarged, causing tissue to become blistered and redd ened followed by necrosis (Baker and Neunzig 1970). If mite infestation is not high enough to kill the bud, the number of berries produced is limited and those that are produced are blistered (Baker and Neunzig 1970). For chemical control to be effective acaricides need to be applied at high pressure to penetrate bud scales and reach bud mites. Endosulfan is registered for blueberry bud mite control in Florida (Turner and Liburd 2007). Flower Thrips: Frankliniella spp. (Thysanoptera: Thripidae). In Florida and southern Georgia, flower thrips are key pests of early -season blueberries (Arevalo 2006). Three species of thrips are pests of blueberries in the southeastern United States: Frankliniella bispinosa

PAGE 22

22 (Morgan), F. tritici (Lindeman) and F. o ccidentalis (Pergrande) (Turner and Liburd 2007). The adult and first two larval instars feed on pollen, flower parts (styles, ovaries and petals) and fruit (Turner and Liburd 2007). Damage causes reduced fruit set and deformed or softened berries. Acet amiprid and spinosad are effective insecticides for flower thrips control (Arevalo 2006). Diseases of Blueberry Important diseases include phytophthora root rot ( Phytophthora cinnamomi ), mummy berry (Monilinia vaccinii-corymbosi ), stem blight ( Botryosphaer ia spp. ), phomopsis canker and twig blight ( Phomopsis vaccinii ), and stem canker ( Botryosphaeria corticis ) (Cli ne and Schilder 2006). Blueberry Gall Midge Blueberry gall midge is indigenous to North America and has a wide geographic range. It is found in most blueberry and cranberry growing regions throughout the southeast, midwest and west coast. Blueberry gall midge feeds on the buds of Vaccinium species (blueberries and cranberries) and is a serious pest in both crops (Sampson et al. 2006). I t is also known as the c ranberry tipworm. Blueberry gall midge has recently been reported on blueberries in the Pacific Northwest (Yang 2005). It also appears to have been introduced into Europe where it was identified on blueberries in Italy (Bosio et al. 1998). Taxonomy The blueberry gall midge ( Dasineura oxycoccana (Johnson)) belongs to the family Cecidomyiidae. There are over 1200 named species in this family known from North America with 900 species known or presumed to be associated with plants (Gagn 1989). The genus Dasineura is poorly defined and the type species is unknown (Gagn 1989). Two of the most serious pests in Cecidomyiidae are the Hessian fly, Mayetiola destructor Say, and the sorghum midge, Contarinia sorghicola (Coquillett). Other pest cecidomyiids include the swede midge

PAGE 23

23 (Contarinia nasturtii (Kieffer)), pea midge (C. pisi Winn.), and wheat midge ( Sitodiplosis mosellana (Ghin)) (Hillbur et al. 2005; Hillbur et al. 2001; Hinks and Doane 1988). Other pests in the genus Dasineura include the brassica pod midge ( D. brassicae Winn.); blackcurrant midge (D. tetensi (Rbsaamen)), apple leafcurling midge (D. mali Keiffer), and pear leafcurling mid ge (D. pyri (Bouch)) (Williams and Martin 1986; Garthwaite and Wall 1986; Harris et al. 1996; Gagn and Harris 1998). Blueberry gall midge was first described under the name Cecidomyia vaccinii by Smith in 1890, and later renamed Cecidomyia oxycoccana (Jo hnson 1900, Barnes 1948). Felt re described the species and placed it in the genus Dasyneura (= Dasineura) (Barnes 1948). Life Cycle In Florida, blueberry gall midge overwinters as a pupa in the soil below blueberry bushes. Adults emerge early in the sea son beginning in late January to early February (Dernisky et al. 2005). In Mississippi, adults may emigrate from nearby overwintering hosts such as wild Vaccinium species (Sampson et al. 2002). Adult gall midges are typically short lived. Most only live for a day or two, long enough to mate and lay eggs (Gagn 1989). A female may deposit a few (10 15) eggs per blueberry bud (Bosio et al. 1998). A bud may contain eggs from more than one female when the infestation level is high (Bosio et al. 1998). Ad ults are 2 3 mm in length with females being slightly larger than males (Bosio et al. 1998). Females have expanded abdomens, which are orange in color and terminate in an elongate ovipositor (Bosio et al. 1998). When extended, the ovipositor is nearly as long as the body (Bosio et al. 1998). Males have slender abdomens, which are yellow in color and terminate in a forceps str ucture which is part of the geni talia (Bosio et al. 1998). Antennae are moniliform and bear hair like sensory receptors. Silfer and Sekhon (1971) described four sensillum types on the antennal flagellum of the sorghum midge: tactile hairs, thin -walled chemoreceptor pegs,

PAGE 24

24 circumfila and short stubby pegs. Circumfila, which are unique to Cecidomyiidae, encircle each flagellomere and in some species form long loops (Gagn 1989). Males have longer antennae and sensory receptors than females. The flagellomeres of male antennae are each made up of two separate nodes joined by short cylindrical stalks (Gagn 1989). The segments of a females antenna do not possess the cylindrical node and the segments appear to fit tightly together. Wing venation is reduced (only three major veins), and wi ngs are covered with small hairs (microtrichia) (Bosio et al. 1998). Eggs are deposited only after the buds begin to expand. Dormant flower buds are not attacked by blueberry gall midge (Lyrene and Payne 1995, Dernisky et al. 2005). The eggs have a cylin drical -elliptical shape and are approximately 0.25 mm long (Bosio et al. 1998). Early bud -swell appears to be a period of high susceptibility (Lyrene and Payne 1995). A high degree of variability exists in the susceptibility of flower buds to gall midge damage (Lyrene and Payne 1995). The flower buds of most southern highbush clones show high resistance to blueberry gall midge, but the vegetative meristems are vulnerable (Lyrene and Payne 1995, Sarzynski and Liburd 2003). Most southern highbush flower b uds expand before blueberry gall midge populations reach peak levels (Sarzynski and Liburd 2003). Lyrene and Payne (1995), however, dismiss timing of flower bud expansion as an explanation of resistance. The difference in susceptibility to gall midge betw een southern highbush and susceptible rabbiteye cultivars such as Premier and Climax could be due to differences in structural or nutritional values of the buds (Sarzynski and Liburd 2003). Resistance could also vary based on surface lipid composition of the blueberry buds (Sarzynski and Liburd 2003). Plant epicuticular lipids influence a number of factors in plant insect interactions including insect movement,

PAGE 25

25 feeding, host selection, and oviposition behavior (Eigenbrode and Espelie 1995). T he mecha nism of resistance is unknown (Lyrene and Payne 1995). Blueberry gall midge larvae feed on interior bud tissues inducing necrosis and bud abortion (Sampson et al. 2002). Contrary to the common name blueberry gall midge, this insect does not form galls. L arvae are spindleform and when newly hatched they are colorless. There are three larval instars (Gagn 1989). In successive instars, the larvae become darker in color, going from white to orange. Mature larvae range from 1.75 1.85 mm in length (Bosio et al. 1998). Under laboratory conditions, larvae require approximately ten days to develop (Bosio et al. 1998). The third instar larva changes shape slightly as pupal tissues begin to develop. This is sometimes called a prepupal stage (Gagn 1989). Ma ture larvae leave the blueberry buds and fall to the ground where they pupate (Bosio et al. 1998). Later instars possess an elongate, sclerotized, epidermal structure on the venter of the prothorax called a spatula. This structure is characteristic of ga ll midge species that pupate outside of a plant host (Gagn 1989). The spatula is used to dig in the soil and for movement, enabling larvae to vault several centimeters (Gagn 1989). Gall -forming species use the spatula to cut exit channels in galls (Gagn 1989). According to Milne (1961), the main function of the sterna spatula is to excavate a pupal chamber in the soil. Fully grown blueberry gall midge larvae form silken cocoons in which they pupate (Lyrene and Payne 1995; Mahr and Kachadoorian 1990). Pupae are approximately 1.1 1.3 mm long and are orange in color with black appendages (Bosio et al. 1998). The head possesses a pair of elongate processes (Bosio et al. 1998). Pupation time is variable. According to Bosio et al. (1998), the pupal sta ge lasts for eight to ten days, but Cockfield and Mahr (1994) state that pupation takes only three days.

PAGE 26

26 Under ideal conditions, blueberry gall midge takes from two to three weeks to complete its lifecycle (Cockfield and Mahr 1994, Bosio et al. 1998). The re are multiple overlapping generations each year (Sarzynski and Liburd 2003). Climate influences the number of generations. On rabbiteye blueberries in Florida, there are five to six generations of blueberry gall midge each year (Bosio et al. 1998). Sa mpson et al. (2002) reported up to 11 generations per year in southeastern Mississippi. Pest Status Initially, the frequent occurrence of high levels of flower bud abortion was attributed to lack of chilling, freeze damage, or high temperatures during flow er bud development. The problem became more severe following mild winters. The association between blueberry gall midge and flower bud abortion was made in 1992 by researchers at the University of Florida (Lyrene and Payne 1995). During the past two deca des, blueberry gall midge has become a severe pest of cultivated blueberries along the Gulf Coast (Sampson et al. 2002). Heavy infestations have also jeopardized the production of rabbiteye blueberries in Florida (Sampson et al. 2002). Although blueberry gall midge infests highbush and rabbiteye blueberries, damage to rabbiteye is more severe (Dernisky et al. 2005). Blueberry gall midge can destroy over 80 90% of flower buds in susceptible cultivars (Sampson et al. 2002, Sarzynski and Liburd 2003, Lyre ne and Payne 1992). Larvae feed in rabbiteye flower buds at the stage of development where inflorescences are still covered by bracts (Dernisky et al. 2005). In Georgia, lower temperatures sometimes prevent early gall midge emergence allowing flower buds to pass through the vulnerable stages before gall midge populations build up (Lyrene and Payne 1995). In Florida, however, warmer winter temperatures allow blueberry gall midge to emerge when buds are highly susceptible to infestation (Lyrene and Payne 1 995). Blueberry gall midge infests leaf buds following

PAGE 27

27 vegetative bud break (Lyrene and Payne 1992). Larvae feed on the youngest leaves, damaging the inner surface and causing a cupping of the leaves (Barnes 1948, Lyrene and Payne 1992). As the larvae c ontinue to feed, they kill the vegetative meristem resulting in growth of lateral meristems (Lyrene and Payne 1992). As a result, more vegetative buds are produced than flower buds, limiting fruit production the following year (Bosio et al. 1998). Emergen ce In those species of Cecidomyiidae where seasonal activity has been studied, emergence appears to be periodic (Pivnick and Labbe 1992). Peak emergence occurs generally at dawn or dusk and varies from species to species (van Lenteren et al. 2002). Blueb erry gall midge adults emerge from the soil in late January to early February and move to host plants (Dernisky et al. 2005). The sorghum midge emerges in the morning shortly after sunrise while Aphidoletes aphidimyza (Rondani), an aphid predator, emerges at dusk (Sharma and Vidyasagar 1992; van Lenteren et al. 2002). Protandry (males emerging before females) has been observed in several gall midge species (Heath et al. 2005; Pivnick and Labbe 1992; Sharma and Vidyasagar 1992). Swede midge males emerge on e day in advance of females, and sorghum midge males emerge one -half to one hour earlier than females (Hillbur et al. 2005; Sharma and Vidyasagar 1992). Both males and females are capable of mating shortly after emerging (Pivnick 1993). Adult gall midges only live for a few days so a synchronous pattern of emergence is important for mating success (Pivnick and Labbe 1992). Apple leaf curling midge adults remain on the ground until mated. This keeps virgin adults close together in space and decreases mate finding time (Heath et al. 2005). Sorghum midge adults form swarms at the crop canopy where mating takes place (Sharma and Vidyasagar 1992). In addition to synchronized emergence, sex pheromones are used to improve mate finding efficiency.

PAGE 28

28 Mating Behavior and Pheromones Few detailed studies have been performed on the reproductive behavior of cecidomyiids (McKay and Hatchett 1984). Adults of most species are short lived, difficult to observe in the field, and hard to rear in the laboratory (McKay a nd Hatchett 1984). Sex pheromone communication may be common in Cecidomyiidae, but has only been demonstrated behaviorally in a small number of species (Harris et al 1996). Out of the approximately 5000 described species in Cecidomyiidae, there is eviden ce of female-produced pheromones in twelve species (Heath et al. 2005). These species tend to be economically important crop pests. Clearly much more work can be done in this area as it is likely that many more gall midge species produce pheromones. Ori entation to pheromones over long distances is more common in Nematocera than in the other suborders of Diptera. In Cyclorrhapha, sex pheromones appear to elicit short range courtship and copulating behaviors (Harris and Foster 1991). Female gall midge cal ling behavior is characterized by full extension of the ovipositor (Hillbur 2001). Ovipositor extension by virgin females prior to mating occurs has been observed for Dasineura alopecuri D. leguminicola Rhabdophaga triadperd, D. violae Rhopalomyia hirtipes Contarinia oregonensis and Mayetiola destructor (Williams and Martin 1986). Female calling behavior follows a diurnal rhythm similar to the pattern of adult emergence (McKay and Hatchett 1984). Females begin calling shortly after emergence, and ca lling is most intense during evening and night and declines during the day (Pivnick 1993). Male gall midges respond to sex pheromones with rapid wing vibration a nd directed flight toward calling females Males generally do not show courtship behavior (McK ay and Hatchett 1984). First contact with calling females generally results in successful copulation. Females are monogamous while males may mate several times during their lifespan (Solinas and Isidoro 1991; van Lenteren et al. 2002). Females do not ex tend their ovipositors after mating, and males

PAGE 29

29 do not respond to mated females (McKay and Hatchett 1984; Williams and Martin 1986; van Lenteren et al. 2002). The gall mi dge ovipositor consists of the eighth, ninth and 10th abdominal segments (Solinas and Isidoro 1991). Sex pheromones are produced in secretory cells found in the epidermis of the interseg mental membrane connecting the eighth and ninth abdominal segments (van Lenter en et al. 2002). These cells possess large nuclei and the cytoplasm contains very extensive smooth endoplasmic reticulum, abundant ribosomes, numerous mitochondria with whorled cristae, well developed Golgi apparatus, and numerous electron -dense membrane -bound vesicles (Solinas and Isidoro 1991). These are characteristics typical of pheromone producing cells. Overlying cuticle does not show any obvious aperture for pheromone release. The outer surface of the membrane has regular longitudinal striations made up of parallel series of minute protuberances separated by grooves (Soli nas and Isidoro 1991). Secretions appear to be stored in these grooves when the ovipositor is retracted (Solinas and Isidoro 1991). These cuticular features have also been found in the pheromone glands of Lepidoptera (Solinas and Isidoro 1991). Sex pheromones have been chemically identified for five gall midge species: Mayetiola destructor Contarinia pisi C. oregonensis Sitodiplosis mosellana, and Aphidoletes aphidimyza (Heath et al. 2005). The Hessian fly, Mayetiola destructor was the first cecidomyiid for which a sex pheromone component was identified ((2S )( E ) 10-tridecen 2 yl acetate) (Foster et al. 1991). The pea midge, Contarinia pisi was the first gall midge for which the full sex pheromone blend was identified (Hillbur et al. 2001). The pea midge pheromone blend consists of three components: 2acetoxytridecane, (2 S ,11 S )diacetoxytridecane, and (2 S ,12 S ) -diacetoxytridecane (Hillbur et al. 2001). As a general rule, sex pheromones consist of chemical blends not single

PAGE 30

30 components (Harri s and Foster 1991). All the identified components of gall midge pheromones so far are chiral compounds. The molecular structure of pheromone components appears to be highly conserved within Cecidomyiidae (Hillbur et al. 2005). Monitoring Monitoring is an important component of any successful pest management program. Blueberry gall midge can be difficult to detect. Sticky boards have been used in some plantings to catch adult midges, but have proven ineffective (Sarzynski and Liburd 2003). Incorporating a lure based on a pheromone or host -volatile could improve trap catch numbers, but no such lure has been developed for blueberry gall midge. Key morphological features can be damaged or obscured by the traps sticky surface making subsequent identificati on difficult (Sarzynski and Liburd 2003). Traps have been designed to catch adults of some cecidomyiid species as they emerge from the soil, but this has not been tried for blueberry gall midge (Smith and Chapman 1996; Pea and Duncan 1992; Akar and Osgood 1987; NDSU n.d.). Effective methods for monitoring blueberry gall midge include an emergence technique and bud dissection. The emergence technique involves removing buds, holding them in a growth chamber, and observing any emerging larvae or adults. This technique proved most effective for detecting adults (Sarzynski and Liburd 2003). However, there is a lag time to detect gall midge infestation since growers have to wait until larvae or adults emerge. An emergence trap that can detect adults in the f ield at an earlier stage will give growers more time to implement a management program. The emergence technique and bud dissection were equally effective for detecting larvae. The only method for detecting eggs was bud dissection (Sarzynski and Liburd 2003).

PAGE 31

31 Management Insecticides Three main pesticides are used for blueberry gall midge control on blueberries: diazinon, malathion, and spinosad (Liburd 2004). To be effective, insecticides have to be applied early. Blueberry gall midge lays eggs shortly a fter emerging and the eggs are deposited in areas that may be sheltered from insecticides (Lyrene and Payne 1995). Larvae feed within developing buds and are sheltered from aerially applied insecticides. Control of larvae may be possible using systemic i nsecticides (Bosio et al. 1998). The short development time of blueberry gall midge makes timing of insecticide application critical for effective control. Insecticide resistance management is important because blueberry gall midges short life cycle and multiple generations per year make the development of resistance a possibility (Sampson et al. 2002). Biological c ontrol Native natural enemies of blueberry gall midge include larval predators like Toxomerus marginatus (Diptera: Syrphidae) and parasitoid Hymenoptera (Sampson et al. 2002). Parasitoids of blueberry gall midge found in Wisconsin cranberry bogs include Aprostocetus sp. (Hymenoptera: Eulophidae), Aphanogmus sp. and Ceraphron pallidiventris Ashmead (Hymenoptera: Ceraphronidae) (Sampson et al. 2 002). In the southern United States, parasitoids have been identified from the families Eulophidae ( Aprostocetus sp. and Quadrastichus sp.) and Platygastridae ( Synopeas sp. and Platygaster sp.) (Sampson et al. 2002, 2006). Sampson et al. (2006) found 30 to 40% parasitism in midge larvae collected from blueberry nurseries in Mississippi, Louisiana, and Alabama. Most parasitoids of blueberry gall midge do not arrive early enough in the season to protect floral buds from damage (Sampson et al. 2002). An ear ly insecticide treatment may be necessary to reduce blueberry gall midge populations until parasitoids can reach sufficient numbers. Post -

PAGE 32

32 bloom insecticide applications could, however, spur gall midge infestations by removing parasitoids and other natural enemies (Sampson et al. 2002). An IPM strategy for blueberry gall midge in blueberry should, therefore, include preservation of natural enemies and augmentation by safe, precise, and effective timing of insecticides (Sampson et al. 2002). Cultural c ontro l Shallow disking beneath blueberries may kill diapausing blueberry gall midge larvae in the soil (Steck et al. 2000). Spreading sand over cranberry beds interferes with blueberry gall midge adult emergence (Lyrene and Payne 1995). Much research is neede d in the areas of cultural control and biological control of blueberry gall midge. Host plant resistance could prove to be effective in reducing blueberry gall midge damage. Screening for resistance to blueberry gall midge has been included in blueberry b reeding programs in Florida (Sampson et al. 2002).

PAGE 33

33 CHAPTER 3 EVALUATION OF EMERGE NCE TRAPS FOR MONITORING BLUEBERRY GALL MIDGE DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE) Monitoring is an important component of any successful pest ma nagement program. Effective monitoring of midge populations should detect adult emergence and population fluctuations. The blueberry gall midge, Dasineura oxycoccana (Johnson), is a key pest of rabbiteye blueberry ( Vaccinium ashei Reade) in Florida (Sampson et al. 2002). In susceptible cultivars, D. oxycoccana feeding can destroy 80 90% of the flower buds (Lyrene and Payne 1992, Sampson et al. 2002, Sarzynski and Liburd 2003). Larvae feed inside developing buds, and are shelter ed from foliar insecticides. Control of larvae may be possible using systemic insecticides (Bosio et al. 1998). However, the short development time of D. oxycoccana (two to three weeks) makes timing of systemic insecticide application critical for effect ive control (Bosio et al. 1998). Dasineura oxycoccana adults are difficult to de tect due to their small size (2 3 mm). Sticky traps have been used in some plantings to catch adult midges, but have proven ineffective (Sarzynski and Liburd 2003). Furthermore, key morphological features can be damaged or obscured by the traps sticky surface, making subsequent identi fication difficult (Sarzynski and Liburd 2003). Incorporating a lure based on a pheromone or host volatile could improve trap catch numbers, but no such lure has been developed for D. oxycoccana Traps have been designed to catch adults of some cecidomyii d species as they emerge from the soil. Specialized traps have been developed for apple leafcurling midge ( Dasineura mali Kieffer), balsam gall midge ( Paradiplosis tumifex Gagn), orange wheat blossom midge (Sitodiplosis mosellana Ghin), and Prodiplosis longifila Gagn, but these traps have not been used for monitoring D. oxycoccana (NDSU n.d, Akar and Osgood 1987, Pea and Duncan 1992, Smith and Chapman 1996).

PAGE 34

34 Sarzynski and Liburd (2003) investigated methods for detecting D. oxycoccana and found that an emergence technique and bud dissections were useful for monitoring populations in the field. The emergence technique involved removing flower buds, holding them in a growth chamber for 10 to 14 days, and counting larvae or adults as they emerge. With thi s technique, however, there is a time lag between oviposition and mature larvae emerging from buds. Larvae begin emerging from buds within hours of being collected, but we have observed that the buds need to be held for about 10 days to obtain accurate counts of population density. An emergence trap that can detect adults in the field at an earlier stage would give growers more time to implement a management program before midge populations reach damaging levels. Although bud dissection is a good technique for detecting eggs (Sarzynski and Liburd 2003), it is time consuming requires visual aid, and is not practical for growers (Finn 2003). This study was conducted to test some emergence traps that can be used for rapid detection of D. oxycoccana adults i n the field. The objectives were to identify the most effective trap, in terms of the number of adults caught, and determine if trap catch can be used to predicted outbreaks of larvae infesting blueberry flower buds. Materials and Methods This study was c onducted at an organic rabbiteye b lueberry farm in Gainesville, Florida All plants were at least 1.5 m tall. Two plots (designated A and B), each 0.15 ha in size, were established in two separate 1 -ha plantings. Two experiments were conducted, one in 2007 and the other in 2008. Three different trap designs were evaluated in each experiment, described separately for each experiment (see below). Each year the same experiment was conducted in both plots for the purpose of replication. In each test, expe rimental design was randomized complete block with four replications (blocked by variety) and three treatments (trap types) for

PAGE 35

35 each plot. Trap designs were modifications based on descriptions from Smith and Chapman (1996), who constructed traps from 10 l iter plastic buckets and adhesive -coated Petri dish tops or plastic funnels with specimen containers. In all experiments, traps were placed on the soil surface approximately 0.3 m from the trunk of a blueberry bush. Traps were rotated and moved to a new location within the same replication each week to avoid bias and prevent trapping out all the midges beneath each trap position. Soil was piled over the edges of the traps to prevent midges from escaping. Trap tops were removed each week and replaced wi th fresh adhesive or a clean glass jar. The trap tops were brought to t he Small Fruit and Vegetable Integrated Pest Management Laboratory at the University of Florida Department of Entomology and Nematology ( Gainesville FL) where adult midges were counte d, removed and placed in vials wi th 70% ethyl alcohol Adults caught on Tangle Trap (The Tanglefoot Company, Grand Rapids, MI) were soaked in Histo-clear (National Diagnostics, Atlanta, GA) to remove the adhesive before transferring to alcohol. In order to compare bud infestation with trap captures, 25 flower buds (development stage tw o or three according to Spiers 1978) were collected each week from the two blueberry plants adjacent to each trap position. Buds were placed in Petri dishes with moistened filter paper and held in a growth chamber (Percival Model I 35 LL, Percival Mfg. Co ., Boone, IA) for two weeks at 30 2 C (day) and 20 2 C (night) with 14L:10D photoperiod for larval emergence. 2006 Pilot Test This pilot test was conducted in one 0.15-ha plot (plot A) located at an organic blueberry farm in Gainesville, FL. The thr ee trap designs evaluated were 1) Glass jar trap, 2) Petri dish trap, and 3) Wheat blossom midge trap (Fig ure 3 1). All traps were constructed from 3 L plastic food containers. The exterior of the traps were painted white (Krylon Interior Exterior, 1502 Flat White, Krylon Products Group, Cleveland, OH). The only light entering the traps came

PAGE 36

36 from the top in order to exploit the positive phototactic behavior of D. oxycoccana The glass jar trap consisted of an inverted plastic funnel and a 473-ml (1 pint ) glass jar for collecting midges. The top of the Petri dish trap was a 14-cm -diameter Petri dish, the underside of which was coated with Tangle Trap (The Tanglefoot Company, Grand Rapids, MI). Alternatively, the wheat blossom midge trap had a transluce nt plastic lid with a 7.6 cm screen covered opening in the lid and two 5.1 cm diamete r openings in the sides (NDSU n.d.). The underside of the lid was coated with vegetable oil spray. All traps received fresh adhesive or a new jar each week. Traps were d eployed in plot A from 3 February to 10 March. Within each replication, traps were placed beneath every fifth blueberry bush (10 15 m between traps). 2007 Experiment The same three trap designs used in the 2006 pilot test were evaluated again in 2007. The experiment was conducted twice using plot A and plot B (both 0.15 ha in area) Captured midges were sexed, and sex ratios were calculated. Experimental design was the same as in 2006. Within each replication, traps were spaced 15 m apart. Traps wer e deployed in the field from 17 January to 11 April. 2008 E xperiment This experiment was conducted in the same two plots used in 2007. The three trap designs evaluated were: 1) Petri dish trap, 2) Modified Petri dish trap, and 3) Bucket trap (Fig ure 3 1 ). Both the glass jar and wheat blossom midge traps from 2007 were excluded in 2008 because captures of D. oxycoccana were very low in these traps in 2007. The Petri dish trap was the same as the one used the previous years (2006 & 2007). The modified Pet ri dish trap was similar to the one used in 2007 except the interior was painted black (Krylon Interior -Exterior 1613 Semi Flat Black, Krylon Products Group, Cleveland, OH) to determine if darkening the trap interior would increase trap captures due to an increased response to light coming from the top of

PAGE 37

37 the trap. The bucket trap, a new trap design, was constructed from the bottom of a white 18.9liter (5 -gallon) bucket. The interior was painted black, and a 19.4 -cm -diameter hole was cut in the bottom an d covered with a 21.3-cm -diameter acrylic sheet (3 -mm thick) coated on the underside with Tangle Trap. Experimental design was the same as in 2006 and 2007. In both plot A and B, traps were spaced 10 m apart within each replication. Traps were deployed in the field from 9 January to 30 April. Statistical Analysis The number of adults captured per week was compared across trap type s using repeated measures ana lysis of variance (ANOVA) (PROC MIXED, SAS Institute 2003) Data were transformed by log10(x+1) to satisfy model assumptions. Untransformed means and standard errors are reported in tables and figures. D ifferences among me ans were determined using the least -squares m eans (LSMeans ) mean separation test ( P = 0.05). Differences between the number of males and females captured for each trap type was analyzed using t -tests For each trap type, the zero frequency of trap catch was calculated. Zero frequency is the percentage of traps that caught zero adults over a one week sample period. Results 2006 Pilot Test The Petri dish trap caught significantly more midge adults than the glass jar trap or wheat blossom midge trap (Table 3 1). Interaction effects for trap captures by time were not significant (F = 1.03; df = 8, 45; P = 0.427). The percentage of traps that caught zero midges was lowest for the Petri dish trap (Table 3 1). Peak trap catch occurred on 24 February for the Petri dish and glass jar traps, and on 3 March for the wheat blossom midge trap.

PAGE 38

38 2007 Experiment Again, the Petri dish trap caught significantly more midge adults than the other trap designs at both plots (Table 3 2). Interaction effects for plot A traps by time were not significant (F = 1.28; df = 22, 103; P = 0.202). Interaction effec ts for plot B traps by time were significant for two out of the 12 weeks ( F = 2.34; df = 22, 69; P = 0.004). The percentage of traps with zero catches was lowest for the Petri dish trap and highest for the wheat blossom midge trap (Table 3 2). The sex ra tio of captured adult midges was not significantly different from 1:1 for any of the trap types (Table 3 3). Petri dish traps captured the most adults in the first week of sampling (Fig ure 3 2 ). Peak trap catch occurred on 21 March for all three trap des igns. Trap catch and bud infestation were compared over a six -week period (24 January to 28 February). Over this period, no pattern between trap catch and bud infestation was observed (Fig ure 3 3 ). Larval infestation increased over the six weeks while tr ap catch for adults decreased. 2008 Experiment The number of adult midges caught in the bucket trap was significantly higher than the other trap designs in plot A ( F = 13.84; df = 2, 94.5; P < 0.001), but no difference due to trap type was observed in plot B ( F = 2.26; df = 2, 95.5; P = 0.110) (Table 3 4). Interaction effects for trap type by time were not significantly different ( F = 1.49; df = 30, 94; P = 0.075 and F =1.22; df = 30, 95; P = 0.236; plots A and B, respectively). The percentage of traps with zero catches was lowest for the bucket traps in plot A (Table 3 4). In plot B, all three trap types had similar percentages of zero catches. The original and modified Petri dish traps tended to capture more males than females, but sex ratios were not significantly different from 1: 1 (Table 3 5). The bucket trap, however, caught significantly more males than females. Sex ratios for the bucket traps were 3.8: 1 and 2.4: 1 males to fema les (plots A and B, respectively).

PAGE 39

39 All traps caught adults in the first week of sampling. In plot A, bucket trap catch peaked on 12 March and original Petri dish trap catch peaked on 2 April (Figure 3 4 ). In plot B, bucket trap catch peaked on 12 March and original Petri dish catch peaked on 5 March. The modified Petri dish traps did not have any large peaks in trap catch. Peak trap catch was compared with peak numbers of larvae collected from flower buds (Fig ure 3 5 ). The peak in trap catch occurred one week before peak in larval infestation for the original Petri dish traps in plot B (Figure 3 5 D), modified Petri dish traps in plot A (Figure 3 5 B) and bucket traps in plot A (Fig ure 3 5 C). For the modified Petri dish traps in plot B, peak trap catc h occurred two weeks before the peak in larval numbers (Fig ure 3 5 E). Peaks in trap catch and larval infestation occurred on the same week for original Petri dish traps in plot A (Fig ure 3 5 A) and bucket traps in plot B (Fig ure 3 5 F). Discussion In 2006 and 2007, the Petri dish trap caught more midge adults than the other trap designs. Trap effectiveness may be due to the type of trapping surface. The glass jar trap was not as effective as the Petri dish possibly because of the greater distance midges have to travel to reach the collecting jar. In addition, midges could possibly escape the jar by crawling back through the funnel. As a result, this treatment was eliminated from subsequent studies. The wheat blossom midge trap and Petri dish trap used adhesive surfaces to trap midges. The top of the Petri dish trap was transparent while the top of the wheat blossom midge trap was translucent. The transparent top allows more light through and may have provided a stronger positive phototactic cue. Thi s was the justification for darkening the interior of the modified Petri dish trap in 2008. A high percentage of the wheat blossom midge traps did not catch any midges (86.8% in plot A, 74.3% in plot B) in 2007, so this trap was eliminated from subsequent studies. The wheat

PAGE 40

40 blossom midge trap performed better in the 2006 pilot study than in 2007. The 2006 study period, however, was shorter than in 2007 (five and seven weeks, respectively), and traps were removed from the field prior to peak adult activi ty in mid to late March as indicated by trap captures in 2007 and 2008 (Figures 3 3 and 3 5). Modifying the Petri dish trap in 2008 did not significantly affect trap catch. The response to light coming through the transparent top was not enhanced by darke ning the trap interior. The higher trap catch in the bucket trap may be due to the greater soil area covered and larger trapping surface. The soil area covered was 14% greater than the Petri dish trap, and the trapping area was 78% larger than that of the Petri dish trap. The increase in trap catch, however, is not proportional to either increase in area. The amount of light entering the trap top may explain trap efficacy. The acrylic panel of the bucket trap appeared to allow more light through than t he Petri dish, which became more translucent after repeated applications of Tangle Trap even though tops were cleaned. In plot A, the bucket trap caught significantly more D. oxycoccana adults than the other trap types, but in plot B it was not significan tly different. This could reflect a difference in trap performance, however, the proportion of adults trapped in bucket traps compared to the other trap types was the same for A and B. In both plots, the bucket trap caught approximately twice as many ad ults as the original Petri dish trap and two anda half times as many as the modified Petri dish trap. Non -significant differences in plot B appear to be the result of high variance in the data, increasing the probability of a Type II error. Side -by-side comparison of graphs of trap captures and larval infestation for 2007 did not show any pattern of increased trap captures preceding increased larval infestation (Fig ure 3 3 ). This could not be explained by fluctuations in soil and air temperatures. The l ack of fit between

PAGE 41

41 trap captures and larval infestation in 2007 may be the result of ending the experiment too early. The sampling period for larvae and adults in 2007 covered five weeks (24 January to 28 February). In 2008, however, the sampling period covered 10 weeks (16 January to 26 March); with peak trap catch occurring in March. Therefore, it appears that in 2007, traps were removed from the field too early thereby missing peak adult emergence. The predictive ability of emergence traps appears to be limited to major peaks in insect population density. The 2007 traps could still prove useful in detecting the onset of D. oxycoccana flight early in the season. Emergence trap catch could be used to predict peaks in larval infestation. In most cases the peak in trap catch came one week before the peak in larval density. Gall midge eggs hatch within a few days of deposition (Lyrene and Payne 1995), so it is possible that larvae collected from flower buds in a given week hatched from eggs deposited by females flying the previous week. The number of traps used in this study, although adequate for comparing trap efficacy, may be too few to make accurate predictions of larval infestation. Further evaluation of these trap designs would be needed to correct for inconsistencies and make better predictions. Yet these traps may have some use in early season detection or detection of major peaks in D. oxycoccana population levels.

PAGE 42

42 Table 3 1. Emergence trap evaluation, pilot experiment 2006. Number of Dasineur a oxycoccana adults per trap (mean SEM) and frequency of traps with zero captures per week. Number of D. oxycoccana adults trapped Trap type T otal Weekly mean Zero frequency (%) Petri dish 142 7.15 1.90 a 10.0 Glass jar 40 2.00 0.96 b 70.0 Wheat midge 54 2.65 0.92 b 40.0 Means followed by the same letter are not significantly different ( F = 9.46; df = 2, 45; P < 0.001). Zero frequency refers to the percentage of traps that did not catch any D. oxycoccana adults over a one week sampling period. Table 3 2. Emergence trap evaluation for 2007. Number of Dasineura oxycoccana adults per trap (mean SEM) and frequency of traps with zero captures per week. Number of D. oxycoccana adults trapped T otal Weekly mean Zero frequency (%) Trap type Plot A Plot B Plot A Plot B Plot A Plot B Petri dish 71 193 1.48 0.35 a 4.02 1.16 a 56.3 52.1 Glass jar 39 48 0.83 0.43 b 1.00 0.36 b 78.5 54.2 Wheat midge 13 22 0.30 0.15 b 0.48 0.16 b 86.8 74.3 Means in columns followed by the same letter are not significantly different ( Plot A: F = 7.70; df = 2, 103; P < 0.001; Plot B: F = 14.0; df = 2, 70; P < 0.001). Zero frequency refers to the percentage of traps that did not catch any D. oxycoccana adults over a one week sampling period. Table 3 3. Mean number of Dasineura oxycoccana females and males per trap ( SEM) and sex ratio of captured adults in 2007. Mean D. oxycoccana adults trapped Plot Trap type Females Males Sex Ratio (F:M) A Petri d ish 0.73 0.17 0.75 0.21 1:1 .0 Glass j ar 0.38 0.22 0.45 0.20 1:1 .2 Wheat m idge 0.16 0.07 0.14 0.10 1:0.9 B Petri d ish 1.94 0.44 2.08 0.87 1:1.1 Glass j ar 0.38 0.11 0.63 0.28 1:1.7 Wheat m idge 0.30 0.11 0.17 0.08 1:0.6

PAGE 43

43 Table 3 4. Emergence trap evaluation for 2008. Number of Dasineura oxycoccana adults per trap ( mean SEM) and frequency of traps with zero captures per week. Trap catch Weekly mean Zero frequency (%) T rap type Plot A Plot B Plot A Plot B Plot A Plot B Bucket Trap 154 123 2.44 0.35 a 1.95 0.48 a 35.9 55.8 Original Petri 83 59 1.30 0.28 b 0.92 0.22 a 54.7 57.8 Modified Petri 57 51 0.89 0.19 b 0.80 0.15 a 62.5 57.8 Means in columns followed by the same letter are not significantly differe nt (Plot A: F = 13.8; df = 2, 95; P < 0.001; Plot B: F = 2.26; df = 2, 9 6 ; P = 0.110). Zero frequency refers to the percentage of traps that did not catch any D. oxycoccana adults over a one week sampling period. Table 3 5. Mean number of Dasineura oxycoccana females and males per trap ( SEM) and sex ratio of captured adults in 2008. Mean D. oxycoccana adults trapped Sex ratio P values Plot Trap type Females Males (F:M) from t tests A Bucket Trap 0.51 0.10 1.94 0.30 1:3.8 P < 0.001 Original Petri 0.45 0.12 0.84 0.19 1:1.9 P = 0.087 Modified Petri 0.38 0.10 0.52 0.13 1:1.4 P = 0.385 B Bucket Trap 0.57 0.16 1.38 0.35 1:2.4 P = 0.039 Original Petri 0.47 0.14 0.45 0.11 1:1.0 P = 0.930 Modified Petri 0.34 0.09 0.45 0.10 1:1.3 P = 0.409

PAGE 44

44 Figure 3 1 Emergence traps evaluated for effectiveness in catching Dasineura oxycoccana adults in 2007 and 2008. A) Glass jar trap ; B) Petri dish trap; C) Wheat blossom midge trap ; D) Bucket trap Traps A, B and C were evaluated in 2007; traps B, D, and a modified version of trap B were evaluated in 2008. B C D A

PAGE 45

45 Figure 3 2 Mean number of Dasineura oxycoccana adults captured in emergence traps in plots A and B 2007.

PAGE 46

46 A D B E C F Figure 3 3 Mean number of Dasineura oxycoccana adult s per emergence trap and mean number of larvae per blueberry flower bud collected at an organic blueberry farm in Gainesville FL in 2007. A -C: plot A traps; D -F: plot B traps.

PAGE 47

47 Figure 3 4 Mean number of Dasineura oxycoccana adults captured in emergence traps in plots A and B in 2008.

PAGE 48

48 A D B E C Figure 3 5 Mean number of Dasineura oxycoccana adults per emergence trap and mean number of larvae per blueberry flower bud collected at an organic blueberry farm in Gainesville, FL in 2008. A C: plot A traps; D F: plot B traps.

PAGE 49

49 CHAPTER 4 WITHIN FIELD DISTRIB UTION OF BLUEBERRY G ALL MIDGE, DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE) The distribution of insect pests is important information for assessing the risk of plant injury when developing management programs for a particular pest. Although the distribution of a population varies spatially and temporally, these factors are not always considered when managing pests. Insecticides, as well as other agricultural inputs, are traditionally applied uniformly to entire fields with little consideration for within -field variations (Weisz et al. 1996). S ite -specific integrated pest management is a system in which inputs are adjusted to match within -field requirements (Weisz et al. 1996). For pest management, this means implementing management strategies (such as insecticide use) only in those areas of the field where pest population density has reached the economic threshold. This not only has the potential to reduce the cost and ecological impact of excess insecticide applications, but also promotes resistance management by preserving refugia for susceptible individuals (Weisz et al. 1996). Increasing the number of refugia promotes natural enemy conservation (Weisz et al. 1996) and allows for more natural control. Blueberry gall midge, Dasineura oxycoccana (Johnson ), is a key pest of rabbiteye blueberry Vaccinium ashei Reade, in the southeastern United States (Dernisky et al. 2005). D asineura oxycoccana larvae feed in developing flower and leaf buds. In susceptible cultivars, over 80 90% of flower buds can be destroyed, reducing potential fruit yield (Sampson et al. 2002, Sarzynski and Liburd 2003, Lyrene and Payne 1992). Rabbiteye is more susceptible than southern highbush ( V. corymbosum L. V. darrowi Camp ) to D. oxycoccana damage, and production in Florida has been jeopardized due to heavy i nfestations (Sampson et al. 2002). A better understanding of D. oxycoccana distribution within blueberry plantings would make development of site -specific management programs possible. Currently, D. oxycoccana

PAGE 50

50 management in blueberries is limited to broad applications of insecticides such as diazinon, malathion, and spinosad (Liburd 2004). The organophosphates diazinon and malathion face increased scrutiny and restriction under the Food Quality Protection Act of 1996 (F ood Q uality P rotection A ct 1996). S ite -specific applications could prolong the usefulness of these products. The main objective of this study was to determine if D. oxycoccana infestation was greater at field borders compared with interior rows. Dispersal from overwintering sites or altern ative host plants into the main plant crop is common among agricultural pests (Kolesik 2000). This tends to result in a higher pest density in border rows early in the season. This edge effect is well known in the family Cecidomyiidae and has been observ ed in a number of economically important species including Contarinia lentis Aczl, Contarinia sorghicola (Coquillett), Contarinia schultzi Gagn, Dasineura mali (Kieffer) and Mayetiola destructor (Say) (Kolesik 2000; Lampo and Medialdea 1994; Hodgson et a l. 2004; Suckling et al. 2007; Withers et al. 1995). As a result, Kolesik (2000) was able to develop the basis for a control strategy that reduced the area in which chemical inputs were required to 10 300 m wide strips along the edges of lentil fields. Although D. oxycoccana overwintering and alternative host plants are currently unknown, it is hypothesized that during blueberry flower bud development, infestation is highest near field margins. Materials and Methods This study was conducted at an organi c blueberry far m in Gainesville, Florida The farm consisted entirely of rabbiteye blueberry plants which were established around 1983 All blueberry bushes in the study were at least 1.5 m tall. Bushes were planted 1.5 m apart within rows, and row centers were 3.7 m apart. At the time of the study, no pruning or weed control had been practiced for several years. Blueberry gall midge infestation was determined by

PAGE 51

51 counting the number of larvae, pupae, and adults that emerged from blueberry flower buds, and counting the number of eggs and larvae found in dissected flower buds. Flower bud samples were taken from the field in re -sealable plastic bags and processed in t he Small Fruit and Vegetable Integrated Pest Management Laboratory at the University of Florida Department of Entomology and Nematology (Gainesville, FL) All flower buds collected were in development stages two or three (Spiers 1978). Buds were placed in 9 -cm diam Petri dishes with moistened filter paper, sealed with Parafilm (Pechiney Pl astic Packaging, Menasha, WI), and placed in a growth chamber at 30 2 C daytime temperature, 20 2 C nighttime temperature, and 14:10 L:D photoperiod (Sarzynski and Liburd 2003). Flower bud samples were kept for two weeks before being discarded. 2006: Experiment to I nvestigate Dasineura oxycoccana Distribution Patterns with Variation in Field B orders The study was conducted in a 0.6 -ha plot (A) that was part of a 1.3 -ha field at the southwest corner of the farm. The field was bordered to the west and south by a city park consisting of diversified flora A drainage ditch was located 8 m to the west of the field (Fig ure 4 1). To the north was an uncultivated 0.5-ha field with mixed vegetation. The plot was 11.5 m from the northern boundary of the field and 7 m from the southern boundary. The plot consisted of the following three varieties of rabbiteye blueberries: Climax, Alice blue and Becky blue Rows were planted in pairs according to variety and alternated Climax, Becky blue and Alic e blue from north to south. Each row contained approximately 100 blueberry bushes. Samples were collected from bushes 1 5, 20 25, and 46 50. These correspond to distances from the western border of 0 5.5 m, 24.5 30 m, and 60.0 65.5 m, respec tively. For each distance, the bud collection area consisted of ten plants (five from two adjacent rows). Thirty

PAGE 52

52 flower buds were collected randomly from all ten plants once a week from 10 February to 10 March from each collection area. Twenty -five of t he buds were processed as previously described by allowing larva to exit the fruit (for larva collection) while the remaining five were dissected to look for eggs, as well as larvae that did not emerge. All samples were taken from the variety Climax. There were four replications each consisting of different rows of Climax and separated by four rows of the other varieties. 2007: Experiment to I nvestigate Dasineura oxycoccana Distribution Patterns in Isolated Plots with Potential Edge E ffects This study was conducted in a 0.13-ha plot (B) consisting of nine rows located on the north side of the farm. This plot was used because it was isolated by uncultivated fields and access roads from other blueberry plantings on the farm and, therefore, it was expect ed that edge effects would be easier to observe. The closest blueberry planting was 80 90 m to the southeast. Samples were taken from three rows: the southern border row, northern border row, and center row (Fig ure 4 2). Sampled rows were separated by three unused rows (14.6 m). Each week from 18 January to 8 March, five bushes were selected in each of the three sampled rows and 25 flower buds were collected from each bush. Individual bushes were treated as replications. Flower bud samples were take n from bushe s all along the north and south border rows. In the center row, sampling was restricted to plants in an 18 -m section near the center of the row to avoid edge effects in the east -west direction. 2008: Experiment to I nvestigate Dasineura oxycocc ana Distribution Patterns in I solated Plots with Potential Edge Effects: Second Y ear The second year of the study to investigate D asineura oxycoccana distribution patterns in isolated plots was conducted at the same site used in 2007 (B) [Figure 4 2]. The sampling

PAGE 53

53 protocol was the same as in 2007 except that 15 flower buds were collected per plant rather than 25, and samples were taken twice per week (as opposed to once per week in 2007) to determine if any rapid changes in D. oxycoccana populations occurr ed in isolated plots. Sampling began on 9 January. The increased frequency of sampling required that fewer buds be taken per sample in order to have sufficient buds to last until bloom. Statistical Analysis Dasineura oxycoccana infestation levels at field borders and centers were compared by analyzing the differences in the numbers of larvae per flower bud at different field locations. The data were analyzed using repeated measures ana lysis of variance (ANOVA) (PROC MIXED, SAS Institute 2003) Data were (x+1)1/2 transformed to satisfy model assumptions. Differences among means were estimated using the l east -s quares m eans (LSMeans) procedure (P < 0.05). In cases where significant interaction effects were observed, differences in infestation were tested by week using a two -way ANOVA (PROC GLM, SAS Institute 2003). Differences among means for each week were estimated using Least Significant Differences (LSD) ( P < 0.05). Untransformed means and standard errors are reported in t ables and figures. Results 2006: Experiment to I nvestigate Dasineura oxycoccana Distribution Patterns with Variation in Field B orders The effect of distance from the field border on D asineura oxycoccana infestation was not significant ( F = 2.46; df = 2, 30; P = 0.103). The number of D. oxycoccana larvae per flower bud in the 60 65 m plots was not statistically different (at P < 0.05) from the number of larvae in the edge plot 0 5 m from the field border (Fig ure 4 3). Interaction effects between distan ce and time were not significant ( F = 0.43; df = 8, 30; P = 0.894). Dasineura oxycoccana

PAGE 54

54 infestation increased from 17 February to its highest numbers on 10 March when sampling was concluded because flower buds became scarce (Fig ure 4 4). 2007: Experiment to I nvestigate Dasineura oxycoccana Distribution Patterns in Isolated Plots with Potential Edge E ffects Dasineura oxycoccana infestation was significantly different among the rows sampled ( F = 98.3; df = 2, 64; P < 0.001). Mean number of D. oxycoccana la rvae per bud was highest in the south border row and lowest in the north border row (Fig ure 4 5). There was a significant interaction between time and row ( F = 3.47; df = 14, 64; P < 0.001) (Table 4 1). At the beginning of the sample period, from 18 Janu ary to 8 February, midge infestation in the center and northern rows was low and differences between these rows were not statistically significant (Fig ure 4 6). Peak in D. oxycoccana larval density occurred during the week of 15 February. From 15 Februar y to 1 March, infestations in the center and southern border rows were not significantly different. 2008: Experiment to I nvestigate Dasineura oxycoccana D istribution Patterns in Isolated P lots Infestation of D asineura oxycoccana over the entire sampling pe riod was significantly different in each of the areas sampled ( F = 32.7; df = 2, 79; P < 0.001) (Fig ure 4 7). As in 2007, the southern border had the highest number of larvae per bud while the northern border had the lowest. There was a significant inter action between time and sample area ( F = 2.33; df = 12, 79; P = 0.013) (Table 4 2). Samples were collected consistently (twice per week) during mid January, but sampling had to be suspended from 23 January to 10 March due to a shortage of flower buds. The highest number of larvae was collected on 10 March.

PAGE 55

55 Discussion Although the edge effect is well documented for several species of Cecidomyiidae (Kolesik 2000), it was not observed in the 2006 experiment. The field in which the 2006 experiment was conducted was adjacent to other blueberry fields on the same farm. Fields were only separated by alleyways that were 8 10 m wide. Perhaps dispersing midges are moving between fields with well-established populations. Lamb et al. (1999) found that Sitodiplosis mosellana (Ghin) infestation did not differ within Manitoba w heat fields at distances of 1, 50, or 100 m from the edge. They did observe, however, that proximity of wheat fields accounted for 75% of the variation in S. mosellana infestation among fields (Lamb et al. 1999). The field used in 2007 and 2008 was an i solated plot (not adjacent to other blueberry fields), and D. oxycoccana infestations varied significantly by position in the field. There was a clear trend of increasing density of larvae per flower bud from south to north. The weekly collection data fr om 2007 show that not only do mean densities differ with sampling areas, but dynamics differ in the areas sampled as well. Larval density remained fairly constant in the northern border row over the entire sampling period (Figure 4 6). In the center row and southern border rows, however, the number of larvae increased at mid-season. Higher numbers of D. oxycoccana larvae in the southern border row may be the result of midges moving into this plot from other parts of the farm. The nearest blueberries wer e approximately 90 m to the south and southeast of this field. To the north and west there was only open ground, and to the east there was mostly woods. It appears that D. oxycoccana originated from some source to the south of the plot and moved north into the plot causing the population to increase in the southern border row followed by the center row. The sample period was not long enough to see if this trend would continue into the northern border row.

PAGE 56

56 Sample collection was halted prematurely in 2008 due to lack of flower buds in the center and south rows. Plants had not been pruned for a few years so there was little new growth and consequently few flower buds. The pattern of total infestation by area was the same as 2007, but due to intermittent sa mpling, the seasonal pattern could not be determined. The edge effect is contingent upon a dispersing population. Most studies of gall midge distribution are done either in temperate regions where midges overwinter, with annual host plants, or both. Blue berries, however, are perennial and can support D. oxycoccana populations through most of the year. Blueberry gall midge larvae feed in flower buds from bud break until bloom and in leaf buds from bloom until plants become dormant During the period when blueberry leaf and flower buds are not available, D. oxycoccana either overwinter or switch to an alternate host plant such as a wild Vaccinium species It is possible that in a large blueberry field, midges could be well -established in the center of the field throughout the entire year. This may be due to a more favorable microclimate (temperature, humidity, etc.). Conditions at the field borders may be more variable; therefore, midge populations are not as established. If midges were already present in the center of the field, an edge effect would not be observed if and when adults dispersed from nearby fields or alternate hosts. The edge effect was clearly seen in the isolated field used in 2007 and 2008. This field was in close proximity, but not a djacent to, previously infested blueberry fields. The edge effect was only observed from the south which was in the direction of infested blueberry fields. To the north of the field there were no blueberry plants or potential alternate hosts. The infest ation in the northern border row was much lower. This suggests that this field was re -infested each season. If D. oxycoccana simply overwintered in the soil beneath the blueberry plants, the infestation should have been more

PAGE 57

57 uniform. Adults could have r eached the northern border row through trivial movements from plant to plant while most remained in the first row encountered, the southern border row. An edge effect on infestation does not necessarily mean that site -specific pest management is possible. Other factors must also be considered. Hodgson et al. (2004) observed that most sunflower midge damage was concentrated in sunflower field edges, yet argued that site -specific management was not feasible. Adult sunflower midges have a prolonged emergenc e and infest and disperse through sunflower fields over a relatively long period of time (Hodgson et al. 2004). Management using insecticides would require multiple applications, especially if the products used have a short residual time. In Florida, D. oxycoccana adults also emerge over a long period of time with multiple overlapping generations each year (Sarzynski and Liburd 2003). At this point, it does not appear to be possible to develop a site -specific pest management program for D. oxycoccana in blueberries. These studies were carried out at a farm with a high, well established D. oxycoccana infestation. With a better understanding of D. oxycoccana flight behavior and dispersal combined with careful monitoring, site -specific pest management may be useful in preventing the establishment of D. oxycoccana in new blueberry plantings.

PAGE 58

58 Table 4 1. Weekly mean Dasineura oxycoccana infestation (larvae per flower bud) ( SEM) in different rows of plot B in 2007. Row 18 Jan. 25 Jan. 1 Feb. 8 Feb. 15 Feb. 22 Feb. 1 Mar. 8 Mar. North 0.39 (0.14) a 0.48 (0.16) a 0.50 (0.15) a 0.46 (0.10) a 0.36 (0.07) a 0.50 (0.03) a 0.84 (0.15) a 0.92 (0.20) a Center 0.74 (0.10) b 0.66 (0.14) ab 0.84 (0.15) a 0.63 (0.05) a 3.10 (0.54) b 2.58 (0.53) b 1.78 (0.13) b 1.18 (0.31) a South 1.67 (0.20) c 1.70 (0.36) b 1.52 (0.27) b 1.71 (0.29) b 3.64 (0.46) b 3.00 (0.30) b 1.85 (0.23) b 2.27 (0.22) b Means within a column followed by the same letter are not significantly different ( P > 0.05; LSD). Table 4 2. Mean Dasineura oxycoccana infestation (larvae per flower bud) ( SEM) in different rows of plot B by sample date in 2008. Row 9 Jan. 11 Jan. 13 Jan. 15 Jan. 17 Jan. 23 Jan. 10 Mar. North 0.0 (0.0) a 0.0 (0.0) a 0.03 (0.02) a 0.0 (0.0) a 0.01 (0.01) a 0.05 (0.05) a 1.57 (0.39) a Center 0.01 (0.01) a 0.0 (0.0) a 0.16 (0.13) a 0.07 (0.07) a 0.09 (0.05) a 0.16 (0.10) ab 5.20 (0.79) b South 0.75 (0.40) b 0.99 (0.40) b 0.85 (0.65) a 0.75 (0.36) b 1.41 (0.45) b 0.37 (0.06) b 2.93 (0.54) ab Means within a column followed by the same letter are not significantly different ( P > 0.05; LSD).

PAGE 59

59 Figure 4 1 Map of Plot A from a blueberry farm in Gainesville, FL. Shaded box indicates area containing rows from which samples were collected in 2006. 8 m 4 m Park Park fence fence Drainage Ditch Blueberry Field Uncultivated Field 65.5 m

PAGE 60

60 Figure 4 2 Map of Plot B from a blueberry farm in Gainesville, FL. Sample rows used in 2007 and 2008 experiments are highlighted in dark gray. Rectangle indicates approximate area from which samples were collected in the center row. Southern Border Row: 13 bushes, 26.2 m Center Row: 27 bushes, 41.8 m 6.7 m Drainage Ditch 29.3 m Northern Border Row: 42 bushes, 59.7 m

PAGE 61

61 Figure 4 3 Mean number of Dasineura oxycoc cana larvae per bud ( SEM) in Plot A over the 2006 sampling period. Figure 4 4 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot A at each sampling date in 2006.

PAGE 62

62 Figure 4 5 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot B over the 2007 sampling period. Bars with the same letter are not significantly different ( P < 0.05; LSMeans). Figure 4 6 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot B at each sampling date in 2007. a b c

PAGE 63

63 Figure 4 7 Mean number of Dasineura oxycoccana larvae per bud ( SEM) in Plot B over the 2008 sampling period. Bars with the same letter are not significantly different ( P < 0.05; LSMeans). a b c

PAGE 64

64 CHAPTER 5 BLUEBERRY GALL MIDGE DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE), PUPATION AND EMERGENCE UNDER VARYING TEM PERATURE CONDITIONS Environmental factors, such as temperature and humidity influence rate responses of development. Because insects are poikilot hermic, temperature has a great effect on their developmental rates (Pedigo 2002). The effect of temperature on development rate is a key component of insect population dynamics and can be used to develop forecasting models for predicting insect activity (Diaz et al. 2007). The timing of management activities can significantly improve pest management programs. Forecasting models can indicate the appropr iate time to start monitoring and implement control measures (Diaz et al. 2007). Temperature developme nt models have been calculated for many important insect pests, such as the Hessian fly ( Mayetiola destructor (Say)), fall armyworm ( Spodoptera frugiperda (Smith)), and codling moth ( Cydia pomonella (L.)) (Foster and Taylor 1975, Barfield et al. 1978, Rock and Shaffer 1983). Blueberry gall midge, Dasineura oxycoccana (Johnson), is an early -season pest of blueberries, Vaccinium spp. In north -central Florida, it emerges as early as January and can cause serious damage to developing blueberry flower buds (Der nisky et al. 2005). Rabbiteye blueberries, V. ashei Reade, are more susceptible to flower bud damage than southern highbush blueberries, V. corymbosum L. V. darrowi Camp (Dernisky et al. 2005). There are several overlapping generations each year (Sarzynski and Liburd 2003). Monitoring blueberry gall midge is challenging because, like many other gall midge species, adults are very short lived (2 3 days) and larvae are hidden within plant tissues (Gagn 1989). Sarzynski and Liburd (2003) developed a m ethod for sampling blueberry gall midge larvae in flower buds. Unfortunately, by the time larvae can be detected, damage is already being done. Furthermore, eggs and larvae

PAGE 65

65 within blueberry buds are sheltered from foliar insecticides (Lyrene and Payne 19 95). The most vulnerable stages of blueberry gall midge are mature larvae that leave blueberry buds to pupate in the soil and adults after emerging from the soil A temperature based development model could predict the onset of these events and tell us w hen to begin monitoring or implement control strategies. Furthermore, we could estimate the number of generations of blueberry gall midge possible in a given season. Several temperature-based phenology models, both linear and nonlinear, have been develope d for insects (Diaz et al. 2007). Insect development is nonlinear at extreme high and low temperatures and fits a sigmoid curve (Diaz et al. 2007). Campbell et al. (1974) argued that it is not practical to examine development rates at low temperatures ne ar the developmental threshold or high temperatures near the maximum because of high mortality at these extremes. The temperature range that insects are most likely encounter in nature lies on the linear portion of the developmental curve (Campbell et al. 1974). Temperature ranges and developmental thresholds vary between different species and populations of the same species (Campbell et al. 1974). The choice of model depends on the objectives of the investigator. The most common and widely used method i n pest management for estimating physiological time is the degree -day method. Degree days are defined as the accumulation of heat units above a developmental threshold over a 24-hour period (Pedigo 2002). The amount of heat required over time for an inse ct to complete a specific aspect of development is called a thermal constant (Campbell et al. 1974). Degree days can be computed using several methods, but the most common methods used are the simple mean method and the sine wave method (Pruess 1983). Pr oponents of degree -day models have argued that the simple mean method is less precise than the sine wave method. Nevertheless, it is easier to calculate degree days using

PAGE 66

66 the simple mean method, and this is adequate for most pest management programs (Pedi go 2002). The objectives of this experiment were to determine the developmental threshold and thermal constant for blueberry gall midge pupation, and apply this information to field trap data to estimate optimal times for control measures. No artificial diet has been developed for blueberry gall midge, which makes it difficult to rear in the laboratory. As a result, calculations for our temperature development model were made for only the pupal stage Materials and Methods Developmen t of Midges in the Lab oratory Larvae were collected from infested rabbiteye blueberry flower buds from an organic b lueberry farm in Gainesville, Florida Flower buds were taken to t he Small Fruit and Vegetable Integrated Pest Management Laboratory at the University of Florida Entomology and Nematology Department (G ainesville, FL) All flower buds collected were in development stages two or three (Spiers 1978). Buds were placed in 9 -cm plastic Petri dishes with moistened filter paper, sealed with Parafilm, and kept in a growt h chamber (Percival Model I 35 LL, Percival Mfg. Co., Boone, IA) at 30 2 C (day) and 20 2 C (night) with 14L:10D photoperiod. Petri dishes were checked twice a week for two weeks, and larvae that had exited buds were removed for use in the experimen ts. Onl y late third instar larvae (sternal spatula present and dark orange color) were used in the experiment. This was determined by examination unde r a dissecting microscope (Olympus SZ60, Olympus America, Center Valley, PA) at 10 magnification. Blueberry gall midge pupates in the soil; therefore, containers were prepared with substrates into which larvae could burrow and pupate. Two experiments were conducted, each using a different type of pupation container: cups or vials. Cups were 947 ml pl astic deli cups

PAGE 67

67 (Fabri -Kal Corp., Kalamazoo, MI) with a layer (4 cm deep) of potting soil (Jungle Growth Products, Statham, GA) and vermiculite (1:1 by volume) moistened with approximately 83 ml deionized water. Twenty larvae were placed in each cup. Vials were 5 -ml polystyrene roundbottom tubes (Falcon 2054, Becton Dickenson Labware, Lincoln Park, NJ) filled with vermiculite (2 cm deep) moistened with deionized water. One larva was placed in each vial. The pupal stage was determined to start when matu re larvae were introduced to the pupation containers and burrowed into the substrate, and the pupal stage ended when adults emerged from the substrate (Weston and Diaz 2005). The temperature chambers used in the study were Florida Reach In Chambers (Walker et al. 1993). Six constant temperatures were used: 10, 15, 20, 25, 30, and 35 ( 0.1 ) C, with a photoperiod of 16:8 (L:D). These experiments were conducted in 2008 and repeated in 2009. Data from both years were pooled for each container type to incre ase the number of replications and, consequently, the accuracy of the analysis. Pooling the data was justified because experimental conditions in both years were the same. For each temperature, six cups were prepared (120 larvae total) and 180 vials (one larva per vial). The number of adults emerging at each temperature and time to emergence, measured in days from the time containers were placed in temperatu re chambers, was recorded. Containers were checked daily up to 28 days. The 28day limit is thre e to four times the length of the pupal stage under laboratory conditions as determined by Bosio et al. (1998). Temperature Models Two models were used to estimate the lower temperature threshold, c and the thermal constant, K for the blueberry gall midg e pupal stage. The first method was the x intercept method, which is based on the linear regression equation: r(T) = a + bT (5 1 )

PAGE 68

68 where r(T) is the rate of development (1/days to adult emergence) T is the temperature ( C), a is the intercept a nd b is the slope of the line. The lower temperature threshold ( c ) was found by extrapolating the regression line to the point where it intercepts the x axis (Campbell et al. 1974). The second method was the Lactin model which is a modified version of the Logan model, a nonlinear regression model (Logan et al. 1976). The Lactin model equation is: r(T) = e e[ max ( T max -T (5 2 ) where T is the temperature and T max (the upper temperature threshold), parameters (Lactin et al. 1995). The modified Logan model can estimate upper and lower temperature thresholds, but it cannot estimate the thermal constant. Linear regressions of development rates on temperature were calculated using PROC REG (Littell et al. 2002). Lactin model curves were fit by iterative nonlinear regression (PROC NLIN SAS Institute 2003). The Marquardt algorithm was used in fitting the nonlinear model because parameter estimates were highly correlated (Lactin et al. 1995). Once the lower tempera ture thresholds were estimated, the thermal constants were calculated using the following equation: K = DTi (Ti c ) (5 3 ) where DTi is the duration of the pupal stage at temperature Ti in days and c is the lower temperature threshold (Van Kirk an d AliNiazee 2005). Trapping Midge Adults in the Field Blueberry gall midge adult emergence data in the field were collected using emergence traps placed at a blueberry farm in Gainesville, FL. All blueberry bushes at the farm were rabbiteye, Vaccinium ashei Reade. The trap design was based on descriptions from Smith and Chapman (1996). The emergence trap was constructed from a 3 L plastic food container which

PAGE 69

69 was painted white (Krylon Interior Exterior, 1502 Flat White, Krylon Products Group, Cleveland, OH). The opening at the top of the trap was covered by a 14 cm Petri dish lid, the underside of which was coated with Tangle Trap (The Tanglefoot Company, Grand Rapids, MI). Traps were placed beneath blueberry bushes approximately 0.3 m from the trunk. Traps were checked once a week throughout the period of blueberry flower bud development in 2007 and 2008. Eight emergence traps were used each year. Soil temperature at 10 cm below the surface was measured on site using a data logger with external soil temperature sensor (WatchDog Model 450, Spectrum Technologies, Plainfield, IL). In the case of missing data points, the data set was supplemented with readings from a University of Florida weather station at the Dept. of Agronomy Forage Research Unit th rough the Florida Automated Weather Netwo rk ( http://www.fawn.ifas.ufl.edu ) located 14.7 km from the blueberry farm. Daily maximum and minimum temperatures (recorded at hourly intervals) were used in degree day ca lculations. Degree days were calculated using the following formula: degree -days = [( T max + T min) / 2] c (5 4 ) where T max is the maximum daily temperature, T min is the minimum daily temperature and c is the lower developmental threshold (Arnold 1960). Results For both pupation containers, the greatest number of adults emerged at 20 and 25 C (Tables 5 1 and 5 2). No adults emerged at 10 C and only one adult emerged at 35 C in both years. These temperatures were, therefore, not included in the analysis. A higher percentage of adults emerged in the vials than in the cups. Minimum and mean days for development were similar ( 1 day) for both containers, but the range of days over which adults eme rged was longer in the vials.

PAGE 70

70 Linear Model : Estimates of the lower developmental threshold and thermal constant were calculated using linear regression for both the vials (Figure 5 1) and cups (Fig ure 5 2). D evelopment rates were linear betwe en 15 and 30 C for the vials [ y = 0.006 T 0.037; r2 = 0. 628 ( F = 371.0; df = 221; P < 0.0 0 1) ] and the cups [( y = 0.007 T 0.045; r2 = 0.749 ( F = 221.0; df = 75; P < 0.001)]. Excluding 30 C from the regression improved the fit of the line for the vial data (r2 = 0.753; F = 604.5; df = 199; P < 0.001) but not for the cup data ( r2 = 0.731; F = 192.8; df = 72; P < 0.001) From the vial data using all four temperatures c was estimated as 6.1 C and using three temperatures (15, 20, and 25 C) was estimated as 8. 9 C From the cup data using all four temperatures, c was estimates as 4.7 C, and using three t emperatures was estimated as 7.4 C. Values of K for each temperature are reported in Table 5 3. Nonlinear Model : Parameter estimates and statistics from the Lactin regressions are reported in Table 5 4. The lower temperature threshold was calculated by solving equation 5 2 for rate equal to zero. Using the vial data, c was estimated as 6.9 C. This threshold temperature was used in calculating estimates of K (Table 5 3). For the cup data, no reasonable temperature threshold could be calculated based on the estimated parameters. All estimates of model parameters based on cup data had high standard errors. The accuracy of c was determined by comparing the K values (equation 5 3) calculated for each temperature. The correct c should give the same K at each temperature (Howell and Neven 2000, Weston and Diaz 2005). K values calculated for c equal to 8.9 C had the least variation over the temperature range 1 5 25 C (Table 5 3). These were the temperatures used in calculating the threshold. Field Data : Degree day approximations were applied to first and peak adult midge emergence observed in 2007 and 2008 using emergence traps. Estimates based on cup data were

PAGE 71

71 not used due to concerns over accuracy. The K values for 15 C (134.2 and 178.2, from linear and nonlinear models, respectively) were used because this was close to the average soil temperature over the sampling period in both years. In 2007, the fi rst adults were collected on 24 January and emergence peaked on 21 March. Based on the two models, it was estimated that the first midges collected began pupation around 6 January. Both models predicted the same starting time for pupation. For peak emer gence, midges began pupation around 9 March (linear) or 7 March (nonlinear). In 2008, the first adults were collected on 16 January; adult emergence peaked on two dates, 5 March and 2 April. It was estimated that the first midges collected began pupation around 28 December (linear) or 27 December (nonlinear). For peak emergence, midges began pupation around 16 February (linear) or 14 February (nonlinear) for the first peak and 19 March (linear) or 17 March (nonlinear) for the second peak. Discussion Camp bell et al. (1974) cautioned that determining the threshold temperature by extrapolation of the regression line is very inaccurate. To compensate for this they recommended that, in constant temperature experiments, at least 50 individuals should be reared (Campbell et al. 1974). Dent (1997) recommended that for accurate estimates a minimum of 30 to 40 insects should survive to the end point of each temperature treatment. The number of adults emerging in the vials exceeded these levels, but this was not t he case for the cups. Only three adults emerged at 30 C in the cups. Replication would need to be increased to cope with mortality. This calls the accuracy of the estimates from cup data into question. Baxendale and Teetes (1983) and Baxendale et al. ( 1984) used the same six temperatures used in these experiments, plus 40 C, to model temperature development of sorghum midge, Contarinia sorghicola. The small number of D. oxycoccana adults emerging at 30 C and failure

PAGE 72

72 of any adults to emerge at 35 C j ustify the decision to omit 40 C. The maximum temperature for D. oxycoccana development in the pupal stage appears to be between 30 and 35 C. Determining development rates at extreme temperatures provides useful information about midge biology, but they should not be used in a linear temperature model. In this case, omitting 30 C from the analysis was justified. Removing these data improved the fit of the linear model for the 5 -ml vial pupation container data The rate of development is approximately linear above the threshold, but at high temperatures, the rates decline from linearity (Campbell et al. 1974). In nature D. oxycoccana does not encounter these high temperatures. Soil temperatures do not reach 30 C in the spring. In fact, soil tempera tures at the blueberry farm rarely exceeded 25 C during the sam pling period in 2007 and 2008. All estimates of c were below 10 C, but no adults were observed emerging at this temperature. At 15 C, adults emerged up to the 28 day limit. It is possible that this limit was too short and the experiment was terminated before pupation was complete at the lower temperatures. In future phenology model experiments, the time limit should be extended to determine if blueberry gall midge can pupate at temperatures below 15 C. Larvae that fail to emerge at low temperatures should be checked to determine if the low temperature is lethal. The thermal un it comparison indicated that 8.9 C was the best estimate of c The accuracy can only be determined over the ran ge of temperatures used in calculating the lower threshold. This threshold was close to the minimum thresholds for the gall midges Feltiella acarisuga (Vallot) (8.4 C) and Cystiphora schmidti (Rbsaamen) (10.0 C) (Moore 1987, Gillespie et al. 2000). Th e threshold for Hessian fly, Mayetiola destructor puparia development was considerably lower at 1.6 C (Foster and Taylor 1975).

PAGE 73

73 The linear model was the best fit for the data (Figures 5 1 and 5 2). Fitted parameters from the nonlinear model could not be accurately estimated for the cup data. Only three adults emerged in cups at temperatures above 25 C. Therefore, the rate does not remain constant at the upper temperature limit. More temperatures with a shorter interval are required at the upper and l ower limits to determine the shape of the curve and fit the model. The challenge of using a temperature model for forecasting pest events in Florida is setting a biofix, the time to begin adding degree days. In colder climates, winter temperatures are bel ow threshold for a prolonged period of time. Degree days can be calculations would begin at some point during this period. In Florida, however, daily temperatures are often above threshold throughout the winter. It was estimated that for blueberry gall midge adults to emerge in mid January 2008, pupation would have to begin no later than late December 2007. This does not necessarily mean that blueberry gall midge is developing all winter long, but that temperature is not the limiting factor. Other clim ate factors such as rainfall may also affect the timing of emergence (Fisher and Teetes 1982, Chen and Shelton 2007). Data from emergence traps were used to approximate the duration of the pupal stage. The time immediately preceding and following pupatio n are the periods of greatest vulnerability for blueberry gall midge. Unfortunately, this information cannot be used to forecast these events. This was the first experiment looking at the effect of temperature on the development rate of blueberry gall mid ge. The lack of effective laboratory rearing methods limited the stage that could be studied to the pupa. The thresholds calculated can only be applied to the pupal stage. Gillespie et al. (2000) were able to rear a cecidomyiid, Feltiella acarisuga fro m egg to adult, but this was a predatory midge that feeds on spider mites (Acari: Tetranychidae) so the food source

PAGE 74

74 was not an issue. Development rates for all life stages would be needed before temperature data can be used to forecast blueberry gall midg e infestations.

PAGE 75

75 Table 5 1. Time to emergence of Dasineura oxycoccana adults reared in 5 -ml vial pupation containers Days to adult emergence Development rate b Temperature No. emerged a % emerged Min Max Mean Mean 95% CI 10 0 0.0 -----15 60 33.3 14 28 22.0 0.047 0.045, 0.04 8 20 71 39.4 8 26 11.8 0.089 0.085, 0.093 25 69 38.3 6 15 8.3 0.12 5 0.120, 0.13 1 30 22 12.2 6 16 9.1 0.119 0.105, 0.13 2 35 1 0.6 8 8 8.0 0.1 25 -a For each temperature, total number of larvae evaluated = 180 b Development rate is the inverse of the number of days to adult emergence. Table 5 2. Time to emergence of Dasineura oxycoccana adults reared in 947 ml deli cup pupation containers Days to adult emergence Development rate b Temperature No. emerged a % emerged Min Max Mean Mean 95% CI 10 0 0.0 -----15 15 12.5 14 23 20.1 0.05 1 0.047 0.05 5 20 32 26.7 9 13 11.2 0.09 1 0.087, 0.094 25 26 21.7 7 14 8.6 0.122 0.112, 0.131 30 3 2.5 7 7 7.0 0.14 3 -35 0 0.0 -----a For each temperature, total number of larvae evaluated = 120 b Development rate is the inverse of the number of days to adult emergence. Table 5 3. Estimates of the thermal constant for Dasineura oxycoccana pupation under constant temperatures. Thermal constants are reported in degree days. Thermal Constant (K) a Model Method Threshold (C) 15 C 20 C 25 C 30 C Linear 5 ml vial 4 b 6.1 195.8 164.0 165.9 217.5 5 ml vial 3 c 8.9 134.2 131.0 133.6 192.0 Deli cup 4 b 4.7 207.0 171.4 17 4.6 17 7.1 Deli cup 3 c 7. 4 15 2.8 141.1 15 1 .4 158.2 Nonlinear 5 ml vial 6.9 178.2 154.6 150.2 210.2 a formula for calculating K = DTi( Ti c ) b four temperatures used (15 30 C) c th ree temperatures used (15 25 C)

PAGE 76

76 Table 5 4. Parameter estimates (mean SEM) and statistics from the nonlinear Lactin regression. Pupation Container df T max Lower Threshold (C) R 2 RSS a 5 ml vial 221 0.1298 0.03 34.9882 1.71 7.6624 1.71 0.0481 0.05 6.9 0.7194 0.0861 Deli cup 75 0.0514 0.09 50.4300 36.63 17.9644 25.76 0.2492 0.77 -0.7525 0.0168 a residual sum s of squares

PAGE 77

77 Figure 5 1 D evelopment rate s of D asineura oxycoccana pupae in 5 ml vials at four constant temperatures. Rates are expressed as the inverse of the duration of the pupal stage in days. Regression on four temperatures ( solid line): r2 = 0.628, root mean square error (RMSE) = 0.022. Regression on three tempe rature s (dashed line): r2 = 0.753, RMSE = 0.018. Figure 5 2 D evelopment rate s of D asineura oxycoccana pupae in deli cups at four constant temperatures. Rates are expressed as the inverse of the duration of the pupal stage in days. Regress ion on four t emperatures (solid line): r2 = 0.749, root mean square error (RMSE) = 0.016. Regression on thre e temperatur es (dashed line): r2 = 0.731, RMSE = 0.016.

PAGE 78

78 CHAPTER 6 EVALUATION OF REDUCE D -RISK INSECTICIDES FO R MANAGEMENT OF BLUEBERRY GALL MIDGE DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE) Blueberry gall midge, Dasineura oxycoccana (Johnson ), is a key pest of blueberries in the southeastern United States (Dernisky et al. 2005). Adults emerge in the spring as blueberry flower bu ds break dormancy, and females lay eggs between bud scales. Larvae feed on bud tissue s resulting in bud death and subsequent loss of yield. The flower buds of rabbiteye blueberry, Vaccinium ashei (Reade), are more susceptible to blueberry gall midge dama ge than southern highbush ( V. corymbosum L. V. darrowi Camp) flower buds (Dernisky et al. 2005). As much as 80% of flower bud loss can occur in susceptible varieties (Lyrene and Payne 1992). Management of blueberry gall midge has been limited to foliar applications of insecticides. The conventional products available for blueberry gall midge control in blueberry include malathion, diazinon, and S pinT or (Liburd 2004). Under the Food Quality Protection Act of 1996, malathion and diazinon, both organophos phates, can face greater restrictions (F ood Quality Protection Act 1996 ). New products with different modes of action will be needed for management of blueberry gall midge. Reduced risk pesticides are chemicals which have limited negative effects on huma n health or the environment (Pedigo 2002). Many conventional insecticides are broad -spectrum. Although broad -spectrum insecticides may be useful for controlling multiple insect pests, they negatively impact beneficial insects such as pollinators and natu ral enemies. Loss of natural enemies can lead to secondary pest outbreaks or resurgence of the primary insect pest (Pedigo 2002). Sampson et al. (2002, 2006) have identified parasitoids and predators of blueberry gall midge in Mississippi which appear t o play a role in regulating midge populations. Reducedrisk insecticides should be used in order to conserve these natural enemies.

PAGE 79

79 Insecticide efficacy trials were conducted in north -central Florida in 2006, 2007, and 2009. The objective of these trials was to identify effective reduced risk insecticides for the management of blueberry gall midge during the period of blueberry flower bud development. Identifying multiple insecticides with different modes of action will also be useful in resistance manag ement. The development of insecticide resistance in blueberry gall midge is a risk due to its short life cycle and multiple generations per year (Sampson et al. 2002). Direct contact is required for many conventional insecticides to be effective. Bluebe rry gall midge larvae feed within flower buds and are sheltered from foliar insecticides (Lyr ene and Payne 1995). Control may be possible using systemic insecticides (Bosio et al. 1998). Materials and Methods The insecticide evaluation trials were conduct ed at the North Florida Research and Education Center in Live Oak, Florida The treatment area consisted of a 0.16 ha planting of rabbiteye blueberry plants planted in rows 37 m long with plant spacing of 1.8 m within rows There were 14 rows of plants; each row planted with a single variety ( Austin Beckyblue, Brightwell Choice, Climax and Premier ). Plants were 2 m tall. Treatments were applied with a CO2-pressurized hand -held sprayer using Tee Jet D3 hollow cone tips with DC25 cores (Spraying Systems, Wheaton, IL) on a 4 -nozzle boom (1 m wide) delivering 47.3 ml per second at 32 psi Individual plants within each treatment were sprayed for three seconds on each side to adequately cover foliage. Control plants were sprayed with water only. All insecticides used and their modes of action are listed in Table 6 1. Treatment efficacy was based on number of blueberry gall midge larvae infesting developing flower buds. Flower buds in stages two or thre e (bud scales not abscised) (Spiers 1978) were collected and placed in re -sealable plastic bags and transport ed to the Small Fruit and

PAGE 80

80 Vegetable Integrated Pest Management Laboratory at the University of Florida Entomology and Nematology Department ( Gaines ville, FL ) In the laboratory, buds were transferred to 9 cm Petri dishes with a disc of moistened filter paper, sealed with Parafilm (Pechiney Plastic Packaging, Menasha, WI), and placed in a growth chamber (Percival Model I 35 LL, Percival Mfg. Co., Bo one, IA) at 30 2 C (day) and 20 2 C (night) with 14L:10D photoperiod for larval emergence (Sarzynski and Liburd 2003). Larvae emerging from the buds were counted twice a week, and buds were discarded after two weeks. Pre treatment samples were coll ected on the same day as the first insecticide application. 2006 Efficacy Trial The four treatments used in this trial were: 1) malathion (Malathion 5EC at 1.17 L / ha, Micro Flo, Memphis, TN); 2) thiamethoxam (Actara at 0.32 kg / ha, Syngenta C rop Protection, Greensboro, NC); 3) acetamiprid (Assail 30SG at 0.38 kg / ha, Cerexa gri -Nisso, King of Prussia, PA); and 4) control (water only). Treatment plot s were arranged in a randomized complete block design with four replications blocked by row. Each treatment was applied to four blueberry bushes in each replicate. Two insecticide applications were made, the first on 20 February and the second on 8 March. Following treatment application, weekly samples of 25 flower buds were collected per treatment in each replicate. 2007 Efficacy Trial Based on the findings in 2006, additional reducedrisk insecticides needed to be evaluated. SpinTor was included as a standard reduced risk insecticide, and malathion was included as a standard conventional insecticide. The five treatments used in this trial were: 1) malathion ( 2.9 L / ha); 2) spinosad (SpinTor 2SC at 0.44 L / ha, Dow AgroSciences, Indianapolis, IN); 3) spinetoram (Delegate at 0.28 kg / ha, Dow Agro Sciences, Indianapolis, IN); 4 ) cyantraniliprole (Rynaxypyr at 0.09 kg a i / ha, DuPont, Wilmington, DE); and 5) control (water

PAGE 81

81 only). Treatments were arranged in a randomized complete block design with four replications blocked by row. Each treatment wa s applied to three blueberry bushes in each replicate. Two insecticide applications were made, the first on 8 February and the second on 22 February. Following treatment application, weekly samples of 25 flower buds were collected per treatment in each r eplicate. 2009 Efficacy Trial The five treatments used in this trial were: 1) acetamiprid low rate ( Assail 30 SG at 0.28 kg / ha); 2) acetamiprid high rate ( 0.37 kg / ha); 3) cyantraniliprole low rate ( Cyazypyr at 0.10 kg a i / ha, DuPont, Wilmington, DE) ; 4) cyantraniliprole high rate ( Cyazypyr at 0.15 kg ai / ha); and 5) control (water only). Treatments were assigned to different rows of blueberry plants in a nested design with three replications. Replications were separated based on variety. Each tr eatment was applied to two blueberry bushes in each replicate. Two insecticide applications were made, the first on 6 March and the second on 16 March. Twenty flower buds per treatment in each replicate were collected three and ten days following each ap plication. Statistical Analysis Treatment differences for the entire trial were analyzed using repeated measures analysis of variance (ANOVA) (PROC MIXED, SAS Institute 2003). Differences among means for the 2006 and 2007 trials were separated using the l east -squares means (LSMeans) procedure ( P < 0.05). For the 2009 data, pairwise contrasts were made comparing the two products, the two rates within each product, and each treatment to the control. Data were square root transformed to satisfy model assumptions. Untransformed means are reported in the figures. Pre -treatment infestation was compared with the first and second post -treatment samples (before the second insecticide application). Pairwis e comparisons of post treatment dates to pre treatment dates

PAGE 82

82 were made using Dunnetts Test (PROC GLM SAS Institute 2003) (P < 0.05) (Littell et al. 2002). Results 2006 Efficacy Trial The mean number of gall midge larvae per bud was not significantly diff erent ( F = 1.09; df = 3, 24; P = 0.372) among insecticide treatments (Figure 6 1). None of the treatments were significantly lower than the control. There were no interaction effects between treatment and sample date ( F = 0.88; df = 6, 24; P = 0.521). V ery few gall midge larvae were detected during pre treatment sampling so comparisons between pre treatment and post -treatment infestations were not carried out. 2007 Efficacy Trial Significant differences among treatments were observed in the 2007 trial ( F = 5.05; df = 4, 48; P = 0.002). Flower buds from bushes treated with malathion had significantly fewer gall midge larvae than bushes treated with any of the reduced risk insecticides (Figure 6 2). This was also the only treatment that was significantly lower than the control. There were no interaction effects between treatment s and sample date s ( F = 1.03; df = 12, 48; P = 0.439). Alt hough post treatment means for m alathion and cyantraniliprole appear lower than pre treatment means, differences were not significant due to high standard error. 2009 Efficacy Trial Significant differences among treatments were observed in the 2009 trial ( F = 3.30; df = 4, 32; P = 0.023) (Figure 6 3) The number of gall midge larvae from acetamiprid treated bushes was not significantly different from cyantraniliprole treated bushes ( F = 1.91; df = 1, 32; P = 0.177). Flower buds from bushes treated with acetamiprid at the low rate had significantly fewer gall midge larvae than bushes treated with acetamiprid at the high rate ( F = 8.62; df = 1, 32; P =

PAGE 83

83 0.006) The number of gall midge larvae from bushes treated with the high or low rate of cyantraniliprole was not significant difference ( F = 2.46; df = 1, 32; P = 0.126). Acetamiprid at the low rate was the o nly treatment significantly different from the control ( F = 5.31; df = 1, 32; P = 0.028) There were no interaction effects between treatment s and sample date s ( F = 0.50; df = 12, 32; P = 0.900). Pre and post treatment means were not significantly diffe rent for any of the insecticides evaluated. Discussion With the exception of acetamiprid (Assail) at the low rate, these trials failed to identify any reduced risk insecticides that effectively reduced blueberry gall midge infestation in flower buds. Any differences among treatments were obscured by high variance. Means were influenced to some degree by outliers; however, using a trimmed mean or removing outliers altogether did not affect the outcomes of the analyses. In the 2009 trial, it was expected t hat bushes treated with high rates of acetamiprid would have lower infestations than those treated with low rates. Bushes treated with acetamiprid at the low rate, however, had significantly fewer larvae per bud than bushes treated at the high rate. It i s doubtful that acetamiprid actually performed better at the low rate than the higher rate. When comparing pre and post treatment means, there were no significant differences for acetamiprid at the low rate. Also there was no significant difference betw een the high and low rate of cyantraniliprole Timing of application is very important. Chen et al. (2007) used acetamiprid (Assail to control Contarinia nasturtii Keiffer on cauliflower transplants. They observed that older larvae were more difficult t o control and recommended spraying close to oviposition in order to target eggs and young larvae (Chen et al. 2007). Blueberry gall midge has many overlapping

PAGE 84

84 generations per year (Sarzynski and Liburd 2003). After the population increases to levels warr anting control, larvae of various sizes are present in buds through the remainder of the flowering period. Many larvae removed from flower buds were in the third instar. This could explain some of the difficulty in controlling blueberry gall midge in the se trials. The label for spinetoram ( Delegate ), for example, recommends targeting eggs at h atch or small larvae (Dow AgroS ciences 2007). Spinetoram is labeled for blueberry gall midge suppression on blueberry. Effective reduced risk insecticides for use in blueberry gall midge management need to be identified. Ideal products would have low toxicity to natural enemies and systemic activity so that larvae feeding within buds will not escape exposure. Timing of a pplication is critical. Future trials should include monitoring of midge adults to estimate peak oviposition periods and optimum timing for insecticide application.

PAGE 85

85 Table 6 1. Insecticides used in efficacy trials for Dasineura oxycoccana control in rabbi teye flower buds. Trade Name Active Ingredient Class Mode of Action Labeled for Blueberry Malathion 5EC malathion organophosphate cholinesterase inhibitor Yes Actara thiamethoxam neonicotinoid ACh a receptor agonist Yes Assail 30SG acetamiprid neonicotinoid ACh a receptor agonist Yes SpinTor 2SC spinosad spinosyn disrupts ACh a binding Yes Delegate spinetoram spinosyn disrupts ACh a binding Yes Rynaxypyr cyantraniliprole anthranilic diamide disrupts ryanodine receptor No Cyazypyr cyantraniliprole anthranilic diamide disrupts ryanodine receptor No a acetylcholine

PAGE 86

86 Figure 6 1. Mean ( SEM) D asineura oxycoccana larvae per flower bud from 2006 insecticide evaluation (North Florida Research and Education Center, Live Oak, FL). F = 1.09; df = 3, 24; P = 0.3 72. Figure 6 2. Mean ( SEM) D asineura oxycoccana larvae per flower bud from 2007 insecticide evaluation (North Florida Research and Education Center, Live Oak, FL). Cyan. = cyantraniliprole. F = 5.05; df = 4, 48; P = 0.002. Significant differences with control are marked by [***] when compared using LS Means. ***

PAGE 87

87 Figure 6 3 Mean ( SEM) D asineura oxycoccana larvae per flower bud from 2009 insecticide evaluation (North Florida Research and Education Center, Live Oak, FL). Acet. = acetamiprid; Cyan. = cyantraniliprole. F = 3.30; df = 4, 32; P = 0.023. Significant differences with control are marked by [***] when compared using pairwise contrasts. ***

PAGE 88

88 CHAPTER 7 PARASITOIDS OF BLUEB ERRY GALL MIDGE DASINEURA OXYCOCCANA (JOHNSON) (DIPTERA: CECIDOMYIIDAE) IN NORTH CENTRAL FLORIDA Blueberry gall midge, Dasineura oxycoccana (Johnson), is an important pest of rabbiteye blueberries, Vaccinium ashei Reade, in the southeastern United States (Dernisky et al. 2005). Dasineura oxycoccana is native to eastern North America and feeds exclusively on Vaccinium buds (Sampson et al. 2006). For many years, it has been known to feed on blueberry and cranberry leaf buds (Barnes 1948, Marucci 1977, Lyrene and Payne 1992). In cranberries ( V. macrocarpon L. ), D. oxycoccana feeds on young leaves causing them to become distorted and cupped (Co ckfield and Mahr 1994). Continuous feeding kills the apical meristem in cranberries and the same type of injury occurs in blueberries (Barnes 1948). There is evidence that Vaccinium darrowi Camp meristems are resistant to D. oxycoccana (Lyrene and Payne 1995). In 1992, it was discovered that D. oxycoccana also causes serious injury to blueberry flower buds (Lyrene and Payne 1992). Before this, loss of flower buds had been erroneously attributed to lack of chilling or freeze damage (Lyrene and Payne 1995) Females lay eggs in flower buds early in the season as bud scales begin to separate ( Lyrene and Payne 1995, Dernisky et al. 2005). When the eggs hatch, larvae crawl deeper into the buds and begin feeding on the innermost meristomatic tissue resulting i n necrosis and bud abortion (Sampson et al. 2002). In susceptible rabbiteye blueberry cultivars, D. oxycoccana can destroy 80 90% of the flower buds ( Lyrene and Payne 1992). Similarly, in cranberries, D. oxycoccana causes indirect reduction of flower buds by damaging growing shoots before flower buds are produced. Vigorous cranberry plants can compensate for damage and form flower buds, but if a plant is weakened or damage occurs late enough in the season, vegetative rather than flower buds will be prod uced (Cockfield and Mahr 1994).

PAGE 89

89 Little research has been done on the natural enemies of D. oxycoccana The hover flies Toxomerus marginatus (Say) and T. geminatus (Say) are predators of D. oxycoccana larvae in Wisconsin and Mississippi, respectively (Mahr and Kachadoorian 1990, Sampson et al. 2002). Other natural enemies of D. oxycoccana include several parasitoids (Hymenoptera) from various subfamilies. Some of the parasitoids observed in cranberry bogs include one eulophid (Tetrastichus sp. [= Aprostocetus (Perkins)]) and two ceraphronids (Aphanogmus sp. and Ceraphron pallidiventris Ashmead) (Barnes 1948). Sampson et al. (2006) found parasitoids of D. oxycoccana larvae in the following four genera: Aprostocetus (Eulophidae: Tetrastichinae), Syn opeas Inostemma and Platygaster (Platygastridae: Platygastrinae). These parasitoids were collected in Mississippi, Louisiana and Alabama. The species of Aprostocetus collected was newly discovered (Sampson et al. 2006). Parasitic Platygastridae replac ed Ceraphronidae as key parasitoids of gall midges in the southern United States (Sampson et al. 2006). The hosts of Aprostocetus are found in several orders of insects (Burks 1967). Species from this genus are key parasitoids of several gall midges such as the balsam gall midge (Dasineura balsamicola (Lint.) sorghum midge (Contarinia sorghicola (Coquillett)) African rice gall midge (Orseolia oryzivora Harris and Gagn), and raspberry cane midge (Resseliella theobaldi (Barnes)) (MacGown and Osgood 1972, Wiseman et al. 1978, Williams et al. 1999, Vtek et al. 2006). The Platygastrinae (including Synopeas Inostemma and Platygaster ) parasitize Cecidomyiidae exclusively (Austin et al. 2005). The apple leaf curling midge, Dasineura mali Kieffer, and pear le af curling midge, D. pyri (Bouch), are parasitized by Platygaster demades Walker (Sandanayaka and Charles 2006). Five species of Platygaster were found associated with the balsam gall midge, D. balsamicola (Lintner) in Maine (MacGown and Osgood 1972).

PAGE 90

90 No surveys of D. oxycoccana natural enemies have been conducted in Florida. The objectives of this research were to collect and identify endoparasitoids of D. oxycoccana in Florida blueberries, and determine patterns of parasitoid activity. Materials and Me thods Dasineura oxycoccana larva e were collected randomly from blueberry farm s in Gainesville, Florida (Alachua County) and the University of Florida North Florida Research and Education Center (NFREC) in Live Oak, FL (Suwannee Co unty ). The farm in Gainesville consisted of 4.5 ha of rabbiteye blueberries. The farm had been managed organically, and no insecticides had been used for many years. The blueberry planting at NFREC consisted of 0.16 ha of rabbiteye blueberries. This planting was managed u sing conventional pesticides. Identification of Parasitoids During the flowering period, blueberry flower buds were sampled weekly. Buds were placed in re -sealable plastic bags and taken to t he Small Fruit and Vegetable Integrated Pest Management Laborato ry at the University of Florida Entomology and Nematology Department (Gainesville, FL) for processing. Samples of 20 or 25 buds were placed in 9 -cm -diam eter Petri dishes with moistened filter paper, sealed with Parafilm (Pechiney Plastic Packaging, Menas ha, WI), and placed in a growth chamber at 30 2 C daytime temperature, 20 2 C nighttime temperature, and 14:10 L:D photoperiod (Sarzynski and Liburd 2003). Flower bud samples were kept for two weeks before being discarded. Midge larvae that had exi ted the buds were examined under a microscope to determine if they were parasitized. Parasitized larvae appeared rigid and capsule -shaped rather than spindleform (personal observation). As the parasitoid develops, the internal tissues of the host larva a re digested, leaving the transparent cuticle. At that point, the parasitoid pupa or adult could be clearly seen. Parasitized larvae were removed from the Petri dishes and placed in 59 ml plastic cups until adult parasitoids emerged. Adult

PAGE 91

91 parasitoids we re stored in 70% ethanol or slide -mounted in CMC 10 (Masters Company, Wood Dale, IL). Adult parasitoids were also collected in situ on vegetative buds using an aspirator. Parasitoids were identified to family using Triplehorn and Johnson (2005) or to gen us using Schauff et al. (1997) for Eulophidae and Rajmohana (2006) for Platygastroidea. Patterns of Parasit oid Activity To determine parasitism rates, midge larvae were collected from blueberry leaf buds at the farm in Gainesville. Samples (15 to 20 buds) were taken on 25 February, 4 March and 11 March to determine the onset of midge infestation. T he first day on which parasitism was observed was 11 March, so sampling commenced the following week. Fifty to 60 leaf buds were collected each week from 18 Ma rch to 8 April. Only leaf buds that appeared damaged were sampled. Leaf buds were placed in 59 -ml plastic cups and taken to the Small Fruit and Vegetable IPM Laboratory at the University of Florida Entomology and Nematology Department (Gainesville, FL) for dissection. Bud length was measured prior to dissection using calipers. Each leaf layer was cut away and the number of midge larvae was counted. Live midge larvae were placed in a drop of water on a microscope slide, compressed under a glass cover sl ip, and examined under a compound microscope ( Olympus CX41, Olympus America, Center Valley, PA). Host instars and the number of parasitoid eggs or larvae were recorded. Percent parasitism was calculated using the formula: % parasitism = P / ( M + P ) 100, where P is the number of parasitized D. oxycoccana larvae in a sample and M is the number of unparasitized larvae (Sampson et al. 2002). Statistical Analysis The differences in mean infestation of host larva by sample date and differences in percent parasitism by sample date were each analyzed using a one -way analysi s of variance (ANOVA) (PROC GLM, SAS Institute 2003). The effects of host instar and host location (leaf layer) on

PAGE 92

92 the number of parasitoid eggs were analyzed using a two-way ANOVA (PROC GLM). The number of host larvae and parasitoid eggs were (x+1)2 transformed to satisfy model assumptions. Percentage data were log10(x+1) transformed to satisfy model assumptions. Differences among means were determined using the Least Significant Diff erence (LSD) mean separation test ( P < 0.05). The chi -square goodness of -fit analysis (PROC FREQ SAS Institute 2003) was used to test if host location and incidence of parasitism were distributed uniformly among leaf layers and if incidence of parasitism was distributed uniformly among host instars. The effect of leaf bud length on the total number of hosts and number of parasitized hosts was analyzed using PROC CORR (SAS Institute 2003) Untransformed means are reported in figures. Results Identificati on of Parasitoids Most parasitoids reared from D asineura oxycoccana larvae were collected at the farm in Gainesville, FL. These parasitoids were identified as Platygaster and Aprostocetus Adults from both genera were reared from larvae collected in late October 2008. In January 2009, parasitoids in the genera Platygaster Synopeas and Telenomus were reared from host larvae. Two midge larvae were found parasitized by Ceraphronids (adul ts not identified to genus) in February 2009. During the first two weeks of March 2009, as D. oxycoccana began to infest leaf buds, most wasps reared from host larvae were Platygaster sp. Aprostocetus sp. adults were collected in situ beginning on 11 Mar ch, and were reared from host larvae on 18 March. The only parasitoids reared from host larvae collected at NFREC in Live Oak, FL were Synopeas sp. Patterns of Parasit oid Activity Dasineura oxycoccana larvae dissected from leaf buds were almost exclusively parasitized by Aprostocetus sp. The oblong ovoid eggs were clearly visible within the host midgut. Three

PAGE 93

93 D. oxycoccan a larvae were found parasitized by Platygaster sp. Two were collected on 18 Mar ch, and both midge larvae also contained Aprostocetus eggs. The other parasitized larva was collected on 25 March. The analysis was conducted using only the Aprostocetus data. The number of D. oxycoccana larvae per bud was significantly higher on 18 Marc h and 8 April than on 25 March and 1 April ( F = 8.71; df = 3, 210; P < 0.001) (Fig ure 7 1). Fifty two percent of larvae collected were second instar, and the remainder was evenly divided between first and third instar. Mean percent parasitism was signifi cantly different for most of the sample dates ( F = 9.05; df = 3, 122; P < 0.001) (Fig ure 7 2). It was highest on 1 April and lowest on 8 April. Host instar had a significant effect on parasitism ( 2 = 111.03; df= 2; P < 0.001). Thirdinstar larvae had 6 2.9 % parasitism, second -instar larvae had 44.0%, and only 7.6 % of first instar larvae were parasitized. Host location by leaf layer also had a significant effect on parasitism ( 2 = 25.30; df= 2; P < 0.001). I n the outermost leaf layer, 52.0% of host lar vae were parasiti zed compared with 39.8% and 22.6% in the middle and innermost layers, respectively. Most host larvae were parasitized only once (first instar: 75%, second instar: 61%, third instar: 51%). Second and third instar larvae were found with up to five and six parasitoid eggs, respectively. The number of parasitoid eggs was significantly different for each leaf layer ( F = 4.63; df = 2, 514; P = 0.010) and host instar ( F = 45.3; df = 2, 514; P < 0.001). There was no interaction between leaf laye r and host instar ( F = 0.76; df = 4; 514; P = 0.553). There was a weak correlation between the number of host larva e and leaf bud length (r = 0.183; P = 0.014). Leaf bud length and incidence of parasitis m were not correlated (r = 0.09 0; P = 0.286).

PAGE 94

94 Disc ussion This was the first survey of Dasineura oxycoccana parasitoids conducted in Florida. It was confirmed that species of Platygaster and Aprostocetus actively parasitize D. oxycoccana larvae infesting rabbiteye flower and leaf buds. The wasps collecte d in Florida should be compared with those from Mississippi to determine if the species complex is the same. Sampson et al. (2006) developed a mitochondrial DNA analysis for adult parasitoids which could make comparisons between regions faster and more re liable. Platygastrid parasitoids ( Platygaster and Synopeas ) appear to be most common early in the season (January), and are capable of parasitizing larvae in blueberry flower buds. Eulophids appeared later in March as flower buds became scarce and D. oxyc occana began infesting leaf buds. Sampson et al. (2006) observed the opposite trend in Mississippi. Aprostocetus was the first parasitoid attacking D. oxycoccana during bloom, while platygastrids became active at the end of bloom (Sampson et al. 2006). In Florida, Aprostocetus appears to completely supersede Platygaster as the primary parasitoid of D. oxycoccana by the middle of March. Differences in climate between Florida and Mississippi may explain these differences in parasitoid activity. A longer survey will be required to more accurately determine the phenology of these parasitoids. The differences in D. oxycoccana infestation from week to week indicate the presence of distinct generations. On 25 March and 1 April over 40% of infesting larvae we re in the third instar. These larvae would have been ready to leave the buds to p upate in the soil. Less than 25% of larvae from buds collected on 18 March and 8 April were in the third instar. In addition, leaf buds collected on 8 April conta ined a hig h proportion of first instar larvae as well as many eggs. This also explains the lower incidence of parasitism on 8 April despite higher midge infestation.

PAGE 95

95 Levels of parasitism were similar in Florida and Mississippi (30 40%) (Sampson et al. 2006). Per cent parasitism increased with each successive host instar. One hypothesis fo r this difference is that first instar larvae may be difficult to locate within the bud and, therefore, escape being parasitized. Aprostocetus females walk along the length of leaf buds probing with their ovipositor until they find a midge larva (personal observation). The fact that parasitism also varied with leaf bud layer l ends support to this host appare ncy hypothesis. An alternative hypot hesis is that female parasitoids avoid first i nstar larvae. Second and third instar larvae may be preferred hosts because they are larger and provide more nutrients and space for wasp development. For this hypothesis to be valid, female parasitoids must be able to assess host quality through ovipositor probing. Alternatively, sensillae on the ovipositor may respond to stimuli that initiate egg -laying rather than assessing host quality. In this case, second and third instar midge larvae ma y induce egg la ying while first instar larvae do not. Sampson et al. (2006) also observed that most female Aprostocetus oviposited in third instar midges. Further work needs to be done on Aprostocetus host selection behavior. Aprostocetus develops singly within each ho st larva (Sampson et al. 2002). It is not surprising; therefore, that most parasitized midge larvae contained only one parasitoid egg. In the cases of superparasitism, it could not be determined if the eggs came from one or multiple females. This survey confirmed the presence of D. oxycoccana parasitoids in north -central Florida. Further studies need to be conducted to determine their impact on D. oxycoccana populations over the blueberry growing season. High parasitism rates suggest that Platygastrid a nd Eulophid parasitoids could play an important role in D. oxycoccana population regulation in blueberries. Parasitism experiments should be done with Platygaster to compliment the experiments with

PAGE 96

96 Aprostocetus In cases where insecticides must be used, applications should be carefully timed and reduced risk insecticides used in order to avoid disrupting natural enemies (Sampson et al. 2002).

PAGE 97

97 Figure 7 1 Mean number of D asineura oxycoccana larvae per leaf bud collected in 2009 at a blueberry farm in Ga inesville, FL. Bars with the same letter are not significantly different ( P > 0.05). Figure 7 2 Mean percent parasitism of D asineura oxycoccana larvae by Aprostocetus sp. Midge larvae were collected from leaf buds in 2009 at a blueberry farm in Gainesville, FL. Bars with the same letter are not significantly different ( P > 0.05).

PAGE 98

98 CHAPTER 8 GALL MIDGE SPECIES C OMPLEX IN FLORIDA RA BBITEYE BLUEBERRIES Blueberry gall midge, Dasineura oxycoccana (Johnson), and blueberry tipworm, Prodiplosis vaccinii (Felt), are two species of gall midges (Diptera: Cecidomyiidae) that feed on Vaccinium buds (Sampson et al. 2006). Barnes (1948) mentioned both b lueberry tipworm and blueberry gall midge, but reported that P. vaccinii was the only gall midge of economic importance on blueberries in North America, while D. oxycoccana was the only economically important on cranberry. Leaf damage caused by cranberry tip worm on cranberry was described as identical to blueberry tip worm damage on blueberry (Barnes 1948). Prodiplosis vaccinii feeds on immature leaves causing the leaves to cup and the tips to become blackened and brittle (Driggers 1926). According to a ll published reports, P. vaccinii only feeds in developing leaf buds, not flower buds. D asineura oxycoccana feeds in leaf buds and flower buds (Sarzynski and Liburd 2003). Dasineura oxycoccana emerges earlier in the season than P. vaccinii As floral bud s become scarce, D. oxycoccana moves to leaf buds. In Mississippi, D. oxycoccana is superseded by P. vaccinii following vegetative bud break (Sampson et al. 2006). Understanding the pest complex on blueberries is important for making management decisions. If D. oxycoccana is the only gall midge on blueberries in Florida, controlling adults after pollination might reduce midge populations in flower buds the following year (Lyrene and Payne 1992). If two species are present, however, and P. vaccinii replaces D. oxycoccana as in Mississippi, then managing late season midge populations would have no effect on midge populations infesting the following years flower buds.

PAGE 99

99 Materials and Methods Gall midge larvae and adults were collected in floral and vegetative bud samples at multiple sites throughout the year in north -central Florida. These sites included a blueberry farm in Gainesville, the North Florida Research and E ducation Center in Live Oak, and the Plant Science Research and Education Unit in Citra. Samples were taken to t he Small Fruit and Vegetable Integrated Pest Management Laboratory at the University of Florida Entomology and Nematology Department (Gainesvil le, FL ) Flower buds were held in 9 -cm diameter Petri dishes with moistened filter paper until larvae emerged, and leaf buds were dissected to remove larvae. Larvae were reared to adults in the laboratory. In addition to larval samples, adult midges we re captured using emergence traps. These traps were constructed from 3 L plastic food containers with an opening in the top of the trap. Each trap was covered with an inverted 14 -cm diameter Petri dish the underside of which was coated with an adhesive, T angle Trap (The Tanglefoot Company, Grand Rapids, MI). Traps were placed beneath blueberry bushes to capture adult midges emerging from pupation sites in the soil. All samples collected from floral and vegetative buds as well as traps were preserved in 70% ethanol for subsequent identification. Species Identification : Dasineura oxycoccana and P rodiplosis vaccinii can be differentiated morphologically without many difficulties because they belong to two different tribes within the subfamily Cecidomyiinae (Oligotrophini and Cecidomyiini, respectively) (Gagn 1989). Wing venation was used because wing veins were easy to see and the differences were less subjective than other morphological characteristics. In Dasineura the R5 vein is shorter than the wing (joins costa before wing apex) (Figure 8 1 A) and in Prodiplosis the R5 vein is longer, joining the costa beyond the wing apex (Figure 8 1 B) (Gagn 1994). Wings were

PAGE 100

100 examined under a dissecting microscope (Olympus SZ60, Olympus America, Center Valley, PA) at 60 magnification. In 2009, midge larvae were collected from blueberry leaf buds and placed in 0.02 ml of water on a microscope slide and examined at 400 magnification. Third-instar Prodiplosis and Dasineura larvae can be differentiated based on t he length of the terminal papillae (Gagn 1994). Papillae are present in second instar larvae, but they are much smaller and difficult to see. Leaf buds were collected once a week at the blueberry farm in Gainesville from 18 March to 8 April. Results and Discussion From 2006 to 2009 a total of 598 midge larvae were collected and reared to adults. From these, 576 were identified as D asineura oxycoccana and 22 as P rodiplosis vaccinii Prodiplosis vaccinii was collected at the farm in Gainesville in March 2006 and at the Plant Science Research and Education Unit in May 2007. The emergence traps captured 684 adult midges; all were identified as D. oxycoccana In 2007, traps were used until 11 April, and in 2008, traps were used until 30 April. D. oxycoccan a adults were captured at these later dates. Dasineura oxycoccana larvae were collected from in Gainesville as late as mid July. All midges collected at the North Florida Research and Education Center in Live Oak were D. oxycoccana In 2009, 124 thirdinstar larvae were collected from leaf buds. All of these larvae were identified as D. oxycoccana Prodiplosis vaccinii was found in Florida blueberries, but in this survey it constituted only 3.7% of the total number of midges collected. Based on emerge nce trap data, D. oxycoccana was active through March and April, the period in which bloom in completed and blueberry leaf buds begin to open. From the larvae collected in 2009, it was shown that D. oxycoccana is the

PAGE 101

101 dominant midge infesting the first flu shes of vegetative growth following bloom. This survey did not confirm that P. vaccinii supplants D. oxycoccana in leaf buds as observed in Mississippi. Additional collecting will be required to determine the phenology and abundance of P. vaccinii in Florida blueberries and its interaction with D. oxycoccana Lyrene and Payne (1992) observed heavy midge infestations in vegetative meristems after flowering was complete. This survey shows that the midges were most likely D. oxycoccana Directing insectici de sprays at D. oxycoccana following bloom could reduce midge populations in the following years flower buds.

PAGE 102

102 Figure 8 1. Wing margins of gall midges collected at blueberry farms in North Central Florida. A) Dasineura oxycoccana wing. B) Prodiplosis vaccinii wing The solid line indicates the wing margin and the dashed line the position of the R5 vein A B

PAGE 103

103 CHAPTER 9 GENERAL CONCLUSIONS Relatively little research has been conducted on blueberry gall midge biology, ecology, and management. Although blueberry gall midge has historically been overlooked as a pest in blueberries, it has the potential to cause significant losses. Other early -season pests of blueberries in Florida are flower thrips, Frankliniella spp., and the blueberry bud mite, Acalitus vaccinii (Keifer) (Finn 2003). This research focused on monitoring and methods of control, essential components for developing an integrate d pest management program. Finn (2003) evaluated methods for monitoring blueberry gall midge in blueberries. Effective methods such as flower bud dissection and larval emergence were developed for immature stages (Finn 2003). Sampson et al. (2002, 2006) investigated methods of controlling blueberry gall midge, specifically biological control. The current research expanded on these previous studies. Experiments on improved monitoring included designing a trap for monitoring adult midges and calculating a temperature -dependent development rate model. Experiments on midge management included insecticide efficacy trials and a survey of natural enemies (hymenopteran parasitoids). The results of these experiments brought us closer to developing a comprehens ive integrated pest management program for insect pests in early -season blueberries. Emergence Trap for Adult Midge Monitoring The larva emergence technique and bud dissection are effective for estimating population density of immature gall midges (Finn 2003, Sarzynski and Liburd 2003). Larvae, however, are detected only after damage has begun, and larvae in buds are largely sheltered from pesticides. Monitoring adults would provide earlier detection of blueberry gall midge activity. I evaluated differen t types of emergence traps for capturing adult midges emerging from pupation sites in the soil beneath blueberry bushes. Traps were constructed based on designs of emergence traps used

PAGE 104

104 for other small dipterans. The most effective traps were the Petri di sh and bucket traps. These traps both had transparent trapping surfaces coated with adhesive (Tangle Trap). Traps indicated increases in midge population density one week before increases in the number of larvae were observed in flower buds. The predic tive ability of traps appears to be limited to peaks in midge numbers typically occurring in late February to early March. Emergence traps could be useful for detecting the first midges emerging during flowering, but more research will be required before traps can be used to make accurate predictions of midge infestation. Development Rate Model for Midge Pupae Determining insect development rates has applications in directing the timing of monitoring and management activities. This experiment was the firs t time a degree -day model was applied to blueberry gall midge. Because there is no method for rearing blueberry gall midge in the laboratory, the entire lifecycle could not be modeled. The pupal stage was used because it was easy to determine the end poi nt of development (adult eclosion). Mature larvae were reared to adults at six constant temperatures. The minimum temperature for pupation was estimated using a linear and nonlinear model. The linear model fit the data better, and estimated a mini mum th reshold temperature of 8.9 C and thermal constant of 134 degree days at 15 C. In 2007 and 2008, the first adult midges were trapped in midto late January. Using these values, it was estimated that these adults began pupation at the end of December or beginning of January. Mean daily temperatures in north -central Florida were high enough to allow blueberry gall midge development through December and January, the period immediately prior to blueberry flower bud break.

PAGE 105

105 Managing Blueberry Gall Midge wi th Reduced-Risk Insecticides Few studies have been done investigating methods for blueberry gall midge control. Insecticides have been listed as the conventional control by Sampson et al. (2002) and Liburd (2004). Larvae feeding in buds, however, are she ltered from many foliar applied insecticides. Systemic insecticides may be more effective at controlling midge larvae (Bosio et al. 1998). Many of the new reduced risk insecticides labeled for use in blueberries have systemic or translaminar activity. I conducted three efficacy trials comparing reduced -risk to conventional insecticides and untreated controls. The reduced risk insecticides used were acetamiprid, thiamethoxam, spinosad, spinatoram, and cyantraniliprole. Treatment efficacy was based on th e number of midge larvae collected from flower buds following applications. The results from all three trials were inconclusive. Two treatments, malathion and acetamiprid resulted in reduced larva densities, however for acetamiprid, the number of larvae per bud was not significantly different between pre and post -treatment samples. Finn (2003) also did not observe any differences in larval infestation between treated bushes and the untreated control. Many larvae collected from flower buds were in the s econd or third instar. According to label recommendations, the reduced risk insecticides included in these trials are most effective against eggs or newly hatched larvae. Control failure, therefore, was likely the result of improper timing of application. Identifying Natural Enemies of Blueberry Gall Midge in Florida Little research has been done focusing on the natural enemies of blueberry gall midge. Sampson et al. (2006) conducted a survey of parasitoids naturally occurring in the southeastern United States. Key parasitoids were found in the families Eulophidae and Platygastridae (Hymenoptera) (Sampson et al. 2006). No survey, however, had been conducted in Florida. I

PAGE 106

106 reared adult wasps from parasitized gall midge larvae and collected adult wasps in situ The wasps were identified to genus, the two most common being Platygaster and Aprostocetus It appeared that Platygaster was more prevalent earlier in the season with Aprostocetus becoming common after bloom. I collected gall midge larvae from b lueberry leaf buds and counted the number parasitized, number of parasitoid eggs, location of host, and host instar. Almost all parasitism was attributed to an unidentified species of Aprostocetus Three larvae had been parasitized by a species of Platyg aster Third instar larvae had the highest incidence o f parasitism (62.9 %), and first -i nstar larvae had the lowest (7.6 %). Also, percent parasitism was lowest at the innermost layer of the leaf buds. The incidence of parasitism depended on host location and size, suggesting that parasitoid success depends on the females ability to locate hosts. Female Aprostocetus probe leaf buds with their ovipositors to locate midge larvae. Hymenopteran parasitoids may play a key role in regulating blueberry gall mi dge populations in blueberries T herefore w hen managing this pest, efforts should be made to conserve its natural enemies through use of reducedrisk insecticides and careful timing of applications.

PAGE 107

107 LIST OF REFERENCES Akar, H., and E.A. Osgood. 1987. Emergence trap and collecting apparatus for capture of insects emerg ing from the soil. Entomol. News. 98: 35 3 9. Arevalo -Rodriguez, H.A. 2006. A study of the behavior, ecology, and control of flower thrips in blueberries towards the development of a n integrated pest management (IPM) program in Florida and Southern Georgia. Ph.D. dissertation, University of Florida, Gainesville, FL. Arnold, C.Y. 1960. Maximum -minimum temperatures as a basis for computing heat units. Proc. Am. Soc. Hort ic Sci. 76: 682692. Austin, A.D., N.F. Johnson, and M. Dowton. 2005. Systematics, evolution, and biology of scelionid and platygastrid wasps. Annu. Rev. Entomol. 50: 553582. Baker, J.R., and H.H. Neunzig. 1970. Biology of the blueberry bud mite. J. Econ. E ntomol. 63: 7479. Baados, M.P. 2009. Expanding blueberry production into nontraditional production areas: Northern Chile and Argentina, Mexico and Spain. Acta Hortic. 810: 439444. Barfield, C.S., E.R. Mitchell, and S.L. Poe. 1978. A temperature depende nt model for fall armyworm development. Ann. Entomol. Soc. Am. 71: 70 74. Barnes, H.F. 1948. Gall midges of economic importance, vol. 3, pp. 7180. In Gall M idges of fruit Crosby .Lockwood & Son Ltd London, U nited K ingdom Baxendale, F.P., G.L. Teetes, a nd P.J.H. Sharpe. 1984. Temperature -dependent model for sorghum midge (Diptera: Cecidomyiidae) spring emergence. Environ. Entomol. 13: 15661571. Baxendale, F.P., and G.L. Teetes. 1983. Factors influencing adult emergence from diapausing sorghum midge, Con tarinia sorghicola (Diptera: Cecidomyiidae). Environ. Entomol. 12: 10641067. Beckwith, C. S. 1941. Control of cranberry fruitworm on blueberries. J. Econ. Entomol. 34: 169171. Blizzard, S. 2006. Blueberries in California. pp. 211 213. In N.F. Childers an d P.M. Lyrene [eds ], Blueberries for growers, gardeners, promoters. E.O. Painter Printing Company, Inc., DeLeon Springs, FL. Bosio, G., C. Bogetti, G. Brussino, F. Gremo, and F. Scarpelli 1998. Dasineura oxycoccana, a new pest of blueberry in Italy. Informatore Fitopatologico. 11: 36 41. Burks, B.D. 1967. The North American species of Aprostocetus Westwood (Hymenoptera: Eulophidae). Ann. Entomol. Soc. Am. 60: 756760.

PAGE 108

108 Campbell, A, B.D. Frazer, N. Gilbert, A.P. Gutierrez, and M. Mackauer. 1974. Tempera ture requirements of some aphids and their parasites. J. Appl. Ecol. 11: 431438. Chen, M. and A.M. Shelton. 2007. Impact of soil type, moisture, and depth on swede midge (Diptera: Cecidomyiidae) pupation and emergence. Environ. Entomol. 36: 13491355. Che n, M., J.Z. Zhao, and A.M. Shelton. 2007. Control of Contarinia nasturtii Keiffer (Diptera: Cecidomyiidae) by foliar sprays of acetamiprid on cauliflo wer transplants. Crop Prot 26: 15741578. Cline, W.O., and A. Schilder. 2006. Identification and control of blueberry diseases. pp. 115138. In N.F. Childers and P.M. Lyrene [ eds ], Blueberries for growers, gardeners, promoters. E.O. Painter Printing Company, Inc., DeLeon Springs, FL. Cockfield, S.D., and D.L. Mahr. 1994. Phenology of oviposition of Dasyneura oxycoccana (Diptera: Cecidomyiidae) in relation to cranberry plant growth and flowering. Great Lakes Entomol 27: 185188. Crocker, T.E., A.J. Rose, J.G. Williamson, and P.C. Andersen. 1994. Deciduous fruit for the home garden in Central Florida. http://edis.ifas.ufl.edu/MG366 Crocker, T.E., and L. Willis. 1989. Survey of southern highbush and rabbiteye blueberries in Florida. Proc. Fla. State Hortic Soc. 102: 204206. Dent, D. R. 1997. Quantifying insec t populations: estimates and parameters, pp. 57 110. In D.R. D ent, and M.P. Walton [eds.] Methods in ecological and agricultural entomology. CAB International, New York, NY. Dernisky, A.K., R.C. Evans, O.E. Liburd, and K. MacKenzie. 2005. Characterization of early floral damage by cranberry tipworm ( Dasineura oxycoccana Johnson) as a precursor to reduced fruit set in rabbiteye blueberry ( Vaccinium ashei Reade). Intl. J. Pest Manage. 51: 143148. Diaz, B.M., M. Muiz, L. Barrios, and A. Fereres. 2007. Temperature thresholds and thermal requirements for development of Nasonovia ribisnigri (Hemiptera: Aphididae). Environ. Entomol. 36: 681688. Dow AgroSciences. 2007. Delegate Insecticide (specimen label). In ChemSearch CDMS Crop Data Management Systems, Inc. http://www.cdms.net. Driggers, B.F. 1926. The blueberry tip worm ( Contarinia vaccinii Felt.), a new species of midge attacking cultivated blueberries. J. N.Y. Entomol. Soc. 34: 8285. Eigenbrode, S.D., and K.E. Espelie. 1995. Effects of plant epicuticular lipids on insect herbivores. Annu Rev. Entomol. 40: 171194.

PAGE 109

109 Finn, E. 2003. Developing integrated pest management (IPM) techniques for managing keyinsect pests of blueberries in the Southeastern United States. M.S. The sis, University of Florida, Gainesville, FL. Fisher, R.W. and G.L. Teetes. 1982. Effect of moisture on sorghum midge (Diptera: Cecidomyiidae) emergence. Environ. Entomol. 11: 946948. Food Quality Protection Act. 1996. P. L. 104170. United States Congressional Board. Vol. 142: 110 Stat. 14891538. Foster, S.P., M.O. Harris, and J.G. Millar. 1991. Identification of the sex pheromone of the Hessian fly, Mayetiola destructor (Say). Naturwissenschaften. 78: 130131. Foster J.E., and P.L. Taylor. 1975. Thermal unit requirements for development of the Hessian fly under controlled conditions. Environ Entomol. 4: 195 202. Fulton, B. 1940. The blueberry bud mite, a new pest. J. Econ. Entomol. 33:699. Gagn, R.J. 1994. The gall midges of the neotropical region. Cornell University Press, Ithaca, NY. Gagn, R.J. 1989. The plant -feeding gall midges of North America. Cornell University Press, Ithaca, NY. Gagn, R.J., and M.O. Harris. 1998. The distinction between Dasineura spp. (Dip tera: Cecidomyiidae) from apple and pear. Proc. Entomol. Soc. Wash. 100: 445448. Garthwaite, D.G., and C. Wall. 1986. Further evidence for a female sex pheromone in the black currant leaf midge Dasineura tetensi pp. 355357. In Proceedings, British Crop Protection Conference Pest and Diseases, 17 20 November 1986, Brighton, UK, BCPC Publications, Hampshire, UK. Gillespie, D.R., G. Opit, and B. Roitberg. 2000. Effects of temperature and relative humidity on development, reproduction, and predation in Fel tiella acarisuga (Vallot) (Dip tera: Cecidomyiidae). Biol. Control. 17: 132138. Harris, M.O., and S.P. Foster. 1991. Wind tunnel studies of sex pheromone -mediated behavior of the Hessian fly (Diptera: Cecidomyiidae). J. Chem. Ecol. 17: 24212435. Harris, M .O., S.P. Foster, K. Agee, and S. Dhana. 1996. Sex pheromone communication in the apple leafcurling midge ( Dasineura mali ). pp. 52 58. In M. OCallaghan [ed.] Proceedings, 49th New Zealand Plant Protection Conference, 13 15 August 1996, Nelson, NZ. The Ne w Zealand Plant Protection Society Inc., Auckland, New Zealand. Heath, J.J., A.Zhang, W.L. Roelofs, and R.F. Smith. 2005. Flight activity and further evidence for a femaleproduced sex pheromone of the apple leaf midge, Dasineura mali in Nova Scotia. Nort heast. Nat 12: 93102.

PAGE 110

110 Hillbur, Y. 2001. Tracking the tiny identification of the sex pheromone of the pea midge as a prerequisite for pheromone -based monitoring. Ph.D. dissertation, Swedish University of Agricultural Sciences, Uppsala, Sweden. Hillbur, Y., M. Bengtsson, J. Lfqvist, A. Biddle, O. Pillon, E. Plass, W. Francke, and E. Hallberg. 2001. A chiral sex pheromone system in the pea midge, Contarinia pisi J. Chem Ecol. 27: 13911407. Hillbur, Y., M. Celander, R. Baur, S. Rauscher, J. Haftmann, S. Franke, and W. Fr ancke. 2005. Identification of the sex pheromone of the swede midge, Contarinia nasturtii J. Chem. Ecol. 31: 18071828. Hinks, C.F., and J.F. Doane. 1988. Observations on rearing and diapause termination of Sitodiplosis mosellana (Diptera: Cecidomyiidae) in the laboratory. J. Econ. Entomol. 81: 18161818. Hodgson, E.W., I.V. Macrae, and G.J. Brewer. 2004. Within -field distribution of the sunflower midge (Diptera: Cecidomyiidae). Environ. Entomol. 33: 10371044. Howell, J.F., and L .G. Neven. 2000. Physiological development time and zero development temperature of the codling moth (Lepidoptera: Tortricidae). Environ. Entomol. 29: 766772. Johnson, C.W. 1900. Some notes and descriptions of seven new species and one new genus of Dipter a. Entomol. News. 11: 323328. Kolesik, P. 2000. Distribution of infestation by lentil gall midge Contarina lentis (Diptera Cecidomyiidae) in lentil field s: statistical model. J. Appl. Entomol. 124: 7 10. Lactin, D.J., N.J. Holliday, D.L. Johnson, and R. C raigen. 1995. Improved rate model of temperature-dependent development by arthropods. Environ. Entomol. 24: 6875. Lamb, R.J., I.L. Wise, O.O. Olfert, J. Gavloski, and P.S. Barker. 1999. Distribution and seasonal abundance of Sitodiplosis mosellana (Dipter a: Ceci domyiidae) in spring wheat. Can. Entomol. 131: 387397. Lampo, M., and V. Medialdea. 1994. Dispersal pattern of the sorghum midge (Diptera: Cecidomyiidae) in sorghum plots. Environ. Entomol. 23: 551 555. Li, Y., W.Q. Yang, W. Lin, and Z. Zhang. 2006. Blueberries in China. pp. 243247. In N.F. Childers and P.M. Lyrene [ eds. ], Blueberries for growers, gardeners, promoters. E.O. Painter Printing Company, Inc., DeLeon Springs, FL. Liburd, O.E. 2004. The effectiveness of various insecticides to control bl ueberry gall midge. Berry/Vegetable Times. 4: 2 4.

PAGE 111

111 Liburd, O.E., and H.A. Arevalo. 2006. Insects and mites in blueberries. pp. 99110. In N.F. Childers and P.M. Lyrene [eds. ], Blueberries for growers, gardeners, promoters. E.O. Painter Printing Company, In c., DeLeon Springs, FL. Liburd, O.E., E.M. Finn, K.L. Pettit, and J.C. Wise. 2003. Response of blueberry maggot fly (Diptera: Tephritidae) to imidacloprid -treated spheres and selected insecticides. Can. Entomol. 135: 427438. Liburd, O.E., S.Polavarapu, S.R. Alm, and R.A. Casagrande. 2000. Effect of trap size, placement, and age on captures of blueberry maggot flies (Diptera: Tephritidae). J. Econ. Entomol. 93: 14521458. Littell, R.C., W.W. Stroup, and R.J. Freund. 2002. SAS for Linear Models, 4th ed. SAS Institute Inc., Cary, NC. Logan, J.A., D.J. Wollking, S.C. Hoyt, and L.K. Tanigoshi. 1976. An analytic model for description of temperature development rate phenomena in arthropods. Environ. Entomol. 5: 11331140. Lyrene, P. M. 1989. The Florida climate as it relates to blueberry production. Proc. Fla. State Hort ic Soc. 102: 208212. Lyrene, P.M. and C. Muoz. 1997. Blueberry production in Chile. J. Small Fruit Vitic. 5: 1 20. Lyrene, P.M., and J.A. Payne. 1992. Blueberry gal l midge: a pest on rabbiteye blueberry in Florida. Proc. Fla. State Hortic Soc. 105: 297300. Lyrene, P.M., and J.A. Payne. 1995. Blueberry gall midge: a new pest of rabbiteye blueberries. J. Small Fruit Vitic. 3: 111 124. MacGown, M.W. and E.A. Osgood, Jr. 1972. Taxonomy and biology of chalcidoid and proctotrupoid Hymenoptera associated with Dasineura balsamicola (Diptera: Cecidomyiidae) in Maine. Can. Entomol. 104: 1259 1269. Mahr, D.L., and R. Kachadoorian. 1990. Cranberry tipworm. Proc. Wisconsin Cranb erry School. Pp 1722. Marucci, P.E. 1966. Insects and their control. pp. 199235. In P. Eck and N.F. Childers [ eds. ], Blueberry culture. Rutgers University Press, New Brunswick, NJ. Marucci, P.E. 1977. Cranberry ins ects in New Jersey. Acta Hortic 61: 231-239. McKay, P.A., and J.H. Hatchett. 1984. Mating behavior and evidence of a female sex pheromone in the Hessian fly, Mayetiola destructor (Say) (Diptera: Cecidomyiidae. Ann. Entomol. Soc. Am. 77: 616620. Milne, D.L. 1961. The function of the sternal spatula in gall midges. Proc. R. Ent omol Soc. Lond. (A) 6: 126131.

PAGE 112

112 Moore, A.D. 1987. Effects of temperature and length of photophase on development and diapause in Cystiphora schmidti (Rbsaamen) (Diptera: Cecidomyiidae). J. Aust. Ent omol Soc. 26: 3493 54. Moore, J.N. 1994. The blueberry industry of North America. HortTechnology. 4: 96 102. (NASS -USDA) National Agricultural Statistics Service and United States Department of Agriculture. 2008. Non -citrus fruits and nuts 2007 summary. http://usda.mannlib.cornell.edu/usda/nass/NoncFruiNu//2000s/2008/NoncFruiNu0708 2008_revision.pdf Naumann, W.D. 1993. Overview of the Vaccinium industry in We st ern Europe. Acta Hortic 346: 5358. (NDSU) North Dakota State University Extension Service. n.d. Emergence traps for detecting orange wheat blossom midge. NDSU Extension Service, Department of Entomology. http://www.ag.ndsu.nodak.edu/aginfo/entomology/entupdates/Wheat_Midge/mdgtrp2.htm NeSmith, D.S. 2006. Blueberries in Georgia. pp. 219 221. In N.F. Childers and P.M. Lyrene [eds. ], Blueberries for growers, gardeners, promoters. E.O. Painter Printing Company, Inc., DeLeon Springs, FL. Neunzig, H.H., and G.J. Galletta. 1977. Abundance of blueberry bud mite on various species of blueberry. J. G a Entomol. Soc. 12: 182184. Patel, N. 19 93. Blueberry culture in Australasia. Acta Hortic 346: 4552. Pedigo, L.P. 2002. Entomology and pest management, 4th ed. Prentice Hall, Upper Saddle River, NJ. Pea, J.E., and R. Duncan. 1992. Sampling methods for Prodiplosis longifila (Diptera: Cecidomyi idae) in limes. Environ. Entomol. 21: 996 1001. Pivnick, K.A. 1993. Response of males to female sex pheromone in the orange wheat blossom midge, Sitodiplosis mosellana (Ghin) (Diptera: Cecidomyiidae). J. Chem. Ecol. 19: 16771689. Pivnick, K.A., and E. La bb. 1992. Emergence and calling rhythms, and mating behavior of the orange wheat blossom midge, Sitodiplosis mosellana (Ghin) (Diptera: Cecidomyiidae). Can. Entomol. 124: 501507. Pliszka, K. 1993. The blueberry industry and research in Eastern Eu rope: r eview. Acta Hortic 346: 4143. Prior, R.L., G. Cao, A. Martin, E. Sofic, J. McEwen, C. OBrien, N. Lischer, M. Ehlenfeldt, W. Kalt, G. Krewer, and C.M. Mainland. 1998. Antioxidant capacity as influenced by total phenolic and anthocyanin content, maturity, and variety of Vaccinium species. J. Agric. Food Chem. 46: 26862693.

PAGE 113

113 Pruess, K.P. 1983. Day -degree methods for pest management. Environ. Entomol. 12: 613619. Rajmohana, K. 2006. Studies on Proctotrupoidae and Platygastroidea (Hymenoptera: Insecta) of Ke rala. Memoirs of the Zoological Survey of India. 21: 1 153. Rock, G.C. and P.L. Shaffer. 1983. Developmental rates of codling moth (Lepidoptera: Olethreutidae) reared on apple at four constant temperatures. Environ. Entomol. 12: 831834. Sampson, B.J., and J.H. Cane. 2000. Pollination efficiencies of three bee (Hymenoptera: Apoidea) visiting rabbiteye blueberry. J. Econ. Entomol. 93: 17261731. Sampson, B.J., S.J. Stringer, and J.M. Spiers. 2002. Integrated pest management for Dasineura oxycoccana (Diptera: Cecidomyiidae) in blueberry. Environ. Entomol. 31: 339347. Sampson, B.J., S.J. Stringer, J.H. Cane, and J.M. Spiers. 2004. Screenhouse evaluations of a mason bee Osmia ribifloris (Hymenoptera: Megachilidae) as a pollinator for blueberries in the Sou theas tern United States. Small Fruits R eview. 3: 381392. Sampson, B.J., T.A. Reinhart, O.E. Liburd, S.J. Stringer, and J.M. Spiers. 2006. Biology of parasitoids (Hymenoptera) attacking Dasineura oxycoccana and Prodiplosis vaccinii (Diptera: Cecidomyiidae) in c ultivated blueberries. Ann. Entomol. Soc. Am. 99: 113120. Sandanayaka, W.R.M., and J.G. Charles. 2006. Host location and ovipositional behavior of Platygaster demades Walker (Hymenoptera: Platygastridae), an egg parasitoid of apple and pear leaf cu rling m idges. J. Insect Behav 19: 99 113. Sarzynski, E.M., and O.E. Liburd. 2003. Techniques for monitoring cranberry tipworm (Diptera: Cecidomyiidae) in rabbiteye and southern highbush blueberries. J. Econ. Entomol. 96: 18211827. SAS Institute. 2003. SAS s ys tem for Window s version 9.1. SAS Institute, Inc., Cary, NC. Schauff, M.E., J. LaSalle, and L.D. Coote. 1997. Chapter 10. Eulophidae. In G.A.P. Gibson, J.T. Huber, and J.B. Woolley [eds.], Annotated keys to the genera of Nearctic Chalcidoidea (Hymenoptera). NRC Research Press, Ottawa, Ontario, Canada. Schick, C.M. 1993. Overview of the blueberry industry in S outh America. Acta Hortic 346: 2732. Sharma, H.C., and P. Vidyasagar. 1992. Orientation of males of sorghum midge, Contarinia sorghicola to sex pheromones from virgin females in the field. Entomol. Exp. App l 64: 2329. Shutak, V.G., and P.E. Maruci, 1966. Plant and fruit development, pp. 179 198. In P. Eck and N.F. Childers [eds.] Blueberry culture. Rutgers University Press, New Brunswick, NJ.

PAGE 114

114 Slifer, E.H. and S.S. Sekhon. 1971. Circumfila and other sense organs on the antenna of the sorghum midge (Diptera: Cecidomyiidae). J. Morph ol 133: 281302. Smith, J.T., and R.B. Chapman. 1996. Comparison between emergence and egg laying in apple leafcurling midge ( Dasyneura mali ), pp. 4851. In M. OCallaghan [ed.] Proceedings, 49th New Zealand Plant Protection Conference, 13 15 August 1996, Nelson, NZ. The New Zealand Plant Protection Society Inc., Auckland, New Zealand. Solinas, M., and N. Isodoro. 1991. Identification of the female sex pheromone gland in the sorghum midge, Allocontarinia sorghicola (Coq.) Solinas (Diptera: Cecidomyiidae). Redia. 74: 441466. Spiers, J.M. 1978. Effect of stage of bud development on cold injury in rabbiteye blueberry. J. Am. Soc. Hortic. Sci. 103: 452455. Steck, G.J., P.M. Lyrene, and J.A. Payne. 2000. Blueberry gall midge, Dasineura oxycoccana (Johnson) (Ins ecta : Diptera: Cecidomyiidae). http://edis.ifas.ufl.edu/IN293 Steck, G.J., and J.A. Payne. 1998. Blueberry maggot, Rhagoletis mendax Curran. http://edis.ifas.ufl.edu/IN198 Strick, B.C., and D. Yarborough. 2005. Blueberry production trends in North America, 1992 to 2003, and predictions for growth. HortTechnology. 15: 391398. Suckling, D.M., J.T.S. Walker, P.W. Shaw, L.A. Manning, P. Lo, R. Wa llis, V. Bell, W.R. Manoharie Sandanayaka, D.R. Hall, J.V. Cross, and A. M. El Sayed. 2007. Trapping Dasineura mali (Diptera: Cecidomyiidae) in apples. J. Econ. Entomol. 100: 745751. Tamada, T., and H. Iwagaki. 1977. The pollination of rabbiteye blueberr i es in Tokyo. Acta Hortic 61: 335341. Taquini, I.L. 2006. Argentine Blueberries. pp. 257259. In N.F. Childers and P.M. Lyrene [eds. ], Blueberries for growers, gardeners, promoters. E.O. Painter Printing Company, Inc., DeLeon Springs, FL. Trehane, J. 2004. Blueberries, cranberries and other Vacciniums Timber Press, Portland, OR. Trinka, D.L. 1997. Production trends in the cultivated blueberry industry of N orth America. Acta Hortic 446: 3740. Triplehorn, C.A., and N.F. Johnson. 2005. Borror and DeLongs Introduction to the study of insects, 7th ed. Thomson Brooks/Cole, Belmont, CA. Turner, J.C., and O.E. Liburd. 2007. Insect management in blueberries in the eastern United States. http://edis.ifas.ufl.edu/IG070

PAGE 115

115 Van Kirk, J.R. and M.T. Aliniazee. 1981. Determining low -temperature threshold for pupal development of the western cherry fruit fly for use in phenology models. Environ. Entomol. 10: 96871. van Lenteren, J.C., M. Schettino, N. Isidoro, R. Romani, and J. van Schelt. 2002. Morphology of putative female sex pheromone glands and mating behavior in Aphidoletes aphidimyza. Entomol. Exp. App l 102: 199209. Vtek, G., C. Thurczy, and B. Pnzes. 2006. Interrelationship between the raspberry cane midge, Resseliella theobaldi (Diptera: Cecidomyiidae) and its parasitoid, Aprostocetus epicharmus (Hymenoptera: Eulophidae). Bull. Entomol. Res. 96: 367372. Walker, T.J., J.J. Gaffney, A.W. Kidder and A.B. Ziffer. 1993. Florida reach ins: environmental chambers for entomological research. Am. Entomol. 39: 177-1 82. Weisz, R., S. Fleischer, and Z. Smilowitz. 1996. Site -specific integrated pest management for high -value crops: impact on potato pest management. J. Econ. Entomol. 89: 501 509. Weston, P.A. and M.D. Dia z. 2005. Thermal requirements and development of immature stages of viburnum leaf beetle, Pyrrhalta viburni (Paykull) (Coleoptera: Chrysomelidae). Environ. Entomol. 34: 985 989. Williams, I.H., and A.P. Martin. 1986. Evidence for a fema le sex pheromone in the brassica pod midge Dasineura brassicae Physiol. Entomol. 11: 353356. Williams, C.T., O. Okhidievbie, K.M. Harris, and M.N. Ukwungwu. 1999. The host range, annual cycle and parasitoids of the African rice gall midge Orseolia oryzivora (Diptera: Cecidomyiidae) in central and southeast Nigeria. Bull. Entomol. Res 89: 5895 97. William son, J.G. and P.M. Lyrene. 2004a Blueberry varieties for Florida. http://edis.ifas.ufl.edu/HS215 Williamson, J.G., and P.M. Lyrene. 2004b. Blueberr y gardeners guide. http://edis.ifas.ufl.edu/MG359 Williams on, J.G., and P.M. Lyrene. 2004c Reproductive growth and development of blueberry. http://edis.ifas.ufl.edu/pdffiles/HS/HS22000.pdf Williamson, J.G., P.M. Lyrene, and E.P. Miller. 2000. A survey of blueberry acreage in Florida. Proc. Fla. State Hortic Soc. 113: 2425. Wiseman, B.R., H.R. Gross, Jr., and W.W. McMillian. 1978. Seasonal distribution of the sorghum midge and its hymenopterous parasites, 1975 77. Environ. Entomol. 7: 820822. Withers, T.M., M.O. Harris, and L.K. Davis. 1995. The incidence of Hessian fly and other pests in New Zealand wheat crops, pp. 165169. In A.J. Popay [ed.] Proceedings, 48th New Zealand Plant Protection Conference, 8 10 August 1995, Hastings, NZ. The New Zealand Plant Protection Society Inc., Auckland, New Zealand.

PAGE 116

116 Yang, W.Q. 2005. Blueberry gall midge: a possible new pest in the Northwest. http://extension.oregonstate.edu/catalog/html/em/em8889/ Zheng, W., and S.Y. Wang. 2003. Oxygen radical absorbing capacity of phenolics in blueberries, cranberries, chokeberries, and ligonberries. J. Agric. Food Chem. 51: 502509.

PAGE 117

117 BIOGRAPHICAL SKETCH Craig R. Rou bos was born in Grand Rapids, Michigan in 1979. He attended Calvin College in Grand Rapids, MI where he majored in biology. After receiving his bachelors degree, he began working on a masters degree at Purdue University in West Lafayette, Indiana studying entomology and integrated pest management. The subject of Craigs masters research was the movement of plum curculio from overwintering sites into apple orchards. He received his masters degree in entomology in 2004 then moved to Florida to pursue a Doctor of Philosophy degree at the University of Florida, Gainesville, FL. His Ph.D. research focused on improving monitoring methods and m anagement practices for blueberry gall midge, a key pest of blueberries in the southeastern United States. He plans to pursue a career as a research entomologist. In 2009, Craig married Gayle Cohen, a fellow graduate of the University of Florida.