Citation
Developing Integrated Pest Management (IPM) Techniques for Managing Key Insect Pests of Blueberries in the Southeastern United States

Material Information

Title:
Developing Integrated Pest Management (IPM) Techniques for Managing Key Insect Pests of Blueberries in the Southeastern United States
Creator:
FINN, ERIN
Copyright Date:
2008

Subjects

Subjects / Keywords:
City of Gainesville ( local )
Blueberries ( jstor )
Planting ( jstor )
Infestation ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Erin Finn. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
6/1/2004
Resource Identifier:
53207968 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

DEVELOPING INTEGRATED PEST MANAGEMENT (IPM) TECHNIQUES FOR MANAGING KEY INSECT PESTS OF BL UEBERRIES IN THE SOUTHEASTERN UNITED STATES By ERIN FINN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Erin Finn

PAGE 3

Dedicated to my fianc Michael Sarzynski.

PAGE 4

ACKNOWLEDGMENTS I thank Jessica Kostarides, Rhiannon O’Brien, and Caitlin Clements (Michigan State University) for their help collecting field data in Michigan. I am grateful to Gisette Seferina, Scott Weihman, Jon Hamill, Carolyn Mullin, Rajya Pandey, Ashley Johnson, Daniel Frank, and Jeff White (University of Florida) for their help in collecting field data and making insecticide applications in Florida. I am especially grateful to Dr. Kenna MacKenzie (Agriculture and Agri-Food Canada) for her help and guidance over the past 2 years. I thank Dan Evarts (USDA) for making insecticide applications and collecting bud samples in Georgia. I am indebted to Ken Lyles, Alto Straughn, Donna Miller, Jerry Mixon, and Ches Bennett for allowing us to conduct my experiments in their plantings. I thank Dr. Kenneth Portier (University of Florida) for his assistance in developing a subsampling protocol for counting thrips. I thank Dr. Raymond Gagn (USDA-ARS, Beltsville, MD) and Dr. Blair Sampson (USDA-ARS, Poplarville, MS) for their assistance in identifying Cecidomyiid specimens. I thank Dr. Heather McAuslane and Dr. Paul Lyrene for critical review of the thesis. Above all, I thank my advisor Dr. Oscar Liburd for his guidance and support throughout the project. The Michigan Blueberry Growers Association, the Florida Blueberry Growers Association, and a USDA Pest Management Alternatives (P-MAP) #03774 grant supported this study. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES...........................................................................................................ix ABSTRACT.........................................................................................................................x CHAPTER 1 INTRODUCTION........................................................................................................1 Blueberry Production in Florida...................................................................................1 Literature Review.........................................................................................................3 Cranberry Fruitworm, Acrobasis vaccinii Riley...................................................3 Blueberry Gall Midge, Dasineura oxycoccana (Johnson)....................................6 Flower Thrips, Frankliniella spp...........................................................................8 2 IMPROVING MONITORING TECHNIQUES FOR MANAGING CRANBERRY FRUITWORM (LEPIDOPTERA: PYRALIDAE) IN HIGHBUSH BLUEBERRIES....................................................................................13 Materials and Methods...............................................................................................16 Trap Height..........................................................................................................16 Trap Geographical Location................................................................................17 Effects of Tebufenozide on A. vaccinii Infestation.............................................19 Data Analysis.......................................................................................................20 Results.........................................................................................................................21 Trap Height..........................................................................................................21 Trap Geographical Location................................................................................22 Effects of Tebufenozide on A. vaccinii Infestation.............................................24 Discussion...................................................................................................................24 3 TECHNIQUES FOR MONITORING BLUEBERRY GALL MIDGE (DIPTERA: CECIDOMYIIDAE) IN RABBITEYE AND SOUTHERN HIGHBUSH BLUEBERRIES....................................................................................35 Materials and Methods...............................................................................................37 Unbaited Colored Traps......................................................................................38 Monitoring Techniques.......................................................................................39 v

PAGE 6

Bud Type Comparison.........................................................................................41 Data Analysis.......................................................................................................41 Results.........................................................................................................................41 Unbaited Colored Traps......................................................................................41 Monitoring Techniques.......................................................................................42 Bud Type Comparison.........................................................................................44 Discussion...................................................................................................................45 4 EVALUATION OF MONITORING TECHNIQUES FOR FLOWER THRIPS (THYSANOPTERA: THRIPIDAE) IN SOUTHERN HIGHBUSH AND RABBITEYE BLUEBERRIES..................................................................................55 Materials and Methods...............................................................................................57 Unbaited Colored Traps......................................................................................59 Sampling Techniques..........................................................................................60 Data Analysis.......................................................................................................62 Results.........................................................................................................................62 Unbaited Colored Traps......................................................................................62 Sampling Techniques..........................................................................................64 Discussion...................................................................................................................65 5 EVALUATION OF SELECTED INSECTICIDES FOR MANAGING BLUEBERRY GALL MIDGE (DIPTERA: CECIDOMYIIDAE) AND FLOWER THRIPS (THYSANOPTERA: THRIPIDAE) IN SOUTHERN HIGHBUSH AND RABBITEYE BLUEBERRIES...................................................75 Materials and Methods...............................................................................................77 Blueberry Gall Midge..........................................................................................77 Flower Thrips......................................................................................................79 Data Analysis.......................................................................................................81 Results.........................................................................................................................82 Blueberry Gall Midge..........................................................................................82 Flower Thrips......................................................................................................83 Discussion...................................................................................................................84 LIST OF REFERENCES...................................................................................................93 BIOGRAPHICAL SKETCH.............................................................................................99 vi

PAGE 7

LIST OF TABLES Table page 2-1. Captures of male A. vaccinii moths in pheromone-baited traps placed at different heights with respect to blueberry bush canopy..........................................29 2-2. Captures of male A. vaccinii moths in pheromone-baited traps placed at different locations within blueberry plantings with respect to adjacent woodlands.................................................................................................................30 2-3. Percentage of blueberry clusters infested by A. vaccinii at different geographic locations within an organic planting in Holland, MI...............................................31 2-4. Percentage of blueberry clusters infested by A. vaccinii after a single application of tebufenozide within an abandoned planting in Holland, MI.............32 3-1. Comparison of techniques for monitoring D. oxycoccana in rabbiteye blueberries in Windsor, FL (2002)...........................................................................49 3-2. Comparison of techniques for monitoring D. oxycoccana in southern highbush blueberries in Windsor, FL (2002)...........................................................................50 3-3. Comparison of techniques for monitoring D. oxycoccana in rabbiteye blueberries in Windsor, FL (2003)...........................................................................51 3-4. Comparison of techniques for monitoring D. oxycoccana in southern highbush blueberries in Windsor, FL (2003)...........................................................................52 3-5. Infestation of rabbiteye and southern highbush blueberry buds by D. oxycoccana, Windsor, FL (2002)........................................................................53 3-6. Infestation of rabbiteye and southern highbush blueberry buds by D. oxycoccana, Windsor, FL (2003)........................................................................54 4-1. Comparison of various colors of sticky board traps for monitoring F. bispinosa in rabbiteye and southern highbush blueberries in north-central Florida (2002)...........................................................................................................68 4-2. Comparison of various colors of colored sticky board traps for monitoring F. bispinosa in rabbiteye and southern highbush blueberries in north-central Florida (2003)...........................................................................................................69 vii

PAGE 8

4-3. Comparison of sampling techniques for monitoring F. bispinosa in rabbiteye and southern highbush blueberries in north-central Florida (2003).........................70 5-1. Effect of selected insecticides on infestation of rabbiteye floral buds by D. oxycoccana, Alma, GA (2003)............................................................................87 5-2. Effect of selected insecticides on infestation of southern highbush flowers by F. bispinosa, Haines City, FL (2003).......................................................................88 viii

PAGE 9

LIST OF FIGURES Figure page 1-2. Results of a survey distributed to blueberry growers in Florida and South Georgia (2001). ......................................................................................................12 2-1. Spatial orientation of pheromone-baited traps for monitoring male A. vaccinii moths in the trap geographical location study (Holland and Covert, MI)................33 2-2. Captures of male A. vaccinii moths in pheromone-baited traps in Holland, MI (2001). .................................................................................................................34 4-1. Results of a pilot study to determine how many quadrats on an individual sticky board trap had to be counted to reasonably estimate true counts of F. bispinosa. ...........................................................................................................71 4-2. Sticky board trap overlaid with a transparency of a grid system to facilitate accurate counting of flower thrips. .........................................................................72 4-3. Abundance of F. bispinosa throughout the flowering period of a rabbiteye blueberry planting in Windsor, FL (2002 and 2003). ............................................73 4-4. Abundance of F. bispinosa throughout the flowering period of a southern highbush blueberry planting in Inverness, FL (2002 and 2003). ...........................74 5-1. Infestation of untreated floral buds by D. oxycoccana in Alma, GA (2003). ........89 5-2. Effect of selected insecticides on infestation of rabbiteye cv. ‘Climax’ floral buds by D. oxycoccana, Alma, GA (2003). ...........................................................90 5-3. Effect of selected insecticides on infestation of rabbiteye cv. ‘Tifblue’ floral buds by D. oxycoccana, Alma, GA (2003). ...........................................................91 5-4. Effect of selected insecticides on infestation of southern highbush blueberry flowers by F. bispinosa in Haines City, FL (2003). ...............................................92 ix

PAGE 10

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science DEVELOPING INTEGRATED PEST MANAGEMENRT (IPM) TECHNIQUES FOR MANAGING KEY INSECT PESTS OF BLUEBERRIES IN THE SOUTHEASTERN UNITED STATES By Erin Finn August 2003 Chair: Oscar Liburd Major Department: Entomology and Nematology In 2001, a survey was administered to blueberry growers in Florida and South Georgia to determine which insect pests were present in their plantings. Cranberry fruitworm, Acrobasis vaccinii Riley; blueberry gall midge, Dasineura oxycoccana (Johnson); and flower thrips, Frankliniella spp. were cited most frequently. Methods for monitoring these pests were evaluated. Winged traps baited with synthetic sex pheromone lures ((E, Z)-8, 10-pentadecadien-1-ol and (E)-9-pentadecen-1-ol acetate) were evaluated for their effectiveness in monitoring A. vaccinii at various heights and geographical locations within highbush blueberry (Vaccinium corymbosum L.) plantings. In height studies, three positions were evaluated: 15 cm below the uppermost branch 60 cm below the uppermost branch 20 cm above ground level x

PAGE 11

Traps placed 15 and 60 cm below the uppermost branch captured significantly more male moths compared with traps placed 20 cm above ground level. In geographical location studies, four treatments were evaluated based on trap location relative to adjacent woodlands. Traps located within 1 m of woodlands adjacent to blueberry plantings captured significantly more male moths compared with traps located 15 and 75 m within plantings. Several techniques were evaluated for their effectiveness in detecting D. oxycoccana, in rabbiteye (V. ashei Reade) and southern highbush (V. corymbosum L. x V. darrowi Camp) blueberry plantings. Three techniques (including yellow unbaited sticky boards, larval emergence from infested buds, and bud dissection) were evaluated for detecting D. oxycoccana eggs, larvae, and adults. The emergence technique detected significantly more D. oxycoccana adults than did other techniques evaluated. Emergence and dissection techniques performed equally well for detecting D. oxycoccana larvae. Dissection was the only technique capable of detecting D. oxycoccana eggs. Using a subsampling technique, several colors of sticky board traps were evaluated for their effectiveness in monitoring flower thrips in rabbiteye and southern highbush plantings. Significantly more flower thrips were detected on blue and white sticky traps compared with yellow or green traps. In other studies, I compared three additional sampling techniques with white sticky boards Dipping flower clusters into alcohol Tapping flowers over white cardboard Destructively sampling flower clusters Significantly more flower thrips were detected on white sticky boards compared with other techniques evaluated. xi

PAGE 12

CHAPTER 1 INTRODUCTION Blueberry Production in Florida The Florida blueberry industry is currently undergoing a conversion from long-established rabbiteye (Vaccinium ashei Reade) plantings to commercial production of southern highbush (V. corymbosum L. x V. darrowi Camp)cultivars, with acreage of southern highbush increasing 23% from 1989 to 2000 (Williamson et al. 2000). The trend toward commercial production of southern highbush plantings in Florida is a response to an increased market for shipping early-season blueberries, which have a higher market value compared with later-ripening blueberries from primary blueberry-producing states (Williamson et al. 2000). In 2002, export sales of southern highbush blueberries exceeded $18 million (Florida Agricultural Statistics Service 2003). Despite this trend, blueberry production in Florida may be threatened by an expected increase in insect problems (Mizell and Johnson 2001). In a recent crop profile for blueberries in the southeastern United States, cranberry fruitworm, Acrobasis vaccinii Riley; blueberry gall midge, Dasineura oxycoccana (Johnson); and flower thrips, Frankliniella spp. were listed as major pests warranting attention, particularly with regard to monitoring and evaluation of alternative chemical management protocols (NeSmith 1999). During fall 2001, a survey was distributed to blueberry growers across the states of Florida and South Georgia to gather information regarding their most critical insect management concerns. The survey included questions concerning the size, type, and age 1

PAGE 13

2 of their plantings; and questions about their typical management strategies (Figure 1-1). Growers were also asked to comment on the presence or absence of several key pests in their plantings, including cranberry fruitworm; blueberry gall midge; flower thrips; blueberry bud mite, Aceria vaccinii (Keifer); and blueberry maggot, Rhagoletis mendax Curran. Overall, twenty surveys were returned. At least 25% of growers reported problems with cranberry fruitworm, blueberry gall midge, flower thrips, and blueberry bud mite (Figure 1-2). These data were used as an index of grower perception regarding the presence of various insect pests in their plantings, although these pests were probably more abundant than survey results imply. After compiling survey responses and visiting several plantings in north-central Florida, I narrowed the focus of my research to cranberry fruitworm, blueberry gall midge, and flower thrips. On October 31, 2001 the Environmental Protection Agency (EPA) announced their decision to reduce and restrict the use of two broad-spectrum organophosphates (OPs)—azinphos-methyl and phosmet—in several crops including blueberries. The OPs are acetylcholine esterase inhibitors, which make this class of insecticides more toxic than reduced-risk insecticide compounds with different modes of action. The EPA decision heightens awareness for reducing insecticide use and intensifies the need for alternative control tactics. Organophosphates are broad-spectrum insecticides that negatively affect non-target insects, pose danger to agricultural workers, and leave residues on fruit (Dinham 1993). Developing effective monitoring programs for key pests affecting blueberry production (including cranberry fruitworm, blueberry gall midge, and flower thrips) could improve timing of sprays and thus reduce the number of

PAGE 14

3 insecticide applications needed to provide effective control. Effective monitoring programs would also provide opportunities to use reduced-risk insecticides. The overall goal of my research was to develop and refine techniques for monitoring selected key pests of blueberries, including cranberry fruitworm, blueberry gall midge, and flower thrips. I also wanted to identify reduced-risk insecticides that could be used in blueberry integrated pest management programs. Literature Review Cranberry Fruitworm, Acrobasis vaccinii Riley The cranberry fruitworm is a major pest of Vaccinium spp. in the eastern United States (Neunzig 1986). Cranberry fruitworm is distributed throughout North America, extending from Nova Scotia southward to Florida, and westward into parts of Wisconsin and Texas (Neunzig 1986). Small populations of cranberry fruitworm have also been recorded in the Pacific Northwest after an accidental introduction in the early 1920s (Crowley 1954). The damage caused by cranberry fruitworm feeding may approach 50% in unmanaged blueberry plantings (Pritts and Hancock 1992). Research aimed at reducing larval feeding by cranberry fruitworm is justified because its range appears to be expanding; and outbreaks have been reported for Vaccinium spp. as far south as northern Florida. Biology and behavior. Cranberry fruitworm belongs to the family Pyralidae and is generally considered a univoltine pest throughout most of its described range (Murray et al. 1996). Adult moths lay their eggs at the calyx end of developing fruit. Newly hatched larvae exit the fruit near the oviposition site and re-enter the berry either at the stem end or at the calyx to begin their feeding cycle (Averill and Sylvia 1998). Larvae develop through six instars before overwintering as hibernacula within the groundcover (Averill

PAGE 15

4 and Sylvia 1998; Godin et al. 2002). Larvae feed on multiple berries to complete their development. Thus a single cranberry fruitworm larva may damage an entire cluster of berries (Simser 1994). Cranberry fruitworm damage is characterized by silken enclosures that larvae construct while webbing berries together. These silk webbings may provide protection for larvae against predators or inclement weather conditions. The amount of silk that larvae produce correlates with instar development (Averill and Sylvia 1998). Also, the frass excreted by cranberry fruitworm larvae often become entangled in webbed fruit, which is one characteristic used to identify the species during infestation. Monitoring. McDonough et al. (1994) extracted glands of female cranberry fruitworm moths and identified several components of their sex pheromone blend. Identification of specific components was conducted by high performance liquid chromatography (HPLC) and gas chromatography-mass spectroscopy (GCMS) procedures (McDonough et al. 1994). Profiled compounds were then tested in laboratory flight-tunnel experiments to determine male moth response. Electroantennographic (EAG) response of male cranberry fruitworm moths was highest for (E, Z)-8, 10-pentadecadien-1-ol, although this component alone did not elicit upwind flight. A combination of (E, Z)-8, 10-pentadecadien-1-ol and (E)-9-pentadecen-1-ol acetate (100:4) was determined to direct male upwind flight at the same rate as extracts from glands of female cranberry fruitworm moths (McDonough et al. 1994). Field trials also confirmed the effectiveness of this ratio in luring male moths, while traps baited with (E, Z)-8, 10-pentadecadien-1-ol alone failed to capture male moths. As a result of this work, lures composed of (E, Z)-8, 10-pentadecadien-1-ol and (E)-9-pentadecen-1-ol acetate were developed for monitoring cranberry fruitworm populations.

PAGE 16

5 Since the development of the cranberry fruitworm sex pheromone lure, experiments have been conducted to evaluate its potential role in blueberry integrated pest management (IPM) programs. Polavarapu et al. (1997 unpublished data) determined that mating disruption would be too costly for growers to implement in the field, since the average cost of cranberry fruitworm trap and lure systems is $6; and traps would need to be spaced approximately 10 to 15 m apart to provide effective control. Although it may be too costly to adopt mating disruption programs using pheromone-baited traps for control of cranberry fruitworm, these traps may work well for improving monitoring programs for cranberry fruitworm. Unfortunately, guidelines for use of synthetic lures to monitor cranberry fruitworm have not yet been established for pest-management programs in either cranberry or blueberry production (Averill and Sylvia 1998). Also, no research has been published on trap deployment within blueberry plantings. Data obtained from effective cranberry fruitworm monitoring programs could aid growers in determining insecticide application dates. The ability to accurately time insecticide sprays may reduce the number of applications needed for adequate control of A. vaccinii in infested plantings. Information on trap height and geographic location of traps within plantings are needed to improve monitoring effectiveness, so that growers can more accurately time insecticide applications. Management. Several reduced-risk compounds developed in the past decade offer potential for managing lepidopteran fruit pests. Specifically, the insect growth regulator tebufenozide has been shown to be effective against early instars of the obliquebanded leafroller (OBLR) Choristoneura rosaceana (Harris) in laboratory bioassays (Waldstein and Reissig 2001a). In field studies, tebufenozide was shown to significantly reduce

PAGE 17

6 apple damage by OBLR at 10 days post-application compared with untreated fruit (Waldstein and Reissig 2001b). Tebufenozide may also be effective against other neonate lepidopterans, including cranberry fruitworm larvae. Blueberry Gall Midge, Dasineura oxycoccana (Johnson) The blueberry gall midge Dasineura oxycoccana (Johnson) is a recently discovered pest of rabbiteye and southern highbush blueberries in the southeastern United States (Lyrene and Payne 1992). Dasineura oxycoccana also occurs in the northern United States, including Maine; New Jersey; Michigan; Wisconsin; and the Pacific Northwest, where it is referred to as cranberry tipworm (Lyrene and Payne 1992). Unmanaged blueberry gall midge infestations can destroy up to 80% of floral buds in susceptible rabbiteye cultivars (Lyrene and Payne 1992, 1995). To date, damage to southern highbush cultivars by blueberry gall midge has not been quantified. In north-central Florida, southern highbush blueberries flower from late January throughout February. This early flowering may enable southern highbush cultivars to avoid high populations of blueberry gall midge that occur later in the spring. By contrast, high populations of blueberry gall midge in the late spring are destructive to vegetative and flowering buds of many rabbiteye cultivars. In Florida, blueberry gall midge has caused severe damage to several rabbiteye plantings throughout the state. Biology and behavior. Blueberry gall midge belongs to the family Cecidomyiidae, which are small insects (approximately 3 mm). These midges have long, slender legs; globular cylindrical antennae with sensory hairs; and transparent wings with fine, black microtrichia and reduced venation (Bosio et al. 1998). Until recently, the presence of blueberry gall midge in Florida blueberry plantings was misdiagnosed as freeze damage (Lyrene and Payne 1995). This discrepancy has

PAGE 18

7 complicated accurate assessment of floral and vegetative bud loss in blueberry plantings throughout Florida. Lyrene and Payne (1992) were the first to identify blueberry gall midge affecting floral buds within rabbiteye blueberries. Until then, only vegetative buds were considered targets for blueberry gall midge infestation. Limited information is available regarding the biology and behavior of D. oxycoccana with respect to blueberry production in the southeast. Dasineura oxycoccana is known to have overlapping generations throughout its described range. Mahr and Kachadoorian (1990) reported that D. oxycoccana completes up to five generations in Wisconsin cranberry bogs, although later work by Cockfield and Mahr (1994) suggested that distinct broods are not easy to distinguish. Sampson et al. (2002) reported that blueberry gall midge is capable of completing up to 11 generations per year in southeastern Mississippi. In Florida, first emergence of blueberry gall midge begins in late January to early February. Milder temperatures appear to favor early emergence. Midge populations appear to be most problematic when peak density coincides with susceptible stages of bud development, particularly during bud swell (Lyrene and Payne 1995). According to Gagn (1989), adult D. oxycoccana live only long enough to mate and lay eggs, approximately 2-3 days. Female flies oviposit into susceptible buds, and their larvae feed on plant juices, which ultimately kill developing buds (Mahr and Kachadoorian 1990). Larvae develop through 3 instars and drop to the ground to pupate in the soil. The lifecycle varies from 2 to 3 weeks, depending on environmental conditions.

PAGE 19

8 Monitoring. Reliable monitoring techniques to identify and quantify the presence of blueberry gall midge in blueberry plantings have not yet been developed. Currently, yellow sticky boards are used within some plantings to monitor the presence of adult midges and the abundance of their natural enemies (Bosio et al. 1998; Sampson pers. comm.). Larval emergence is another common technique that has been recommended for monitoring blueberry gall midge populations in affected plantings (Sampson 2001 pers. comm.). This strategy requires that samples be observed over a period of time to allow midge larvae to emerge, a tactic that may offer little benefit to growers since infestation would already be established. An alternative approach is bud dissection, which may reduce sampling delay. Management. Recently, a few reduced-risk compounds, including naturalytes and neonicotinoids, have been shown to be effective against dipteran insects (Wise et al. 2002; Liburd et al. 2003). Spinosad (SpinTor 2 SC) is a naturalyte compound that has recently received registration for major pests in blueberries, including flower thrips, although further evaluation is needed to test its effectiveness against D. oxycoccana. Imidacloprid, a neonicotinoid, is another reduced-risk compound that may show promise for control of D. oxycoccana, particularly because of its translaminar effects (Elbert et al. 1990). In addition to reduced-risk compounds, several organic-approved insecticides, including azadirachtin (Ecozin 3% EC) and kaolin clay (Surround WP), should be evaluated for potential control of D. oxycoccana since many growers in the southeastern US choose to market their berries organically. Flower Thrips, Frankliniella spp. Flower thrips belonging to the genus Frankliniella are well-documented pests of plants in the genus Vaccinium (Wood 1956; Braman et al. 1996). There are a number of

PAGE 20

9 thrips species that cause damage to blueberries in North America. In the Southeast, Frankliniella tritici Fitch [eastern flower thrips], F. bispinosa (Morgan) [Florida flower thrips] and F. occidentalis (Pergrande) [western flower thrips] have been identified as pests of both rabbiteye and southern highbush blueberries (Reitz, unpublished data). These three species are known to have wide host ranges and cause damage to many different crop plants, with feeding primarily on lush tissue such as buds, flowers and young leaves. In blueberry, feeding on pollen, styles, and developing berries can initiate major yield losses. Biology and behavior. Flower thrips are small insects (approximately 1 to 1.3 mm long) with yellowish-orange coloration and banding across the abdomen. When viewed under a microscope, adults have long thin wings with fine hairs and distinctive antennae. Females are generally larger than males; and have more distinctive characteristics. In general, flower thrips species have a very short life cycle (approximately 18 to 22 days under ideal conditions); and complete multiple generations per year (Lewis 1997). Females oviposit within plant tissue, making them difficult to see. Thrips development progresses through several nymphal instars, each resembling the adult morph without wings. Adults and nymphs have rasping mouthparts, which are used to extract cell sap from plant tissues. Feeding occurs within the flowers or on mature fruit, causing floral abortion and blemished fruit, respectively. In addition, flower thrips may feed on pollen, which may also lead to fruit abortion. Monitoring. Sticky board traps have been useful for monitoring thrips populations in various citrus crops (Childers and Brecht 1996), although it is uncertain whether this technique will work well in blueberry plantings. Other sampling techniques for thrips

PAGE 21

10 require taking flower clusters from the field to monitor populations. These techniques include dipping flower clusters into alcohol for microscopic evaluation, destructive sampling, or tapping clusters over a sheet of paper. The development of an effective technique for monitoring flower thrips in blueberry plantings would enable growers to make informed management decisions and reduce the number of insecticide applications required for control. Management. Some insecticides targeted for control of D. oxycoccana may also work well for managing flower thrips. For instance, spinosad has a high level of activity against Thysanoptera, with minimal activity against beneficial insects (Eger and Lindenberg 1998). Spinosad also holds promise for control of D. oxycoccana. The translaminar effects of imidacloprid make it suitable for a range of sucking insects, including flower thrips and larvae of dipteran insects (Groves et al. 2001; Wise et al. 2002). Recently, activated garlic extract has been marketed to blueberry growers in the southeast for control of flower thrips. This compound is novel and should also be evaluated for its effectiveness against Frankliniella spp.

PAGE 22

11 Figure 1-1. Pest survey administered to Florida blueberry growers (2001).

PAGE 23

12 25%10%25%30%25%0%10%20%30%40%CranberryFruitwormBlueberryMaggotBlueberry GallMidgeBlueberry BudMiteFlower ThripsInsect Pest% Growers Reporting Problem Figure 1-2. Results of a survey distributed to blueberry growers in Florida and South Georgia (2001). Growers were asked to comment on the presence or absence of various insect pests in their plantings. A total of twenty survey responses were used to compile this chart.

PAGE 24

CHAPTER 2 IMPROVING MONITORING TECHNIQUES FOR MANAGING CRANBERRY FRUITWORM (LEPIDOPTERA: PYRALIDAE) IN HIGHBUSH BLUEBERRIES The cranberry fruitworm, Acrobasis vaccinii Riley, is a major pest of Vaccinium spp. in the eastern United States (Neunzig 1986). In cultivated plantings of highbush blueberries, V. corymbosum L., feeding damage may exceed 50% (Pritts and Hancock 1992). Cranberry fruitworm larvae feed internally, consuming between five and eight berries to complete development (Murray et al. 1996). Newly hatched larvae exit the fruit near the oviposition site and crawl over the berry surface before re-entering the same berry at either the stem end or the calyx to begin feeding (Beckwith 1941; Averill and Sylvia 1998). Damaged blueberry clusters often exhibit silk webbing, rendering the entire cluster unmarketable (Simser 1994; Murray et al. 1996). Currently, commercial blueberry production relies on the use of broad-spectrum insecticides, mainly organophosphates (phosmet and malathion) and carbamates (carbaryl and methomyl), applied from fruit set until full green to control A. vaccinii populations (Eck 1988). The use of organophosphates and some carbamates for control of A. vaccinii is relatively effective, but these insecticides often threaten non-target organisms (Wilson et al. 1991), pollute the environment (Coppage and Braidech 1976), and increase the potential for resistance development (Knight and Hull 1989; Pree et al. 1998). In addition, recent Food Quality Protection Act regulations target many of these compounds for elimination by 2005. Alternatives to broad-spectrum insecticides are being 13

PAGE 25

14 developed, but until new strategies are identified, techniques that reduce the amount of insecticides in a cropping system should be implemented. Pheromone-baited traps and insect growth regulators may improve monitoring effectiveness and provide alternatives to broad-spectrum insecticides for control of A. vaccinii in blueberry plantings. Effective monitoring programs for cranberry fruitworm would enable growers to better time insecticide applications rather than adhering to a traditional calendar spray program. McDonough et al. (1994) identified several components of the cranberry fruitworm sex pheromone from gland extracts of female moths. A combination of (E, Z)-8, 10-pentadecadien-1-ol and (E)-9-pentadecen-1-ol acetate in a 100 to 4 ratio was determined to direct male upwind flight as effectively as female gland extracts in wind tunnel studies (McDonough et al. 1994). Field trials confirmed that this ratio was effective in luring male moths to baited traps. Despite this knowledge, specific guidelines for use of synthetic lures in blueberry integrated pest management programs have not been established. Trap position and location are important factors to consider when refining monitoring programs. Tomlinson (1970) reported that black light traps located 90 cm above vine tips captured more male A. vaccinii moths than traps located at vine level in cranberries, V. macrocarpon Aiton. The effect of trap height may similarly influence attraction of A. vaccinii males to sex pheromone traps in highbush blueberry plantings. Currently, there are no published data to support the hypothesis that trap height influences the attraction of A. vaccinii in highbush blueberries. Averill and Sylvia (1998) suggested the need to establish guidelines for use of synthetic A. vaccinii pheromone

PAGE 26

15 lures in cranberries. Similar guidelines are also needed for the use of synthetic lures in blueberry IPM programs. Another factor to consider when monitoring for A. vaccinii is geographical location of traps within blueberry plantings. In an insecticide efficacy trial, Beckwith (1941) noted higher infestations of A. vaccinii at the edge of blueberry plantings compared with the center, regardless of treatment assignment. More recently, Mallampalli and Isaacs (2002) compared single blueberry plants and individual fruit clusters as sampling units for detecting A. vaccinii and found significantly more eggs in blueberries located adjacent to wooded habitats compared with blueberries located further away. Currently there are no published reports indicating a direct preference for A. vaccinii moths to pheromone-baited traps within various locations of highbush blueberry plantings. Improved monitoring protocols using pheromone-baited traps may allow for the timely use of reduced-risk insecticides for control of A. vaccinii. Insect growth regulators are reduced-risk insecticides that affect insect development primarily by regulating titers of either 20-hydroxyecdysone or juvenile hormone (Croft 1990). Specifically, tebufenozide is a nonsteroidal 20-hydroxyecdysone agonist that selectively targets lepidopteran pests, with decreased toxicity to non-target insects relative to conventional insecticides (Dhadialla et al. 1998; Brown 1994). Tebufenozide has been shown to provide ovicidal activities against other pyralids, including the European corn borer, Ostrinia nubilalis (Hubner), in laboratory assays (Trisyono and Chippendale 1997). Tebufenozide has also been shown to be effective against early instars of the obliquebanded leafroller, Choristoneura rosaceana (Harris), in laboratory and field studies (Waldstein and Reissig 2001a,b). Tebufenozide may also prove to be effective

PAGE 27

16 against other neonate lepidopterans, including cranberry fruitworm, if applied prior to peak moth flight and before large numbers of eggs are laid in the field. The primary goal of this research was to improve IPM protocols for monitoring A. vaccinii in highbush blueberries. My hypothesis was that both trap height and trap location within a planting would influence captures of male moths in pheromone-baited traps. In addition, I wanted to investigate the effects of a single application of tebufenozide shortly after peak moth flight for control of A. vaccinii. Materials and Methods Trap Height Experiments to investigate the effect of trap height on captures of male A. vaccinii moths were conducted at three sites in western Michigan from 10 May to 17 July 2001. Individual sites consisted of 2-ha blocks of V. corymbosum cultivar ‘Jersey’. There were two organic plantings (Holland and Fennville, MI) and one conventional planting (Covert, MI). Management practices in the conventional planting included the application of organophosphate insecticides (phosmet and malathion) on a calendar schedule. Blueberry bushes at each site were 2 to 2.5 m in height and spaced 1.5 m between bushes and 2.4 m between rows. I also monitored activity of male A. vaccinii moths in a 2-ha section of woodlands located adjacent to the conventional planting that contained predominantly oak, pine, and broadleaf weeds. White cardboard winged sticky traps (model # IPM101-00, Great Lakes IPM, Vestaburg, MI) baited with a single 300 mg cranberry fruitworm lure (model # IPM-CFW-L1500, Great Lakes IPM) were used to monitor A. vaccinii activity. Three treatments were evaluated based on trap height:

PAGE 28

17 At the top of the bush canopy, 15 cm below the uppermost branch (high) Centrally within the bush canopy, 60 cm below the uppermost branch (medium) At the bottom of the bush, 20 cm above ground level (low). At the woodland site, A. vaccinii populations were monitored at three distinct heights spaced further apart vertically: 4.5 m above ground level (high) 1.5 m above ground level (medium) 20 cm above ground level (low). At the woodland site, my aim was to monitor A. vaccinii activity and to document any differences in flight activity between woodland areas and the adjacent highbush blueberry planting. Treatments were arranged in a randomized complete block design with four replicates per treatment at each of the three blueberry plantings. Monitoring traps were spaced 30 m apart within rows and 35 m between blocks in the plantings. The same design and spacing protocol was adopted in the woodland area. Monitoring. Cranberry fruitworm moths were counted and removed from traps once per week at the conventional site and woodland area, and twice per week at the two organic sites where cranberry fruitworm pressure was greater. Traps and lures were replaced every three weeks. Trap Geographical Location Experiments to investigate the effects of geographical location of pheromone-baited traps on captures of A. vaccinii were conducted at an organic (Holland, MI), and a conventional (Covert, MI) planting from 10 May to 17 July 2001. Each site consisted of a 2-ha block of V. corymbosum cultivar ‘Jersey.’ Four treatments were evaluated based on geographical location. Treatments were arranged relative to distance from woodlands

PAGE 29

18 with four replicates per treatment (Figure 2-1). Treatments included pheromone-baited traps placed: In trees within 1 m of the woodland boundary In blueberry bushes adjacent to woodlands, 15 m from the woodland boundary In blueberry bushes in the center of the planting, 75 m from the woodland boundary In blueberry bushes furthest away from woodlands, 150 m from the woodland boundary At both sites, traps located 150 m from the woodland boundary were adjacent to open (non-woodland) fields (Figure 2-1). Pheromone-baited traps were spaced 50 m between treatment blocks. All traps within blueberry plantings were positioned within the canopy of the bushes, 60 cm below the uppermost branch (medium height). Monitoring. Cranberry fruitworm moths were counted and removed from traps once per week at the conventional site and twice per week at the organic site where cranberry fruitworm pressure was greater. Traps and lures were replaced every three weeks. Fruit sampling. Twenty-seven blueberry clusters of 10 berries each (nine clusters from each of three replicates; 270 berries per treatment) were harvested from each treatment twice per week starting 28 June and ending 19 July 2001. Samples were taken randomly from bushes located within a 3-m radius of pheromone-baited traps corresponding to individual treatments, 15, 75, and 150 m from the woodland boundary. No fruit was collected from the woodland area due to the absence of berries along the woodland boundary. Individual cluster samples were placed into re-sealable plastic bags, then placed into a cooler and transported back to the laboratory for visual analysis and dissection. Clusters were categorized as uninfested, or exhibiting single berry infestation

PAGE 30

19 or multiple berry infestation, depending on the presence of A. vaccinii larva and/or whether the cluster exhibited no feeding (uninfested), feeding in one berry (single infestation), or feeding in more than one berry (multiple infestation). The presence of silk webbing and frass was also used to identify multiple berry infestation. Data were reported as the percentage of clusters exhibiting one of the three categories of infestation (uninfested, single infestation, or multiple infestation). In addition, I determined total infestation, which was calculated as the sum of single and multiple infestation of blueberry clusters. Effects of Tebufenozide on A. vaccinii Infestation The experiment to evaluate a single application of tebufenozide for suppressing A. vaccinii infestations was conducted in a 2-ha block of an untreated (abandoned) V. corymbosum cultivar ‘Jersey’ planting (Holland, MI). Bushes were spaced 1.5 m apart and 2.4 m between rows. My hypothesis was that one application of tebufenozide applied shortly after peak moth flight (before the majority of eggs were laid) could reduce infestations of A. vaccinii to tolerable levels. Monitoring data from previous years had indicated that peak moth flight occurred in mid-June in southwestern Michigan (Wise et al. 1999). In addition, my 2001 monitoring data from pheromone-baited traps suggested that male A. vaccinii populations peaked on 18 June in the test planting (Figure 2-2). Therefore, an application of tebufenozide was made on 26 June, one week after peak moth flight, which was estimated to coincide with peak egg hatch. Two treatments were evaluated: An application of tebufenozide (Confirm 2F, Rohm and Haas, Philadelphia, PA), applied at a rate of 1.2 L/ha in 233.6 L of water on 26 June 2001 An untreated control.

PAGE 31

20 A surfactant, Latron B1956 (Dow Agrosciences, Indianapolis, IN), was added to the tank-mix of the tebufenozide treatment. Tebufenozide was applied using a Solo backpack sprayer (Newport News, VA) with a delivery pressure of 448.2 kPa. Treatments were applied in a randomized complete block design with four replicates. A 20-m buffer zone was allocated between treatments to adjust for potential spray drift. Fruit sampling. Thirty-six blueberry clusters of 10 berries each (nine clusters per replicate; 360 berries per treatment) were harvested from each treatment twice per week starting 5 July and ending 12 July 2001. Samples were transported back to the laboratory for visual analysis and dissection as previously described. Data were reported as the percentage of clusters exhibiting one of the three categories of infestation (uninfested, single infestation, or multiple infestation). Again, I determined total infestation, which was calculated as the sum of single and multiple infestation of blueberry clusters. Data Analysis Data collected from the trap height and geographical location studies were square-root transformed to account for deviations from normality and then subjected to an analysis of variance (ANOVA) followed by mean separation using least significant difference (LSD) tests (SAS Institute 2001). Data were also subjected to repeated measures analysis (using PROC MIXED, SAS Institute 2001) to examine the interaction effect between treatment and time (sampling date) throughout the monitoring period. Percentage data collected from the two fruit sampling analyses (geographical location study and tebufenozide evaluation) were arcsine-square-root transformed to account for deviations from normality and then subjected to an ANOVA followed by mean separation using LSD tests (SAS Institute 2001) for each category (uninfested,

PAGE 32

21 single, and multiple infestation). Percentages were considered significant when P values were 0.05. The untransformed percentages and standard errors are presented in tables. Results Trap Height At the Holland organic site, I found that traps placed 15 and 60 cm below the uppermost branch were significantly (F = 4.2; df = 2,6; P = 0.05) more effective in detecting male A. vaccinii moths compared with traps placed 20 cm above the ground (Table 2-1). Specifically, traps placed 15 and 60 cm from the uppermost branch captured 1.6 times as many A. vaccinii compared with traps placed 20 cm above the ground (Table 2-1). Similar observations were recorded at the Fennville organic site. Traps placed 15 and 60 cm from the uppermost branch captured significantly (F = 25.8; df = 2,6; P < 0.01) more male moths compared with traps placed 20 cm above ground level (Table 2-1). Traps placed 15 and 60 cm from the uppermost branch captured 6.4 and 5.6 times as many A. vaccinii compared with traps placed 20 cm above the ground, respectively (Table 2-1). At the conventional site, male moth captures for each of the three trap heights were not significantly different (Table 2-1). The interaction effect between treatment (trap height) and time (sample date) was significant (F = 4.45; df = 14,63; P < 0.01) only at the Fennville organic site. Treatment separation was consistent when moth captures were high. Specifically, traps placed 15 and 60 cm from the uppermost branch captured significantly more A. vaccinii moths compared with traps placed 20 cm above the ground. Treatments were not significantly different when moth populations were low, prior to 4 June and after 9 July. In the woodland area where I monitored A. vaccinii activity, moth distributions among treatments followed a trend similar to those observed at the two organic sites.

PAGE 33

22 Pheromone-baited traps located 4.5 m above ground level captured significantly (F = 75.0; df = 2,6; P < 0.01) more male moths compared with traps 1.5 m and 20 cm above the ground. In addition, traps located 1.5 m above the ground were significantly (F = 75.0; df = 2,6; P < 0.01) more effective in detecting male moths compared with traps located 20 cm above ground level. Overall, traps located 4.5 and 1.5 m above ground level attracted 6.3 and 2.5 times, respectively, more male moths compared with traps positioned 20 cm above ground level. The interaction effect between treatment and time was significant (F = 3.2; df = 22,99; P < 0.01) in the woodland area. Although trap height did not affect male moth capture prior to 14 June and after 10 July, traps placed 4.5 m above ground level consistently captured significantly (F = 18.7; df = 2,6; P < 0.01) more moths compared with traps located 20 cm above ground level during the period corresponding to peak moth flight. Trap Geographical Location Trap captures of male A. vaccinii moths were approximately three times higher at the organic site compared with the conventional site (Table 2-2). In the organic planting, I recorded significantly (F = 7.0; df = 3,9; P = 0.01) higher male moth captures on pheromone-baited traps located within 1 m of the woodland boundary compared with traps located 15, 75, and 150 m within the plantings (Table 2-2). Overall, traps located along the woodland boundary captured 1.8 times as many A. vaccinii compared with all other trap locations (within the planting) I evaluated. A similar trend was observed at the conventional site. Again, traps located within the woodland boundary captured significantly (F = 11.5; df = 3,9; P 0.01) more male A. vaccinii moths compared with

PAGE 34

23 traps located 15 and 75 m within the planting (Table 2-2). However, A. vaccinii captures in traps located at the far edge of the planting (150 m away) did not differ significantly from captures in traps located along the woodland boundary. Interaction effects between treatment (geographic location of traps) and time were significant at both the organic (F = 2.5; df = 24,95; P < 0.01) and the conventional (F = 1.9; df = 24,95; P < 0.01) sites. As in my trap height study, treatment differences were observed only during the sampling dates corresponding peak moth flight. At both sites, male A. vaccinii moths were captured more frequently in pheromone-baited traps located along the woodland boundary compared with traps located within the planting during peak flight. Treatment separation for traps located within the planting varied throughout the same period, although traps located 150 m from the woodland boundary frequently captured more moths compared with traps located at 15 and 75 m. Fruit sampling. Data from my fruit infestation analysis followed a similar trend as my trap catches from the organic planting. The percentage of blueberry clusters with single infestation (larval feeding in only one berry of the cluster) within 15 m of the woodland boundary was significantly (F = 5.5; df = 2,4; P = 0.05) higher compared with clusters located 150 m from the woodland boundary (Table 2-3). There were no significant differences among clusters exhibiting multiple berry infestation, regardless of distance relative to the woodland boundary (Table 2-3). The total infestation percentage of blueberry clusters harvested from bushes located 15 and 75 m within the planting was significantly (F = 32.0; df = 2,4; P 0.01) higher compared with clusters harvested from bushes 150 m away from the woodland boundary (Table 2-3).

PAGE 35

24 Effects of Tebufenozide on A. vaccinii Infestation The incidence of single berry infestation was not significantly different among tebufenozide-treated and untreated blueberry clusters. However, multiple berry infestation was significantly (F = 26.3; df = 1, 3; P 0.01) reduced for the tebufenozide-treated clusters compared with the control (Table 2-4). Overall, untreated blueberry clusters were 11 times more likely to exhibit multiple berry infestation compared with tebufenozide-treated clusters (Table 2-4). Total percentage infestation of blueberry clusters was significantly (F = 106.8; df = 1, 3; P 0.01) reduced by a single treatment of tebufenozide applied one week after peak moth flight compared with the control (Table 2-4). There was a two-fold reduction in total berry infestation between tebufenozide-treated berries and untreated berries. Discussion My study indicated that the most efficient trap height for monitoring male A. vaccinii activity in highbush blueberry plantings ranges between 15 and 60 cm below the uppermost branch. Data supporting this finding were especially pronounced at the Fennville organic site, where the relative separation between treatments was greater because bushes in that planting were slightly taller than bushes at the Holland organic and conventional (Covert) plantings. Traps located close (20 cm) to ground level did not appear as attractive to male moths as traps located higher within the blueberry bush canopy. These observations were particularly clear during the interval corresponding to peak moth flight (mid-June). A number of factors may explain the elevated moth captures in traps located higher within the bush canopy: pheromone concentration and release rate, microclimate, and trap placement, which may affect the efficacy of

PAGE 36

25 monitoring programs that employ sex pheromone technologies (Shorey 1973). AliNiazee (1983) found that male filbertworm, Melissopus latiferreanus Walshingham, moths respond preferentially to pheromone-baited traps located within or slightly above the canopy of cultivated filbert trees compared with lower traps, regardless of absolute height. The observed differences were attributed to little activity below the tree canopies and the mate-seeking advantages associated with localizing in the upper canopy where females were found. Ahmad (1987) suggested that reduced catches of the almond moth, Cadra cautella (Walker), in pheromone-baited traps located close to ground level could have been due to inadequate dispersal of the pheromone plume. These theories may help to explain the observations I recorded in my experiments. My hypothesis that height influences preference of male moths to pheromone-baited traps was further supported by data collected in the woodland monitoring area, where captures of A. vaccinii were significantly different among all three trap heights. I presume that the use of broad-spectrum insecticides and the resulting lower moth population at the conventional planting may have prevented clear treatment separation for captures of A. vaccinii in pheromone-baited traps at the various heights. Perhaps moths flying within the canopy of bushes had a higher risk of encountering pesticide residues than moths flying below the canopy. High trap captures below the canopy in the conventional planting may reflect flight behavior to avoid these residues. Regarding the geographic location of traps, my results suggest that traps positioned at the edge of woodlands adjacent to plantings may serve as a prime location for monitoring male A. vaccinii activity. The high trap captures for woodland treatments in my experiments indicate that male moths may be moving between plantings and adjacent

PAGE 37

26 woodlands, perhaps to increase mating encounters, to seek refuge, or to escape daytime predators. In my geographical location study, traps located along the woodland periphery and in blueberry bushes at the edge of the planting adjacent to open fields (150 m away from woodland boundary) captured more male moths than traps located centrally (75 m) within the planting. These results imply that high populations of male A. vaccinii moths aggregate near planting boundaries. Mallampalli and Isaacs (2002) recorded significantly more A. vaccinii eggs in blueberries located adjacent to woodland boundaries and hypothesized that woodlands may serve as a reservoir for individuals that affect commercial plantings because of the wild hosts they may contain. The high incidence of moths captured in traps located adjacent to open fields may be the result of pheromone dispersal patterns, where the plume may have been carried over longer distances in the field than within plantings. Regardless of these hypotheses, my recommendation to growers is to monitor for A. vaccinii moths both within the interior of plantings as well as at the periphery (woodlands and open fields). Inherent limitations exist when using sex pheromone traps for monitoring A. vaccinii activity, primarily because only male moths can be monitored. Female moths may or may not exhibit similar movements between plantings and woodlands as males, although indices such as infestation can be used to predict their behavior and preferences within a planting. The infestation trends I observed at the organic site paralleled my monitoring trap data, indicating that females prefer to oviposit in blueberries adjacent to wooded areas. My evaluation of the insect growth regulator tebufenozide indicates that a single application sprayed one week after peak flight can significantly reduce total infestation of

PAGE 38

27 highbush blueberries by A. vaccinii. Whether the reduction obtained from one application of tebufenozide is sufficient to market injury-free blueberries in a commercial planting is a question that warrants further research. My results from a highly infested abandoned planting indicate that more than one application of tebufenozide would be necessary to market blueberries free of A. vaccinii larvae in a high-pressure system. Although a single application of tebufenozide reduced the incidence of multiple berry infestation in my study, single berry infestation was not reduced. The observed differences among total infestation rates for blueberry clusters taken from untreated versus tebufenozide-treated plots were largely due to incidence of multiple rather than single berry infestation. Cadogan et al. (2002) postulated that the presence of tebufenozide residues on egg-laying surfaces might be a primary factor in deterring oviposition by spruce budworm, Choristoneura fumiferana (Clemens). However, it was unclear at what concentration tebufenozide must be present on a substrate to exert its effectiveness. Regarding my study, the presence of tebufenozide residues on fruit may have been responsible for the decrease in A. vaccinii infestation in treated blueberry clusters. Highbush blueberry growers may need to consider additional applications of tebufenozide for adequate control of A. vaccinii in their plantings. Furthermore, the inclusion of a surfactant (such as Latron B-1956) may increase the efficacy of tebufenozide. Waldstein et al. (2001b) suggested that the addition of a surfactant to tebufenozide might decrease adverse precipitation effects. The usefulness of reduced-risk insecticides such as tebufenozide may depend on their time of application. Legaspi et al. (1999) suggested that efficacy of tebufenozide

PAGE 39

28 against the Mexican rice borer, Eoreuma loftini (Dyar), may be improved if timed during a window of susceptibility based on pest lifecycle or plant phenology. In the future, blueberry IPM programs may be able to utilize information gained from effective monitoring programs to predict the optimal time(s) to apply tebufenozide or other reduced-risk insecticides to reduce A. vaccinii infestation.

PAGE 40

29 Table 2-1. Captures of male A. vaccinii moths in pheromone-baited traps placed at different heights with respect to blueberry bush canopy. Treatment Mean SEM* of A. vaccinii Trap height Organic (Holland, MI) Organic (Fennville, MI) Conventional (Covert, MI) 15 cm from uppermost branch 77.3 7.6a 78.3 14.5a 46.5 5.5 60 cm from uppermost branch 78.3 8.6a 68.8 3.9a 47.0 16.3 20 cm above ground level 47.5 10.4b 12.3 5.5b 45.5 21.8 *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted once per week at the conventional site and twice per week at the two organic sites from 10 May to 17 July 2001. Treatment means are the sum of data collected over all sampling dates.

PAGE 41

30 Table 2-2. Captures of male A. vaccinii moths in pheromone-baited traps placed at different locations within blueberry plantings with respect to adjacent woodlands. Treatment Mean SEM* of A. vaccinii Trap position Organic (Holland, MI) Conventional (Covert, MI) Woodland boundary 233.8 42.2a 89.0 22.1a 15 m from woodland boundary 132.0 9.2b 19.5 2.2b 75 m from woodland boundary 73.3 0.6b 31.5 6.6b 150 m from woodland boundary 124.8 32.5b 71.7 7.9a *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted once per week at the conventional site and twice per week at the organic site from 10 May to 17 July 2001. Treatment means are the sum of data collected over all sampling dates.

PAGE 42

31 Table 2-3. Percentage of blueberry clusters infested by A. vaccinii at different geographic locations within an organic planting in Holland, MI. % Clusters infested by A. vaccinii SEM* Treatment Level of infestation Distance from woodland boundary Uninfested Single berry infestation Multiple berry infestation Total infestation 15 m 35.7 3.8b 24.3 3.8a 40.0 7.5 64.3 3.8a 75 m 42.7 1.3b 22.0 1.2ab 35.0 1.0 57.3 1.3a 150 m 62.7 0.7a 14.0 2.1b 23.0 2.7 37.3 0.7b *Percentages within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on arcsine-square-root transformed data, but percentages shown reflect untransformed data. Blueberry clusters were collected twice per week from 28 June to 19 July 2001 and sampled for A. vaccinii infestation.

PAGE 43

32 Table 2-4. Percentage of blueberry clusters infested by A. vaccinii after a single application of tebufenozide within an abandoned planting in Holland, MI. % Clusters infested by A. vaccinii SEM* Level of infestation Treatment Uninfested Single berry infestation Multiple berry infestation Total infestation Tebufenozide (Confirm 2F) 64.5 2.4a 31.8 3.5 3.8 1.4b 35.5 2.4b Untreated control 31.0 5.0b 26.5 1.4 42.3 5.2a 69.0 5.0a *Percentages within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on arcsine-square-root transformed data, but percentages shown reflect untransformed data. Blueberry clusters were collected twice per week from 5 to 12 July 2001 and sampled for A. vaccinii infestation.

PAGE 44

33 = Cranberry Fruitworm Pheromone-Baited Trap Woodlands Open Field 15 m (Trt 2) 75 m (Trt 3) 150 m(Trt 4) 1 m (Trt 1) 50 m Figure 2-1. Spatial orientation of pheromone-baited traps for monitoring male A. vaccinii moths in the trap geographical location study (Holland and Covert, MI).

PAGE 45

34 051015202513-Jun18-Jun23-Jun28-Jun3-Jul8-Jul13-Jul18-JulDateMean # A. vaccinii in pheromone-baited traps Figure 2-2. Captures of male A. vaccinii moths in pheromone-baited traps in Holland, MI (2001). Trap catch data were used to determine a spray date for a single application of tebufenozide for managing A. vaccinii. The application was made one week after peak moth flight on 26 June, which was estimated to coincide with peak egg hatch.

PAGE 46

CHAPTER 3 TECHNIQUES FOR MONITORING BLUEBERRY GALL MIDGE (DIPTERA: CECIDOMYIIDAE) IN RABBITEYE AND SOUTHERN HIGHBUSH BLUEBERRIES The blueberry gall midge, Dasineura oxycoccana (Johnson), is a recently discovered pest of rabbiteye (Vaccinium ashei Reade) and southern highbush (Vaccinium corymbosum L. x V. darrowi Camp) blueberries in the southeastern United States (Lyrene and Payne 1992). Female flies oviposit several eggs into susceptible floral and leaf buds, and during heavy infestations, more than one female may oviposit into a single bud (Bosio et al. 1998). Larvae feed on plant juices, which ultimately kill developing buds (Mahr and Kachadoorian 1990). Unmanaged D. oxycoccana infestations can destroy up to 80% of floral buds in susceptible rabbiteye cultivars (Lyrene and Payne 1992, 1995). Until recently, the presence of D. oxycoccana in Florida blueberry plantings was misdiagnosed as freeze damage (Lyrene and Payne 1995). This discrepancy has complicated accurate assessment of floral and vegetative bud loss in blueberry plantings throughout Florida and much of the southeastern United States. Lyrene and Payne (1992) were the first to identify D. oxycoccana affecting floral buds within rabbiteye blueberries. Infestations have since been identified in blueberry plantings in Italy and other European nations (Bosio et al. 1998). Currently, the Florida blueberry industry is undergoing a conversion from long-established rabbiteye plantings to commercial production of southern highbush cultivars, with acreage of southern highbush increasing 23% from 1989 to 2000 (Williamson et al. 2000). The trend toward commercial production of southern highbush 35

PAGE 47

36 plantings in Florida is a response to an increased market for early-season blueberries, which have a higher market value than later-ripening blueberries from primary blueberry-producing states (Williamson et al. 2000). Despite this trend toward increasing southern highbush acreage, blueberry production in Florida may become limited by an expected increase in insect problems (Mizell and Johnson 2001). In a recent crop profile for blueberries in the southeastern United States, D. oxycoccana was listed as a major pest warranting attention, particularly with regard to monitoring and evaluation of alternative chemical management protocols (NeSmith 1999). According to Gagn (1989), adult D. oxycoccana live only long enough to mate and lay eggs, approximately two to three days. Blueberry gall midge is known to have overlapping generations throughout its described range, which includes the southeast as well as cranberry and blueberry producing areas in the north (Liburd and Finn 2002). Mahr and Kachadoorian (1990) reported that D. oxycoccana completed up to five generations in Wisconsin cranberry bogs, although later work by Cockfield and Mahr (1994) suggested that distinct broods are not easy to distinguish. In Mississippi, D. oxycoccana are capable of completing up to 11 generations per year (Sampson et al. 2002). Reliable monitoring techniques to detect and quantify populations of D. oxycoccana in blueberry plantings have not been thoroughly evaluated. Currently, yellow sticky boards are used within some plantings to monitor the presence of adult midges and the abundance of their natural enemies (Bosio et al. 1998; Sampson pers. comm.). Other colors of sticky boards have not been evaluated in the United States, despite knowledge that dipteran insects can be monitored effectively using various

PAGE 48

37 colored traps (Vernon and Broatch 1996; Liburd et al. 1998). Other frugivorous insects such as flower thrips, Frankliniella spp. (Thysanoptera: Thripidae), respond to blue and white (Cho et al. 1995; Childers and Brecht 1996). Since flower thrips are another key pest in Florida blueberry plantings, and because they affect blueberry production during approximately the same time interval as D. oxycoccana (Finn and Liburd, unpublished data), the development of a trapping technique that may simultaneously monitor the presence of both pests is desirable. My objectives were to determine if color influences captures of D. oxycoccana adults on unbaited sticky traps in the field and to evaluate several sampling strategies for detecting various life stages of D. oxycoccana. Finally, I wanted to compare rabbiteye and southern highbush floral and leaf buds for the presence of D. oxycoccana eggs and larvae. Materials and Methods All experiments were conducted in 2-ha blocks of rabbiteye or southern highbush plantings in north-central Florida. Bushes in the rabbiteye planting were unmanaged for at least three years prior to the start of studies and were ~ 2 to 3 m tall in 2002 and ~ 1.5 to 2 m tall in 2003 (after pruning in July 2002). The planting was located in Windsor, FL and consisted of cvs. ‘Beckyblue,’ ‘Bonita,’ and ‘Climax.’ Bush spacing was 1.5 m apart and 2.4 m between rows. Bushes in the southern highbush plantings were managed commercially and were ~ 1.5 to 2 m tall at each of two distinct sites. One southern highbush planting was located in Inverness, FL, and contained cvs. ‘Misty,’ ‘Jewel,’ and ‘Sharpblue.’ Bush spacing was 1.5 m apart and 2.4 m between rows. A second southern highbush planting was located in Windsor, FL, within 500 m of my rabbiteye planting. Bushes in the second southern highbush planting were clones of similar parentage (no

PAGE 49

38 cultivar identification) and spaced 1.5 m apart and 1.5 m between rows (high density bed). Unbaited Colored Traps Experiments to evaluate the effect of color for monitoring D. oxycoccana adults using unbaited sticky traps were conducted in rabbiteye and southern highbush plantings in Windsor, and Inverness, FL, respectively. In 2002, D. oxycoccana adults were monitored once per week for a 6-wk period, from 5 March to 12 April at the rabbiteye planting and 31 January to 6 March at the southern highbush planting. In 2003, monitoring was conducted only in the rabbiteye planting since data from the previous year indicated that D. oxycoccana populations were too low in the southern highbush planting to warrant sampling again. Monitoring was conducted twice per week from 6 to 21 March (five sampling dates in total). The monitoring period for each planting included the seven stages of floral bud development as outlined by Spiers (1978). Various colors of commercially produced rectangular unbaited sticky board traps (treated area 394 cm 2 , Great Lakes IPM, Vestaburg, MI) were used to monitor D. oxycoccana adults in commercial blueberry plantings. Four treatments were evaluated: Standard pantone yellow Safety white Walnut husk green Thrips blue Experimental design was randomized complete block with 4 replicates per treatment in each planting. Traps were hung within the canopy of bushes in a vertical position and spaced 10 m apart and 15 m between blocks. Traps were re-randomized and replaced once per week in 2002 and twice per week in 2003. Traps that were removed from the

PAGE 50

39 field were covered with a layer of plastic wrap and transported back to the laboratory for a two-stage visual analysis. In the initial analysis, traps were screened for D. oxycoccana adults using a bench-top illuminated magnifier (3D, Cole-Palmer, Vernon Hills, IL). All specimens with morphological characteristics resembling D. oxycoccana were encircled on the plastic and subjected to a second analysis. Positive identification of D. oxycoccana was conducted using a 10-X dissecting microscope. Dr. Raymond Gagn at USDA-ARS, Beltsville, MD, and Dr. Blair Sampson at USDA-ARS, Poplarville, MS, confirmed the identification of specimens. The most effective trap color was determined based on the highest mean captures of D. oxycoccana adults throughout the season. Monitoring Techniques Experiments to evaluate different sampling techniques for detecting various life stages of D. oxycoccana were conducted in both rabbiteye and southern highbush plantings in Windsor, FL. Three monitoring techniques were evaluated for their effectiveness in detecting D. oxycoccana: Use of unbaited pantone yellow sticky boards Collection of bud samples for emergence Collection of bud samples for dissection In techniques two and three, 140 floral buds and 140 vegetative buds were collected for analysis from each planting. Experimental design was randomized complete block with four replicates per treatment in each planting. Treatments were blocked by cultivar in the rabbiteye planting and by clone (no cultivar identification) in the southern highbush planting. Experimental plots consisted of 15 bushes and were designated for the application of each treatment.

PAGE 51

40 In both rabbiteye and southern highbush plantings, buds were collected over three dates when floral and leaf buds were presumed to be most susceptible to D. oxycoccana infestation. This range was estimated to be stage 2 to 4 of floral bud development (Spiers 1978) and stage 2 to 4 of leaf bud development (NeSmith et al. 1998). In the rabbiteye planting, floral bud and leaf bud development occurred simultaneously. In 2002, sampling was conducted on 8, 15, and 19 March, and in 2003, sampling was conducted on 3, 6, and 10 March in the rabbiteye planting. In the southern highbush planting, stages 2 to 4 of floral bud development preceded vegetative bud development by approximately 3 wk in both years. In 2002, floral buds were collected on 7, 11, and 18 February, and leaf buds were collected on 5, 11, and 18 March. In 2003, floral buds were collected on 3, 7, and 13 February, and leaf buds were collected on 24 February and 3 and 6 March. Monitoring boards were sampled using the two-stage visual analysis previously described. One hundred forty buds per treatment (35 buds from each replicate) were collected at random for larval emergence and dissection techniques. Floral and vegetative buds were collected separately and placed into re-sealable transparent containers for transportation back to the laboratory. Buds collected for monitoring larval emergence were transferred to 15-cm plastic petri dishes containing moistened filter paper and held at 27 o C under 14L:10D conditions for 10 days. The total number of emergent larvae and adults was recorded. The remaining 140 buds were dissected under a dissecting microscope; all life stages of D. oxycoccana were recorded. Treatments were rotated among experimental plots on a weekly basis. The most efficient sampling technique was determined based on ability to detect various life stages of D. oxycoccana.

PAGE 52

41 Bud Type Comparison Data collected from the dissection technique were used to compare the infestation rates of floral buds and leaf buds collected from rabbiteye and southern highbush plantings. Due to variations in plant phenology, the 4 bud types (rabbiteye floral, rabbiteye leaf, southern highbush floral, and southern highbush leaf) were collected over different time periods. Nonetheless, overall comparisons were made to determine which buds were most susceptible to D. oxycoccana infestation. In 2002 and 2003, bud types were compared for the presence of eggs and larvae. In addition, larvae were also categorized by instar in 2003. Data Analysis Data from all three studies were square-root transformed to account for deviations from normality and then subjected to an analysis of variance (ANOVA) followed by mean separation using least significant difference (LSD) tests (SAS Institute 2001). Data were also subjected to repeated measures analysis (using PROC MIXED, SAS Institute 2001) to examine the interaction effect between treatment and time (sampling date) throughout the duration of each experiment. When significant interaction effects were noted, further analysis was conducted to determine the order of treatment efficacy for each sampling date. Means were considered significant when P values were 0.05. The untransformed means and standard errors are presented in tables. Results Unbaited Colored Traps In 2002, there were no significant differences in captures of D. oxycoccana adults on unbaited sticky boards of various colors in either the rabbiteye or the southern highbush plantings. Overall, the number of D. oxycoccana adults caught on traps

PAGE 53

42 averaged less than 1 fly per trap per sampling date in 2002. In 2003, trap captures were higher, averaging 8 to 9 flies per trap per sampling date. Still, I found no differences in captures of adults on the colored traps. Monitoring Techniques 2002. The emergence technique was significantly (F = 53.5; df = 2, 6; P < 0.01) more effective in detecting D. oxycoccana adults in rabbiteye floral buds compared with unbaited yellow sticky boards or dissection techniques (Table 3-1). No adults were detected in my dissection technique (Table 3-1). Emergence and dissection techniques performed equally well for detecting D. oxycoccana larvae in rabbiteye floral buds, and both techniques detected significantly (F = 57.1; df = 2, 6; P < 0.01) more larvae compared with unbaited yellow sticky boards (Table 3-1). There were no significant differences among sampling techniques for detecting adults in rabbiteye leaf buds (Table 3-1). Among all of the plant material I analyzed, the only sampling technique capable of detecting D. oxycoccana eggs was dissection (Tables 3-1 3-4). The number of D. oxycoccana adults detected using my three techniques did not differ in southern highbush floral and leaf buds (Table 3-2). However, the dissection technique was significantly (F = 38.0; df = 2, 6; P < 0.01) more effective in detecting D. oxycoccana larvae in southern highbush floral buds compared with unbaited yellow sticky boards or emergence techniques (Table 3-2). Six times more D. oxycoccana larvae were detected in leaf buds by dissection than by emergence techniques (Table 3-2). Again, D. oxycoccana eggs were detected only by the dissection technique. 2003. Infestation of flower buds by D. oxycoccana in the rabbiteye planting was slightly higher in 2003 than in 2002. As in 2002, emergence techniques were significantly (F = 101.0; df = 2, 6; P < 0.01) more effective in detecting D. oxycoccana

PAGE 54

43 adults in floral buds compared with either yellow boards or dissection (Table 3-3). However, in my analysis of leaf buds, yellow sticky boards were significantly (F = 28.2; df = 2, 6; P < 0.01) more effective in detecting adults compared with emergence or dissection techniques (Table 3-3). Importantly, there were significant (F = 8.6; df = 6, 18; P < 0.01) interaction effects between treatment and time (sampling date) for the detection of D. oxycoccana adults in rabbiteye floral buds. That is, treatment efficacy varied over the three sampling dates. Specifically, yellow sticky boards were ineffective for detecting D. oxycoccana adults early in the season (3 and 7 March), but were significantly (F = 47.9; df = 2, 6; P < 0.01) more effective than dissection techniques later in the season (10 March). Again, significant (F = 6.6; df = 6, 18; P < 0.01) interaction effects between treatment and sampling date were observed for rabbiteye leaf buds, where yellow sticky boards were ineffective for detecting adults on 3 March, but were more effective than dissection and emergence techniques on 6 and 10 March. Similar numbers of D. oxycoccana were detected using emergence and dissection techniques in rabbiteye floral and leaf buds (Table 3-3). As in 2002, some larvae were detected on sticky boards, although their incidence was sporadic. Again, eggs were only detected by the dissection technique. In my southern highbush planting, emergence techniques were significantly (F = 169.8; df = 2, 6; P < 0.01) more effective in detecting D. oxycoccana adults in floral and leaf buds compared with yellow sticky boards or dissection (Table 3-4). Similarly, my emergence technique was significantly (F = 31.4; df = 2, 6; P < 0.01) more effective in detecting D. oxycoccana larvae in my southern highbush floral buds compared with the dissection technique (Table 3-4). However, sampling techniques did not differ for

PAGE 55

44 detecting larvae in southern highbush leaf buds. In 2003, I did not record any D. oxycoccana eggs in southern highbush floral buds. By contrast, a substantial number of D. oxycoccana eggs were detected in southern highbush leaf buds. On several occasions I noted the presence of an adult parasitic wasp (Hymenoptera: Platygastridae) in buds collected for my emergence and dissection treatments. Although I never witnessed the emergence of this wasp from D. oxycoccana larvae, its role as a potential parasitoid was recognized. In addition, D. oxycoccana adults reared in the laboratory for this study indicated a sex ratio of approximately 60 70% females. Bud Type Comparison In 2002, there were significantly (F = 9.9; df = 3, 9; P < 0.01) more D. oxycoccana eggs in southern highbush leaf buds compared with any other bud types (Table 3-5). Eggs were found least often in southern highbush floral buds in both 2002 and 2003, although the incidence of eggs in southern highbush floral buds was similar to that of rabbiteye floral and leaf buds in 2002. Similar numbers of D. oxycoccana eggs were recorded in rabbiteye floral and leaf buds in both years (Tables 3-5, 3-6). I recorded significantly (F = 39.7; df = 3, 9; P < 0.01) more D. oxycoccana larvae in rabbiteye floral buds compared with any other bud type in 2002. In 2003, the highest number of D. oxycoccana eggs was recorded in southern highbush leaf buds (Table 3-6). Significantly (F = 18.2; df = 3, 9; P < 0.01) more first instar larvae were detected in rabbiteye floral buds compared with any other bud type, followed by southern highbush floral buds, which had significantly (F = 18.2; df = 3, 9; P < 0.01) more first instar larvae compared with either species of leaf buds (Table 3-6). Second instar larvae were significantly (F = 485.0; df = 3, 9; P < 0.01) more abundant in

PAGE 56

45 rabbiteye floral buds compared with any other bud type (Table 3-6). Similar results were recorded for third instar larvae. Discussion My studies indicated that D. oxycoccana adults are not easily monitored using any one of various colors of unbaited sticky traps in the field. The relative increase in mean trap counts from 2002 to 2003 was likely due to higher pressure of D. oxycoccana in the field and also to the better condition of board-preserved specimens, which was facilitated by replacing traps more frequently. Overall, identification of D. oxycoccana was difficult due to several key factors: small size (~ 2 mm), the sticky surface of the traps, specimen degradation in the field, poor preservation of distinguishing characteristics, and the presence of other flies within the family Cecidomyiidae. The sticky surface on the monitoring boards often damaged the integrity of key morphological features, including wing venation, microtrichia, and antennal segmentation, which are useful for accurately identifying D. oxycoccana adults. Identification was marginally enhanced on the white boards, largely because of the color contrast with the insect abdomen, which is often bright orange in D. oxycoccana females. Currently, unbaited sticky traps appear to be an ineffective tactic for monitoring D. oxycoccana populations. However, the use of sticky traps for monitoring D. oxycoccana adults may be improved by incorporating a luring device, either a sex pheromone or a host-volatile compound, or by changing trap height or type. A better understanding of olfactory stimuli and responses in D. oxycoccana may compliment the use of visual stimuli in future monitoring efforts. In my sampling techniques study, I found that emergence techniques generally performed better than unbaited yellow sticky traps or dissection techniques for detecting D. oxycoccana adults. Emergence and dissection techniques performed equally well for

PAGE 57

46 detecting D. oxycoccana larvae. The presence of larvae on sticky boards was likely due to wind disturbances as the larvae were dropping from buds to pupate in the soil. In contrast, eggs were only detected by carefully dissecting infested buds. Although bud dissection is time consuming, the ability to detect D. oxycoccana eggs is important because this information could be used by blueberry growers to make insecticide applications either before eggs hatch or before additional females begin laying eggs. Ultimately, it may be possible to manage D. oxycoccana infestations effectively using a limited number of properly timed sprays. The phenological differences between rabbiteye and southern highbush blueberry plants may account for differences in infestation by D. oxycoccana. In general, flower bud development of most southern highbush cultivars occurs before leaf bud development and before populations of D. oxycoccana peak in the field. Perhaps this phenological difference is the reason why southern highbush floral buds had the fewest number of D. oxycoccana eggs in both years that I sampled. Unfortunately, climatic conditions vary greatly from year-to-year during the spring growing season in Florida, and susceptible stages of floral bud development may coincide with D. oxycoccana pressure. For instance, a mild winter may allow D. oxycoccana populations to build up early in the blueberry production season, whereas a hard freeze in February or March may kill any D. oxycoccana that are already developing inside infested buds. Overall, I recorded the highest incidence of D. oxycoccana eggs in southern highbush leaf buds, both in 2002 and 2003. However, larval infestation of southern highbush leaf buds never exceeded that of the other bud types I evaluated. This phenomenon may indicate that survivorship of D. oxycoccana infesting southern

PAGE 58

47 highbush cultivars may be reduced compared with rabbiteye cultivars. The reason why southern highbush flower buds contained more eggs yet fewer larvae is unknown, but it is possible that the structure or the nutritional value of rabbiteye buds is more conducive to the development of D. oxycoccana than southern highbush buds. The significantly higher number of D. oxycoccana larvae in rabbiteye floral buds compared with other bud types further supports this idea. However, laboratory studies must be conducted to prove or disprove these hypotheses. The number of eggs in the various types of buds I compared may lend insight regarding the oviposition preference by D. oxycoccana females. One factor that may vary from one bud type to another is surface lipid composition, which may influence plant examination and egg-laying behavior. Eigenbrode and Espelie (1995) discussed the variation in epicuticular waxes of plants within and among species, indicating that their presence may not only prevent plant dehydration, but also mediate interactions between plants and insects. In another cecidomyiid, the Hessian fly, Mayetiola destructor (Say), laboratory studies showed that the extractable surface lipids of wheat increase the number of eggs laid by females compared with chloroform controls (Harris and Rose 1990). Comparisons between rabbiteye and southern highbush surface lipid composition, as well as various cultivars within these species, may compliment future laboratory studies to determine oviposition preference by D. oxycoccana females. Overall, my studies demonstrated that D. oxycoccana adults show no preference to various colors of sticky boards in the field. The study demonstrated that the emergence technique was most effective for detecting D. oxycoccana adults, while emergence and dissection techniques were equally effective for detecting D. oxycoccana larvae.

PAGE 59

48 Dasineura oxycoccana eggs were detected by my dissection technique, which may hold implications for timely management if growers are able to make insecticide applications before eggs hatch or before more females begin laying eggs.

PAGE 60

49 Table 3-1. Comparison of techniques for monitoring D. oxycoccana in rabbiteye blueberries in Windsor, FL (2002). Mean SEM* D. oxycoccana Life stage detected Sampling technique Adults Larvae Eggs Floral buds Yellow sticky boards** 4.0 2.0b 1.3 0.8b — Emergence 72.8 12.2a 104.3 17.5a — Dissection 0.0 0.0b 135.5 32.4a 17.3 6.3 Leaf buds Yellow sticky boards** 4.0 2.0 1.3 0.8b — Emergence 2.3 1.3 3.5 1.9ab — Dissection 0.0 0.0 12.0 6.4a 13.3 2.0 *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted on 8, 15, and 19 March for both floral and leaf buds. Means given are the total number of D. oxycoccana detected by each technique over all three sampling dates. **Data for yellow sticky boards were identical for floral and leaf bud comparisons since sampling was conducted on the same dates for these bud types.

PAGE 61

50 Table 3-2. Comparison of techniques for monitoring D. oxycoccana in southern highbush blueberries in Windsor, FL (2002). Mean SEM* D. oxycoccana Life stage detected Sampling technique Adults Larvae Eggs Floral buds Yellow sticky boards 1.0 0.7 0.0 0.0b — Emergence 1.4 0.9 2.0 1.2b — Dissection 0.0 0.0 16.8 4.7a 7.3 4.6 Leaf buds Yellow sticky boards 0.8 0.5 0.0 0.0b — Emergence 0.5 0.3 0.8 0.5ab — Dissection 0.0 0.0 5.0 2.6a 56.0 11.3 *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted on 7, 11, and 18 February for floral buds and 5, 11, and 18 March for leaf buds. Means given are the total number of D. oxycoccana detected by each technique over all three sampling dates.

PAGE 62

51 Table 3-3. Comparison of techniques for monitoring D. oxycoccana in rabbiteye blueberries in Windsor, FL (2003). Mean SEM* D. oxycoccana Life stage detected Sampling technique Adults Larvae Eggs Floral buds Yellow sticky boards** 12.8 3.4b 0.3 0.3b — Emergence 122.5 13.0a 175.0 18.7a — Dissection 0.0 0.0c 182.3 15.5a 27.5 7.9 Leaf buds Yellow sticky boards** 12.8 3.4a 0.3 0.3b — Emergence 4.3 1.7b 6.0 2.1a — Dissection 0.0 0.0c 4.0 2.0ab 31.3 8.2 *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted on 3, 6, and 10 March for both floral and leaf buds. Means given are the total number of D. oxycoccana detected by each technique over all three sampling dates. **Data for yellow sticky boards were identical for floral and leaf bud comparisons since sampling was conducted on the same dates for these bud types.

PAGE 63

52 Table 3-4. Comparison of techniques for monitoring D. oxycoccana in southern highbush blueberries in Windsor, FL (2003). Mean SEM* D. oxycoccana Life stage detected Sampling technique Adults Larvae Eggs Floral buds Yellow sticky boards 0.3 0.3b 0.0 0.0c — Emergence 19.8 3.5a 28.0 5.1a — Dissection 0.0 0.0b 13.8 4.4b 0.0 0.0 Leaf buds Yellow sticky boards 0.3 0.3b 0.0 0.0 — Emergence 4.0 2.7a 5.8 3.8 — Dissection 0.0 0.0b 6.3 3.6 52.5 15.5 *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted on 3, 7, and 13 March for floral buds, and 24 February and 3 and 6 March for leaf buds. Means given are the total number of D. oxycoccana detected by each technique over all three sampling dates.

PAGE 64

53 Table 3-5. Infestation of rabbiteye and southern highbush blueberry buds by D. oxycoccana, Windsor, FL (2002). Mean SEM* of D. oxycoccana per 420 buds Life stage detected Eggs Larvae Bud type Rabbiteye floral 17.3 6.3b 133.5 32.4a Rabbiteye leaf 13.3 2.0b 12.0 6.4b Southern highbush floral 7.3 4.6b 16.8 4.7b Southern highbush leaf 56.0 11.3a 5.0 2.6b *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Bud samples were dissected under a dissecting microscope to determine the presence of D. oxycoccana eggs and larvae. Treatment means are the sum of data collected over three sampling dates for each bud type.

PAGE 65

54 Table 3-6. Infestation of rabbiteye and southern highbush blueberry buds by D. oxycoccana, Windsor, FL (2003). Mean SEM* of D. oxycoccana per 420 buds Life stage detected Eggs First instar Second instar Third instar Bud type Rabbiteye floral 27.5 7.9a 61.0 16.9a 108.0 6.3a 12.3 4.7a Rabbiteye leaf 31.3 8.2a 3.8 1.9bc 0.3 0.3b 0.0 0.0b Southern highbush floral 0.0 0.0b 13.5 4.3b 0.3 0.3b 0.0 0.0b Southern highbush leaf 52.5 15.5a 1.0 0.4c 0.0 0.0b 0.0 0.0b *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Bud samples were dissected under a dissecting microscope to determine the presence of various life stages of D. oxycoccana. Treatment means are the sum of data collected over three sampling dates for each bud type.

PAGE 66

CHAPTER 4 EVALUATION OF MONITORING TECHNIQUES FOR FLOWER THRIPS (THYSANOPTERA: THRIPIDAE) IN SOUTHERN HIGHBUSH AND RABBITEYE BLUEBERRIES Since the 1990s, the Florida blueberry industry has been expanding production of southern highbush (Vaccinium corymbosum L. x V. darrowi Camp) blueberries throughout the state. This movement is a dramatic shift from the long-established rabbiteye (V. ashei Reade) plantings (Williamson et al. 2000). The increased acreage of southern highbush plantings in Florida is a response to an increased market for shipping early-season blueberries, which have a higher market value compared with later-ripening blueberries from primary-blueberry producing states. In 2002, sales of southern highbush blueberries exceeded $18 million (Florida Agricultural Statistics 2003). Despite this trend, further growth and development of the blueberry industry in Florida is impeded by the high incidence of flower thrips, Frankliniella spp. In a recent survey conducted by the University of Florida Fruit and Vegetable IPM Laboratory in Gainesville, in cooperation with several county extension agents, Florida blueberry growers cited flower thrips as their most important insect pest concern that warrants management (Chapter 1). Blueberry growers in neighboring southeastern states share similar concerns regarding flower thrips in their plantings (NeSmith 1999). Flower thrips belonging to the genus Frankliniella are well-documented pests of plants in the genus Vaccinium (Wood 1956; Braman et al. 1996). There are a number of thrips species that damage blueberries in North America. On lowbush blueberry in Maine and Atlantic Canada, Frankliniella vaccinii Morgan and Catinathrips kainos 55

PAGE 67

56 O’Neill are minor pests causing leaf distortion and discoloration in small isolated patches in fields (Langille and Forsythe 1972). In the past few years, thrips have caused considerable damage to highbush plantings in New Jersey. Two species in particular, Frankliniella tritici (Fitch) [eastern flower thrips] and Scirtothrips ruthveni Shull were causing extensive damage to leaves and flowers with potential to damage fruit (Polavarapu 2001). Typically, leaf curling and malformation were observed, with damage to styles and surrounding green tissue in the flowers. In the Southeast, F. tritici, Frankliniella bispinosa (Morgan) [Florida flower thrips] and Frankliniella occidentalis (Pergrande) [western flower thrips] have been identified as pests of both rabbiteye and southern highbush blueberries (Liburd et al. 2002). These three species are known to have a wide host range and cause extensive damage to many different crop plants, with feeding primarily on lush tissue such as buds, flowers and young leaves (Kirk 1995; Lewis 1997). In blueberry, feeding on pollen, styles, and developing berries can initiate major yield losses. Little is known about flower thrips management with regard to blueberry production in the southeastern United States. In general, flower thrips species have a very short life cycle, approximately 18 to 22 days under ideal conditions, and will complete multiple generations per year (Lewis 1997). These factors, as well as the ability of thrips to move long distances along air currents, may complicate management strategies for many Florida blueberry growers. Although sticky board traps have been a useful tool for monitoring thrips populations in other crops (Childers and Brecht 1996), it is uncertain whether this monitoring technique will work in blueberry plantings. Two other techniques that are becoming increasingly popular in food crops include immersing

PAGE 68

57 flower clusters into alcohol for microscopic evaluation and dissection of leaf and floral buds (Funderburk, pers. comm.). Currently, Florida blueberry growers are using the ‘beating sheet’ method, whereby a 30 x 30-cm white board is placed under the foliage or flower clusters. These plant parts are then tapped a few times to dislodge thrips from the plant onto the white board. One of the problems with this technique is that thrips may remain in the flowers protected by the corolla, giving an inaccurate estimation of density in plantings. The development of an effective technique for monitoring flower thrips in blueberry plantings would enable growers to make informed management decisions and potentially reduce the number of insecticide applications required for control. My objective was to evaluate various sampling techniques for monitoring flower thrips in southeastern blueberry plantings. Since blue and white sticky boards were found to be effective for monitoring F. bispinosa in citrus groves (Childers and Brecht 1996), I wanted to investigate various colors of sticky traps for monitoring thrips in blueberries. I also wanted to investigate other techniques, including dipping flower clusters into alcohol, tapping floral clusters over white boards, and dissecting floral clusters under a microscope as alternatives to using sticky traps for monitoring thrips populations in infested plantings. Finally, I wanted to study the phenology of thrips in rabbiteye and southern highbush blueberry plantings. Materials and Methods In 2002, thrips populations increased dramatically over the course of the monitoring period, making sampling time-consuming and prone to error. After one week in the field, sticky board traps often contained sufficient numbers of thrips that the entire surface of the trap was covered, making counting difficult. Therefore, a pilot study was

PAGE 69

58 conducted to develop a subsampling protocol for counting flower thrips on sticky board traps. The pilot study was used to determine: how thrips were spatially distributed on sticky traps containing a grid system the minimum number of quadrats (of 63 total) that had to be counted to yield precise estimates of total thrips per trap Pilot Study. Unbaited yellow sticky traps with a grid system (63 quadrats, size 2.5 x 2.5 cm) [Great Lakes IPM, Vestaburg, MI] were used to develop my subsampling protocol. The experiment was conducted in a 1-ha section of a southern highbush planting in Windsor, FL. Four yellow sticky traps were hung in a vertical position within the canopy of bushes at random and spaced 20 m apart. Sampling was conducted weekly from 24 January to 11 March 2002. Traps were re-randomized and replaced on each sampling date. During sampling, traps were removed from the field and covered with a layer of plastic wrap, then transported back to the laboratory for visual analysis. In the laboratory, traps were thoroughly examined with respect to thrips distribution. With the assistance of a statistician Dr. Kenneth Portier, University of Florida, Gainesville, Florida, I determined variance estimates for thrips counts among the 63 quadrats and between trap replicates using the nested procedure (SAS Institute Inc. 2001). I maximized relative precision estimates for moderate numbers of thrips per trap (intermediate population levels), where count estimates varied 37% from the true count with 95% confidence. In future estimations, I assumed (based on data from the pilot study) that flower thrips were distributed randomly on the traps and adopted a systematic sampling protocol using 15 quadrants per trap to estimate total thrips counts. This subsampling method was used exclusively to determine treatment efficacy for subsequent experiments. For traps that did not possess a grid system, I reproduced an identical grid

PAGE 70

59 system on a transparency and overlaid it in on the traps before sampling (Figure 4-2). All traps collected in the pilot study were analyzed prior to analyzing traps collected for subsequent studies. Unbaited Colored Traps Experiments to evaluate the effect of color for monitoring flower thrips using unbaited sticky traps were conducted in both rabbiteye and southern highbush plantings in Windsor, and Inverness, FL, respectively. Individual sites consisted of 2-ha blocks. The monitoring period for each planting was carefully planned to include the 7 stages of floral bud development as outlined by Spiers (1978). In 2002, monitoring was conducted weekly from 5 March to 12 April at the rabbiteye planting and 31 January to 6 March at the southern highbush planting. In 2003, monitoring was conducted weekly from 4 February to 11 March at the southern highbush planting and twice per week from 6 to 21 March at the rabbiteye planting (five sampling dates). Blueberry bushes in the rabbiteye planting were ~ 2.5 to 3 m tall and contained the following cultivars: ‘Beckyblue,’ ‘Bonita,’ and ‘Climax.’ Bushes in the southern highbush planting were ~ 1.5 to 2 m tall and contained the following cultivars; ‘Jewel,’ ‘Misty,’ ‘Sharpblue,’ and ‘Star.’ Bush spacing was 1.5 m apart and 2.4 m between rows in each planting. Various colors of commercially produced rectangular unbaited sticky board traps (treated area 394 cm 2 , Great Lakes IPM) were used to monitor flower thrips in commercial blueberry plantings. Four treatments were evaluated: Standard pantone yellow Safety white Walnut husk green Thrips blue

PAGE 71

60 Experimental design was randomized complete block with four replicates per treatment in each planting. Traps were hung within the canopy of blueberry bushes in a vertical position and spaced 10 m apart and 15 m between blocks. Traps were re-randomized and replaced on each sampling date. During sampling, traps were removed from the field and covered with a layer of plastic wrap, then transported back to the laboratory for visual analysis. The most effective trap color was determined based on the highest mean captures of flower thrips throughout the season using my newly developed subsampling protocol. During the sampling operation, I did not distinguish between life stages or sex due to time constraints. Climatic conditions were recorded throughout each of the sampling periods at both plantings to study thrips phenology. Sampling Techniques Experiments to evaluate different sampling techniques for detecting flower thrips were conducted in both rabbiteye and southern highbush plantings in Windsor, and Inverness, FL, respectively. Individual sites consisted of a 1-ha blocks and blueberry bushes were ~ 1.5 to 2 m tall. Four techniques were evaluated for their effectiveness in detecting flower thrips populations: Using unbaited white sticky boards Dipping flower clusters into alcohol Tapping floral clusters onto a white surface Collecting floral clusters for dissection Experimental design was randomized complete block with 4 replicates per treatment in each planting. Treatments were blocked by cultivar. Each experimental unit consisted of approximately 15 bushes, which were spaced 1.5 m apart and 2.4 m between rows in each planting.

PAGE 72

61 Sampling. White sticky boards were sampled using the subsampling protocol that was developed from the pilot study. Briefly, 15 of the 63 total quadrats were systematically sampled to determine thrips populations. In the latter three treatments, 40 flower clusters per treatment (10 flower clusters per replicate) were collected at random and sampled for the presence of flower thrips. Sampling was conducted in the early afternoon when thrips activity was presumed to be high. In my alcohol dip technique, clusters were immediately placed into 237-ml white polyethylene jars (B & A Products, Ltd. Co., Bunch, Oklahoma) containing ~ 80 ml of 70% ethyl alcohol and returned to the laboratory for further analysis. Jar contents were separated using a funnel with a 2 mm screen. The remaining plant material was rinsed with ~ 200-ml water to recover any remaining thrips. The total number of thrips in both alcohol and water aliquots was counted and recorded using a 10X-dissecting microscope. In my floral tap technique, individual clusters were tapped 5 times (50 taps of 10 flower clusters per replicate) over a 12 x 12-cm white cardboard surface, and the total number of thrips collected on the cardboard was counted and recorded. In my floral dissection technique, 10 clusters per replicate were immediately placed into 60-ml plastic containers and returned to the laboratory for dissection. Overall, the total number of thrips in each sample was determined by counting the number of thrips per 10 clusters during dissection. In addition, I inspected floral organs (n = 40 clusters), including ovaries, pistils, and stamens, to determine which floral organs were most susceptible to flower thrips damage. Treatments (monitoring techniques) were rotated weekly among experimental plots. The most efficient sampling technique was determined based on the ability and ease for

PAGE 73

62 detecting flower thrips. Again, no effort was made to distinguish between life stage and sex due to time constraints. Data Analysis Data from both studies were square-root transformed to account for deviations from normality and then subjected to an analysis of variance (ANOVA) followed by mean separation using least significant difference (LSD) tests (SAS Institute 2001). Data were also subjected to repeated measures analysis (using PROC MIXED, SAS Institute 2001) to examine interaction effects between treatment and time (sampling date) throughout the duration of each experiment. When significant interaction effects were noted, further analysis was conducted to determine the order of treatment efficacy for each sampling date. Means were considered significant when P values were 0.05. The untransformed means and standard errors are presented in tables and figures. Results In each of my studies, F. bispinosa was the most abundant species of flower thrips I encountered, comprising more than 95% of the total thrips in my samples. The other species observed were F. tritici and F. occidentalis. Pilot study. Based on variance estimates as well as standard equations for sample size estimation, I determined that counting 15 of the 63 quadrats per sticky trap would provide adequate precision for estimating total thrips per trap (Figure 4-1). As the number of quadrats to sample increases beyond 15, there was very little reduction in the variance component for among quadrat-sampling. Unbaited Colored Traps In 2002, white and blue sticky boards captured significantly (F = 6.34; df = 3, 9; P = 0.01) more F. bispinosa compared with yellow and green boards in my rabbiteye

PAGE 74

63 planting (Table 4-1). Similar results were recorded in my southern highbush planting, where white and blue boards captured significantly (F = 41.2; df = 3, 9; P < 0.01) more F. bispinosa compared with yellow boards, which captured more than green boards. Overall, blue appears to be most the most effective color for monitoring F. bispinosa, followed by white, yellow, and green, respectively. There were significant interaction effects between treatments and sampling date in the rabbiteye planting, where yellow was the most effective trap color early in the season (5 and 13 March) when populations of thrips were low. However, as the season progressed and thrips populations increased (22 March through 12 April), blue and white were the most effective trap colors. In 2003, white, blue, and yellow sticky traps captured significantly (F = 10.3; df = 3, 9; P < 0.01) more thrips than green in the rabbiteye planting (Table 4-2). Similar results were recorded in the southern highbush planting, where white, blue, and yellow sticky traps captured significantly (F = 13.9; df = 3, 9; P < 0.01) more thrips than green traps. Again, there were significant (F = 6.9; df = 16, 45; P < 0.01) interaction effects between treatment and sampling dates in the southern highbush planting. As before, yellow was the most effective trap color for monitoring thrips when populations were low (5 March in the southern highbush planting and 6 to 18 March in the rabbiteye planting), but when thrips populations peaked (11 and 21 March in the southern highbush and rabbiteye plantings, respectively), blue and white traps were more effective. Overall, thrips populations were higher in 2002 than in 2003 (Figure 4-2, 4-3). In 2002, thrips were a serious pest in my rabbiteye planting for three weeks, with peak density (2270 540 thrips per trap) occurring on 12 April (Figure 4-3). In 2003, peak thrips density (125 10 thrips per trap) occurred on 21 March, approximately three

PAGE 75

64 weeks earlier than in the previous year in the same planting (Figure 4-3). By contrast, in the southern highbush planting, thrips populations peaked approximately three weeks later in 2003 (22 5 thrips per trap) than in 2002 (130 30 thrips per trap) [Figure 4-4]. The number of days between peak thrips densities in the southern highbush and rabbiteye plantings was 40 in 2002 and 10 in 2003 (Figure 4-3, 4-4). Sampling Techniques In my rabbiteye planting, white sticky boards were significantly (F = 33.6; df = 3, 9; P < 0.01) more effective in detecting F. bispinosa than the other techniques (dipping flower clusters into alcohol, tapping floral clusters over white cardboard, or dissecting flower clusters) [Table 4-3]. Although there were significant (F = 52.9; df = 12, 36; P < 0.01) interaction effects between treatment and time, white boards were always the most effective technique for sampling thrips in my rabbiteye planting, regardless of sampling date. Similar results were recorded in the southern highbush planting, where white boards were significantly (F = 21.6; df = 3, 9; P < 0.01) more effective in detecting thrips than the other techniques I evaluated. Overall, alcohol dip and floral dissection techniques were equivalent in their ability to detect thrips in both of my plantings. Tapping floral clusters over a flat white surface was the least effective technique, regardless of planting type. While dissecting the floral organs, I noted that nymphs and adults were present on the ovary, style, filaments, and anthers. I also noted the presence of feeding scars on the tissues of the ovary that develop into the calyx cup. In general, feeding scars were brown and contrasted the green color of the developing fruit.

PAGE 76

65 Discussion Overall, my results indicated that blue and white sticky boards were the most efficient colors for monitoring thrips, specifically F. bispinosa, in Florida blueberry plantings. These results parallel work by Childers and Brecht (1996) who found that blue and white were effective colors for monitoring F. bispinosa using sticky boards during the flowering cycles of citrus in Florida. In my studies, yellow sticky boards were found to be effective for monitoring thrips early in the flowering season when populations were low. The reason why there is a shift in trap captures from yellow (early in the season) to white or blue later in the flowering period is unknown, but it may be related to changes in perception from increased vegetation or changes in thrips maturity in relation to their phenology. Another hypothesis is that males and females of F. bispinosa are attracted to different ranges of the visual spectrum. While I did not systematically quantify specimens based on sex throughout my studies, it appeared that the ratio of females to males was low early in the sampling period when yellow traps captured the most thrips (personal observation). This pattern appeared to reverse later in the flowering period when the ratio of females to males increased. In greenhouse studies with F. occidentalis, Vernon and Gillespie (1990) found that the ratio of females to males was consistently higher on blue traps than on yellow or white traps. Future studies with F. bispinosa may address this hypothesis. The presence and abundance of thrips in blueberry plantings appears to be heavily influenced by climatic conditions. In 2002, thrips populations were much higher than in 2003. The variation in population density from 2002 to 2003 is probably the result of a milder winter in 2002. Temperatures in north-central Florida exceeded 27 o C in mid-February, which may have allowed thrips populations to build up earlier in the

PAGE 77

66 southern highbush planting. However, there was a late freeze in early March 2002 that delayed the build-up of thrips in rabbiteye plantings later in the season. Consequently, in 2002 I observed a long period between peak thrips densities in the southern highbush and rabbiteye plantings. A cold winter in 2003 allowed for a short flowering period in both rabbiteye and southern highbush blueberries, which may partially explain the reduced duration and abundance of thrips in my plantings. The development of a subsampling protocol for sticky boards has facilitated the speed and precision with which I can count flower thrips in blueberry plantings. I determined that there is no increase in precision by counting more than 15 quadrats on a sticky trap with a 63-quadrat grid system. In my studies, thrips were distributed randomly on traps, and therefore I was able to adopt a systematic approach for counting flower thrips within quadrats. The major advantage with this system is that it allows for more precise estimates of thrips densities during peak flight. In my sampling techniques study, I found that sticky boards were the most effective technique for monitoring flower thrips in blueberry plantings. Overall, this technique is less labor intensive and more cost effective than other monitoring techniques I evaluated. The standard technique of tapping flower clusters was also cost effective and simple, but it was less reliable for detecting flower thrips than sticky boards, particularly early in the season when populations were low. The dipping of infested flower clusters into alcohol was largely impractical, since the structure of blueberry flowers, specifically the shape and length of the corolla, often prevented easy separation of plant and insect material. My final technique, flower dissection, was the most labor-intensive strategy for monitoring thrips in my plantings and is not recommended for practical purposes.

PAGE 78

67 Nonetheless, through my dissections I noted activity of F. bispinosa nymphs and adults on the ovary, style, filaments, and anthers of blueberry flowers. This observation is similar to that reported by Childers and Achor (1991), who found that F. bispinosa was active on the same floral organs in citrus. In summary, I developed an effective subsampling protocol for monitoring thrips populations in blueberry plantings using sticky board traps. My studies demonstrated that white and blue sticky traps are more effective colors for monitoring thrips than yellow or green traps. Furthermore, white sticky boards were more effective than other techniques for monitoring thrips populations in affected blueberry plantings. In the future, blueberry growers may benefit from combining the use of white sticky board traps and a subsampling protocol for quickly and accurately monitoring flower thrips in their plantings.

PAGE 79

68 Table 4-1. Comparison of various colors of sticky board traps for monitoring F. bispinosa in rabbiteye and southern highbush blueberries in north-central Florida (2002). Mean SEM* F. bispinosa Color Rabbiteye Southern highbush Standard pantone yellow 1106.0 342.0b 167.8 21.2b Safety white 3285.0 931.7a 277.8 28.0a Walnut husk green 465.3 192.4b 53.8 5.9c Thrips blue 4061.5 2131.6a 281.5 24.4a *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted on 3/5, 3/13, 3/22, 3/28, and 4/5 in the rabbiteye planting, and 1/31, 2/6, 2/13, 2/21, 2/27, and 3/6 in the southern highbush planting. Means given are the total number of F. bispinosa over all sampling dates as determined by my subsampling protocol.

PAGE 80

69 Table 4-2. Comparison of various colors of colored sticky board traps for monitoring F. bispinosa in rabbiteye and southern highbush blueberries in north-central Florida (2003). Mean SEM* F. bispinosa Color Rabbiteye Southern highbush Standard pantone yellow 229.8 20.5a 83.3 26.2a Safety white 193.0 35.2a 56.0 9.7a Walnut husk green 76.8 11.8b 14.0 1.9b Thrips blue 211.8 38.0a 65.8 11.1a *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted on 3/6, 3/10, 3/13, 3/18, and 3/21 in the rabbiteye planting, and 2/4, 2/11, 2/18, 2/26, 3/5, and 3/11 in the southern highbush planting. Means given are the total number of F. bispinosa over all sampling dates as determined by my subsampling protocol.

PAGE 81

70 Table 4-3. Comparison of sampling techniques for monitoring F. bispinosa in rabbiteye and southern highbush blueberries in north-central Florida (2003). Mean SEM* F. bispinosa Sampling technique Rabbiteye Southern highbush White sticky boards 168.8 9.3a 271.5 72.8a Alcohol dip 101.0 9.8b 50.0 9.6b Floral tap 57.5 10.2c 30.0 4.9b Floral dissection 100.0 13.5b 11.8 2.1b *Means within columns followed by the same letter are not significantly different, P = 0.05, LSD Test. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Sampling was conducted on 3/6, 3/13, 3/18, and 3/21 in the rabbiteye planting, and 2/4, 2/18, 2/26, and 3/5 in the southern highbush planting. Means given are the total number of F. bispinosa over all sampling dates.

PAGE 82

71 0.00.10.20.30.40.50.60.70.802468101214161820Number of Quadrats per TrapAmong-Quadrat Variance Component wk 1 wk 2 wk 3 wk 4 wk 5 wk 6 wk 7 Figure 4-1. Results of a pilot study to determine how many quadrats on an individual sticky board trap had to be counted to reasonably estimate true counts of F. bispinosa. The variance components on the y-axis were calculated by dividing the between-quadrat variance component by the total number of quadrats to be sampled. Overall, the variance among quadrats was not greatly reduced upon sampling more than 15 of the 63 quadrats per trap.

PAGE 83

72 Figure 4-2. Sticky board trap overlaid with a transparency of a grid system to facilitate accurate counting of flower thrips. Fifteen of 63 quadrats were systematically sampled to estimate true counts of thrips on traps.

PAGE 84

73 05001000150020002500300028-Feb5-Mar10-Mar15-Mar20-Mar25-Mar30-Mar4-Apr9-Apr14-Apr19-AprDateMean # F. bispinosa on White Sticky Board Traps 2002 2003 Figure 4-3. Abundance of F. bispinosa throughout the flowering period of a rabbiteye blueberry planting in Windsor, FL (2002 and 2003). Thrips populations were monitored for a 6-week period in each planting using white sticky board traps. The mean number of F. bispinosa was determined by systematically sampling specimens on 15 of the 63 quadrats on each of four white sticky traps.

PAGE 85

74 02040608010012014016018029-Jan3-Feb8-Feb13-Feb18-Feb23-Feb28-Feb5-Mar10-Mar15-MarDateMean # F. bispinosa on White Sticky Board Traps 2002 2003 Figure 4-4. Abundance of F. bispinosa throughout the flowering period of a southern highbush blueberry planting in Inverness, FL (2002 and 2003). Thrips populations were monitored for a 6-week period in each planting using white sticky board traps. The mean number of F. bispinosa was determined by systematically sampling specimens on 15 of the 63 quadrats on each of four white sticky traps.

PAGE 86

CHAPTER 5 EVALUATION OF SELECTED INSECTICIDES FOR MANAGING BLUEBERRY GALL MIDGE (DIPTERA: CECIDOMYIIDAE) AND FLOWER THRIPS (THYSANOPTERA: THRIPIDAE) IN SOUTHERN HIGHBUSH AND RABBITEYE BLUEBERRIES Several new insecticide chemistries have been introduced for use in fruit crops, largely in response to the potential loss and/or restriction of organophosphates (OPs) and carbamates due to the Food Quality Protection Act of 1996 (FQPA). Many of these new chemistries are classified as reduced-risk because they pose less threat to the environment, humans, and non-target organisms. Applications of reduced-risk insecticides, including neonicotinoids and naturalytes, may hold promise for control of key blueberry pests, including blueberry gall midge, Dasineura oxycoccana (Johnson); and flower thrips, Frankliniella spp. Field trials are necessary to determine the potential of using selected reduced-risk compounds in blueberry integrated pest management (IPM) programs. The blueberry gall midge is a primary pest affecting rabbiteye (Vaccinium ashei Reade) blueberry plantings in the southeastern United States (Lyrene and Payne 1995). Prior to 1992, floral bud abortion caused by D. oxycoccana in rabbiteye blueberries had not been correctly diagnosed, and therefore chemical control had not been recommended (Lyrene and Payne 1992). Since then, D. oxycoccana infestations have increased significantly, destroying up to 80% of floral buds on susceptible rabbiteye cultivars. Dasineura oxycoccana also infests southern highbush (V. corymbosum x V. darrowi Camp) blueberries, which ripen early in the season and thus have a high market value. 75

PAGE 87

76 The potential threat imposed by D. oxycoccana to southern highbush plantings may be very devastating to the Florida blueberry industry. Flower thrips are another key pest of blueberries in the southeastern United States. Several species, including Frankliniella tritici (Fitch) [eastern flower thrips], Frankliniella bispinosa (Morgan) [Florida flower thrips] and Frankliniella occidentalis (Pergrande) [western flower thrips] have been identified as pests of both rabbiteye and southern highbush plantings. These species are known to have wide host ranges and cause damage to various crop plants, primarily by feeding on lush tissues such as buds, flowers, and young leaves (Lewis 1997). Neonicotinoids are a relatively new class of systemic insecticides that interfere with the transmission of nerve impulses by mimicking acetylcholine and binding to nicotinergic receptors on nerve cells to prolong nerve stimulation. Neonicotinoids are contact and stomach poisons, and are particularly effective against sucking insects because of their translaminar effects (Elbert et al. 1990; Isaacs et al. 1999). Imidacloprid and thiamethoxam are two neonicotinoids that have shown promise for control of other dipteran fruit pests including blueberry maggot, Rhagoletis mendax Curran, in field and laboratory studies (Liburd et al. 2003). Like neonicotinoids, naturalytes are a class of reduced-risk insecticides that alter nerve stimulation, specifically by enhancing the activity of acetylcholine at a separate binding site (Horowitz et al. 1998). Spinosad is one naturalyte compound that has recently received registration for blueberry pests, including flower thrips. In laboratory studies, spinosad was as effective as phosmet (an OP) in inducing mortality of D. oxycoccana larvae after 24 hours (Sampson et al. 2002). Additional studies are needed to evaluate the effectiveness of spinosad for control of D. oxycoccana in the field.

PAGE 88

77 In addition to these chemistries, several organic-approved insecticides, including neem (azadirachtin) and kaolin clay, have shown effectiveness against a variety of fruit pests (VanRanden and Roitberg 1998; Knight et al. 2000). In laboratory studies, VanRanden and Roitberg (1998) found that incorporation of azadirachtin into the diet of western cherry fruit fly, Rhagoletis indifferens Curran, larvae resulted in decreased pupal formation and subsequent adult emergence. Kaolin clay, a particle protectant, is another organic-approved compound that has shown effectiveness in fruit crops. In choice and no-choice tests, Lapointe (2000) reported suppression of oviposition by citrus root weevil, Diaprepes abbreviatus (L.), on leaves treated a kaolin formulation (M-97-009). Similarly, in choice tests with the silverleaf whitefly, Bemisia argentifolii Bellows and Perring, melon leaves treated with kaolin had significantly fewer eggs than water-treated control leaves (Liang and Liu 2002). Given the importance of blueberries as a special commodity in the state of Florida and throughout the southeastern United States, along with the threat to production imposed by blueberry gall midge and flower thrips populations, effective compounds need to be identified for inclusion into blueberry IPM programs. My objective was to evaluate several insecticide chemistries for their effectiveness against D. oxycoccana and Frankliniella spp. in blueberry plantings throughout the southeastern United States. Materials and Methods Blueberry Gall Midge Experiments to evaluate the effectiveness of several insecticide chemistries for managing D. oxycoccana were conducted in heavily infested rabbiteye plantings in Windsor, FL (2002) and Alma, GA (2003). The planting in Florida was a 2-ha block and contained the following cultivars: ‘Beckyblue,’ ‘Bonita,’ and ‘Climax.’ The planting in

PAGE 89

78 Georgia was a 4-ha block and contained ‘Climax’ and ‘Tifblue’ cultivars. Blueberry bushes were spaced 1.5 m apart and 2.4 m between rows in each planting. 2002. Five treatments were evaluated: Spinosad (SpinTor 2SC, Dow Agrosciences, Carmel, IN) at a rate of 0.4 L / ha (high rate) Spinosad at a rate of 0.3 L / ha (low rate) Imidacloprid (Provado 1.6F, Bayer, Kansas City, MO) at a rate of 0.3 L / ha Phosmet (Imidan 70W, Gowan, Yuma, AZ) at a rate of 1.5 kg / ha An untreated control Treatments were applied on 13 March, during stage 3 of floral bud development (Spiers 1978). All treatments were applied using a Solo backpack sprayer (Newport News, VA) with a delivery pressure of 448.2 kPa. Experimental design was randomized complete block with 4 replicates per treatment. Treatments were blocked by cultivar. A 20 m buffer zone was allocated between treatments to adjust for potential spray drift. Insecticide effectiveness for suppressing D. oxycoccana was evaluated by randomly selecting and dissecting 200 floral buds per treatment (50 buds per replicate). Bud samples were collected on 15 and 18 March (400 floral buds per treatment). Results from the dissections were separated categorically based on life stage present (egg, larva, puparia, or adult) for each treatment and compared separately. 2003. Seven treatments were evaluated: Diazinon (Diazinon AG500, Syngenta Crop Protection, Greensboro, NC) at a rate of 1.8 L / ha Malathion (Malathion 5 EC, Gowan, Yuma, AZ) at a rate of 1.8 L / ha Thiamethoxam (Actara 25 WG, Bayer) at a rate of 0.3 L / ha Spinosad (SpinTor 2SC) at a rate of 0.4 L / ha (high rate)

PAGE 90

79 Azadirachtin (Ecozin 3% EC, Amvac, Los Angeles, CA) at a rate of 0.6 L / ha Kaolin clay (Surround WP, Engelhard Corporation, Iselin, NJ) at a rate of 28 kg / ha An untreated control Treatments were applied on 14 and 28 February and 14 March, during stages 2 to 4 of bud development (Spiers 1978). All treatments were applied using an airblast sprayer (Gyro-speed 260) fitted with a ceramic disc (#12) nozzle head (Berthoud/Excel GSA, Villefanche s/Sane, France). Experimental design was randomized complete block with 4 replicates per treatment. A 40 m buffer zone was allocated between treatments and blocks to adjust for potential spray drift. Samples were collected from both ‘Climax’ and ‘Tifblue’ cultivars. In 2003, treatment efficacy for suppressing D. oxycoccana infestation was evaluated by randomly selecting 200 floral buds per treatment per cultivar (50 buds per replicate). Bud samples from the cultivar ‘Climax’ were collected on 17 and 23 February and 2 March, for a total of 600 floral buds per treatment. Buds from the cultivar ‘Tifblue’ developed approximately one week after ‘Climax,’ and therefore sampling was conducted on the same dates (17 and 23 February and 2 March) as well as 10 March, for a total of 800 floral buds per treatment. All bud specimens were transferred to 15-cm plastic petri dishes containing moistened filter paper and held at 27 o C under 14L:10D conditions for 10 days to allow larvae to emerge. The total number of emergent larvae was recorded. Flower Thrips Experiments to evaluate the efficacy of several insecticide chemistries for managing thrips were conducted in a 2-ha block of a heavily infested southern highbush

PAGE 91

80 planting in Haines City, FL. Blueberry bushes were spaced 0.5 m apart and 1.5 m between rows (high-density planting). The planting contained the following cultivars in each bed: ‘Jewel,’ ‘Misty,’ ‘Sharpblue,’ and an experimental clone FL 93-204. Seven treatments were evaluated: Phosmet (Imidan 70 W) at a rate of 1.5 kg / ha Imidacloprid (Provado1.6F) at a rate of 0.3 L / ha Spinosad (SpinTor 2SC) at a rate of 0.4 L / ha (high rate) Azadirachtin (Ecozin 3% EC) at a rate of 0.6 L / ha Activated garlic extract (Cropmaster Repel, U. A. S. of America, Inc., Hudson, Florida) at a rate of 2% final volume Kaolin clay (Surround WP) at a rate of 28 kg / ha An untreated control Treatments were applied on 30 January, and 12 and 25 February using Solo backpack sprayers (Newport News, VA) with a delivery pressure of 448.2 kPa. Total volume was 45 L per treatment. Experimental design was randomized complete block with 4 replicates per treatment. Beds containing several cultivars each were assigned as blocks. A 10 m buffer zone was allocated between treatments to adjust for potential spray drift. Insecticide effectiveness for suppressing thrips was evaluated using two separate techniques. In the first technique, white sticky board traps were hung in the center of each experimental plot within the canopy of bushes in a vertical position. Traps were replaced on each sampling date. During sampling, traps were removed from the field and covered with a layer of plastic wrap, then transported back to the laboratory for analysis. Traps were counted with the aid of a 10X-dissecting microscope using the subsampling protocol developed in Chapter 4. Briefly, a transparency containing a grid system of 63

PAGE 92

81 equally sized 2.5 x 2.5 cm quadrats was overlaid onto each trap. A total of 15 quadrats were systematically sampled from each board and these counts were then used to estimate thrips populations within treatments. In my second technique, 40 flower clusters per treatment (10 flower clusters per replicate) were collected at random and sampled for the presence of flower thrips. Sampling was conducted in the early afternoon when thrips activity was presumed to be high. Clusters were taken from the field and immediately placed into 237-ml white polyethylene jars (B & A Products, Ltd. Co., Bunch, OK) containing ~ 80 ml of 70% ethyl alcohol. Jar contents were separated using a funnel with a 2-mm screen in the laboratory. The remaining plant material was rinsed with ~ 200 ml water to recover any remaining thrips. The total number of thrips in both alcohol and water aliquots was counted and recorded using a 10X-dissecting microscope. Treatment evaluation based on the number of flower thrips on white sticky traps was conducted on 6, 12, 18, and 25 February, and 5 March 2003. Evaluation based on the number of thrips in flower clusters preserved in alcohol was conducted on the same dates, with the exception of the last date because flowering had already finished at that time. Since each block contained several cultivars, sampling on a particular date was restricted to the cultivar containing the highest percentage of open bloom. Overall, insecticide efficacy was determined by calculating the mean number of flower thrips using each technique. Data Analysis Data from each study were square-root transformed to account for deviations from normality and then subjected to an analysis of variance (ANOVA) followed by mean separation using least significant difference (LSD) tests (SAS Institute 2001). Data were

PAGE 93

82 also subjected to repeated measures analysis (using PROC MIXED, SAS Institute 2001) to examine the interaction effect between treatment and time (sampling date) throughout the duration of each experiment. When significant interaction effects were noted, LS-means values were computed and pairwise comparisons were made to determine the order of treatment efficacy for each sampling date. Means were considered significant when P values were 0.05. The untransformed means and standard errors are presented in tables and figures. Results Blueberry Gall Midge 2002. There were no differences in the number of D. oxycoccana eggs detected in floral buds among any of the treatments I evaluated. Similarly, none of the treatments I evaluated reduced larval infestation of D. oxycoccana compared with the control. Although not significantly, a single application of imidacloprid did reduce larval infestation by approximately 25% compared with the control. A single treatment of floral buds with spinosad was ineffective in suppressing larval infestation by D. oxycoccana, regardless of application rate (high or low). 2003. Larval infestation of floral buds by D. oxycoccana occurred earlier in ‘Climax’ than in ‘Tifblue’ cultivars (Figure 5-1). These data corresponded to the rate of floral bud development, where second stage bud development (Spiers 1978) was observed on the first sampling date for ‘Climax,’ and on the second sampling date for ‘Tifblue’ buds. In the cultivar ‘Climax,’ there were significant (F = 2.4; df = 14, 54; P = 0.01) interaction effects between treatment and sampling date, with significant treatment differences only observed on the first and third sampling dates (Table 5-1). Overall, diazinon-treated floral buds had significantly (F = 6.0; df = 6, 18; P < 0.01) fewer

PAGE 94

83 D. oxycoccana larvae compared with buds treated with other compounds (Figure 4-2). Furthermore, none of the other insecticides I evaluated significantly suppressed larval infestation compared with the control, with the exception of malathion, which was as effective as diazinon on the first sampling date (Table 5-1). Similar results were recorded for ‘Tifblue’ floral buds, with diazinon significantly (F = 5.4; df = 6, 18; P < 0.01) reducing larval infestation compared with all other insecticide treatments (Figure 5-3). Again, there were significant (F = 11.9; df = 21, 72; P = 0.01) interaction effects, and treatment differences were only observed on the third sampling date (Table 5-1). Flower Thrips In each of my studies, F. bispinosa was the most abundant species of flower thrips I encountered, comprising more than 95% of the total thrips in my samples. The other species observed were F. tritici and F. occidentalis. Overall, treatment effectiveness for managing flower thrips varied with sampling strategy. In my analysis based on dipping flower clusters into alcohol, kaolin clay was the only insecticide that significantly (F = 1.6; df = 6, 18; P = 0.21) reduced infestation of F. bispinosa compared with the control (Figure 5-4). However, there were significant (F = 58.4; df = 21, 72; P < 0.01) interaction effects between treatments and sampling dates. Among the 4 sampling dates, treatment differences were observed only on the second sampling date (Table 5-2). On this date, phosmet, imidacloprid, and kaolin clay were equivalent to the control, whereas samples taken from spinosad, azadirachtin, and garlic -treated plots actually contained significantly (F = 2.89; df = 6, 18; P = 0.04) more F. bispinosa (Table 5-2). Treatment comparisons made via the white sticky board method indicated that there were no overall differences among treatments. Again, there were significant (F = 5.1; df

PAGE 95

84 = 28, 90; P < 0.01) interaction effects between treatments and sampling dates. Among the five sampling dates, treatment differences were observed only on the first sampling date (Table 5-2). On this date, kaolin clay and spinosad were significantly (F = 2.7; df = 6, 18; P = 0.05) more effective in suppressing F. bispinosa populations compared with the control (Table 5-2). The remaining treatments were no different than the control. Discussion In my insecticide evaluation for managing D. oxycoccana, the lack of differences among treatments in 2002 was probably due to time of application, where applications were made after larvae began developing inside the buds. It is interesting, however, that buds treated with imidacloprid had 25% fewer D. oxycoccana larvae compared with untreated buds. Imidacloprid can be taken up by plant tissue (translaminar effect) [Elbert et al. 1990], and therefore it is possible that survivorship of D. oxycoccana could have been negatively affected, reducing infestations by 25%. Since D. oxycoccana larvae are well protected in buds while they are feeding, contact insecticides such as OPs may not have been as effective. In 2003, diazinon was the most effective compound for control of D. oxycoccana, followed by malathion, in both ‘Climax’ and ‘Tifblue’ cultivars. Unfortunately, since these compounds are OPs, their future in blueberry IPM programs may be in jeopardy due to FQPA regulations. Many broad-spectrum insecticides, including the OPs, have serious health and environmental consequences, including danger to agricultural workers, toxicity to non-target insects, and residues on fruit (Dinham 1993). In my studies, spinosad, imidacloprid, and thiamethoxam demonstrated minimal suppression of D. oxycoccana, though not at the levels needed for adequate control. Interestingly, kaolin clay appeared to increase infestation of D. oxycoccana in one of my

PAGE 96

85 trials. It is uncertain why I saw this phenomenon, although the color of the kaolin residues on the buds may have increased their attractiveness to D. oxycoccana. It is also possible that rain and other elements may have washed the compound off the foliage, limiting its activity against D. oxycoccana. Future studies should focus on ways to improve the effectiveness of these compounds. For instance, the addition of a sticker/spreader may allow for better adhesion of the particles to plant surfaces (Puterka et al. 2000). In contrast to my studies with D. oxycoccana, kaolin clay was the only compound that showed promise for managing flower thrips. Although I do not know how kaolin clay reduced thrips populations, it is possible that it may reduce thrips accessibility to blueberry flowers. Unfortunately, the structure of the blueberry flower allows thrips to feed in a protected environment. This detail may be one of the reasons why I did not record suppression of F. bispinosa in many of my treatments. Nonetheless, it is possible that a compound with systemic or translaminar effects such as imidacloprid or spinosad may demonstrate more effectiveness against thrips in future studies. One of the ways to improve treatment effectiveness is to adjust the intervals between spray dates depending on the residual activities for a particular chemistry. For instance, although diazinon and malathion are both OPs, malathion must be applied more frequently than diazinon because it has less residual activity. Here, monitoring strategies would play an important role in determining pest pressure over time as it relates to treatment effectiveness, dictating when subsequent applications are necessary. Unfortunately, assessing treatment effectiveness for managing thrips may be hindered by their quick movement within plantings and from adjacent hosts. Several

PAGE 97

86 factors may influence the increase of flower thrips populations in blueberry plantings, including the availability of alternate host plants, specifically citrus. Patterns of host migration have been noted for many Frankliniella spp. that affect cultivated crops. For instance, Chellemi et al. (1994) reported that F. occidentalis populations are often well established on wild hosts prior to the availability of cultivated hosts. The study also noted that, in general, Frankliniella spp. were more abundant and utilized more wild hosts compared with any other genera of flower-inhabiting thrips, appearing to be the most abundant genera of flower-inhabiting thrips in north-central Florida. The versatile feeding habits of Frankliniella spp. should be a concern for Florida blueberry growers, since many grow a number of other crops. Cultural practices such as proper weed management may be integrated with reduced-risk compounds that have demonstrated potential for reducing flower thrips populations. Future management of D. oxycoccana and F. bispinosa in blueberry plantings should include monitoring strategies to facilitate informed decisions regarding insecticide applications.

PAGE 98

87 Table 5-1. Effect of selected insecticides on infestation of rabbiteye floral buds by D. oxycoccana, Alma, GA (2003). Mean* number D. oxycoccana Sampling date Treatment 17 Feb 23 Feb 2 Mar 10 Mar Climax Diazinon 2.3a 6.8 4.8a — Malathion 2.0a 25.3 20.5b — Thiamethoxam 10.5b 8.3 27.5b — Spinosad 11.5b 19.5 24.5b — Azadirachtin 16.3b 30.0 35.5bc — Kaolin clay 20.0b 47.8 72.0c — Untreated control 15.5b 15.5 35.3bc — Tifblue Diazinon 0.0 2.3 0.8a 10.0 Malathion 0.0 2.0 11.8b 20.0 Thiamethoxam 0.0 2.0 19.8b 29.0 Spinosad 0.3 3.8 18.3b 17.0 Azadirachtin 0.3 5.3 31.3b 20.5 Kaolin clay 0.0 15.8 28.3b 23.3 Untreated control 0.3 13.0 29.3b 17.3 *LS-Means within columns followed by the same letter are not significantly different, P = 0.05. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Treatment effectiveness was evaluated by allowing larvae to emerge from infested buds.

PAGE 99

88 Table 5-2. Effect of selected insecticides on infestation of southern highbush flowers by F. bispinosa, Haines City, FL (2003). Mean* number D. oxycoccana Sampling date Treatment 6 Feb 12 Feb 18 Feb 25 Feb 5 Mar Alcohol dip Phosmet 0.0 0.5ab 5.3 136.8 — Imidacloprid 0.0 0.5abc 4.8 105.0 — Spinosad 0.0 1.3bc 8.0 101.0 — Azadirachtin 0.0 1.8c 9.3 123.0 — Garlic 0.0 1.3bc 7.8 98.8 — Kaolin clay 0.3 0.8abc 5.5 71.3 — Untreated control 0.0 0.0a 12.3 136.0 — White board Phosmet 2.5bc 9.0 28.0 244.3 1304.5 Imidacloprid 4.0bc 9.0 26.8 266.3 1438.8 Spinosad 1.5ab 7.0 27.0 278.0 2080.3 Azadirachtin 1.8bc 9.8 35.3 259.3 2157.5 Garlic 2.5bc 8.8 28.5 222.5 2124.5 Kaolin clay 0.8a 5.0 16.3 173.5 1320.0 Untreated control 4.8c 11.3 55.8 215.8 1322.5 *LS-Means within columns followed by the same letter are not significantly different, P = 0.05. Analysis was performed on square-root transformed data, but means shown reflect untransformed data. Treatment effectiveness was evaluated by two distinct techniques, alcohol dip and white boards.

PAGE 100

89 0102030405016-Feb20-Feb24-Feb28-Feb4-Mar8-Mar12-MarDateMean # Larvae Emerging from 50 Randomly Selected Floral Buds Climax Tifblue Figure 5-1. Infestation of untreated floral buds by D. oxycoccana in Alma, GA (2003). Buds were placed into 15-cm diameter petri dishes containing moistened filter paper and incubated at 27 o C under 14L:10D for 10 days.

PAGE 101

90 04080120160200DiazinonMalathionThiamethoxamSpinosadAzadirachtinKaolin ClayUntreated ControlTreatmentMean # Larvae per 150 Floral Buds c bbbabab Figure 5-2. Effect of selected insecticides on infestation of rabbiteye cv. ‘Climax’ floral buds by D. oxycoccana, Alma, GA (2003). Buds were placed into 15-cm diameter petri dishes containing moistened filter paper and incubated at 27 o C under 14L:10D for 10 days. Treatment means are the sum of larvae from buds collected on 17 and 23 February, and 2 March, 2003. Means followed by the same letter are not significantly different, P = 0.05, LSD Test.

PAGE 102

91 020406080100DiazinonMalathionThiamethoxamSpinosadAzadirachtinKaolin ClayUntreatedControlTreatmentMean # Larvae per 200 Floral Budscbabababaab Figure 5-3. Effect of selected insecticides on infestation of rabbiteye cv. ‘Tifblue’ floral buds by D. oxycoccana, Alma, GA (2003). Buds were placed into 15-cm diameter petri dishes containing moistened filter paper and incubated at 27 o C under 14L:10D for 10 days. Treatment means are the sum of larvae from buds collected on 17 and 23 February, and 2 and 10 March, 2003. Means followed by the same letter are not significantly different, P = 0.05, LSD Test.

PAGE 103

92 04080120160200PhosmetImidaclopridSpinosadAzadirachtinGarlicKaolin ClayUntreatedControlTreatmentMean # F. bispinosa per 40 Flower clustersabaaababab Figure 5-4. Effect of selected insecticides on infestation of southern highbush blueberry flowers by F. bispinosa in Haines City, FL (2003). Flower clusters from treated plots were placed into jars containing 70% ethanol for preservation and later examined in the laboratory for the presence of F. bispinosa. Treatment means are the total number of F. bispinosa from clusters collected on 6, 12, 18, and 25 February, 2003. Means followed by the same letter are not significantly different, P = 0.05, LSD Test.

PAGE 104

LIST OF REFERENCES Ahmad, T. R. 1987. Effects of pheromone trap design and placement on capture of almond moth, Cadra cautella (Lepidoptera: Pyralidae). J. Econ. Entomol. 80: 897-900. AliNiazee, M. T. 1983. Monitoring the filbertworm, Melissopus latiferreanus (Lepidoptera: Olethreutidae), with sex attractant traps: effect of trap design and placement on moth catches. Environ. Entomol. 12: 141-146. Averill, A. L., and M. M. Sylvia. 1998. Cranberry insects of the northeast: a guide to identification, biology, and management. Univ. Mass. Cranberry Exp. Sta. Bull. Beckwith, C. S. 1941. Control of cranberry fruitworm on blueberries. J. Econ. Entomol. 34: 169-171. Bosio, G., C. Bogetti, G. Brussino, F. Gremo, and F. Scarpelli. 1998. Dasineura oxycoccana, a new pest of highbush blueberry in Italy. Informatore Fitopatologico 11: 36-41. Braman, S. K., R. J. Beshear, J. A. Payne, and A. A. Amis. 1996. Population dynamics of thrips (Thysanoptera: Thripidae: Phlaeothripidae) inhabiting Vaccinium (Ericales: Ericaceae) galls in Georgia. Environ. Entomol. 25: 327-332. Brown, J. J. 1994. Effect of a nonsteroidal ecdysone agonist, tebufenozide, on host parasitoid interactions. Arch. Insect Biochem. Physiol. 26: 235-248. Cadogan, B. L., R. D. Scharbach, R. E. Krause, and K. R. Knowles. 2002. Evaluation of tebufenozide carry-over and residual effects on spruce budworm (Lepidoptera: Tortricidae). J. Econ. Entomol. 95: 578-586. Chellemi, D. O., J. E. Funderburk, and D. W. Hall. 1994. Seasonal abundance of flower-inhabiting Frankliniella species (Thysanoptera: Thripidae) on wild plant species. Environ. Entomol. 23: 337-342. Childers, C. C. and D. S. Achor. 1991. Feeding and oviposition injury to flowers and developing floral buds of ‘Navel’ orange by Frankliniella bispinosa (Tysanoptera: Thripidae) in Florida. Ann. Entomol. Soc. Am. 84: 272-282. Childers, C. C., and J. K. Brecht. 1996. Colored sticky traps for monitoring Frankliniella bispinosa (Morgan) (Thysanoptera: Thripidae) during flowering cycles in citrus. J. Econ. Entomol. 89: 1240-1249. 93

PAGE 105

94 Cho, K., C. S. Eckel, J. F. Walgenbach, and G. G. Kennedy. 1995. Comparison of colored sticky traps for monitoring thrips populations (Thysanoptera: Thripidae) in staked tomato fields. J. Entomol. Sci. 30: 176-190. Cockfield, S. D., and D. L. Mahr. 1994. Phenology of oviposition of Dasineura oxycoccana (Diptera: Cecidomyiidae) in relation to cranberry plant growth and flowering. Great Lakes Entomol. 27: 185-188. Coppage, D. L., and T. E. Braidech. 1976. River pollution by anticholinesterase agents. Wat. Res. 10: 19-24. Croft, B. A. 1990. Arthropod biological control agents and pesticides. Wiley, New York, NY. 723 pp. Crowley, D. J. 1954. Cranberry growing in Washington. Wash. Agric. Exp. Sta. Bull. 554. 30pp. Dhadialla, T. S., G. R. Carlson, and D. P. Lee. 1998. New insecticides with ecdysteroidal and juvenile hormone activity. Annu. Rev. Entomol. 43: 545-569. Dinham, B. 1993. The pesticide hazard: a global health and environmental audit. Zed Books: Atlantic Highlands, New Jersey. 228 p. Eck, P. 1988. Blueberry science. Rutgers University Press, New Brunswick, New Jersey. 284 p. Eger, J. E. and L. B. Lindenberg. 1998. Utility of spinosad for insect control in Florida vegetables. Proc. Fla. State Hort. Soc. 111: 55-57. Eigenbrode, S. D., and Espelie, K. E. 1995. Effects of plant epicuticular lipids on insect herbivores. Ann. Rev. Entomol. 40: 171-194. Elbert, A., H. Overbeck, K. Iwaya, and S. Tsuboi. 1990. Imidacloprid, a novel systemic nitromethylene analogue insecticide for crop protection. Proc. Brighton Crop Prot. Conf. – Pests and Diseases. 3: 21-28. Florida Agricultural Statistics. 2003. Florida Agricultural Statistics Service, 1222 Woodward St., Orlando, FL 32803. Food Quality Protection Act. 1996. P. L. 104-170. United States Congressional Board. Vol. 142: 110 Stal. 1489-1538. Gagn, R. J. 1989. The plant-feeding gall midges of North America. Cornell University Press, Ithaca, New York. 356 pp. Godin, J., P. Maltais, and S. Gaudet. 2002. Head capsule width as an instar indicator for larvae of the cranberry fruitworm (Lepidoptera: Pyralidae) in southeastern New Brunswick. J. Econ. Entomol. 95: 1308-1313.

PAGE 106

95 Groves, R. L., C. E. Sorenson, J. F. Walgenbach, and G. G. Kenedy. 2001. Effects of imidacloprid on transmission of tomato spotted wilt tospovirus to pepper, tomato, and tobacco by Frankliniella fusca Hinds (Thysanoptera: Thripidae). Crop Protection. 20: 439-445. Harris, M. O. and S. Rose. 1990. Chemical, color, and tactile cues influencing oviposition behavior of the Hessian fly (Diptera: Cecidomyiidae). Environ. Entomol. 19: 303-308. Horowitz, A. R., Z. Mendelson, P. G. Weintraub, and I. Ishaaya. 1998. Comparative toxicity of foliar and systemic applications of acetamiprid and imidacloprid against the cotton whitefly, Besmia tabaci (Hemiptera: Aleyrodidae). Bull. Entomol. Res. 88: 437-442. Isaacs, R., M. Cahill, and D. N. Byrne. 1999. Host plant evaluation behaviour of Besmia tabaci and its modification by external or internal uptake of imidacloprid. Physiol. Entomol. 24: 101-108. Kirk, W. D. J. 1995. Feeding behavior and nutritional requirements. In: B. L. Parker, M. Skinner, and T. Lewis (eds.) Thrips Biology and Management. Plenum Press, New York, NY, pp. 21-29. Knight, A. L., and L. A. Hull. 1989. Response of tufted apple bud moth (Lepidoptera: Tortricidae) neonates to selected insecticides. J. Econ. Entomol. 82: 1027-1032. Knight, A. L., T. R. Unruh, B. A. Christianson, G. J. Puterka, and D. M. Glenn. 2000. Effects of a kaolin-based particle film on obliquebanded leafroller (Lepidoptera: Tortricidae). J. Econ. Entomol. 93: 744-749 Langille, R. H. and H. Y. Forsythe. 1972. Biology of the blueberry thrips Frankliniella vaccinii and Catinathrips kainos in Maine (Thysanoptera: Thripidae). Can. Entomol. 104: 1781-1786. Lapointe, S. L. 2000. Particle film deters oviposition by Diaprepes abbreviatus (Coleoptera: Curculionidae). J. Econ. Entomol. 93: 1459-1463. Legaspi, J. C., B. C. Legaspi, Jr., and R. R. Saldaa. 1999. Laboratory and field evaluations of biorational insecticides against the Mexican rice borer (Lepidoptera: Pyralidae) and a parasitoid (Hymenoptera: Braconidae). J. Econ. Entomol. 92: 804-810. Lewis, T. (ed.) 1997. Thrips as crop pests. CAB International, New York, NY, 740pp. Liang, G. and T. Liu. 2002. Repellancy of a kaolin particle film, Surround, and a mineral oil, Sunspray Oil, to silverleaf whitefly (Homoptera: Aleyrodidae) on melon in the laboratory. J. Econ. Entomol. 95: 317-324.

PAGE 107

96 Liburd, O. E., S. R. Alm, R. A. Casagrande, and S. Polavarapu. 1998. Effect of trap color, bait, shape, and orientation in attraction of blueberry maggot (Diptera: Tephritidae) flies. J. Econ. Entomol. 91: 243-249. Liburd, O. E., and E. M. Finn. 2002. The status of blueberry gall midge in the southeastern United States. Florida Cooperative Extension Service ENY-825, Institute of Food and Agricultural Sciences, University of Florida. Liburd, O. E. and E. M. Finn. 2003. Small fruit pests and their management. In: J. Capinera (ed). Encyclopedia of entomology. Kluwer Academic Press, AADordrecht, The Netherlands. (in press) 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: 1-12. Liburd, O. E., H. J. McAuslane, B. J. Sampson, and K. E. MacKenzie. 2002. A multifaceted approach for control of blueberry pests in southeastern United States. United States Department of Agriculture Crops at Risk Grant. Gainesville, FL. 46 pp. Lyrene, P. M., and J. A. Payne. 1992. Blueberry gall midge: a new pest of rabbiteye blueberry in Florida. Proc. Fla. State Hort. Soc. 105: 297-300. Lyrene, P. M., and J. A. Payne. 1995. Blueberry gall midge: a new pest in rabbiteye blueberries. J. Small Fruit and Viticulture 2/3: 111-124. Mahr, D. L., and R. Kachadoorian. 1990. Cranberry tipworm. Proc. Wisconsin Cranberry School, Madison, WI. pp. 17-22. Mallampalli, N., and R. Isaacs. 2002. Distribution of egg and larval populations of cranberry fruitworm (Lepidoptera: Pyralidae) and cherry fruitworm (Lepidoptera: Tortricidae) in highbush blueberries. Environ. Entomol. 31: 852-858. McDonough, L. M., A. L. Averill, H. G. Davis, C. L. Smithhisler, D. A. Murray, P. S. Chapman, S. Voerman, L J. Dapsis, and M. M. Averill. 1994. Sex pheromone of cranberry fruitworm, Acrobasis vaccinii Riley (Lepidoptera: Pyralidae). J. Chem. Ecol. 20: 3269-3279. Mizell, R. F, and F. Johnson. 2001. Insect management in blueberries. Florida Cooperative Extension Service ENY-411, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL. Murray, D. A., R. D. Kriegel, J. W. Johnson, and A. J. Howitt. 1996. Natural enemies of cranberry fruitworm, Acrobasis vaccinii Riley (Lepidoptera: Pyralidae) in Michigan highbush blueberries. Great Lakes Entomol. 29: 81-86.

PAGE 108

97 NeSmith, D. S. (ed). Blueberry research: 1999 annual research update. 1999. University of Georgia. March/2002 http://www.ces.uga.edu/ES-pubs/RR662.htm NeSmith, D. S., G. Krewer, and J. G. Williamson. 1998. A leaf bud development scale for rabbiteye blueberry (Vaccinium ashei Reade). HortScience. 33: 757. Neunzig, H. H. 1986. The Moths of America North of Mexico. Fascicle 15.2 Pyraloidae: Pyralidae (part): Phycitinae (Part-Acrobasis and allies). Wedge Entomol. Res. Foundation. pp. 12-24. Polavarapu, S. 2001. The blueberry bulletin: a weekly update to growers. Rutgers Cooperative Extension of Atlantic County, 6260 Old Harding Highway, NJ 08330. 17: 1-4. Pree, D. J., K. J. Whitty, and L. Van Driel. 1998. Resistance to insecticides in oriental fruit moth populations (Grapholita molesta) from the Niagara peninsula of Ontario. Can. Entomol. 130: 245-256. Pritts, M. P., and J. F. Hancock (eds.). 1992. Highbush blueberry production guide. Northeast Regional Agricultural Engineering Service. Ithaca, New York. 200 pp. Puterka, G. J., D. M. Glenn, D. G. Sekutowski, T. R. Unruh, and S. K. Jones. 2000. Progress toward liquid formulations of particle films for insect and disease control in pear. Environ. Entomol. 29: 329-339. Sampson, B. J., S. J. Stringer, and J. M. Spiers. 2002. Integrated pest management for Dasineura oxycoccana (Diptera: Cecidomyiidae) in blueberry. Environ. Entomol. 31: 339-347. SAS Institute Inc. 2001. SAS System for Windows Rel. 8.2. SAS Institute, Inc. Cary, North Carolina, USA. Shorey, H. H. 1973. Behavioral responses to insect pheromones. Annu. Rev. Entomol. 18: 349-380. Simser, D. 1994. Parasitism of cranberry fruitworm (Acrobasis vaccinii; Lepidoptera: Pyralidae) by endemic or released Trichogramma pretiosum (Hymenoptera: Trichogrammitidae). Great Lakes Entomol. 27: 189-196. Spiers, J. M. 1978. Effect of stage of bud development on cold injury in rabbiteye blueberry. J. Am. Soc. Hort. Sci. 103: 452-455. Tomlinson, W. E. 1970. Effect of blacklight trap height on catches of moths of three cranberry insects. J. Econ. Entomol. 63: 1678-1679. Trisyono, A. and G. M. Chippendale. 1997. Effect of the nonsteroidal ecdysone agonists, methoxyfenozide and tebufenozide, on the European corn borer (Lepidoptera: Pyralidae). J. Econ. Entomol. 90: 1486-1492.

PAGE 109

98 VanRanden, E. J. and B. D. Roitberg. 1998. The effect of a neem (Azadirachta indica) based insecticide on survival and development of juvenile western cherry fruit fly (Rhagoletis indifferens) (Diptera: Tephritidae). Can. Entomol. 130: 869-876. Vernon, R. S. and J. S. Broatch. 1996. Responsiveness of Delia spp. (Diptera: Anthomyiidae) to colored sticky traps in flowering and rosette stage canola. Can. Entomol. 128: 1077-1085. Vernon, R. S. and D. R. Gillespie. 1990. Spectral responsiveness of Frankliniella occidentalis (Thysanoptera: Thripidae) determined by trap catches in greenhouses. Environ. Entomol. 19: 1229-1241. Waldstein, D. E. and W. H. Reissig. 2001a. Apple damage, pest phenology, and factors influencing the efficacy of tebufenozide for control of obliquebanded leafroller (Lepidoptera: Tortricidae). J. Econ. Entomol. 94: 673-679. Waldstein, D. E. and W. H. Reissig. 2001b. Effects of field applied residues and length of exposure to tebufenozide on the obliquebanded leafroller (Lepidoptera: Tortricidae). J. Econ. Entomol. 94: 468-475. Williamson, J. G., P. M. Lyrene, and E. P. Miller. 2000. A survey of blueberry acreage in Florida. Proc. Fla. State Hort. Soc. 113: 24-25. Wilson, B. W., M. J. Hooper, E. E. Littrel, P. J. Detrich, M. E. Hansen, C. P. Weisskopf, and J. N. Seiber. 1991. Orchard dormant sprays and exposure of red-tailed hawks to organophosphates. Bull. Environ. Contam. Toxicol. 47: 717-724. Wise, J. C., L. Gut, R. Isaacs, A. Schilder, C. Garcia-Salazar, and G. Thornton. 2002. Tree and small fruits insecticide/fungicide evaluation studies. Michigan State University, East Lansing, MI. Wise, J. C., L. Gut, R. Isaacs, A. Schilder, O. Liburd, and G. Thornton. 1999. Tree and small fruit insecticide/fungicide evaluation studies. Michigan State University, East Lansing, MI. Wood, G. W. 1956. Note on injury to blueberry sprouts by the blueberry thrips, Frankliniella vaccinii Morgan (Thysanoptera: Thripidae). Can. J. Agric. Sci. 36: 510.

PAGE 110

BIOGRAPHICAL SKETCH Erin Finn was born in Lansing, Michigan on September 25, 1979. She graduated from Grand Ledge High School, Grand Ledge, Michigan, in May 1997. Erin attended Lyman Briggs School at Michigan State University where she earned her BS in biochemistry in May 2001. While at Michigan State, Erin worked as an undergraduate research assistant with Dr. Oscar Liburd in the Entomology Department. After receiving her BS degree, Erin followed Dr. Liburd to Gainesville, Florida to pursue a master’s degree in entomology at the University of Florida. After completing her degree requirements, Erin will return home to Michigan and attend medical school at Michigan State University. 99