<%BANNER%>

A Study of the Behavior, Ecology, and Control of Flower Thrips in Blueberries Towards the Development of An Integrated P...

HIDE
 Title Page
 Dedication
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Abstract
 List of Figures
 Literature review
 Sampling techniques and dispersion...
 Pest phenology and species assemblage...
 Host status, injury description,...
 Efficacy of reduced-risk insecticides...
 Effectiveness of preventive and...
 General conclusions and experiences...
 References
 Biographical sketch
 

PAGE 1

A STUDY OF THE BEHAVIOR, ECOLOGY, AND CONTROL OF FLOWER THRIPS IN BLUEBERRIES TOWARDS THE DEVELOPMENT OF AN INTEGRATED PEST MANAGEMENT (IPM) PROGRAM IN FLORIDA AND SOUTHERN GEORGIA By HECTOR ALEJANDRO AREVALO RODRIGUEZ A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Hector Alejandro Arevalo Rodriguez

PAGE 3

To my family Hector, Olga and Pedro for standing with me all the way.

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank all the people that were involved in the development of this dissertation, principally my major professor Dr Oscar E. Liburd, and the members of my committee, Drs. J. Howard Frank, Frank Slansky Jr., and Paul M. Lyrene, for their close involvement. I would like to thank the sta ff of the Small Fruit and Vegetable IPM Laboratory at the University of Florida for their collaboration and support. Thanks to Aimee Fraulo for her collaboration with drawi ngs and maps. Special thanks go to all the blueberry growers who allowed me to use th eir farms to collect data. Thank you Florida Blueberry Growers Association, Sustainabl e Agriculture Research and Education Program (SARE), and the Environmental Prot ection Agency (EPA) for partially funding this project. Finally, I will like to acknowledge my famil y, Hector, Olga y Pedro who were with me all the way.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES.............................................................................................................x ABSTRACT.....................................................................................................................xiii CHAPTER........................................................................................................................ ..... 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................6 Blueberry History and Production Practices.................................................................6 Plant Selection and Common Varieties.................................................................7 Soil Management and Preparation.......................................................................10 Blueberry Pollination...........................................................................................11 Pest Complex in Blueberries......................................................................................12 Arthropod Pests...................................................................................................12 Blueberry Diseases..............................................................................................15 Thrips: Diversity and Ecology....................................................................................15 Behavior and Ecology.........................................................................................16 Dispersal behavior of thrips.........................................................................17 Population dynamics of thrips......................................................................18 Thrips as Crop Pests............................................................................................25 Thrips as Tospovirus vectors........................................................................26 Thrips in blueberries.....................................................................................27 Thrips control...............................................................................................28 Reduced-risk insecticides...........................................................................................30 Economic Injury Levels (EIL)....................................................................................31 Tables and Figures......................................................................................................35 3 SAMPLING TECHNIQUES AND DISPE RSION OF FLOWER THRIPS IN BLUEBERRY FIELDS..............................................................................................41 Materials and Methods...............................................................................................42 Methodology to Determine Thrips Popul ation Inside Blueberry Flowers..........42

PAGE 6

vi Vertical Distribution of Flower Thrips in Blueberry Fields................................43 Thrips Dispersion................................................................................................45 Results........................................................................................................................ .46 Methodology to determine thrips popul ation inside blueberry flowers..............46 Vertical Distribution of Flower Thrips................................................................46 Thrips Dispersion................................................................................................48 2004 farm FL02............................................................................................48 2004 farm FL03............................................................................................49 2005 farm FL02............................................................................................49 2005 farm FL03............................................................................................50 Discussion...................................................................................................................50 Tables and Figures......................................................................................................54 4 PEST PHENOLOGY AND SPECIES ASSEMBLAGE OF FLOWER THRIPS IN FLORIDA AND SOUTHERN GEORGIA IN EARLY-SEASON BLUEBERRIES.........................................................................................................69 Materials and Methods...............................................................................................70 Results........................................................................................................................ .72 Pest phenology:...................................................................................................72 Species assemblage:............................................................................................74 Rapid Determination of the Most co mmon Species Found in Early-Season Blueberry Fields...............................................................................................76 Discussion...................................................................................................................80 Tables and Figures......................................................................................................83 5 HOST STATUS, INJURY DESCRIPTION, AND DETERMINATION OF ECONOMIC INJURY LEVELS OF FL OWER THRIPS FOR EARLY-SEASON BLUEBERRIES.........................................................................................................89 Material and Methods.................................................................................................90 Host status of early-season blue berry bushes for flower thrips...........................90 Economic Injury Level and injury description....................................................91 Correlation between the number of thri ps inside the flowers and on sticky traps..................................................................................................................93 Results........................................................................................................................ .94 Host status of early-season blue berry bushes for flower thrips...........................94 Economic Injury Level and injury description....................................................94 Correlation between the number of thri ps inside the flowers and on sticky traps..................................................................................................................97 Discussion...................................................................................................................97 Tables and Figures....................................................................................................101 6 EFFICACY OF REDUCED-RISK INSECTICIDES TO CONTROL FLOWER THRIPS IN EARLY-SEASON BLUEBERRIES AND THEIR EFFECT ON ORIUS INSIDIOSUS, A NATURAL ENEMY OF FLOWER THRIPS..................104

PAGE 7

vii Material and Methods...............................................................................................106 Field Trials.........................................................................................................106 Laboratory Trials...............................................................................................109 Results and Discussion.............................................................................................112 Field trials..........................................................................................................112 Laboratory Trials...............................................................................................113 Thrips bioassay...........................................................................................113 Orius insidiosus bioassay...........................................................................114 Tables and Figures....................................................................................................118 7 EFFECTIVENESS OF PREVENTIVE AND INUNDATIVE BIOLOGICAL CONTROL TACTICS TO MANAGE FL OWER THRIPS POPULATIONS IN EARLY-SEASON BLUEBERRIES........................................................................123 Materials and Methods.............................................................................................124 Results and Discussion.............................................................................................125 Tables and Figures....................................................................................................129 8 GENERAL CONCLUSIONS AND EXPERIENCES WORKING WITH FLOWER THRIPS IN EARLY-SEASON BLUEBERRY FIELDS.......................134 Flower Thrips Monitoring and Sampling.................................................................135 Flower Thrips in Blueberries....................................................................................136 Flower Thrips Control..............................................................................................138 LIST OF REFERENCES.................................................................................................141 BIOGRAPHICAL SKETCH...........................................................................................153

PAGE 8

viii LIST OF TABLES Table page 21 List of diseases reported in blueberries in the United States....................................35 22 Number of pollen grains per flower for four plant species, and the extrapolated percentage of the grains th at could be eaten by five or 100 thrips per flower in three days (95% confidence limits)..........................................................................38 23 Some estimates of populati on parameters of pest thrips..........................................38 24 Known tospoviruses and thrips vectors in the world ..............................................39 25 Reduced-risk, biopesticides and OP a lternative insecticides registered or pending registration for use in blueberries...............................................................40 31 Distribution indices, Gr eens index (Cx) and Standard ized Morisitas index (Ip), used to describe the level of aggregat ion of thrips population on farm FL02 in Florida in 2004.........................................................................................................54 32 Distribution indices, Gr eens index (Cx) and Standard ized Morisitas index (Ip), used to describe the level of aggregat ion of thrips population on farm FL03 in Florida in 2004.........................................................................................................54 33 Distribution indices, Gr eens index (Cx) and Standard ized Morisitas index (Ip), used to describe the level of aggregat ion of thrips population on farm FL02 in Florida in 2005.........................................................................................................55 41 Pearson correlation coefficients for the relationship between percentage of opened flowers and thrips population captu red in sticky traps and inside five blueberry inflorescences...........................................................................................83 42 Dates, latitude, and pr incipal characteristics of fl ower thrips population in 2004 and 2005 from the samples taken from south-central Florida to southern Georgia.....................................................................................................................83 43 Distribution of the thri ps species complex in Flor ida and southern Georgia...........84 61 Proportion of Frankliniella bispinosa surviving at various times after the release of the insects in the bioassay aren as in essays conducted in 2004.........................118

PAGE 9

ix 62 Proportion of Orius insidiosus surviving at various times after release into bioassay arenas.......................................................................................................119

PAGE 10

x LIST OF FIGURES Figure page 31 Vertical distribution of thrips cap tured with respect to southern highbush blueberry bushes in south Florida............................................................................56 32 Vertical distribution of thrips captured with resp ect to rabbiteye blueberry bushes in southern Georgia......................................................................................56 33 Map of farm FL02 located at N 28 04 W 81 34 in north central Florida............57 34 Map of farm FL03 located at N 28 04 W 81 34 in north central Florida............58 35 Population dynamics inside the hot-s pot in coordinates (4, 4) of Figure 36 for 2004 on farm FL02.............................................................................................58 36 Number of thrips captured at 2 (a), 6 (b ), 8 (c), 10 (d), 14 (e), 16 (f), 18 (g), and 22 (h) days after bloom began on farm FL02...........................................................59 37 Number of thrips captured on farm FL03 at 2 (a), 4 (b), 8 (c), 14 (d), 16 (e), and 20 (f), days after bloom in 2004...............................................................................62 38 Population dynamics inside the hot-spot s in coordinates ( 2, 2), and (0, 4) of Figure 37 in 2004 on the fa rm FL03 in Florida.....................................................64 39 Number of thrips captured on farm FL02 at 2 (a), 4 (b), 8 (c), 10 (d), 14 (e), 16 (f), 18 (g), and 22 (h) days after bloom in 2005.......................................................65 310 Population dynamics inside the hot-spot in coordinates (2, 3), and (5, 2) of Figure 39 on farm FL02 in Florida in 2005...........................................................68 4-1 Phenology of thrips population on farm SFL01 in 2004..........................................85 4-2 Phenology of thrips population on farm SFL01 in 2005..........................................86 43 Phenology of thrips population on farm NCFL01 in 2004......................................86 44 Phenology of thrips population on farm NCFL in 2005..........................................87 45 Phenology of thrips population on farm SGA01 in 2004.........................................87 46 Phenology of thrips population on farm SGA01 in 2005.........................................88

PAGE 11

xi 47 Dates of first and last captures of fl ower thrips in the va rious blueberry sites. SFL01 represents south-central Florida, NCFL01 represents north central Florida and SGA01 represen ts southern Georgia.....................................................88 51 Average number of larvae emerged fr om individual tissues of 10 flowers and fruits.......................................................................................................................101 52 Linear regression showing the average number of thrips released per flower and the percentage of formed fruits in two principal cultivars of rabbiteye blueberries Climax and Tifblue........................................................................102 53 Various thrips injuries inflicted by flower thrips in blueberry flowers. a) Represents a healthy fruit, b) S hows feeding injury, and c) Shows oviposition/emergence injuries..............................................................................102 54 Regression illustration the number of thrips capture d on white sticky traps for a week in relation to the number of flower thrips captured in five inflorescences collected in the same bush......................................................................................103 61 Average growth rate ( r ) between the week before treatment application and the week after the application of the trea tments. Thrips populations correspond to the thrips recovered from the flowers collected in IFL04......................................120 62 Average growth rate ( r ) between the week before treatment application and the week after the application of the trea tments. Thrips populations correspond to the thrips captured in white s ticky traps collected in IFL04..................................120 63 Average growth rate ( r ) between the week before treatment application and the week after the application of the trea tments. Thrips populations correspond to the thrips captured in white s ticky traps collected in IGA04.................................121 64 Average growth rate ( r ) between the week before treatment application and the week after the application of the trea tments. Thrips populations correspond to the thrips captured in white s ticky traps collected in IFL05..................................121 65 Average growth rate ( r ) between the week before treatment application and the week after the application of the trea tments. Thrips populations correspond to the thrips captured in white s ticky traps collected in IFL06..................................122 66 Average growth rate ( r ) between the week before treatment application and the week after the application of the trea tments. Thrips populations correspond to the thrips recovered from in side the flowers in IFL06...........................................122 71 Average number of thrips captured per week after the release of natural enemies, as a preventive measure, in white stic ky traps located inside the blueberry bush in 2005....................................................................................................................129

PAGE 12

xii 72 Average number of thrips captured per week after the release of natural enemies, as curative measure, on white sticky trap s located inside the blueberry bush in 2006........................................................................................................................130 73 Average number of thrips captured per week after the release of natural enemies, as curative measure, inside five flowe r-clusters collected from blueberry bushes.131 74 Growth rate ( r ) of thrips populations captured in white sticky traps during the 2006 flowering season one and two weeks after the release of natural enemies as a curative alternative..............................................................................................132 75 Growth rate ( r ) of thrips populations collected inside blueberry flowers during the 2006 flowering season, one and two w eeks after the release of natural enemies as a curative alternative............................................................................133

PAGE 13

xiii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy A STUDY OF THE BEHAVIOR, ECOLOGY, AND CONTROL OF FLOWER THRIPS IN BLUEBERRIES TOWARDS THE DEVELOPMENT OF AN INTEGRATED PEST MANAGEMENT (IPM) PROGRAM IN FLORIDA AND SOUTHERN GEORGIA By Hctor Alejandro Arvalo-Rodrguez December 2006. Chair: Oscar E. Liburd Major Department: Entomology and Nematology Flower thrips are considered by growers as one of the key insect pests for earlyseason blueberries in Florida and southern Georgia. The objective of this study was to understand interactions between thrips a nd early-season blueberries and to develop strategies that can be used in an Integrated Pest Manage ment (IPM) program to control flower thrips in blueberries. This study in cluded the two blueberry species cultivated in Florida and southern Georgia, rabbiteye and southern highbush. The investigation began with the refinement of thrips sampling techniques and a study of dispersion of thrips in blueberry plantings. From these observations, I concluded that the distribution of flower thrips was highly aggregated in blueberry fi elds, which is an im portant factor when considering management strategies. I also deve loped a new system to collect thrips from inside blueberry flowers, which is more e fficient than flower dissection. This work continued with the analysis of population dyna mics and descriptions of thrips species

PAGE 14

xiv assemblage in blueberries. The results obtai ned showed a high correla tion between thrips populations and the latitude at which blueberry plantings are located as well as with the phenology of the flowers in blueberry bushes. I developed a dichotomous key for the six most common species of thrips found in blue berry fields during flowering. The damage inflicted by flower thrips on blueberries was also described. An ec onomic injury level analysis for two of the most popular cultivar s of rabbiteye blueberries was completed. A correlation between the number of thrips capture d in sticky traps and the number of thrips found inside the flowers was developed to improve monitoring efficiency. For the chemical control of thrips, I screened nine commercial and three experimental insecticides. From these trials, I concluded that acetamiprid is the most effective insecticide for thrips control. The reduced-risk insecticide spinosad is as effective as any of the other insecticides (except for a cetamiprid) and it is compatible with Orius insidiosus Say, one of the main natural enemies of thrips. Biocontrol trials did not show any advantages of mass releasing natural en emies as preventive or curative methods to control flower thrips in blueberries.

PAGE 15

1 CHAPTER 1 INTRODUCTION Blueberries belong to the genus Vaccinium in the family Ericaceae. This family also includes azaleas, cranberries, and huckleber ries as their most economically important species. There are more than 400 species of Vaccinium in the world, 26 of them are found in North America (Pritts and Strik 1992). Cultivated blueberries are native to North America, and have been dispersed around the world, principally to Europe and South and Central America. The United States (US) blueberry industr y started in 1908 when breeding programs were developed to improve wild species pres ent in New Hampshire and New Jersey. By 1916 the first harvest of the new crosses were released to the market, but it was not until the 1930s when several new blueberry cul tivars were released and commercial production of blueberries became popular. Despite the new and more productive cultivars today, only 70% of the fruits come from commercial blueberry cultivars, the remaining 30% are wild bluebe rries (Pollack and Perez 2004). Worldwide production of blueberries st arted around 1990. The world production of blueberries for 2003 was estimated at 238,358 metric tons (t), with the U.S. producing approximately 51.3% of the total production, followed by Canada (32.9%), and Poland (7%). Other countries that produce blueberries are Netherlands, Ukraine, and Chile, Argentina (Food and Agriculture Organizati on of the United Nations (FAO) 2004). In the United States, blueberry pr oduction was estimated at 113,800 t for 2004 (NASS-USDA 2006a). Michigan leads the na tional production with one third of the

PAGE 16

2 highbush blueberries in the country. Main e produces 80% of the total low-bush blueberries in the United States. In general, most of the states are increasing their overall production with exception of New Jersey, wh ich had a small reduction in the production since 2003 (NASS-USDA 2006a). For the 2004 and 2005 season, there was an increase in the production of fresh market blueberrie s, apparently due to a general increase in production in low producing states. USDA an ticipates a reducti on on the price of blueberries despite the incr ease in production over the next couple of years. The consumption of fresh blueberrie s per-capita is expected to re main close to 0.17 Kg / year in the United States. During the off-season (D ecember to March) the U.S. imports fruit principally from Chile, which is the principal provider of winter blueberries (NASSUSDA 2006a). There are only two cultivated bluebe rry species with low chill requirement, southern highbush ( Vaccinium corymbosum L.) and rabbiteye blueberries ( Vaccinium ashei Reade). Due to the environmental conditio ns these two species are the only species cultivated in Florida and sout hern Georgia. These states al ong with California, are the only producers of early-season blueberries (A pril to May) in the U.S. Early-season blueberries have prices that can be five to six times higher than mid-season blueberries. Although Florida represented only 1.18% of th e national production of fresh blueberries for 2003, it collected 8.24% of the money produced from blue berries in the nation. In Florida, the total acreage has increased 25% since 2001 and overall revenue has increased to more than $50,000,000 /year (NASS-USDA 2006a). There are 26 insect pests reported in blueberries for Florida (Mizell 2003). However, only four of them are considered as key pests for early -season blueberries.

PAGE 17

3 These species include blueberry gall midge, cranberry fruitworm, flower thrips, and blueberry maggot (Liburd and Arvalo 2006) The blueberry gall midg e [cranberry tipworm], Dasineura oxycoccana (Johnson) is the primary early-season pest in blueberry pl antings affecting up to 80% of the floral buds in susceptible cultivars (Lyrene and Payne 1996). Damage resulting from D. oxycoccana in blueberry plantings has increased significantly in the last 10 years. Floral and leaf bud injuries caused by D. oxycoccana were previously misdiagnosed as frost damage (Lyrene and Payne 1996). Many growers in Florida are replacing the susceptible rabbiteye blueberries with southern highbush, which is more tolerant to this pest (Williamson et al. 2000). In a recent study, Sarzynski and Liburd (2003) found that allowing adults to emerge from buds kept in storage bags at room temperature was the most effective technique for monitoring populatio ns in highly infested blueberry fields. Blueberry gall midge can damage developing vegetative buds after the harvest, which may affect the yield in the followi ng year (Liburd and Arvalo 2006). Flower thrips of the genus Frankliniella are another important pest of early-season blueberries. In (1999), the USDA reported that 40% of the losses in blueberries in Georgia were attributed to flower thrips. Three species of flower thrips have been reported repeatedly in blue berries throughout Florida and southern Georgia, and are F. bispinosa, F. occidentalis and F. tritici (Finn 2003, Liburd and Arvalo 2005). Thrips populations are known to move rapidly into blueberry fields with the help of wind currents and farm workers (Lewis 1997a). Th eir life cycles are extr emely short, taking only 15 days if environmental conditions are conducive for their growth and development. The short life cycles, as well as overlapping ge nerations during the

PAGE 18

4 blueberry flowering cycle, make this insect a dangerous pest that can reach economic damaging levels in a very short period (F inn 2003). Generally, information on the types of damage and behavior of flower thrips in early-season blueberries is limited. Sarzynsky and Liburd (2003) were only ab le to obtain initial but lim ited information on monitoring techniques for thrips. Cranberry fruitworm, Acrobasis vaccinii Riley, is found from Nova Scotia in Canada to Florida in the U.S. In the norther n states and Canada it has only one generation per year but a second generation is possible in the southern states. The larva feeds on the fruits and each one can damage as many as 10 fruits by feeding on them and secreting a web around the fruits, which make them unmarketable (Liburd et al. 2005). A very important late-season pest that has the potentia l to be a problem is the blueberry maggot, Rhagoletis mendax Curran. The females oviposit under the fruits exocarp. Seven to ten days later the larvae emerge and feed on the pulp of the blueberry for two to three weeks and th en drop to the ground to pupate. After overwintering, 80% of the pupae emerge the next season, 19% emerge in the second season after pupation, and the remaining 1% emerges four to five s easons later. Its damage is so severe that berries produced on farms in eastern and midw estern states must be certified maggot free to be able to be transported. There is zero tolerance to blueberry maggot in most fresh markets and a very low tolerance in proc essing markets. Some of the practices to reduce blueberry maggot infestations include sa nitation, collection of fruits after harvest and weed control (Maund et al. 2003, Liburd et al. 2005). To control these pests, farmers have reli ed on the intensive use of insecticides. However, as a response to the excessive us e of highly toxic pesticides and the public

PAGE 19

5 concern for a cleaner environment and hea lthier products, the Envi ronmental Protection Agency (EPA) created the Reduced-Risk Pe sticide Program in 1993, but not until 1996 was it formalized by the Food Quality Prot ection Act (FQPA) (1996). Some of the characteristics of reduced-risk insecticides include. Low effect on human health Lower toxicity for non-target organisms Low potential for groundwater contamination Low use rates Low pest resistance potential Compatibility with IPM practi ces as defined by EPA (2003) The main objective of this work was to study the biology, movement and the effects of flower thrips in commercial plantings of early-season blueberr ies. It is my goal to develop the fundation for an IPM program to control thrips in commercial early-season blueberry plantings in the southeastern Unite d States. The knowledge gained will be used to establish a management program i nvolving monitoring, use of reduced-risk insecticides and their effect on specific natura l enemies, use of biol ogical control, and the calculation of an Economic Injury Level to help farmers to make the decisions in controlling this pest.

PAGE 20

6 CHAPTER 2 LITERATURE REVIEW Blueberry History and Production Practices Blueberries and cranberries are some of the few fruit crops native to North America, along with blackberries, grapes, pa wpaw, and mulberry. Blueberries belong to the genus Vaccinium in the family Ericaceae and have been part of the North American tradition for centuries (Pritts and Strik 1992). Na tive Americans used almost every part of the plant for their consumption. Roots and leaves were used to make teas, the fruits were used for fresh consumption, and dried as seas oning for meats, incl uding beef jerky, and the juice was used as dye fo r various items including cove rs and clothing items. During the seventeenth century, English settlers le arned to cultivate blueberries from the Wampanoag Indians and to preserve the fruit (sun-dried) for the winter as a nutritional supplement. However, it was not until th e 1880s that the canned-blueberry industry started in the northeast United States (U.S. Highbush Blueberry Council 2002). In the early 1900s, Elizabeth White and Dr. Frederic Coville started the efforts to domesticate wild highbush blueberries in New Hampshire and New Jersey. By the 1930s the first group of domesticated blueberry cultivars was released. Today, there are 3 principal sources of germplasm for the blueberry cultivars: Vaccinium corymbosusm L. (northern highbush), V. ashei Reade (southern rabbiteye), and V. angustifolium Ait (lowbush blueberry). Despite the developmen t of many new cultivars and the effort put into improvement of blueberry quality, only 70% of the total producti on of blueberries in

PAGE 21

7 the U.S. is the product of commercial cultiv ars. The other 30% ar e the product of wild blueberries (U.S. Highbush Blueberry Council 2002, Pollack and Perez 2004). Interest in blueberries as a minor crop influenced horticultural departments to develop breeding programs at several land gr ant universities and colleges including Rutgers and Michigan State Universities. Th e breeding program at the University of Florida started in 1949. The objective of this program was initially to develop blueberry cultivars that could be produced commercially in Florida where the winter temperatures on average are above 13C. Two main t ypes of blueberries are produced by this program: the tetraploid highbush also known as southern highbush, based on crosses between V. corymbosum V. darrowi V. ashei. Several native blueberr ies have been used as gene sources for adaptation to the part icular conditions in the region (i.e., a low number of chill hours, warm conditions most of the year, soils with low organic matter and high bicarbonates). The most important spec ies used in this program as gene sources are V. darrowi, V. arboreum, V. corymbosum, among others (Lyrene 1997). The program has been very successful, allowing the Florida blueberry industry to develop as the principal producer of ear ly-season blueberrie s between April and May. This window of opportunity gave producers of early-season blueberries a price advantage of 3 5 USD more per pound than producers of midseason blueberries ( NASS-USDA 2006a). Plant Selection and Common Varieties Blueberries are a perennia l crop and with proper care can be productive for many years. The development of a successful crop st arts with varietal se lection. According to Lyrene (2005), there are four main obstacles that producers encounter when selecting the adequate variety for their farms. 1) The newe r varieties have the be st potential to be highly productive but a lot of information is still unknown. On the other hand, the older

PAGE 22

8 varieties are obsolete, but most of the information about their productive capabilities and needs in the field are well known. 2) The c hosen plants might not be available when needed and in the quantities needed. In some cases the order for new plants needs to be placed a year or more in advance to ensure th e availability of the varieties selected. 3) Blueberries in Florida need cross-pollination. It is necessary to select two or more varieties that are highly compatible to ensure maximum fruit-set. At the same time, it is necessary to have alternating rows of the varieties in the field, which complicates the management and harvest. 4) Finally, there ar e difficulties in using Dormex (Dormex Co. USA. LLC Parsippany, NJ) in the varieties. Th e use of Dormex is variety-specific, some varieties respond posit ively to the use of this plant gr owth regulator, while some present phytotoxicity, and the buds can be destroyed. Principal blueberry varie ties used in Florida are Star: This variety is planted from Ocala, FL to North Carolina. It is in the lateflowering group of blueberries in Florida. Th e fruit quality is high, with desirable size, firmness and flavor. In north-centr al Florida it is harvested in three pickings. This variety is usually considered as low-yield. However, it is very responsive to care during the previous fall, and requires applications of fungicides and fertilizers from the beginning of the crop. Emerald: It is considered as a high-yield vari ety. However, it ripens 7 to 10 days later than Star. Due to its productivity, a careful winter pruning is n eeded to ensure that the fruits left on the bush will set properly. This variety responds well to Dormex in productive areas south of Orlando, FL (Lyrene 2005).

PAGE 23

9 Jewel: This variety is especially popular in north and centr al Florida. Jewel fruits start ripening about five days after Star, approximately at the same time as Emerald. In north and central Florida, Jewel may produce too many flower buds; therefore, it needs winter-pruning to remove weak branches that may fail to carry the fruits during harvest. Jewel responds well to a pplications of Dormex TM. When harvesting, it is necessary to leave the fruits on the bush for extended periods of time because the fr uit has a tart flavor. Leaving the fruits on the bush allows them to accumulate sugars, which eventually decreases their tartness. Millennia and Windsor: These two varieties used to be popular in Florida. However, their popularity was reduced due to several factors. Millenia and Windsor are early-ripening varietie s with good yields. The problem w ith Windsor is that, if the weather is too hot or the fruits are not picked on time, fruit scars develop, which could be a problem. Ideally, fruits should have a sma ll and dry scar between the pedicle and the fruit, which will prolong their shelf life. Millennia has excellent fruit characteristics, but has problems with fruit-setting and is highly su sceptible to botrytis, a key fungal disease, during flowering. The recommendations for varieties in Florida depend on the tests and knowledge gathered year after year. However, varie ties have been divided into three groups: Obsolete, Core, and New. The actual recomme ndation is to plant 75 % of the area with core varieties (Star, Emerald, and Jewel for Florida) and 25% with new varieties that show good potential. Among these new variet ies: Springhigh, Springwide, Abundance, Sapphire, and Southern Belle are the ones with the highest potential. All of them are low-

PAGE 24

10 chill varieties with desirable characteristics of fruit quality, ripening and fruit set (Lyrene 2005). Soil Management and Preparation Blueberries prefer porous, well-drained soils (sandy-loam or loamy-sand) with high organic matter and a low pH. It appears that there is a positive co rrelation between good growth of southern highbush and the amount of sand in the soil a nd a negative correlation with the amount of clay and silt (Korkac 1986). In heavy soils, such as southern Georgia, blueberries might take longer to reach matu rity but once they ar e mature the production will be similar to that of other soil types (Williamson et al. 2006). For blueberry bushes to achieve their life expectancy, 50 years, it is necessary to se lect a good place to plant this crop. This place must have low risk of fr eeze injury, adequate soil conditions, and easy access to water. The soil management pract ices described by Williamson et al. (2006) and the actual recommendation for soil preparati on of blueberries in the southeastern U.S. region are as follows: 1. Collect and analyze soil samples, determ ine drainage and options to improve it 2. Clear and drain the land 3. Incorporate sulfur, phosphorus, and organi c matter as recommended by the soil analysis 4. Construct beds so that plants will have at least 18 inches of well-drained soil Plants should be spaced 0.6 to 1.2 m appart in the row with 2.7 to 3.3 m between rows; this density will average 2,400 plants per ha. In the southern U.S. the use of pine bark has become a common practice. The high cost of using pine bark mulch makes it exclusive for early-season blueberry` producer s in Florida and sout hern Georgia. The pine needs to be replaced every 3 to 4 year s. In most cases the roots of the plants concentrate in the pine bark and usually do not penetrate the soil. This method makes it

PAGE 25

11 possible to produce blueberry in soils that are otherwise not adequate for its production (Williamson et al. 2006). Blueberry Pollination Due to structure and genetics, early-season blueberries need to be cross-pollinated to achieve an adequate fruit set. Blueberry plants are known to be entomophilous and depend mostly on hymenopterans for their pol lination. Adequate pollination must occur between three and six days after the stigma is receptive, otherwise the fruit set will be poor. In North America, some hymenopterans have co-evolved with blueberries. This is the case of bees from the genera Osmia and Habropoda which pollinate only blueberries and their entire life cycle is coordina ted with blueberry bloom (Yarborough 2006). However, the commonest bee used by commercial growers is the honey bee, Apis mellifera L. Because A. mellifera is commercially available, it reduces the growers dependency on fragile natural populations of bees for their blueberry pollination, ensuring a high fruit setting. The recommendation for the number of beehives in blueberries is between 50,000 and 150,00 bees per ha for northern highbush (Yarborough 2006). In the case of rabbiteye blueberries, three main pollinators can be used successfully in the field: Osmia ribifloris Cockerell, Habropoda laboriosa (F.), the southeastern blueberry b ee, and the honey bee, A. mellifera. Sampson and Cane (2000), compared the pollination efficiency of O. ribifloris H. laboriosa and A. mellifera in three of the core cultivars for rabbiteye, Ti fblue, Premier, and Climax. They found no significant difference in fruit set among the va rieties. However, when all the varieties were averaged, there were significant differen ces in terms of which species prefers which variety. Tifblue was reported to have a good response to O. ribifloris and H. laboriosa

PAGE 26

12 but not to A. mellifera. Premier has a high percentage of fruit set when H. laboriosa and A. mellifera are used but not when O. ribifloris is the pollinator Finally, Climax seems to have a good response to all the bee species used in the trial (Sampson and Cane 2000). Experiments conducted by Sampson (unpublished data) found that the optimal number of bees needed for maximum fruit set is between 5 and 6 bees per 1,000 opened flowers, this means 5 to 6 bees per bush during peak polli nation. However, the number of bees needed should be studied on a farm-to-farm basis. Pollinators tend to ge t distracted by other nectar sources around the blueberry crops, ther efore the distribution of the hives and the number of hives needed depend on the conditi ons of the farm and the natural population of bees in the surrounding areas. Unlike honey bees, indigenous genera of bees including Andrena Halictus Bombus Lasioglossum use sonication to harvest the po llen from certain plant species including plants of the genus Vaccinium This sonication or buzz-pollination significantly increases the amount of pollen collected and thus the amount of pollen transported to other flowers (Cane and Payne 1988, Javorek et al. 2002). Javorek et al. (2002), compared various bee genera to determin e which ones were the most efficient in collecting and pollinating lowbush blueberry fl owers. They found that, for example, a honey bee, which does not use sonication, will n eed to visit a flower 4 times to harvest the same amount of pollen than a bumble bee, which uses sonication; at the same time, bumble bees (97%) pollinate close to 4 tim es more flowers than honey bees (24%). Pest Complex in Blueberries Arthropod Pests Seven arthropod species are c onsidered as major pests in blueberries. However, only four of them are considered as key pest s for early-season bluebe rries in Florida and

PAGE 27

13 southern Georgia. These in sects are blueberry maggot, Rhagoletis mendax Curran, blueberry gall midge, Dasineura oxycoccana Johnson, cranberry fruitworm, Acrobasis vaccinii Riley, and flower thrips Frankliniella spp. (Liburd and Arvalo 2006), which are discussed below. Blueberry maggot: Rhagoletis mendax Curran (Diptera: Tephritidae). This insect is the principal pest in blueberries in the eas tern U.S. (Liburd et al. 1999). It is found in all blueberry regions east of the Rocky M ountains. This species belongs to the same genus as the apple maggot, Rhagoletis pomonella (Walsh). Both of them are host-specific and concentrate their damage on the fruits, making these unmarketable (Liburd et al. 1999). Their damage is so devastating th at USDA has imposed restrictions on the transport of blueberries, and some stat es, such as Florida, that do not have R. mendax reported, have a zero-tolerance policy for th is insect (Liburd et al. 1998, Liburd and Arvalo 2006). Monitoring for this particular species is based on the use of yellow sticky boards, green sticky spheres and red sticky sp heres. There must be a minimum of one trap per ha and one of the traps should be placed within 18 m from the border of the crop. The economic threshold for R. mendax has been defined as 2 flies per trap per week (Liburd et al. 2000, Liburd et al. 2006). Manageme nt strategies for this tephritid include reduced-risk insecticides, insecticide-treated spheres, and natural enemies such as Diachasma alloeum Muesebeck (Hymenoptera: Braconidae) which is being researched in order to improve its efficiency controlli ng this pest (Liburd and Arvalo 2006) Blueberry gall midge: Dasineura oxycoccana (Diptera: Cecidomyiidae). This species, formerly known as cranberry tipworm, was recently reported a pest of blueberries in southeastern crops (Lyren e and Payne 1992, Sarzyns ki and Liburd 2003).

PAGE 28

14 After their discovery, it was es tablished that, if left unman aged, blueberry gall midge can destroy up to 80% of the crop production (L yrene and Payne 1996). Eggs hatch between 2 and 3 days after oviposition, then larvae feed on the young buds, killing them and preventing the formation of new flowers and leaves. Post-harvest damage could potentially affect the yield for subsequent years (Liburd and Arvalo 2006). To monitor blueberry gall midge, Sarzynski and Liburd (200 3) established that collecting buds from the field, 20 buds per ha, and placing them in zip-lock bags for 14 days is the most accurate way to determine the presence of this species. The recommendations to control this insect rely on the use of reduced-risk insecticides for fields with a history of blueberry gall midge or, even if the presence of this species has been confirmed (Liburd and Arvalo 2006). However, six natural enemies of D. oxycoccana have been identified, but none of them has been made commercial and studies about th eir efficacy managing the pest are still under wa y (Sampson et al. 2006). Cranberry fruitworm: Acrobasis vaccinii (Lepidoptera: Pyrali dae) is present in all the places where its host plants are pres ent in North America. Its host plants include huckleberries, dangle-berries, beach plumbs apples, cranberries, and blueberries (Beckwith 1941). It is consid ered a key pest for blueberries. Each larva can damage between 5 and 10 fruits during its development, and the affected blueberry clusters will present webbing and berry deformation. The use of pheromones in sticky traps is recommended to monitor the activity of th is insect, and based on the observations on these traps insecticide applications might be necessary at the beginning of the flying season (Liburd and Arvalo 2006).

PAGE 29

15 Blueberry Diseases Blueberries are susceptible to many diseases Most of them can be prevented with optimal management of the crop and with good decisions at the moment of selecting varieties and plots to be used for blueberry production. Blueberries are susceptible to diseases caused by fungus, viruses, phytoplasma s, bacteria, nematodes, dodder, and some physiological disorders. Some diseases have b een attributed to abio tic factors such as deficient nutrition, freeze, and poor water ma nagement, among others. A summary of the diseases reported for blueberries is found in Table 21 Thrips: Diversity and Ecology Thrips belong to the order Thysanoptera, which literally means fringed wings. However, the English name for thrips is derived from the Greek word for woodworm, because early naturalists found various species in dead branches (Mound 2005). Thysanoptera are characterized by fringed wi ngs in the adult stage, and asymmetric mouthparts (Triplehorn and Johnson 2005). The left mandible is the only one that develops because the right one is resorbed by the embryo (Heming 1993). The mouthparts of this order have been describe d as punch and suck. The mandible is used to break the external layer of plant cells or pollen grains and the contents are sucked through the maxillary stylets, which are joined to form a tube (Triplehorn and Johnson 2005). The order Thysanoptera is divided into tw o suborders, Tubulifera, one family, and Terebrantia, eight families worldwide. Th e females in Tubulifera do not have an ovipositor and the distal abdominal segment is similar to the males. This segment is tubular in shape and ends in a series of se tae. The forewings in Tubulifera have neither venation nor setae except for the base. Terebr antia are the most common suborder and the

PAGE 30

16 one that has the greatest effect on agriculture. Close to 94% of the total pest species are in this suborder, all of them in the fa mily Thripidae (Moritz et al. 2004b). The metamorphosis of thrips is intermediate. There have been some discussions as to whether thrips should be classified as holometabolous or hemimetabolous. The first two instars do not have external wings, b ecause they are being developed internally. Usually these two instars are called larvae and resemble hol ometabolous metamorphosis. Thrips display two more distin ct immature stages, which sh ow vestigial wings but they do not feed. The first of these stages is called propupa, which shows vestigial wings (except in Tubulifera). Following the propupa stag e, a real pupa is formed, similar to the adults. The main differences between pupa and adult are that the pupa are not mobile, has two pairs of vestigial wings, and the antenna has fewer segm ents, while the adults are mobile, macropterous have well formed wi ngs and have between 6 and 9 antennal segments. The pupa is also inactive but ha s external developmen t of the wings and morphological resemblance to the adult st age, which refers to a hemimetabolous metamorphosis (Triplehorn and Johnson 2005) Propupa and pupa differ from the adult stage in the size, morphology a nd functionality of the wings, and the segments in the antennae and legs are reduced. Pupa a nd propupa are immobile (Moritz 1997). Behavior and Ecology Based on thrips alimentary preferences, they can be divided into fungivorous, phytophagous, predacious, and omnivorous sp ecies. The damage caused by thrips in agricultural crops is primarily due to feeding on leaves, flowers or fruits, and secondarily to oviposition in these same structures (Kirk 1995). Howeve r, little is known about the feeding behavior of these insects in the fiel d including diets, host-sw itching behavior, etc. (Kirk 1997a). In the case of Te rebrantia, more than 95% of the species are associated

PAGE 31

17 with green plants. However, in most of the cases the host report is based on the places where thrips are found and not on the place s where they breed, making most of the records confusing and the definition of host plant very subjective (Mound 2005). Dispersal behavior of thrips Thrips in general have two means of disp ersal, artificial and natural dispersal. Artificial dispersal is usually human-assist ed and is facilitated by the increasing international transportation of agricultural products. Thrips are easily transported in various products including potte d and cut flowers and several fruits and vegetables that are imported and exported. Accidentally transp orting thrips across borders is relatively easy. They are difficult to spot in a port insp ection due to their small size. Furthermore, the eggs of these insects ar e found inside plant tissues and the signs left by the ovipositing female are minimal. The second method, natural dispersal, is accomplished by thrips using natural means and the most common method is flying. Ju st before to flying, thrips have a very complicated preparation for takeoff. Duri ng this period macropterous forms bend their abdomen and use setae located on abdominal te rgites V to VIII to comb those located on the wings. The objective of this movement is to increase the surface-area of the wings, facilitating take-off (Ellington 1980). Thrips ha ve been reported to fly at 6 to 30 ms-1 depending on the species. However, it is known that thrips disperse to distances further than that which they would be able to indepe ndently fly. One of the explanations for this phenomenon is the use of wind currents (Lew is 1997a). Thrips, like many small insects, potentially use wind currents to move l ong distances. This phenomenon is summarized by Gatehouse (1997). Small insects might ta ke advantage of the convective upper currents developed by warm air-pockets, which have speeds of ascension measured up to

PAGE 32

18 3m s-1 (Drake and Farrow 1988). These currents he lp the insects to reach the Flight Boundary Level (FBL) for each insect species, which might run into the Planetary Boundary Level (PBL), which is the layer be tween the ground and the free atmosphere. This PBL is located between 100 and 3000m a bove the ground and once the insects break it, it facilitates their dispersal. Radar data s how that most of the ma ssive flights are shortlived; but some of the populat ions can travel overnight (Gatehouse 1997). One of the main disadvantages of this mode of transporta tion is that insects have little to no control of the direction that they are being transported. However, despite the small size of thrips and their apparent lack of cont rol of their flight patterns due to wind interaction, there is good evidence indicating th at thrips have a certain amount of control in the landing. Field observations indicate that th rips land on their feet on indi vidual plants, showing some amount of control (Lewis 1997a). Kirk (1984) de monstrated that thri ps have control of their landing selection. The author used va rious colored traps on the ground separated by 5 m from each other to show that there wa s a 20-fold difference between flower thrips and grass-dwelling thrips in their color selec tion for landing. Flower thrips were attracted to bright colors such as white while grass-dw elling thrips were attracted to colors that were closer to green (Kirk 1984, Teulon and Pe nman 1992). There is evidence that thrips are attracted to various odors, the use of anisaldehyde (for flower thrips), or ethyl nicotinate (for Thrips obscuratus ) increased the trapping of the respective thrips compared with the controls (Kirk 1985, Teulon 1988). Population dynamics of thrips To understand the population dynamics of thri ps, it is necessary to understand their relation with their host plants. There are two types of plants where thrips have been reported. The first type is a provisional or alternate host, which might offer temporary

PAGE 33

19 shelter or food, but in the vast majority of cases thrips do not repr oduce in these plants. The second group of hosts might be called prope r hosts; these plants offer food, shelter, a reproductive substrate and alimentation for the immature thrips. Unfortunately, there is a controversy about whether the pl ants reported as hosts in the literature are proper hosts or alternate hosts, and if these alte rnate hosts should be defined as hosts or if they are just accidental relationships (Mound 2005). Ther e are approximately 50 economically important pest species among 5,300 known species of thrips. Some thrips species are considered to be very host-specific. Those th rips species that are considered as crop pests are usually very prolific a nd non-host-specific. For example Frankliniella occidentalis (Pergande), the western flower thrips, is re ported on more than 500 plant species within 50 families. However, it is necessary to reme mber there is controversy about reports of host plants (Moritz et al. 2004b). Thrips are ideal for population dynamics st udies. Their populations are large and are generally easy to sample. However, thrips sampling presents some challenges, such as the difficulty of finding dead thrips and the f act that big migrations go unnoticed most of the time, sometimes for unknown reasons. Some species are very common in one year and very rare in the next. To study their popu lation dynamics, it is necessary to consider feeding and reproductive behavior, migrati on, short and long term effects of the environment, and the effect of management techniques in the fi eld populations (Kirk 1997b). Feeding behavior: The mouthparts of Thysanopter a are one of the identifying characters of this order. The mouthparts are located on the unders ide of the head and form a mouthcone. This structure is form ed by a single mandible (characteristic of

PAGE 34

20 Thysanoptera) and two maxillary stylets. In order to feed, thrips use their mandible to punch a hole in the external walls of th e tissue that they are going to feed upon and then use the stylets to suck the liquids from inside these tissues (Kir k 1997a). In the past, feeding behavior of thrips was considered to be rasping or gashing and sucking. However, this observation has been re-evaluat ed and the feeding is considered to be the piercing and sucking type (Hunter and Ullman 1992). Thrips in general, can feed on diverse plan t tissues (leaves, flow ers, fruits, pollen) and some fungal tissues such as spores a nd hyphae. Most of the attention has been focused on the feeding behavior of phytophagous species, thus this is the group of which we have broader knowledge of their pref erences (Kirk 1995). The feeding behavior displayed by thrips is similar for all plant tissues. Once the thrips have landed on what seems to be an appropriate substrate on which to feed on they start the process of probing the tissue. They start using the legs and an tennae, walking in circles or forming figure eights on the tissue. Once they find a spot th at seems to be adequate, thrips use their mandible to probe and open a small hole in the cell wall. A small amount of liquid comes from this small puncture. Using their palps, th rips test the liquid for the correct nutrient compositions. If the tissues and nutrient composition are adequate, they use their mandible and head to punch a bigger hole in the tissue and start feeding. This causes nearby cells to collapse. If the damage occurs in the ovary in the flower these marks will become magnified during the fruit development and the scars will be very noticeable, reducing fruit quality (Kirk 1997a, Liburd and Arvalo 2006). Pollen-feeding is another behavior th at is common, princi pally among flower thrips. These thrips feed on individual pollen grains one by one. The time spent on each

PAGE 35

21 pollen grain varies between 3 to 120 s depe nding on thrips species and instar, grain volume, and temperature (Kirk 1987). Thrips can ingest pollen from the anthers or the grains found around the flowers and leaves. The potential damage that flower thrips can have on pollen quantity depends upon the plant production of pollen and thrips populations present in the field as observed in Table 22. As observed in Table 22 there is the potential that thrips may affect the availability of pollen for fertilization. Ho wever, thrips populations will need to be extremely high and the pollen production by the pl ant very low for this to occur. Based on calculations presented by Ki rk (1987), one thrips could potentially destroy between 0.2-0.7% of the pollen in a flower per day, assuming that it fed exclusively on pollen. Furthermore, thrips might be responsible for the destruction of anthers or the destruction of pollen on stigmas, which would affect pollination. The damage caused by thrips on plant fertilization depends on many factors such as timing, amount of pollen produced by the plant, amount of pollen destroyed by thrips effectiveness of pollinators, temperature, etc. (Kirk 1987). In addition to interfering with pollen availability a nd fertilization, thrips balance their diet by consuming other plant tissues (Kirk 1997a). Because thrips are usually a ssociated with plant pests, they have been overlooked as pollinators, and there are no studies about their efficienc y. However, to determine the correlation between pollinators and flowers, a chart of pollination syndromes describes the characteristics of flowers that may attr act certain types of pollinators (Kirk 1997a). Thrips pollination syndrome is called thri pophily (Kirk 1988). Thrips flowers are described by Kirk (1997a) and Mondal et al. (1993) as medium size, white to yellow, sweetly scented, with or wit hout nectar, with compact flor al structures or globose or

PAGE 36

22 urceolate blossoms providing shelter, and with small to medium-sized pollen grains, possibly with nocturnal pollen pr esentation. This description is very close to blueberry plants, which have medium-sized white fl owers with nectar, gl obose blossoms that provide good shelter for thrips. So blueberry fl owers meet the criteria to be pollinated by thrips; however, more research is needed to de termine the role thrips play in pollination. It is difficult to determine the net eff ect of thrips on the flowers taking into consideration the benefits of pollination and the damage to floral structures. Several species of plants have been re ported to be pollinated by thrips. Peltophorum inerme (Roxb.) Llanos, is pollinated by two species of thrips, Thrips hawaiiensis (Morgan) and Haplothrips ceylonicus Schmutz, in addition to various hymenopteran species (Mondal et al. 1993). Erica tetralix L. is not only pollinated by Taeniothrips ericae (Haliday), but they have a close mutualisti c relationship. Flowers of E. tetralix offer protection to thrips from the environment and a place to reproduce, the insect offers the plant self and cross pollination (Hagerup and Hagerup 1953). Another feeding behavior shown by thrips is predation. There are a few specialist predators among thrips that ha ve some behavioral adaptations such as speed or color among others. Among the specialists the most common prey are mite motiles and eggs. Some of the most common speci es of predatory thrips are Haplothrips kurdjumovi Karny, which feed on moth and mite eggs (Putnam 1942), Scolothrips sexamaculatus (Pergande), which feed on mites that fo rm webs (Trichilo and Leigh 1986), and Trichinothrips breviceps Bagnall which feed exclusively on psocids (Kirk 1997a). Some species of thrips feed upon other thrips larvae. Including Aeolothrips intermedius Bagnall, which feeds on thrips immatures thr ough their abdomen (Kir k 1997a). Some of

PAGE 37

23 the polyphagous thrips are well known as pests but they can switch their preferences and become predatory. For instance, Frankliniella occidentalis (Pergande) feed on mites in cotton (Trichilo and Leigh 1986) and prey on twospotted spider mites. Thrips tabaci (Lindeman) is considered to be a pest of ve getables susceptible to tospoviruses, but preys on twospotted spider mites in Au stralia (Wilson et al. 1996). Reproductive behavior: Thrips in general have short life cycles. Many environmental factors can affect the reproduction rate and the length of their life cycle. One of the most important factors are host pl ants. Plant species and quality (age, vigor, phenological stage, etc.) affect the net reproductive rate (R0) of thrips populations (Table 23). Abiotic conditions aff ect the reproduction of thrips as well. These include light regimen, temperature, and humi dity among others. Kirk (199 7b) presents a summary of the effects of plant species and conditions that affect thrips reproducti ve behavior. The effect of plant quality on th rips populations is very important. As observed in Table 23, plant species influe nce the life history of some species. Some thrips have particular preferences towards the qualit y of the tissues used for oviposition. For example, Taeniothrips inconsequens (Uzel) will only lay its eggs on convex structures like veins in the leaves or stems (Teulon et al. 1994). Bates and Wei ss (1991) showed that Limothrips denticornis Haliday only lay their eggs on the intervein space of barley leaves, limiting the oviposition to mature leaves. Furthermore, Chau et al. (2005) showed a close correlation betw een the populations of F. occidentalis and the level of nitrogen fertilization in chrysanthemum. The experi ment reported an increase in the number of thrips correlated to the level of nitrogen used up to 100% of the recommended dosage for this crop.

PAGE 38

24 Many intrinsic behaviors help us to unders tand the relationships among thrips from the same or from different species. The firs t one is the use of semiochemicals such as alarm pheromones, aggregation pheromones, defensive mechanisms, etc (Terry 1997). Kirk and Hamilton (2004) demonstrated the existence of some type of substance produced by males of F. occidentalis that has an attractive eff ect on females of the same species. Unfortunately, identification of th e compounds in the pheromones is still in progress, but the description of the behavior of females, virgin females and males in a Ytube bioassay indicate that this might be a sex-pheromone Milne et al. (2002) observed what seems to be some type of attraction pheromone produced by males for females from the same species, Frankliniella schultzei (Trybom), and a direct correlation between the number of females per male attracted and the number of males present. There is not yet evidence of sex-pheromones that work at long distances, but there is some indication of short-range attractants that might help th rips to locate their mates (Terry 1997). Apparently due to the bias in the female: male ratio, males and females have exhibited different behaviors to locate each other. Frankliniella occidentalis males tend to aggregate on the external side of floral st ructures where the females might be attracted, probably by the substances descri bed in Kirk and Hamilton (2004) Some females are attracted while others ignore the signals a nd move themselves towards the food sources inside the flowers. Because females do not need to mate to lay fertile eggs, mating behavior in this order of insects is comple x and not completely understood; their behavior is very inconsistent and species specific, ma king it difficult to stat e generalizations about this topic (Terry 1997).

PAGE 39

25 Unlike mating, oviposition behavior is more general and well-described at least for terebrantian species. These females raise the ti p of the abdomen, test the tissues using the setae in the last abdominal segment, and inse rt the ovipositor into selected plant tissues. While in this position the saw-like ovipositor cuts a space for the egg in the tissue, which is pushed out by a contraction of the abdomen. Thrips prefer to lay their eggs in mature non-expanding tissues to avoid having the e ggs crushed by the expanding cells (Terry 1997). Oviposition preferences depend on the speci es. Most species prefer to oviposit on leaves or on floral tissue. In citrus, F. bispinosa oviposit in the floral tissues, it has a preference for the pistilcalyx area followed by the petals and finally, filaments and anthers (Childers and Anchor 1991). In apples, F. occidentalis prefers to lay its eggs in blossoms of any age, although most adu lts are found in opened blossoms, the egg concentration is higher in peta lless clusters mainly in th e king bud (Terry 1991). Other thrips species lay their eggs close to the inne r veins of the leaves or in the fruits. The damage caused by these thrips due to oviposit ion depends on the place and plant stage selected for oviposition. Thrips th at lay their eggs and feed in the commercial part of the plant, flower or fruits, are th e ones that are considered as ma jor threats to the agricultural industry independent of their role as virus vectors. Thrips as Crop Pests Monophagous thrips are rarely considered as pests. Only a few examples of this interaction are known: Liothrips karnyi Bagnall, which damages Asian piper, L. adisi Strassen, which feeds on Brazi lian guarana trees, and Sciothrips cardamomi Ramakrishna a common pest of cardamom (Mound 2005). Most of the thrips that are considered as severe pests are polyphagous. Due to their hi gh adaptability, they can feed on various resources and modify their larval stages, adap ting to various temperature ranges, etc. Due

PAGE 40

26 to their high plasticity, agricultural systems should not be looked at as if they were isolated islands (Altieri 1988). Thrips ar e notorious for moving their population to alternate hosts during the season when the main hosts are not very conducive. This is the case of flower thrips, which reproduce and f eed in the flowers of our crops and then during the season when flowers are not presen t, they migrate to nearby crops and wild flowers to continue th eir cycle (Kirk 1997b). Thrips as Tospovirus vectors Tospoviruses are one of the most damaging groups of pathogens in agriculture. In recent decades, due to the increase in international trad e, the spread of infected plants and vectors has increased worldwide. Thrips and viruses are probably two of the most difficult things to detect in the ports of entr y. Thrips eggs inside the plant tissues as well as asymptomatic plants infected with the viruses are virtually impossible to detect (Lathman and Jones 1997). There are 16 species of viruses in the genus Tospovirus family: Bunyaviridae, recognized as plant pest s, and they are transmitted by 11 species of thrips, of the family Thripidae. However, th e list of viruses and v ectors changes due to the complicated genetics of the virus and the discovery of new relationships with various thrips species (Ullman 2005). Thrips acquire viruses in the first or ea rly second instar when there is a close relationship between mid-gut, visceral mu scles and salivary glands. Once the wing muscles start developing and the supra-oes ophageal ganglion moves towards the head the connection between the salivary glands, the mid-gut, and the visceral muscles is ended stopping the movement of virus particles into the salivary glands. If the thrips did not acquire the virus during this sh ort period, it will not be able acquire the virus due to the lack of connection between the salivary glands and the mid-gut. In ad ult thrips the virus

PAGE 41

27 is located in the malpighian tubes, in the lumen, the hemocoel, and in the salivary glands. Until recently, the only proven way thrips transmit the virus is through the salivary glands during feeding. However, there is e nough evidence to support th e possibility that the virus might be transmitted through excrements and oviposition wounds, but more research is needed (Moritz et al. 2004a). Thrips in blueberries To study the relationship between thrips and blueberries we can divide thrips into three groups, which is a very broad division f ound in the literature: leaf thrips, flower thrips, and flower and leaf thrips (Kirk 1997a, Liburd and Arvalo 2006). Leaf thrips: Frankliniella vaccinii Morgan and Catinathrips kainos ONeill are the two main leaf pests of blueberr ies in northeastern U.S. They feed on the leaves right after pruning and their larvae are found feeding inside curled leaves, which prevent them from developing properly. In Maine, these thrips ar e found during the summer from late July to early August after the pruning. Af ter the damage is done, the larvae mature and the adult thrips migrate to other hosts disappear ing until next season (Collins et al. 1995). Flower thrips: Unfortunately, the relationship between flower thrips and blueberries is not well known. In an in terview conducted by Finn (2003), growers of early-season blueberries consider ed flower thrips as one of the most important pests of his crop along with blueberry gall midge a nd blueberry maggot. The USDA reported in (1999) that 40% of the losses in blueberries in Georgia were attributed to flower thrips. Thrips populations are known to rapidly move in to blueberry fields w ith the help of wind currents and workers. Their life cycles ar e extremely short, taking between 15 and 20 days if environmental conditions are conduciv e for their growth and development. The short life cycles as well as overlapping gene rations during the bluebe rry flowering cycle,

PAGE 42

28 make this insect a dangerous pest that can reach economically damaging levels in a very short period. The reduced amount of knowledge and the importance of this pest for blueberry growers is the main reason to deve lop an Integrated Pest Management (IPM) program to control the populations of thrips. The results shown in this dissertation are the beginning of this IPM program answering some of the basic questions about the relationship between bluebe rries and flower thrips. Thrips control Due to their behavior, quick reproduction rate and potential to inflict great damage even at low populations (in the case of virus vectors), there are 236 products registered to control thrips in the U.S. listed by Crop Data Management Systems Inc. (CDMS) (Marysville, CA). In blueberr ies there are 24 insecticides labeled to control thrips, but only 8 active ingredients(CDMS 2006). Due to high pressure of these insects, growers have a high dependence on chemical contro l for fast management of the pest. Thrips resistance to tartar emeric ins ecticides was detected as early as 1941. However, the typical example of thrips re sistance is described by Morse and Brawner (1986). They described how in four years thri ps became resistant to DDT and dieldrin, 18 years to dimethoate, seven years to be resi stant to malathion. Othe r tests in the same species showed how Scirtothrips citri (Moulton) increased its resistance by 428-fold to fluvalinate after only 10 selections, at the same time that the resistance to other pyrethroids increased by 10 or more (Morse and Brawner 1986). Today there is a greater adoption of Integr ated Pest Management (IPM) initiatives to control thrips. The IPM approach is base d on five techniques: host plant resistance, chemical control, mechanical control, cultu ral control, and biologi cal control (Parrella and Lewis 1997). The only type of plant resistan ce that has been achieved is resistance to

PAGE 43

29 certain tospoviruses vectored by thrips, wh ich considerably re duces their damage (Ullman 2005). In the case of direct resistance to the insect most of the work is being conducted in non-preference changes in morphologi cal characteristics such as form of the leaf or even color of the product (Parrella and Lewis 1997). Various forms of mechanical control have being tested to control thrips; the most ev aluated ones are mechanical barriers such a screens in gr eenhouses and filtration systems, and the use of UV reflective mulches. Barriers in the field have proven to be not economically viable. Yudin et al. (1991) used 1.5 m tall plastic barriers ar ound the crop. The results showed that the barriers only reduced the movement of F. occidentalis by 10% while they had no effect on the intra-crop movement of the insects. The use of reflective mulch to reduce the population of thrips in field crops has proven to be effective. However, the reduction was only evident when sticky traps were used in the study (Scott et al. 1989, Kring and Schuster 1992). When the number of thrips in the flowers was counted by Kring and Schuster (1992), they found no differe nces between the treatments. Biological control of thrips is very successful in closed environments such as greenhouses. However, in field crops the use of biocontrol agents has not been very successful (Parrella and Lewis 1997). Hoy a nd Glenister (1991) tried to control T. tabaci by inoculating and inundating the field with Amblyseius spp. but it failed to show positive results. The reason why biological contro l is not effective in the field might be due to the fact that thrips populations move very fast and in large numbers. Also, thrips might cause significant damage before the bene ficial organisms have time to react and achieve the appropriate control (Pa rrella and Lewis 1997)

PAGE 44

30 Reduced-risk insecticides The concept of reduced-risk insecticides was introduced by the Environmental Protection Agency (EPA)s Office of Pestic ide Programs (OPP) in July 1992. In this public notice there are incentives for the deve lopment and registration of new chemistries that comply with the following characteri stics to be registered as reduced-risk insecticides (Environmental Protection Agency (EPA) 1997). Human health effects Very low mammalian toxicity Between 10 to 100 times less toxic than alternatives Displace chemistries with known lethal effects on human health such as organophosphates Reduce exposure to workers Non-target organisms Very low toxicity to birds, fish, honey bees, and other benefi cial insects, and non-target organisms in general (calcula ted as direct toxi city of degree of exposure) Highly selective to target pests Groundwater (GW) Low potential for GW contamination Low drift and runoff Lower use rates than the alternatives Low pest resistance potential Highly compatible with IPM Effective to control target pests Currently there are approximately 60 new chemistries that are considered as reduced-risk insecticides, or biopesticides as described by the Food Quality Protection Act of 1996 (United States Congress (104th) 1996, IR 4 project 2006). Nine of these chemistries are registered or pending registration for use in blueberries, its chemistry and

PAGE 45

31 status is summarized in Table 25 as well as four Orga nophosphate (OP) alternatives that can be used to reduced the effect of OPs in the enviroment. Reduced-risk insecticides are considered a fundament al part of IPM programs independent of the commodity in questi on (Environmental Prot ection Agency (EPA) 1997, Atanassov et al. 2002, Environmental Protection Agency (EPA) 2003, Finn 2003, Hamill et al. 2003, Liburd and Finn 2003, Li burd et al. 2003, Mizell 2003, Liburd and Arvalo 2005, IR 4 project 2006). Economic Injury Levels (EIL) Economic Injury Level (EIL) is one of the most discu ssed topics in entomology. The reason for this is that the EIL gives us the most basic information needed for a successful IPM program. How many insect s will cause significant damage? The answer to this question is usually the star ting point for decision making in a commercial crop (Pedigo et al. 1986). Stern et al. (1959) developed the first concepts of economic damage, EIL, and the majority of these have not changed since then (Pedigo et al. 1986). In entomology, Stern et al. ( 1959) defined the EIL as The lowest population density that will cause economic damage. This concept assumes the possibility of scouting, evaluation and use of control tac tics as needed. For this reaso n, it is very practical in the case of arthropod pests since this is the root of IPM programs. Several authors have criticized the simplicity of the EIL, arguing th e lack of a more comprehensive view of the farm as a system. Variation on commodity pr ices, interaction with other arthropods and climate conditions made some EILs stationa ry, obsolete, and only valid for one season (Poston et al. 1983, Pedigo et al. 1986). Despite these critiques, EI L is the most used method of decision-making for arthropod pests. That new authors incl ude their ideas and

PAGE 46

32 suggestions to improve it, makes the EIL a dynamic concept (Pedigo et al. 1986, Pedigo 2003). I would like to define some of the concepts as used in this dissertation. Some authors have determined EIL as a level of in jury (Shelton et al. 1982) but in most of the cases standardization of the in jury is difficult to determine in such a way that might be practical for growers. Pedigo et al. (1986) uses the term injury equivalent to determine the injury level caused by one pest through its life cycle and the term equivalence as the total injury equiva lents inflicted by a population at a given moment. The term EIL in this dissertation will correspond to the insect density causing economic damage as defined by Pedigo (2003) and describe d in Equation 2-1. In this case ( C ) is the cost of management per production unit, ( V ) is the market value per production unit, ( I ) defines the injury un it per pest, ( D ) is the damage per injury unit and ( K ) is the proportional reduction in pest attack or iginated by the control. Another important value complementary to the EIL at the moment of taking decisions is the gain threshold (GT). To define GT it is necessary to understand the concept of economic damage (ED), which refers to the equilibrium point where the cost of controlling the pest is e qual to the damage caused by the pest, it is determined as monetary value and it is described in terms of ( C(a)), the cost of the control, ( Y ), yield, ( P ) price per unit of yield, ( s ) level of pest injury, and ( a ) control action (Equation 2-1 (a)). Stone and Pedigo (1972) defined GT in function of the same terms that Stern et al. (1959) had defined as ED, but the GT is described in terms of loss of marketable product per cultivated unit (Equation2-1 (b)).

PAGE 47

33 Equation 2-1: Some producers might use this number as indicator to make decisions. However, it is too risky to take actions when the pest has reached the EIL or the GT, because by the time that the control practices are in place th e pest might have reached the point where the cost of controlling is higher than the value of the crop and it would not be economically wise to take any actions. For th is reason the concept of economic threshold (ET) was included. Economic threshold is defi ned as the practical or operational pest density when control must be taken in order to keep the crop as a pr ofitable business. The ET includes variables such as EIL, pest and host phenology, population growth rates (variable depending on the conditions for each farm), and interaction with other organisms or chemicals applied for other purpo ses. Due to the practical and mathematical complexity of calculating this ET, most ET are relatively crude as expressed in Pedigo (2003). There are some limitations to the con cept of EIL expressed by Pedigo (2003). Some of the limitations mentioned are: 1. Lack of mathematical definition for ET 2. Lack of more comprehensive EIL 3. Reduced ability of make cost-effective, accu rate population analysis in the field 4. Inability to predict mark et, population trends and other variables for the ET 5. Difficulty to quantify variables such as weather, environmental cost etc. ) ( ) ( ) ( ) ( ) (] [ ] [s s a a aP Y s P s Y C ] [) ( a aS P C GT (a) (b)

PAGE 48

34 However, despite these limitations, this concep t it is still the best tool at present for growers to decide pest management strategies Some of the values used by the growers are empirical due to the lack of rese arch in some of the commodities. The relationship between blueberries and flow er thrips is still unexplored. There is a lot of knowledge about thes e two species generated through out years of research. Our objective is to use all this information to guide our research and understand the relationship between flower thrips and blue berries. The following dissertation is an attempt to generate and compile informati on regarding this relationship. The ultimate goal is to develop an IPM program for early -season blueberries in Florida and Southern Georgia.

PAGE 49

35Tables and Figures Table 21: List of dise ases reported in blueberries in the United States. Type Agent Distribution Symptoms Transmission Management Phomopsis vaccinii Southeastern U.S. Phomosis canker Dieback of fruit bearing stems Yield reduction up to 70% Rotting of fruits Spores are released from infected case that overwintered in the field Remove infected and suspicious branches during winter pruning Botryosphaeria dothidea Southeastern U.S. 1-2 year old bushes Blueberry stem blight Blight of individual branches Brown r red branches flags Spores are released from infected stems (blueberry and alternate hosts) and need an injury in the stem to be able to penetrate Timely pruning of infected branches far from the start if the symptom. Resistant varieties are available Fungi Botryosphaeria corticis Southeastern U.S. (NJ, GA, FL, AL, MS) Stem Canker Swelling at the point of infection Spore-producing structures emerge through the bark Spores are released during the wet season and are wind transported. Only young stems are susceptible Sanitation, avoidance and the use of resistant cultivars Viruses and Phytoplasmas Blueberry scorch carlavirus Northern coastal states in the U.S. Scorch Rapid necrosis of leaves and flowers Small Chlorosis of the leaves and stems Vectored by aphids Infected bushes should be removed, burned and replaced with tolerant cultivars

PAGE 50

36Type Agent Distribution Symptoms Transmission Management Blueberry shock ilarvirus Western U.S. Shock Sudden necrosis of flowers and leafs A second vegetative flush is present but no flowers produced In well-managed fields close to normal production is possible 1 to 4 years after infection It is vectored by pollinators carrying infected pollen Infected bushes should be destroyed before bloom Blueberry shoestring Northeastern U.S. Shoestring Symptoms appear 4 years after infection Reddish lines in the stems Leaves become red and deformed Ripe fruits are red It is transmitted by the blueberry aphid, Illinoia pepperi MacGillivray Use of resistant cultivars, use of virus free plantings, control of the vector to reduce spread of the virus Viruses and Phytoplasmas Tobacco ringspot virus Necrotic ringspot Leaves have 2-3 mm necrotic spots, the center of the injury may fall off Vectored by the dagger nematode Xiphinema americanum Cobb Avoid the nematode, fumigate in case the vector-nematode is present and use of virus free plantings Table 2-1 continued

PAGE 51

37Type Agent Distribution Symptoms Transmission Management Viruses and Phytoplasmas Blueberry stunt phytoplasma Midwestern and eastern U.S. Stunt Short bushy canes Leaves with yellowing in the margins Reduced internodal distance Vectored by leafhoppers Eradication and destruction of the infected plants Agrobacterium tumefaciens Cosmopolitan Crown gall Potted plants and new plantings present galls in the roots up to 2.5 cm in diameter It is a soil-borne bacterium. Enters through wounds and cuts on the roots Avoid planting material with obvious galls. If site is infested plant nonhost materials for 3 to 4 years before planting blueberries Bacteria Pseudomonas syringae Pacific northwest U.S. Bacterial canker Die back of young branches The bacterium penetrates through wounds caused by insects, wind, or manual labor Cut infected canes in late fall during the dry season and sterilize the equipment used with bleach Modified from (2006)Table 2-1 continued

PAGE 52

38 Table 22. Number of polle n grains per flower for f our plant species, and the extrapolated percentage of the grains that could be eaten by five or 100 thrips per flower in three days (95% confidence limits).* Extrapolated consumption Plant species Grains per flower 5 thrips / 3 days 100 thrips/3 days Echium plantagineum 156,6007.6% (4-11)152% (77-227) Actinidia deliciosa 2,000,0000.5% (0.2-0.7)9% (5-14%) Brassica napus 140,0003.2% (2-4)64% (47-80) Jacaranda acutifolia 13,4003.2% (2-5)64% (35-94) From Table 2 in Kirk (1987). Table 23: Some estimates of popul ation parameters of pest thrips* Thrips species / Crop Temperature C L:D R0 rm day T days rc day-1 Tc days Frankliniella fusca Peanut 20 14:105.07 0.05 31.2 Peanut 30 14:1016.0 0.16 17.5 Peanut 35 14:101.67 0.04 13.2 F. occidentalis Bean 23 04:203.70.062 21.2 Bean 23 16:0812.20.140 17.9 Chrysanthemum 15 42.20.056 66.5 Chrysanthemum 35 2.70.056 17.6 Cotton pollen 27 14:1030.10.157 21.6 Cotton + pollen 27 14:10111.80.220 23.4 Peanut 20 14:101.1 0.02 19.7 Peanut 30 14:102.3 0.02 15.6 R0 is the net reproductive rate; rm is the intrinsic rate of natural increase; T is the mean generation time; rc is the capacity for increase; and Tc is the cohort generation time; L:D is hours of light and dark per day. Note that rc and Tc are approximations of rm and T. Modified from table 7.1 in Kirk (1997b)

PAGE 53

39 Table 24: Known tospovi ruses and thrips vect ors in the world Virus Thrips vector Tomato Spotted Wilt Virus Frankliniella bispinosa F. fusca F. intosa F. occidentalis F. schultzei Thrips setosus T. tabaci Impatiens Necrotic Spot F. occidentalis F. schultzei F. intosa Zucchini Lethal Chlorosis F. zucchini Watermelon Bud Necrosis T. palmi Watermelon Silver Mottle T. palmi Melon Yellow Spot T. palmi Capsicum Chlorosis Ceratothrips claratis Groundnut Ringspot F. occidentalis F. schultzei F. intosa Tomato Chlorotic Spot F. intosa F. occidentalis F. schultzei Peanut Chlorotic Fan-spot Scirtothrips dorsalis Groundnut Bud Necrosis S. dorsalis Peanut Yellow Spot S. dorsalis Iris Yellow Spot T. tabaci Chrysanthemum Stem Necrosis F. occidentalis F. schultzei Physalis Severe Mottle Not described Table modified from Naidu et al. (2005) and Ullman (2005).

PAGE 54

40 Table 25: Reduced-risk, biopesticides and OP alternative insecticides registered or pending registration for use in blueberries*. Chemical name Trade name Chemistry Status Classification Acetamiprid Assail 70 WP Adjust Chloronicotinyl Pending Reduced-risk Azadirachtin Neemix Niblecidine Extract from neem oil Registered Biopesticide Bacillus thuringensis Dipel Bacteria Registered Biopesticide Cinnamaldehyde Cinnacure Cinnamite Natural product Registered Biopesticide Flonicamid 1 Carbine 50 WG Beleaf Nicotinamide Potential OP alternative Imidacloprid Admire Provado Gaucho I Chloronicotinyl Registered OP alternative Indoxacarb Avaunt Steward Oxadiazine Pending Reduced-risk OP alternative Metaflumizone BAS 320 I Semicarbazone Potential Reduced-risk Methoxyfenozide Intrepid Runner Diacylhydrazine Pending Reduced-risk OP alternative Novaluron 1 Diamond Rimon Benzoylphenyl Pending Reduced-risk OP alternative Spinosad 1 Success Spintor Entrust Macrocyclic lactone Registered Reduced-risk OP alternative Thiamethoxam 1 Actara Platinum Centric Cruiser Helix Second generation Neonicotinoid Registered OP alternativeZeta-cypermethrin 1 Mustang Mustang Maxx Pyrethroid Pending OP alternative Table Modified from IR 4 Project (2006) 1 Indicates chemistries registered or that have the potential to control thrips

PAGE 55

41 CHAPTER 3 SAMPLING TECHNIQUES AND DISPE RSION OF FLOWER THRIPS IN BLUEBERRY FIELDS Blueberries are one of the fastest growi ng crops in Florida. Due to climatic conditions and the early-season varieties produc ed, Florida and most of southern Georgia blueberries mature during April and May, maki ng these two states the principal producers in the world during this time. This window of production gives the growers a price advantage of 3 to 5 USD per pound compared to regular season blueberries produced between the end of May and August (NASS-USDA 2006a). Thrips have been identified by blueberry growers as insect pests that require immediate management (Finn 2003). Twenty fi ve percent of blueberry growers from southern Georgia and Florida identified flow er thrips as one of their main problems surpassed only by blueberry bud mite, Acalitus vaccinii (Keifer). Other major pests of concern for the growers include cranberry fruitworm, Acrobasis vaccinii Riley, and blueberry gall midge, Dasineura oxycoccana (Johnson) (Finn 2003). For these reasons the Small Fruit and Vegetable IPM Laboratory at the University of Florida started a project to understand th e relationship between blueberries and flower-thrips as an initial step to develop an IPM program for thrips in early-season blueberries. Finn (2003) initiated preliminary work in sampling techniqu es for flower thrips in blueberries. As a common practice, flower thrips have been monitored by using sticky traps of various colors. The two colors most commonly used are yellow and blue (Diraviam and Uthamasamy 1992, Cho et al. 1995, Hoddle et al. 2002, Finn 2003). Finn (2003) found

PAGE 56

42 non significant differences betw een the numbers of thrips captured in yellow, blue, or white sticky cards in blueberry plantings. Due to the contrast between the thrips and the white background on the sticky cards, I decided to use white sticky cards for monitoring populations in my experiments. Because there is little information about the relationship between flower thrips and blueberries, I must determine some of the basic characteristics of this relationship. One of the basic char acteristics I are look ing at is dispersion. Ordinarily there are three types of di spersion: random, uniform, and clumped. Distribution depends on the mobility of the inse cts. Highly mobile insects tend to have a more random distribution than insects with low mobility, which tend to form hot-spots in highly clumped populations (Flint and Gouveira 2001). My objectives were to select an efficien t and effective sampling method to monitor flower thrips inside blueberry flowers, a nd to determine the vertical and horizontal distribution of thrips popul ations in blueberry plantings. My final goal was to characterize thrips populations depending on their level of aggregation in blueberry fields. Materials and Methods To address my objectives, I have conducte d a series of experiments on private blueberry farms in Florida and southern Georgia. Methodology to Determine Thrips Population Inside Blueberry Flowers Due to the high volume of flower samples, a more efficient system to determine the number of thrips inside the flowers was developed by combin ing procedures from various researchers (Finn 2003, Funderbur k and Stavisky 2004). Flower clusters were collected using Corning 50 ml plastic tubes (Fisher Scientific, Pittsburg, PA). Flowers were collected in the field by cutti ng the pedicels using the rim of the vial and allowing them

PAGE 57

43 drop inside the plastic tubes containing 70% ethanol. Each vial was manually shaken for approximately one minute. The contents were emptied into a 300 ml white polyethylene jar (B & A Products, Ltd. Co., Bunch, OK) and filtered through a plastic screen with 6.3 x 6.3 mm openings to ensure that thrips pa ss through, leaving the fl owers on the screen. The remains left on the screen were rinsed with water from a polyethylene wash bottle into a white container. The flowers left on the screen were placed in a 300 ml white polyethylene jar while the corollas and the calyxes of the flowers were manually separated. This procedure was repeated thre e times to ensure the collection of the maximum number of thrips. After each rinsing, the thrips found in the rinsing water were collected inside a white container and counted. This water was then transferred to another container with black background to ensure I collected the maximum number of thrips. To determine the efficiency of this system, I decided to compare the results obtained with this method with standard flower dissection, which is the method commonly used in the laborator y to determine thrips popula tions inside flowers (Finn 2003). I selected 20 samples from the weeks th at had the highest number of thrips. After following the shake and rinse procedure, samples were dissected and observed under a microscope to determine how many thrips were missed. I then used a t -test to compare the total number of thrips collected using th e shaking and rinsing pr ocedure with the total number obtained by shaking and rinsing pl us the number of thrips found under the microscope by dissecting the samples after this procedure (SAS Institute Inc. 2002). Vertical Distribution of Flower Thrips in Blueberry Fields To determine thrips distribution within blueberry bushes, I placed ten sampling stations in each of my two farms. The fi rst was located on farm FL01 in south-central Florida (N 28 04 W 81 34). This farm was planted with Southern highbush. A second

PAGE 58

44 farm, Farm GA01, located in southern Georgi a (N 31 31 W 82 27), was planted with rabbiteye blueberries. Samples were ta ken during the 2004 and 2005 flowering seasons. To collect the samples I randomly establishe d 10 sampling stations around the selected plot in each farm. Each sample station cons isted of a blueberry bush where I took four samples: three white sticky traps (23 x 17 cm of effective area) (Great Lakes IPM Vestaburg, MI) and one flower sample. One of the traps was placed on the ground in an inverted V shape with the sticky surface to wards the ground, a second trap was located inside the canopy approximat ely in the middle of the bush, and the third one was approximately 40 cm above the canopy. The num ber of thrips in the sticky traps was determined by counting the number of thrips in 16 out of the 63 squa res (each square is 6.45 cm2) that the trap is divided into (2003). Flower samples were taken from the same bush containing the sampling station and consiste d of five flower clusters collected in Corning 50 ml plastic tubes filled with 70% ethanol. I cut the flowers using the thumb and the rim of the vial, thereby reducing th e manipulation of the flowers. The flower samples were processed using the shake and rinse method described above and the total number of thrips was recorded. The sampling stations were randomly placed in the field each week by using random number tables based on the number of rows and the number of plants in each row. The data were co llected from bloom to petal-fall 2004 and 2005. Data were analyzed using the repeated m easures analysis (SAS Institute Inc. 2002). I decided not to use the soil traps in Georgia since the results in Florida showed that the number of thrips captured was too low for an alysis. On average, 0.66 0.3 thrips per trap per week in 2004 and 0.38 0.1 thrips per trap per week in 2005 were collected in the soil samples. Data were transformed to comply with the assumptions of the analyses.

PAGE 59

45 The data for 2004 and 2005 in Florida and 2005 in Georgia were transformed using the natural logarithm of th e original data plus one; for 2004 in Georgia, the transformation used was the square root of the number of thrips captured. Thrips Dispersion We selected two fields in north central Florida, Farm FL02 (N 29 40 W 82 11) and Farm FL03 (N 29 43 W 82 08). Both farms were planted with southern highbush blueberries during 2005. However, during 2004 FL02 was planted half on rabbiteye and half in southern highbush as indicated in Figure 3-3. Two grids one of 5 x 6 and the second one of 8 x 7 traps were respectively deployed in each one of the selected plots. The traps were spaced 30.48 m from each othe r, which covered blueberries and adjacent non-cultivated areas. These traps were re placed every other da y starting from bloom initiation and finishing at petal fall. The tota l number of thrips trapped was recorded to monitor the movement of thri ps into and out of blueberry fields for two flowering seasons. In 2004 sampling begun on March 3, while in 2005 it begun on February 20 at both locations. To determine degree of aggregation I se lected the standardized Morisitas coefficient of dispersion (Ip) (Smit-Gill 1975) and Greens coefficient of dispersion (Cx) (Green 1966), because they have low or no correlation with the mean (Myers 1978, Taylor 1984, Schexnayder Jr. et al. 2001). The aggregation indices were calculated for each day that the sticky traps were collected. I graphed the number of thrips captured in each trap using Sigma Plot (SYSTAT Software Inc. 2006). Once the hot-spots were graphically identified, I conducted a Gaussian regression to describe the population behavior in each one of the hot-spots.

PAGE 60

46 In this study populations were considered to be clumped if Greens index Cx > 0, random if Cx = 0, or uniform if Cx < 0 (M yers 1978, Schexnayder Jr et al. 2001). In the case of standardized Mo risitas index (Ip) populations were considered to be significantly clumped ( = 0.05) if Ip > 0.5, not significantly clum ped if 0.5 > Ip > 0, random if Ip = 0, not-significantly uniform if 0 > Ip > -0.5, a nd significantly uniform if Ip < 0.5 (SmitGill 1975). Overall comparisons were conducted by averaging all the indices calculated and comparing them to 0 for Cx, and 0.5 for Ip using a ttest (SAS Institute Inc. 2002). Equation 31 Results Methodology to determine thrips po pulation inside blueberry flowers I found no significant differences between the dissecting (35.7 4.3) and the shake and rinse (34.7 4.3) methods ( t = 0.17; df = 1, 38; P = 0.869), when comparing the number of thrips obtained with each method. Th ese results allow us to use the shake and rinse procedure in the following experiments with confidence in the data collected. Vertical Distribut ion of Flower Thrips The thrips distribution with in the bushes follows the same pattern independent of the year and the location. In Florida, th ere are no significant di fferences between 2004 and 2005 when the same positions within the bush were compared. For soil ( t = 0.531; df = 1, 325; P = 0.595), for flowers ( t = 1.474; df = 1, 325; P = 0.142), for the traps in the bushes ( t = 0.308; df = 1, 325; P = 0.7582) and for the traps above the canopy ( t = 0.438; df = 1, 325; P = 0.662). However, when comparing the treatments within each one of the ) 1 ( 12 q q X s Cx ) 1 ( 1 X X x x q Ipq i i i

PAGE 61

47 years, I found significant differences among th e positions with respect to the bush (For 2004, F = 291.13; df = 3, 157; P < 0.0001, and for 2005 F = 197.51; df = 3, 164; P < 0.0001). In both years, the number of thrips cap tured was significantly higher within the canopy compared with all other positions eval uated. The second highest number of thrips captured was found in the traps deployed above the canopy, followed by the number of thrips inside the flowers, and fi nally by traps on top of the soil (Figure 31). Our results on farm GA01 in southern Geor gia were different from the situation presented in Florida. As in Florida, I found no significant differences between the number of thrips captured in the flowers between 2004 and 2005 ( t = 0.682; df = 1, 144; P = 0.496) in Georgia. However, I found that in 2004 I captured significantly more thrips within the canopy ( t = 7.345; df = 1, 144; P <0.0001), and above the canopy ( t = 8.563; df = 1, 144; P < 0.0001) than in 2005. When comparing the number of thrips captured at the various positions, I found no significant di fferences between the number of thrips within the canopy and above the canopy during 2004. However, these values were both significantly higher than the num ber of thrips captured in th e flowers sampled during the same year. In 2005, there was a reduction in the number of thrips captured on farm GA01 compared with 2004 ( Figure 32). However, despite the redu ced numbers, I found a significantly higher number of thrips captured in the canopy than above the cano py of the bushes. Both of which were significantly higher than the num ber of thrips capture d in the flowers ( Figure 32).

PAGE 62

48Thrips Dispersion 2004 farm FL02 Thrips aggregation increased over time and p eaked 12 to 14 d after bloom initiation. This peak coincides with the highest population de nsity of thrips 14.7 d after bloom initiation (Table 31 and Figure 35). Table 31 shows that thrips population can be considered clumped from day 4 based on Cx. This obser vation is reinforced by Ip, which shows a significant level of aggregation from the beginning. After recording a clumped-type distribu tion, I plotted thrips population and the coordinates where the traps were located to determine the position and number of hotspots based on the locations presented in Figure 33. During the 2004 field-season I found only one hot-spot located at the coordinate (4, 4) in Figure 36. When analyzing this hot-spot during 2004, I found that th e dynamics of thrips population could be described by a Gaussian non-lin ear regression (Equation 3-2 a.). The pattern for the hotspot on farm FL02 in 2004 is described by the equation represented in Equation 3-2 b. Overall Cx (0.467 0.147) is significantly higher than 0 ( t = 3.17; df = 8; P = 0.013), and Ip (0.521 0.004) is si gnificantly higher than 0.5 ( t = 7.48; df = 8; P <0.0001), which shows a significant level of ag gregation of the flower thrips on farm FL02 for 2004. Equation 3-2: a. b. 26 2 7 14 5 017 194xe y 25 02 1 ) ( xe x f

PAGE 63

492004 farm FL03 The distribution of traps on farm FL03 is shown in Figure 34. Thrips population for farm FL03 was considerably lower compared with farm FL02 during 2004. However, it appears that there are two main areas, id entified using the graphic method, where thrips tended to aggregate. Tw o hot-spots were (Figure 37). One hot-s pot was located at coordinate (0, 4) and a sec ond at coordinate (2, 2) of Figure 37. The peak population for these hot-spots occurred on different da ys. For the spot found at (2, 2) the peak occurs at 11.3 d after bloom and for the hot-s pot located at (0, 4) aggregation occurred 17 d after bloom (Figure 38 and Equation 3-3). Equation 3-3: Greens index (Cx) and the standardized Morisitas index (Ip) showed a tendency towards a random distribution of thrips on farm FL03 in 2004. The overall Cx for farm FL03 (0.046 0.041) was not significantly different from 0 ( t = 1.14; df = 6; P = 0.298), and the Ip value (0.020 0.006) was significantly higher than 0 ( t = 3.21; df = 6; P = 0.0184) but still in the region .5< Ip < 0.5. Ho wever, this aggregation appears not to be significant on farm FL03 in 2004. The distribu tion appears to be more aggregated for days 7 to 15, which again coincide s with the peak of the population (Figure 38). 2005 farm FL02 The same set up used in 2004 was used in 2005 to describe the dispersion of flower thrips in the field. During this year the fa rmer replaced the rabbite ye blueberries with a 1 4 3 11 5 032 5xe y 4 2 17 5 036 4xe ya. b.

PAGE 64

50 new planting of southern highbush. Thrips population on farm FL02 was lower in 2005 than in 2004. During this year I found two hot-spots located at coordinates (2, 3) and (5, 2) in Figure 39. The highest aggregation wa s between days 10 and 14, which again coincides with the days of maximum population (Table 33 and Figure 310). The hotspots reached their maximum population at 13.8 d after bloom initiation for the coordinate (2, 3) and 12.1 d for coordinate (5, 2) ( Figure 310 and Equation 34). The overall indices show a highly signif icant aggregation for 2005. Greens index, Cx = 0.24 0.06, was significantly greater than 0 ( t = 3.87; df = 9; P = 0.0047), and the overall Standardized Morisi tas index, Ip = 0.52 0.01, was significantly gr eater than 0.5 ( t = 4.94; df = 9; P = 0.0011). Equation 3-4: 2005 farm FL03 Thrips population for this farm was too lo w to make a robust an alysis. The highest number of insects captured in one trap was three thrips 15 d after bloom initiation. Most of the other traps captured no thrips, and the data were not considered significant. Discussion The literature has discussed several types of sampling methods for thrips inside flowers. Finn (2003) mentions the use of alc ohol dipping, tapping the fl oral clusters on a white surface, and flower disse ction as methods to determine the number of thrips inside the flowers. Finns study showed no signifi cant differences in the number of thrips 267 2 05 12 5 058xe y 244 3 78 13 5 049 63xe ya. b.

PAGE 65

51 captured among the various methods for southern highbush blueberries. However, the Finn (2003) found that sampling by tapping the fl owers in rabbiteye blueberries resulted in significantly fewer thrips captured than did other treatments. A large variation was observed probably due to the distribution of the thrips in blueberry fields. High aggregation and random sampling usually produc es high variance in the results. Palumbo (2003) compared trapping at canopy level, pl ant beating, direct observations and whole plant washes. Whole plant washes (very sim ilar to the shake and rinse method used in this study) were used as absolute sample s. Palumbo (2003) found a significantly higher number of thrips in his absolute method wh en compared with the other methods used. The fact that I found no significant differen ces between the shake and rinse method and the dissection method, which is considered to be an absolute count of the thrips in the flowers (Hollingsworth et al 2002), indicates that the shake and rinse method is appropriate to estimate thrips population insi de the flowers. The shake and rinse method might be too time-consuming and not very useful for growers who need to determine the population rapidly and accurately, but it might be useful for research purposes, as it is as accurate and less time-consuming than flower dissections. The vertical distribution of thrips remained the same independent of location and year. The highest number of thrips was cons istently captured in or above the canopy of the blueberry bushes using sticky traps. Howeve r, the most damaging population is inside the flowers (Arevalo and Liburd unpublished data). The average number of thrips captured va ries from year to year. In 2005, thrips population was lower on farm FL01 located in Florida than the farm GA01 located in

PAGE 66

52 Georgia. However, only in GA01 were the diffe rences in the number of thrips captured between the years significant, having a lower population in 2005 than in 2004. The analysis of the distribution based on Morisitas and Greens indices described the distribution of thrips in the field as aggregated. However, the level of aggregation seems to be lower in cases when peak populat ions are lower. For instance, on farm FL03 in 2004 (peak population = 5.3 thrips per tr ap) the average Cx (0.046 0.041) and average Ip (0.02 0.006) were lower than for farm FL02 in 2005 (peak population = 63.49 thrips per trap) Cx (0.24 0.06) and Ip (0.52 0.01) and lower than Farm FL02 (Peak population = 194.1 thrips per trap), Cx (0.467 0.147) and Ip (0.521 0.004). In Figures 3-5 and 3-11, I observed that the hot -spots start forming between 7 and 10 days after bloom initiation and in both cases the population at these spots grew beyond 20 thrips per trap every two days. After this in itial period, thrips populat ion captured in the traps increased exponentially, reaching a maximum population between 12 and 15 days after bloom initiation. The population then decl ined at the same ra te that it increased, virtually disappearing about 22 days after bl oom started, after most of the flowers had become fruits. Apparently, formation of hot-spots on bl ueberry farms is random. I did not find any correlation among the locations where hot -spots were formed between the years. However, several variables such as flower concentration, soil type, fertilization methods, and wind direction, might be studied to de termine a correlation among these variables and hot-spot locations to create a pred ictive dispersion model of flower thrips on blueberry farms. For now, sampling me thods that consider highly aggregated populations should be explored to reduce the variability of the data when sampling flower

PAGE 67

53 thrips in blueberry farms (Southwood 1989, Wang and Shipp 2001). Despite that until now no sex pheromones have been isolated, some behavioral observations show the presence of a mating or an aggregation phero mone in thrips (Milne et al. 2002, Kirk and Hamilton 2004). Kirk and Hamilton (2004) show ed how virgin females are attracted to the smell of males and not to the smell of other females. This situation was interpreted as the presence of some type of sex pheromone in Frankliniella occidentalis (Pergande). Milne at al. (2002) found a hi gh correlation between the number of males in hibiscus flowers and the number of females landing on thes e flowers. Salguero-Navas et. al (1994) found indications of aggregati on in tomato plants for vi rus thrips species including F. occidentalis, and F. tritici. However, differences in the degree of aggregation were also found between years in their experiment as it was also found in my trials. Thrips populations were demonstrated to be variable within the same region. For instance, farms FL02 and FL03 are 7.02 km from each other an d the peak populations are significantly different, 194.2 for FL02 and 5.3 for farm FL03. However, there appears to be a correlation between the population dynamics and the year; in my observations, 2005 had a lower number of thrips captured than 2004 in all farms, and studies to determine the reasons for this pattern are needed.

PAGE 68

54Tables and Figures Table 31: Distributi on indices, Greens index (Cx) a nd Standardized Morisitas index (Ip), used to describe the level of aggregation of thrips population on farm FL02 in Florida in 2004 Table 32: Distributi on indices, Greens index (Cx) a nd Standardized Morisitas index (Ip), used to describe the level of aggregation of thrips population on farm FL03 in Florida in 2004. Index Days after blooming 2 4 6 8 10 12 14 18 22 Cx -0.005 0.438 0.223 0.138 0.227 0.967 1.336 0.207 0.670 Ip 0.518 0.549 0.525 0.517 0.524 0.543 0.539 0.520 0.538 Index Days after blooming 2 4 6 8 14 16 19 Cx -0.011 -0.065-0.013 0.129 0.062 0.246 -0.021 Ip 0.041 0.0420.024 0.013 0.006 0.014 -0.0002

PAGE 69

55 Table 33: Distributi on indices, Greens index (Cx) a nd Standardized Morisitas index (Ip), used to describe the level of aggregation of thrips population on farm FL02 in Florida in 2005. Index Days after blooming 2 4 6 8 10 14 16 18 22 Cx 0.124 0.098 0.142 0.185 0.314 0.713 0.320 0.233 0.112 Ip 0.051 0.513 0.518 0.522 0.528 0.556 0.529 0.525 0.511

PAGE 70

56 0 10 20 30 40 50 60 70 80 90 SoilFlowersAbove the canopyIn the canopy Location in the sampling station Average No. of thrips captured per week 2004 2005 d c b a Figure 31: Vertical distri bution of thrips captured with respect to southern highbush blueberry bushes in south Florida. Different letters represent significant differences among the groups usi ng LSD mean separation test, = 0.05. 0 20 40 60 80 100 120 140 160 180 200 FlowersIn the canopyAbove the canopy Location in the sampling stationAverage No. of thrips captured per week 2004 2005 B A A c b a Figure 32: Vertical distribution of thrips captured with respect to rabbiteye blueberry bushes in southern Georgia. Different letters in capitals represent significant differences among the groups in 2004, small letters represent significant differences among groups in 2005 us ing LSD mean separation test = 0.05.

PAGE 71

57 Figure 33: Map of farm FL02 located at N 28 04 W 81 34 in north central Florida

PAGE 72

58 Figure 34: Map of farm FL 03 located at N 28 04 W 81 34 in north central Florida Days after blooming 0510152025No. of thrips captured 0 50 100 150 200 250 Figure 35: Population dyna mics inside the hot-spot in coordinates (4, 4) of Figure 36 for 2004 on farm FL02.

PAGE 73

59 0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips captured X D a t aY D a t a 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX Da t aY D a t a 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 80 100 120 140 160 Figure 36: Number of thrips captured at 2 (a), 6 (b), 8 (c), 10 (d), 14 (e), 16 (f), 18 (g), and 22 (h) days after bloom began on farm FL02. a. b. c.

PAGE 74

60 0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 80 100 120 140 160 Figure 3-6: Continued d. e. f.

PAGE 75

61 0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips captured per trapX D a t aY D a t a 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 80 100 120 140 160 Figure 3-6: Continued g. h.

PAGE 76

62 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3No. of thrips capturedX D a t aY D a t a 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3No. of thrips capturedX D a t aY D a t a 0 1 2 3 4 5 0 1 2 3 4 5 1 2 3 4 5 0 1 2 3Z DataX D a t aY D a t a 0 1 2 3 4 5 Figure 37: Number of thrips captured on farm FL03 at 2 (a), 4 (b), 8 (c), 14 (d), 16 (e), and 20 (f), days after bloom in 2004 a. b. c.

PAGE 77

63 0 1 2 3 4 5 1 2 3 4 5 0 1 2 3No of thrips capturedX D a t aY D a t a 0 1 2 3 4 5 0 1 2 3 4 5 1 2 3 4 5 0 1 2 3No. of thrips capturedX D a t aY D a t a 0 1 2 3 4 5 0 1 2 3 4 5 1 2 3 4 5 0 1 2 3No of thrips capturedX D a t aY D a t a 0 1 2 3 4 5 Figure 3-7: Continued d. e. f.

PAGE 78

64 Days after bloom 05101520 No. of thrips captured 0 1 2 3 4 5 6 Hot spot in (2, 2) Hot spot in (0, 4) Regression hot spot (2, 2) Regression hot spot (0, 4) Figure 38: Population dynamics inside the hot-spots in co ordinates (2, 2), and (0, 4) of Figure 37 in 2004 on the farm FL03 in Florida.

PAGE 79

65 0 20 40 60 0 1 2 3 4 5 6 7 0 1 2 3 4 5No of thrips capturedX Da t aY D a t a 0 20 40 60 0 20 40 60 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX Da t aY D a t a 0 20 40 60 0 20 40 60 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX Da t aY D a t a 0 20 40 60 Figure 39: Number of thrips captured on farm FL02 at 2 (a), 4 (b), 8 (c), 10 (d), 14 (e), 16 (f), 18 (g), and 22 (h) days after bloom in 2005. a. b. c.

PAGE 80

66 0 20 40 60 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX Da t aY D a t a 0 20 40 60 0 20 40 60 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 0 20 40 60 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 Figure 3-9: Continued d. e. f.

PAGE 81

67 0 20 40 60 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 0 20 40 60 0 1 2 3 4 5 6 7 0 1 2 3 4 5No. of thrips capturedX D a t aY D a t a 0 20 40 60 Figure 3-9: Continued g. h.

PAGE 82

68 Days after bloom 0510152025 No. of thrips captured 0 20 40 60 80 Hot spot in (2, 3) Hot spot in (5, 2) Regression for (2, 3) Regression for (5, 2) Figure 310: Population dynamics inside the hot-spot in coordinates (2, 3), and (5, 2) of Figure 39 on farm FL02 in Florida in 2005.

PAGE 83

69 CHAPTER 4 PEST PHENOLOGY AND SPECIES ASSEMBLAGE OF FLOWER THRIPS IN FLORIDA AND SOUTHERN GEORGIA IN EARLY-SEASON BLUEBERRIES The relationship between blueberries and flower thrips has not been studied because blueberries are a relatively new crop for Florida. Characterizations of thrips assemblage have been conducted in citrus, tomatoes, mangoes, and other crops. In north Florida tomatoes, Reitz (2002) found that the most common ly encountered species is Frankliniella occidentalis (Pergande), but during the spring and the fall, F. tritici (Fitch) was the most abundant. Other thrips species found in north Florida tomatoes include F. bispinosa (Morgan) and F. fusca (Hinds). Discussing his findings, Reitz (2002) emphasized the importance of studying individu al species populations of thrips when developing a new sampling protoc ol or a management program for specific crops due to their differences in behavior, damage, and importance. To understand the phenology of a pest, it is necessary to understand its relationship with its host plants. Plants can be divide d into two groups depending on their relationship with thrips. The first type is a provisional or alternate host, which offers temporary shelter or food, but thrips do not reproduce in these plants. The s econd type is called a proper host because it offers f ood, shelter, a reproductive substrate, and alimentation for immature thrips (Mound 2005). Of the 5,300 thrips species worldwide, only 50 are recognized as economically important pest species. These species considered as crop pests are prolific and nonhost-specific. For instance, F. occidentalis, western flower

PAGE 84

70 thrips, has been reported on more than 500 pl ant species within 50 families (Moritz et al. 2004b). In this study, I describe d the phenology of flower thri ps in relation to plantings across the early-season blueberr y production regions of Florid a and southern Georgia. I described the main species collected during the trials to facilitate the identification of these species. These results are part of th e thrips management strategy that the Small Fruit and Vegetable IPM Laboratory is trying to develop as part of an IPM program for blueberries. Materials and Methods Two farms in Florida planted with southe rn highbush blueberries, SFL01 located in south-central Florida (N 28o 04 W81o 35), and NCFL01 located in north central Florida (N 29o 41 W 82o 11), as well as one farm in southern Georgia, SGA01 (N 31o 32 W 82o 28), which was planted with rabbiteye blueberries, were selected to conduct the trials. These farms were sampled during th e 2004 and 2005 blueberry flowering season to monitor thrips activity. In each of these farms, I randomly placed 10 white sticky traps (Great Lakes IPM, Vestaburg, MI) in one hectare of blueberries. The traps were collected weekly from flower opening to fru it set. I also collected five flower-clusters from the same bushes where traps were deployed. The traps and the flower samples were processed at the Small Fruit and Vegetable IP M Laboratory at the Univ ersity of Florida. The number of thrips captured in th e sticky traps was counted using a 20x magnifying glass, and the thrips inside the flowers were extracted using the shake and rinse procedure described in Chapter 3. I used the Pears on correlation coefficient to quantify the relationship between the number of thrips captured in flow ers and sticky traps and the observed percentage of opened flowers in the field. I quantified the relationship between the dates of first thrips capture and dates of maximum capture with the latitude at which

PAGE 85

71 these farms were located. To be able to co rrelate the dates, I tr ansformed dates into a numerical system and used the general form at of dates in Microsoft-Excel 2000 in which January 1, 1900 corresponds to 1, Ja nuary 2, 1900 corresponds to 2 and so on, increasing by one with each day. The ten tr aps and ten flower samples per week were averaged to determine the population in the fi eld. The results were graphed to determine any trends and show correlations. A sample of 100 thrips per week from sticky traps was randomly collected to determine the thrips species assemblage presen t at each of the sampli ng sites. In the case of the flowers, I used as many thrips as I could extract from the five flower-cluster samples, to a maximum of 100 per week. To detach the thrips fr om the sticky traps I submerged the traps in 500 ml of CitroSolv (Fisher Scientif ic, Pittsburgh, PA) for four days. I used a squirt-bottle c ontaining CitroSolv to rinse the insects that were still attached to the trap. The C itroSolv along with the thrips were then filtered through a basket-style coffee filter (Publix Supermarkets Lakeland, FL). The insects in the filter were placed in a Quilted Crystal Jelly Ja r (Jarden Corporation, Muncie, IN) containing CitroSolv for five more days to dissolve the remaining glue from the traps. Thrips were allowed to air-dry at room temperature and then re-hydrated us ing deionized water. The thrips collected from the flowers were pr eserved in 50% alcohol until I was able to slidemount them for identification. The thrips from both traps and flowers were individually mounted on microscope slides using CMC-10 media (Masters Chemical Company, Elk Grove, IL). Vouchers of these specimens were sent to the Florida State Collection of Arthropods (FSCA) in Gainesvi lle, FL. The thrips collected were divided into mature and immature. Mature thrips were then identified to species using taxonomic

PAGE 86

72 keys (Mound and Marullo 1996, Moritz et al. 2001, Moritz et al. 2004b, Edwards Unpublished data ). The percentage of th rips of each species was tabulated for comparison. A short key of the most common flower thrips species present in earlyseason blueberries in Florida and Ge orgia was constructed. Results Pest phenology: Flower thrips populations in blueberry pl antings were highly correlated with the latitude and percentage of opened flowers. The Pearson correlation coefficient for the relationship between the latitude and the date of first capture was 0.971 for the 2004 season and 0.957 for the 2005 season. I also de termined the same coefficient for the relationship between the latitude and th e date when the maximum population was recorded. The coefficients for this rela tionship were 0.999 for the 2004 season and 0.955 for the 2005 season. Independent of the latitude flower thrips were first captured when 10 15 % of the flowers in the field were ope ned (Figures 4-1 to 4-6). Thrips populations peaked when 80 to 90 % of the flowers were opened. The population began to decline once the fruits started to form and the peta ls fell, leaving a reduced number of opened flowers in the field. For the farm in south-central Florida SFL01, the first thrips were recorded between February 11, 2004; three weeks later (March 3, 2004) thrips populations reached their highest on sticky traps and flowers (Figure 41). However, for 2005 the first capture was registered on February 15, on March 8, the population reaches its highest point. The highest populations of thrips inside the flowers were record ed only two weeks after their first capture on the sticky traps (Figure 42). The Pearson correlation coefficient between the average number of thrips captured in flow ers and the sticky traps with respect to the

PAGE 87

73 percentage of opened flowers was very high for both years (Table 4-1). An exception occurred for flowers in 2005 when it appear ed that the highest population was reached a week earlier in the flowers (March 2, 2005) than in the traps (March 9, 2005). The observations in north-central Florida fo llow the same trend as for south-central Florida. The first thrips ca ptured on sticky traps in NCF 01 was recorded, on average, 22 days after the observa tions in south Florida. The p eak population was recorded two weeks after the first captures in both years (Figure 4-3 a nd 4-4). After thrips population reached its peak, the number of thrips captu red declined with the decreasing number of opened flowers due to fruit formation. Correla tion coefficients are high for sticky traps and flowers in both years in relation to the percentage of opened fl owers (Table 4-1). In southern Georgia, SGA01, the first thrips were captured between March 14th and 17th, on average 9 days after NCFL01 and 31 days after SFL01. Maximum populations during both years we re very different from those in Florida. In the 2004 season, the maximum population was reache d three weeks after first detection and averaged 255.4 43.6 thri ps per trap per week (Figure 45). During the 2005 season, the maximum number of thrips captured on stic ky traps per week reached 22.0 3.3 two weeks after the fi rst detection (Figure 46). Despite these differences, Pearsons coefficients measuring the correlation between the percentage of opened flowers and the number of thrips captured in the traps and in side the flowers were high (Table 4-1). A summary of the main vari ables observed at the various sites is presented in Table 42. This table shows the high variability in the number of thrips captured among the various years and sites. For example, for south-central Florida the maximum population captured in traps was 65.4 14.7 thrips pe r sticky trap per we ek while in 2005 the

PAGE 88

74 amount of thrips captured w ith the same method was almost double, 123.4 29.7 thrips per trap per week. Due to the variability in the populations and the distribution of thrips (Chapter 3), mean compar isons were not conducted si nce there was no homogeneity among the observed sites. At the same time Table 42 and Figure 47 show the correlation between latitude and dates of fi rst capture and of maximum capture. Species assemblage: I repeatedly collected four species of thrips in Flor ida blueberry fields. These species were F. bispinosa, F. fusca, F. occidentalis, and Thrips hawaiiensis (Morgan). Other species were also recorded in Florida: Haplothrips victoriensis Bagnall, F. kelliae Salimura, and T. pini Uzel. The species with the highe st number of individuals was F. bispinosa in the flowers as well as in the sticky traps deployed in Florida (Table 4-3). The thrips species assemblage in Georgia is different from the one in Florida. The predominant species is F. tritici, followed by F. occidentalis and finally T. pini. These species were found in flowers as well as in sticky traps during th e two years that the sampling was conducted (Table 43). There is no apprec iable difference in the species assemblage between the samples taken in 2004 and 2005 at the various sites. Between 14.75 and 27.8 % of the thrips recorded inside the flowers are immatures. The percentage of immature thrips was highest in SGA 01, followed by SFL01 and finally NCFL01. Few immature thrips were found on sticky traps. This might be due to wind currents or immature thrips emerging from some of the flower materials and accidentally being caught in traps. Frankliniella bispinosa (Morgan): This species was the most abundant in Florida. It accounted for 78.89% (2005-NCFL01) to 88.96% (2004-NCFL01) of the adults captured inside the blueberry flowers. Fu rthermore, it represented between 82.51%

PAGE 89

75 (2005-SFL01) to 95.37% (2005-NCFL01) of the to tal number of thrips captured in sticky traps (Table 43). Frankliniella bispinosa was not captured in Geor gia in flowers or in sticky traps. Frankliniella tritici (Fitch): This species replaced F. bispinosa in southern Georgia (SGA01) as the most commonly en countered species in blueberry fields. Between 60.01 % (in 2004), and 49.58 % (in 2005) of the adults captured in the flowers belong to this species. The percentage of F. tritici captured in traps was overwhelmingly higher than the other spec ies. In 2004 adults of F. tritici accounted for 94.00 % of the population captured in sticky traps, while in 2005 they represented 92.00 %. Frankliniella occidentalis (Pergande): This species was found in Florida and Georgia. It was the second mo st abundant species in Georgi a in sticky traps and flowers in both years, as well as in the flowers of north central Florida (NCF01) in 2005. It was the third most abundant in south-central Fl orida (SFL01) and in sticky traps at the NCFL01 site. In Georgia, the relative abundance of F. occidentalis as percentage of the adults captured was higher inside the flowers (36.61% for 2004 and 38.78 % for 2005) than in the sticky traps (4.4 % in 2004 and 6.0% in 2005). Frankliniella fusca (Hinds): This species was the second most abundant species in SFL01 and the third most abundant in NC FL01 Flowers. Only two individuals of F. fusca were captured in Georgia (Other species Table 4-3). Thrips hawaiiensis (Morgan): This species was recorded only at the Florida sites. It was captured inside the flowers as well as on the sticky traps and apparently is more abundant in the southe rn regions (SFL01).

PAGE 90

76 Rapid Determination of the Most common Species Found in Early-Season Blueberry Fields. Based on the information from (Mound a nd Marullo 1996, Moritz et al. 2001, Moritz et al. 2004b, Edwards Unpublished data ) and from personal observations a key is presented that can be used with a 40x compound microscope for the most commonly encountered species of thri ps in blueberry fields. Identification key to the flower thrips of blueberries in Florida and Georgia 1. Four or five pairs of elongated pronotal se tae (Figure 1). Abdominal sternite VII has no discal setae. The first-vein setal-row in the forewing is complete and the setae are uniformly spaced (Figure 2)....(Frankliniella spp.) 2 Figure 1 Figure 2 1. Three or fewer pairs of elongated pronot al setae. Abdomina l sternite VII has discal setae. The first vein setal row in the forewing is incomplete (Figure 3)................................................................................................(Thrips spp.) 5

PAGE 91

77 Figure 3 2. The postero-marginal comb of microtri chia is complete in the middle. The microtrichia are long and irre gular and their base s are broadly tria ngular (Figure 4). Major post-ocular seta are more than of the length of ocellar setae III, and usually extending clearly to the outside of the head (Figure 5)......... F. occidentalis Figure 4 Figure 5

PAGE 92

78 2. With a different combination of char acters from the ones described above. 3 3. Base of the first antennomere restri cted (Figure 6). The reticulation on the metanotum media area is equiangular. No postocular seta I. Wings might be absent in the adult stage... F. fusca Figure 6 3. Base of the first antennomere is swo llen (Figures 7 and 8). The metanotum has no equiangular reticulations, but ir regular longitudinal ones. Post-o cular seta I is present and adults always have wings ...4 4. Base of the first antennomere is swo llen and the edges are more or less sharp (Figure 7). It presents two well developed and sclerotised setae in the second antennal segment. This species is the most abundant in Florida, and very rarely found in Georgia...F. bispinosa Figure 7

PAGE 93

79 4 Base of the first antennomere is swolle n but the edges are not sharp. Setae are less developed and less sclerotized on antennal segment II than in F. bispinosa (Figure 8). This species is very common in blueberry plantings in Georgia ......F. tritici Figure 8 5. Seven or eight antennal segments. Th e postero-marginal micr otrichia are short and irregular in length, they appeared to have their bases fused or more than one microtrichia per base (Figure 9). Sternite V has between 10 and 13 discal setae (Figure 10). Has no equiangular reticulation on the metanotum median area ..T. hawaiiensis Figure 9 Figure 10

PAGE 94

80 5. Always with eight antennal segments. The postero-marginal microtrichia are long slender and irregular. Their base s are broad and clearly not fuse d at the base (Figure 11). Sternite V has between 3 and 9 discal se tae (Figure 12). The metanotum median area presents some equiangular reticulation .......T. pini Figure 11 Figure 12 Discussion This is the first time that a description of the thrips species assemblage and the phenology of flower thrips in relation to early -season blueberries have been investigated. Previous research on the rela tionship between flower thrips and blueberries has been limited to adequate monitoring techniques (Finn 2003). It is clear now that the presence of flower thrips on blueberry fields is highl y correlated with the latitude and with the percentage of opened flowers in the field. Thrips were captured for the first time when close to 10 to 15% of the flowers were opene d, independent of the latitude. Flowers and thrips appeared later in the season for north ern sites compared with those recorded in southern areas. Despite the differences in the dates of blooming among the various sites, thrips populations followed the same pattern with respect to flower opening. The number of thrips captured inside the flowers, as well as the number of thrips captured in sticky

PAGE 95

81 traps increased up to the point where the ma ximum number of flowers is opened in the field. After that time, thrips populations starte d to decrease when the petals of the flowers began to fall and the fruits start forming. Th is correlation between thrips populations with flower phenology as well as with latitude indicates that it mi ght be a correlation between thrips populations and degree-days accumulation. To determine this correlation it will be necessary to establish a base-date or condi tions to begin the accumulation of the degreedays. Future research could be conducted to establish a more accura te model to predict thrips populations in blueberries based on weat her patterns and temperature. Among the thrips captured in Florida, F. bispinosa is the most commonly encountered species. This results complement the species descripti ons made in citrus where 80 to 95% of the thrips captured bel ong to this species (C hilders et al. 1990, Childers et al. 1994, Toapanta et al. 1996, Ch ilders et al. 1998). In Georgia, I did not capture F. bispinosa. I found the most common species to be F. tritici followed by F. occidentalis. This difference might be related to environmental conditions. Other species found in the samples include: F. fusca, T. hawaiiensis, T. pini. A taxonomic key to distinguish the species commonl y found in blueberries was de veloped. This key can be used by professionals who are trying to dete rmine the species assemblage found in earlyseason blueberries. Almost all the thrips found belong to thes e six species. Other thrips encountered sporadically, fewer than 2 specimens collected in total, include Haplothrips victoriensis, F. kelliae, and F. schultzei (Trybom). Blueberry growers will be able to use this information to predict when thrips are more likely to arrive to their fields, base d on the percentage of opened flowers and the latitude at which the farm is located. The dur ation of the flowering season in blueberries

PAGE 96

82 is on average 25 days and coinci des with the time that flower thrips are captured in the fields. This period is not long enough for th rips to have multiple generations in the blueberry flowers because the life cycle is almost as long as th e flowering period in blueberries (Childers et al. 1994, Moritz 1997). This led me to believe that most of the adult thrips collected in the flowers and in the sticky traps migr ated into blueberry plantings from adjacent fields. Previous resear ch indicated that the flower thrips species that were collected in blueberry fields are also found in citrus, wheat, and non-crop plants such as hairy vetch (Vicia villosa Roth), and crimson clover (Trifolium incarnatum L.) during both winter and spring (Toapanta et al. 1996). Adequate management of these alternative hosts for the flower thrips around blueberry fields might help to reduce the immigration of thrips into the plots.

PAGE 97

83 Tables and Figures Table 41. Pearson correlation coefficients for the relationship between percentage of opened flowers and thrips population captu red in sticky traps and inside five blueberry inflorescences. Sample SFL01 (n = 5) NCFL01 (n = 4) SGA01 (n = 5) 2004 2005 2004 2005 2004 2005 Sticky traps 0.807 0.838 0.784 0.953 0.994 0.874 Flowers 0.918 0.052 0.780 0.840 0.986 0.868 Table 42. Dates, latitude, and principal characteristics of flower thrips population in 2004 and 2005 from the samples taken from south-central Florida to southern Georgia. SFL01 represents the farm in south Florida, NCFL01 represents the farm located in north-central Florida, and SGA01 is the farm located in southern GA. Farm Date of first capture* Latitude ( o ) Date of max. population1 Max. population per 5 flower clusters Max. population per trap SFL01 11-Feb-2004 N 28o 043-Mar-2004 16.6 4.7 65.4 14.7 SFL012 15-Feb-2005 N 28o 049-Mar-2005 123.4 29.7 SFL013 2-Mar-2005 8.9 2.5 NCFL01 4-Mar-2004 N 29o 4118-Mar-2004 5.2 1.8 31.6 8.4 NCFL01 10-Mar-2005 N 29o 4124-Mar-2005 23.0 7.5 86.5 10.5 SGA01 14-Mar-2004 N 31o 32 4-Apr-2004 25.2 8.3 255.5 43.3 SGA01 17-Mar-2005 N 31o 321-Apr-2005 4.5 0.4 22.3 3.40 1 Refers to the collection date for the traps which were placed in the field 7 days prior to this date. 2 Information for the thrips capture d on sticky traps in SGL01 in 2005 3 Information for the thrips captured in flower-clusters in SGL01 in 2005

PAGE 98

84 Table 43. Distribution of the thrips species assemblage in Florida and southern Georgia. Percentage of thrips captured per season 2004 2005 Farm Species Flowers Sticky traps Flowers Sticky traps SFL01 Immature 20.4 0.0 23.0 0.0 F. bispinosa 67.2 83.6 61.2 82.4 F. fusca 6.6 10.4 8.4 12.2 F. occidentalis 4.4 5.8 5.8 3.6 T. hawaiiensis 1.0 0.0 1.6 1.2 Other species 0.4 0.2 0.0 0.6 NCFL01 Immature 18.5 0.0 14.7 0.3 F. bispinosa 72.5 93.3 67.3 95.2 F. fusca 3.7 5.0 8.5 3.0 F. occidentalis 4.7 1.3 9.5 1.0 T. hawaiiensis 0.5 0.2 0.0 0.3 Other species 0.0 0.3 0.0 0.3 SGA01 Immature 26.8 0.0 27.8 0.2 F. tritici 44.0 94.0 35.8 92.0 F. occidentalis 26.8 4.4 28.0 6.0 T. pini 2.4 1.2 8.4 0.4 Other species 0.0 0.4 0.0 1.4

PAGE 99

85 Figure 4-1. Phenology of thri ps population on farm SFL01 in 2004. The graph represents average number of thrips captured per sticky trap and the average number of thrips in five blueberry flower-clusters SEM. 0 10 20 30 40 50 60 70 80 90 10-Feb-0417-Feb-0424-Feb-042-Mar-049-Mar-04 DateAverage No. of thrips capture d 10508020 Percentage of opened flowers Sticky Traps SEM Flower SEM 30

PAGE 100

86 Figure 4-2 Phenology of thrips population on farm SFL01 in 2005. The graph represents average number of thrips captured per sticky trap and the average number of thrips in five blueberry flower-clusters SEM. Figure 43 Phenology of thrips popula tion on farm NCFL01 in 2004. The graph represents average number of thrips cap tured per sticky trap and the average number of thrips captured in five blueberry flower-clusters SEM. 0 5 10 15 20 25 30 35 40 45 3-Mar-0410-Mar-0417-Mar-0424-Mar-04 DateAverage No. of thrips captured .109020 Percentage of opened flowers Sticky trap SEM Flower SEM 60 0 20 40 60 80 100 120 140 160 180 15-Feb-0522-Feb-051-Mar-058-Mar-0515-Mar-05 DateAverage No. of thrips captured10508020 Percentage of opened flowers Sticky Traps SEM Flower SEM 30

PAGE 101

87 Figure 44. Phenology of thri ps population on farm NCFL in 2005. The graph represents average number of thrips captured per sticky trap and the average number of thrips captured in five bluebe rry flower-clusters SEM. Figure 45. Phenology of thrips popula tion on farm SGA01 in 2004. The graph represents average number of thrips cap tured per sticky trap and the average number of thrips captured in five blueberry flower-clusters SEM. 0 50 100 150 200 250 300 350 13-Mar-0420-Mar-0427-Mar-043-Apr-0410-Apr-04 DateAverage No. of thrips capture d 15 508030 Percentage of opened flowers 30 Sticky Traps SEM Flower SEM 0 20 40 60 80 100 120 9-Mar-0516-Mar-0523-Mar-0530-Mar-05 DateAverage No. of thrips captured .108020 Percentage of opened flowers Sticky trap SEM Flower SEM 50

PAGE 102

88 Figure 46. Phenology of thrips popula tion on farm SGA01 in 2005. The graph represents average number of thrips cap tured per sticky trap and the average number of thrips captured in five blueberry flower-clusters SEM. 7-Feb14-Feb21-Feb28-Feb6-Mar13-Mar20-Mar27-Mar3-Apr10-Apr17-Apr SFL01 N CFL01 SGA01 Figure 47: Dates of first and la st captures of flower thrips in the various blueberry sites. SFL01 represents south-central Florida, NCFL01 represents north central Florida and SGA01 represents southern Georgia. Solid lines represent 2004 and dotted lines represent 2005. 0 5 10 15 20 25 30 16-Mar-0523-Mar-0530-Mar-056-Apr-0513-Apr-05 DateAverage No. of thrips capture d 10806010 Percentage of opened flowers 40 Sticky Traps SEM Flower SEM

PAGE 103

89 CHAPTER 5 HOST STATUS, INJURY DESCRIPTION, AND DETERMINATION OF ECONOMIC INJURY LEVELS OF FLOW ER THRIPS FOR EARLY-SEASON BLUEBERRIES The recognition of injury and determination of Economic Injury Levels (EIL) are of major importance for the success of integrated pest management (IPM) programs. These concepts are the foundation of the decision-m aking process in agri culture. Neve rtheless, few studies have been publishe d concerning EIL for blueberries. This may be due to the variability of the factors invol ved in the calculation of EIL (Poston et al. 1983). Most EIL variables depend on market values which can exhibit considerable variation. It is useful to have an EIL for flower thrips in earlyseason blueberries because this is a high-value crop and growers need to make important deci sions, before economic losses are incurred. For this study, I used definitions publis hed by Stern et al. (1959). The authors define Economic Damage (ED) as the amount of injury that will justify the cost of artificial control measures and EIL as the lowest population density that will cause this damage. Economic threshold (ET) was defi ned as the density at which control measures should be initiated to prevent an increasing pest population from reaching the EIL. Thrips are considered to be key pests fo r many crops and have been reported to affect more than 500 plant species (Morit z et al. 2001, Moritz et al. 2004b). Among blueberry producers, flower thrips is consider ed as one the key arthropod pests. However, there have been no descriptions of the injury and damage in flicted by flower thrips in blueberries. Mound (2005) expl ains the difference between proper hosts a nd alternative

PAGE 104

90 hosts for thrips and why it is important to determine the type of hosts and their interactions with thrips. The author describes primary hosts or real hosts as plants where thrips feed, shelter, and reproduce. Secondary hosts are plants used by thrips as shelter and food but not as reproduc tive sites in the field. There are a few publications dealing with th e EIL of flower thrips in economically important plants. Shipp et al. (2000), determin ed the EIL for wester n flower thrips on greenhouse cucumber. Their results indicated that greenhouse cucumbers have a period of tolerance, which lasts until the eighth-w eek of production. Cucumbers appear to be more sensitive to primary damage inflicted by thrips (damage to the fruit) than by secondary damage (damage to the leaves). The objectives of this study were 1) to determine the type of host, primary or secondary, that blueberries repres ent for flower thrips, 2) to describe for the first time the type of damage inflicted by flower thrips on early-season blueberries, and 3) to determine the EIL of flower thrips for the two most popul ar rabbiteye blueberry cultivars, Climax and Tifblue. Material and Methods Host status of early-season bl ueberry bushes for flower thrips To clarify the host status of early-season blueberries 100 opened flowers from the two main species of early-season blueberr ies (50 southern highbush and 50 rabbiteye) were collected in commercial farms lo cated in north-central Florida (N 29o 41 W 82o 11) and in south-cen tral Florida (N 28o 04 W 81o 34). This 2 x 4 factorial experiment has as main effects the blueberry species (rabbiteye and southern highbush), and the blueberry tissues (fruit, ovary, styles, petals). The flower s were brought for processing to the Small Fruit and Vegetable IPM Laboratory at the University of Florida in Gainesville.

PAGE 105

91 The flowers were manually dissected into pe tals, ovary, and styles The floral tissues were then placed into groups of ten in 100 x 15 mm polystyrene Petri-dishes (Fisher Scientific, Pittsburgh, PA). In addition, one wet circle of filter paper Qualitative P8 (Fisher Scientific, Pittsburgh, PA) was placed in each Petr i-dish to keep the humidity high. Petri-dishes were left at room temperature (27C) for 15 days, and all of the thrips that emerged over this period were counted a nd moved to another Petri-dish with green beans and Bee pollen (Y.S. Organic bee farm s, Yonkers, NY) as food sources to allow them mature. The Petri-dishes were seal ed using Parafilm M (Pechiney Plastic Packaging, Chicago, IL) to prevent thrips from escaping. At the end of the blueberry flowering season, I collected green fruits a nd processed these using the same procedure as with the flowers to provide informati on about emergence of flower thrips. Every second day the Petri-dishes were opened and observed for the presence of first instar thrips. These larvae were count ed and taken from the Petridishes. The number of larvae emerged in each Petri-dish was recorded a nd analyzed according with the design using LSD as test for mean separation. ( = 0.05) (SAS Institute In c. 2002). Economic Injury Level and injury description Two of the most popular cultivars of rabbite ye blueberries, Climax and Tifblue, were selected for this study. Both cultivars we re located in a commerci al farm in southern Georgia (N 31o 32 W 82o 31). The farm is managed with the standard practices used by commercial growers to maintain a blueberr y crop. At the beginning of the flowering season (March 9, 2005 and March 8, 2006) I se lected five flower-clusters with five flowers each per bush. Five non-consecutive bushes from each cultivar were used as blocks in a completely randomized block de sign. There were five replicates for each cultivar. The flowers were protected with bags made of antivirus/nothrips screen (thread

PAGE 106

92 count 81 x 81, BioQuip Products, Rancho Dominguez, CA). The screened bags were handmade measuring 10 x 10 cm, and were placed on the selected inflorescences. These bags were sealed using small binder cl ips (ACCO Lincolnshire, IL), by folding the opened side three times and then placing the clip over the folding. Two weeks later when the flowers opened and were receptive, I manually pollinated them using a mixture of pollens collected earlier the same day from various cultivars present in the field to ensure cross-pollination and a ppropriate fruit setting. A know n number of adult flower thrips, Frankliniella spp., was released inside the bags when pollination was conducted. For 2005, the treatments used were a control (no thrips), and 2, 10, or 20 thrips per flower. In 2006, my treatments include a contro l (no thrips), 4, 10, 20 thrips per flower. The thrips released in the bags were collect ed from various wild flowers and blueberry flowers the day before the releases. The thrips collected were separated into adults and immatures. Only the mature stages belonging to the genus Frankliniella were used in this trial. The formed fruits were collected in the green stage two to three weeks after the release of the thrips inside the bags and brought to the Small Fruit and Vegetable IPM Laboratory at the University of Florida for observation. To determine the EIL, I used the defi nitions and formulas described by Pedigo (1986) (Equation 5-1). Based on the data collected, I conducted a linear regression between the number of thrips pe r flower released in the bags and the percentage of fruits formed inside the bags. The assumptions (cost of control, price, market values etc.) used to calculate the EIL were based on the growe rs and experts experience and on various publications (Food and Agriculture Organiza tion of the United Nations (FAO) 2004, Pollack and Perez 2004, Lyrene 2005, NASSUSDA 2006b). These sources enabled us to

PAGE 107

93 establish the most accurate and functional vari ables to estimate the EIL. I compared the slopes of the linear regressions between the number of thrips per flower and the number of flowers formed using a Student t-test ( = 0.05) (Ott and Longnecker 2004). Fruits that presented signs of thrips injury were photographed and described. (Equation 5-1) K D I V C EIL 5-1 Correlation between the number of thrips in side the flowers and on sticky traps To facilitate the use of economic injury levels (EIL) by commercial growers and to find a non-destructive way to determine thrips populations affecting blueberries, I decided to correlate the number of thrips captured in sticky traps with the number of thrips recorded inside blueberry flowers. Tw o farms were selected in southern Georgia (N 31o 32 W 82o 28) and (N 31o 32 W 82o 31). In each farm, I deployed 10 white sticky traps, in 2004 and 2005, located inside th e blueberry-bushs ca nopy (Chapter 3) and randomly distributed them in one ha. Traps were collected every week, simultaneously five inflorescences were collected from th e same blueberry bush where the trap was placed. Traps were randomly rotated within th e selected hectare every week for 4 weeks in 2004 and 3 weeks in 2005. The total num ber of thrips captured on the traps was counted. Thrips taken from the flowers were processed using the shake and rinse method described in chapter 3. The number of thrips from the flowers was tabulated. To linearize the relationship between the num ber of thrips captured in st icky traps for a week and the number of thrips inside the flowers, the data had to be transformed using the natural logarithm. The regression was performed using SigmaPlot (SYSTAT Software Inc. 2006).

PAGE 108

94 Results Host status of early-season bl ueberry bushes for flower thrips The emergence of larvae from the blueberr y tissues confirms that blueberries are a proper host for flower thrips. The analysis of the data collected showed no effect of the blueberry species on the number of thrips emerged from the flowers (F = 0.47; df = 1,92; P = 0.49) and no significant interaction betw een the blueberry species and the tissue from where the thrips emerged (F = 0.16; df = 3,92; P = 0.9235). The resu lts obtained from this analysis and illustrated in Figure 51 indicated that significantly more thrips emerged from the petals (11.3 2.5 thrips per ten corollas) than from any other tissues in the flower (F = 13.28; df = 3, 92; P < 0.0001). The number of thrips that emerged from the ovaries (4.4 1.2 thrips per ten ovar ies) was significantly lower than the number emerged from the petals, but it was signifi cantly higher than the number of thrips emerging from the styles (0.1 0.1 thrips per ten styles) and the fruits (0.0 0.0 thrips per ten fruits). Finally, there was no differe nce between the number of thrips emerging from styles and fruits, but both of these we re significantly lower than the number of larvae emerging from petals and ovaries (Figure 51). There is no differences between the number of thrips emerging from rabbiteye flowers (4.2 1.1 thrips per ten flowers) and highbush flowers (3.1 0.8 thrips per ten flowers) (F = 0.47; df = 1,92; P = 0.49). Economic Injury Level and injury description The results of the regression correlating th e average number of thrips released per flower and the number of fruits formed in each treatment (Figure 52), provides the information needed to calculate the value of the injury per insect I. The value of the slope in the equation indicates a 0.4032 % reduction in production pe r adult thrips per flower in Climax and 0.4515 % reduction in production per adult th rips per flower in

PAGE 109

95 D I V C EIL Tifblue. Using average densitie s (2,471 plants / ha), fruit wei ght (1.2 g per fruit), flower setting (60%), and average production (16, 812 kg / ha), and 9,500 flowers per plant (P. M. Lyrene, personal communica tion) for rabbiteye blueberr y commercial production. The value for injury per insect in Climax (Ic) is approximately 1.71x10-10 proportion damage/[thrips/ha]. The value for the damage per unit injured for Climax (Dc), is 16,812 (Kg/ha)/proportion damaged. For Tifblue the value for It is approximately 1.92 x 10-10 proportion damage/[thrips/ha] and the va lue for the damage per unit injured for Tifblue (Dt) is equivalent to 16,812(K g/ha)/proportion damaged. Equation 51: The value for D in both cases was calcul ated under the assump tions that flower thrips adults arrive at the flowers at the mo ment of pollination and that the plants have no mechanisms of compensation for fruits not form ed due to thrips injury. The value of D and the value of the total produc tion per ha are equal. If the proportion of damage is equal to 1 (100% of the fruits) the damage per unit in jured is equivalent to the total production. The injury per insect value I is not very variable and depends on the tolerance of the inflorescences to thrips damage. When I co mpared the slopes of the regressions between the average number of thrips released inside the bags and the number of fruits formed, I found non-significant differences between th e slope for Climax and Tifblue (F = 0.66; df = 1, 98; P = 0.4205). Two of the most unpredictabl e variables that change fr om season to season in the calculation of EIL are the cost of the cont rol C and the value of the product V. However, using the average cost of an applic ation of Malathion 5EC, which is the most

PAGE 110

96 commonly used insecticide for controlling flower thrips, the value for C is approximately $185.32/ha (O. E. Liburd, Pe rsonal communication). The average value of blueberries for processing in Georgi a for 2005 was $1.95 / Kg (NASS-USDA 2006b). Based on the collected data, the EIL for Climax is 33,057,666 thrips / ha or approximately 14 thrips per 10 flowers. In the case of Tifblue, the EIL is 29,441,983 thrips per ha or approximately 13 thrips pe r 10 flowers. I also calculated the EIL for SpinTor 2SC because it is the most popular reduced-risk insecticide used to control flower thrips in blueberries (Liburd and Finn 2003). The value to control thrips using SpinTor 2SC (Vc) is approximately $249.17 / ha when all the other variables were kept constant. The value for the EIL using SpinTor 2SC as the control is approximately 44,447,327 thrips / ha or 19 thrips per 10 fl owers for Climax and 40,204,430 thrips / ha or 17 thrips per 10 flowers for Tifblue. After analyzing the fruits a nd the injuries inflicted by thrips, I found a wide range of injuries. The injuries inflicted by thrips could be divided into three categories: fruit dehydration, feeding injuries and oviposition in juries (which include larval emergence). Fruit dehydration damage was only found in in florescences where 20 thrips per flower were released. The damage on the flowers ovary during fruit form ation was severe and the fruits formed were dehydrat ed. A higher than usual portion of the fruits failed to set, which might explain in part the reduction in the blueberry production in inflorescences with flower thrips. Thrips feeding damage is inflicted by adults and the two larval stages. The symptoms shown by this damage are similar to those found on mangoes when F. occidentalis have feed on the flowers [similar to the picture taken by M. Wysoky and shown in Childers (1997)] (Figure 53b). Oviposition a nd emergence injuries are

PAGE 111

97 inflicted when the females lay their eggs and the larvae emerge from the ovaries, respectively. This damage might be magnified with the cell division of the tissue and be more evident in formed fruits than in the ovary of the flowers (Figure 53c) (Childers 1997, Kirk 1997a). Correlation between the number of thrips in side the flowers and on sticky traps Our regression shows a high correlation be tween the number of thrips in sticky traps and the number of thrips inside blueberry flowers. Th e equation that describes the line (Figure 54) corresponds to Ln (Tf+1) = -0.06 + 0.709 x Ln (Tt+1) (F = 177.37; df = 1, 128; P < 0.0001). In the equation Tf represents the number of thrips recovered from 5 blueberry flower-clusters using the sh ake and rinse method (Chapter 3); Tt represents the total number of thrips capt ured in white sticky traps located inside the canopy of blueberry. Pearsons coefficient (r = 0.7621) indicates a high correlation between these two variables. Based on the information presented in hi s study I was able to approximate the economic injury level to between 45 (Tifblue ) and 50 (Climax) thri ps per trap at the beginning of the pollination period in the 2005 se ason in the case of Malathion 5EC. In the case of SpinTor the EIL is between 64 (T ifblue) and 73 (Climax) thrips per trap at the beginning of pollination. Discussion These results confirmed that blueberry bus hes are a true host for flower thrips. These thrips are reproducing in the flowers of blueberries and emerge before the full formation of the fruit. Thrips prefer to la y their eggs in mature non-expanding tissue to avoid having their eggs crushed by growi ng cells (Terry 1997). Flower thrips on blueberries have a similar ovipos ition behavior as flower thri ps in citrus. As described by

PAGE 112

98 Childers (1991), flower thrips lay their eggs principally in the pi stil-calyx area and the petals in oranges. A simila r situation was found on bluebe rry flowers where most eggs were laid either in the ovary or in the peta ls, while no larvae emerged from the fruits. The data indicate that flower thri ps in blueberries prefer to la y their eggs in young flowers so that the eggs will have time to develop into larvae before fruit formation starts. I also found non-significant differences between the number of larvae emerging from the two blueberry species, southern highbush and rabbiteye, cultivat ed in southeastern U.S., which indicates that flower thrips do not s how significant ovipositi on preference to either blueberry species in the field. Once I dem onstrated that early-season blueberries are primary hosts of flower thrips, I studied the implications of the presence of these insects in commercial plantings. I began with the calc ulation of an EIL for which I used the two most popular cultivars of rabbiteye blueberr ies, Climax and Tif blue. I found that flower thrips are capable of reducing the yi eld of blueberries to economically damaging levels when high populations (2 thrips pe r flower on average) are found inside the flowers during the beginning of the pollination period. However, this EIL was calculated based on information from the 2005 season, whil e some of the variables used correspond to this year and will change in the future depending on market values. The cost of treatment C varies depending on the cost of ma nual labor, the type of product applied, and the type of equipment used for the appl ication (assuming that only chemical control is used). The value of V changes with supply and demand, harvest time, product quality etc. Due to this variation, the actual valu e for an EIL is farm-, and year-specific. The most important variable calculated and the least likely to change is the slope in the correlation between flower thrips density inside the fl owers and the reduction in the

PAGE 113

99 yield. This slope is the key factor in the calcu lation of injury per in sect (I), which is the variable less likely to change as it does not depend on the market. Some of the reasons for which I may vary include changes that affect flower thrips interest in feeding or oviposition on blueberry flower s. This might include genetic improvement of the cultivars to produce feeding or oviposition deterrents, clima tic conditions, presence of natural enemies etc. I found non-significant di fferences between the levels of injury inflicted by thrips in relation to the pest density when Climax and Tifblue where compared. This might indicate a reduced variation between cultivars of the same species. However, the evaluation of more cultivars is necessary to strengthen this conclusion. I also determined that the damage associated with flower thrips in blueberries is not only linked to yield. Although reduction in the yield is the most important consequence of an unmanaged population of fl ower thrips, it seems that the quality of the fruits might also be compromised. Injuries to fruit, principally feeding inju ries, might reduce the grade of the fruit because of their visibility. Oviposition-emer gence injuries due to their color and size are almost invisible when the fruits mature, thus might not represent a big problem for growers and customers. However, these injuries have the potential to facilitate the acquisition of secondary infection such as Botrytis cinerea, which is the principal decaying agent for blueberries in post harvest (Sargent et al. 2006). We found a high correlation between the num ber of thrips captu red in sticky traps and the number of thrips found in blueberry flowers. This correlation will increase the accuracy for monitoring flower thrips in bl ueberry fields and re duce the effect of destructive sampling in commercial settings. Th e values for the EIL in flowers as well as in traps are very high. From field observations described in other chapters the number of

PAGE 114

100 thrips captured during the week in which flowers are opening is very low. The usual values obtained for flowers are approximately 5 to 10 thrips per fi ve flower-clusters, which means an average of 0.3 thrips per fl ower (Chapter 3). The differences observed in the EIL when calculated for Malathion 5EC and for SpinTor 2SC illustrates the variability of its value. Du e to this variability, the cal culations of EIL should be performed almost individually for each farm, depending on type of control, plant density, varieties, price of their product, etc.

PAGE 115

101 Tables and Figures Figure 51: Average number of larvae emerge d from individual tissues of 10 flowers and fruits. Comparisons were conducted using LSD test = 0.05 0 2 4 6 8 10 12 14 16 PetalsOvaryStyleFruit Blueberry flower tissuesAverage number of thrips emerging per ten flowersab cc

PAGE 116

102 Figure 52: Linear re gression showing the average number of thrips released per flower and the percentage of formed fruits in two principal cultivars of rabbiteye blueberries Climax and Tifblue Figure 53: Various thrips injuries inflicte d by flower thrips in blueberry flowers. a) Represents a healthy fruit, b) S hows feeding injury, and c) Shows oviposition/emergence injuries y = -0.4032x + 62.726 r2 = 0.9377 y = -0.4515x + 65.088 r2 = 0.912 45 55 65 05101520 No. of thrips per flowerPercentageof fruits .Climax Tifblue x a b c

PAGE 117

103 Figure 54: Regression illust ration the number of thrips captured on white sticky traps for a week in relation to the number of flower thrips captured in five inflorescences collected in the same bush. The graph represents the tendency SEM y = 0.7087x 0.0667 r2 = 0.5808 0 1 2 3 4 5 6 01234567 Ln (No. of thrips in traps)Ln ( No. of thrips in flowers)

PAGE 118

104 CHAPTER 6 EFFICACY OF REDUCED-RIS K INSECTICIDES TO CONTROL FLOWER THRIPS IN EARLY-SEASON BLUEBERRIES AND THEIR EFFECT ON ORIUS INSIDIOSUS, A NATURAL ENEMY OF FLOWER THRIPS Flower thrips are among the most damaging in sect pest in the pr oduction of early-season blueberries (Finn 2003, Liburd and Finn 2003, Liburd and Arval o 2005, 2006, Liburd et al. 2006). The use of chemical means to control th rips began in the early 1900s. Since then, no insecticide has been developed to target exclusively thysanopteran pests. All of the insecticides used for thrips management were designed to co ntrol other insects and then tested on thrips. However, their effect on thrips might be limited du e to differences in behavior and feeding habits compared with other major pests (Lewis 1997b). In secticides recommended for thrips control by the manufacturers are usually systemic and st omach poisons. There are a few insect-growth regulators and contact-only insecticides but these are very limited (Lewis 1997b). The United States congress (1996), warned of the use of traditional chemistries and encouraged the development of new chemistries th at are less toxic to non-ta rget organisms and to the environment in general. There are currently 60 chemistries that are considered as reducedrisk insecticides, biopesticides, and organophosphate s (OP) alternatives as described in by the IR-4 project (IR 4 project 2006). Nine of these chemistries are labele d or in process to be labeled to be used in blueberries. With in this group, only five have sh own potential to control thrips: flonicamid, novaluron, spinosad, thiamethoxam, a nd zeta-cypermethrin (United States Congress (104th) 1996, IR 4 project 2006). Most of the chemical control for flower thri ps in blueberries has relied on the use of malathion, an OP insecticide, and on spinosad, a reduced-risk naturalyte (O. E. Liburd, personal

PAGE 119

105 communication). Malathion has been used to cont rol flower thrips in various crops including garlic, leek, onion, cucumbers, endive, etc. (Micro Flo Compa ny 2006). Spinosad is being used as an organic OMRI-listed insecticide in its form ulation Entrust, or as non-organic formulation as SpinTor 2SC. Spinosad is labeled to control thrips in bushberries, citrus, Brassica vegetables, cucurbits, etc. (Dow AgroSciences 2006b, 2006a). However, concerns for the development of insect resistance due to the use of one chemis try and a limited number of modes of action, prevails. Several studies have repo rted various degrees of thrips re sistance to several insecticides. Jensen (2000), compiled a list of Frankliniella occidentalis (Pergande) wild and laboratory populations, that have been reporte d as resistant to insecticides. The list includes seven OP, nine carbamates, eight pyrethroids and six other chemistries, which include endosulfan, DDT, and the OP-alternative imidacloprid. In a recent study in Australia, Herron a nd James (2005) reported resistance of F. occidentalis to chlorpyrifos, dichlorvos, and malathion. At the same time resistant individuals where detected for acephate, dimethoate, endosulfan, fipronil, mathemidophos, methidathion, and spinosad. Labor atory selections using fipronil and spinosad were successful rearing increasingly resistant popul ations of thrips, this was the first report of induced resistance to spinosad and fipronil to Frankliniella thrips. Another consideration when using chemical altern atives to control thrips in integrated pest management (IPM) is the effect of these insectic ides on non-target organisms, principally natural enemies. The effect of traditi onal and reduced-risk in secticides to non-target organisms is well documented. To mention a few examples, studi es on natural enemies of thrips include Geocoris punctipes (Mizell and Sconyers 1992, Elzen et al. 1998, Elzen 2001, Myers et al. 2006), Orius spp. (Elzen et al. 1998, Ludwig and Oetting 2001, Studebaker and Kring 2003), and in Amblyseius spp. (Castagnoli et al. 2002). All these studies demonstrated the lethal and sub-lethal

PAGE 120

106 effects of these chemical products on natural enemies. Selective inse cticides will facilitate adequate biological control. Insecticides that have little effect on non-target organisms are desirable due to their compatibility with natural enemies, which is the ba sis for a successful IPM program. The present study shows the relationshi p between selected reduced-risk insecticides that have the potential to cont rol flower thrips in early-s eason blueberries and to replace traditional insecticides in commercial fields. The results from the field were complemented with laboratory bioassays to determine the effect of these insecticides under controlled conditions on flower thrips and on Orius insidiosus Say, one of the most common and effective natural enemies for thrips control in Flor ida (Funderburk et al. 2000). Material and Methods Field Trials Field experiments were conducted from 2004 to 2006 in four commercial blueberry farms located in Florida and southern Georgia. Three farms in Florida (IFL04 located at N 28o 04 W 81o 35, IFL05 located at N 27o 30 W 81o 31, and IFL06 N 29o 40 W 82o 11) and one farm located in Georgia (IGA04 located at N 31o 31 W 82o 27) were selected for these trials. Each year the insecticides that s howed the most potential from pr evious season and a few more insecticides were compared. A selection of reduced -risk insecticides, traditional insecticides, and OP alternatives was selected to determine thei r efficacy. Treatments were sprayed after sunset because of the concerns about th e effect of these insecticides on populations of honeybees used for blueberry pollination. The plot sized varied depending on the farm conditions and the areas where we were allowed to work during each one of the seasons. To determine the effect of insecticides on th rips populations, I deploye d a white sticky trap in the middle of each treatment and randomly collected 5 flower-clusters from the blueberry bushes holding the traps. Sticky traps were depl oyed in the field one week prior to the

PAGE 121

107 application of the treatm ents. These traps were collected befo re the application and replaced by a new one, which was located in the same position a nd collected one week after the treatment. The experimental design in all cases was a complete ly randomized extended block design. The block in my experiments was represented by the applic ation (first application and second application) and each one of the blocks had four replicates of each of treatment evaluated. Due to the highly aggregated pattern shown by flower thrips in ea rly-season blueberry fields (chapters 3 and 4), I decided to analyze the effect of these insecticides using the growth rate (r). This growth rate (r) represents the change in thrips population due the inse cticide application. An r value of 1 represents no changes in the population afte r the application of the treatment. If r > 1, then more thrips were captured during the week after the application than the week prior to the application; and a value of r < 1 represents a fall in the number of thrips in the week after the application compared with the week before the application. In the case of sticky traps, I used the total number of thrips captured duri ng the week immediately after treatment divided by the total number of thrips captured at the same location th e week prior to the trea tment. To analyze the number thrips inside the blueberr y flowers, I used the shake a nd rinse method (C hapter 3) and divided the number of thrips ex tracted from five flower-cluster s one week after the treatment by the number of thrips extracted from the same am ount of flower-clusters co llected the day of the treatment in the same blueberry bush. Comp arisons were conducted using a one-way ANOVA and LSD tests for mean comparison (SAS Institute Inc. 2002). The experimental design in all cases was a completely randomized extended bloc k design. The block in my experiments was represented by the application (f irst application and second app lication) and each one of the blocks had four replicates of each of treatment evaluated.

PAGE 122

108 IFL04: This farm located in south central Flor ida was used for the trials conducted in 2004. Treatments were sprayed two-times during flowering of blueberry bushes with a 14 day interval between sprays as recommended by the Small Fruit and Vegetable IPM Laboratory at the University of Florida (O.E Liburd Persona l communication). Insecticide treatments were applied at the doses recommended by the manufacturer. Treatments were sprayed using fourgallon backpack sprayers (Lowes North Wilkesbor o, NC). The sprayers were manually pumped to a maximum capacity and the handle was pumped once every five seconds to maintain a homogeneous pressure. A comple tely randomized extended block design with eight treatments, four replicates and two blocks were used to ev aluate insecticides. The treatments used were: Malathion 5 EC at 139.8 g a.i. / ha, Calypso 480C (thiacloprid) at 116.92 g a.i. / ha, Assail 70WP(acetamiprid) at 112.78 g a.i. / ha, Novaluron 174.08 g a.i. / ha, SpinTor 2SC (spinosad) 105 g a.i. / ha, Knack (Pyriproxyfen) at 120 g a.i. / ha, GF968 (spinosad experimental insecticide) at 52.63 g a.i. / ha, and an untreated contro l. Each plot measured 84 m2 with a 15 m buffer zones between plots. The selected plot wa s covered with black bird netting to reduce the attack of birds during harvest. IGA04: This farm located in southern Georgia wa s planted in rabbiteye blueberries and the trial was conducted during the 2004 flowering season. Selected plots for this trial covered 193 m2 with 12 m buffer zones between the plots. There were eight treatments and four replicates of each treatment using the same experimental desi gn as above. Insectic ides treatments were applied to plots using a 400 gal., tractor-mounted orchard airblast sprayer. Treatments evaluated included Malathion 5 EC at 139.8 g a.i. / ha, Ca lypso 480C (thiacloprid) at 116.92 g a.i. / ha, Assail 70WP(acetamiprid) at 112.78 g a.i. / ha, Pe destal (novaluron) 174.08 g a.i. / ha, SpinTor

PAGE 123

109 2SC (spinosad) 105 g a.i. / ha, Knack (pyrip roxyfen) at 120 g a.i. / ha, Actara 25WG (thiamethoxam) 78.80 g / ha, and an untreated control. IFL05: Located in south Florida, this farm used low density southern highbush blueberries. Treatments were sp rayed during the 2005 flowering seas on using an airblast sprayer as described above at the ma nufacturer recommended doses. Th e treatments included were Malathion 5 EC at 139.8 g a.i. / ha, SpinTor 2SC (spinosad) 105 g a.i. / ha, Assail 30SG (used as TD 2480 acetamipridexperimental insecticide) 5.4 oz/acre, Assail 70WP (acetamiprid) 112.77 g / ha, Diamond .83EC (novaluron) 145.35 g / ha Actara 25WG (thiamethoxam) 78.80 g / ha. Treatments were applied in plots of 471 m2 with 9 m of buffer zone between treatments and 41 m between blocks of 7 treatments. E ach treatment had four replicates. IFL06: This farm was located in north-central Florida and was planted with southern highbush blueberries. Treatments were spraye d during the 2006 flowering season using a CO2 sprayer calibrated at 23 PSI and us ing a Teejet hollow cone spray core D3 disk DC 25 (Spraying systems Co. Keystone Heights, FL). I used four replicates per treatment and the plots were distributed in a completely randomized exte nded block design. Each plot covered 219.46 m2 and buffers between treatments were 30.4 m long with 4 m between rows. The treatments used in this farm were: Malathion 5 EC at 139.8 g a.i. / ha, SpinTor 2SC (spinosad) 105 g a.i. / ha, Diamond .83EC (novaluron) 145.35 g / ha, Assail 30 SG (acetami prid) 113.48 g / ha, Actara 25WG (thiamethoxam) 78.80 g / ha, Coragen 20S C (rynaxpyr formally known as DPX-E2Y45) 98.63 g / ha, and an untreated control Laboratory Trials During the laboratory trials, I ev aluated the eight most promis ing treatments studied in the field. Laboratory trials we re divided into two parts 1) toxicity of insecticides to flower thrips and 2) the effect of the insecticides on non-target or ganisms. With regard to non-target organisms, I

PAGE 124

110 selected O. insidiosus because it is one of the principal natural enemies of flower thrips in North America (Van de Veire and Degheele 1995, Funderburk et al. 2000, Shipp and Wang 2003, Liburd and Arvalo 2005). Both organisms were tested in similar aren as, a 300 ml white pol yethylene jar (B & A Products, Ltd. Co., Bunch, Oklahoma). The lids of these jars were modified in such a way that only the rim remained. The jars were cove red with non-thrips mesh (Bioquip. Rancho Dominguez, CA) and the modified lids were screw-on over the mesh to prevent the insects from escaping. This modification allowe d ventilation of the arenas. Fo r both insects, two green beans were used as the substrate for each treatment. Th e green beans were cleaned using a solution of 0.6% sodium hypochlorite and de-ionized (DI) wate r for ten minutes, then rinsed with DI water and air dry for 2 h. Green beans were sprayed usi ng a hand-held spray bottle that released 2 ml of solution per spray (two sprays per treatment). Th e green beans were then air-dried for two hours and then placed inside the arenas. Ten O. insidiosus were released into each one of the arenas. The number of live O. insidiosus was recorded at selected times after the release. Orius insidiosus individuals were selected from a laboratory colony which started approximately 2 mont hs prior to the start of the experiments. Orius insidiosus was initially obtained form Ko ppert Biological Systems (Romolus, MI). During this trial, I added five live adult thrips, whic h were not exposed to insecticides, to each one of the replicates every two hours. Thrips that were found dead inside the O. insidiosus arenas were removed. The thrips addition was done with the objective of feeding the predators as if they were under field conditions. Treatments were prepared at the same concentrations as in field experiments assuming 935 liters / ha (100 U. S gallons / acre) unless otherwise specified on the label. The treatments selected for O. insidiosus were Actara 25WG (thiamethoxam) at 78.80

PAGE 125

111 g a.i. / ha, Assail 70WP (acetamiprid) at 112.77g a. i. / ha, Malathion 5EC (malathion) 139.8 g a.i / ha, and SpinTor 2SC (spinosad) 105 g a.i. / ha. These insecticides were either as the most commonly used or they have the highest potential to control thrips in blueberry fields as observed in the field experiments. This experime nt was designed as a completely randomized experiment with six replicates. and analyzed hour by hour using a one-way ANOVA table and the treatments were compared using LSD mean separation tests ( = 0.05). Flower thrips were collected in blueberry fields in north-cen tral Florida and brought to the laboratory for identification. Frankliniella bispinosa (Morgan) was selected because it was the most abundant thrips recorded in Florida accordi ng to our field survey (C hapter 4). Thrips were kept under laboratory conditions (27oC and 70% RH) for two days be fore the trials began. The colonies were fed with a mixt ure of honey and pollen spread over green beans, which were cleaned using a 0.6% sodium hypochlorite and de-i onized (DI) water as described above. Activeadult thrips were selected for the trials and di vided into groups of ten. These were randomly released into the arenas once th e treated green beans were dry a nd in place. Experiments were organized in completely randomized designs with five replicates. the treatment tested on the thrips included Actara, Assail 70WP 112.7 g a.i. / ha, Calypso 480C at 116.9 g a.i. / ha, Kanck at 120 g a.i. / ha, GF 968 (an spinosad experimental ) at 52.63 g a.i. / ha, Malathion at 5EC 139.8 g a.i. / ha, Novaluron, SpinTor 2SC at 105 g a.i. / ha and an untreated control. Analysis was conducted using ANOVA tables and LS D mean separation analysis ( = 0.05) (SAS Institute Inc. 2002). The experiments lasted until no more mortality was observed in the treatments, for thrips the laboratory experiment lasted 12 h and for O. insidiosus lasted 24 h. Preliminary experiments with O. insidiosus had indicated that the insecticides ha d a delay effect on these predatory

PAGE 126

112 insects. During the preliminary experiments c onducted under the same conditions, most of the deaths occurred between 10 and 20 h after exposure to the treated green beans. For this reason, most of the observations were taken during this inte rval. Results and Discussion Field trials IFL04: When observing the effectiveness of the insecticide on the populations located inside the flowers in this farm, only tw o insecticides had a significantly lower r value than the control: Novaluron (r = 1.1 0.8) and Assail 70WP (r = 0.55 0.08) (F = 9.24, df = 7, 53; P < 0.0001). Novalurons r value was significantly lower than Malathion 5EC, Knack, and GF968, but it was significantly higher than Assail 70W Assail 70 WP (acetamiprid) was the only compound that reduced thrips population in side blueberry flowers as shown by r < 1(Figure 61). In the floating populations, t hose collected in the sticky traps, similar results were observed. Assail 70WP had the lowest r value and it showed to be significantly lower than all the other treatments including the control. The growth rate (r) for Novaluron and Malathion was significantly lower than SpinTor 2SC but not si gnificantly different from any of the other treatments with exception of Assail, (F = 5.22, df = 7, 53; P < 0.0001). SpinTor had the highest r value in the sticky traps (Figure 62). IGA04: The results obtained in Georgia in 2004 did not show significant differences among the treatments (F = 0.66, df = 7, 54; P = 0.7456). However, the floating population in all of treatments decreased during this season in si milar proportions, which might indicate that the reduction in thrips population was indepe ndent of the insecticide treatments. IFL05: There was very limited activity of thri ps during this flowering season on the selected farm. For this reason, it was not possibl e to collect enough data for a robust analysis to show the effect of these insectic ides on thrips population inside the flowers. However, I observed

PAGE 127

113 increasing number of thrips captured in the wh ite sticky traps. In all the treatments, the populations increased possibly as a result of immigration, since th e amount of thrips inside the flowers where they reproduce (Chapter 4) was ve ry limited. None of treatments was significantly different from the control. However, Actara 25WG and Assail 70WPsi gnificantly reduced the population growth when compar ed with Malathion 5 EC, (F = 4.79, df = 6, 48; P = 0.0004), but they were not significantly different from any of the other treatments including the control (Figure 64). IFL06: On the experiments conducted in Fl orida in 2006, none of treatments was significantly different from th e control when comparing the r value for the sticky traps (F = 4.01, df = 6, 48; P = 0.0016) or inside the flowers (F = 0.48, df = 6, 48; P = 0.7275). The r value for the thrips captured in the sticky traps increased overtime independent of treatment applications, with exception of Coragen 20SC (r = 0.93 0.21), which was the only treatment with r < 1. Coragen 20SC had a significantly lower r value than Diamond .83EC. However, the comparisons of all other treatments were non-significant (Figure 65). On the established populations collected from the flowers, the situation was si milar to the situation encountered in the sticky traps. None of treatments significa ntly reduced the population growth (r) when they were compared (Figure 66). Laboratory Trials Thrips bioassay All the insecticides were effective against thrips in the laboratory. From 1h after the release of the thrips in the arenas, all the treatmen ts had significantly less thrips surviving when compared with the control (F = 13.86; df = 8, 36; P < 0.001). After 1 h of contact with the treated green beans, all the insec ticides killed half of the insect s exposed to the treatments. Three of the treatments, Actara 25 WG, Assail 70WP, and Malathion 5EC, reduced the population to

PAGE 128

114 less than one third of the original size w ithin the first hour (Table 6-1). The only insecticide that killed all the thrips exposed to it was SpinTor 2SC. This insec ticide in less than four hours reduced the population to zero. However, this result was not significantly different to Assail 70WP, Knack, and Malathion 5EC. Six hours after the thrips were exposed to the treated green beans, all the treatments were not significantly different from each other except for the control. After 6 h of exposure to insecticid es, all of the insecticides virtua lly eliminated the thrips in the arena. Orius insidiosus bioassay Three of the four treatments tested killed more than 70% of the O. insidiosus used in the bioassays after 24 h. SpinTor 2SC was the inse cticide that killed th e lowest proportion of predators (Table 62). The most lethal insecticide wa s Actara 25WG, which killed almost all the insects in the first 20 h. Thr oughout the duration of the experi ment, Actara 25WG and Assail 70WP, had consistently the lowest survival rate compared with all the other treatments. Neither treatment (Actara 25WG and Assail 70WP) were si gnificantly different from each other. Fast acting treatments, Actara 25WG and Assail 70WP started showing significant differences with the control one hour after exposure (F = 4.87; df = 4, 25; P = 0.0016). Ten hours later only one treatment, SpinTor 2SC, was not signif icantly different from the control (F = 8.09; df = 4, 25; P < 0.001). This situation was true for the initial 12 h after exposure (F = 11.87; df = 4, 25; P < 0.001). Fifteen hours after exposure SpinTor 2S C showed significant differences with the control, but was the chemical treatment with th e highest proportion of survivors (0.73 0.04) (Table 62). In this study, I tested nine commercial ins ecticides and three experimental insecticides under various locations, methods of application, and cultivar s. Among the treatments used,

PAGE 129

115 Assail (acetamiprid) in its two formulations, A ssail 70WP (used in IFL04, IGA04, and IFL05), and Assail 30SG used in IFL05 (as experiment al TD 2480) and in IFL06, were the most consistent of the insecticides in co ntrolling the population growth rate (r) of flower thrips in blueberries. Assail 70WP is a redu ced-risk insecticide labeled for thrips control in cotton, Cole crops and fruiting vegetables but it is not yet labeled for use in bl ueberries, but it is a potential candidate to be labeled to be used in blueberries (OE Liburd personal communication). Acetamiprid had been mentioned by Morishita (2 001) as effective in thrips control under laboratory conditions. In the laboratory bioassay s, Assail 70WP demonstrated to be fast acting insecticide reducing the number of thrips alive to one-forth of the initial population on the first hour of exposure. After 6 h of exposure the numbe r of surviving thrips was virtually zero (Table 6-1). Assail 70WP app ears to be toxic to O. insidiosus during the first ten hours of exposure. After 10 hours after the release th e population declines rapidly fr om 63 % to 20 % in the next five hours. After 15 h, Assail 70 WP seem s to have no effect on the surviving O. insidiosus. SpinTor 2SC is another reduced-risk insecticid e that has been used on thrips control in several crops including be rries, cotton, grapes, and tubers am ong several others. In the laboratory experiments, SpinTor was the first insecticide th at killed 100% of the thrips exposed to treated green beans. At the same time SpinTor wa s the most compatible treatment with O. insidiosus. In the laboratory, 70% of O. insidiosus survived after 24 h of exposure. Despite that the treatment had significantly less survivors than the control, SpinTor 2SC was the insecticide that had the least effect among the insecticides screened having significantly mo re survivors than any of the other treatments (Table 62). The number of survivors was significantly lower than the control after 15 h of exposure to treate d green beans. This lag time has been observed with other

PAGE 130

116 species such as Geocoris punctipes Say (Myers et al. 2006), and in blueberry maggot, Rhagoletis mendax Curran (Liburd et al. 2003). In the field SpinTor 2SC did not have a signific ant effect on thrips populations. In all cases both populations exposed to SpinTor, floating an d established, increased and were among the populations with the highest r values, with exception of the es tablished population, collected from flowers, in 2004 at IFL04. However, it was not significantly different from the control or any of treatments used that year with excepti on of Assail 70WP, which was significantly lower (Figure 61). The reduced effectiveness in the fiel d might be related to the low residual activity of this insecticide. In field experiments in egg plants spinosad lost all activity after six days (McLeod et al. 2002). If the residua l activity of SpinTor 2SC is si milar in blueberries this will indicate that there is not e nough residual time to protect the blueberry bushes for the time between applications (two weeks). Malathion, a conventional insecticide, was not significantly different from the control in any of the observations conducted. In the laboratory bioassays, ma lathion was one of the fastest acting products. It killed more than two thirds of the population in the fi rst hour. Four hours after the treatment, malathion killed almost all the thrips exposed to the treated green beans (Table 61). In the O. insidiosus bioassay, malathion killed significantly less predators than Actara at all times, but significantly more than SpinTor. On average malathion killed close to 70% of the O. insidiosus individuals after 24 h (Table 62). The use of growth rate (r) in the field, as the ratio of the population after the treatments divided by the population before the treatments, reduces the risk of misinterpretation of the data in cases where insect populations are not uniform or randomly dist ributed. In the case of flower thrips in blueberries, the distri bution is highly aggregated (chapter 3), thus analysis of the final

PAGE 131

117 number of thrips captured after the applications might be affected by the presence of a hot-spot or aggregation sites in places where samples are taken. The use of r will reduce this problem by taking into consideration the initi al populations inside the treatm ents before the applications. However, it assumes that external conditions, such as temperature immigra tion, rain, affect all the plots with in a block similarly. Apparently th e population growth of thrips in blueberry fields is associated more with immigration than w ith the populations found in the fields. Every day more and more thrips arrive to the fields and th e insecticides are active for short periods of time, for this reason in most of the cases the population rates after / be fore the application were more than 1. From my results, any of the reduced-risk insecticides test ed in these experiments were as or more effective than malathion to control flower thrips in blueberries. Assail 70WP and Assail 30 SG were the most effective and consistent of treatments reduci ng the growth rate (r) of flower thrips in the field. These results were also co rroborated by the results in the laboratory. SpinTor 2SC was as effective as any of the other treatme nts and at the same time the most compatible with O. insidiosus which is one of the main natural enem ies, so it will be a good alternative in places where natural enemies are a high priority.

PAGE 132

118 Tables and Figures Table 6-1. Proportion of Frankliniella bispinosa surviving at various times after release into the bioassay arenas in 2004. Hours after th e release (HAR) of Frankliniella bispinosa in 2004 Treatments 1 4 6 12 Control 1.00 0.00 (a) 1.00 0.00 (a) 0.92 0.02 (a) 0.92 0.02 (a) Actara 0.15 0.04 (e) 0.02 0.02 (d) 0.02 0.02 (bc) 0.02 0.02 (b) Assail 0.25 0.02 (de) 0.06 0.03 (cd) 0.02 0.02 (bc) 0.02 0.02 (b) Calypso 0.39 0.04 (bcd) 0.11 0.05 (bc) 0.06 0.02 (bc) 0.02 0.02 (b) Knack 0.38 0.05 (bcd) 0.04 0.02 (cd) 0.04 0.03 (bc) 0.02 0.02 (b) GF 968 0.41 0.04 (bc) 0.12 0.03 (bc) 0.03 0.03 (bc) 0.03 0.03 (b) Malathion 0.29 0.07 (cde) 0.04 0.02 (cd) 0.02 0.02 (bc) 0.02 0.02 (b) Novaluron 0.49 0.07 (b) 0.16 0.04 (b) 0.07 0.04 (b) 0.03 0.03 (b) SpinTor 0.39 0.08 (bcd) 0.00 0.00 (d) 0.00 0.00 (c) 0.00 0.00 (b) Means followed by the same letter within each column are not significantly different from each other when compared using LSD test ( = 0.05).1 HAR (F = 13.86; df = 8, 36; P < 0.001); 4 HAR (F = 47.08; df = 8, 36; P < 0.001); 6 HAR (F = 84.56; df = 8, 36; P < 0.001); 12 HAR (F = 115.00; df = 8, 36; P < 0.001).

PAGE 133

119 Table 62. Proportion of Orius insidiosus surviving at various times after release into bioassay arenas. Hours after the release (HAR) of Orius insidiosus Treatments 1 10 12 15 20 24 Control 1.00 0.00 (a) 1.00 0.00 (a) 0.96 0.03 (a) 0.96 0.03 (a) 0.96 0.03 (a) 0.93 0.06 (a) Actara 0.56 0.06 (b) 0.56 .06 (c) 0.36 0.06 (c) 0.33 0.04 (c) 0.06 0.06 (d) 0.03 0.03 (d) Assail 0.63 0.03 (b) 0.63 0.03 (bc) 0.36 0.13 (c) 0.20 0.07 (c) 0.20 0.07 (d) 0.20 0.07 (cd) Malathion 0.87 0.04 (a) 0.80 0.00 (b) 0.73 0.04 (b) 0.66 0.04 (b) 0.50 0.07 (c) 0.30 0.04 (c) SpinTor 0.90 0.04 (a) 0.90 0.04 (a) 0.90 0.04 (a) 0.73 0.04 (b) 0.70 0.04 (b) 0.70 0.04 (b) Means with the same letter with in each column are not significantly different fr om each other when compared using LSD test ( = 0.05). for 1 HAR (F = 4.87; df = 4, 25; P = 0.0016); 10 HAR (F = 8.09; df = 4, 25; P < 0.001); 12 HAR (F = 11.87; df = 4, 25; P < 0.001); 15 HAR (F = 18.24; df = 4, 20; P < 0.001); 20 HAR (F = 17.42; df = 4, 20; P < 0.001); 24 HAR (F = 26.31; df = 4, 20; P< 0.001);

PAGE 134

120 Figure 61: Average growth rate ( r ) between the week before treatment application and the week after the application of the treatments. Thrips populati ons correspond to the thrips recovered from the flowers collected in IFL04. Significant differe nces with control are marked by [*] when compared using LSD test ( = 0.05). Figure 62: Average growth rate ( r ) between the week before treatment application and the week after the application of the treatments. Thrips populati ons correspond to the thrips captured in white sticky traps collected in IFL04. Significant di fferences with control are marked by [*] when compared using LSD test ( = 0.05). 0 0.5 1 1.5 2 ControlMalathion Calypso 480C KnackGF968NovaluronSpinTor 2SC Assail 70WP TreatmentGrowth rate ( r) after / before treatment applicationa a ab a a bab c 0 0.5 1 1.5 2 2.5 3 3.5 4 ControlSpinTor 2SC GF-968CalypsoEsteemNovaluronMalathion 5EC Assail 70WP TreatmentGrowth rate ( r) after / before treatment applicationab a ababab b bc

PAGE 135

121 Figure 63: Average growth rate (r) between the week before trea tment application and the week after the application of the treatments. Thrips populations correspond to the thrips captured in white sticky tr aps collected in IGA04. Figure 64: Average growth rate (r) between the week before trea tment application and the week after the application of the treatments. Thrips populations correspond to the thrips captured in white sticky traps collected in IFL05. Significant differences are represented by different letters wh en compared using LSD test (= 0.05). 0 0.5 1 1.5 2 ControlCalypso 480EC Assail 70WP KnackSpinTor 2SC PedestalActara Malathion TreatmentGrowth rate ( r) after / before treatment application 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 ControlMalathion 5EC SpinTor 2SCDiamond .83EC TD 2480Actara 25 WG Assail 70W TreatmentGrowth rate ( r) after / before treatment applicationa ab ab ab abb b

PAGE 136

122 Figure 65: Average growth rate ( r ) between the week before treatment application and the week after the application of the treatments. Thrips populations corr espond to the thrips captured in white sticky traps collected in IFL06. Significant di fferences are represented by different letters when compared using LSD test ( = 0.05). 1/sqrt(x) Figure 66: Average growth rate ( r ) between the week before treatment application and the week after the application of the treatments. Thrips populations correspond to the thrips recovered from inside the flowers in IFL06. Significant differences are repres ented by different letters when compared using LSD test ( = 0.05). ( F = 0.48, df = 6, 48; P = 0.7275) 0 0.5 1 1.5 2 2.5 3 ControlDiamond 0.83EC SpinTor 2SCAssail 30 SGMalathion 5EActara 25WG Coragen 20SC TreatmentGrowth rate ( r) after / before treatment applicationa ab ab ab abab b 0 10 ControlSpinTor 2SCActara Diamond 0.83EC Coragen 20SC Malathion 5EC Assail 30SG TreatmentGrowth rate ( r) after / before treatment application

PAGE 137

123 CHAPTER 7 EFFECTIVENESS OF PREVENTIVE AN D INUNDATIVE BIOLOGICAL CONTROL TACTICS TO MANAGE FLOWER THRI PS POPULATIONS IN EARLY-SEASON BLUEBERRIES Thrips have many natural enemies that he lp to manage their populations keeping them under the economic injury level. There ar e 23 families distribu ted in eight insect orders and nine families of mites that have been reported as natural enemies of thrips (Sabelis and VanRijn 1997). However, the vast majority of these natural enemies is polyphagous and feed on a diversity of small insects including th rips. Most of the research studying these pred atory arthropods has been conducted in greenhouses where the predator-prey interaction is localized and the environmental complexity is limited (Sabelis and VanRijn 1997). Hoy and Glenister (1991) showed that i nundative releases of Amblyseius sp. are not effective en ough to reduce thrips populations below the damage thresholds on field cabbage, despite the time of release and number of mites released. Thrips are known to be prolific, po lyphagous, have short generation time, a tendency to aggregate and they tend to expl oit localized ephemeral optimal conditions, which characterize them as r-selected insects (Mound and Teulon 1995). Due to these characteristics, principally a ggregation, flower thrips are able to dramatically increase their population in short periods of time (Chapt ers 3 and 4). One of the main objectives of biological control is to achiev e long-term pest regulation, ke eping the populations of the target organisms below injury levels (VanDr iesche and Bellows 2001). My objective is to test the efficacy of commercially available natural enemies of flower thrips in earlyseason blueberry fields. Inunda tive-preventive and inundati ve-curative methods were

PAGE 138

124 evaluated to determine if O. insidiosus and A. cucumeris are able to reduce the number of flower thrips in the field during the blue berry flowering season. Materials and Methods A farm located in north-central Florida (N 28o 54 W 82o 14) was selected to conduct the biological cont rol trials. This farm uses mini mum applications of pesticides and is planted with southern highbush blueberries. Orius insidiosus Say (Hemiptera: Anthocoridae) as Thripor-I and Amblyseius cucumeris (Oudemans) (Acari: Phytoseiidae) as Thripex-plus were select ed as they are commercially recommended for flower thrips control (Koppert biological systems Romulus, M I). The farm was divided in 16 plots organized in four blocks of four 283 m2. There were buffer zones of 17m between blocks and 5 m between plots with in a block. The experiment was design as a completely randomized block with four replicates and four treatments. Treatments used in the experiments were 1) O. insidiosus, 2) A. cucumeris, 3) combination of both O. insidiosus and A. cucumeris in half doses, and 4) untreated control. The first year of the expe riment (2005), I released the natural enemies as preventive tactic. Natural enemies were re leased one week after flowering started, before thrips population began to build-up and the hot-spots were de fined (Chapter 3). I used the doses recommended for preventive control: O. insidiosus (Thripor-I) was released at 0.5 insects per m2, A. cucumeris (Thripex-Plus) at 0.5 sachets of 1000 mites per m2. For the combination treatment, I released half doses of each biocontrol agent. The analysis conducted compared the population of flower thrips in each one of the sampling dates using one-way ANOVA tabl es and LSD mean separation test ( = 0,05) for each one of the sampling dates (SAS Institute Inc. 2002).

PAGE 139

125 During the second season, 2006, I tested the use of biocontrol agents as curative method for the control of flower thrips. I used the same set-up as the experiments conducted in 2005. A new randomization of trea tments was conducted. The manufacturer recommended doses for curative control were: O. insidiosus (Thripor-I) 10 insects per m2, and A. cucumeris (Thripex-Plus) 1.3 sachets of 1000 mites per m2. The treatments were released on February 15, 2006 when the number of thrips on s ticky traps was above 100 thrips per trap (Figure 72). These treatments were compared with a control where no natural enemies were released. To disc ourage the movement of natural enemies between plots SpinTor 2SC (spinosad) 105 g a. i. / ha was sprayed in the buffer zones whenever natural enemies were released. In the center of each treatment a white sticky trap was deployed weekly. A sample of five flower-clusters was collected every week from each repetition and processed using the shake and rinse method (chapter 3) to determine thrips population inside the flowers. Differences between the treatments were analyzed using the LSD mean separation test ( = 0.05) for each one of the sampling date s(SAS Institute Inc. 2002). At the same time an analysis of the population growth rate (r) comparing the increment of the population one week after the rele ase, and two weeks after the release of natural enemies, with the population of thrips be fore the release (Chapter 6). Results and Discussion The trials conducted in 2005, indicated that releases of O. insidiosus or A. cucumeris, as well as the combination of both trea tments as a preventive tactic does not reduce thrips populations in blueberries duri ng the flowering period. One week after the release of natural enemies, February 511, 2005, I found th at thrips population in the control were on average significantly lower than in treatments of O. insidiosus, and A.

PAGE 140

126cucumeris alone. However, no significant differences were detected between the control and the combination treatment duri ng the first week after release (F = 7.13; df = 3, 9; P = 0.016) (Figure 71). The same situation was obs erved during the second week after the release February 1218, 2005 (F = 2.83; df = 3, 9; P = 0.0988). During the last week of sampling February 1925, 2005 the control ha d significantly less th rips than any of treatments in sticky traps (F = 7.95; df = 3, 9; P = 0.0067) (Figure 71). Due to the low population of thrips in 2005, I could not coll ect enough thrips from in side the flowers to make a robust analysis. In 2006 when curative releases of O. insidiosus, A. cucumeris, and the combination treatment of both natural enemies were conducted, I found no significant differences between the treatments of natural enemies and the control in the number of thrips caught on the sticky traps (Figure 72) or inside the flowers (Figure 73). I calculated the growth ratio (r) of thrips population for one and two weeks after the re lease of natural enemies in relation with the week before th e release (Chapter 6). The results of this analysis showed no significant differences among the treatments in sticky traps (Figure 74) or in flowers (Figure 75). Despite that there was no a significant change in the r value inside flowers ( = 0.05), we detected a significant reduction of r in all treatments when compared with the control one week afte r the release of natural enemies when = 0.1, but no differences among the treatments using natural enemies (Figure 75). My results were consistent with obser vations made by Mound and Teulon (1995) and by Parella and Lewis (1997). These au thors concluded that the biological characteristics of thrips overcome the attributes of natural enemies in such a way that the participation of natural enemies in the regula tion of field populations of thrips is minor.

PAGE 141

127 Other authors argued that the use of natu ral enemies is enough to control thrips populations (Van de Veire and De gheele 1995, Shipp and Wang 2003). Most of the studies that conclude that Orius spp. and A. cucumeris are efficient in controlling flower thrips were conducted unde r greenhouse conditions (Van de Veire and Degheele 1995, Jacobson 1997, Shipp and Wang 2003). One of the few successes in control of flower thrips under field conditi ons was reported by Funderburk et al. (2000). The authors showed that unt reated fields and field tr eated with spinosad had a significantly higher population of Orius spp. and lower population of flower thrips than fields treated with acephate and fenopropa thrin, which excluded the predator. The reduction in thrips population began between 55 and 60 days after transplanting, approximately 10 days after the first sampli ng. This period of time allowed the natural enemies to build their populati on, and have an effect on thrips population. The situation in blueberries is different i. e. thrips are only present for an average for 20 to 25 days, which correspond to the flowering period in bl ueberries (Chapter 3 and 4). This short period of time might not be long enough for th e natural enemies to establish and reach a significant level of control. When the natu ral enemies were released as preventive method, I found that the control had significantly less thrips than the other treatments at the end of the season. I believe the reason fo r this difference relies in the polyphagy of natural enemies used. The natural enemies were released before thrips were present in the field. Hulsof and Vnninen (2001) mentioned that the presence of Orius spp. depends on the crop characteristics principally the viabil ity of pollen and the pr esence of alternative preys. Since there was no pollen or thrips pr esent in the blueberries at the time of the release of Orius spp. this natural enemies may ha ve preyed on natural enemies and

PAGE 142

128 competitors of flower thrips and emigrated from the field before thrips population was present. This situation might have created a niche free of natural enemies where thrips can reproduce more efficiently. At the same time, the low temperatures during the days following the release ( -1o C on February 5, 2005 and 3 o C on February 6, 2005) and the lack of pollen in the field might have had an effect on the survival of mites. These situations combined might have influenced an increase of the number of thrips in those treatments were the predators were released These opinions are rather speculative. The exact reason for the increase in thrips popul ations in the non-cont rol treatments is not known. More research about the interaction of flower thrips and their natural enemies in early season-blueberries n eeds to be conducted.

PAGE 143

129Tables and Figures Figure 71: Average number of thrips captu red per week after th e release of natural enemies, as a preventive measure, in white sticky traps located inside the blueberry bush in 2005. Treatments fo llowed by the same letter are not significantly different when compared using LSD ( = 0.05). The arrow represents the date of re lease. February 4, 2005 (F = 1.52; df = 3, 9; P = 0.62), February 11, 2005 (F = 7.13; df = 3, 9; P = 0.016), February 18, 2005 (F = 2.83; df = 3, 9; P = 0.0988), February 25, 2005 (F = 7.95; df = 3, 9; P = 0.0067) 0 50 100 150 200 250 300 28-Jan4-Feb11-Feb18-Feb25-Feb DateNo. of thrips per trap per wee k Control A. cucumeris O. insidiosus O.i / A.c a a a b b ab a a b a ab a

PAGE 144

130 Figure 72: Average number of thrips captu red per week after th e release of natural enemies, as curative measure, on white sticky traps located inside the blueberry bush in 2006. The arrow indicates the date of the release of natural enemies. Treatments followed by the same letter are not significantly different when compared using LSD ( = 0.05). February 15, 2006 (F = 0.95; df = 3, 9; P = 0.4549), February 22, 2006 (F = 0.54; df = 3, 9; P = 0.6639), March 1, 2006 (F = 1.79; df = 3, 9; P = 0.2185). 0 100 200 300 400 500 600 700 800 900 8-Feb15-Feb22-Feb1-Mar DateNo. of thrips per trap per week Control A. cucumeris O. insidiosus O.i / A.c

PAGE 145

131 Figure 73: Average number of thrips captu red per week after th e release of natural enemies, as curative measure, inside five flower-clusters collected from blueberry bushes. The arrow indicates the date of the release of natural enemies. Treatments followed by the same letter are not significantly different when compared using LSD ( = 0.05). February 15, 2006 (F = 0.05; df = 3, 9; P = 0.9821), February 22, 2006 (F = 1.42; df = 3, 9; P = 0.0.2992), march 1, 2006 (F = 0.06; df = 3, 9; P = 0.9781) 0 5 10 15 20 25 30 35 40 8-Feb15-Feb22-Feb1-Mar DateNo. of thrips per trap per week Control A. cucumeris O. insidiosus O.i / A.c

PAGE 146

132 Figure 74: Growth rate (r) of thrips populations captured in white sticky traps during the 2006 flowering season one and two weeks af ter the release of natural enemies as a curative alternative. February 16 22, 2006 (F = 1.01; df = 3, 9; P = 0.4339), February 23March 1, 2006 (F = 1.53; df = 3, 9; P = 0.2737) 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 February 16-22February 23March 1 DateGrow rate ( r ) after release Control A. cucumeris O. insidiosus O.i / A.c

PAGE 147

133 Figure 75: Growth rate (r) of thrips populations collect ed inside blueberry flowers during the 2006 flowering season, one and two weeks after the release of natural enemies as a curative alternative. February 16 22, 2006 (F = 2.05; df = 3, 9; P = 0.1779), February 23March 1, 2006 (F = 0.38; df = 3, 9; P = 0.7728) 0 5 10 15 20 25 30 February 16-22February 23 March 1 DateGrow rate ( r ) after release Control A. cucumeris O. insidiosus O.i / A.c

PAGE 148

134 CHAPTER 8 GENERAL CONCLUSIONS AND EXPERIENCES WORKING WITH FLOWER THRIPS IN EARLY-SEASON BLUEBERRY FIELDS. The development of an Integrated Pest Management (IPM) program is a complicated process than needs to consider all of the relevant information regarding the pests biology, ecology behavior and the crop response to its presence. In this case, the relationship between blueberrie s and flower thrips was st udied. Finn (2003) conducted a survey among blueberry grower s in Florida and southern Georgia, where early-season blueberries are produced, to prioritize the arthro pod pests that concerned them. One of the top three arthropods were flower th rips, along with the blueberry bud mite, Acalitus vaccinii (Keifer) and the blueberry gall midge, Dasineura oxycoccana (Johnson) (Diptera: Cecidomyiidae). Finn (2003) re search involving monitoring and sampling formed the basis for this work. The current work refined some of the monitoring and sampling techniques evaluated by Finn (2003) To gather as much information as possible I decided to divide the work into two. The first group of experiments was designed to understand the biology, ecology and be havior of thrips with respect to earlyseason blueberries. The second set of experime nts was designed to explore methods of control for flower thrips. Finally, it is necessary to organize all the information in such a way that farmers can apply the knowledge acq uired during these three years to improve their management techniques and increase the efficiency managing early-season blueberries.

PAGE 149

135 Flower Thrips Monitoring and Sampling The use of white sticky traps has proven to be the more efficient method to monitor flower thrips. Finn (2003) demonstrated th at there were no differences between white traps and other colors such as blue and yellow to monitor thrips in blueberry fields. Therefore, I selected white traps because the background contrast with the insects captured and it appears to attract predominantly thrips. Unlike white traps, yellow traps captured a high number of dipterans and othe r non-target organisms. The horizontal distribution of flower thrips with respect to the blueberry bush is also of importance to determine the best place to hang the traps. After comparing traps above the canopy, inside the canopy and above the soil, results s howed that the highest number of thrips was captured inside the canopy. Traps inside th e canopy were located in the middle of the bush between 1.50 and 1.60 m from the ground, a nd this is the position that was selected for the remaining of the experiments. I deve loped a procedure to extract thrips from inside blueberry flowers. This procedure is referred as the shake and rinse method (described in chapter 3). This method proved to be as efficient as manual dissection of the flowers, but less time consuming. This method could be a good alternative for researchers. As for growers, white sticky traps are a valuable tool and an efficient way of monitoring thrips populations in the field. The reliability of these sticky traps was improved when we were able to correlate th e number of thrips inside the flowers and thrips captured on the trap. A regression betw een the number of thrips in five flowerclusters whit the number of thrips captured by a sticky trap located inside the canopy of the same bush allowed me to calculate the eco nomic injury level (EIL) for flower thrips in sticky traps based on the EIL calculated in thrips per flower. Based on the number of

PAGE 150

136 thrips captured in the traps, growers can make an informed decision of the timing to use the adequate controlling methods. Flower Thrips in Blueberries The observation of flower thri ps in blueberries coincide with the remarks published by Mound (1997). Mound described thrips as r selected insects. This classification was given due to their success coloni zing habitats that are suitable but brief, their successful use of a wide range of hosts (proper and provisional), their short generation time, vagility, parthenogenesis, and polyphagy. Flow ering blueberry fields appear to be a perfect target for flower thrips. Blueberry blooming period lasts for approximately 25 days, from the beginning of flower openi ng to petal drop. During this period, the predominant color in the field is white to whic h several studies have pointed flower thrips are attracted to (Kir k 1984, Teulon and Penman 1992). This research demonstrated that flower thrips use blueberry plants as proper hosts. I was successful at obtaining larvae that emerged from flowers collected in the field. Flower thrips prefer to lay their eggs in the petals of the flowers, but I also obta ined larvae from ovaries and stiles. While describing and comparing the injuries inflicte d by flower thrips on the blueberry fruits, two types of injuries were found: feeding and emergence injuries. Thes e injuries are very similar to injuries observed by other researcher s in other fruits such as avocados, citrus, plums (Childers 1997, Kirk 1997a). Captures of flower thrips in blueberrie s, on white sticky traps and flowers, begin with the opening of the flowers when they change from pink to white color, between stages five and six as defined by Spiers (1978). The hot spots or sites of high aggregation begin to form between seven to ni ne days after the beginning of the captures and the peak of the population is reached seven to 10 days later. After th is point, the

PAGE 151

137 petals of the flowers begin to fall, the fruits begin to form, and thrips population starts to decline. The results obtained from various fields in Florida an d southern Georgia indicated that flower thrips ha ve highly aggregated populations in the field. Their level of aggregation is correlated w ith the number of individua ls or the maximum population observed, when populations where high I observe d the highest levels of aggregation.Very low populations show low levels of aggreg ation with a tendency to randomness. During this work calculations of an economic injury level (EIL) for two of the most popular cultivars of rabbiteye blueberries Climax and Tifblue were conducted. However, due to the dynamic nature of EI L and its dependency on market values only one variable are dependent on plant health and thrips density and damage. These variables are the Injury per insect I and da mage per unit injured D. Two of the main rabbiteye cultivars were selected to conduc t the EIL determinations, Climax and Tifblue. For Climax we determin ed that the value of I was 1.71x10-10 proportion damage / [thrips/ha] and the value of D was 16,812 (kg/ha)/proportion damaged. In the case of Tifblue I = 1.92 proportion damage /[thrips/ha] and D had the same value as in Climax. Based on the variables observed and the variables de scribed in Figure 5-1 each producer will be able to calculate the EIL for the farm. Nine thrips species in total were found in early-season blueberr ies. I developed an identification key for these six species to be used in future research. Frankliniella bispinosa (Morgan) was the predominant species in Florida captured inside the flowers and in the sticky traps. Frankliniella tritici (Fitch), eastern flower thrips, was the most abundant species in southern Georgia. Frankliniella occidentalis (Pergande), western flower thrips, was the second most abundant species in Florida and in Georgia. Other

PAGE 152

138 species encountered include F. fusca (Hinds), Thrips hawaiiensis (Morgan), and T. pini Uzel. Flower Thrips Control The EIL calculated indicated that approxima tely 2 thrips per flower will be enough to reach the economic damage level in rabbi teye blueberries. Ba sed on the correlation between the number of thrips in flowers and the number of thrips captured on sticky traps the EIL was between 45 (in Tifblue) and 50 (in Climax) if Malathion is used for control, and between 64 (in Tifblue) and 73 (in Climax) if SpinTor is used to control thrips. These numbers might seem very low; however, the timing for this EIL is also important. The time for which the EIL was calcu lated was at the opening of the flowers. During flower opening, or week one, in most of the trials conducted during this work the maximum number of thrips captured on sticky traps during the first week was 37.7 5.5 (Figure 4-2). This number rather than the rule is an exception of an extremely high thrips population in this site. Most of the observati ons of week one during the three years that the trials were conducted were very low and rarely more than 20 thrips per trap were observed in the first week. None of the farm s surveyed during these three years reached the EIL. Early monitoring is very important to use this EIL. Monitoring for flower thrips should begin when more than 50% of the flow ers are in stage 5 and some are beginning stage six according to the descriptions made by Spiers (1978). Grow ers should not allow thrips populations to reach these values by using the adequate management techniques before the thrips reach this threshold. I explored the use of chemical and biological control to manage the populations of fl ower thrips. I screened 12 commercial and experimental insecticides with potentia l to control flower thrips. Among these insecticides, conventional ins ecticides, reduced-risk insec ticides, and insect growth

PAGE 153

139 regulators with the potential to be labeled fo r blueberry use were included. We conducted field and laboratory experiment to determin e the efficacy of those insecticides that demonstrated to be successful in the field and compared them with non-treated controls and conventional insecticides. The active ingred ient that demonstrated to be the most consistent and effective managing thrips in the field and in the laboratory was acetamiprid. Two formulations of acetamiprid were tested Assail 30SG and Assail 70W. Both formulations consistently demonstrated to control the growth of thrips populations better than the other chemistries evaluated in the field (measured as growth rate r). These low r values indicate that th rips population did not increas ed the same way as in the other treatments and in some cases th e population decreased af ter the application (r < 1). The reasoning for the use of population grow th rate is based on the high levels of aggregation of flower thrips, which in th e field means high variances. The use of r as variable reduces the variance due to the pres ence of hot-spots in some treatments and not in others by taking into c onsideration the populat ion before the applications instead of assuming that all the treatments started with the same population. SpinTor 2SC (spinosad) demonstrated in the laboratory to be efficient in controlling flower thrips and at the same time was the most compatible chemistry towards Orius insidiosus. In the biological control experiments we explored the use of commercial, O. insidiosus and Amblyseius cucumeris to control flower thrips. We evaluated preventive and curative releases of these natural enemies. Results showed that these natural enemies were not effective managing thrips populations. The s hort period that thrips are present in blueberry fields, the high mobility of O. insidiosus and slow action of A. cucumeris, combined with their polyphagy could explain in part the reason for the failure in to

PAGE 154

140 manage flower thrips populations in co mmercial blueberry plantings. The success of these predators to manage flower thrips had been widely demonstrated mostly under greenhouse conditions (Van de Veire a nd Degheele 1995, Jacobson, 1997 #219, Shipp and Wang 2003). Some authors had found that biol ogical control is not the best approach when dealing with flower thrips in op en field situations (Mound and Teulon 1995, Parrella and Lewis 1997). Funderburk (2000) found O. insidiosus to be successful controlling thrips populations in peppers.

PAGE 155

141 LIST OF REFERENCES Altieri, M. A. 1988. The dynamics of insect populations in crop systems subject to weed interference, pp. 433-452. In E. A. Heinrichs [ed.], Insect interactions. Wiley & Sons, New York, NY. Atanassov, A., P. W. Shearer, G. Hamilton, and D. Polk. 2002. Development and implementation of a reduced risk peach arthropod management program in New Jersey. J Econ Entomol 95: 803-812. Bates, B. A., and M. J. Weiss. 1991. The spatial distribution of Limothrips denticornis Haliday (Thrysanoptera: Thripidae) e ggs on spring barley. Can Entomol 123: 205-210. Beckwith, C. S. 1941. Control of cranberry fruitwor m on blueberries. J Econ Entomol 34: 169-171. Cane, J. H., and J. A. Payne. 1988. Foraging ecology of the bee Habropoda laboriosa (Hymenoptera: Anthophoridae), an o ligolege of blueberries (Ericaceae: Vaccinium) in the southeastern United States. Ann Entomol Soc Am 81: 419-427. Castagnoli, M., G. Angeli, M. Ligu ori, D. Forti, and S. Simoni. 2002. Side effects of botanical insecticides on predatory mite Amblyseius andersoni (Chant). J Pest Sci 75: 122-127. CDMS. 2006. ChemSearch. CDMS inc. Chau, A., K. M. Heinz, and F. T. Davies, Jr. 2005. Influences of fertilization on population abundance, distri bution, and control of Frankliniella occidentalis on chrysanthemum. Entomol Exp Appl 117: 27-39. Childers, C. C. 1997. Feeding and oviposition injuries to plants, pp. 505-538. In T. Lewis [ed.], Thrips as crop pests. CAB International, Wallingford, UK. Childers, C. C., and D. S. Anchor. 1991. Feeding and oviposition injury to flowers and developing floral buds of 'navel' oranges by Frankliniella bispinosa (Thysanoptera: Thripidae) in Florid a. Ann Entomol Soc Am 84: 272-282. Childers, C. C., S. Nakahara, and R. J. Beshear. 1994. Relative abundance of Frankliniella bispinosa and other species of Thys anoptera emerging from soil beneath 'Navel' orangetrees in Florida dur ing spring flowering. J Entomol Sci 29: 318-329.

PAGE 156

142 Childers, C. C., S. Nakahara, and R. J. Beshear. 1998. Thysanoptera collected during bloom on white and colored disposable st icky cards in Florida Citrus groves. J Entomol Sci 33: 49-71. Childers, C. C., R. J. Beshear, J. R. Brushwein, and H. A. Denmark. 1990. Thrips (Thysanoptera) species, their occurr ence and seasonal abundance on developing buds and flowers of Florida citr us. J Entomol Sci 25: 6001-614. 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. Cline, W. O., and A. Schilder. 2006. Identification and control of blueberry diseases, pp. 115-138. In N. F. Childers and P. M. Lyrene [eds.], Blueberries for growers gardeners, promoters. Dr. Norman F. Childers Publications, Gainesville FL. Collins, J. A., H. Y. Forsythe, and D. Yarborough. 1995. Blueberry thrips (Frankliniella vaccinii Morgan & Catinathrips kainos O'Neill). University of Maine Fact Sheet No. 202 UMCE No 2373. Diraviam, J., and S. Uthamasamy. 1992. A new sampling technique involving yellow sticky traps for monitoring of thrips infes ting different crops. J Entomol Res (New Delhi) 16: 78-81. Dow AgroSciences. 2006a. Entrust Naruralyte insect control (original label + suplemental labels). In ChemSearch CDMS [ed.]. Crop Data Management Systems, Inc. Dow AgroSciences. 2006b. SpinTor Naruralyte insect control (original label + suplemental labels). In ChemSearch CDMS [ed.]. Crop Data Management Systems, Inc. Drake, V. A., and R. A. Farrow. 1988. The influence of atmosferic structure and motions on insect migration. Annu Rev Entomol 33: 183-210. Edwards, G. B. Unpublished data Thripidae (Thysanoptera: Terebrantia) of Canada and United States, pp. 18. Depart ment of Plant Industry, FL. Ellington, C. P. 1980. Wing mechanics and take-o ff preparation of thrips (Thysanoptera). J Exp Biol 85: 129-136. Elzen, G. W. 2001. Lethal and sublethal effects of insecticide residues on Orius insidiosus (hemiptera: Anthocoridae) and Geocoris punctipes (Hemiptera: Lygaeidae). J Econ Entomol 94: 55-59. Elzen, G. W., P. J. Elze n, and E. G. King. 1998. Laboratory toxicity of insecticide residues to Orius insidiosus, Geocoris punctipes, Hippodamia convergens, and Chrysoperla carnea. Southwest Entomol 23: 335-342.

PAGE 157

143 Environmental Protection Agency (EPA). 1996. Food Quality Protection Act of 1996. In USDA-EPA [ed.]. [http://www.epa.gov/ oppfead1/fqpa/]. Last accessed; Nov2006 Environmental Protection Agency (EPA). 1997. Notice to manufacturers, formulators, producers and registrants of pes ticide products. (PR) Notice 97-3. [http://www.epa.gov/opppmsd1/ PR_Notices/pr97-3.html] Last accessed; Nov2006 Environmental Protection Agency (EPA). 2003. Integrated pest management (IPM) and food production. Finn, E. 2003. Developing integrated pest manage ment (IPM) techniques for managing key insect pests of blueberries in th e southern United States, pp. xi+99, Entomology and Nematology Department. University of Florida, Gainesville,FL. Flint, M. L., and P. Gouveira. 2001. IPM in practice, principles and methods of integrated pest management. Univer sity of California, Oakland, CA. Food and Agriculture Organization of the United Nations (FAO). 2004. FAOStatAgriculture. Funderburk, J., and J. Stavisky. 2004. Biology and economic importance of flower thrips, pp. 6, Series of the Entomo logy and Nematology Department, Florida Cooperative Extension Service, Institut e of Food and Agricultural Sciences. University of Florida, Gainesville, FL. Funderburk, J., J. Stavisky, and S. Olson. 2000. Predation of Frankliniella occidentalis (Thysanoptera: Thripidae) in field peppers by Orius insidiosus (Hemiptera: Anthocoridae). Environ Entomol 29: 376382. Gatehouse, A. G. 1997. Behavior and ecological geneti cs of wind-borne migration by insects. Annu Rev Entomol 42: 475-502. Green, R. H. 1966. Measurement of non-randomness in spatial distributions. Res Popul Ecol (Kyoto) 8: 1-7. Hagerup, E., and O. Hagerup. 1953. Thrips pollination of Erica tetralix. New Phytol 52: 1-7. Hamill, J. E., O. E. Liburd, and S. R. Alm. 2003. Comparison of biodegradable plastic wooden imidacloprid-treated spheres for control of Rhagoletis mendax (Diptera:Tephritidae) flie s. Fla Entomol 86: 206-210. Heming, B. S. 1993. Structure, function, ontogeny and evolution of feeding in thrips (Thysanoptera), pp. 3-41. In C. W. Schaefer and R. Leschen [eds.], Functional morphology of insect feeding. Tomas Say, Lanham, MD.

PAGE 158

144 Herron, G. A., and T. M. James. 2005. Monitoring insecticide resistance in Australian Frankliniella occidentalis Pergande (Thysanoptera: Th ripidae) detects fipronil and spinosad resistance. Aust J Entomol 44: 299-303. Hoddle, M. S., L. Robinson, and D. Morgan. 2002. Attraction of thrips (Thysanoptera: thripidae and Aleolothripidae) to colo red sticky cards in California avocado orchard. Crop Prot 21: 383 388. Hollingsworth, R. G., K. T. Sewake, and J. W. Armstrong. 2002. Scouting methods for detection of thrips (Thysanoptera: Thripidae) on dendrobium orchids in Hawaii Environ Entomol 31: 523-532. Hoy, C. W., and C. S. Glenister. 1991. Releasing Amblyseius spp. (Acarina: Phytoseiidae) to control Thrips tabaci (Thysanoptera: Thri pidae) on cabbage. Entomophaga 36: 561-573. Hulshof, J., and I. Vnninen. 2001. Western flower thrips feeding on pollen, its implications for control, pp. 173-178. In R. Marullo and L. Mound [eds.], Thrips, plants, tospoviruses: The millennial review. Proceedings of the 7th international symposium on thysanoptera, Reggio Calabria, Italy. Hunter, W. B., and D. E. Ullman. 1992. Anatomy and ultrastructure of the piercingsucking mouthparts and paraglossal sensilla of Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae) Int J Morphol Embryol 21: 17-35. IR 4 project. 2006. New Pest Control Technologies. Rutgers University. Jacobson, R. J. 1997. Integrated pest management (IPM) in glasshouses, pp. 639-666. In T. Lewis [ed.], Thrips as crop pests. CAB International, Wallingford, UK. Javorek, S. K., K. E. Mackenzie, and S. P. Vander Kloet. 2002. Comparative pollination effectiveness among bees (Hymenoptera: Apoidea) on lowbush blueberry (Erucaceae: Vaccinium angustifolium). Ann Entomol Soc Am 95: 345351. Jensen, S. E. 2000. Insecticide resistance in the western flower thrips, Frankliniella occidentalis. Integrated Pest Manage Rev 5: 131-146. Kirk, W. D. J. 1984. Ecologically selective coloured traps. Ecol Entomol 9: 35-41. Kirk, W. D. J. 1985. Effect of some floral scents on host feeding by thrips (Insceta: Thysanoptera). J Chem Ecol 11: 35-43. Kirk, W. D. J. 1987. How much pollen can thrips destroy? Ecol Entomol 12: 31-40. Kirk, W. D. J. 1988. Thrips and pollination biology pp. 129-135. In T. N. Ananthakrishnan and A. Raman [eds.], D ynamics of Insect-plant interaction. Recent advances and future trends Oxford, and IBH, New Delhi.

PAGE 159

145 Kirk, W. D. J. 1995. Feeding behavior and nutri tional requirements, pp. 21-29. In B. L. Parker, M. Skinner and T. Lewis [eds.] Thrips biology and management. Plenum Press, New York. Kirk, W. D. J. 1997a. Feeding, pp. 119-174. In T. Lewis [ed.], Th rips as crop pests. CAB International, Wallingford, UK. Kirk, W. D. J. 1997b. Distribution, abundance and population dynamics pp. 217-258. In T. Lewis [ed.], Thrips as Crop Pests. CAB International, Wallingford, UK. Kirk, W. D. J., and J. G. C. Hamilton. 2004. Evidence for a male-produced sex pheromone in the western flower thrips Frankliniella occidentalis. J Chem Ecol 30: 167-174. Korkac, R. F. 1986. Adaptability of blueberry species to various soil types: growth and initial fruting. J Am So c Hortic Sci 111: 816-821. Kring, J. B., and D. J. Schuster. 1992. Management of insects on pepper and tomato with UV-reflective mulches. Fla Entomol 75: 119-129. Lathman, L. J., and R. A. C. Jones. 1997. Ocurrence of tomato spotted wilt tospovirus in native flora, weeds, and horticult ural crops. Aust J Agric Res 48: 359-369. Lewis, T. 1997a. Flight and dispersal, pp. 175-197. In T. Lewis [ed.], Thrips as crop pests. CAB International, Wallingford, UK. Lewis, T. 1997b. Chemical control, pp. 567-594. In T. Lewis [ed.], Thrips as crop pests. CAB International, Wallingford, UK. Liburd, O. E., and E. Finn. 2003. Evaluation of conventional and reduced-risk insecticides for control of flower th rips. Berry/Vegetable Times 3: 3-4. Liburd, O. E., and H. A. Arvalo. 2005. Integrated strategies for controlling flower thrips in southern highbush blueberries. University of Florida. IFAS extension IPM-103: 4. Liburd, O. E., and H. A. Arvalo. 2006. Insects and mites in blueberries, pp. 99-110. In N. F. Childers and P. M. Lyrene [eds.] Blueberries for growers, gardeners, promoters. Dr. Norman F. Childers Hor ticultural Publications, Gainesville FL. Liburd, O. E., C. Mullin, and J. C. Turner. 2005. Monitoring blueberry maggot in blueberry varieties. Citrus & Vegetable Magazine 69: 18-19. Liburd, O. E., S. R. Alm, R. A. Casagrande, and S. Polavarapu. 1998. Effect of trap color bait, shape and orientation in a ttaction of blueberry maggot (Diptera: Tephritidae). J Econ Entomol 91: 243-249.

PAGE 160

146 Liburd, O. E., S. Polavarapu, S. R. Alm, and R. A. Casagrande. 2000. Effect of trap size, placement and age on captures of blueberry maggot flies (Diptera: Tephritidae). Hort Entomol 93: 1452-1458. Liburd, O. E., E. Finn, K. L. Pettit, and J. C. Wise. 2003. Response of blueberry maggot fly (Diptera: Tephritidae) to im idacloprid-treated spheres and selected classes of insecticides. Can Entomol 135: 427-438. Liburd, O. E., H. A. Arvalo, H. J. McAuslane, and G. Krewer. 2006. Thrips and gall midge control, pp. 135-138. In G. Krewer and J. Martin [eds.], Proceedings for the Georgia Blueberry Conference, Ge orgiaSouth Carolina Muscadine Conference, North American Brambl e Growers Association Conference Savannah, GA. Liburd, O. E., L. J. Gut, L. L. Stelinski, M. E. Whalon, M. R. McGuire, J. C. Wise, J. L. Willett, X. P. Hu, and R. J. Prokopy. 1999. Mortality of Rhagoletis species encountering pestic ide-treated spheres (Diptera: Tephritidae). J Econ Entomol 92: 1151-1156. Ludwig, S., and R. Oetting. 2001. Effect of spinosad on Orius insidiosus (Hemiptera: Anthocoridae) when used for Frankliniella occidentalis (Thysanopter: Thripidae) control on green houses. Fla Entomol 84: 311-313. Lyrene, P. 1997. Value of various taxa in breeding tetraploid blueberries in Florida. Euphytica 94: 15-22. Lyrene, P. M. 2005. Blueberry varieties for commerc ial production in Florida, Fall meeting of the Florida Blueberry Grow ers Association. Florida Blueberry Growers Association, Bartow, Florida. Lyrene, P. M., and J. A. Payne. 1992. Blueberry gall midge: a pest on rabbiteye blueberry in Florida. Proc Ann Meet ing Fla State Hortic Soc 105: 197-300. Lyrene, P. M., and J. A. Payne. 1996. Blueberry gall midge: a new pest of rabbiteye blueberries, pp. 111-124. In R. E. Gough and R. F. Korcak [eds.], Blueberries: a century of research. Haworth Press, Inc. Maund, C., E. Estabrooks, and G. Chiasson. 2003. Blueberry Maggot. Species: Rhagoletis mendax Curran, C2.3.0. Department of Agriculture, Fisheries and Aquaculture, Fredericton, N.B, Canada. McLeod, P., F. J. Diaz, and D. T. Johnson. 2002. Toxicity, persistence, and efficacy of spinosad, chlorfenapyr, and thiamethoxa m on eggplant when applied against the eggplant flea beetle (Coleoptera: Ch rysomelidae). J Econ Entomol 95: 331-335. Micro Flo Company. 2006. Malathion 5EC. In ChemSearch CDMS [ed.]. Crop Data Management Systems, Inc. [http://www.cdms.net ]

PAGE 161

147 Milne, M., G. H. Walter, and J. R. Milne. 2002. Mating aggregations and mating success in the flower thrips, Frankliniella schultzei (Thysanoptera: Thripidae), and a possible role for pheromone s. J Insect Behav 15: 351-368. Mizell, R. F. 2003. Insect Management in Blueberri es. pp. 2, Fact sheet IFAS, University of Florida, Gainesville. Mizell, R. F., and M. C. Sconyers. 1992. Toxixity of imidacloprid to selected arthropod predators in the laboratory Fla Entomol 75: 277-280. Mondal, S., K. N. Bhattacharya, and S. Mandal. 1993. Thrips pollinator of Peltophorum inerme (Roxb.) Llanos. Indian Biol 25: 54-57. Morishita, M. 2001. Toxicity of some insecticides to larvae of Flankliniella occidentalis (Pergande) (Thysanoptera: Thripidae) eval uated by the petri dish-spraying tower method. Appl Entomol Zool 36: 137-141. Moritz, G. 1997. Structure, growth and development, pp. 15-64. In T. Lewis [ed.], Thrips as crop pests. CAB International, Wallingford, UK. Moritz, G., D. C. Morris, and L. A. Mound. 2001. Thrips ID Pest of the world. An interactive identification and info rmation system. CSIRO Publishing. Collingwood, Victoria, Australia. Moritz, G., S. Kumm, and L. Mound. 2004a. Tospovirus transmission depends on thrips ontogeny. Virus Research 100: 143-149. Moritz, G., L. A. Mound, D. C. Morris, and A. Goldarazena. 2004b. Pest thrips of the world. Visual and Molecular identificati on of pests thrips. An identification and information system using molecular a nd microscopical methods. LucidCSIRO publishing, Collingwood, Victoria, Australia. Morse, J. G., and O. L. Brawner. 1986. Toxicity of pesticides to Scirtothrips citri (Thysanoptera: Thripidae) and implica tions to resistance management. J Econ Entomol 79: 565-570. Mound, L. A. 1997. Biological diversity, pp. 197-216. In T. Lewis [ed.], Thrips as crop pests. CAB Internationa, Wallingford, UK. Mound, L. A. 2005. Thysanoptera: Diversity and in teractions. Annu Rev Entomol 50: 247-269. Mound, L. A., and D. A. J. Teulon. 1995. Thysanoptera as phytophagous opportunists, pp. 2-19. In B. L. Parker, M. Skinner and T. Lewis [eds.], Thrips biology and management. Plenum, New York, NY.

PAGE 162

148 Mound, L. A., and R. Marullo. 1996. The thrips of Central and South America: an introduction (Insecta: Thysanoptera) pp. 40-185, Memoirs on Entomology, International. Associated Publishers, Gainesville, FL. Myers, J. H. 1978. Selecting a measure of disper sion. Environ Entomol 7: 619-621. Myers, L., O. E. Liburd, and H. A. Arvalo. 2006. Survival of Geocoris punctipes Say (Hemiptera: Lygaeidae) following exposure to selected reduced-risk insecticides. J Entomol Sci 41: 57-64. Naidu, R. A., S. Adkins, K. S. Ravi, P. Ch iemsombat, R. K. Jain, H. S. Savithri, O. Gajanandana, V. Muniyappa, and D. J. Riley. 2005. Epidemiology of Tospovirusesin South and Southeast Asia Current status and future prospects, VIII International symposium on thysanoptera and tospoviruses Pacific grove, CA NASS-USDA. 2006a. Non-Citrus and nuts, 2005 preliminary summary pp. 82. In NASSUSDA [ed.]. National Agricultural Statistics Service (NASS) and United States Department of Agriculture (USDA), Arlington VA. NASS-USDA. 2006b. Non-Citrus and nuts, 2005 summary. July 2006 pp. 82. In NASSUSDA [ed.]. National Agricultural Statistics Service (NASS) and United States Department of Agriculture (USDA), Arlington VA. Ott, R. L., and M. T. Longnecker. 2004. A first course in sta tistical methods. Thomson Learning, Belmont, CA. Palumbo, J. C. 2003. Comparison of sampling methods for estimating western flower thrips abundance on lettuce. In D. N. Byrne and P. Baciewicz [eds.], 2003 Vegetable report. College of Agriculture & Life Sciences, The University of Arizona. University of Arisona, CALS, Tucson, AR. Parrella, M. P., and T. Lewis. 1997. Integrated pest management (IPM) in field crops pp. 595-614. In T. Lewis [ed.], Thrips as crop pests CAB International, Wallingford, UK. Pedigo, L. P. 2003. Economic thresholds and economic injury levels In E. B. Radcliffe and W. D. Hutchison [eds.], Radcliffe 's IPM World Textbook. University of Minnesota, St. Paul, MN. Pedigo, L. P., S. H. Hutchins, and L. G. Higley. 1986. Economic injury levels in theory and practice. Annu Rev Entomol 31: 341-368. Pollack, S., and A. Perez. 2004. Fruit and tree nuts outlook / FTS-311/July 29 2004. United States Department of Agriculture, Washington DC. Poston, F. L., L. P. Pedigo, and M. Welch. 1983. Economic injury level: Reality and practicality. Bull Entomol Soc Am 29: 49-53.

PAGE 163

149 Pritts, M., and B. Strik. 1992. The highbush blueberry, pp. 1-8. In M. Pritts and J. Hancock [eds.], Highbush blueberry pr oduction guide. Northeast Regional Agriculture Engineering Service, Ithaca, NY. Putnam, W. L. 1942. Notes on the predaceous thrips Haplothrips subtilissimus Hal. and Aeolothrips meleleucus Hal. Can Entomol 74: 37-43. Reitz, S. R. 2002. Seasonal and with in plant distribution of Frankliniella thrips (Thysanoptera: Thripidae) in northcentra l Florida tomatoes Fla Entomol 85: 431439. Sabelis, M. W., and P. C. J. VanRijn. 1997. Predation by insects and mites, pp. 259354. In T. Lewis [ed.], Thrips as crop pest s. CAB International, Wallingford, UK. Salguero-Navas, V. E., J. Funderburk, T. P. Mack, R. J. Beshear, and S. M. Olson. 1994. Aggregation indices and sample si ze curves for binomial sampling of flower-inhabiting Frankliniella species (Thysanoptera: Thripidae) on tomato. J Econ Entomol 87: 1622-1626. Sampson, B. J., and J. H. Cane. 2000. Pollination efficiencies of three bee (Hymenoptera: Apoidea) species visiting rabbiteye blueberry. Hort Entomol 93: 1726-1731. Sampson, B. J., T. A. Rinehart, O. E. Li burd, S. J. Stringer, and J. M. Spiers. 2006. Biology of parasitoids (H ymenoptera) attacking Dasineura oxycoccana and Prodiplosis vaccinii (Diptera: Cecidomyiidae) in cultivated blueberries. Ann Entomol Soc Am 99: 113-120. Sargent, S. A., J. K. Brecht, and C. F. Forney. 2006. Blueberry harvest and postharvest operations: quality, maintenance, and food safety, pp. 139-151. In N. F. Childers and P. M. Lyrene [eds.], Blueberr ies for growers, gardeners, promoters. Dr. Norman F. Childers Horticultura l Publications Gainesville, FL. Sarzynski, E. M., and O. E. Liburd. 2003. Techniques for monitoring cranberry tipworm (Diptera: Cecidomy iidae) in rabbiteye bluebe rry and southern highbush blueberries. J Econ Entomol 96: 1821-1827. SAS Institute Inc. 2002. SAS system for Windows com puter program, version 9.00. SAS Institute Inc., Cary, NC. Schexnayder Jr., H. P., T. E. Reagan, and D. R. Ring. 2001. Sampling for the sugar cane borer (Lepidoptera: Crambidae) in sugarcane in Louisiana. J Econ Entomol 94: 766-771. Scott, S. J., P. J. McLeod, F. W. Montgomery, and C. A. Hander. 1989. Influence of reflective mulch on incidence of th rips (Thysanoptera: Thripidae: Phlaeothripidae) in staked tomatoes. J Entomol Sci 4: 422-427.

PAGE 164

150 Shelton, A. M., J. T. Andaloro, and J. Barnard. 1982. Effects of cabbage looper, imported cabbageworm, and diamondback moth on fresh market and processing cabbage. J Econ Entomol 75: 742-745. Shipp, J. L., and K. Wang. 2003. Evaluation of Amblyseius cucumeris (Acari:Phytoseiidae) and Orius insidiosus (Hemiptera: Anthocoridae) for control of Frankliniella occidentalis (Thysanoptera: Thripidae) on green house tomatoes. Biol Control 28: 271-281. Shipp, J. L., K. Wang, and M. R. Binns. 2000. Economic injury level for western flower thrips (Thysanoptera: Thri pidae) on greenhouse cucumber. J Econ Entomol 93: 1732-1740. Smit-Gill, S. J. 1975. Cytophysiological basis of disr uptive pigmentary patterns in leopard frog Rana pipiens. II. Wild type and mutant cell specific patterns. J Morphol 146: 35-54. Southwood, T. R. E. 1989. Ecological methods with partic ular reference to the study of insect populations. Chapman and Hall, New York. Spiers, J. M. 1978. Effect of stage of bud development on cold injury in rabbiteye blueberry. J Am Soc Hortic Sci 103: 452-455. Stern, V. M., R. F. Smith, R. Va n den Bosh, and K. S. Hagen. 1959. The integrated control concept. H ilgardia 29: 81-101. Stone, J. D., and L. P. Pedigo. 1972. Development and economic-injury level of the green cloverworm on soybean in Iowa. J Econ Entomol 65: 197-2001. Studebaker, G. E., and J. B. Kring. 2003. Effects of insecticides on Orius insidiosus (Hemiptera: Anthocoridae), measured by field, greenhouse and Petri dish bioassays. Fla Entomol 86: 178-184. SYSTAT Software Inc. 2006. Sigma Plot for windows 2000 V6.00 computer program, version. SYSTAT Software Inc., Chicago, IL. Taylor, L. R. 1984. Assessing and interpreting the sp atial distributions of insect populations Annu Rev Entomol 29: 321-357. Terry, L. I. 1991. Frankliniella occidentalis (Thysanoptera: Thripi dae) oviposition in apple buds: role of blossom state, bl ossom phenology, and population density. Environ Entomol 20: 1568-1576. Terry, L. I. 1997. Host selection, communication and reproductive behavior, pp. 65-118. In T. Lewis [ed.], Thrips as crop pests. CAB International Wallingford, UK.

PAGE 165

151 Teulon, D. A. J. 1988. Pest management of the New Zealand flower thrips (Thrips obscuratus (Crawford) Thysanoptera: Thripida e) on stone fruit in Canterbury, New Zealand Lincoln College Universi ty of Canterbury, Canterbury, New Zealand. Teulon, D. A. J., and D. R. Penman. 1992. Colour preferences of New Zealand thrips (Terebrantia: Thysanoptera) N Z Entomol 15: 8-13. Teulon, D. A. J., J. W. Groni nger, and E. A. Cameron. 1994. Distribution and host plant associations of Taeniothrips inconsequens (Thysanoptera: Thripidae) Environ Entomol 23: 587-611. Toapanta, M., J. Funderburk, S. We bb, D. Chellemi, and J. Tsai. 1996. Abundance of Frankliniella spp. (Thysanoptera: Thripidae) on winter and spring host plants. Environ Entomol 25: 793-800. Trichilo, P., and T. F. Leigh. 1986. Predation on spider mite eggs by the western flower thrips, Frankliniella occidentalis, (Thysanoptera: Thripidae) an opportunist in a cotton agroecosystem. Environ Entomol 15: 821-825. Triplehorn, C. A., and N. F. Johnson. 2005. Borror and Delong's Introduction to the study of insects. Thomson Brooks/Cole, Belmont, CA. U.S. Highbush Blueberry Council. 2002. The high bush blueberry, history. [http://www.blueberry.org/blueberries.htm] last accessed; Nov2006 Ullman, D. E. 2005. Thrips vectors & tospoviruses: On the edge of adaptation, VIII International symposium on Thysanoptera and tospoviruses Pacific grove, CA United States Congress (104th). 1996. Food Quality Protection Act of 1996. [http://www.epa.gov ] last accessed: Nov: 2006 USDA. 1999. Crop Profile for Blueberries in Georgi a. NSF Center for Integrated Pest Management, North Carolina Stat e University, Raleigh, NC Van de Veire, M., and D. Degheele. 1995. Comparative laboratory experiment with Orius insidiosus and Orius albipidipennis (Het: Anthocoridae), two candidates for biological control in green houses Entomophaga 40 (3): 341-344. VanDriesche, R. G., and T. S. J. Bellows. 2001. Biological control. Kluwer Academic Publishers, Norwell, MA. Wang, K., and J. L. Shipp. 2001. Sequential sampling plans for western flower thrips (Thysanoptera: Thripidae) on greenhous e cucumbers. J Econ Entomol 94: 579585. Williamson, J. G., P. M. Lyrene, and E. P. Miller. 2000. A survey of blueberry acreage in Florida. Proc Ann Meeting Fla State Hortic Soc 123: 24-25.

PAGE 166

152 Williamson, J. G., G. Krewer, G. Pavlis, and C. Mainland. 2006. Blueberry soil management, nutrition and irrigation pp. 60-74. In N. F. Childers and P. M. Lyrene [eds.], Blueberries for growers, gardeners and promoters. Dr. Norman F Childers Publications Gainesville, FL. Wilson, L. J., L. R. Bauer, and G. H. Walter. 1996. 'Phytophagous' thrips are predators of twospotted spider mites (Acari: Tetran ychidae) on cotton in Australia. Bull Entomol Res 86: 297-305. Yarborough, D. 2006. Blueberry pruning and pollination, pp. 75-83. In N. F. Childers and P. M. Lyrene [eds.], Blueberries fo r growers, gardeners and promoters. Dr. Norman F Childers Publications Gainesville, FL. Yudin, L. S., B. E. Tabashnik, W. C. Mitchell, and J. J. Cho. 1991. Effects of mechanical barriers on distribution of thrips (Thysanoptera: Thripi dae) in lettuce. J Econ Entomol 84: 136-139.

PAGE 167

153 BIOGRAPHICAL SKETCH Hector Alejandro was born in Bogota, Colombia in 1977. After graduating from high school he began a career studying ag ronomy at the Universidad Nacional de Colombia in Bogota. During this degree hi s specialization and focus was in plant protection. In 2000, he moved to the United Stat es to pursue his gra duate education. He began studying English at the English Language Institute at The University of Florida, and six months later he was admitted into the masters program at the same university. During his masters, he worked in mole cricke t biocontrol. He graduated with his masters in entomology, and he accepted a position to co ntinue with his studies at the Small Fruit and Vegetable IPM Laboratory at the same university. During his PhD his research was focused in the interaction between flower th rips and early-season bl ueberries. The study included biology, population dynamics and control of this pest. He is planning to pursue a career as an entomologist in academia.


Permanent Link: http://ufdc.ufl.edu/UFE0017368/00001

Material Information

Title: A Study of the Behavior, Ecology, and Control of Flower Thrips in Blueberries Towards the Development of An Integrated Pest Management (IPM) Program in Florida and Southern Georgia
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0017368:00001

Permanent Link: http://ufdc.ufl.edu/UFE0017368/00001

Material Information

Title: A Study of the Behavior, Ecology, and Control of Flower Thrips in Blueberries Towards the Development of An Integrated Pest Management (IPM) Program in Florida and Southern Georgia
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0017368:00001


This item has the following downloads:


Table of Contents
    Title Page
        Page i
        Page ii
    Dedication
        Page iii
    Acknowledgement
        Page iv
    Table of Contents
        Page v
        Page vi
        Page vii
    List of Tables
        Page viii
        Page ix
    List of Figures
        Page x
        Page xi
        Page xii
    Abstract
        Page xiii
        Page xiv
    List of Figures
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
    Literature review
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
    Sampling techniques and dispersion of flower thrips in blueberry fields
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
    Pest phenology and species assemblage of flower thrips in Florida and southern Georgia in early-season blueberries
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
    Host status, injury description, and determination of economic injury levels of flower thrips for early-season blueberries
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
    Efficacy of reduced-risk insecticides to control flower thrips in early-season blueberries and their effect on Orius insidiosus, a natural enemy of flower thrips
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
    Effectiveness of preventive and inundative biological control tactics to manage flower thrips populations in early-season blueberries
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
    General conclusions and experiences working with flower thrips in early-season blueberry fields
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
    References
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
    Biographical sketch
        Page 153
Full Text











A STUDY OF THE BEHAVIOR, ECOLOGY, AND CONTROL OF FLOWER THRIPS
IN BLUEBERRIES TOWARDS THE DEVELOPMENT OF AN INTEGRATED PEST
MANAGEMENT (IPM) PROGRAM IN FLORIDA AND SOUTHERN GEORGIA














By

HECTOR ALEJANDRO AREVALO RODRIGUEZ


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

UNIVERSITY OF FLORIDA


2006





























Copyright 2006

by

Hector Alejandro Arevalo Rodriguez
































To my family Hector, Olga and Pedro for standing with me all the way.















ACKNOWLEDGMENTS

I would like to thank all the people that were involved in the development of this

dissertation, principally my major professor Dr. Oscar E. Liburd, and the members of my

committee, Drs. J. Howard Frank, Frank Slansky Jr., and Paul M. Lyrene, for their close

involvement. I would like to thank the staff of the Small Fruit and Vegetable IPM

Laboratory at the University of Florida for their collaboration and support. Thanks to

Aimee Fraulo for her collaboration with drawings and maps. Special thanks go to all the

blueberry growers who allowed me to use their farms to collect data. Thank you Florida

Blueberry Growers Association, Sustainable Agriculture Research and Education

Program (SARE), and the Environmental Protection Agency (EPA) for partially funding

this project. Finally, I will like to acknowledge my family, Hector, Olga y Pedro who

were with me all the way.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ................................................................................................. iv

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

LIST O F FIG U RE S ..........................................................................................x..........

ABSTRACT ............................................ ............. ................. xiii

C H A P T E R .............................................................................................................................

1 IN TR O D U C T IO N ...................................................................... .... ..... ...............

2 LITER A TU R E R EV IEW .................................................................... ...............6...

Blueberry H history and Production Practices............................................ ...............6...
Plant Selection and Com m on V arieties............................................ ...............7...
Soil M anagem ent and Preparation.................................................. ............... 10
B lueberry Pollination................................................ ............... .. .. .............. 11
P est C om plex in B lueberries ....................................... ....................... ............... 12
Arthropod Pests .................................. .......... ........................ 12
B lueberry D diseases .............. .................. .............................................. 15
Thrips: D diversity and E cology................................... ...................... ............... 15
B behavior and E cology ......................................... ........................ ............... 16
D ispersal behavior of thrips .................................................... ................ 17
Population dynam ics of thrips.................................................... 18
T hrip s as C rop P ests ......................................... .. ....................... .. ........... .. 2 5
Thrips as Tosp ovirus vectors................................................... ............... 26
T hrip s in blu eb erries....................................... ...................... ................ 27
T hrips control ...................................................................................... 28
R educed-risk insecticides .......................................... ......................... .................. 30
E conom ic Injury Levels (EIL) ......................................................... ..................... 31
Tables and Figures .................... .. ........... ............................. 35

3 SAMPLING TECHNIQUES AND DISPERSION OF FLOWER THRIPS IN
B L U E B ER R Y FIE LD S ..................................................................... ................ 41

M materials an d M eth od s ............................................................................. ............... 42
Methodology to Determine Thrips Population Inside Blueberry Flowers ..........42


v









Vertical Distribution of Flower Thrips in Blueberry Fields...............................43
T hrips D ispersion .............. .................. ................................................ 45
R e su lts............................................. .............................................................. ........ 4 6
Methodology to determine thrips population inside blueberry flowers ..............46
V ertical D distribution of Flow er Thrips........................................... ................ 46
T hrips D ispersion .............. .................. ................................................ 48
2004 farm F L 02 ... ................................................................ .. ........... .. 48
2004 farm F L 03 ... ................................................................ .. ........... .. 49
2005 farm FL02.. .................................................................... ........... 49
2005 farm F L 03 ....................................................................... . .......... 50
D isc u ssio n ............................................................................................................... ... 5 0
Tables and Figures .................... .. ........... ............................. 54

4 PEST PHENOLOGY AND SPECIES ASSEMBLAGE OF FLOWER THRIPS
IN FLORIDA AND SOUTHERN GEORGIA IN EARLY-SEASON
B L U E B E R R IE S ............................................................ ............................................. 6 9

M materials and M ethods .. ..................................................................... ................ 70
R e su lts....................................................................................................... ....... .. 7 2
P est p h en o lo g y : ................................................................................................... 7 2
S p ecies assem b lag e : .......................... .. ............ ........................ ....................... 74
Rapid Determination of the Most common Species Found in Early-Season
Blueberry Fields ....................................................................... 76
D isc u ssio n ............................................................................................................... ... 8 0
Tables and Figures .................... .. ........... ............................. 83

5 HOST STATUS, INJURY DESCRIPTION, AND DETERMINATION OF
ECONOMIC INJURY LEVELS OF FLOWER THRIPS FOR EARLY-SEASON
B L U E B E R R IE S ............................................................ ............................................. 8 9

M material and M methods ............................................................ .......................... 90
Host status of early-season blueberry bushes for flower thrips........................ 90
Economic Injury Level and injury description...............................................91
Correlation between the number of thrips inside the flowers and on sticky
tr a p s .............................................................................................................. .. 9 3
R e su lts.................................................. ......................................................... ........ 9 4
Host status of early-season blueberry bushes for flower thrips........................ 94
Economic Injury Level and injury description...............................................94
Correlation between the number of thrips inside the flowers and on sticky
tra p s .............................................................................................................. .. 9 7
D isc u ssio n ............................................................................................................... ... 9 7
Tables and Figures .................. .. ........... ............................. 101

6 EFFICACY OF REDUCED-RISK INSECTICIDES TO CONTROL FLOWER
THRIPS IN EARLY-SEASON BLUEBERRIES AND THEIR EFFECT ON
ORIUS INSIDIOSUS, A NATURAL ENEMY OF FLOWER THRIPS .................. 104









M material and M ethods ................................................................... ............... 106
F ie ld T ria ls .........................................................................................................1 0 6
Laboratory Trials .................................................................. 109
R results and D discussion ............................ .......................................... 112
F ie ld tria ls ..........................................................................................................1 1 2
Laboratory Trials ......................................................... .. ............... 113
Thrips bioassay ..................................................................................... 113
Orius insidiosus bioassay ...... ......... .. ........ ...................... 114
Tables and Figures .................. ............. ............................. 118

7 EFFECTIVENESS OF PREVENTIVE AND INUNDATIVE BIOLOGICAL
CONTROL TACTICS TO MANAGE FLOWER THRIPS POPULATIONS IN
EARLY -SEA SON BLUEBERRIES ........................................................................ 123

M materials and M ethods ................... .............................................................. 124
R results and D discussion ................ .............. ............................................ 125
Tables and Figures .................. ............. ............................. 129

8 GENERAL CONCLUSIONS AND EXPERIENCES WORKING WITH
FLOWER THRIPS IN EARLY-SEASON BLUEBERRY FIELDS .....................134

Flower Thrips Monitoring and Sampling .......... ...................... 135
Flow er Thrips in B blueberries ................................... ...................... ..................... 36
Flow er Thrips C control .................... .............................................................. 138

LIST O F REFEREN CE S .. .................................................................... ............... 141

BIOGRAPH ICAL SKETCH .................. .............................................................. 153















LIST OF TABLES


Table page

2- 1 List of diseases reported in blueberries in the United States...............................35

2- 2 Number of pollen grains per flower for four plant species, and the extrapolated
percentage of the grains that could be eaten by five or 100 thrips per flower in
three days (95% confidence lim its) .................................................... ................ 38

2- 3 Some estimates of population parameters of pest thrips.....................................38

2- 4 Known tospoviruses and thrips vectors in the world .........................................39

2- 5 Reduced-risk, biopesticides and OP alternative insecticides registered or
pending registration for use in blueberries ......................................... ................ 40

3- 1 Distribution indices, Green's index (Cx) and Standardized Morisita's index (Ip),
used to describe the level of aggregation of thrips population on farm FL02 in
F lo rid a in 2 0 0 4 ......................................................................................................... 5 4

3- 2 Distribution indices, Green's index (Cx) and Standardized Morisita's index (Ip),
used to describe the level of aggregation of thrips population on farm FL03 in
F lo rid a in 2 0 0 4 ........................................................................................................ 5 4

3- 3 Distribution indices, Green's index (Cx) and Standardized Morisita's index (Ip),
used to describe the level of aggregation of thrips population on farm FL02 in
F lo rid a in 2 0 0 5 ........................................................................................................ 5 5

4- 1 Pearson correlation coefficients for the relationship between percentage of
opened flowers and thrips population captured in sticky traps and inside five
blueberry inflorescences ......................................... ......................... ................ 83

4- 2 Dates, latitude, and principal characteristics of flower thrips population in 2004
and 2005 from the samples taken from south-central Florida to southern
G e o rg ia ................................................................................................................ .... 8 3

4- 3 Distribution of the thrips species complex in Florida and southern Georgia...........84

6- 1 Proportion of Frankliniella bispinosa surviving at various times after the release
of the insects in the bioassay arenas in essays conducted in 2004 .......................118









6- 2 Proportion of Orius insidiosus surviving at various times after release into
b io assay aren as ....................................................................................................... 1 19















LIST OF FIGURES


Figure page

3- 1 Vertical distribution of thrips captured with respect to southern highbush
blueberry bushes in south Florida ....................................................... ................ 56

3- 2 Vertical distribution of thrips captured with respect to rabbiteye blueberry
bushes in southern G eorgia ...................................... ....................... ................ 56

3- 3 Map of farm FL02 located at N 280 04' W 81 34' in north central Florida............57

3- 4 Map of farm FL03 located at N 280 04' W 81 34' in north central Florida............58

3- 5 Population dynamics inside the "hot-spot" in coordinates (4, 4) of Figure 3- 6
for 2004 on farm FL 02 .......................................................................... ................ 58

3- 6 Number of thrips captured at 2 (a), 6 (b), 8 (c), 10 (d), 14 (e), 16 (f), 18 (g), and
22 (h) days after bloom began on farm FL02...................................... ................ 59

3- 7 Number of thrips captured on farm FL03 at 2 (a), 4 (b), 8 (c), 14 (d), 16 (e), and
20 (f), days after bloom in 2004 .......................................................... ................ 62

3- 8 Population dynamics inside the "hot-spots" in coordinates (2, 2), and (0, 4) of
Figure 3- 7 in 2004 on the farm FL03 in Florida. ............................... ................ 64

3- 9 Number of thrips captured on farm FL02 at 2 (a), 4 (b), 8 (c), 10 (d), 14 (e), 16
(f), 18 (g), and 22 (h) days after bloom in 2005.................................. ................ 65

3- 10 Population dynamics inside the "hot-spot" in coordinates (2, 3), and (5, 2) of
Figure 3- 9 on farm FL02 in Florida in 2005. ..................................... ................ 68

4-1 Phenology of thrips population on farm SFL01 in 2004....................................85

4-2 Phenology of thrips population on farm SFL01 in 2005 ................ ..................... 86

4- 3 Phenology of thrips population on farm NCFL01 in 2004 .. ................................. 86

4- 4 Phenology of thrips population on farm NCFL in 2005 .. ............. ..................... 87

4- 5 Phenology of thrips population on farm SGA01 in 2004 ..................................... 87

4- 6 Phenology of thrips population on farm SGA01 in 2005 ..................................... 88









4- 7 Dates of first and last captures of flower thrips in the various blueberry sites.
SFL01 represents south-central Florida, NCFL01 represents north -central
Florida and SGA01 represents southern Georgia................................ ................ 88

5- 1 Average number of larvae emerged from individual tissues of 10 flowers and
fru its ............................................................................................................... . 1 0 1

5- 2 Linear regression showing the average number of thrips released per flower and
the percentage of formed fruits in two principal cultivars of rabbiteye
blueberries 'Climax' and 'Tifblue' ....... ... ...... ...................... 102

5- 3 Various thrips injuries inflicted by flower thrips in blueberry flowers, a)
Represents a healthy fruit, b) Shows feeding injury, and c) Shows
oviposition/em ergence injuries ....... ........ .......... ...................... 102

5- 4 Regression illustration the number of thrips captured on white sticky traps for a
week in relation to the number of flower thrips captured in five inflorescences
collected in the sam e bush ................................... ....................... ............... 103

6- 1 Average growth rate (r) between the week before treatment application and the
week after the application of the treatments. Thrips populations correspond to
the thrips recovered from the flowers collected in IFL04.................................120

6- 2 Average growth rate (r) between the week before treatment application and the
week after the application of the treatments. Thrips populations correspond to
the thrips captured in white sticky traps collected in IFL04. ................................120

6- 3 Average growth rate (r) between the week before treatment application and the
week after the application of the treatments. Thrips populations correspond to
the thrips captured in white sticky traps collected in IGA04 ...............................121

6- 4 Average growth rate (r) between the week before treatment application and the
week after the application of the treatments. Thrips populations correspond to
the thrips captured in white sticky traps collected in IFL05.. ............................. 121

6- 5 Average growth rate (r) between the week before treatment application and the
week after the application of the treatments. Thrips populations correspond to
the thrips captured in white sticky traps collected in IFLO6. ................................122

6- 6 Average growth rate (r) between the week before treatment application and the
week after the application of the treatments. Thrips populations correspond to
the thrips recovered from inside the flowers in IFL06................ ...................122

7- 1 Average number of thrips captured per week after the release of natural enemies,
as a preventive measure, in white sticky traps located inside the blueberry bush
in 2 0 0 5 ................................................................................................................ ... 1 2 9









7- 2 Average number of thrips captured per week after the release of natural enemies,
as curative measure, on white sticky traps located inside the blueberry bush in
2 0 0 6 ...................................................................................................... ........ .. 13 0

7- 3 Average number of thrips captured per week after the release of natural enemies,
as curative measure, inside five flower-clusters collected from blueberry bushes. 131

7- 4 Growth rate (r) of thrips populations captured in white sticky traps during the
2006 flowering season one and two weeks after the release of natural enemies as
a curative alternative. ............................ ........................................... 132

7- 5 Growth rate (r) of thrips populations collected inside blueberry flowers during
the 2006 flowering season, one and two weeks after the release of natural
enem ies as a curative alternative ....... .......... ............. ................... 133















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

A STUDY OF THE BEHAVIOR, ECOLOGY, AND CONTROL OF FLOWER THRIPS
IN BLUEBERRIES TOWARDS THE DEVELOPMENT OF AN INTEGRATED PEST
MANAGEMENT (IPM) PROGRAM IN FLORIDA AND SOUTHERN GEORGIA


By

Hector Alejandro Arevalo-Rodriguez

December 2006.

Chair: Oscar E. Liburd
Major Department: Entomology and Nematology

Flower thrips are considered by growers as one of the key insect pests for early-

season blueberries in Florida and southern Georgia. The objective of this study was to

understand interactions between thrips and early-season blueberries and to develop

strategies that can be used in an Integrated Pest Management (IPM) program to control

flower thrips in blueberries. This study included the two blueberry species cultivated in

Florida and southern Georgia, rabbiteye and southern highbush. The investigation began

with the refinement of thrips sampling techniques and a study of dispersion of thrips in

blueberry plantings. From these observations, I concluded that the distribution of flower

thrips was highly aggregated in blueberry fields, which is an important factor when

considering management strategies. I also developed a new system to collect thrips from

inside blueberry flowers, which is more efficient than flower dissection. This work

continued with the analysis of population dynamics and descriptions of thrips species









assemblage in blueberries. The results obtained showed a high correlation between thrips

populations and the latitude at which blueberry plantings are located as well as with the

phenology of the flowers in blueberry bushes. I developed a dichotomous key for the six

most common species of thrips found in blueberry fields during flowering. The damage

inflicted by flower thrips on blueberries was also described. An 'economic injury level'

analysis for two of the most popular cultivars of rabbiteye blueberries was completed. A

correlation between the number of thrips captured in sticky traps and the number of thrips

found inside the flowers was developed to improve monitoring efficiency. For the

chemical control of thrips, I screened nine commercial and three experimental

insecticides. From these trials, I concluded that acetamiprid is the most effective

insecticide for thrips control. The reduced-risk insecticide spinosad is as effective as any

of the other insecticides (except for acetamiprid) and it is compatible with Orius

insidiosus Say, one of the main natural enemies of thrips. Biocontrol trials did not show

any advantages of mass releasing natural enemies as preventive or curative methods to

control flower thrips in blueberries.














CHAPTER 1
INTRODUCTION

Blueberries belong to the genus Vaccinium in the family Ericaceae. This family

also includes azaleas, cranberries, and huckleberries as their most economically important

species. There are more than 400 species of Vaccinium in the world, 26 of them are found

in North America (Pritts and Strik 1992). Cultivated blueberries are native to North

America, and have been dispersed around the world, principally to Europe and South and

Central America.

The United States (US) blueberry industry started in 1908 when breeding programs

were developed to improve wild species present in New Hampshire and New Jersey. By

1916 the first harvest of the new crosses were released to the market, but it was not until

the 1930s when several new blueberry cultivars were released and commercial

production of blueberries became popular. Despite the new and more productive

cultivars today, only 70% of the fruits come from commercial blueberry cultivars, the

remaining 30% are wild blueberries (Pollack and Perez 2004).

Worldwide production of blueberries started around 1990. The world production

of blueberries for 2003 was estimated at 238,358 metric tons (t), with the U.S. producing

approximately 51.3% of the total production, followed by Canada (32.9%), and Poland

(7%). Other countries that produce blueberries are Netherlands, Ukraine, and Chile,

Argentina (Food and Agriculture Organization of the United Nations (FAO) 2004).

In the United States, blueberry production was estimated at 113,800 t for 2004

(NASS-USDA 2006a). Michigan leads the national production with one third of the









highbush blueberries in the country. Maine produces 80% of the total low-bush

blueberries in the United States. In general, most of the states are increasing their overall

production with exception of New Jersey, which had a small reduction in the production

since 2003 (NASS-USDA 2006a). For the 2004 and 2005 season, there was an increase

in the production of fresh market blueberries, apparently due to a general increase in

production in low producing states. USDA anticipates a reduction on the price of

blueberries despite the increase in production over the next couple of years. The

consumption of fresh blueberries per-capita is expected to remain close to 0.17 Kg / year

in the United States. During the off-season (December to March) the U.S. imports fruit

principally from Chile, which is the principal provider of "winter" blueberries (NASS-

USDA 2006a).

There are only two cultivated blueberry species with low chill requirement,

southern highbush (Vaccinium corymbosum L.) and rabbiteye blueberries (Vaccinium

ashei Reade). Due to the environmental conditions these two species are the only species

cultivated in Florida and southern Georgia. These states along with California, are the

only producers of early-season blueberries (April to May) in the U.S. Early-season

blueberries have prices that can be five to six times higher than mid-season blueberries.

Although Florida represented only 1.18% of the national production of fresh blueberries

for 2003, it collected 8.24% of the money produced from blueberries in the nation. In

Florida, the total acreage has increased 25% since 2001 and overall revenue has increased

to more than $50,000,000 /year (NASS-USDA 2006a).

There are 26 insect pests reported in blueberries for Florida (Mizell 2003).

However, only four of them are considered as key pests for early-season blueberries.









These species include blueberry gall midge, cranberry fruitworm, flower thrips, and

blueberry maggot (Liburd and Arevalo 2006)

The blueberry gall midge [cranberry tipworm], Dasineura oxycoccana (Johnson) is

the primary early-season pest in blueberry plantings affecting up to 80% of the floral

buds in susceptible cultivars (Lyrene and Payne 1996). Damage resulting from D.

oxycoccana in blueberry plantings has increased significantly in the last 10 years. Floral

and leaf bud injuries caused by D. oxycoccana were previously misdiagnosed as frost

damage (Lyrene and Payne 1996). Many growers in Florida are replacing the susceptible

rabbiteye blueberries with southern highbush, which is more tolerant to this pest

(Williamson et al. 2000). In a recent study, Sarzynski and Liburd (2003) found that

allowing adults to emerge from buds kept in storage bags at room temperature was the

most effective technique for monitoring populations in highly infested blueberry fields.

Blueberry gall midge can damage developing vegetative buds after the harvest, which

may affect the yield in the following year (Liburd and Arevalo 2006).

Flower thrips of the genus Frankliniella are another important pest of early-season

blueberries. In (1999), the USDA reported that 40% of the losses in blueberries in

Georgia were attributed to flower thrips. Three species of flower thrips have been

reported repeatedly in blueberries throughout Florida and southern Georgia, and are F.

bispinosa, F. occidentalis and F. tritici (Finn 2003, Liburd and Arevalo 2005). Thrips

populations are known to move rapidly into blueberry fields with the help of wind

currents and farm workers (Lewis 1997a). Their life cycles are extremely short, taking

only 15 days if environmental conditions are conducive for their growth and

development. The short life cycles, as well as overlapping generations during the









blueberry flowering cycle, make this insect a dangerous pest that can reach economic

damaging levels in a very short period (Finn 2003). Generally, information on the types

of damage and behavior of flower thrips in early-season blueberries is limited. Sarzynsky

and Liburd (2003) were only able to obtain initial but limited information on monitoring

techniques for thrips.

Cranberry fruitworm, Acrobasis vaccinii Riley, is found from Nova Scotia in

Canada to Florida in the U.S. In the northern states and Canada it has only one generation

per year but a second generation is possible in the southern states. The larva feeds on the

fruits and each one can damage as many as 10 fruits by feeding on them and secreting a

web around the fruits, which make them unmarketable (Liburd et al. 2005).

A very important late-season pest that has the potential to be a problem is the

blueberry maggot, Rhagoletis mendax Curran. The females oviposit under the fruit's

exocarp. Seven to ten days later the larvae emerge and feed on the pulp of the blueberry

for two to three weeks and then drop to the ground to pupate. After overwintering, 80%

of the pupae emerge the next season, 19% emerge in the second season after pupation,

and the remaining 1% emerges four to five seasons later. Its damage is so severe that

berries produced on farms in eastern and midwestern states must be certified "maggot

free" to be able to be transported. There is zero tolerance to blueberry maggot in most

fresh markets and a very low tolerance in processing markets. Some of the practices to

reduce blueberry maggot infestations include sanitation, collection of fruits after harvest

and weed control (Maund et al. 2003, Liburd et al. 2005).

To control these pests, farmers have relied on the intensive use of insecticides.

However, as a response to the excessive use of highly toxic pesticides and the public









concern for a cleaner environment and healthier products, the Environmental Protection

Agency (EPA) created the Reduced-Risk Pesticide Program in 1993, but not until 1996

was it formalized by the Food Quality Protection Act (FQPA) (1996). Some of the

characteristics of reduced-risk insecticides include.

* Low effect on human health
* Lower toxicity for non-target organisms
* Low potential for groundwater contamination
* Low use rates
* Low pest resistance potential
* Compatibility with IPM practices as defined by EPA (2003)


The main objective of this work was to study the biology, movement and the

effects of flower thrips in commercial plantings of early-season blueberries. It is my goal

to develop the foundation for an IPM program to control thrips in commercial early-season

blueberry plantings in the southeastern United States. The knowledge gained will be used

to establish a management program involving monitoring, use of reduced-risk

insecticides and their effect on specific natural enemies, use of biological control, and the

calculation of an Economic Injury Level to help farmers to make the decisions in

controlling this pest.














CHAPTER 2
LITERATURE REVIEW

Blueberry History and Production Practices

Blueberries and cranberries are some of the few fruit crops native to North

America, along with blackberries, grapes, pawpaw, and mulberry. Blueberries belong to

the genus Vaccinium in the family Ericaceae and have been part of the North American

tradition for centuries (Pritts and Strik 1992). Native Americans used almost every part of

the plant for their consumption. Roots and leaves were used to make teas, the fruits were

used for fresh consumption, and dried as seasoning for meats, including beef jerky, and

the juice was used as dye for various items including covers and clothing items. During

the seventeenth century, English settlers learned to cultivate blueberries from the

Wampanoag Indians and to preserve the fruit (sun-dried) for the winter as a nutritional

supplement. However, it was not until the 1880s that the canned-blueberry industry

started in the northeast United States (U.S. Highbush Blueberry Council 2002).

In the early 1900s, Elizabeth White and Dr. Frederic Coville started the efforts to

domesticate wild highbush blueberries in New Hampshire and New Jersey. By the 1930s

the first group of domesticated blueberry cultivars was released. Today, there are 3

principal sources of germplasm for the blueberry cultivars: Vaccinium corymbosusm L.

(northern highbush), V ashei Reade (southern rabbiteye), and V angustifolium Ait

(lowbush blueberry). Despite the development of many new cultivars and the effort put

into improvement of blueberry quality, only 70% of the total production of blueberries in









the U.S. is the product of commercial cultivars. The other 30% are the product of wild

blueberries (U.S. Highbush Blueberry Council 2002, Pollack and Perez 2004).

Interest in blueberries as a minor crop influenced horticultural departments to

develop breeding programs at several land grant universities and colleges including

Rutgers and Michigan State Universities. The breeding program at the University of

Florida started in 1949. The objective of this program was initially to develop blueberry

cultivars that could be produced commercially in Florida where the winter temperatures

on average are above 130C. Two main types of blueberries are produced by this

program: the tetraploid highbush also known as southern highbush, based on crosses

between V. corymbosum, V. darrowi, V ashei. Several native blueberries have been used

as gene sources for adaptation to the particular conditions in the region (i.e., a low

number of chill hours, warm conditions most of the year, soils with low organic matter

and high bicarbonates). The most important species used in this program as gene sources

are V darrowi, V. arboreum, V. corymbosum, among others (Lyrene 1997). The

program has been very successful, allowing the Florida blueberry industry to develop as

the principal producer of early-season blueberries between April and May. This window

of opportunity gave producers of early-season blueberries a price advantage of 3 5 USD

more per pound than producers of mid-season blueberries (NASS-USDA 2006a).

Plant Selection and Common Varieties

Blueberries are a perennial crop and with proper care can be productive for many

years. The development of a successful crop starts with varietal selection. According to

Lyrene (2005), there are four main obstacles that producers encounter when selecting the

adequate variety for their farms. 1) The newer varieties have the best potential to be

highly productive but a lot of information is still unknown. On the other hand, the older









varieties are obsolete, but most of the information about their productive capabilities and

needs in the field are well known. 2) The chosen plants might not be available when

needed and in the quantities needed. In some cases the order for new plants needs to be

placed a year or more in advance to ensure the availability of the varieties selected. 3)

Blueberries in Florida need cross-pollination. It is necessary to select two or more

varieties that are highly compatible to ensure maximum fruit-set. At the same time, it is

necessary to have alternating rows of the varieties in the field, which complicates the

management and harvest. 4) Finally, there are difficulties in using Dormex (Dormex Co.

USA. LLC Parsippany, NJ) in the varieties. The use of Dormex is variety-specific, some

varieties respond positively to the use of this plant growth regulator, while some present

phytotoxicity, and the buds can be destroyed.

Principal blueberry varieties used in Florida are

Star: This variety is planted from Ocala, FL to North Carolina. It is in the late-

flowering group of blueberries in Florida. The fruit quality is high, with desirable size,

firmness and flavor. In north-central Florida it is harvested in three pickings. This variety

is usually considered as low-yield. However, it is very responsive to care during the

previous fall, and requires applications of fungicides and fertilizers from the beginning of

the crop.

Emerald: It is considered as a high-yield variety. However, it ripens 7 to 10 days

later than Star. Due to its productivity, a careful winter pruning is needed to ensure that

the fruits left on the bush will set properly. This variety responds well to Dormex in

productive areas south of Orlando, FL (Lyrene 2005).









Jewel: This variety is especially popular in north and central Florida. Jewel fruits

start ripening about five days after Star, approximately at the same time as Emerald. In

north and central Florida, Jewel may produce too many flower buds; therefore, it needs

winter-pruning to remove weak branches that may fail to carry the fruits during harvest.

Jewel responds well to applications of Dormex TM. When harvesting, it is necessary to

leave the fruits on the bush for extended periods of time because the fruit has a tart flavor.

Leaving the fruits on the bush allows them to accumulate sugars, which eventually

decreases their tartness.

Millennia and Windsor: These two varieties used to be popular in Florida.

However, their popularity was reduced due to several factors. Millenia and Windsor are

early-ripening varieties with good yields. The problem with Windsor is that, if the

weather is too hot or the fruits are not picked on time, fruit scars develop, which could be

a problem. Ideally, fruits should have a small and dry scar between the pedicle and the

fruit, which will prolong their shelf life. Millennia has excellent fruit characteristics, but

has problems with fruit-setting and is highly susceptible to botrytis, a key fungal disease,

during flowering.

The recommendations for varieties in Florida depend on the tests and knowledge

gathered year after year. However, varieties have been divided into three groups:

Obsolete, Core, and New. The actual recommendation is to plant 75% of the area with

core varieties (Star, Emerald, and Jewel for Florida) and 25% with new varieties that

show good potential. Among these new varieties: Springhigh, Springwide, Abundance,

Sapphire, and Southern Belle are the ones with the highest potential. All of them are low-









chill varieties with desirable characteristics of fruit quality, ripening and fruit set (Lyrene

2005).

Soil Management and Preparation

Blueberries prefer porous, well-drained soils (sandy-loam or loamy-sand) with high

organic matter and a low pH. It appears that there is a positive correlation between good

growth of southern highbush and the amount of sand in the soil and a negative correlation

with the amount of clay and silt (Korkac 1986). In heavy soils, such as southern Georgia,

blueberries might take longer to reach maturity but once they are mature the production

will be similar to that of other soil types (Williamson et al. 2006). For blueberry bushes to

achieve their life expectancy, 50 years, it is necessary to select a good place to plant this

crop. This place must have low risk of freeze injury, adequate soil conditions, and easy

access to water. The soil management practices described by Williamson et al. (2006)

and the actual recommendation for soil preparation of blueberries in the southeastern U.S.

region are as follows:

1. Collect and analyze soil samples, determine drainage and options to improve it
2. Clear and drain the land
3. Incorporate sulfur, phosphorus, and organic matter as recommended by the soil
analysis
4. Construct beds so that plants will have at least 18 inches of well-drained soil


Plants should be spaced 0.6 to 1.2 m appart in the row with 2.7 to 3.3 m between

rows; this density will average 2,400 plants per ha. In the southern U.S. the use of pine

bark has become a common practice. The high cost of using pine bark mulch makes it

exclusive for early-season blueberry' producers in Florida and southern Georgia. The

pine needs to be replaced every 3 to 4 years. In most cases the roots of the plants

concentrate in the pine bark and usually do not penetrate the soil. This method makes it









possible to produce blueberry in soils that are otherwise not adequate for its production

(Williamson et al. 2006).

Blueberry Pollination

Due to structure and genetics, early-season blueberries need to be cross-pollinated

to achieve an adequate fruit set. Blueberry plants are known to be entomophilous and

depend mostly on hymenopterans for their pollination. Adequate pollination must occur

between three and six days after the stigma is receptive, otherwise the fruit set will be

poor.

In North America, some hymenopterans have co-evolved with blueberries. This is

the case of bees from the genera Osmia and Habropoda, which pollinate only blueberries

and their entire life cycle is coordinated with blueberry bloom (Yarborough 2006).

However, the commonest bee used by commercial growers is the honey bee, Apis

mellifera L. Because A. mellifera is commercially available, it reduces the growers'

dependency on fragile natural populations of bees for their blueberry pollination,

ensuring a high fruit setting. The recommendation for the number of beehives in

blueberries is between 50,000 and 150,00 bees per ha for northern highbush (Yarborough

2006). In the case of rabbiteye blueberries, three main pollinators can be used

successfully in the field: Osmia ribifloris Cockerell, Habropoda laboriosa (F.), the

southeastern blueberry bee, and the honey bee, A. mellifera. Sampson and Cane (2000),

compared the pollination efficiency of 0. ribifloris, H. laboriosa, and A. mellifera in

three of the core cultivars for rabbiteye, Tifblue, Premier, and Climax. They found no

significant difference in fruit set among the varieties. However, when all the varieties

were averaged, there were significant differences in terms of which species prefers which

variety. Tifblue was reported to have a good response to 0. ribifloris and H. laboriosa,









but not to A. mellifera. Premier has a high percentage of fruit set when H. laboriosa and

A. mellifera are used but not when 0. ribifloris is the pollinator. Finally, Climax seems to

have a good response to all the bee species used in the trial (Sampson and Cane 2000).

Experiments conducted by Sampson (unpublished data) found that the optimal number of

bees needed for maximum fruit set is between 5 and 6 bees per 1,000 opened flowers, this

means 5 to 6 bees per bush during peak pollination. However, the number of bees needed

should be studied on a farm-to-farm basis. Pollinators tend to get distracted by other

nectar sources around the blueberry crops, therefore the distribution of the hives and the

number of hives needed depend on the conditions of the farm and the natural population

of bees in the surrounding areas.

Unlike honey bees, indigenous genera of bees including Andrena, Halictus,

Bombus, Lasioglossum use sonication to harvest the pollen from certain plant species

including plants of the genus Vaccinium. This sonication or buzz-pollination significantly

increases the amount of pollen collected and thus the amount of pollen transported to

other flowers (Cane and Payne 1988, Javorek et al. 2002). Javorek et al. (2002),

compared various bee genera to determine which ones were the most efficient in

collecting and pollinating lowbush blueberry flowers. They found that, for example, a

honey bee, which does not use sonication, will need to visit a flower 4 times to harvest

the same amount of pollen than a bumble bee, which uses sonication; at the same time,

bumble bees (97%) pollinate close to 4 times more flowers than honey bees (24%).

Pest Complex in Blueberries

Arthropod Pests

Seven arthropod species are considered as major pests in blueberries. However,

only four of them are considered as key pests for early-season blueberries in Florida and









southern Georgia. These insects are blueberry maggot, Rhagoletis mendax Curran,

blueberry gall midge, Dasineura oxycoccana Johnson, cranberry fruitworm, Acrobasis

vaccinii Riley, and flower thrips Frankliniella spp. (Liburd and Arevalo 2006), which are

discussed below.

Blueberry maggot: Rhagoletis mendax Curran (Diptera: Tephritidae). This insect

is the principal pest in blueberries in the eastern U.S. (Liburd et al. 1999). It is found in

all blueberry regions east of the Rocky Mountains. This species belongs to the same

genus as the apple maggot, Rhagoletispomonella (Walsh). Both of them are host-specific

and concentrate their damage on the fruits, making these unmarketable (Liburd et al.

1999). Their damage is so devastating that USDA has imposed restrictions on the

transport of blueberries, and some states, such as Florida, that do not have R. mendax

reported, have a zero-tolerance policy for this insect (Liburd et al. 1998, Liburd and

Arevalo 2006). Monitoring for this particular species is based on the use of yellow sticky

boards, green sticky spheres and red sticky spheres. There must be a minimum of one

trap per ha and one of the traps should be placed within 18 m from the border of the crop.

The economic threshold for R. mendax has been defined as 2 flies per trap per week

(Liburd et al. 2000, Liburd et al. 2006). Management strategies for this tephritid include

reduced-risk insecticides, insecticide-treated spheres, and natural enemies such as

Diachasma alloeum Muesebeck (Hymenoptera: Braconidae) which is being researched in

order to improve its efficiency controlling this pest (Liburd and Arevalo 2006)

Blueberry gall midge: Dasineura oxycoccana (Diptera: Cecidomyiidae). This

species, formerly known as cranberry tipworm, was recently reported a pest of

blueberries in southeastern crops (Lyrene and Payne 1992, Sarzynski and Liburd 2003).









After their discovery, it was established that, if left unmanaged, blueberry gall midge can

destroy up to 80% of the crop production (Lyrene and Payne 1996). Eggs hatch between

2 and 3 days after oviposition, then larvae feed on the young buds, killing them and

preventing the formation of new flowers and leaves. Post-harvest damage could

potentially affect the yield for subsequent years (Liburd and Arevalo 2006). To monitor

blueberry gall midge, Sarzynski and Liburd (2003) established that collecting buds from

the field, 20 buds per ha, and placing them in zip-lock bags for 14 days is the most

accurate way to determine the presence of this species. The recommendations to control

this insect rely on the use of reduced-risk insecticides for fields with a history of

blueberry gall midge or, even if the presence of this species has been confirmed (Liburd

and Arevalo 2006). However, six natural enemies ofD. oxycoccana have been identified,

but none of them has been made commercial, and studies about their efficacy managing

the pest are still under way (Sampson et al. 2006).

Cranberry fruitworm: Acrobasis vaccinii (Lepidoptera: Pyralidae) is present in

all the places where its host plants are present in North America. Its host plants include

huckleberries, dangle-berries, beach plumbs, apples, cranberries, and blueberries

(Beckwith 1941). It is considered a key pest for blueberries. Each larva can damage

between 5 and 10 fruits during its development, and the affected blueberry clusters will

present webbing and berry deformation. The use of pheromones in sticky traps is

recommended to monitor the activity of this insect, and based on the observations on

these traps insecticide applications might be necessary at the beginning of the flying

season (Liburd and Arevalo 2006).









Blueberry Diseases

Blueberries are susceptible to many diseases. Most of them can be prevented with

optimal management of the crop and with good decisions at the moment of selecting

varieties and plots to be used for blueberry production. Blueberries are susceptible to

diseases caused by fungus, viruses, phytoplasmas, bacteria, nematodes, dodder, and some

physiological disorders. Some diseases have been attributed to abiotic factors such as

deficient nutrition, freeze, and poor water management, among others. A summary of the

diseases reported for blueberries is found in Table 2- 1

Thrips: Diversity and Ecology

Thrips belong to the order Thysanoptera, which literally means "fringed wings."

However, the English name for thrips is derived from the Greek word for "woodworm,"

because early naturalists found various species in dead branches (Mound 2005).

Thysanoptera are characterized by fringed wings in the adult stage, and asymmetric

mouthparts (Triplehorn and Johnson 2005). The left mandible is the only one that

develops because the right one is resorbed by the embryo (Heming 1993). The

mouthparts of this order have been described as "punch and suck". The mandible is used

to break the external layer of plant cells or pollen grains and the contents are sucked

through the maxillary stylets, which are joined to form a tube (Triplehorn and Johnson

2005).

The order Thysanoptera is divided into two suborders, Tubulifera, one family, and

Terebrantia, eight families worldwide. The females in Tubulifera do not have an

ovipositor and the distal abdominal segment is similar to the males. This segment is

tubular in shape and ends in a series of setae. The forewings in Tubulifera have neither

venation nor setae except for the base. Terebrantia are the most common suborder and the









one that has the greatest effect on agriculture. Close to 94% of the total pest species are in

this suborder, all of them in the family Thripidae (Moritz et al. 2004b).

The metamorphosis of thrips is intermediate. There have been some discussions as

to whether thrips should be classified as holometabolous or hemimetabolous. The first

two instars do not have external wings, because they are being developed internally.

Usually these two instars are called larvae and resemble holometabolous metamorphosis.

Thrips display two more distinct immature stages, which show vestigial wings but they

do not feed. The first of these stages is called propupa, which shows vestigial wings

(except in Tubulifera). Following the propupa stage, a real pupa is formed, similar to the

adults. The main differences between pupa and adult are that the pupa are not mobile, has

two pairs of vestigial wings, and the antenna has fewer segments, while the adults are

mobile, macropterous have well formed wings and have between 6 and 9 antennal

segments. The pupa is also inactive but has external development of the wings and

morphological resemblance to the adult stage, which refers to a hemimetabolous

metamorphosis (Triplehorn and Johnson 2005). Propupa and pupa differ from the adult

stage in the size, morphology and functionality of the wings, and the segments in the

antennae and legs are reduced. Pupa and propupa are immobile (Moritz 1997).

Behavior and Ecology

Based on thrips' alimentary preferences, they can be divided into fungivorous,

phytophagous, predacious, and omnivorous species. The damage caused by thrips in

agricultural crops is primarily due to feeding on leaves, flowers or fruits, and secondarily

to oviposition in these same structures (Kirk 1995). However, little is known about the

feeding behavior of these insects in the field including diets, host-switching behavior, etc.

(Kirk 1997a). In the case of Terebrantia, more than 95% of the species are associated









with green plants. However, in most of the cases the host report is based on the places

where thrips are found and not on the places where they breed, making most of the

records confusing and the definition of host plant very subjective (Mound 2005).

Dispersal behavior of thrips

Thrips in general have two means of dispersal, artificial and natural dispersal.

Artificial dispersal is usually human-assisted and is facilitated by the increasing

international transportation of agricultural products. Thrips are easily transported in

various products including potted and cut flowers and several fruits and vegetables that

are imported and exported. Accidentally transporting thrips across borders is relatively

easy. They are difficult to spot in a port inspection due to their small size. Furthermore,

the eggs of these insects are found inside plant tissues and the signs left by the

ovipositing female are minimal.

The second method, natural dispersal, is accomplished by thrips using natural

means and the most common method is flying. Just before to flying, thrips have a very

complicated preparation for takeoff During this period macropterous forms bend their

abdomen and use setae located on abdominal tergites V to VIII to comb those located on

the wings. The objective of this movement is to increase the surface-area of the wings,

facilitating take-off (Ellington 1980). Thrips have been reported to fly at 6 to 30 ms-1

depending on the species. However, it is known that thrips disperse to distances further

than that which they would be able to independently fly. One of the explanations for this

phenomenon is the use of wind currents (Lewis 1997a). Thrips, like many small insects,

potentially use wind currents to move long distances. This phenomenon is summarized

by Gatehouse (1997). Small insects might take advantage of the convective upper

currents developed by warm air-pockets, which have speeds of ascension measured up to









3m s-1 (Drake and Farrow 1988). These currents help the insects to reach the Flight

Boundary Level (FBL) for each insect species, which might run into the Planetary

Boundary Level (PBL), which is the layer between the ground and the free atmosphere.

This PBL is located between 100 and 3000m above the ground and once the insects break

it, it facilitates their dispersal. Radar data show that most of the massive flights are short-

lived; but some of the populations can travel overnight (Gatehouse 1997). One of the

main disadvantages of this mode of transportation is that insects have little to no control

of the direction that they are being transported. However, despite the small size of thrips

and their apparent lack of control of their flight patterns due to wind interaction, there is

good evidence indicating that thrips have a certain amount of control in the landing. Field

observations indicate that thrips land on their feet on individual plants, showing some

amount of control (Lewis 1997a). Kirk (1984) demonstrated that thrips have control of

their landing selection. The author used various colored traps on the ground separated by

5 m from each other to show that there was a 20-fold difference between flower thrips

and grass-dwelling thrips in their color selection for landing. Flower thrips were attracted

to bright colors such as white while grass-dwelling thrips were attracted to colors that

were closer to green (Kirk 1984, Teulon and Penman 1992). There is evidence that thrips

are attracted to various odors, the use of anisaldehyde (for flower thrips), or ethyl

nicotinate (for Thrips obscuratus) increased the trapping of the respective thrips

compared with the controls (Kirk 1985, Teulon 1988).

Population dynamics of thrips

To understand the population dynamics of thrips, it is necessary to understand their

relation with their host plants. There are two types of plants where thrips have been

reported. The first type is a provisional or alternate host, which might offer temporary









shelter or food, but in the vast majority of cases thrips do not reproduce in these plants.

The second group of hosts might be called proper hosts; these plants offer food, shelter, a

reproductive substrate and alimentation for the immature thrips. Unfortunately, there is a

controversy about whether the plants reported as hosts in the literature are proper hosts or

alternate hosts, and if these alternate hosts should be defined as hosts or if they are just

accidental relationships (Mound 2005). There are approximately 50 economically

important pest species among 5,300 known species of thrips. Some thrips species are

considered to be very host-specific. Those thrips species that are considered as crop pests

are usually very prolific and non-host-specific. For example Frankliniella occidentalis

(Pergande), the western flower thrips, is reported on more than 500 plant species within

50 families. However, it is necessary to remember there is controversy about reports of

host plants (Moritz et al. 2004b).

Thrips are ideal for population dynamics studies. Their populations are large and

are generally easy to sample. However, thrips sampling presents some challenges, such as

the difficulty of finding dead thrips and the fact that big migrations go unnoticed most of

the time, sometimes for unknown reasons. Some species are very common in one year

and very rare in the next. To study their population dynamics, it is necessary to consider

feeding and reproductive behavior, migration, short and long term effects of the

environment, and the effect of management techniques in the field populations (Kirk

1997b).

Feeding behavior: The mouthparts of Thysanoptera are one of the identifying

characters of this order. The mouthparts are located on the underside of the head and

form a mouthcone. This structure is formed by a single mandible (characteristic of









Thysanoptera) and two maxillary stylets. In order to feed, thrips use their mandible to

"punch" a hole in the external walls of the tissue that they are going to feed upon and

then use the stylets to suck the liquids from inside these tissues (Kirk 1997a). In the past,

feeding behavior of thrips was considered to be "rasping" or "gashing" and sucking.

However, this observation has been re-evaluated and the feeding is considered to be the

"piercing and sucking" type (Hunter and Ullman 1992).

Thrips in general, can feed on diverse plant tissues (leaves, flowers, fruits, pollen)

and some fungal tissues such as spores and hyphae. Most of the attention has been

focused on the feeding behavior of phytophagous species, thus this is the group of which

we have broader knowledge of their preferences (Kirk 1995). The feeding behavior

displayed by thrips is similar for all plant tissues. Once the thrips have landed on what

seems to be an appropriate substrate on which to feed on they start the process of probing

the tissue. They start using the legs and antennae, walking in circles or forming figure

eights on the tissue. Once they find a spot that seems to be adequate, thrips use their

mandible to probe and open a small hole in the cell wall. A small amount of liquid comes

from this small puncture. Using their palps, thrips test the liquid for the correct nutrient

compositions. If the tissues and nutrient composition are adequate, they use their

mandible and head to punch a bigger hole in the tissue and start feeding. This causes

nearby cells to collapse. If the damage occurs in the ovary in the flower these marks will

become magnified during the fruit development and the scars will be very noticeable,

reducing fruit quality (Kirk 1997a, Liburd and Arevalo 2006).

Pollen-feeding is another behavior that is common, principally among flower

thrips. These thrips feed on individual pollen grains one by one. The time spent on each









pollen grain varies between 3 to 120 s depending on thrips species and instar, grain

volume, and temperature (Kirk 1987). Thrips can ingest pollen from the anthers or the

grains found around the flowers and leaves. The potential damage that flower thrips can

have on pollen quantity depends upon the plant production of pollen and thrips

populations present in the field as observed in Table 2- 2.

As observed in Table 2- 2 there is the potential that thrips may affect the

availability of pollen for fertilization. However, thrips populations will need to be

extremely high and the pollen production by the plant very low for this to occur. Based

on calculations presented by Kirk (1987), one thrips could potentially destroy between

0.2-0.7% of the pollen in a flower per day, assuming that it fed exclusively on pollen.

Furthermore, thrips might be responsible for the destruction of anthers or the destruction

of pollen on stigmas, which would affect pollination. The damage caused by thrips on

plant fertilization depends on many factors such as timing, amount of pollen produced by

the plant, amount of pollen destroyed by thrips, effectiveness of pollinators, temperature,

etc. (Kirk 1987). In addition to interfering with pollen availability and fertilization, thrips

balance their diet by consuming other plant tissues (Kirk 1997a).

Because thrips are usually associated with plant pests, they have been overlooked

as pollinators, and there are no studies about their efficiency. However, to determine the

correlation between pollinators and flowers, a chart of "pollination syndromes" describes

the characteristics of flowers that may attract certain types of pollinators (Kirk 1997a).

Thrips pollination syndrome is called thripophily (Kirk 1988). Thrips flowers are

described by Kirk (1997a) and Mondal et al. (1993) as "medium size, white to yellow,

sweetly scented, with or without nectar, with compact floral structures or globose or









urceolate blossoms providing shelter, and with small to medium-sized pollen grains,

possibly with nocturnal pollen presentation". This description is very close to blueberry

plants, which have medium-sized white flowers with nectar, globose blossoms that

provide good shelter for thrips. So blueberry flowers meet the criteria to be pollinated by

thrips; however, more research is needed to determine the role thrips play in pollination.

It is difficult to determine the net effect of thrips on the flowers taking into

consideration the benefits of pollination and the damage to floral structures. Several

species of plants have been reported to be pollinated by thrips. Peltophorum inerme

(Roxb.) Llanos, is pollinated by two species of thrips, Thrips hawaiiensis (Morgan) and

Haplothrips ceylonicus Schmutz, in addition to various hymenopteran species (Mondal et

al. 1993). Erica tetralix L. is not only pollinated by Taeniothrips ericae (Haliday), but

they have a close mutualistic relationship. Flowers of E. tetralix offer protection to thrips

from the environment and a place to reproduce, the insect offers the plant self and cross

pollination (Hagerup and Hagerup 1953).

Another feeding behavior shown by thrips is predation. There are a few specialist

predators among thrips that have some behavioral adaptations such as speed or color

among others. Among the specialists the most common prey are mite motiles and eggs.

Some of the most common species of predatory thrips are Haplothrips kurdjumovi Karny,

which feed on moth and mite eggs (Putnam 1942), .,h,i/dl i1,% sexamaculatus

(Pergande), which feed on mites that form webs (Trichilo and Leigh 1986), and

Trichinothrips breviceps Bagnall, which feed exclusively on psocids (Kirk 1997a). Some

species of thrips feed upon other thrips larvae. Including Aeolothrips intermedius

Bagnall, which feeds on thrips immatures through their abdomen (Kirk 1997a). Some of









the polyphagous thrips are well known as pests but they can switch their preferences and

become predatory. For instance, Frankliniella occidentalis (Pergande) feed on mites in

cotton (Trichilo and Leigh 1986) and prey on twospotted spider mites. Thrips tabaci

(Lindeman) is considered to be a pest of vegetables susceptible to tospoviruses, but preys

on twospotted spider mites in Australia (Wilson et al. 1996).

Reproductive behavior: Thrips in general have short life cycles. Many

environmental factors can affect the reproduction rate and the length of their life cycle.

One of the most important factors are host plants. Plant species and quality (age, vigor,

phenological stage, etc.) affect the net reproductive rate (Ro) of thrips populations (Table

2- 3). Abiotic conditions affect the reproduction of thrips as well. These include light

regimen, temperature, and humidity among others. Kirk (1997b) presents a summary of

the effects of plant species and conditions that affect thrips reproductive behavior.

The effect of plant quality on thrips populations is very important. As observed

in Table 2- 3, plant species influence the life history of some species. Some thrips have

particular preferences towards the quality of the tissues used for oviposition. For

example, Taeniothrips inconsequens (Uzel) will only lay its eggs on convex structures

like veins in the leaves or stems (Teulon et al. 1994). Bates and Weiss (1991) showed that

Limothrips denticornis Haliday only lay their eggs on the intervein space of barley

leaves, limiting the oviposition to mature leaves. Furthermore, Chau et al. (2005) showed

a close correlation between the populations ofF. occidentalis and the level of nitrogen

fertilization in chrysanthemum. The experiment reported an increase in the number of

thrips correlated to the level of nitrogen used up to 100% of the recommended dosage for

this crop.









Many intrinsic behaviors help us to understand the relationships among thrips from

the same or from different species. The first one is the use of semiochemicals such as

alarm pheromones, aggregation pheromones, defensive mechanisms, etc (Terry 1997).

Kirk and Hamilton (2004) demonstrated the existence of some type of substance

produced by males ofF. occidentalis that has an attractive effect on females of the same

species. Unfortunately, identification of the compounds in the pheromones is still in

progress, but the description of the behavior of females, virgin females and males in a Y-

tube bioassay indicate that this might be a sex-pheromone. Milne et al. (2002) observed

what seems to be some type of attraction pheromone produced by males for females from

the same species, Frankliniella schultzei (Trybom), and a direct correlation between the

number of females per male attracted and the number of males present. There is not yet

evidence of sex-pheromones that work at long distances, but there is some indication of

short-range attractants that might help thrips to locate their mates (Terry 1997).

Apparently due to the bias in the female: male ratio, males and females have

exhibited different behaviors to locate each other. Frankliniella occidentalis males tend

to aggregate on the external side of floral structures where the females might be attracted,

probably by the substances described in Kirk and Hamilton (2004). Some females are

attracted while others ignore the signals and move themselves towards the food sources

inside the flowers. Because females do not need to mate to lay fertile eggs, mating

behavior in this order of insects is complex and not completely understood; their behavior

is very inconsistent and species specific, making it difficult to state generalizations about

this topic (Terry 1997).









Unlike mating, oviposition behavior is more general and well-described at least for

terebrantian species. These females raise the tip of the abdomen, test the tissues using the

setae in the last abdominal segment, and insert the ovipositor into selected plant tissues.

While in this position the saw-like ovipositor cuts a space for the egg in the tissue, which

is pushed out by a contraction of the abdomen. Thrips prefer to lay their eggs in mature

non-expanding tissues to avoid having the eggs crushed by the expanding cells (Terry

1997). Oviposition preferences depend on the species. Most species prefer to oviposit on

leaves or on floral tissue. In citrus, F. bispinosa oviposit in the floral tissues, it has a

preference for the pistil- calyx area followed by the petals and finally, filaments and

anthers (Childers and Anchor 1991). In apples, F. occidentalis prefers to lay its eggs in

blossoms of any age, although most adults are found in opened blossoms, the egg

concentration is higher in petalless clusters mainly in the 'king bud' (Terry 1991). Other

thrips species lay their eggs close to the inner veins of the leaves or in the fruits. The

damage caused by these thrips due to oviposition depends on the place and plant stage

selected for oviposition. Thrips that lay their eggs and feed in the commercial part of the

plant, flower or fruits, are the ones that are considered as major threats to the agricultural

industry independent of their role as virus vectors.

Thrips as Crop Pests

Monophagous thrips are rarely considered as pests. Only a few examples of this

interaction are known: Lithil ip, karnyi Bagnall, which damages Asian piper, L. adisi

Strassen, which feeds on Brazilian guarana trees, and ,Y ii/lu i, cardamomi Ramakrishna

a common pest of cardamom (Mound 2005). Most of the thrips that are considered as

severe pests are polyphagous. Due to their high adaptability, they can feed on various

resources and modify their larval stages, adapting to various temperature ranges, etc. Due









to their high plasticity, agricultural systems should not be looked at as if they were

isolated islands (Altieri 1988). Thrips are notorious for moving their population to

alternate hosts during the season when the main hosts are not very conducive. This is the

case of flower thrips, which reproduce and feed in the flowers of our crops and then

during the season when flowers are not present, they migrate to nearby crops and wild

flowers to continue their cycle (Kirk 1997b).

Thrips as Tospovirus vectors

Tospoviruses are one of the most damaging groups of pathogens in agriculture. In

recent decades, due to the increase in international trade, the spread of infected plants and

vectors has increased worldwide. Thrips and viruses are probably two of the most

difficult things to detect in the ports of entry. Thrips eggs inside the plant tissues as well

as asymptomatic plants infected with the viruses are virtually impossible to detect

(Lathman and Jones 1997). There are 16 species of viruses in the genus Tospovirus,

family: Bunyaviridae, recognized as plant pests, and they are transmitted by 11 species of

thrips, of the family Thripidae. However, the list of viruses and vectors changes due to

the complicated genetics of the virus and the discovery of new relationships with various

thrips species (Ullman 2005).

Thrips acquire viruses in the first or early second instar when there is a close

relationship between mid-gut, visceral muscles and salivary glands. Once the wing

muscles start developing and the supra-oesophageal ganglion moves towards the head the

connection between the salivary glands, the mid-gut, and the visceral muscles is ended

stopping the movement of virus particles into the salivary glands. If the thrips did not

acquire the virus during this short period, it will not be able acquire the virus due to the

lack of connection between the salivary glands and the mid-gut. In adult thrips the virus









is located in the malpighian tubes, in the lumen, the hemocoel, and in the salivary glands.

Until recently, the only proven way thrips transmit the virus is through the salivary

glands during feeding. However, there is enough evidence to support the possibility that

the virus might be transmitted through excrements and oviposition wounds, but more

research is needed (Moritz et al. 2004a).

Thrips in blueberries

To study the relationship between thrips and blueberries we can divide thrips into

three groups, which is a very broad division found in the literature: leaf thrips, flower

thrips, and flower and leaf thrips (Kirk 1997a, Liburd and Arevalo 2006).

Leaf thrips: Frankliniella vaccinii Morgan and Catinathrips kainos O'Neill are the

two main leaf pests of blueberries in northeastern U.S. They feed on the leaves right after

pruning and their larvae are found feeding inside curled leaves, which prevent them from

developing properly. In Maine, these thrips are found during the summer from late July to

early August after the pruning. After the damage is done, the larvae mature and the adult

thrips migrate to other hosts disappearing until next season (Collins et al. 1995).

Flower thrips: Unfortunately, the relationship between flower thrips and

blueberries is not well known. In an interview conducted by Finn (2003), growers of

early-season blueberries considered flower thrips as one of the most important pests of

his crop along with blueberry gall midge and blueberry maggot. The USDA reported in

(1999) that 40% of the losses in blueberries in Georgia were attributed to flower thrips.

Thrips populations are known to rapidly move into blueberry fields with the help of wind

currents and workers. Their life cycles are extremely short, taking between 15 and 20

days if environmental conditions are conducive for their growth and development. The

short life cycles as well as overlapping generations during the blueberry flowering cycle,









make this insect a dangerous pest that can reach economically damaging levels in a very

short period. The reduced amount of knowledge and the importance of this pest for

blueberry growers is the main reason to develop an Integrated Pest Management (IPM)

program to control the populations of thrips. The results shown in this dissertation are the

beginning of this IPM program answering some of the basic questions about the

relationship between blueberries and flower thrips.

Thrips control

Due to their behavior, quick reproduction rate, and potential to inflict great damage

even at low populations (in the case of virus vectors), there are 236 products registered to

control thrips in the U.S. listed by Crop Data Management Systems Inc. (CDMS)

(Marysville, CA). In blueberries there are 24 insecticides labeled to control thrips, but

only 8 active ingredients(CDMS 2006). Due to high pressure of these insects, growers

have a high dependence on chemical control for fast management of the pest.

Thrips resistance to tartar emeric insecticides was detected as early as 1941.

However, the typical example of thrips resistance is described by Morse and Brawner

(1986). They described how in four years thrips became resistant to DDT and dieldrin, 18

years to dimethoate, seven years to be resistant to malathion. Other tests in the same

species showed how i, i/I nth i/, citri (Moulton) increased its resistance by 428-fold to

fluvalinate after only 10 selections, at the same time that the resistance to other

pyrethroids increased by 10 or more (Morse and Brawner 1986).

Today there is a greater adoption of Integrated Pest Management (IPM) initiatives

to control thrips. The IPM approach is based on five techniques: host plant resistance,

chemical control, mechanical control, cultural control, and biological control (Parrella

and Lewis 1997). The only type of plant resistance that has been achieved is resistance to









certain tospoviruses vectored by thrips, which considerably reduces their damage

(Ullman 2005). In the case of direct resistance to the insect most of the work is being

conducted in non-preference changes in morphological characteristics such as form of the

leaf or even color of the product (Parrella and Lewis 1997). Various forms of mechanical

control have being tested to control thrips; the most evaluated ones are mechanical

barriers such a screens in greenhouses and filtration systems, and the use of UV reflective

mulches. Barriers in the field have proven to be not economically viable. Yudin et al.

(1991) used 1.5 m tall plastic barriers around the crop. The results showed that the

barriers only reduced the movement ofF. occidentalis by 10% while they had no effect

on the intra-crop movement of the insects. The use of reflective mulch to reduce the

population of thrips in field crops has proven to be effective. However, the reduction was

only evident when sticky traps were used in the study (Scott et al. 1989, Kring and

Schuster 1992). When the number of thrips in the flowers was counted by Kring and

Schuster (1992), they found no differences between the treatments.

Biological control of thrips is very successful in closed environments such as

greenhouses. However, in field crops the use of biocontrol agents has not been very

successful (Parrella and Lewis 1997). Hoy and Glenister (1991) tried to control T tabaci

by inoculating and inundating the field with Amblyseius spp. but it failed to show

positive results. The reason why biological control is not effective in the field might be

due to the fact that thrips populations move very fast and in large numbers. Also, thrips

might cause significant damage before the beneficial organisms have time to react and

achieve the appropriate control (Parrella and Lewis 1997)









Reduced-risk insecticides

The concept of reduced-risk insecticides was introduced by the Environmental

Protection Agency (EPA)'s Office of Pesticide Programs (OPP) in July 1992. In this

public notice there are incentives for the development and registration of new chemistries

that comply with the following characteristics to be registered as reduced-risk

insecticides (Environmental Protection Agency (EPA) 1997).

Human health effects
Very low mammalian toxicity
Between 10 to 100 times less toxic than alternatives
Displace chemistries with known lethal effects on human health such as
organophosphates
Reduce exposure to workers

Non-target organisms
Very low toxicity to birds, fish, honey bees, and other beneficial insects, and
non-target organisms in general (calculated as direct toxicity of degree of
exposure)
Highly selective to target pests

Groundwater (GW)
Low potential for GW contamination
Low drift and runoff

Lower use rates than the alternatives

Low pest resistance potential

Highly compatible with IPM

Effective to control target pests

Currently there are approximately 60 new chemistries that are considered as

reduced-risk insecticides, or biopesticides as described by the Food Quality Protection

Act of 1996 (United States Congress (104th) 1996, IR 4 project 2006). Nine of these

chemistries are registered or pending registration for use in blueberries, its chemistry and









status is summarized in Table 2- 5 as well as four Organophosphate (OP) alternatives that

can be used to reduced the effect of OPs in the environment.

Reduced-risk insecticides are considered a fundamental part of IPM programs

independent of the commodity in question (Environmental Protection Agency (EPA)

1997, Atanassov et al. 2002, Environmental Protection Agency (EPA) 2003, Finn 2003,

Hamill et al. 2003, Liburd and Finn 2003, Liburd et al. 2003, Mizell 2003, Liburd and

Arevalo 2005, IR 4 project 2006).

Economic Injury Levels (EIL)

Economic Injury Level (EIL) is one of the most discussed topics in entomology.

The reason for this is that the EIL gives us the most basic information needed for a

successful IPM program. "How many insects will cause significant damage?" The

answer to this question is usually the starting point for decision making in a commercial

crop (Pedigo et al. 1986). Stem et al. (1959) developed the first concepts of economic

damage, EIL, and the majority of these have not changed since then (Pedigo et al. 1986).

In entomology, Stem et al. (1959) defined the EIL as "The lowest population density that

will cause economic damage." This concept assumes the possibility of scouting,

evaluation and use of control tactics as needed. For this reason, it is very practical in the

case of arthropod pests since this is the root of IPM programs. Several authors have

criticized the simplicity of the EIL, arguing the lack of a more comprehensive view of the

farm as a system. Variation on commodity prices, interaction with other arthropods and

climate conditions made some ElLs stationary, obsolete, and only valid for one season

(Poston et al. 1983, Pedigo et al. 1986). Despite these critiques, EIL is the most used

method of decision-making for arthropod pests. That new authors include their ideas and









suggestions to improve it, makes the EIL a dynamic concept (Pedigo et al. 1986, Pedigo

2003).

I would like to define some of the concepts as used in this dissertation. Some

authors have determined EIL as a level of injury (Shelton et al. 1982), but in most of the

cases standardization of the injury is difficult to determine in such a way that might be

practical for growers. Pedigo et al. (1986) uses the term "injury equivalent" to determine

the injury level caused by one pest through its life cycle and the term "equivalence" as

the total injury equivalents inflicted by a population at a given moment. The term EIL in

this dissertation will correspond to the insect density causing economic damage as

defined by Pedigo (2003) and described in Equation 2-1. In this case (C) is the cost of

management per production unit, (V) is the market value per production unit, (1) defines

the injury unit per pest, (D) is the damage per injury unit and (K) is the proportional

reduction in pest attack originated by the control.

Another important value complementary to the EIL at the moment of taking

decisions is the gain threshold (GT). To define GT it is necessary to understand the

concept of economic damage (ED), which refers to the equilibrium point where the cost

of controlling the pest is equal to the damage caused by the pest, it is determined as

monetary value and it is described in terms of (C(a)), the cost of the control, (Y), yield, (P)

price per unit of yield, (s) level of pest injury, and (a) control action (Equation 2-1 (a)).

Stone and Pedigo (1972) defined GT in function of the same terms that Stern et al. (1959)

had defined as ED, but the GT is described in terms of loss of marketable product per

cultivated unit (Equation2-1 (b)).













(b) GT= C Equation 2-1:
(b) GT= a
P [S(a)]

Some producers might use this number as indicator to make decisions. However, it

is too risky to take actions when the pest has reached the EIL or the GT, because by the

time that the control practices are in place the pest might have reached the point where

the cost of controlling is higher than the value of the crop and it would not be

economically wise to take any actions. For this reason the concept of economic threshold

(ET) was included. Economic threshold is defined as the practical or operational pest

density when control must be taken in order to keep the crop as a profitable business. The

ET includes variables such as EIL, pest and host phenology, population growth rates

(variable depending on the conditions for each farm), and interaction with other

organisms or chemicals applied for other purposes. Due to the practical and mathematical

complexity of calculating this ET, most ET are "relatively crude" as expressed in Pedigo

(2003).

There are some limitations to the concept of EIL expressed by Pedigo (2003).

Some of the limitations mentioned are:

1. Lack of mathematical definition for ET
2. Lack of more comprehensive EIL
3. Reduced ability of make cost-effective, accurate population analysis in the field
4. Inability to predict market, population trends and other variables for the ET
5. Difficulty to quantify variables such as weather, environmental cost etc.


C(,) = Y Is(,) I X P Is(,) I Y(S) X P(S)









However, despite these limitations, this concept it is still the best tool at present for

growers to decide pest management strategies. Some of the values used by the growers

are empirical due to the lack of research in some of the commodities.

The relationship between blueberries and flower thrips is still unexplored. There is

a lot of knowledge about these two species generated throughout years of research. Our

objective is to use all this information to guide our research and understand the

relationship between flower thrips and blueberries. The following dissertation is an

attempt to generate and compile information regarding this relationship. The ultimate

goal is to develop an IPM program for early-season blueberries in Florida and Southern

Georgia.













Tables and Figures


Table 2- 1: List of diseases reported in blueberries in the United States.


Type


Agent
Phomopsis vaccinii


Botryosphaeria dothidea


Distribution
Southeastern U.S.


Southeastern U.S.
1-2 year old bushes


Fungi


Botryosphaeria corticis


Viruses and
Phytoplasmas


Blueberry scorch
carlavirus


Southeastern U.S.
(NJ, GA, FL, AL,
MS)


Northern coastal
states in the U.S.


Symptoms
Phomosis canker
Dieback of fruit bearing
stems
Yield reduction up to
70%
Rotting of fruits

Blueberry stem blight
Blight of individual
branches
Brown r red branches
"flags"


Stem Canker
Swelling at the point of
infection
Spore-producing
structures emerge
through the bark

Scorch
Rapid necrosis of leaves
and flowers
Small Chlorosis of the
leaves and stems


Transmission
Spores are released
from infected case
that overwintered in
the field


Spores are released
from infected stems
(blueberry and
alternate hosts) and
need an injury in the
stem to be able to
penetrate

Spores are released
during the wet season
and are wind
transported. Only
young stems are
susceptible

Vectored by aphids


Management
Remove infected and
suspicious branches
during winter pruning


Timely pruning of
infected branches far
from the start if the
symptom. Resistant
varieties are available


Sanitation, avoidance
and the use of
resistant cultivars


Infected bushes
should be removed,
burned and replaced
with tolerant cultivars











Table 2-1 continued


Type


Agent Distribution
Blueberry shock ilarvirus Western U.S.


Viruses and
Phytoplasmas


Symptoms
- Shock
- Sudden necrosis of
flowers and leafs
- A second vegetative
flush is present but no
flowers produced
- In well-managed fields
close to normal
production is possible 1
to 4 years after infection


Transmission
It is vectored by
pollinators carrying
infected pollen


Management
Infected bushes
should be destroyed
before bloom


Blueberry shoestring


Northeastern U.S.


Tobacco ringspot virus


Shoestring
Symptoms appear 4
years after infection
Reddish lines in the
stems
Leaves become red and
deformed
Ripe fruits are red

Necrotic ringspot
Leaves have 2-3 mm
necrotic spots, the
center of the injury may
fall off


It is transmitted by
the blueberry aphid,
Illinoia pepper
MacGillivray


Vectored by the
dagger nematode
Xiphinema
americanum Cobb


Use of resistant
cultivars, use of virus
free plantings, control
of the vector to
reduce spread of the
virus


Avoid the nematode,
fumigate in case the
vector-nematode is
present and use of
virus free plantings











Table 2-1 continued


Type Agent Distribution Symptoms Transmission Management
Blueberry stunt Midwestern and Stunt Vectored by Eradication and
phytoplasma eastern U.S. Short bushy canes leafhoppers destruction of the
Viruses and Leaves with yellowing infected plants
Phytoplasmas in the margins
Reduced intermodal
distance

Agrobacterium Cosmopolitan Crown gall It is a soil-borne Avoid planting
tumefaciens Potted plants and new bacterium. Enters material with obvious
plantings present galls through wounds and galls. If site is
in the roots up to 2.5 cm cuts on the roots infested plant non-
in diameter host materials for 3 to
4 years before
planting blueberries
Bacteria
Pseudomonas syringae Pacific northwest Bacterial canker The bacterium Cut infected canes in
U.S. Die back of young penetrates through late fall during the
branches wounds caused by dry season and
insects, wind, or sterilize the
manual labor equipment used with
bleach

Modified from (2006)









Table 2- 2. Number of pollen grains per flower for four plant species, and the
extrapolated percentage of the grains that could be eaten by five or 100 thrips
per flower in three days (95% confidence limits).*


Plant species


Echium plantagineum
Actinidia deliciosa
Brassica napus
Jacaranda acutifolia


Grains per
flower
156,600
2,000,000
140,000
13,400


Extrapolated consumption
5 thrips / 3 days 100 thrips/3 days
7.6% (4-11) 152% (77-227)
0.5% (0.2-0.7) 9% (5-14%)
3.2% (2-4) 64% (47-80)
3.2% (2-5) 64% (35-94)


* From Table 2 in Kirk (1987).


Table 2- 3: Some estimates of population parameters of pest thrips*
Thrips species / Crop Temperature L:D Ro rm T re Tc
C day 1 days day-1 days

Frankliniella fusca
Peanut 20 14:10 5.07 0.05 31.2
Peanut 30 14:10 16.0 0.16 17.5
Peanut 35 14:10 1.67 0.04 13.2

F. occidentalis
Bean 23 04:20 3.7 0.062 21.2
Bean 23 16:08 12.2 0.140 17.9
Chrysanthemum 15 42.2 0.056 66.5
Chrysanthemum 35 2.7 0.056 17.6
Cotton -pollen 27 14:10 30.1 0.157 21.6
Cotton + pollen 27 14:10 111.8 0.220 23.4
Peanut 20 14:10 1.1 0.02 19.7
Peanut 30 14:10 2.3 0.02 15.6


RO is the net reproductive rate; rm is the intrinsic rate of natural increase; T is the mean
generation time; re is the capacity for increase; and Tc is the cohort generation time; L:D
is hours of light and dark per day. Note that re and Tc are approximations of rm and T.
* Modified from table 7.1 in Kirk (1997b)










Table 2- 4: Known tospoviruses and thrips vectors in the world *

Virus Thrips vector

Frankliniella bispinosa
F. fusca
F. intosa
Tomato Spotted Wilt Virus F. occidentalis
F schultzei
Thrips setosus
T. tabaci

F. occidentalis
Impatiens Necrotic Spot F schultzei
F. intosa

Zucchini Lethal Chlorosis F zucchini

Watermelon Bud Necrosis T. palmi

Watermelon Silver Mottle T. palmi

Melon Yellow Spot T. palmi

Capsicum Chlorosis Ceratothrips claratis

F. occidentalis
Groundnut Ringspot F schultzei
F. intosa

F. intosa
Tomato Chlorotic Spot F. occidentalis
F schultzei

Peanut Chlorotic Fan-spot Scirtothrips dorsalis

Groundnut Bud Necrosis S. dorsalis

Peanut Yellow Spot S. dorsalis

Iris Yellow Spot T. tabaci

F. occidentalis
Chrysanthemum Stem Necrosis F cul .
F schultzei

Physalis Severe Mottle Not described

Table modified from Naidu et al. (2005) and Ullman (2005).









Table 2- 5: Reduced-risk, biopesticides and OP alternative insecticides registered or
pending registration for use in blueberries*.


Chemical name


Acetamiprid


Azadirachtin


Bacillus
thuringensis

Cinnamaldehyde


Flonicamid 1


Imidacloprid


Indoxacarb

Metaflumizone

Methoxyfenozide


Novaluron 1


Spinosad 1


Thiamethoxam


Trade name
Assail 70 WP
Adjust
Neemix
Niblecidine

Dipel
Cinnacure
Cinnamite
Carbine 50 WG
Beleaf


Admire
Provado
Gaucho I
Avaunt
Steward
BAS 320 I
Intrepid
Runner
Diamond
Rimon
Success
Spintor
Entrust
Actara
Platinum
Centric
Cruiser
Helix


Chemistry


Chloronicotinyl

Extract from neem
oil

Bacteria

Natural product


Nicotinamide


Chloronicotinyl


Oxadiazine


Semicarbazone

Diacylhydrazine


Benzoylphenyl


Macrocyclic
lactone


Second generation
Neonicotinoid


Status


Pending


Registered

Registered

Registered


Classification

Reduced-risk


Biopesticide

Biopesticide

Biopesticide


Potential OP alternative


Registered


OP alternative


Pending Reduced-risk
Pending OP alternative

Potential Reduced-risk
Pending Reduced-risk
Pending OP alternative

Pending Reduced-risk
Pending OP alternative


Registered


Registered


Reduced-risk
OP alternative


OP alternative


Zeta-cypermnethrin


Mustang
Mustang Maxx


Pyrethroid


Pending


OP alternative


* Table Modified from IR 4 Project (2006)
1 Indicates chemistries registered or that have the potential to control thrips














CHAPTER 3
SAMPLING TECHNIQUES AND DISPERSION OF FLOWER THRIPS IN
BLUEBERRY FIELDS

Blueberries are one of the fastest growing crops in Florida. Due to climatic

conditions and the early-season varieties produced, Florida and most of southern Georgia

blueberries mature during April and May, making these two states the principal producers

in the world during this time. This window of production gives the growers a price

advantage of 3 to 5 USD per pound compared to regular season blueberries produced

between the end of May and August (NASS-USDA 2006a).

Thrips have been identified by blueberry growers as insect pests that require

immediate management (Finn 2003). Twenty five percent of blueberry growers from

southern Georgia and Florida identified flower thrips as one of their main problems

surpassed only by blueberry bud mite, Acalitus vaccinii (Keifer). Other major pests of

concern for the growers include cranberry fruitworm, Acrobasis vaccinii Riley, and

blueberry gall midge, Dasineura oxycoccana (Johnson) (Finn 2003). For these reasons

the Small Fruit and Vegetable IPM Laboratory at the University of Florida started a

project to understand the relationship between blueberries and flower-thrips as an initial

step to develop an IPM program for thrips in early-season blueberries. Finn (2003)

initiated preliminary work in sampling techniques for flower thrips in blueberries. As a

common practice, flower thrips have been monitored by using sticky traps of various

colors. The two colors most commonly used are yellow and blue (Diraviam and

Uthamasamy 1992, Cho et al. 1995, Hoddle et al. 2002, Finn 2003). Finn (2003) found









non significant differences between the numbers of thrips captured in yellow, blue, or

white sticky cards in blueberry plantings. Due to the contrast between the thrips and the

white background on the sticky cards, I decided to use white sticky cards for monitoring

populations in my experiments. Because there is little information about the relationship

between flower thrips and blueberries, I must determine some of the basic characteristics

of this relationship. One of the basic characteristics I are looking at is dispersion.

Ordinarily there are three types of dispersion: random, uniform, and clumped.

Distribution depends on the mobility of the insects. Highly mobile insects tend to have a

more random distribution than insects with low mobility, which tend to form "hot-spots"

in highly clumped populations (Flint and Gouveira 2001).

My objectives were to select an efficient and effective sampling method to monitor

flower thrips inside blueberry flowers, and to determine the vertical and horizontal

distribution of thrips populations in blueberry plantings. My final goal was to

characterize thrips populations depending on their level of aggregation in blueberry

fields.

Materials and Methods

To address my objectives, I have conducted a series of experiments on private

blueberry farms in Florida and southern Georgia.

Methodology to Determine Thrips Population Inside Blueberry Flowers

Due to the high volume of flower samples, a more efficient system to determine the

number of thrips inside the flowers was developed by combining procedures from various

researchers (Finn 2003, Funderburk and Stavisky 2004). Flower clusters were collected

using Corning 50 ml plastic tubes (Fisher Scientific, Pittsburg, PA). Flowers were

collected in the field by cutting the pedicels using the rim of the vial and allowing them









drop inside the plastic tubes containing 70% ethanol. Each vial was manually shaken for

approximately one minute. The contents were emptied into a 300 ml white polyethylene

jar (B & A Products, Ltd. Co., Bunch, OK) and filtered through a plastic screen with 6.3

x 6.3 mm openings to ensure that thrips pass through, leaving the flowers on the screen.

The remains left on the screen were rinsed with water from a polyethylene wash bottle

into a white container. The flowers left on the screen were placed in a 300 ml white

polyethylene jar while the corollas and the calyxes of the flowers were manually

separated. This procedure was repeated three times to ensure the collection of the

maximum number of thrips. After each rinsing, the thrips found in the rinsing water were

collected inside a white container and counted. This water was then transferred to another

container with black background to ensure I collected the maximum number of thrips.

To determine the efficiency of this system, I decided to compare the results

obtained with this method with standard flower dissection, which is the method

commonly used in the laboratory to determine thrips populations inside flowers (Finn

2003). I selected 20 samples from the weeks that had the highest number of thrips. After

following the 'shake and rinse' procedure, samples were dissected and observed under a

microscope to determine how many thrips were missed. I then used a t-test to compare

the total number of thrips collected using the shaking and rinsing procedure with the total

number obtained by shaking and rinsing plus the number of thrips found under the

microscope by dissecting the samples after this procedure (SAS Institute Inc. 2002).

Vertical Distribution of Flower Thrips in Blueberry Fields

To determine thrips distribution within blueberry bushes, I placed ten sampling

stations in each of my two farms. The first was located on farm FL01 in south-central

Florida (N 280 04' W 81 34'). This farm was planted with Southern highbush. A second









farm, Farm GA01, located in southern Georgia (N 31 31' W 820 27'), was planted with

rabbiteye blueberries. Samples were taken during the 2004 and 2005 flowering seasons.

To collect the samples I randomly established 10 sampling stations around the selected

plot in each farm. Each sample station consisted of a blueberry bush where I took four

samples: three white sticky traps (23 x 17 cm of effective area) (Great Lakes IPM

Vestaburg, MI) and one flower sample. One of the traps was placed on the ground in an

inverted V shape with the sticky surface towards the ground, a second trap was located

inside the canopy approximately in the middle of the bush, and the third one was

approximately 40 cm above the canopy. The number of thrips in the sticky traps was

determined by counting the number of thrips in 16 out of the 63 squares (each square is

6.45 cm2) that the trap is divided into (2003). Flower samples were taken from the same

bush containing the sampling station and consisted of five flower clusters collected in

Corning 50 ml plastic tubes filled with 70% ethanol. I cut the flowers using the thumb

and the rim of the vial, thereby reducing the manipulation of the flowers. The flower

samples were processed using the shake and rinse method described above and the total

number ofthrips was recorded. The sampling stations were randomly placed in the field

each week by using random number tables based on the number of rows and the number

of plants in each row. The data were collected from bloom to petal-fall 2004 and 2005.

Data were analyzed using the repeated measures analysis (SAS Institute Inc. 2002).

I decided not to use the soil traps in Georgia since the results in Florida showed that the

number of thrips captured was too low for analysis. On average, 0.66 0.3 thrips per trap

per week in 2004 and 0.38 + 0.1 thrips per trap per week in 2005 were collected in the

soil samples. Data were transformed to comply with the assumptions of the analyses.









The data for 2004 and 2005 in Florida and 2005 in Georgia were transformed using the

natural logarithm of the original data plus one; for 2004 in Georgia, the transformation

used was the square root of the number of thrips captured.

Thrips Dispersion

We selected two fields in north central Florida, Farm FL02 (N 290 40' W 820 11')

and Farm FLO3 (N 290 43' W 820 08'). Both farms were planted with southern highbush

blueberries during 2005. However, during 2004 FL02 was planted half on rabbiteye and

half in southern highbush as indicated in Figure 3-3. Two grids one of 5 x 6 and the

second one of 8 x 7 traps were respectively deployed in each one of the selected plots.

The traps were spaced 30.48 m from each other, which covered blueberries and adjacent

non-cultivated areas. These traps were replaced every other day starting from bloom

initiation and finishing at petal fall. The total number of thrips trapped was recorded to

monitor the movement of thrips into and out of blueberry fields for two flowering

seasons. In 2004 sampling begun on March 3, while in 2005 it begun on February 20 at

both locations.

To determine degree of aggregation I selected the standardized Morisita's

coefficient of dispersion (Ip) (Smit-Gill 1975) and Green's coefficient of dispersion (Cx)

(Green 1966), because they have low or no correlation with the mean (Myers 1978,

Taylor 1984, Schexnayder Jr. et al. 2001). The aggregation indices were calculated for

each day that the sticky traps were collected. I graphed the number of thrips captured in

each trap using Sigma Plot (SYSTAT Software Inc. 2006). Once the "hot-spots" were

graphically identified, I conducted a Gaussian regression to describe the population

behavior in each one of the "hot-spots".









In this study populations were considered to be clumped if Green's index Cx > 0,

random if Cx = 0, or uniform if Cx < 0 (Myers 1978, Schexnayder Jr. et al. 2001). In the

case of standardized Morisita's index (Ip) populations were considered to be significantly

clumped (a = 0.05) if Ip > 0.5, not significantly clumped if 0.5 > Ip > 0, random if Ip = 0,

not-significantly uniform if 0 > Ip > -0.5, and significantly uniform if Ip < 0.5 (Smit-

Gill 1975). Overall comparisons were conducted by averaging all the indices calculated

and comparing them to 0 for Cx, and 0.5 for Ip using a t-test (SAS Institute Inc. 2002).



S2
s-

q xx, -1)] 1- 1 Equation 3- 1
Ip = i Cx =
X(X 1) (q 1)

Results

Methodology to determine thrips population inside blueberry flowers

I found no significant differences between the dissecting (35.7 4.3) and the shake

and rinse (34.7 4.3) methods (t = 0.17; df 1, 38; P = 0.869), when comparing the

number of thrips obtained with each method. These results allow us to use the shake and

rinse procedure in the following experiments with confidence in the data collected.

Vertical Distribution of Flower Thrips

The thrips distribution within the bushes follows the same pattern independent of

the year and the location. In Florida, there are no significant differences between 2004

and 2005 when the same positions within the bush were compared. For soil (t = 0.531; df

= 1, 325; P = 0.595), for flowers (t 1.474; df= 1, 325; P = 0.142), for the traps in the

bushes (t = 0.308; df= 1, 325; P = 0.7582) and for the traps above the canopy (t = 0.438;

df= 1, 325; P = 0.662). However, when comparing the treatments within each one of the









years, I found significant differences among the positions with respect to the bush (For

2004, F= 291.13; df= 3, 157; P < 0.0001, and for 2005 F = 197.51; df= 3, 164; P<

0.0001). In both years, the number of thrips captured was significantly higher within the

canopy compared with all other positions evaluated. The second highest number of thrips

captured was found in the traps deployed above the canopy, followed by the number of

thrips inside the flowers, and finally by traps on top of the soil (Figure 3- 1).

Our results on farm GA01 in southern Georgia were different from the situation

presented in Florida. As in Florida, I found no significant differences between the

number of thrips captured in the flowers between 2004 and 2005 (t = 0.682; df= 1, 144;

P = 0.496) in Georgia. However, I found that in 2004 I captured significantly more thrips

within the canopy (t = 7.345; df= 1, 144; P <0.0001), and above the canopy (t = 8.563;

df= 1, 144; P < 0.0001) than in 2005. When comparing the number of thrips captured at

the various positions, I found no significant differences between the number of thrips

within the canopy and above the canopy during 2004. However, these values were both

significantly higher than the number of thrips captured in the flowers sampled during the

same year. In 2005, there was a reduction in the number of thrips captured on farm GA01

compared with 2004 (

Figure 3- 2). However, despite the reduced numbers, I found a significantly higher

number of thrips captured in the canopy than above the canopy of the bushes. Both of

which were significantly higher than the number of thrips captured in the flowers (

Figure 3- 2).









Thrips Dispersion

2004 farm FL02

Thrips aggregation increased over time and peaked 12 to 14 d after bloom initiation. This

peak coincides with the highest population density of thrips 14.7 d after bloom initiation

(Table 3- 1 and Figure 3- 5). Table 3- 1 shows that thrips population can be considered

clumped from day 4 based on Cx. This observation is reinforced by Ip, which shows a

significant level of aggregation from the beginning.

After recording a clumped-type distribution, I plotted thrips population and the

coordinates where the traps were located to determine the position and number of "hot-

spots" based on the locations presented in Figure 3- 3. During the 2004 field-season I

found only one "hot-spot" located at the coordinate (4, 4) in Figure 3- 6. When analyzing

this "hot-spot" during 2004, I found that the dynamics of thrips population could be

described by a Gaussian non-linear regression (Equation 3-2 a.). The pattern for the "hot-

spot" on farm FL02 in 2004 is described by the equation represented in Equation 3-2 b.

Overall Cx (0.467 0.147) is significantly higher than 0 (t = 3.17; df= 8; P =

0.013), and Ip (0.521 0.004) is significantly higher than 0.5 (t = 7.48; df= 8; P

<0.0001), which shows a significant level of aggregation of the flower thrips on farm

FL02 for 2004.


a. b.
)-0.5= ---P 0. yx-14.7 19217
f(x) e y = 194.17e 2 1


Equation 3-2:








2004 farm FL03

The distribution of traps on farm FL03 is shown in Figure 3- 4. Thrips population

for farm FLO3 was considerably lower compared with farm FL02 during 2004. However,

it appears that there are two main areas, identified using the graphic method, where thrips

tended to aggregate. Two "hot-spots" were (Figure 3- 7). One "hot-spot" was located at

coordinate (0, 4) and a second at coordinate (2, 2) of Figure 3- 7. The peak population

for these "hot-spots" occurred on different days. For the spot found at (2, 2) the peak

occurs at 11.3 d after bloom and for the "hot-spot" located at (0, 4), aggregation occurred

17 d after bloom (Figure 3- 8 and Equation 3-3).


a. b.

3 e x-11 .3 0.5(x-17)]
-0.5 -0.5 -11
y = 5.32e 4.1 y y=4.36e[ 2.4


Equation 3-3:

Green's index (Cx) and the standardized Morisita's index (Ip) showed a tendency

towards a random distribution of thrips on farm FL03 in 2004. The overall Cx for farm

FL03 (0.046 0.041) was not significantly different from 0 (t = 1.14; df= 6; P = 0.298),

and the Ip value (0.020 + 0.006) was significantly higher than 0 (t = 3.21; df= 6; P =

0.0184) but still in the region -0.5< Ip < 0.5. However, this aggregation appears not to

be significant on farm FL03 in 2004. The distribution appears to be more aggregated for

days 7 to 15, which again coincides with the peak of the population (Figure 3- 8).

2005 farm FL02

The same set up used in 2004 was used in 2005 to describe the dispersion of flower

thrips in the field. During this year the farmer replaced the rabbiteye blueberries with a









new planting of southern highbush. Thrips population on farm FL02 was lower in 2005

than in 2004. During this year I found two "hot-spots" located at coordinates (2, 3) and

(5, 2) in Figure 3- 9. The highest aggregation was between days 10 and 14, which again

coincides with the days of maximum population (Table 3- 3 and Figure 3- 10). The "hot-

spots" reached their maximum population at 13.8 d after bloom initiation for the

coordinate (2, 3) and 12.1 d for coordinate (5, 2) ( Figure 3- 10 and Equation 3- 4).

The overall indices show a highly significant aggregation for 2005. Green's index,

Cx = 0.24 0.06, was significantly greater than 0 (t = 3.87; df= 9; P = 0.0047), and the

overall Standardized Morisita's index, Ip = 0.52 0.01, was significantly greater than 0.5

(t= 4.94; df= 9; P= 0.0011).



a. b.
0.5 x-13.782 -0.5 x-12.05 )
y 63.49e 3.44 y = 58e[ 22.67


Equation 3-4:

2005 farm FLO3

Thrips population for this farm was too low to make a robust analysis. The highest

number of insects captured in one trap was three thrips 15 d after bloom initiation. Most

of the other traps captured no thrips, and the data were not considered significant.

Discussion

The literature has discussed several types of sampling methods for thrips inside

flowers. Finn (2003) mentions the use of alcohol dipping, tapping the floral clusters on a

white surface, and flower dissection as methods to determine the number of thrips inside

the flowers. Finn's study showed no significant differences in the number of thrips









captured among the various methods for southern highbush blueberries. However, the

Finn (2003) found that sampling by tapping the flowers in rabbiteye blueberries resulted

in significantly fewer thrips captured than did other treatments. A large variation was

observed probably due to the distribution of the thrips in blueberry fields. High

aggregation and random sampling usually produces high variance in the results. Palumbo

(2003) compared trapping at canopy level, plant beating, direct observations and whole

plant washes. Whole plant washes (very similar to the shake and rinse method used in

this study) were used as absolute samples. Palumbo (2003) found a significantly higher

number of thrips in his absolute method when compared with the other methods used.

The fact that I found no significant differences between the shake and rinse method and

the dissection method, which is considered to be an absolute count of the thrips in the

flowers (Hollingsworth et al. 2002), indicates that the shake and rinse method is

appropriate to estimate thrips population inside the flowers. The shake and rinse method

might be too time-consuming and not very useful for growers who need to determine the

population rapidly and accurately, but it might be useful for research purposes, as it is as

accurate and less time-consuming than flower dissections.

The vertical distribution of thrips remained the same independent of location and

year. The highest number of thrips was consistently captured in or above the canopy of

the blueberry bushes using sticky traps. However, the most damaging population is inside

the flowers (Arevalo and Liburd unpublished data).

The average number of thrips captured varies from year to year. In 2005, thrips

population was lower on farm FL01 located in Florida than the farm GA01 located in









Georgia. However, only in GA01 were the differences in the number of thrips captured

between the years significant, having a lower population in 2005 than in 2004.

The analysis of the distribution based on Morisita's and Green's indices described

the distribution of thrips in the field as aggregated. However, the level of aggregation

seems to be lower in cases when peak populations are lower. For instance, on farm FL03

in 2004 (peak population = 5.3 thrips per trap) the average Cx (0.046 0.041) and

average Ip (0.02 0.006) were lower than for farm FL02 in 2005 (peak population =

63.49 thrips per trap) Cx (0.24 0.06) and Ip (0.52 0.01) and lower than Farm FL02

(Peak population = 194.1 thrips per trap), Cx (0.467 0.147) and Ip (0.521 + 0.004). In

Figures 3-5 and 3-11, I observed that the "hot-spots" start forming between 7 and 10 days

after bloom initiation and in both cases the population at these spots grew beyond 20

thrips per trap every two days. After this initial period, thrips population captured in the

traps increased exponentially, reaching a maximum population between 12 and 15 days

after bloom initiation. The population then declined at the same rate that it increased,

virtually disappearing about 22 days after bloom started, after most of the flowers had

become fruits.

Apparently, formation of "hot-spots" on blueberry farms is random. I did not find

any correlation among the locations where "hot-spots" were formed between the years.

However, several variables such as flower concentration, soil type, fertilization methods,

and wind direction, might be studied to determine a correlation among these variables

and "hot-spot" locations to create a predictive dispersion model of flower thrips on

blueberry farms. For now, sampling methods that consider highly aggregated

populations should be explored to reduce the variability of the data when sampling flower









thrips in blueberry farms (Southwood 1989, Wang and Shipp 2001). Despite that until

now no sex pheromones have been isolated, some behavioral observations show the

presence of a mating or an aggregation pheromone in thrips (Milne et al. 2002, Kirk and

Hamilton 2004). Kirk and Hamilton (2004) showed how virgin females are attracted to

the smell of males and not to the smell of other females. This situation was interpreted as

the presence of some type of sex pheromone in Frankliniella occidentalis (Pergande).

Milne at al. (2002) found a high correlation between the number of males in hibiscus

flowers and the number of females landing on these flowers. Salguero-Navas et. al (1994)

found indications of aggregation in tomato plants for virus thrips species including F.

occidentalis, and F. tritici. However, differences in the degree of aggregation were also

found between years in their experiment as it was also found in my trials. Thrips

populations were demonstrated to be variable within the same region. For instance, farms

FL02 and FLO3 are 7.02 km from each other and the peak populations are significantly

different, 194.2 for FL02 and 5.3 for farm FLO3. However, there appears to be a

correlation between the population dynamics and the year; in my observations, 2005 had

a lower number of thrips captured than 2004 in all farms, and studies to determine the

reasons for this pattern are needed.









Tables and Figures

Table 3- 1: Distribution indices, Green's index (Cx) and Standardized Morisita's index
(Ip), used to describe the level of aggregation of thrips population on farm
FL02 in Florida in 2004


Index


Days after blooming

2 4 6 8 10 12 14 18 22


Cx -0.005 0.438 0.223 0.138 0.227 0.967

ID 0.518 0.549 0.525 0.517 0.524 0.543


1.336 0.207 0.670

0.539 0.520 0.538


Table 3- 2: Distribution indices, Green's index (Cx) and Standardized Morisita's index
(Ip), used to describe the level of aggregation of thrips population on farm
FL03 in Florida in 2004.


Index


Days after blooming

2 4 6 8 14 16 19


-0.011


Ip 0.041


-0.065 -0.013 0.129 0.062 0.246 -0.021

0.042 0.024 0.013 0.006 0.014 -0.0002














Table 3- 3: Distribution indices, Green's index (Cx) and Standardized Morisita's index
(Ip), used to describe the level of aggregation of thrips population on farm
FL02 in Florida in 2005.


Index


Days after blooming

2 4 6 8 10 14 16 18 22


Cx 0.124 0.098 0.142 0.185 0.314 0.713 0.320 0.233 0.112

Ip 0.051 0.513 0.518 0.522 0.528 0.556 0.529 0.525 0.511










- 90.
80
70-
Q 60-
o 50-
40-
8 30-
O
< 20-
10.


102004
02005


S d I I I I I


Flowers


Above the canopy In the canopy


Location in the sampling station
Figure 3- 1: Vertical distribution of thrips captured with respect to southern highbush
blueberry bushes in south Florida. Different letters represent significant
differences among the groups using LSD mean separation test, a = 0.05.


02004
D2005 A




B


Flowers In the canopy Above the canopy
Location in the sampling station


Figure 3- 2: Vertical distribution of thrips captured with respect to rabbiteye blueberry
bushes in southern Georgia. Different letters in capitals represent significant
differences among the groups in 2004, small letters represent significant
differences among groups in 2005 using LSD mean separation test a = 0.05.

































Figure 3- 3: Map of farm FL02 located at N 280 04' W 81 34' in north central Florida
















Farm FLO3


U



U


U

r


.1- 4- L


U





* ~1Ii~ ~\.


U


U U


I U


* U


* U


Legend
a Trap
- Fence
Highbush New
Highbush Old
Grass area

S4 nursery 8
( 29 43 WM2 08)


Figure 3- 4: Map of farm FLO3 located at N 280 04' W 81 34' in north central Florida




250 -


-o 200 -


8 150-


100-
0



S0o


0 5 10 15 20 25

Days after blooming


Figure 3- 5: Population dynamics inside the "hot-spot" in coordinates (4, 4) of Figure 3- 6

for 2004 on farm FL02.










59










a.

160
140 0
120 20
40
M60
100 80
1000
80" 1 20
S140
S 60 160
Zo 40
20


41 6

2 34

1 1 2 9




















1,,,-
S120














z2
1 40
460

































an 22 (h"ay ,ftrboombga n am L
1 -,80

0 20
S140
o m 160




0
7
4 6

3

1 0 2







mm C.











57
4 6

2 34
O1 0 2









Figure 3- 6: Number ofthrips captured at 2 (a), 6 (b), 8 (c), 10 (d), 14 (e), 16 (f), 18 (g),

and 22 (h) days after bloom began on farm FL02.





























































































0


Figure 3-6: Continued


M0
20
40
60
80
100
120
140
160


M0
20
40
60
80
m 100
m 120
140
160





7,


0
20
40
60
80
m 100
120
140
160


2 .,0 g g


2























0O
20
40
60
80
100
120
140
160


-0
20
40
60
80
m 100
m 120
140
160


Figure 3-6: Continued







62




a.





















b.
55
4
3




3 5
4













































and 20 (f), days after bloom in 2004
\0




6 C.
4 ? iO

Figr 3-7 ubrotrp atre nfr L3a1 a,4 b,8() 4()6()
an 0(),dyftrbomin200






63




d.



0 ,1














a%-3







-mo


01
zv.. ,
I* ~


v^5


Figure 3-7: Continued









64





6 -

0 Hot spot in (2, 2)
5 O Hot spot in (0, 4) *
-- Regression hot spot (2, 2)
Regression hot spot (0, 4)
4 O

ou 3 0 \
U)

2 0 o \
0 *

0-
Z 1 0


0 G a a-- -



0 5 10 15 20
Days after bloom


Figure 3- 8: Population dynamics inside the "hot-spots" in coordinates (2, 2), and (0, 4)

of Figure 3- 7 in 2004 on the farm FL03 in Florida.






















-O
S20
S40
S60


40
S20
m 40
m60


7>7


Figure 3- 9: Number of thrips captured on farm FL02 at 2 (a), 4 (b), 8 (c), 10 (d), 14 (e),
16 (f), 18 (g), and 22 (h) days after bloom in 2005.








66










-- d.



/ :--
m-R








4 6











2 2r 20
0 m 40
0 60
e.







7
4 36


2 3 4 5










0
Fr f.











4 6

2 3
1 0 2 -9 V,



Figure 3-9: Continued












































Figure 3-9: Continued






















0

0
S 40 O. 0
L- \



z
20

O 0


0 5 10 15 20 25
Days after bloom


Figure 3- 10: Population dynamics inside the "hot-spot" in coordinates (2, 3), and (5, 2)

of Figure 3- 9 on farm FL02 in Florida in 2005.














CHAPTER 4
PEST PHENOLOGY AND SPECIES ASSEMBLAGE OF FLOWER THRIPS IN
FLORIDA AND SOUTHERN GEORGIA IN EARLY-SEASON BLUEBERRIES



The relationship between blueberries and flower thrips has not been studied

because blueberries are a relatively new crop for Florida. Characterizations of thrips

assemblage have been conducted in citrus, tomatoes, mangoes, and other crops. In north

Florida tomatoes, Reitz (2002) found that the most commonly encountered species is

Frankliniella occidentalis (Pergande), but during the spring and the fall, F. tritici (Fitch)

was the most abundant. Other thrips species found in north Florida tomatoes include F.

bispinosa (Morgan) and F. fusca (Hinds). Discussing his findings, Reitz (2002)

emphasized the importance of studying individual species populations of thrips when

developing a new sampling protocol or a management program for specific crops due to

their differences in behavior, damage, and importance.

To understand the phenology of a pest, it is necessary to understand its relationship

with its host plants. Plants can be divided into two groups depending on their relationship

with thrips. The first type is a provisional or alternate host, which offers temporary

shelter or food, but thrips do not reproduce in these plants. The second type is called a

proper host because it offers food, shelter, a reproductive substrate, and alimentation for

immature thrips (Mound 2005). Of the 5,300 thrips species worldwide, only 50 are

recognized as economically important pest species. These species considered as crop

pests are prolific and non-host-specific. For instance, F. occidentalis, western flower









thrips, has been reported on more than 500 plant species within 50 families (Moritz et al.

2004b). In this study, I described the phenology of flower thrips in relation to plantings

across the early-season blueberry production regions of Florida and southern Georgia. I

described the main species collected during the trials to facilitate the identification of

these species. These results are part of the thrips management strategy that the Small

Fruit and Vegetable IPM Laboratory is trying to develop as part of an IPM program for

blueberries.

Materials and Methods

Two farms in Florida planted with southern highbush blueberries, SFL01 located in

south-central Florida (N 280 04' W81 35'), and NCFL01 located in north central Florida

(N 290 41' W 82 11'), as well as one farm in southern Georgia, SGA01 (N 31 32' W

82 28'), which was planted with rabbiteye blueberries, were selected to conduct the

trials. These farms were sampled during the 2004 and 2005 blueberry flowering season to

monitor thrips activity. In each of these farms, I randomly placed 10 white sticky traps

(Great Lakes IPM, Vestaburg, MI) in one hectare of blueberries. The traps were

collected weekly from flower opening to fruit set. I also collected five flower-clusters

from the same bushes where traps were deployed. The traps and the flower samples were

processed at the Small Fruit and Vegetable IPM Laboratory at the University of Florida.

The number of thrips captured in the sticky traps was counted using a 20x magnifying

glass, and the thrips inside the flowers were extracted using the shake and rinse procedure

described in Chapter 3. I used the Pearson correlation coefficient to quantify the

relationship between the number of thrips captured in flowers and sticky traps and the

observed percentage of opened flowers in the field. I quantified the relationship between

the dates of first thrips capture and dates of maximum capture with the latitude at which









these farms were located. To be able to correlate the dates, I transformed dates into a

numerical system and used the general format of dates in Microsoft-Excel 2000 in

which January 1, 1900 corresponds to 1, January 2, 1900 corresponds to 2 and so on,

increasing by one with each day. The ten traps and ten flower samples per week were

averaged to determine the population in the field. The results were graphed to determine

any trends and show correlations.

A sample of 100 thrips per week from sticky traps was randomly collected to

determine the thrips species assemblage present at each of the sampling sites. In the case

of the flowers, I used as many thrips as I could extract from the five flower-cluster

samples, to a maximum of 100 per week. To detach the thrips from the sticky traps I

submerged the traps in 500 ml of CitroSolvTM (Fisher Scientific, Pittsburgh, PA) for four

days. I used a squirt-bottle containing CitroSolvTM to rinse the insects that were still

attached to the trap. The CitroSolvTM along with the thrips were then filtered through a

basket-style coffee filter (Publix Supermarkets. Lakeland, FL). The insects in the filter

were placed in a Quilted Crystal Jelly Jar (Jarden Corporation, Muncie, IN) containing

CitroSolvTM for five more days to dissolve the remaining glue from the traps. Thrips

were allowed to air-dry at room temperature and then re-hydrated using deionized water.

The thrips collected from the flowers were preserved in 50% alcohol until I was able to

slide- mount them for identification. The thrips from both traps and flowers were

individually mounted on microscope slides using CMC-10 media (Masters Chemical

Company, Elk Grove, IL). Vouchers of these specimens were sent to the Florida State

Collection of Arthropods (FSCA) in Gainesville, FL. The thrips collected were divided

into mature and immature. Mature thrips were then identified to species using taxonomic









keys (Mound and Marullo 1996, Moritz et al. 2001, Moritz et al. 2004b, Edwards

Unpublished data). The percentage of thrips of each species was tabulated for

comparison. A short key of the most common flower thrips species present in early-

season blueberries in Florida and Georgia was constructed.

Results

Pest phenology:

Flower thrips populations in blueberry plantings were highly correlated with the

latitude and percentage of opened flowers. The Pearson correlation coefficient for the

relationship between the latitude and the date of first capture was 0.971 for the 2004

season and 0.957 for the 2005 season. I also determined the same coefficient for the

relationship between the latitude and the date when the maximum population was

recorded. The coefficients for this relationship were 0.999 for the 2004 season and 0.955

for the 2005 season. Independent of the latitude, flower thrips were first captured when

10 15 % of the flowers in the field were opened (Figures 4-1 to 4-6). Thrips populations

peaked when 80 to 90 % of the flowers were opened. The population began to decline

once the fruits started to form and the petals fell, leaving a reduced number of opened

flowers in the field.

For the farm in south-central Florida SFL01, the first thrips were recorded between

February 11, 2004; three weeks later (March 3, 2004) thrips populations reached their

highest on sticky traps and flowers (Figure 4-1). However, for 2005 the first capture was

registered on February 15, on March 8, the population reaches its highest point. The

highest populations of thrips inside the flowers were recorded only two weeks after their

first capture on the sticky traps (Figure 4- 2). The Pearson correlation coefficient between

the average number of thrips captured in flowers and the sticky traps with respect to the









percentage of opened flowers was very high for both years (Table 4-1). An exception

occurred for flowers in 2005 when it appeared that the highest population was reached a

week earlier in the flowers (March 2, 2005) than in the traps (March 9, 2005).

The observations in north-central Florida follow the same trend as for south-central

Florida. The first thrips captured on sticky traps in NCF01 was recorded, on average, 22

days after the observations in south Florida. The peak population was recorded two

weeks after the first captures in both years (Figure 4-3 and 4-4). After thrips population

reached its peak, the number of thrips captured declined with the decreasing number of

opened flowers due to fruit formation. Correlation coefficients are high for sticky traps

and flowers in both years in relation to the percentage of opened flowers (Table 4-1).

In southern Georgia, SGA01, the first thrips were captured between March 14th

and 17th, on average 9 days after NCFL01 and 31 days after SFL01. Maximum

populations during both years were very different from those in Florida. In the 2004

season, the maximum population was reached three weeks after first detection and

averaged 255.4 43.6 thrips per trap per week (Figure 4- 5). During the 2005 season, the

maximum number of thrips captured on sticky traps per week reached 22.0 + 3.3 two

weeks after the first detection (Figure 4- 6). Despite these differences, Pearson's

coefficients measuring the correlation between the percentage of opened flowers and the

number of thrips captured in the traps and inside the flowers were high (Table 4-1).

A summary of the main variables observed at the various sites is presented in Table

4- 2. This table shows the high variability in the number of thrips captured among the

various years and sites. For example, for south-central Florida the maximum population

captured in traps was 65.4 14.7 thrips per sticky trap per week while in 2005 the









amount of thrips captured with the same method was almost double, 123.4 + 29.7 thrips

per trap per week. Due to the variability in the populations and the distribution of thrips

(Chapter 3), mean comparisons were not conducted since there was no homogeneity

among the observed sites. At the same time Table 4- 2 and Figure 4- 7 show the

correlation between latitude and dates of first capture and of maximum capture.

Species assemblage:

I repeatedly collected four species of thrips in Florida blueberry fields. These

species were F. bispinosa, F. fusca, F. occidentalis, and Thrips hawaiiensis (Morgan).

Other species were also recorded in Florida: Haplothrips victoriensis Bagnall, F. kelliae

Salimura, and T. pini Uzel. The species with the highest number of individuals was F.

bispinosa in the flowers as well as in the sticky traps deployed in Florida (Table 4-3).

The thrips species assemblage in Georgia is different from the one in Florida. The

predominant species is F. tritici, followed by F. occidentalis and finally T. pini. These

species were found in flowers as well as in sticky traps during the two years that the

sampling was conducted (Table 4- 3). There is no appreciable difference in the species

assemblage between the samples taken in 2004 and 2005 at the various sites. Between

14.75 and 27.8 % of the thrips recorded inside the flowers are immatures. The percentage

of immature thrips was highest in SGA01, followed by SFL01 and finally NCFL01. Few

immature thrips were found on sticky traps. This might be due to wind currents or

immature thrips emerging from some of the flower materials and accidentally being

caught in traps.

Frankliniella bispinosa (Morgan): This species was the most abundant in Florida.

It accounted for 78.89% (2005-NCFL01) to 88.96% (2004-NCFL01) of the adults

captured inside the blueberry flowers. Furthermore, it represented between 82.51%









(2005-SFL01) to 95.37% (2005-NCFL01) of the total number of thrips captured in sticky

traps (Table 4- 3). Frankliniella bispinosa was not captured in Georgia in flowers or in

sticky traps.

Frankliniella tritici (Fitch): This species replaced F. bispinosa in southern

Georgia (SGA01) as the most commonly encountered species in blueberry fields.

Between 60.01 % (in 2004), and 49.58 % (in 2005) of the adults captured in the flowers

belong to this species. The percentage ofF. tritici captured in traps was overwhelmingly

higher than the other species. In 2004 adults ofF. tritici accounted for 94.00 % of the

population captured in sticky traps, while in 2005 they represented 92.00 %.

Frankliniella occidentalis (Pergande): This species was found in Florida and

Georgia. It was the second most abundant species in Georgia in sticky traps and flowers

in both years, as well as in the flowers of north central Florida (NCF01) in 2005. It was

the third most abundant in south-central Florida (SFL01) and in sticky traps at the

NCFL01 site. In Georgia, the relative abundance ofF. occidentalis as percentage of the

adults captured was higher inside the flowers (36.61% for 2004 and 38.78 % for 2005)

than in the sticky traps (4.4% in 2004 and 6.0% in 2005).

Frankliniellafusca (Hinds): This species was the second most abundant species

in SFL01 and the third most abundant in NCFL01 Flowers. Only two individuals ofF.

fusca were captured in Georgia (Other species Table 4-3).

Thrips hawaiiensis (Morgan): This species was recorded only at the Florida sites.

It was captured inside the flowers as well as on the sticky traps and apparently is more

abundant in the southern regions (SFLO 1).









Rapid Determination of the Most common Species Found in Early-Season
Blueberry Fields.

Based on the information from (Mound and Marullo 1996, Moritz et al. 2001,

Moritz et al. 2004b, Edwards Unpublished data ) and from personal observations a key is

presented that can be used with a 40x compound microscope for the most commonly

encountered species of thrips in blueberry fields.

Identification key to the flower thrips of blueberries in Florida and Georgia

1. Four or five pairs of elongated pronotal setae (Figure 1). Abdominal sternite VII
has no discal setae. The first-vein setal-row in the forewing is complete and the setae are
uniformly spaced (Figure 2).................................. ........... (Frankliniella spp.) 2

















Figure 1 Figure 2

1'. Three or fewer pairs of elongated pronotal setae. Abdominal sternite VII has
discal setae. The first vein setal row in the forewing is incomplete (Figure
3 ) ........... ...................................................................................................(T h r ip s sp p .) 5






77

















Figure 3

2. The postero-marginal comb of microtrichia is complete in the middle. The
microtrichia are long and irregular and their bases are broadly triangular (Figure 4).
Major post-ocular seta are more than 1/2 of the length of ocellar setae III, and usually
extending clearly to the outside of the head (Figure 5).................... F. occidentalis


Figure 4


'V


I'
13-


Figure 5







78


2'. With a different combination of characters from the ones described above....... 3

3. Base of the first antennomere restricted (Figure 6). The reticulation on the
metanotum media area is equiangular. No post-ocular seta I. Wings might be absent in
the adult stage ................................................................. ........... F fusca


Figure 6


3'. Base of the first antennomere is swollen (Figures 7 and 8). The metanotum has no
equiangular reticulations, but irregular longitudinal ones. Post-ocular seta I is present and
adults always have wings ............. ....... .................................... 4

4. Base of the first antennomere is swollen and the edges are more or less sharp
(Figure 7). It presents two well developed and sclerotised setae in the second antennal
segment. This species is the most abundant in Florida, and very rarely found in
G eorgia................................................................... ............ F. bispinosa


Figure 7


/-4


.'y-
*"- ^,






79


4' Base of the first antennomere is swollen but the edges are not sharp. Setae are less
developed and less sclerotized on antennal segment II than in F. bispinosa (Figure 8).
This species is very common in blueberry plantings in Georgia ...........F. tritici


Figure 8

5. Seven or eight antennal segments. The postero-marginal microtrichia are short
and irregular in length, they appeared to have their bases fused or more than one
microtrichia per base (Figure 9). Sternite V has between 10 and 13 discal setae (Figure
10). Has no equiangular reticulation on the metanotum median area ........T. hawaiiensis


Figure 10


x2


- 'p


~2 A -
Q st. -
I


Figure 9









5'. Always with eight antennal segments. The postero-marginal microtrichia are long
slender and irregular. Their bases are broad and clearly not fused at the base (Figure 11).
Sternite V has between 3 and 9 discal setae (Figure 12). The metanotum median area
presents some equiangular reticulation ............ ................................ pini

















Figure 11 Figure 12



Discussion

This is the first time that a description of the thrips species assemblage and the

phenology of flower thrips in relation to early-season blueberries have been investigated.

Previous research on the relationship between flower thrips and blueberries has been

limited to adequate monitoring techniques (Finn 2003). It is clear now that the presence

of flower thrips on blueberry fields is highly correlated with the latitude and with the

percentage of opened flowers in the field. Thrips were captured for the first time when

close to 10 to 15% of the flowers were opened, independent of the latitude. Flowers and

thrips appeared later in the season for northern sites compared with those recorded in

southern areas. Despite the differences in the dates of blooming among the various sites,

thrips populations followed the same pattern with respect to flower opening. The number

of thrips captured inside the flowers, as well as the number of thrips captured in sticky









traps increased up to the point where the maximum number of flowers is opened in the

field. After that time, thrips populations started to decrease when the petals of the flowers

began to fall and the fruits start forming. This correlation between thrips populations with

flower phenology as well as with latitude indicates that it might be a correlation between

thrips populations and degree-days accumulation. To determine this correlation it will be

necessary to establish a 'base-date' or conditions to begin the accumulation of the degree-

days. Future research could be conducted to establish a more accurate model to predict

thrips populations in blueberries based on weather patterns and temperature.

Among the thrips captured in Florida, F. bispinosa is the most commonly

encountered species. This results complement the species descriptions made in citrus

where 80 to 95% of the thrips captured belong to this species (Childers et al. 1990,

Childers et al. 1994, Toapanta et al. 1996, Childers et al. 1998). In Georgia, I did not

capture F. bispinosa. I found the most common species to be F. tritici followed by F.

occidentalis. This difference might be related to environmental conditions. Other species

found in the samples include: F. fusca, T. hawaiiensis, T. pini. A taxonomic key to

distinguish the species commonly found in blueberries was developed. This key can be

used by professionals who are trying to determine the species assemblage found in early-

season blueberries. Almost all the thrips found belong to these six species. Other thrips

encountered sporadically, fewer than 2 specimens collected in total, include Haplothrips

victoriensis, F. kelliae, and F. schultzei (Trybom).

Blueberry growers will be able to use this information to predict when thrips are

more likely to arrive to their fields, based on the percentage of opened flowers and the

latitude at which the farm is located. The duration of the flowering season in blueberries









is on average 25 days and coincides with the time that flower thrips are captured in the

fields. This period is not long enough for thrips to have multiple generations in the

blueberry flowers because the life cycle is almost as long as the flowering period in

blueberries (Childers et al. 1994, Moritz 1997). This led me to believe that most of the

adult thrips collected in the flowers and in the sticky traps migrated into blueberry

plantings from adjacent fields. Previous research indicated that the flower thrips species

that were collected in blueberry fields are also found in citrus, wheat, and non-crop plants

such as hairy vetch (Vicia villosa Roth), and crimson clover (Trifolium incarnatum L.)

during both winter and spring (Toapanta et al. 1996). Adequate management of these

alternative hosts for the flower thrips around blueberry fields might help to reduce the

immigration of thrips into the plots.









Tables and Figures

Table 4- 1. Pearson correlation coefficients for the relationship between percentage of
opened flowers and thrips population captured in sticky traps and inside five
blueberry inflorescences.

SFL01 NCFL01 SGA01
Sample (n = 5) (n = 4) (n = 5)

2004 2005 2004 2005 2004 2005

Sticky traps 0.807 0.838 0.784 0.953 0.994 0.874

Flowers 0.918 0.052 0.780 0.840 0.986 0.868


Table 4- 2. Dates, latitude, and principal characteristics of flower thrips population in
2004 and 2005 from the samples taken from south-central Florida to southern
Georgia. SFL01 represents the farm in south Florida, NCFL01 represents the
farm located in north-central Florida, and SGA01 is the farm located in
southern GA.
Max.
Farm Date of first Latitude Date of max. population Max.
capture* ( ) population1 per 5 flower population per
clusters trap

SFL01 11-Feb-2004 N 280 04' 3-Mar-2004 16.6 4.7 65.4 + 14.7
SFL012 15-Feb-2005 N 280 04' 9-Mar-2005 123.4 + 29.7
SFL013 2-Mar-2005 8.9 2.5
NCFL01 4-Mar-2004 N 290 41' 18-Mar-2004 5.2 + 1.8 31.6 8.4
NCFL01 10-Mar-2005 N 290 41' 24-Mar-2005 23.0 + 7.5 86.5 10.5
SGA01 14-Mar-2004 N 310 32' 4-Apr-2004 25.2 + 8.3 255.5 + 43.3
SGA01 17-Mar-2005 N 310 32' 1-Apr-2005 4.5 0.4 22.3 + 3.40


1 Refers to the collection date for the traps which were placed in the field 7 days prior to
this date.
2 Information for the thrips captured on sticky traps in SGL01 in 2005
3 Information for the thrips captured in flower-clusters in SGL01 in 2005














Table 4- 3. Distribution of the thrips species assemblage in Florida and southern Georgia.


Percentage of thrips captured per season
2004 2005
Farm Species Flowers Sticky traps Flowers Sticky traps

SFL01 Immature 20.4 0.0 23.0 0.0
F. bispinosa 67.2 83.6 61.2 82.4
F. fusca 6.6 10.4 8.4 12.2
F. occidentalis 4.4 5.8 5.8 3.6
T. hawaiiensis 1.0 0.0 1.6 1.2
Other species 0.4 0.2 0.0 0.6

NCFL01 Immature 18.5 0.0 14.7 0.3
F. bispinosa 72.5 93.3 67.3 95.2
F. fusca 3.7 5.0 8.5 3.0
F. occidentalis 4.7 1.3 9.5 1.0
T. hawaiiensis 0.5 0.2 0.0 0.3
Other species 0.0 0.3 0.0 0.3

SGA01 Immature 26.8 0.0 27.8 0.2
F. tritici 44.0 94.0 35.8 92.0
F. occidentalis 26.8 4.4 28.0 6.0
T. pini 2.4 1.2 8.4 0.4
Other species 0.0 0.4 0.0 1.4




















Percentage of opened flowers
30 50


90

-z 80

70-

60
S6-

50
--
40-
-
30


20-

10-

0
10-F


eb-04


17-Feb-04


24-Feb-04
Date


2-Mar-04


9-Mar-04


Figure 4-1. Phenology of thrips population on farm SFL01 in 2004. The graph represents
average number of thrips captured per sticky trap and the average number of
thrips in five blueberry flower-clusters + SEM.


- Sticky Traps SEM
* ... Flower SEM


.-4


I- w I












Percentage of opened flowers
30 50


- Sticky Traps SEM
..... Flower SEM


- W 7


15-Feb-05


22-Feb-05


1-Mar-05
Date


8-Mar-05


-, Mar-1
15-Mar-05


Figure 4-2 Phenology of thrips population on farm SFL01 in 2005. The graph represents
average number of thrips captured per sticky trap and the average number of
thrips in five blueberry flower-clusters SEM.


Percentage of opened flowers


- Sticky trap SEM
..... Flower+ SEM









-*


10-Mar-04


17-Mar-04
Date


24-Mar-04


Figure 4- 3 Phenology of thrips population on farm NCFL01 in 2004. The graph
represents average number of thrips captured per sticky trap and the average
number of thrips captured in five blueberry flower-clusters SEM.


180

160
140

120
-4 100
80
6 80
I 60

40


45 I
40 -
35 -
30 -
25 -
20 -
15 -
10 -
5
0 -
3-Mar-


04