<%BANNER%>

Host Susceptibility and Population Dynamics of Scirtothrips Dorsalis Hood (Thysanoptera

Permanent Link: http://ufdc.ufl.edu/UFE0024628/00001

Material Information

Title: Host Susceptibility and Population Dynamics of Scirtothrips Dorsalis Hood (Thysanoptera Thripidae) on Select Ornamental Hosts in Southern Florida.
Physical Description: 1 online resource (136 p.)
Language: english
Creator: Derksen, Andrew
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: chili, diel, dispersal, dorsalis, dynamics, flight, florida, ipm, ornamental, pest, population, rose, scirtothrips, south, thripidae, thrips, thysanoptera
Entomology and Nematology -- Dissertations, Academic -- UF
Genre: Entomology and Nematology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Scirtothrips dorsalis Hood (1919) is an invasive pest thrips that recently arrived in Florida. The growth and dispersal of S. dorsalis populations are influenced by environmental factors. Populations show a positive relationship with increasing temperature, and months with greater accumulated degree-days show larger thrips populations on all hosts tested. Increasing humidity and rainfall are associated with negative population growth. Periods of peak population growth also appear to be periods of peak population dispersal. Temperature and humidity are also important to dispersal, with higher temperatures increasing thrips flights throughout the day, and higher humidity reducing the number of flying thrips. Dispersal of S. dorsalis is heavily influenced by the direction of the wind. This thrips appears capable of choosing between hosts, appearing in greater density on schefflera than buttonwood, but this may be an effect of the denser foliage of the schefflera trapping more thrips in the wind. It was the predominant thrips detected on all hosts year round, but while all life stages were detected on hosts such as schefflera and buttonwood, populations were significantly larger on rose. Population demographics of rose also suggested a growing population. When comparing different cultivars of rose, all tested appear to be exploited in a similar fashion and suffer similar damage from the thrips. Larger populations of S. dorsalis resulted in greater damage to host plants, with a reduction in blossom size by almost half. Thrips were more abundant in flowers than on leaves, but the density of thrips per square centimeter was similar between the two parts. Thrips density was fairly even between plants within a plot, suggesting that the thrips has density dependent distribution. The nutrient quality of plants also impacted thrips populations. While higher rates of fertilization did not result in significantly larger numbers or sizes of flowers, there were trends suggesting that over time it would result in significantly larger populations of thrips.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Andrew Derksen.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Mannion, Catharine M.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024628:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024628/00001

Material Information

Title: Host Susceptibility and Population Dynamics of Scirtothrips Dorsalis Hood (Thysanoptera Thripidae) on Select Ornamental Hosts in Southern Florida.
Physical Description: 1 online resource (136 p.)
Language: english
Creator: Derksen, Andrew
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: chili, diel, dispersal, dorsalis, dynamics, flight, florida, ipm, ornamental, pest, population, rose, scirtothrips, south, thripidae, thrips, thysanoptera
Entomology and Nematology -- Dissertations, Academic -- UF
Genre: Entomology and Nematology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Scirtothrips dorsalis Hood (1919) is an invasive pest thrips that recently arrived in Florida. The growth and dispersal of S. dorsalis populations are influenced by environmental factors. Populations show a positive relationship with increasing temperature, and months with greater accumulated degree-days show larger thrips populations on all hosts tested. Increasing humidity and rainfall are associated with negative population growth. Periods of peak population growth also appear to be periods of peak population dispersal. Temperature and humidity are also important to dispersal, with higher temperatures increasing thrips flights throughout the day, and higher humidity reducing the number of flying thrips. Dispersal of S. dorsalis is heavily influenced by the direction of the wind. This thrips appears capable of choosing between hosts, appearing in greater density on schefflera than buttonwood, but this may be an effect of the denser foliage of the schefflera trapping more thrips in the wind. It was the predominant thrips detected on all hosts year round, but while all life stages were detected on hosts such as schefflera and buttonwood, populations were significantly larger on rose. Population demographics of rose also suggested a growing population. When comparing different cultivars of rose, all tested appear to be exploited in a similar fashion and suffer similar damage from the thrips. Larger populations of S. dorsalis resulted in greater damage to host plants, with a reduction in blossom size by almost half. Thrips were more abundant in flowers than on leaves, but the density of thrips per square centimeter was similar between the two parts. Thrips density was fairly even between plants within a plot, suggesting that the thrips has density dependent distribution. The nutrient quality of plants also impacted thrips populations. While higher rates of fertilization did not result in significantly larger numbers or sizes of flowers, there were trends suggesting that over time it would result in significantly larger populations of thrips.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Andrew Derksen.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Mannion, Catharine M.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024628:00001


This item has the following downloads:


Full Text

PAGE 1

1 HOST SUSCEPTIBILITY AND POPULATION DYNAMICS OF Scirtothrips dorsalis HOOD (THYSANOPTERA: THRIPIDAE) ON SELECT ORNAMENTAL HOSTS IN SOUTHERN FLORIDA By ANDREW I. DERKSEN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVE RSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Andrew I. Derksen

PAGE 3

3 Humility before nature

PAGE 4

4 ACKNOWLEDGMENTS I would l ike to that a number of people without whom the completion of this project would have been impossible. I would begin by thanking my family for their unique sense of humor and their pragmatic approach to unconditional love. I would then thank my committee, including Dr. Dakshina Seal, Dr. Bruce Schaffer, and Dr. Lance Osborne for their input, advice, and enthusiasm and my advisor, Dr. Catharine Mannion, for her continued patience and support. I would like to include the staff of the Mannion Lab at TREC, Ho lly Glenn and Julio Almanza, for all their advice on growing greener plants, and for their suggestions to improve experiments. I will definitely extend my thanks to GB Edwards and the staff of DPI for identifying and confirming the species of thrips found in samples collected. I would also thank the Dade County AgriCouncil for a scholarship funding this research, and to Nelsons Roses of Florida and the Conard -Pyle Company for donating a number of the roses used in these studies. I will then extend my thank s to my friends and roommates: Justin Vitullo, Daniel Carrillo, and Luis Barquin. Thanks for knowing when and how to keep me focused and on track and for knowing when to tell me to take a break. Last and not least, I would like to thank Ms. Dana Watson f or restarting the mental engines that got me here in the first place.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 7 LIST OF FIGURES .............................................................................................................................. 8 ABSTRACT ........................................................................................................................................ 11 CHAPTER 1 LITERATURE REVIEW ........................................................................................................... 13 Expansion and Economic Impact ............................................................................................... 13 Nomenclatural and Biogeographic History ........................................................................ 13 Biological Invasi on .............................................................................................................. 14 Outbreak and Establishment ............................................................................................... 15 History of Expansion ........................................................................................................... 16 Econ omic Impact ................................................................................................................. 18 Morphology and Classification .................................................................................................. 20 Ecology and Behavior ................................................................................................................. 21 Integrated Pest Management Strategies for Scirtothrips dorsalis ............................................ 28 Detection and Sampling ...................................................................................................... 28 Mechanical Controls ............................................................................................................ 29 Cultural Controls .................................................................................................................. 30 Chemical Controls ............................................................................................................... 32 Classical Biological Contr ol ............................................................................................... 33 Conclusions ................................................................................................................................. 35 2 DISPERSAL OF Scirtothrips dorsalis HOOD (THYSANOPTERA: THRIPIDAE) FROM KNOCK OUT RADRAZZ ROSE TO TWO ALTERNATIVE ORNAMENTAL HOSTS IN SOUTHERN FLORIDA ........................................................... 38 Introduction ................................................................................................................................. 38 Materials and Methods ................................................................................................................ 40 Population Estimates on Two Hosts ................................................................................... 41 Daily Flight Behavior .......................................................................................................... 42 Analysis ................................................................................................................................ 43 Results .......................................................................................................................................... 43 Stadia .................................................................................................................................... 44 Host Plant Damage .............................................................................................................. 44 Direction ............................................................................................................................... 44 Daily Flight Behavior .......................................................................................................... 45 Discussion .................................................................................................................................... 46 Acknowledgments ....................................................................................................................... 49

PAGE 6

6 3 POPULATION DYNAMICS OF Scirtothrips dorsalis HOOD (THYSANOPTERA: THRIPIDAE) ON TWO ORNAMENTAL HOSTS IN SOUTHERN FLORIDA ................. 58 Introduction ................................................................................................................................. 58 Materials and Methods ................................................................................................................ 60 On plant Measures ............................................................................................................... 61 Flight Measures .................................................................................................................... 62 Thrips Identification ............................................................................................................ 62 Environmental Factors ........................................................................................................ 62 Analysis ................................................................................................................................ 62 Results .......................................................................................................................................... 63 On plant Measures ............................................................................................................... 63 Flight Measures .................................................................................................................... 67 Comparing Flight Measures to On-plant Estimates .......................................................... 69 Discussion .................................................................................................................................... 70 4 EFFECTS OF FERTILIZER RATE AND ROSE CULTIVAR ON POPULATIONS OF Scirtothrips dorsalis HOOD (THYSANOPTERA: THRIPIDAE) IN SOUTHERN FLORIDA .................................................................................................................................... 94 Introduction ................................................................................................................................. 94 Materials and Methods ................................................................................................................ 96 Fertilizer Rate with Radrazz ............................................................................................. 96 Fertilizer Rate with Radsunny and Radrazz Cultivars ................................................. 97 Susceptibility of Six Cultivars ............................................................................................ 98 Results ........................................................................................................................................ 100 Fertilizer Rate with Radrazz ........................................................................................... 100 Fertilizer Rate with Radsunny and Radrazz Cultivars ............................................... 100 Suscept ibility of Six Cultivars .......................................................................................... 101 Discussion .................................................................................................................................. 102 Acknowledgements ................................................................................................................... 108 5 CONCLUSIONS ....................................................................................................................... 120 LIST OF REFERENCES ................................................................................................................. 123 BIOGRAPHICAL SKETCH ........................................................................................................... 136

PAGE 7

7 LIST OF TABLES Table page 1 1 Scirtothrips dorsalis Hood (1919) was first described on chili peppers in India, and as is common with the order, has since undergone considerable synonymization ........... 37 1 2 Scirtothrips dorsalis has been implicated in the transmission of several tospoviruses. .... 37 3 1 Symptoms of feeding damage prevalent at each stage of the infestation by Scirtothrips dorsalis for Conocarpus erectus var erectus L. (green buttonwood) and Rosa Radrazz (Original Knock Out rose). ..................................................................... 77 3 2 Total thrips captures of each life stage on both host plants. ................................................ 77 3 3 Proportion of adult Scirtothrips dorsalis on each plant organ. ........................................... 77

PAGE 8

8 LIST OF FIGURES Figure page 2 1 Mean weekly S. dorsalis nymphs plus standard error on buttonwood foliage (* = significantly different by pairwise t -test comparisons at a = 0.05). .................................... 50 2 2 Weekly flight captures of S. dorsalis plus standard error on sticky cards behind Conocarpus erectus and Schefflera arboricola and weekly samples of S. dorsalis plus standard error washed from plant terminals on both hosts. ......................................... 51 2 3 Mean damage rating plus mean standard error for Conocarpus erectus and Schefflera arboricola (* = significant difference by ANOVA with a = 0.0001). ................................ 52 2 4 Mean thrips density ( S. dorsalis /cm2) plus mean standard error for Conocarpus erectus and Schefflera arboricola (* = significant difference by ANOVA at a = 0.05). ....................................................................................................................................... 53 2 5 Mean captures of S. dorsalis plus mean standard error on yellow sticky cards placed behind Conocarpus erectus and Schefflera arboricola (* = significant difference by ANOVA at a = 0.05). ............................................................................................................. 54 2 6 Me an S. dorsalis captures plus mean standard error on yellow sticky cards by direction (* = significant difference by pairwise t tests at a = 0.0001). ............................. 55 2 7 Mean hourly thrips captures plus sta ndard error for greenhouse and field sites at TREC and MREC (* = significant difference between sites determined using ANOVA at a = 0.05). ............................................................................................................. 56 2 8 Mean thrips captures by degree -hour for greenhouse and field sites at TREC and MREC. .................................................................................................................................... 57 3 1 Biweekly mean total thrips by plant organ plus standard error (organs connected by the same letter are not significantly different by Tukey -Kramer at a = 0. 05). .................. 78 3 2 Mean thrips density ( S. dorsalis /cm2) plus mean standard error by plant organ (organs connected by the same letter are not significantly different by Tukey-Kramer at a = 0. 05). ............................................................................................................................ 79 3 3 Biweekly mean thrips density ( S. dorsalis /cm2) plus mean standard error per rose flower and buttonwood sample over time. ........................................................................... 80 3 4 Mean thrips floral density ( S. dorsalis /cm2) plus mean standard error observed at each damage rating on Knock Out rose (ratings connected by the same letter are not significantly different by Tukey-Kramer at a = 0. 05). .................................................. 81

PAGE 9

9 3 5 Mean total buds and flowers plus mean standard error for Knock Out rose observed at each damage rating (ratings connected by the same letter are not significantly different by Tukey-Kramer at a = 0. 05). ........................................................ 82 3 6 Mean floral area (cm2) plus mean standard error for Knock Out rose observed at each damage rating (ratings connected by the same letter are not significantly different by Tukey-Kramer at a = 0. 05). .............................................................................. 83 3 7 Monthly mean thrips terminal density ( S. dorsalis /cm2) and damage rating on buttonwood over time. ........................................................................................................... 84 3 8 Mo nthly mean thrips floral density ( S. dorsalis /cm2) and damage rating on Knock Out rose over time. ............................................................................................................. 85 3 9 Total number of each species of thrips collected above each host plot from August 2007 to September 2008 using yellow 7.6x12.7 cm sticky traps. ....................................... 86 3 10 Weekly mean captures of genus Frankliniella on yellow sticky traps placed above buttonwood and Knock Out rose. ..................................................................................... 87 3 11 Weekly mean captures of Gynaikothrips uzeli on yellow sticky traps placed above buttonwood and Knock Out rose. ..................................................................................... 87 3 12 Weekly mean captures of genus Thrips on yellow sticky traps placed above buttonwood and Knock Out rose. ..................................................................................... 88 3 13 Weekly mean captures of Merothrips spp. on yellow sticky traps placed above buttonwood and Knock Out rose. ..................................................................................... 88 3 14 Weekly mean captures of Echinothrips americanus on yellow sticky traps placed above buttonwood and Knock Out rose. .......................................................................... 89 3 15 Weekly mean captures of Chirothrips spp. on yellow sticky traps placed above buttonwood and Knock Out rose. ..................................................................................... 89 3 16 Weekly mean captures of Heliothrips haemorrhoidal is on yellow sticky traps placed above buttonwood and Knock Out rose. .......................................................................... 90 3 17 Weekly mean captures of Selenothrips rubrocinctus on yellow sticky traps placed above buttonwood and Knock Out rose. .......................................................................... 90 3 18 Weekly mean captures of Chaetanaphothrips orchidii on yellow sticky traps placed above buttonwood and Knock Out rose. .......................................................................... 91 3 19 Weekly mean captures of Holopothrips inquilinus on yellow sticky traps placed above buttonwood and Knock Out rose. .......................................................................... 91

PAGE 10

10 3 20 Weekly mean captures of Scolothrips sexmaculatus on yellow sticky traps placed above buttonwood and Knock Out rose. .......................................................................... 92 3 21 Weekly mean captures of low abundance low -frequency unidentified thrips on yellow sticky traps placed above buttonwood and Knock Out rose. .............................. 92 3 22 Weekly mean captures of Scirtothrips dorsalis on yellow sticky traps placed above buttonwood and Knock Out rose. ..................................................................................... 93 4 1 Weekly mean total number of thrips plus mean standard error for three fertilizer rates (no significant differences). ................................................................................................. 109 4 2 Weekly mean thrips density ( S. dorsalis /cm2) plus mean standard error for three fertilizer rates (no significant difference). .......................................................................... 110 4 3 Mean total number of Scirtothrips dorsalis per floral sample for each fertilizer rate plus mean stan dard error (different letters among fertilizer treatments are significantly different by muliple pairwise t -test comparisons at a = 0.05). .................... 111 4 4 Mean damage rating recorded for each fertilizer treatment plus mean standard error (different letters among fertilizer treatments are significantly different by multiple pairwise t -test comparisons at a = 0.05). ............................................................................ 112 4 5 Weekly mean total number of thrips plus mean standard error for three fertilizer rates (* = high and maintenance rates are significantly different from control by multiple pairwise t -test comparisons at a = 0.05). ............................................................................ 113 4 6 Weekly mean thrips density ( S. dorsalis /cm2) plus mean standard error for three fertilizer rates (no significant differences detected using multiple pairwise t -tests, with p > 0.25). ...................................................................................................................... 114 4 7 Mean weekly total number of thrips for F. bispinosa and S. dorsalis plus mean standard error over time. ...................................................................................................... 115 4 8 Weekly mean thrips density ( S. dorsalis /cm2) plus mean standard error for each cultivar from mid -September to mid November. ............................................................... 116 4 9 Mean thrips density ( S. dorsalis /cm2) plus mean standard error on each cultivar (cultivars connected by the same letter are not significantly different by TukeyKramer HSD at a = 0.05). .................................................................................................... 117 4 10 Mean thrips density ( S. dorsalis /cm2) plus mean standard error on different plant organs sampled (organs connected by the sam e letter are not significantly different by Tukey-Kramer HSD at a = 0.05). .................................................................................. 118 4 11 Mean damage rating plus mean standard error for each cultivar over nine weeks of exposure (cultivars connect ed by the same letter are not significantly different by Tukey -Kramer HSD at a = 0.05). ........................................................................................ 119

PAGE 11

11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Re quirements for the Degree of Master of Science HOST SUSCEPTIBILITY AND POPULATION DYNAMICS OF Scirtothrips dorsalis HOOD (THYSANOPTERA: THRIPIDAE) ON SELECT ORNAMENTAL HOSTS IN SOUTHERN FLORIDA By Andrew I. Derksen August 2009 Chair: Catharine Mannio n Major: Entomology and Nematology Scirtothrips dorsalis Hood (1919) is an invasive pest thrips that recently arrived in Florida. The growth and dispersal of S. dorsalis populations are influenced by environmental factors. Populations show a positive rela tionship with increasing temperature, and months with greater accumulated degree days show larger thrips populations on all hosts tested. Increasing humidity and rainfall are associated with negative population growth. Periods of peak population growth als o appear to be periods of peak population dispersal. Temperature and humidity are also important to dispersal, with higher temperatures increasing thrips flights throughout the day, and higher humidity reducing the number of flying thrips. Dispersal of S. dorsalis is heavily influenced by the direction of the wind. This thrips appears capable of choosing between hosts, appearing in greater density on schefflera than buttonwood, but this may be an effect of the denser foliage of the schefflera trapping mor e thrips in the wind. It was the predominant thrips detected on all hosts year round, but while all life stages were detected on hosts such as schefflera and buttonwood, populations were significantly larger on rose. Population demographics of rose also suggested a growing population.

PAGE 12

12 When comparing different cultivars of rose, all tested appear to be exploited in a similar fashion and suffer similar damage from the thrips. Larger populations of S. dorsalis resulted in greater damage to host plants, with a reduction in blossom size by almost half. Thrips were more abundant in flowers than on leaves, but the density of thrips per square centimeter was similar between the two parts. Thrips density was fairly even between plants within a plot, suggesting that the thrips has density dependent distribution. The nutrient quality of plants also impacted thrips populations. While higher rates of fertilization did not result in significantly larger numbers or sizes of flowers, there were trends suggesting that over ti me it would result in significantly larger populations of thrips.

PAGE 13

13 CHAPTER 1 LITERATURE REVIEW Expansion and Economic Impact Scirtothrips dorsalis Hood (1919) is a remarkably successful invasive thrips whose range has expanded rapidly from Asia over the last twenty years. It is a polyphagous insect with a broad host range and a rapid rate of development and reproduction. Its small size permits wind dispersal and makes it difficult to detect during agricultural inspection. Together, these traits contribut e to its pest status on many crops of economic significance, and have led to its recent arrival in Florida. Nomenclatural and Biogeographic History The characteristic feeding damage of S. dorsalis was known and recognized as Murda disease (Kulkarni 1922, Ramakrishna Ayyar 1932) on chili long before the thrips was associated with and then determined to be the cause of the blight. The insect was first described in India by J. Douglas Hood of Cornell University (1919b) and then by subsequent authors (Table 1 1 ) as a pest common to Ricinus communus L. ( castor bean ) and Capsicum annuum L. (chili pepper ) (1919a). Synonymy probably occurred due to the global distribution and broad host range of this thrips, its small size and difficulty to characterize, as well as the limited access that earlier authors had to type material (Mound and Palmer 1981) While it is currently believed that the point of biogeographical origin for the thrips is in Southeast Asia or on the Indian subcontinent, the original host s are unk nown. As c hili peppers did not arrive in India until the sixteenth and seventeenth centuries with Portuguese traders (Collingham 2006), and given S. dorsalis broad host profile, it is unlikely that the initial host of description is also the point of orig in. I t has been speculated that a weed such as Ricinus communus L. later adopted for agriculture, may have been one of several hosts of origin. As

PAGE 14

14 agriculture developed in its native range, the thrips may have adapted to opportunity, and began to exploit other hosts as they became available a pattern observed as this insect expands its range. Biological Invasion The success of an invasive species depends on its ability to find and exploit a sustainable ecological niche. The organisms most likely to est ablish a foothold in a new environment are generalist herbivores. As a group, they are more likely to find sufficient substrate at lower trophic levels to meet their basic nutrient needs. This allows their population to subsist until they have the opportun ity to disperse into more ideal environments (Crawley 1986). Thrips are inadvertently well adapted to a lifestyle as an invasive species (Morse and Hoddle 2006), and have shown themselves to be remarkably effective at humanaided dispersal. When transport ed by humans, some species ranges have advanced by almost two hundred kilometers in a year (Kirk and Terry 2003). As a small cryptophilic organism that prefers to feed in the tightly packed meristem of growing plants, they may successfully avoid many contr ol measures. Inspectors may not observe individuals hiding within the blooms, and topical pesticides or vigorous washes may not penetrate the tight folds of buds to remove or exterminate the thrips. Thrips rely upon their host plant for a controlled enviro nment providing moisture and nutrients, remaining insulated against even some extremes of temperature. More severe mechanisms to kill the thrips might also damage or destroy the host, thereby reducing its economic value. The mechanisms by which an organis m achieves dispersal are important to the evolutionary success of any species (Nathan 2001, Nathan et al. 2003), but are particularly important to an invasive species seeking to establish itself in a new environment. Thrips are capable fliers (Ellington 19 80), but usually rely on their wings to carry them in directed flight

PAGE 15

15 only over short distances. More recent research has suggested that their wings play a stronger role in local dispersal than in long distance dispersal (Rhainds et al. 2005). Thrips low mass and the high surface area created by their fringed wings allows them to be caught and carried by the wind over great distances through the upper atmosphere by the rise and fall of daily convection currents (Lewis 1973). Collection experiments conduct ed during the 1930s using stickytraps mounted on airplanes found that thrips occupied a significant proportion of the aeroplankton (Glick 1939). It is believed that this is the normal mechanism by which thrips achieve longdistance dispersal without hum an intermediaries, and populations of S. dorsalis found in St. Vincent (Seal et al. 2006a) appear to support this model. Within a large plot, Seal et al. observed aggregated population clusters randomly distributed through the field, suggesting an initiall y random colonization event within the crop followed by gradual growth and expansion to proximate hosts. Within -plot dispersal presents an interesting avenue of research, for the mechanisms which drive dispersal may provide either mechanisms for control or the ability to anticipate the direction or rate at which the invasion will proceed. Outbreak and Establishment Thrips populations will occasionally explode into outbreak status, increasing their potential for economic damage. The causes driving these outbreaks are unknown, but they follow an established pattern for phytophagous insects (Mattson and Haack 1987b, a). Thrips o utbreaks are usually reported in conjunction with a season of drought (Schultz 1989, Wheatley et al. 1989, Irwin 1991), or recognized with the observation of tospovirus symptoms in a crop. Sensitivity to population density (Rhainds et al. 2005) and intraspecific competition may be increased by conditions affecting host quality, resulting in mass -dispersal to relieve population pressures It may be that lowered host quality triggers conditions for dispersal and outbreak in thrips. There is some evidence supporting this hypothesis in S. dorsalis When Shibao et al. (1993)

PAGE 16

16 physically removed the apical buds of grape plants; they observed i ncreased flight and dispersal of thrips as well as a reduced density of those thrips still feeding on plants within the plot. Invasive organisms may also only achieve outbreak status outside of their native range. In moving to a new environment, they may no longer be restricted or restrained by their competitors or predators, and may no longer be facing hosts which have evolved defenses specific to their feeding (Hoddle et al. 2002). As thrips are parthenogenetic insects with a high generation rate, even i ndividuals can potentially establish a successful colony. While it is highly unlikely that a large number of thrips would successfully avoid inspection and control measures at ports, there is experimental evidence demonstrating that a self -sustaining popul ation of thrips can be established with only a few individuals. When Sericothrips staphylinus was intentionally released to control Ulex europaeus L. ( gorse ) in New Zealand, it was observed that many small releases of less than a hundred thrips were more l ikely to establish a stable breeding population than single large releases of a thousand or more thrips (Memmott et al. 1997). This repetitive release of small populations simulate d regular commerce with an infected area, and implies that unless an invasi ve population is controlled and eliminated at the source, a foothold will eventually be established in any favorable environment (Zaman and Maiti 1994). Additionally, while parthenogenesis might limit the initial genetic diversity of any invading populatio n and retard that populations ability to develop resistance to novel control measures, repeated infection from a source population will provide some form of gene flow and increase the potential to establish resistance ( Morse and Hoddle 2006). History of Expansion The expansion of S. dorsalis has been fairly typical for an invasive thrips. It spread rapidly throughout the contiguous region of southeastern Asia fairly early, and there are many historical

PAGE 17

17 instances of this thrips attacking key crops on a re gular basis. In India, it has been described as a pest of Ricinus communus L. ( castor ) (Raizada 1965), Capsicum spp. L. ( peppe r) (Varadharajan and Veeravel 1995), Gossypium spp. L. (cotton) (Panickar and Patel 2001), Camellia spp. L. (tea) (Dev 1964, M ukhopadhyay et al. 1997), Mangifera spp. L. (mango), and Arachis hypogaea L. (peanut) (Wheatley et al. 1989). Outside of India, it has been reported as a pest in China on tea and Litchi chinensis Sonn. (litchi) (Li et al. 2004), in Taiwan on citrus and vegetab le (Chang 1995), on citrus and tea i n Japan (Okada and Kudo 1982b, Tatara and Furuhashi 1992) many vegetable and fruit crops in Thailand (Bansiddhi and Poonchaisri 1995), peppers and mango in Vietnam, and even seasonally on the Korean peninsula (Lee et al 2001). Under the pressures of globalization and trade, this thrips has expanded its range greatly, and EPPO (1997) recognized this pest as one with significant potential for global expansion. By that point it had already become fairly established beyond its precinctive range, had been intercepted at South African ports in 1986 (Gilbert), noted as a pest in Kenya by 1997 (Lewis), and infesting cotton in Cte d Ivoire by 1999 (Bournier). It was described as a pest in Australia on Anacardium occidentale L. (cashew) in 1998 (Peng et al. 2004), and on Fragaria x annanasa Duchesne (strawberry) and tea only a few years later (Hoddle and Mound 2003, Chin et al. 2007). By 2001, it had been detected in Egypt, and in 2003 it was found in Israel (EPPO 2003). It was only a matter of time before S. dorsalis was intercepted in the United States and the Caribbean region. While the thrips had been reported from port interceptions in Florida in 1995 and in Texas in 2000, surveys failed to detect any established populations or other sightings of the thrips (Skarlinsky 2003, Silagyi and Dixon 2006). However, interceptions of this insect at a Miami port under the calyces of pepper from St. Vincent in 2003 (Skarlinsky) inspired the USDA to act in order to predict and prevent th e pests arrival (Meissner et al. 2005). The USDA

PAGE 18

18 APHIS and the University of Florida responded with investigations of Caribbean islands (Seal and Ciomperlik 2004, Seal et al. 2006a). They found the pest already distributed throughout the Caribbean, and sp eculated that it had almost certainly already spread throughout South and Central America (Klassen et al. 2008). In late 2005, S. dorsalis was reported as a significant pest on Palm Beach County ornamental roses in Florida and reports from other counties of the thrips infesting the Knock Out cultivar of rose and pepper rapidly followed thereafter (Silagyi and Dixon 2006). By January of 2007, the thrips had been found in more than thirty counties from Alachua to Monroe, and has been detected in southern Georgia (Diffie et al. 2008). It has been spotted several times on retail roses in south Texas, and anecdotal evidence suggests that the expansion into Texas counties is probably greatly underreported (Ludwig 2008). Models of climatological and host potent ial suggest that this thrips has the potential to expand its range to cover much of the southeast, the Gulf coast region, and the western coast of the United States (Venette and Davis 2004, Nietschke et al. 2008). Economic Impact Nationally, horticulture is a $16.5 billion industry, and Florida is the nations second largest contributor, adding $1.6 billion to the national economy last year (Jerado 2006). Among the largest of Floridas contributions are the ornamental and cut -flower industries, which provided nearly a billion dollars in sales during 2006. This industry has grown and slowed in step with the housing and development boom, following trends in disposable household income (Jerado 2006). Peppers, strawberries, cucumbers, cotton, and peanuts add a nother $483.5 million in vegetable crops to Floridas economy. All of these crops have the potential to serve as hosts to S. dorsalis and losing even ten percent of these industries would prove catastrophic to the national economy, as well as open the mar ket to foreign competition.

PAGE 19

19 Prolonged feeding by thrips curls tender leaves and buds, and will turn fruits and flowers from bronze to black in color, rendering the plant material unmarketable. Feeding on young tissue stunts further growth, and may induce abortion of fruit (Lewis 1997). Even damaged or scarred vegetables or inflorescences are viewed as unmarketable. When thrips feed in high densities, or in sufficiently dry climates, this process can result in the eventual desiccation and death of their hos t plant. Even low densities of thrips can contribute to the decline in fruit production and plant health, especially during times of drought (Kolb et al. 1991, Fournier et al. 1995). At least ten species of thrips belonging to family Thripidae including S dorsalis have been confirmed as vectors for the transmission of thirteen or more tospoviruses (Jones 2005), further emphasizing the value of studying this group (Mound 2002, Sakurai et al. 2002). Tospoviruses are negative RNA viruses in the family Bunyav iridae. Infection with the virus results in spotting and wilting of the plant, reduced output, and occasionally death. Even at a low density, thrips can transmit the virus and cause mortality out of proportion to their actual feeding damage. They may infec t a broad spectrum of plants with thrips as a common vector, leaping from vegetable to ornamental crops (Chatzivassiliou et al. 2000) The thrips vectors are not closely related, implying an independent origin of infection for each thrips (Mound 2002), pos sibly transmitted horizontally through shared hosts. There may be other species of thrips competent to transmit similar viruses, but they have not been observed on crops of economic significance. S. dorsalis has been implicated in the transmission of seve ral separate tospovirus (Whitfield et al. 2005), as seen in ( Table 1 2 ). However, recent experiments have cast some doubt on the efficiency at which S. dorsalis actually transmits the virus to its hosts. It may be a

PAGE 20

20 member of a class that Nagata et al (20 02) described in a near relative, Thrips palmi Karny: an infected nontransmitter with detectable levels of virus. Morphology and Classification The Thysanoptera were first described in 1744 as a genus Physapus by DeGreer, and then renamed Thrips by Linna eus in 1758. In 1836, Haliday promoted the genus to the rank of order, and named them Thysanoptera after their distinct ciliated wings. The order is currently divided into two suborders: the Terebrantia, and the Tubulifera. Tubulifera can be identified by their characteristic tube -shaped apical abdominal segment, whereas females of the eight families of the Terebrantia all possess the eponymous saw like ovipositor on the anteapical abdominal segment. Of the eight families of Terebrantia, family Thripidae is the most species rich with over 290 genera representing just over two thousand species (Mound 2002). Thripidae are considered to be among the more derived of thrips, having evolved many traits key to specializing as cryptophilous phytovores, living in th e narrow spaces at the bases of leaves and within florescences (Gentile and Bailey 1968, Lewis 1973, Mound 1998). Adult S. dorsalis can usually be distinguished from other thrips collected in a South Florida sample by its small size, yellow coloration, an d a dark transverse ridge on the lower abdominal tergites (Seal et al. 2006a). It can be specifically identified and confirmed through the presence of forked sense cones on antennal segments III and IV, antennomeres I II are pale and III to IX are dark; th ere are three discal setae on the lateral margins of abdominal tergites, with pronotal posteromarginal seta II nearly one and a half times the length of I or III, a complete posteromarginal comb on tergite VII; and three ocellar setae with III between post erior ocelli (Mound and Palmer 1981, Hoddle and Mound 2003, Skarlinsky 2004). Thus far, all observed populations within Florida meet this description (GB Edwards, personal communication).

PAGE 21

21 Ecology and Behavior Thrips are believed to have descended from a my cetophilic ancestor during the Mesozoic (Grimaldi et al. 2004), and many groups still feed upon and inadvertently redistribute fungal spores, but most research has focused on those species feeding on or in association with economically significant crops. S ome thrips are predatory, but the majority are phytophagous insects feeding on pollen and the chloroplasts harvested from the outer layer of plant epidermal and mesophyll cells (Heming 1993, Kirk 1995). These species are small cryptophilic organisms that p refer to feed within the tightly packed apical buds of new growth. Flower -feeding thrips may be responsible for pollination while feeding (Sakai 2001), and S. dorsalis has been documented as a pollination vector encouraging gene flow for pepper plants in the field (Saxena et al. 1996), but their most significant contribution remains the defoliation they can cause during years of outbreak. Feeding usually occurs along the main vein or ribs of leaves and petals (Lewis 1973, 1997). The feeding damage of S. dorsalis is distinctive, and in some cases may be considered diagnostic of the pest (EPPO 2005). Feeding damage for S. dorsalis leaves bronzed scars on the petals, leaves, and surface of fruits (Asaf -Ali et al. 1973, Tatara and Furuhashi 1992, Shibao 1996a, Chandrasekaran 2005). This damage is frequently described as greater than an infestation by other thrips. Some researchers suggest that a protease in the saliva may be responsible for the additional damage (Dev 1964, Raizada 1965). S. dorsalis is often fo und in the flowers and terminal shoo ts of vegetables and ornamental plants (Shibao et al. 1993), and will feed upon pollen (Saxena et al. 1996). Some authors suggest that it is primarily a foliar thrips, but many authors have observed feeding in floral str uctures. Saxena et al. (1996) demonstrated that in an olfactometer test, a greater number of S. dorsalis were attracted to scents of pollen than to nectar or plant -tissue.

PAGE 22

22 Like many thrips, S. dorsalis appears to prefer feeding on new growth and young pla nts (Dev 1964, Shibao et al. 1990, Seal et al. 2006a), and is often found on the newer abaxial surface of leaves on smaller plants. Physically manipulating the host by removing preferred feeding sites has been shown to reduce thrips density on the plant, but also to increase the relative rate of between -plot dispersal (Shibao et al. 1993). Family Thripidae is particularly notorious for members with broad host ranges, and the majority of pest thrips come from this family (Bailey 1940, Ananthakrishnan 1993). Feeding by S. dorsalis has been described on over a hundred species of plants representing over forty separate families, and new hosts are being identified as the creature expands into Floridas environment (Silagyi and Dixon 2006). Many of the hosts ide ntified for S. dorsalis have been common agricultural crops and ornamental plants, but these should not be thought of as the limits of the organisms host range, as these records may represent an error of observation. Many pest organisms are only reported when a commodity is scrutinized on a regular basis, or when feeding damage on a particular host increases to noticeable levels. Other hosts of less economic value may go underreported and serve as a reservoir for establishing an invasive species. Historic ally, S. dorsalis has not been documented on all possible hosts across its global range, and the population density of the thrips on the same species of host may be very different in geographically disparate environments (Hoddle and Mound 2003). As an exam ple, in spite of the availability of citrus hosts throughout S. dorsalis range, it has only been described as a significant pest on citrus in Taiwan (Chang 1995) Japan (Tatara and Furuhashi 1992), and Thailand (Bansiddhi and Poonchaisri 1995) More recen tly, this thrips has been observed as a possible pest on mango in Puerto Rico, but not in nearby Florida (Klassen et al. 2008).This

PAGE 23

23 regional variation may be due to variation in abiotic conditions, competition with established herbivore and predator commun ities, the availability of a given host in an area, or different standards for local IPM. An alternative hypothesis is offered by Hoddle et al (2008), who provide strong evidence that S. dorsalis may represent a cryptic species complex. Regional populations of this thrips may be locally adapted, and therefore present slightly different host profiles. It is also possible that the stated number of hosts actually overestimates the host range of this thrips. Due to their small size and broad dispersal the misi dentification and synonymy of insect samples are problems frequently encountered in the literature of Thysanoptera (Mound and Palmer 1981, Mound and zur Strassen 2001, Diffie et al. 2008). Samples are frequently collected from plants where adults may be fe eding, but upon which the insect may be incapable of completing development (Ananthakrishnan 1993). Masui observed many adult S. dorsalis feeding upon citrus, but found significantly lower numbers of nymphs on citrus relative to other nearby hosts (2007b). These results, and others like it, lead some authors to suggest that this discrepancy could be resolved by specifying whether the strength of the host association in the literature was for a feeding or reproductive host. While host selection is usual ly dependent on many stimuli, the factors leading to thrips attraction are poorly understood, and are frequently species -specific (Jan de Kogel and Koschier 2002). Many leaf and flower -feeding insects have been selected to identify their prey through visua l cues (Lewis and Taylor 1964), with a species -specific response to lighting stimuli of various wavelengths (Lewis 1973). S. dorsalis is no exception, and there is considerable evidence demonstrating that it does rely upon visual stimuli to identify hosts. Studies with chili peppers demonstrated that morphological characters of plants, including height, leaf size, petiole

PAGE 24

24 length, and leaf internode distance can have an impact on populations of S. dorsalis (Pramanick and Mohasin 2004). Experiments conducted by Shibao (1996b) and Mochizuki and Honma (2001) have shown that screens or mulches reflecting light in the ultraviolet spectrum can be used to effectively reduce seasonal populations of S. dorsalis Larger numbers of S. dorsalis adults are captured on yel low (Tsuchiya et al. 1995, Chu et al. 2006) or yellow and green (Tsuchiya et al. 1995) sticky traps. This appears to conflict with the findings of Rani and Sridhar (2003) and Gahukar (2003), in which more adults of S. dorsalis were recovered on red or orange rose petals than yellow ones in a small arena choice-test. Saxena et al. (1996) noted more adults of S. dorsalis were captured on white traps than on yellow, red, green, blue, or black colored traps placed in a pepper field. All of these results may be true at different spatial scales, with S. dorsalis relying on different cues to identify and select hosts at different distances. Both Rani and Sridhar and Gahukars choice -tests occurred in small arenas within the lab, whereas the other tests were perform ed in the field. Saxena et al. speculate that their results may show an ability of the thrips to identify the contrast of brighter flowers against the darker background foliage, while Tsuchiya and Chu suggest that their results may be similar to the colora tion of younger foliage this thrips preferred feeding location. These general cues would be valuable to a thrips descending from great distance. The components used in host selection are probably not exclusively visual, and this may help to explain the apparently contradictory results of Rani and Sridhar to Gahukars findings. Scirtothrips dorsalis may have been attracted to the scent of rose petals as well as to color, and volatile chemical compounds may be of greater importance to local dispersal. Whi le poorly documented, chemical communication is believed to be important to the group (Blum 1991). Anal secretions are produced in the hindgut (Howard et al. 1983), and released along the

PAGE 25

25 posterior setae as predator deterrents (Lewis 1973, Howard et al. 1983, Tschuch et al. 2002) in some thrips, and elements such as floral scent may play a role in host identification. Volatile compounds r eleased by plants during thrips feeding have been documented to act both as an aggregation kairomone (Ananthakrishnan 1993), and as a deterrent (Delphia et al. 2007). Saxena et al. also demonstrated that in an olfactometer choicetest, a greater number of S. dorsalis were attracted to scents of pollen than nectar or plant tissue. Host selection is important to the thrips, a s host characteristics may impact developmental time. Shibao observed 294.1 degree days on Vitis vinifera L. ( grape ) (1996c), while Tatara recorded 26 5 degree -days on Citrus reticulata Blanco ( mandarin orange) (1994), and Raizada (1965) documented 267 degr eedays on castor. At the other extreme, D found only 188 degree days were required for full development on Citrus Over the last thirty years, authors have generally observed at least one peak period of growth in mid-summer (Raizada 1965, Okada and Kudo 1982b, Shibao et al. 1990, Shiba o 1996a, Mukhopadhyay et al. 1997, Duraimurugan and Jagadish 2002, Shibao et al. 2004, Masui 2007a) after approximately sixty to ninety degree -days had passed, with further peaks every three hundred additional degree days (Tatara 1994, Shibao 1996c). Some researchers have observed a second population peak occurring in late fall (Raizada 1965, Okada and Kudo 1982b, Shibao 1996b, Paul and Konar 2005). The developmental rate observed increases linear to temp erature (Tatara 1994, Varadharajan and Veeravel 1995, Saxena et al. 1996, Shibao 1996c, Paul and Konar 2005) and the number of sunlit hours in a day (Saxena et al. 1996, Gahukar 2003, Paul and Konar 2005), and S. dorsalis development continues between critical thresholds of 9.7 32 generat ion rate of fifteen days at 25

PAGE 26

26 Negative linear correlations have been observed with increasing humidity (Varadharajan and Veeravel 1995, Saxena et al. 1996, Lingeri et al. 1998, Paul and Konar 2005) and rainfall (Lewis 1973, Varadharajan and Veeravel 1995, Saxena et al. 1996, Lewis 1997, Lingeri et al. 1998, Paul and Konar 2005, Singh 2005). At least one researcher has described a positive correlation to increasing humidity (Singh 2005), but his observations were made at a lower relative humidity than other researchers cited, and may suggest a lower threshold for humidity beyond which desiccation afflicts the small soft -bodied insect. All researchers generally suggest that hot and dry conditions with low intensity rainfall pro mote the best population growth for this thrips. It is speculated that this is because hard rain will wash adults and nymphs from the leaves to the ground, where many of them will not return successfully to their host plant for feeding after falling (Lewis 1997). Like all Terebrantia, S. dorsalis undergoes two nymphal stages followed by two pupal stages (Dev 1964) but research has shown that length of these life -history stages are flexible within a group depending on an individuals access to nutrients (M ound 2005) and temperature (Tatara 1994). Nymphs begin the metamorphic process by dropping off of the plant, and tend to pupate in loose soil or leaf litter at the base of their host, but have been observed to pupate in any dark and humid crevice low on the plant, including bark and lower leaves (Okada et al. 1981) or flowers. The pupal process can range from two days to a full week (Tatara 1994). In temperate regions where the temperature falls below the critical lower threshold, non-diapausing adults a re reported to over winter in the soil or apical buds (Okada and Kudo 1982a, b, Shibao et al. 1990) Colder temperatures may drive pupation, but this has not been experimentally confirmed (personal observation).

PAGE 27

27 Adults begin to emerge again in early summer in Japan (Shibao 1996c, Shibao et al. 2004), and it is believed that the greatest dispersal and distance propagation occurs during the emergence post -pupation immediately following winter (Masui 2007b). Six flight peaks have been observed in Japan (Masui 200 7a, b), and three peaks in India (Varadharajan and Veeravel 1995), running from early summer to late fall. These flight peaks do not necessarily correlate to population densities or abundances on plant, suggesting that dispersal behavior is not linked excl usively to population density (Shibao 1996c). However, some authors have noted that when generations are tracked, flight peaks do tend to occur after populations have built up to peak densities on hosts (Masui 2007a, b), and it is possible that Shibaos re sults are caused by overlapping generations in the field. Post -emergence, females have a pre -oviposition period of one to two days (Tatara 1994). Using their ovipositor, females lay single eggs within the plant tissue (Lewis 1973), and lay an average of f orty eggs during their lifetime (Tatara 1994). Females of S. dorsalis prefer to lay their eggs inside of young leaves and buds at the apical meristem of plants but as populations increase, will lay their eggs within the surface of mature leaves (Dev 1964, Raizada 1965). Depending on temperature, eggs may gestate for one to three weeks (Tatara 1994). After hatching, nymphs will migrate from older leaves to the newer growth at terminals (Onkarappa and Mallik 1998). S. dorsalis is often found feeding in the n ewer growth and top leaves of younger plants (Dev 1964, Shibao et al. 1990, Seal et al. 2006a), although individual plant morphology and chemistry may result in some distributional differences (Shibao et al. 1990). Nymph populations will continue to increa se so long as new shoots are allowed to grow (Shibao et al. 1993) and adults are permitted to land (Shibao 1996b).

PAGE 28

28 Integrated Pest Management Strategies for Scirtothrips dorsalis As thrips are a significant global pest for greenhouse and field producers, there is already considerable authoritative literature recommending generic management strategies that could be applied to most species of thrips. As S. dorsalis has been a pest for some time in India and East Asia, there is already a quantity of literatu re specific to this species that suggests several strategies by which this thrips might be managed in Florida and the United States. Detection and Sampling The feeding damage of established thrips populations are usually the first sign of their presence ( Lewis 1981). The feeding damage of S. dorsalis is distinctive, even among thrips, and in some cases may be considered diagnostic of the pest (EPPO 2005). This damage may appear superficially similar to spider -mite damage, but it can be distinguished from m ites and other thrips by the characteristic bronzed scars and stippling on the petals, leaves, and surface of fruits (Asaf -Ali et al. 1973, Tatara and Furuhashi 1992, Shibao 1996a, Chandrasekaran 2005). Concerned growers should l ook for flowers with small spots or blemishes, wrinkled and shriveled petals, or curled, wrinkled and dry new growth (Duraimurugan and Jagadish 2004b). If the infestation has proceeded long enough, there should be stippled scarring on the abaxial leaf surface. With sufficiently larg e populations, exuviae may be present on the lower leaf surface. These cast skins are distinct and can be distinguished from those of aphids. Once active populations have been identified, plant populations can be estimated by sampling a number of flowers appropriate to the size of the plot (Lewis 1981, Casey et al. 2007). Remove the infested flower or leaf, rinse it in a solution of 75% ethanol, and pour the runoff through a sieve with a 25 m grating (Seal and Baranowski 1993). The number, stage, and sex of thrips counted can be given a preliminary field identification under a dissection scope at a minimum of 12x magnification (Seal et al. 2006a, Funderburk et al. 2007), but slides should

PAGE 29

29 quickly be prepared using the methods of Bisevac (1997) for a detail ed examination and species confirmation at 200x magnification (Moritz et al. 2001, Moritz et al. 2004, Seal and Ciomperlik 2004). To detect adult populations before they become established and damage becomes visible, it has been recommended to hang one st icky card at flower height per thousand square feet of residential landscape or one per ten thousand square feet if a commercial grower (Lewis 1981, Casey et al. 2007). Historically, sticky traps have been valuable tools used to accurately predict on -plant population estimates of thrips in the field (Pearsall and Myers 2000b, a, Rhainds and Shipp 2003, Rhainds et al. 2005), and have been used to successfully predict populations of S. dorsalis on nearby plant hosts (Shibao et al. 1990). Research suggests th at either yellow (Tsuchiya et al. 1995, Chu et al. 2006) or yellow and green (Tsuchiya et al. 1995) sticky traps are most preferred by this thrips. Casey et al. (2007) recommend a generalized damage threshold of twenty -five to fifty thrips per sticky trap, or less than two thrips per bud in order to maintain flower production, but some authors (Shibao et al. 2004, Seal et al. 2006a, Shibao et al. 2006) suggest that lower numbers of S. dorsalis are recommended for successful agricultural or ornamental plant production. Mechanical Controls Before attempting other measures, certain mechanical processes may be changed in order to discourage or eliminate the thrips. While potentially labour intensive, these changes may be among the most cost -effective and least damaging to the surrounding environment. As S. dorsalis lay the majority of their eggs in the fresh flush of new growth and the opening buds of flowers (Onkarappa and Mallik 1998, Duraimurugan and Jagadish 2004b), Shibao et al (1993) demonstrated that the adult population density in a field of grape could be reduced by removing the flowers and growth terminals of nearby alternate hosts. It has also been

PAGE 30

30 suggested that as with other thrips, regular pruning of older leaves and sweeping the leaf litter benea th infested plants may reduce the number of eggs and pupae (Lewis 1973, 1997). Duraimurugan and Jagadhish demonstrated the effectiveness of this technique in rose, showing a reduction in nymph and adult populations, and a reduction in feeding damage (2004a ). Alternatively, regular streams of high pressure water simulating heavy rains may be used to reduce thrips populations by washing away nymphal thrips or by disturbing adults (Lewis 1973), but this technique may not be dependable. Schuch et al. observed that while this technique reduced adult numbers of Frankliniella sp., it was not as effective against nymphs who may have found hiding spaces on the plant (Schuch et al. 1998). Caging plants may prove effective as a physical barrier, but using an ultraviol et (UV) reflective tarp has an additional effect. The UV reflective barrier reduces the number of adult S. dorsalis landing on grape, and thereby the number of nymphs produced (Shibao 1996b). While the screening may block some landings, the reflection of U V light is presumed to play an additional role by interfering with the thrips ability to locate hosts. UV -reflective mulch was also observed to reduce S. dorsalis feeding damage on tea by as much as eighty percent (Mochizuki and Honma 2001). Cultural Controls Changing human behavior can be a key component to controlling outbreaks of any arthropod pest. Growers should do their best to inspect all plants for thrips before transporting them into an area where the thrips might expand. Ideally, one should kee p plants in a quarantine area for up to two weeks, as this should allow sufficient time for S.dorsalis rapid life cycle to force it to reveal itself (Lewis 1981). Removing alternate hosts from nearby a production area may also provide a control mechanism (Lewis 1973, Shibao et al. 1993), as hidden populations of S. dorsalis may migrate from these alternative hosts to a preferred host (Masui 2007b).

PAGE 31

31 Temporally shifting plantings of agricultural crops may not prove an effective strategy against this thrips, as the peak growth season for many of these crops may also be most adventitious to this thrips, and its broad host ra nge should allow it some refuge This strategy may also prove ineffective to ornamental plant hosts for similar reasons, as the host range of this thrips includes many perennials, and it would be difficult to eradicate it from all potential hosts in an environment. A more compelling management technique would be to develop, discover, or plant existing strains of commercially available crops that are naturally resistant to this thrips. Resistance can be defined as all plant responses resulting in the ability to withstand infestation, and to support insect populations that would severely damage susceptible plants (Horber 1980), and it comes in many forms. Xenobiosis excludes pests by discouraging feeding through physical or chemical barriers. Tolerance can be described as an elevated pest threshold before a plant begins to show additional damage when compared to a susceptible plant (Fery and Schalk 1991). Several varieties of plant have already been screened for resistance to S. dorsalis including peanut (Ekvised et al. 2006), pepper (Krishna Kumar et al. 1996, Pramanick and Mohasin 2004), and rose (Onkarappa and Mallik 1998, Gahukar 2003, Ra ni and Sridhar 2003). An increase in the concentration of free amino acids in plant tissues has also been shown to attract pest insects (Slansky and Rodriquez 1987), and adding nitrogen-based fertilizers to a plant will also allow for a higher growth rate for feeding insects (Mattson 1980). This trend holds true for the Thysanoptera, with elevated concentrations of fertilizer resulting in a higher growth rate of Frankliniella occidentalis (Chau et al. 2005), but this process will be mediated by the endogenous resistant factors produced by specialized cultivars (Schuch et al. 1998) or by particular hosts (Reitz 2002). Duraimurugan and Jagadish (2004a) observed increasingly higher populations of S. dorsalis on rose fertilized with 15 grams of nitrogen. Varghe se and Giraddi

PAGE 32

32 (2005) recommended halving the fertilizer used on chili in conjunction with azadirachtin when controlling for S. dorsalis While nitrogen levels are important, the work of Abro et al. (2004) in cotton suggests that controlling other micronut rients may not significantly affect S. dorsalis populations. Chemical Controls While insecticides are among the fastest and most effective methods by which to deal with a pest, they are not always the first solution selected for pest -management. Off -targe t environmental contamination and resistance should always be considered before application. The evolution of pesticide resistance is of particular importance to thrips, due to their high generation rate, and the ability of an individual resistant female to parthenogenetically reproduce and replace a population. The Rutgers IR 4 (Palmer and Vea 2008) and Seal et al. (Seal et al. 2006b) studies currently provide the most exhaustive and recent tests for insecticide efficacy against S. dorsalis on several ornamental and vegetable hosts. While many of the compounds tested demonstrate success in controlling thrips populations, chlorfenapyr, spinosad, imidacloprid, and azadiractin appear to be the most effective pesticides at suppressing populations of S. dorsali s Permethrin has also been used effectively on citrus (Shibao 1997b) and Nair et al. recommend dimethoate and phosphamidon at 0.05% as the most effective on rose (1991). Traditional methods, such as ground neem seeds, have also proven effective in contr olling S. dorsalis on chili (Varghese and Giraddi 2005, George and Giraddi 2007). A solution of macerated Lepidopteran larvae has been shown to induce volatile compounds that deter other Thysanoptera (Delphia et al. 2007), and might also prove effective ag ainst S. dorsalis Resistance to these pesticides and methods has yet to be documented in the literature, but such a possibility remains a significant concern.

PAGE 33

33 The application of a systemic pesticide such as imidacloprid to seed or young plants early in d evelopment is critical to preventing or stopping an attack that will overwhelm their defenses and destroy a crop (Kannan et al. 2004). Control of S. dorsalis on mature plants can be achieved using an evenly applied foliar spray of a contact insecticide, bu t this topical application may miss some members of the population feeding within closed buds or unfurling leaves. As S. dorsalis are cryptophilic, systemic formulations with high penetration are critical to controlling this pest. Classical Biological Cont rol Invasive species can potentially avoid accumulating a high predator load in a new environment, as local predators may have yet to adapt to the behaviors and chemical signals that signify a new prey resource to exploit. In some situations, an invading species may demonstrate greater fitness in their adventive environment when compared to their precinctive populations of origin. This predator free space may last as long as ten years, and may grant invasive species the competitive advantage they need to outperform established competitors using similar resources (Cornell and Hawkins 1994) The class of predators most likely to accumulate on invasive species early in their expansion is a generalist (Cornell and Hawkins 1994) These are polyphagous predato ry organisms that will utilize any organism that falls within their foraging strategy as a food resource. As thrips appear to have operated under selection favoring small size and cryptophilic behavior within the tight confines of early plant growth, they represent a difficult prey item to exploit, and have accumulated few specialist predators (Loomans et al. 1997). As most of the pest thrips have such broad global ranges, it has been difficult to find a precinctive range where native specialist predators m ight control their growth (Loomans 2003). As a result, generalist predators that specialize in thrips are usually selected as the most promising option for long term suppression of many thrips species (Funderburk et al. 2000, Loomans 2003). Some research h as

PAGE 34

34 already looked at the effectiveness of several generalist predators in suppressing populations of S. dorsalis The Hymenopteran eggparasitoid, Megaphragma (Hymenoptera: Trichogrammatidae) could repress populations of S. dorsalis when released in suffi cient quantities (Takanashi et al. 1996), sometimes parasitizing as much as fifty-one percent of all host eggs, but proved seasonally incompatible (Shibao et al. 2000). The Trichogrammatids never established themselves during periods of peak thrips populat ion growth, and were rejected as a control mechanism in Japan. It remains possible that these wasps would prove more effective in an environment where its time of peak fecundity and activity matched that of the thrips intended as hosts. One of the most rec ommended predators for control of thrips have been bugs of genus Orius (Hemiptera: Anthocoridae) (Funderburk et al. 2000). These insects can be mass -reared commercially, and provide an inexpensive and potentially self -sustaining population. They prey prima rily upon the nymphs of many species of thrips (Kajita 1986), but will feed opportunistically on other small herbivorous pests, and maintain their populations using pollen as a protein source when prey is not available. However, there remains some concern that Orius might ignore S. dorsalis in the presence of alternative prey items. This possibility has already been raised by choice tests for Orius between the larger and more docile Frankliniella occidentalis and the smaller and more mobile F. bispinosa (Re itz et al. 2006). As S. dorsalis is smaller bodied and relatively long -legged thrips (Raizada 1965) noted for its activity (Dev 1964), it is probable that Orius might ignore this prey item in favor of less difficult and more rewarding prey, giving this thr ips a competitive advantage against other thrips already present in the landscape. This avenue of research merits further investigation, as it is possible that certain

PAGE 35

35 smaller species of Orius may engage in niche -specialization, and choose to prey on the s maller thrips. Phytoseiid mites have also enjoyed great success as a biocontrol of thrips (McMurtry and Croft 1997), although primarily in greenhouses. Shibao et al (2004) demonstrated a proportional increase in predator density of Euseius sojaensis in re sponse to an increase in the density of nymphal S. dorsalis on grape, but the seasonal growth peak arrived too late in the season to control S. dorsalis before damage could occur to host plants. Other species of mite, including Neoseiulus cucumeris Oudeman s and Amblyseius swirskii (Athias Henriot), have proven effective predator in studies of field and gree nhouse vegetables (Arthurs et al 2009). Thrips of genus Scolothrips have been observed feeding on juvenile S. dorsalis on castor in India (Dev 1964, Raiz ada 1965), as have thrips of genus Franklinothrips However, mass rearing of predatory thrips has frequently proved more difficult than encouraging naturally established populations. In the end, the most effective biocontrol for S. dorsalis may prove to b e the ubiquitous mycopathogens (Raizada 1976) or a durable strain of nematode (Sims et al. 2005). A sustainable culture of entomopathogenic fungi grown on-plant, such as Fusarium (Mikunthan and Manjunatha 2008), might provide enduring protection. Conclusio ns Thrips are a diverse and abundant group of insects whose life histories are relatively understudied, given the significant potential for economic damage that they represent. Scirtothirps dorsalis has recently invaded Florida, and presents an excellent opportunity to explore the seasonal population dynamics of this insect The literature indicates that this thrips demonstrates variable performance on different hosts, but it remains unanswered whether S. dorsalis actively selects hosts of higher performance or whether it disperses randomly. A broad

PAGE 36

36 range of host plants in Florida have already been reported with significant damage and large populations of these thrips, but the extent to which certain hosts are susceptible has been poorly documented. The fa ctors determining the selection and timing of reported outbreaks of S. dorsalis are also poorly understood. While the literature suggests that this insect prefers warmer weather with reduced rainfall, the annual population dynamics of this thrips in southe rn Florida have not yet been documented. Knowledge of this thrips seasonal performance could be very valuable to the design of a more efficient pes t management program, allowing more preventative plans that anticipate periods of peak activity and dispersa l. The conditions leading to outbreak and dispersal among all thrips and herbivorous insects are poorly understood, and using S. dorsalis to test hypotheses of dispersal and host selection would be valuable to preparing a control regime for this species, and may provide insight into certain broad questions of thrips ecology.

PAGE 37

37 Table 1 1 Scirtothrips dorsalis Hood (1919) was first described on chili peppers in India, and as is common with the order, has since undergone considerable synonymization Genus s pecies Authority Year Scirtothrips dorsalis Hood 1919 Heliothrips minutissimus Bagnall 1919 Anaphothrips andreae Karny 1925 Neophysopus fragariae Girault 1927 Scirtothrips padmae Ramakrishna 1942 Table 1 2 Scirtothrips dorsalis has been implicated in the transmission of several tospoviruses. Virus Abbreviation Reference Peanut yellow spot virus PYSV (Reddy et al. 1991) Groundnut bud necrosis virus GBNV (Reddy et al. 1992) Peanut chlorotic fanspot virus PCFV (Chen and Chiu 1996) Watermelon silve r mottle virus WSMoV (Chiemsombat et al. 2008) Capsicum chlorosis virus CaCV (Chiemsombat et al. 2008) Melon yellow spotted virus MYSV (Chiemsombat et al. 2008)

PAGE 38

38 CHAPTER 2 DISPERSAL OF Scirtothrips dorsalis HOOD (THYSANOPTERA: THRIPIDAE) FROM KNOCK OUT RADRAZZ ROSE TO TWO ALTERNATIVE ORNA MENTAL HOSTS IN SOUTHERN FLORIDA Introduction Scirtothrips dorsalis Hood (1919) is a polyphagous insect pest whose global distribution has expanded rapidly over the last twenty years, becoming established in souther n and central Florida (Silagyi and Dixon 2006). This insect pest shows considerable potential for expansion into the rest of North America (Venette and Davis 2004, Meissner et al. 2005, Nietschke et al. 2008), and impacts several fruit and vegetable hosts of economic significance, including Fragaria x annanasa Duchesne (strawberry), Mangifera spp. L. (mango), Capsicum spp. L. (pepper) Citrus spp. L ., Gossypium spp. L. (cotton), Glycine max (L.) Merr (soy bean) and Arachis hypogaea L. ( peanut ). The fruit an d vegetable crops potentially affected by this thrips added nearly 1.4 billion dollars to the Florida economy in 2005 (Jerado 2006). For many hosts, damage from this thrips results in reduced floral and vegetative growth increased plant stress, scarred fo liage, and can result in death (Tatara and Furuhashi 1992, Shibao 1996). Ornamental hosts are particularly sensitive to feeding damage, as even minimal scarring can render a plant unsalable. Among these host plant s are popular landscaping plants including Rosa spp. L. Raphiolepis indica (L.) Lindl. ex Ker Gawl. (Indian hawthorn), Ligustrum spp L. (ligustrum) Conocarpus erectus var erectus L. ( G reen buttonwood ), and Schefflera arboricola Forst & Forst ( Emerald Green schefflera). Schefflera (Emerald G reen) and buttonwood (Green) were identified as ornamental hosts of significance after early reports by nurseries to extension agents in south and central Florida (C. Mannion, personal communication) B oth plants are common in southern Florida and have a broad distribution because l andscapers make frequent use of both species as hedges or

PAGE 39

39 borders This practice could permit the thrips a natural corridor for establishment throughout the state (Morse and Hoddle 2006). As this thrips is a pest for both orn amental and agricultural crops it may develop resistance to co mmercial pesticides in one area and the resistant strain may spread rapidly to another (Kirk and Terry 2003). G rower s and homeowner s must simultaneously apply a comprehensive integrated pest man agement program that incorporates both of their unique environments to prevent the establishment of reservoir populations or the development of resistance to commercial pesticides. An awareness of the presence of this thrips upon common ornamental plant s will be valuable in constructing a part of a fully integrated pest management program. Environmental cues may also be important to an organisms dispersal behaviors. Natural selection has driven m any leaf and flower -feeding insects to identify their host through visual cues, and therefore tend to fly by day (Lewis and Taylor 1964) Daytime flight is not uncommon for thrips, and other flower feeding thrips such as Frankliniella occidentalis have been documented as diurnal fliers (Pearsall 2002) with a speci e s -specific response to light of various wavelengths (Lewis 1973). Chu et al. (2006) discovered that yellow sticky traps captured more individuals of S. dorsalis than sticky traps of other colors, indicating that this thrips is probably a daytime flier. Se al (personal communication) report ed peak flight activity within greenhouses between 1000 and 1400. Recent research has emphasized the importance of local dispersal to the success of colonization by thrips (Rhainds et al. 2005), and understanding the patt erns by which thrips disperse within a plot are key to establishing a successful long -term control strategy (Groves et al. 2003, Nault et al. 2003) An awareness of peak flight activity may allow growers to better target the application of pesticides when thrips populations are at their highest An improved

PAGE 40

40 understanding of the timing and environmental conditions during which this species of thrips increases flight behaviors may help to predict the overall rate of dispersal for the thrips through the Gulf c oast region. Th e purpose of this study was to determine the influence of host plant and wind conditions on thrips dispersal. Also observed were the time of day and environmental conditions during which the thrips was most likely to disperse. Materials an d Methods Dispersal behaviors of S. dorsalis were evaluated using rose, schefflera, and buttonwood. Schefflera and buttonwood were grown from cuttings harvested at the University of Florida Tropical Research and Education Center ( TREC) campus in Homestead, Florida. Prior to the initiation of the test, all schefflera (Green Emerald) and buttonwood (Green) plants were rigorously inspect ed to be certain that they were clear of thrips. Twelve eight month old Knock Out Radrazz roses originally purchased from Treemendous Landscapes (Homestead, Florida) were randomly selected from an established research plot heavily infested by S. dorsalis (with three S. dorsalis per square centimeter in flowers), and placed in an outdoor growing area on the TREC campus. T hese roses were arranged in fo ur replicates with clusters of three rose s whose canopies overlapped per replicate Each cluster of three rose plants served as a source point for S. dorsalis populations by surrounding them with uninfested alternate host plants (schefflera and buttonwood). One schefflera and one buttonwood were placed at each of the cardinal points at a distance of one meter from the roses. All plants remained exposed to ambient environmental conditions (mean temper ature of 80 0.3 throughout the experiment, and were irrigated with 93 15 mL water using an overhead system running for half an hour at six in the morning and two in the afternoon. All plants were fertilized mont hly with an application of 3.96 grams of

PAGE 41

41 liquid fertilizer (Peters Professional, Scotts Company, Marysville, Ohio) (N H3: NO3: (NH2)2CO: P2O5: K2O = 4:6:10:20:20) and 7.5 grams of pellet fertilizer (NH3: NO3: P2O5: K2O = 15:9:12) (Osmocote Plus, Scotts Co mpany, Marysville, Ohio) once every other month. Population Estimates on Two Hosts Thrips p opulation estimates for buttonwood and schefflera were assessed by randomly sampling one terminal bud from each plant once per week. Plant samples were consistentl y selected from similar positions within each plants canopy, and were of similar age and development. All samples were removed with pruning shears and sealed immediately in small plastic containers with two drops of 95% ethanol. Samples that could not imm ediately be processed were stored at 6 C. The total number and life stage s of thrips in a sample was recorded by washing the samples with 75% ethanol and pouring the runoff through a 25 m mesh as per Seal and Baranowski (1993), and then examining the collected debris under a dissectio n scope at a minimum of 12x magnification, with additional magnification and slide mounting as necessary. The Funderburk et al. (2007) key to species of ornamental thrips was used as a primary aid in identification. S cirtothrips dorsalis could rapidly be d ifferentiated from other thrips present in samples by their relative size and other distinguishing characters, including antennal, pronotal, and abdominal characters. P lant samples were then dried overnight and the area of each dried sample was determined using a leaf area meter (LI 3000, LiCor, Lincoln, Nebraska) to determine the density of thrips populations per unit area (thrips/cm2). To develop an estimate of thrips flights among host plants yellow 7.6x12.7 cm sticky cards were placed on posts six cen timeters behind each pot of the schefflera and buttonwood plants The sticky cards were positioned at one meter in height. Sticky cards were replaced weekly. Cards were removed, wrapped in saran wrap, labeled, and stored at not be immediately processed.

PAGE 42

42 Plant quality and feeding damage was rated weekly, using a damage scale similar to that describe d by Krishna Kumar et al. (1996) for pepper. Plants were rated from zero, no symptoms, to five, defoliated. A Florida Automated Weather Network (FAWN, http://fawn.ifas.ufl.edu/) located a hundred meters from the research plot was used to collect wind speed and direction data throughout the experiment. Samples were collected weekly for eleven weeks from 19 July 2007 until 27 September 2007, and voucher specimens were sent to the Florida D epartment of Plant Industry to confirm species identification throughout the experiment. Daily Flight Behavior Daily thrips flight patterns were evaluated in a set of experiments conducted in greenhouses and field plots at TREC and at the University of Florida Mid -Florida Research and Education Center ( MREC ) in Apopka, Flor ida. The greenhouse at TREC contained chili pepper and cotton, two key hosts of S. dorsalis New chili peppers were planted and placed within the greenhouse every other week to maintain the colony of thrips. The greenhouse at MREC contained other hosts of the thrips including Vaccinium corymbosum L. (Northern highbush blueberry ), cotton, and chili pepper. Four yellow 7.6x12.7 cm sticky cards were placed in each greenhouse on posts at a meter and a half in height, and within one -half meter of one of the known host plants in the greenhouse Preliminary experiments ( A.I.D. unpublished data ) did not detect nighttime flights for this species of thrips, so sticky cards were replaced every hour during daylight, beginning one hour before dawn at six AM and ending one hour after dusk at nine PM An electronic sensor (HOBO H8 Pro Series, Onset Computer Co rporation, Bourne, MA) recorded temperature and relative humidity in each greenhouse throughout the experiment.

PAGE 43

43 Small field p lots of twenty infested Knock Out Rose grown in 11.36 liter containers at MREC and TREC were also used in this experiment. Four sticky traps were placed on posts at approximately half a meter in height a third of a meter from the inside corners of each plot. These cards were collected and replaced every hour during daylight, beginning one hour before dawn at six AM and ending one hour after dusk at nine PM, using a protocol identical to the greenhouse experiment. Collected sticky cards were stored in an identical fashion to the previous ex periment, and S. dorsalis were identified and counted using the same magnification and keys as in the previous experiment. The FAWN stations w ere used at both field sites to record environmental variables including temperature, humidity, and wind speed A nalysis Analysis for the plant to -plant movement experiment focused on the number and stage of thrips and the observed feeding damage for each host plant as well as the adult trap catch A two way ANOVA was performed to determine if there was interaction between host and direction for all variables tested. Treatments were compared using the ANOVA function of JMP ( SAS Institute 2007), and means were separated using Students t test or Tukeys HSD when sample n was not constant for all treatments (due to los s of sticky cards by birds or weather). Degree hours began accumulation at six AM when traps were first placed. Analysis for the daily flight pattern focused on the number and sex of thrips collected each hour. Least square means were used to fit the gener al linear model and determine a relationship between wind direction and thrips captures. Results In the plant to -plant movement experiment, no significant interaction was observed between host plant and direction of dispersal for any variable tested, incl uding damage rating, the total number of thrips of any instar on plant, thrips density, and the number of thrips captures on sticky traps The data were pooled for analyse s.

PAGE 44

44 Stadia The number of nymphs on the plants did not vary significantly throughout the experiment, with the exception of the second week of the experiment (25 August 2007), where significantly more nymphs were collected on buttonwood than for all other sample dates (F = 2.01; df = 10, 313; p = 0.0321) ( Figure 2 1 ). The number of adult th rips captured on plants did not vary significantly from week to week ( F = 1.64, df = 10, 313; p = 0.0951), but the number of thrips captured on sticky cards varied significantly with time ( F = 9.21; df = 10, 274; p < 0.0001). There was no correlation betwe en the number of nymphs or adults captured on the plant and the number of adults captured with flight traps (Figure 2 2 ). Host Plant Damage Although the amount of damage observed on either host plant was very low (< 1), with only minor scarring at the ba ses of petioles and wrinkled new growth, there was a s ignificantly higher damage rating f or schefflera than buttonwood (t = 1.97; df = 350; p < 0.0001) ( Figure 2 3 ). There were n o significant difference s for total numbers of S. dorsalis sampled from the tw o host plant s, but there was a significantly higher density of thrips (thrips/ cm2) on schefflera than on buttonwood ( t = 2.03; df = 182.6717; p < 0.05) ( Figure 2 4 ), and significantly more adult thrips were captured on sticky traps placed behind buttonwood than behind schefflera (t = 3.07; df = 245.0794; p < 0.003) ( Figure 2 5 ). Direction No significant difference was observed for damage rating, total number of thrip s, or thrips density on the host plants located in the four cardinal directions from the infested rose plants. However, a significantly larger number of thrips were captured on western sticky cards ( t = 1.97; df = 281; p < 0.0001), which are in the direction of the prevailing winds (Figure 2 6 ). The mean

PAGE 45

45 wind direction was observed from 1132. 18 (MSE) degrees, from east by slightly southeast to the west by slightly northwest. Daily Flight Behavior Hourly flight activity was measured in two environments (the field and greenhouse) at two locations in south and central Florida (TREC and MREC). S ignificant differences were detected between the two locations ( F = 19.36; df = 1; p < 0.0001), and between the two environments ( F = 101.53; df = 1; p < 0.0001), with a greater number of thrips collected per hour at MREC and in greenhouses. An interaction effect was detected between locations and environment ( F = 19.90; df = 1; p < 0.0001), and so each population was considered independently. T here was no significant difference between hourly thrips captures for field data at TREC or MREC (t = 0.669; df = 706; p = 0.5035) H owever, there was a significant difference in the hourly thrips captures between TREC and MREC greenhouses ( t = 3.17; df = 298; p = 0.0008) at ten AM ( t = 2.17; df = 18; p = 0.0228) and seven PM ( t = 2.57; df = 18; p = 0.0097) ( Figur e 2 7 ). A t a ll four sites there was a rapid increase in captures early in the morning, with peak captures occurring at approximately noon, which then decline d to a constant but low number of flights until sunset. Sites differ ed by what time the daily peak was reached, and the number of thrips captured at each time of day. The greenhouse at MREC and the field site at TREC both reached their peak by ten in the morning, the TREC greenhouse peaked at noon, and the MREC field site peaked at two in the afternoon just prior to daily rainstorms. The TREC greenhouse site had two percent higher humidity than MREC greenhouse ( t = 2.05; df = 278; p = 0.0205) at 75.6%, and was on average four degrees warmer ( t = 5.99; df = 278; p = 0.0001) at 86.0 te was also four percent more humid than the MREC field site ( t = 3.79; df = 690; p = 0.0002) at 76%. Afternoon rainstorms in Apopka

PAGE 46

46 reduced the effectiveness of hourly sampling at MREC, and no thrips were collected during periods of rain. Temperature was significantly correlated with thrips captures for the TREC field site ( r = 0.67; p = 0.0029), the TREC greenhouse ( r = 0.80; p = 0.0002), and the MREC field site ( r = 0.54, p = 0.0205). Across the range observed, increasing temperatures t ended to result i n an increase in the mean number of hourly thrips captures. Relative humidity was significantly correlated with thrips captures at the TREC field site ( r = 0.64; p = 0.0048) and greenhouse ( r = 0 .80 ; p = 0.0002). With increasing humidity, less thrips cap tures were observed. After observing this correlation, flight captures were plotted by degree hours for all four locations ( Figure 2 8 ). Morning flight peaks occurred at all four sites after the accumulation of three to four hundred degree hours. A second peak occurred at the MREC field site after the accumulation of approximately seven hundred degree hours just before a rainstorm. Discussion As S. dorsalis is a minute insect, it was not surprising that the greatest number of flight captures occurred in the direction of the prevailing winds. Thrips low mass and the high surface area created by their fringed wings allow them to easily be carried by the winds (Lewis 1973). However, most local dispersal for minute insects occurs within the boundary layer ne ar to the ground (Taylor 1974). The plant canopy protects this layer of relatively immobilized air, and small insects may fly short distances through this space undisturbed. Populations of S. dorsalis found in St. Vincent (Seal et al. 2006) support ed this hypothesis Within a large plot, researchers observed aggregated population clusters randomly distributed throughout the field, suggesting an initially random colonization event within the crop followed by gradual growth and local expansion to proximate ho sts.

PAGE 47

47 The fact that neither host plant received a damage rating greater than one, or a mean abundance greate r than two thrips per terminal suggests that these plants may not be key hosts in the preference hierarchy of this thrips when compared to the Knock Out rose which was used as the source point of thrips This was surprising, given the poor health of heavily infested plant samples of the same species received at extension centers. This discrepancy might be explained by the choice of infested source plants used in this experiment. It is probable that the infested Knock Out roses used as a source population are more attractive hosts to S. dorsalis than schefflera and buttonwood, and that populations chose to remain on the source host rather than migr ate to alternate hosts. It was assumed that as population densities built up on the rose that the thrips would migrate to alternative hosts where there might be less competition or a greater concentration of nutrients. Recent research ( A.I.D., unpublished data ) suggests that the Knock Out rose can support a large infestation of thrips before reaching critical population thresholds that result in plant death and increased dispersal from the host. Higher dispersal rates might have been achieved by reducing the quality of the rose source host. It is also possible that the damaged plant samples received from nurseries were growing in a greenhouse where an enclosed and growing population of thrips had nowhere else to disperse to, or that they had been heavily fertilized to push plant growth, making them a more nutritionally attractive host than tested plants. In spite of reduced dispersal from the source plants, significant differences remained between the two hosts tested in thrips density, thrips damage, and flight captures. Schefflera had higher damage ratings and thrips densities, but more thrips were captured on sticky traps near the buttonwood than the schefflera One interpretation of the results might be to suggest that schefflera is a more attractive ho st to dispersing S. dorsalis than buttonwood. The higher thrips

PAGE 48

48 density on schefflera may represent a larger total population on the whole plant. It is also possible that plant architecture has resulted in slightly different captures. The canopy of buttonw ood may be looser and allow greater airflow around the host, while the denser canopy of schefflera may act to trap dispersing thrips, as a large hedge used as a windbreak in a crop situation (Lewis 1997). This would also explain the higher numbers on sticky cards behind buttonwood and the greater density of thrips on the schefflera Records of hourly flights at the four test sites have reveal s information about the efficacy of sampling methods in different locations and local thrips populations. The great er numbers collected per hour on greenhouse traps was not surprising, given that the thrips within the greenhouse were operating within an enclosed space with a limited volume within which to disperse. These greenhouses also lacked the strong air currents created by wind in the field setting, and as such a larger number of thrips were probably able to land upon and become ensnared by traps. This suggests that the use of sticky traps as a monitoring device in the field by growers is probably less cost effect ive than in greenhouse environments. T o be effective monitoring devices, field traps should probably be exposed for longer periods of time than one hour The early morning flight peak followed by a low but continuous number of flights is a pattern common to thrips. Thysanoptera are small poikilothermic insec ts requiring the accumulation of a certain number of degree hours to heat flight muscles before takeoff (Ellington 1980), and are then carried over a great distance by the rise and fall of daily convect ion currents (Lewis 1973). It appears that the majority of S. dorsalis that take flight do so as soon as they are able, resulting in an increasing number of flights that peak towards the middle of the day, and then a gradual decline in flights as f ewer mem bers of the population take flight.

PAGE 49

49 The negative relationship observed between increasing humidity and flight captures may suggest a behavior of flight avoidance on rainy days. Both rainfall (Lewis 1973, Varadharajan and Veeravel 1995, Saxena et al. 1996, Lewis 1997, Lingeri et al. 1998, Paul and Konar 2005, Singh 2005) and increasing humidity (Varadharajan and Veeravel 1995, Saxena et al. 1996, Lingeri et al. 1998, Paul and Konar 2005) have been documented to reduce the rate of population growth. It is possible that thrips have adapted to avoid flight when environmental conditions are suggestive of approaching rain fall This information will be important to continuing an integrated pest management program for this insect. Growers will quarantine plants do wnwind of their crops, and will be aware to increase their inspections downwind of infested plants. The management community will also be prepared to spray plants either early in the morning or late in the evening before or after the thrips becomes active and might leave the plant long enough to avoid the effective period of the pesticide. Acknowledgments The authors would like to thank S. NaLampang for his suggestions and advice in data analysis, Ms. M. Peterson and Dr. L. Osborne for sample collection in Apopka, and Drs. D. Seal and L. Osborne for use of their greenhouse and field plots to conduct experiments.

PAGE 50

50 Figure 2 1 Mean weekly S. dorsalis nymphs plus standard error on buttonwood foliage (* = significantly different by pairwise t -test compariso ns at a = 0.05).

PAGE 51

51 Figure 2 2 W eekly flight captures of S. dorsalis plus standard error on sticky cards behind Conocarpus erectus and Schefflera arboricola and weekly samples of S. dorsalis plus standard error washed from plant terminals on both hosts

PAGE 52

52 Figure 2 3 Mean damage rating plus mean standard error for Conocarpus erectus and Schefflera arboricola (* = significant difference by ANOVA with a = 0.0001).

PAGE 53

53 Figure 2 4 Mean thrips density ( S. dorsalis /cm2) plus mean standard error for Conocarpus erectus and Schefflera arboricola (* = significant difference by ANOVA at a = 0.05).

PAGE 54

54 Figure 2 5 Mean captures of S. dorsalis plus mean standard error on yellow sticky cards placed behind Conocarpus erectus and Schefflera arboricola (* = significant difference by ANOVA at a = 0.05).

PAGE 55

55 Figure 2 6 Mean S. dorsalis captures plus mean standard error on ye llow sticky cards by direction (* = significant difference by pairwise t tests at a = 0.0001).

PAGE 56

56 Figure 2 7 Mean hourly thrips captures plus stand ard error for greenhouse and field sites at TREC and MREC (* = significant difference between sites determined using ANOVA at a = 0.05).

PAGE 57

57 Figure 2 8 Mean thrips captures by degree hour for greenhouse and field sites at TREC and MREC.

PAGE 58

58 CHAPTER 3 POPULA TION DYNAMICS OF Scirtothrips dorsalis HOOD (THYSANOPTERA: THRIPIDAE) ON TWO ORNAMENTAL HOSTS IN SO UTHERN FLORIDA Introduction Scirtothrips dorsalis Hood (1919) is a minute insect pest whose global distribution has expanded rapidly over the last twenty yea rs (Morse and Hoddle 2006). It was first identified in the United States at a Palm Beach nursery in 2005, and has since become established in southern and central Florida (Silagyi and Dixon 2006). It is a polyphagous thrips with a broad host range includin g many ornamental and vegetable hosts of economic importance (Mound and Palmer 1981, Venette and Davis 2004, Nietschke et al. 2008). Recent reports suggest that this thrips has the potential to expand from Florida to occupy much of the southeast, the Gulf coast region, and the western coast of the United States (Venette and Davis 2004, Nietschke et al. 2008). The success of an invasive species depends on its ability to find an ecological niche that it can exploit successfully (Crawley 1986). As a polyphago us thrips, S. dorsalis has been documented as a serious pest on fruit and vegetable hosts such as Fragaria x annanasa Duchesne (strawberry) Mangifera spp. L. (mango), Capsicum spp. L. (pepper) Citrus spp. L. Gossypium spp. L. (cotton), Glycine max (L.) Merr (soy bean) and Arachis hypogaea L. ( peanut ) The fruit and vegetable crops potentially affected by this thrips added nearly 1.4 billion dollars to the Florida economy in 2005 (Jerado 2006). For many hosts, damage from this thrips results in reduced fl oral and vegetative output, increased plant stress, scarred foliage, and potentially death (Mound and Palmer 1981, Tatara and Furuhashi 1992, Shibao 1996a). Ornamental plant hosts are particularly sensitive to feeding damage, as even minimal scarring can r ender a plant unsalable Among these host plants are popular landscaping choices including Rosa spp. L. Raphiolepis indica (L.) Lindl. ex Ker Gawl. (Indian hawthorn), Ligustrum spp L. (ligustrum) Viburnum spp. L. (viburnum), and Jasminum multiflorum Lind l. (star jasmine)

PAGE 59

59 Initial reports cited the Knock Out Rose Rosa Radrazz (Radler 2001) as an important host for this thrips in the Florida landscape environment (Silagyi and Dixon 2006), and it appears capable of sustaining large populations of this pest for long periods of time befor e succumbing to pest pressure. Green buttonwood, Conocarpus erectus var erectus L., is another common landscape plant in southern Florida that has been reported with severe thrips damage from S. dorsalis L andscapers make frequent use of both species as ornament al hedges, and so b oth plants are common and have a broad distribution in southern and central Florida. These plants, as well as others, could potentially provide the thrips a natural corridor for establishment throughout the state (Morse and Hoddle 2006). As a group, thrips have demonstrated great adaptive potential for pesticide resistance. Anecdotal reports indicate that certain populations of this species may have already developed limited resistance to neonicot inoid pesticides. The cost to develop novel pesticides is high, and this has resulted in increased expense to management programs stressing chemical maintenance of crops. This underlines the need to develop additional methods of non -chemical management, or to improve the specificity and use of existing strategies. As this thrips is a pest of ornamental and agricultural crops it may develop resistance to commercial pesticides in one area, and then spread to other locations (Kirk and Terry 2003). G rower s and homeowner s must simultaneously apply a comprehensive integrated pest management program that considers both of their unique environments to prevent the establishment of reservoir populations or the development of resistance to commercial pesticides. Know ledge of peak seasons of population increase is critical to the development of a proactive management program, as thrips prevention is frequently less difficult than eradication

PAGE 60

60 or control (Lewis 1997). An awareness of this thrips seasonal presence upon c ommon ornamental plant s will be valuable in constructing a fully integrated pest management program. While much research regarding the control and management of this thrips has begun, there is no literature available on the seasonal population dynamics of this thrips in the Florida environment. This study explores the annual populat ion dynamics of this thrips on Rosa Radrazz, and Conocarpus erectus var erectus two structurally different hosts under ambient environmental conditions in southern Florida. Materials and Methods The S. dorsalis used in this study were obtained from naturally established populations local to southern Florida, and from a colony raised and maintained by Dr. Dakshina Seal at the University of Florida Tropical Research and Educati on Center ( TREC ). T wenty green buttonwood ( Conocarpus erectus var erectus L.) were grown from cuttings harvested at the TREC, and t wenty Knock Out Roses (Rosa Radrazz PP#11836) were purchased from Treemendous Landscapes (Homestead, Florida), and place d in 11.36 liter containers in a medium consisting of 50% Canadian sphagnum peat moss, 25% processed pine bark, and a mixture of perlite and vermiculite (Fafard 3B Mix, Conrad Fafard Inc., Agawam, Massachusetts). These plants were infested with naturally established populations of S. dorsalis beginning on 12 July 2007. Each species plot consisted of a four by five grid with each plant separated from their neighbors by a third of a meter. All plants remained exposed to ambient environmental conditions throughout the experiment, and were irrigated with 93 15 mL using an overhead system at six in the morning and two in the afternoon, according to local water restrictions and extension office recommendations. All plants were fertilized monthly with a n applicat ion of 3.96 grams of liquid fertilizer (Peters Professional, Scotts Company, Marysville, Ohio) (NH3: NO3: (NH2)2CO: P2O5: K2O = 4:6:10:20:20), and 7.5 grams of pellet fertilizer

PAGE 61

61 (NH3: NO3: P2O5: K2O = 15:9:12) (Osmocote Plus, Scotts Company, Marysville, O hio) once every other month. On -plant Measures The plants were sampled from 12 August 2007 to 4 September 2008. The total number of flowers and buds were counted weekly on rose s A damage estimate similar to that described by Krishna Kumar et al (1996) f or pepper was developed during preliminary observations and used to visually rate the feeding damage of plants each week ( Table 3 1 ). Thrips population estimates for the two hosts we re assessed by sampling an organ from each plant in e very other week, alt ernating between plots of buttonwood and rose. Buttonwood samples always included five -centimeter sections of the terminal bud. As roses presented more complex structural morphology, different plant organs were selected based on the availability and devel opment of individual terminals. Flowers of approximately one week in age were always the preferred choice, but if none were available, rosebuds approximately five days old were selected. If no rosebuds were available, a rose growth terminal was taken. If n o terminals were present, then a mature compound leaf was sampled. All samples were removed with pruning shears and sealed immediately in small plastic containers with two drops of 95% ethanol. Samples that could not immediately be processed were stored a t 6 C. T he total number and life stage s of thrips in a sample was recorded by washing the samples with 75% ethanol and pouring the runoff through a 25 m mesh as per Seal and Baranowski (1993), and then examining the collected debris under a dissection scope at a minimum of 12x magnification. P lant samples were then dried overnight and the area of each dried sample was determined using a leaf area meter (LI 3000, LiCor, Lincoln, Nebraska) to determine the density of thrips populations in each sample (thr ips/cm2).

PAGE 62

62 Flight Measures Four 7.6x12.7 cm yellow sticky traps (Chu et al. 2006) were hung at approximately half a meter in height on posts placed approximately half a meter within the inside corner of each plot t o develop an estimate of the thrips popul ation in flight throughout the experiment Sticky traps were replaced weekly. C ollected traps wrapped in clear plastic, labeled, and stored at 6 C if they could not be immediately processed. Thrips Identification All thrips were identified using a diss ection scope at a minimum of 12x magnification, with additional magnification and slide -mounting as necessary. Thrips were identified using the keys provided in Funderburk et al (2007), and Moritz (2001, 2004) when possible. Samples of unknown thrips and voucher specimens were collected and s ent to the Florida Department of Plant I ndustry for confirmation and for further identification. Scirtothrips dorsalis could rapidly be differentiated from other thrips present in the environment by their relative size and other distinguishing characters, including antennal, pronotal, and abdominal characters. Environmental Factors The Homestead station of the Florida Automated Weather Network (FAWN, http://fawn.ifas.ufl.edu/) was located a hundred meters from the res earch plot, and was used to provide environmental data including wind speed and direction, relative humidity, and the ambient temperature at 60 c enti m eters throughout the experiment. Analysis Thrips density was transformed to homogenize variance and norm alize data for analysis by adding one to the total number of thrip s, and then applying the natural log. Host plots were compared using ANOVA and means were separated using t tests or Tukeys HSD when sample n was not constant for all treatments (due to los s of samples by birds or weather). Degreedays

PAGE 63

63 began to accumulate from 12 July 2007, when plants were first exposed to natural infestations. Least square means were used to fit the general linear model and determine the strength of relationship s between t hrips populations, hosts, and environmental factors. All statistical analysis was conducted using JMP (SAS Institute 2007). Results On -plant Measures Sampling resulted in the collection of 560 buttonwood terminals, 470 rose flowers, 131 rosebuds, 41 rose terminals, and 31 rose leaves. Only buttonwood terminals, rose flowers, and rosebuds were collected on all sampling dates. The grid that plants were placed in for each plot was analyzed as separate rows running from north to south, and as columns running from east to west to determine the spatial distribution of the thrips within the plot. No row or column of plants within the plot of buttonwood ( F = 0.54; df = 19, 540; p = 0.9449) or rose ( F = 0.55; df = 19, 672; p = 0.9377) was significantly different f rom any other for total number of thrips or thrips density throughout the year. The total number of thrips collected from the wash samples for both plant hosts was 21,304 thrips, and 95.2% of all thrips collected were S. dorsalis (Table 3 2 ). While 20,610 thrips (96.74% of all thrips) collected occurred on rose, there was no significant difference in the proportion of S. dorsalis relative to other species collected on rose or buttonwood, and all life stages of S. dorsalis were present on both hosts year ro und. The total number of thrips ( t = 13.82; df = 1232; p < 0.0001) and thrips densities ( t = 1831.21; df = 1232; p < 0.0001) for S. dorsalis were significantly different by host, with an average of almost twenty -eight more (2.25 per square centimeter) S. dorsalis counted on rose samples than on buttonwood.

PAGE 64

64 Plant organ was also found to have a significant effect upon the total number of thrips ( F = 120.27; df = 3, 1229; p < 0.0001) ( Figure 3 1 ) and thrips density ( F = 74.35; df = 3, 1229; p < 0.0001) ( Figu re 3 2 ) on the hosts, with rosebuds and rose terminals having the greatest density of thrips, followed by rose flowers, rose leaves, and buttonwood terminals. Because of these distributional differences, plant organs will also be considered separately. Ap proximately 67.2 3.69% of all adults collected were female, and this proportion was not significantly different between hosts ( t = 0.19; df = 638; p = 0.8462) or plant organs ( F = 1.23; df = 3, 636; p = 0.2980). The proportion of females fluctuated sign ificantly among rose over time without an obvious trend ( F = 2.30; df = 25, 425; p = 0.0004) with a range of 28.57% to 100% and an average of 65.32 8.43%. The proportion of females on buttonwood was not affected by time ( F = 1.13; df = 24, 164; p = 0.3143), but ranged from 0% to 100%, with an average of 65.48 19.04%. Thrips density fluctuated significantly without any directional trend over time for most plant organs, including buttonwood terminals ( F = 8.35; df = 26, 533; p < 0.0001), rose flowers (F = 15.55; df = 27, 442; p < 0.0001), rosebuds ( F = 3.64, df = 26, 104; p < 0.0001), and rose terminals ( F = 0.06; df = 10, 30; p = 0.0003). Rose leaves ( F = 2.37, df = 4, 26; p = 0.0782) did not change significantly over the sampling period. However, as onl y buttonwood terminals and rose flowers were collected on all sample dates, only these parts were used to assess changes in the thrips population over time ( Figure 3 3 ). Population densities on both hosts changed throughout the year, with peak density occ urring in November through December of 2007 and July of 2008 for buttonwood, and peaks in August and October of 2007 and January and July of 2008 for rose. Increasing mean rose flower densities were significantly correlated with increasing rosebud densitie s ( r = 0.53; p =

PAGE 65

65 0.0028). Rose flower densities were also significantly correlated with buttonwood terminal densities, but the relationship was not as strong ( r = 0.37; p = 0.0384). Significant differences in the proportion of adult S. dorsalis were detected for host and plant organ, with buttonwood having the largest proportion of adults and rose leaves the smallest number of adults relative to the number of nymphs ( Table 3 3 ). Using the rose flower data, the proportion of the population that was adult ch anged significantly with time ( F = 3.98; df = 27, 431; p = 0.0001), ranging from 8.6% to 54% with a mean of 20.3 1.8%. This value was inversely proportional to the total size of the population at that time. The proportion of adult population also changed significantly with time on buttonwood ( F = 1.56; df = 26, 243; p = 0.0459), ranging from 3.7% to 100% with a mean of 47.7 4.2%, and this percentage was also inversely proportional to the size of the population. While the number of nymphs captured in sam ples ranged from a mean of 5.13 to 65.14 nymphs, the number of adults only ranged from a mean of 0.53 to 25.88 adults. This gap was proportionately larger on buttonwood than on rose. While damage ranged from a rating of zero to two (moderate damage), on buttonwood, no significant relationship could be detected between damage and the total number of thrips or thrips density. The small number of rose leaves and rose terminals collected prevented a significant comparison of damage ratings to population estim ates by parts sampled. As rose flowers were collected on all sample dates, and displayed the most visible elements of floral damage, they were used to compare plant damage to populations of S. dorsalis Roses experienced damage ratings ranging from zero to four (critical damage) ( Figure 3 4 ). Population estimates were lower (but not significantly) for roses at the more severe end of the scale than at moderate damage. A significant but weak positive correlation was observed between the total number of thrips and

PAGE 66

66 damage estimates ( r = 0.29; p = 0.0001), and a significant but somewhat stronger correlation was observed between thrips densities and damage ratings ( r = 0.43; p = 0.0001). There was also a significant but weak correlation between increasing damage a nd a reduction in the total number of flowers and buds produced ( r = 0.15; p = 0.0009) and the mean floral area ( r = 0.33; p = 0.0001). There was no significant relationship between the total number of thrips, density, and the total number of flowers and buds produced. While the mean number of flowers and buds decreased by an average of four flowers as damage ratings increased ( Figure 3 5 ), the mean area of each inflorescence decreased by almost half ( Figure 3 6 ). There was a much stronger correlation for both hosts when biweekly samples were pooled to develop a monthly average for populations and damage. The positive correlation between increasing thrips density and damage could be seen over time for buttonwood ( r = 0.90; p = 0.0001) ( Figure 3 7 ), with pea ks in density and population occurring from December 2007 to January 2008, and again in July of 2008. On rose, peaks in population density and damage occurred in August and October of 2007, and again in January and June of 2008 ( r = 0.78; p = 0.0003) (Figure 3 8). However, no significant correlation could be found between buttonwood and rose populations, suggesting that the two populations are independent of one another. Monthly averages for buttonwood and rose flowers were used also instead of biweekly av erages to relate thrips populations to ambient environmental conditions. On rose, no significant correlations were observed between thrips population estimates and environmental factors. However, positive relationships between mean monthly thrips density a nd mean temperature ( r = 0.42; p = 0.0679), minimum monthly temperatures ( r = 0.45; p = 0.0509), and mean monthly degree -days ( r = 0.43; p = 0.0614) were nearly significant. For buttonwood, a weak but significant negative relationship was observed between thrips density and maximum monthly

PAGE 67

67 temperature ( r = 0.52; p = 0.0412), and between mean the total number of thrips and average monthly maximum wind speed ( r = 0.53; p = 0.0358). Flight M easures No significant differences in the number of adults were detected among sticky traps based on their location within the plot, so all trap data could be pooled. S cirtothrips dorsalis proved to be the predominant thrips collected on yellow sticky traps above both host plots, representing 84.04% of the 31,785 thrips ca ptured. The next two species accounting for the second and third largest proportions were Frankliniella schultzei (Trybom) at 3.59%, and Gynaikothrips uzeli Zimmerman at 3.05% of all captures ( Figure 3 9 ). G. uzeli is the only thrips that occurred more fre quently on sticky traps placed in the buttonwood plot than in the rose plot. Most of the thrips on the sticky traps (83.84%) were captured over rose. S. dorsalis represented 75.39 22.96% of all captures in any given week, and were a significantly higher proportion of weekly captures above rose (78.81 17.96%) than buttonwood (71.64 26.97%) ( t = 3.29; df = 367.2377; p < 0.005). Although the total captures for other species were low relative to the total, certain fluctuations over time were observed for particular genera and species. Genus Frankliniella showed peaks occurring for all three species identified in July 2007, and April through May 2008. F. schultzei showed additional peaks in December 2007, and in July 2008 (Figure 3 10). G. uzeli showed sma ll peaks from August to October 2007, and from February to March 2008. A large number of G. uzeli were captured in July 2008 (Figure 3 11). A peak of Thrips palmi Karny was observed from August to October 2007, and numbers of T. tabaci Lindeman were low th roughout the year (Figure 3 12). A peak of Merothrips spp. was observed from January to February 2008 (Figure 3 13). A peak of Echinothrips americanus Morgan was identified from July to August 2007 (Figure 3 14). Several small peaks of Chirothrips spp. wer e observed from

PAGE 68

68 August to September 2007, from November of 2007 to February 2008, and May to August 2008 (Figure 3 15). A small peak of Heliothrips haemorrhoidalis (Bouch) was observed in May 2008 (Figure 3 16). A small peak of Selenothrips rubrocinctus w as observed in August 2007 (Figure 3 17). Chaetanaphothrips orchidii (Moulton) were present in low but constant numbers throughout the year (Figure 3 18). A small peak of Holopothrips inquilinus (Bournier) was observed from March to June 2008 (Figure 3 19) The predatory thrips, Scolothrips sexmaculatus (Pergande) was the most abundant predatory thrips present, with nineteen individuals intercepted throughout the year (Figure 3 20). Other species of thrips were captured, but as less than fifteen individual s of each species were collected, and some specimens were incomplete or damaged, they could not be identifi ed to species. This group totaling sixty -six thrips were summed together and recorded as unidentified genera (Figure 3 21). This group includes put ative individuals from the genera Aeolothrips Aleurodothrips Franklinothrips and Karnyothrips with no obvious peaks for any individual thrips collected. S. dorsalis adults occurred year round over both buttonwood and rose hosts, with sticky traps over rose capturing significantly more thrips than traps placed over buttonwood ( t = 10.64; df = 461; p < 0.0001). The number of thrips captured changed significantly throughout the year for both hosts ( F = 7.09; df = 59, 403; p < 0.0001), with peak captures o bserved over buttonwood in May and June of 2008, and peak captures observed over rose in August, October, November of 2007 and June 2008. Both rose and buttonwood had their lowest captures from January to early May in 2008 ( Figure 3 22), and showed signifi cant and strong correlation on thrips captures between hosts ( r = 0.78; p = 0.0006).

PAGE 69

69 There was no statistical difference between the proportion of S. dorsalis females captured over buttonwood and rose ( t = 0.04; df = 377; p = 0.9656), but the proportion of females captured over hosts changed significantly over the year ( F = 3.32; df = 47, 331; p < 0.001) ranging from 27.1% in August 2008 to 77.2 % in January 2008 with an average of 49.3 1.6%. There was a positive significant relationship between monthl y mean flight captures and monthly mean temperature ( r = 0.47; p = 0.0459), monthly minimum temperature ( r = 0. 5 4; p = 0.0213), and mean monthly degree days ( r = 0.48; p = 0.0413). No significant relationship was detected between mean monthly buttonwood ca ptures and any environmental factor. Comparing Flight Measures to On -plant E stimates Thrips captures on sticky traps above rose were far more diverse than thrips collected from rose, with only S. dorsalis F. bispinosa, F. occidentalis F. schultzei, and T. palmi showing more than one on -plant capture during the experiment. Both rose ( r = 0.62; p = 0.0079) and buttonwood ( r = 0.50; p = 0.0408) showed moderate and significant correlation between increasingly mean monthly host density and greater mean monthly flight captures, but did not show significant or strong correlation between densities on one host and flight captures above the other ( r < 0.47; p > 0.0936). Both rose ( r = 0.62; p = 0.0079) and buttonwood ( r = 0.50; p = 0.0408) also demonstrated signif icant correlation with greater mean monthly sticky trap captures and higher mean monthly estimates of plant damage for that host, but did not show significant correlation between damage on one host and sticky trap captures from the other plot ( r < 0.06; p > 0.3480). The proportion of female thrips measured on plants (66%) was significantly different from the proportion of females captured (49%) on sticky traps ( t = 4.31; df = 24; p > 0.0001), but there was strong and significant correlation between an incr easing percentage of females on -plant and a larger percentage of females captured in flight ( r = 0.61; p = 0.0346).

PAGE 70

70 Discussion Examining the distribution of thrips within each plot showed that no row of plants from north to south, or column from east to w est of hosts within a plot was significantly different from any other row or column for the total number of thrips of any instar, or for thrips density. This suggests that the thrips have distributed themselves evenly across each host species of without ed ge effects. Weekly flight captures show the same findings, as down-wind traps did not show higher mean captures than those up -wind. This homogenous distribution is similar to the results observed by Rhainds et al. (2005), where adult female flower thrips m aintained a fairly uniform density within a plot of chrysanthemum. This uniform distribution between plants may help to reduce intraspecific competition on plants based on available plant nutrients, and to encourage more active dispersal. While the majori ty of thrips captured on traps and on both host plants were S. dorsalis a far greater diversity and abundance of thrips were captured on rose when compared to buttonwood. A greater diversity of thrips was observed on sticky traps than actually captured on plants. Most relevant to this study, significantly higher numbers and densities of S. dorsalis were recorded on rose than on buttonwood throughout the year, suggesting that the rose is a more attractive host for this pest. These results on plant s we re rep eated with the number of adults capture d in the traps a bove each plot of hosts In addition to the larger numbers of nymphs collected on rose, the relative proportion of adult thrips was greater on buttonwood than on rose, suggesting that this host is less advantageous for S. dorsalis reproduction. However, as the thrips did persist on buttonwood throughout the year, and nymphs were recovered from all buttonwood plants, it is possible for this insect to complete its reproductive cycle on this host. This su ggests that the pest can and will persist in an environment with less optimal hosts, and

PAGE 71

71 then migrate to preferred hosts when the opportunity becomes available, as observed by Masui (2007a). As has been suggested by many other researchers for this thrips, roses most probably represent a preferred host choice for S. dorsalis Given that roses have already featured prominently in extension reports of this thrips, it is probably best that future scouting continue to emphasize this host when surveying for this pest. It remains that buttonwood in isolation can still develop significant levels of infestation ( A.I.D., personal observations and unpublished data). In the absence of other more preferred hosts, the thrips can survive and reproduce on this substrate ye ar -round, and it has the potential to serve as both a reservoir and roadway for migrant populations of thrips moving between primary hosts. Together, these results imply that S. dorsalis has probably become well established in Florida, and even if this th rips were to be completely eliminated from agricultural or landscape environments, it will probably persist somewhere in the environment. These reservoir populations will gradually re -colonize lost territory from unmanaged areas, and a realistic management program would focus on controlling or limiting this thrips during periods of peak growth. Studies of the relationship between damage ratings and thrips populations might have been improved through the use of a more refined scale, allowing for a more spec ific description of damage using more quantitative variables, but the system does provide usable results. While results for buttonwood were inconclusive but suggestive of increasing damage with increasing thrips density, increasing thrips densities on rose resulted in an increase in both damage to the plant, and in the number of thrips dispersing about the plant. However, it should be noted that the relationship between measures of thrips and damage were not exclusively linear. The

PAGE 72

72 observations taken from r ose suggest that thrips populations are slightly smaller at the upper range of the damage scale than at middle values. This suggests that the thrips may have damaged the plant beyond its capacity to maintain maximum thrips populations, or that other factor s play a role in damage. As the greatest damage was measured during the hottest parts of the year, it is possible that the plant stress due to a high -temperature environment also plays a role in mediating damage to the plant. However, only ten plants were ever recorded with a damage rating of four, and all ten observations were made on the same sample date in late October. This coincided with a period during which thrips densities began to decline. It is therefore unlikely that the thrips density observed w ith this damage rating is representative of the populations on a heavily damaged plant. Interestingly, while damage to the plant increased with thrips populations and the total number of flowers and buds produced decreased significantly, the difference wa s a mean of four less flowers. However, the mean floral area of those flowers decreased by almost half. This suggests that while roses continue to produce florets at growth terminals, the nutrient resources available to do so may have been diverted by the thrips Growers concerned that they might have a thrips infestation should therefore be aware of the mean size of flowers more than the total number of flowers produced. Monthly population averages are more correlated to plant damage ratings than biweekly sample averages. This suggests that there is some lag between the level of population growth or decline and the display of damage or healing on the host. As such, use of the damage rating system may not be a good predictive measure by which to proactively estimate populations of S. dorsalis in order to preventatively apply pesticides. By the time a grower detects host damage, populations may already have grown too large for easy management.

PAGE 73

73 Flight populations also showed high correlation to population density on plants, but were only correlated with the host over which they were sampled. This is an important distinction, as it suggests that the populations captured above the host are probably related to the presence of that host and are not merely ambient populations created by some other significant source. This was not an unexpected result, but it is important to realize that indirect sampling of the thrips using regularly replaced sticky traps as a sentinel system might be able to detect and anticipate o utbreaks of S. dorsalis before they became too large and begin to significantly damage hosts in a growing system. Among the more unusual results observed was that the proportion of females measured on each plant was significantly greater than the proporti on of females captured in the air above each host. Masui (2007b) observed an increased proportion of female S. dorsalis dispersing at greater distances from preferred hosts. It is generally believed that male Thysanoptera have not been selected for dispers al, given their relatively small size when compared to females and their inability to independently establish a new colony (Lewis 1997). It is possible that this experiment only recorded local dispersal, as traps were only placed within the plot being stud ied. Had traps been placed at increasing distances from each of the two studied plots, an increasing proportion of female thrips might have been observed at greater distances, as per Masuis results (2007b). Males may be more active in local dispersal with in a plot in order to slightly increase the amount of gene flow within the population, and this would also help to explain our apparently inconsistent results. The low number of rose leaves and terminals collected prevented any realistic use of different plant organ s as a factor when estimating thrips populations, but the data is suggestive of certain trends. For all samples, S. dorsalis was found in significantly higher densities on the

PAGE 74

74 rosebuds and rose terminals than on rose flowers or leaves. The thrip s density of rose leaves was not significantly different from the leaves of buttonwood, or the density from the petals of rose flowers. This is not atypical behavior for thrips, as many Thripidae appear to prefer the upper part of the plant canopy and the outer extremities of their hosts (Reitz 2002, Hansen et al. 2003). It has been suggested that thrips prefer these areas because the flow of free nitrogen in plants is often towards these areas of new growth (Lewis 1997). This is not inconsistent with prior research regarding this thrips on chili pepper (Seal et al. 2006), castor (Raizada 1965), or tea (Dev 1964). Females of S. dorsalis prefer to lay their eggs inside of young leaves and buds at the apical meristem of plants, but as populations increase, wil l lay their eggs within the surface of mature leaves (Dev 1964, Raizada 1965). After hatching, nymphs will migrate from older leaves to the newer growth at terminals (Onkarappa and Mallik 1998). While thrips were not present in high densities or abundance s on the leaves, the nymphs left evidence of their presence through their distinctive feeding damage. More importantly, S. dorsalis was present on leaves in densities that were not significantly different than those within the petals of flowers. This shoul d not be surprising, as the petals of flowers are merely developmentally modified leaves distributed within a much smaller volume. This may help to explain a discrepancy in the literature where this thrips is sometimes referred to as a flower thrips, and i n other references is referred to as an exclusively foliar thrips. S. dorsalis may thrive equally well on both substrates, and exogenous environmental factors such as heavy precipitation, predation, or competition may determine the distribution of populati ons within the plant. There was some correlation between environmental conditions and populations of S. dorsalis Rose and buttonwood populations of thrips appeared to respond to environmental

PAGE 75

75 factors in a similar fashion. Prior studies have shown that po pulation levels are positively associated with increasing temperature (Tatara 1994, Varadharajan and Veeravel 1995, Saxena et al. 1996, Shibao 1996b, Paul and Konar 2005) and the number of sunlit hours in a day (Saxena et al. 1996, Gahukar 2003, Paul and K onar 2005). These results were also observed during this study, with the largest populations occurring in months with the greatest cumulative degree hours. As daily average and maximum temperatures were fairly constant, this suggests that the most variable measurement of temperature, the daily minimum, may be one of the most important factors defining potential populations of S. dorsalis in southern Florida. Other researchers have observed a negative association between population density and increasing relative humidity (Varadharajan and Veeravel 1995, Saxena et al. 1996, Lingeri et al. 1998, Paul and Konar 2005) and rainfall (Lewis 1973, Varadharajan and Veeravel 1995, Saxena et al. 1996, Lewis 1997, Lingeri et al. 1998, Paul and Konar 2005, Singh 2005). At least one researcher (Singh 2005) observed a positive correlation between population density and humidity, but their observations were made at a lower relative humidity than that of other sources cited. Results observed by this study are suggestive of t hese trends, but not conclusive. All researchers suggest that dry conditions with low intensity rainfall promote the best population growth for this thrips. It has been suggested that this is because hard rain will wash adults and nymphs from the leaves t o the ground, where many of them will not return successfully to their host plant for feeding after falling (Lewis 1997). During the period of observation in southern Florida, colder months were drier and warmer months tended to be rainier which may have m ade the effects of these factors less obvious. The predictive limitations of this portion of the study may also be a result of the limited range of environmental variability during the period of the experiment. The subtropical

PAGE 76

76 environment of southern Flor ida does not allow for much environmental variation across the environmental factors tested. A larger study incorporating a longer period of observation or several sites across Florida as a whole might help to refine these conclusions and better explore th e annual performance of this thrips.

PAGE 77

77 Table 3 1 S ymptoms of feeding damage prevalent at each stage of the infestation by Scirtothrips dorsalis for Conocarpus erectus var erectus L. (green buttonwood) and Rosa Radrazz (Original Knock Out rose). Score Conocarpus erectus Rosa Radrazz 0 No symptoms. No symptoms. 1 Scarring along veins of leaves. Scarring along sepals. One or two outer petals of each bloom display scarring. A few newer leaves may show stippling. 2 Scarring at base of petiole. Wrinkle d new growth New growth becomes wrinkled. Recent flush shows scarring along the base of veins. Scarring on most sepals and petals. Some mature leaves may show scarring. 3 Blackened new growth. Scarring on older leaves becomes prominent. Whole surface of newer leaves may be crisscrossed with scar tissue radiating outward from veins. Little new growth, and terminals may be blackened at tips. Shorter mean bud length relative to healthy plants, and smaller petal area when flowers open. 4 Leaves brown and wit hered. Little new growth. Leaves brown and wither, reduced production of flowers and buds No new growth. Mature leaves show significant feeding damage. Bud tips are small and blackened. 5 Defoliation. Defoliation and possible death of the plant. Table 3 2 Total thrips captures of each life stage on both host plants. S. dorsalis Other thrips Host Nymph Pupa Adult Total Nymph Pupa Adult Total Total Knock Out rose 16,203 412 3,000 19,615 707 12 275 994 20,609 Buttonwood 386 2 270 658 23 0 14 37 6 95 TOTAL 16,589 414 3,270 20,273 730 12 289 1031 21,304 Table 3 3 Proportion of adult Scirtothrips dorsalis on each plant organ. Plant organ n Mean (%) SEM Level Rose leaves 31 7.19 0.06 c Rosebuds 131 15.85 0.02 c Rose flowers 470 20.92 0.0 1 bc Rose terminals 41 31.32 0.04 b Buttonwood terminals 560 44.98 0.02 a Plant organ s not connected by the same letter are sign ificantly different by Tukeys at a = 0.05

PAGE 78

78 Figure 3 1 Biweekly m ean total thrips by plant organ plus standard error ( organs connected by the same letter are not significantly different by Tukey -Kramer at a = 0. 05).

PAGE 79

79 Figure 3 2 Mean thrips density (S. dorsalis /cm2) plus mean standard error by plant organ (organs connected by the same letter are not significantly diff erent by Tukey-Kramer at a = 0. 05).

PAGE 80

80 Figure 3 3 Biweekly m ean thrips density ( S. dorsalis /cm2) plus mean standard error per rose flower and buttonwood sample over time.

PAGE 81

81 Figure 3 4 Mean thrips floral density (S. dorsalis /cm2) plus mean standard er ror observed at each damage rating on Knock Out rose (ratings connected by the same letter are not significantly different by Tukey-Kramer at a = 0. 05).

PAGE 82

82 Figure 3 5 Mean total buds and flowers plus mean standard error for Knock Out rose observed a t each damage rating (ratings connected by the same letter are not significantly different by Tukey-Kramer at a = 0. 05).

PAGE 83

83 Figure 3 6 Mean floral area (cm2) plus mean standard error for Knock Out rose observed at each damage rating (ratings connected by the same letter are not significantly different by Tukey-Kramer at a = 0. 05).

PAGE 84

84 Figure 3 7 Monthly mean thrips terminal density ( S. dorsalis /cm2) and damage rating on buttonwood over time

PAGE 85

85 Figure 3 8 Monthly mean thrips floral density ( S. dors alis /cm2) and damage rating on Knock Out rose over time

PAGE 86

86 Figure 3 9. Total number of each species of thrips collected above each host plot from August 2007 to September 2008 using yellow 7.6x12.7 cm sticky traps.

PAGE 87

87 Figure 3 10. Weekly mean capture s of genus Frankliniella on yellow sticky traps placed above buttonwood and Knock Out rose. Figure 3 11. Weekly mean captures of Gynaikothrips uzeli on yellow sticky traps placed above buttonwood and Knock Out rose.

PAGE 88

88 Figure 3 12. Weekly mean ca ptures of genus Thrips on yellow sticky traps placed above buttonwood and Knock Out rose. Figure 3 13. Weekly mean captures of Merothrips spp. on yellow sticky traps placed above buttonwood and Knock Out rose.

PAGE 89

89 Figure 3 14. Weekly mean capture s of Echinothrips americanus on yellow sticky traps placed above buttonwood and Knock Out rose. Figure 3 15. Weekly mean captures of Chirothrips spp. on yellow sticky traps placed above buttonwood and Knock Out rose.

PAGE 90

90 Figure 3 16. Weekly mean captures of Heliothrips haemorrhoidalis on yellow sticky traps placed above buttonwood and Knock Out rose. Figure 3 17. Weekly mean captures of Selenothrips rubrocinctus on yellow sticky traps placed above buttonwood and Knock Out rose.

PAGE 91

91 Figur e 3 18. Weekly mean captures of Chaetanaphothrips orchidii on yellow sticky traps placed above buttonwood and Knock Out rose. Figure 3 19. Weekly mean captures of Holopothrips inquilinus on yellow sticky traps placed above buttonwood and Knock Out rose.

PAGE 92

92 Figure 3 20. Weekly mean captures of Scolothrips sexmaculatus on yellow sticky traps placed above buttonwood and Knock Out rose. Figure 3 21. Weekly mean captures of low abundance low -frequency unidentified thrips on yellow sticky traps placed above buttonwood and Knock Out rose.

PAGE 93

93 Figure 3 22. Weekly mean captures of S cirtothrips dorsalis on yellow sticky traps placed above buttonwood and Knock Out rose.

PAGE 94

94 CHAPTER 4 EFFECTS OF FERTILIZER RATE AND ROSE CULTIVAR ON POPULATIONS OF Sc irtothrips dorsalis H OOD (THYSANOPTERA: T HRIPIDAE) IN SOUTHERN FLORIDA Introduction Scirtothrips dorsalis Hood (1919) is a polyphagous insect pest whose global distribution has expanded rapidly over the last twenty years, becoming established in southern a nd central Florida (Silagyi and Dixon 2006). This insect pest shows considerable potential for expansion into the rest of North America (Venette and Davis 2004, Meissner et al. 2005, Nietschke et al. 2008), and impacts several fruit and vegetable hosts of economic significance, including F ragaria x annanasa Duchesne (strawberry) Mangifera spp. L. (mango), Capsicum spp. L. (pepper) Citrus spp L., Gossypium spp. L. (cotton), Glycine max (L.) Merr (soy bean) and Arachis hypogaea L. ( peanut ). This thrips is also a pest on ornamental hosts, and has proved a key pest of rose for the cut flower industry in India, reducing the quality, number, size, and appearance of the flowers produced (Onkarappa and Mallik 1998, Duraimurugan and Jagadish 2004a). Leaves, sepals and petals develop a characteristic bronze scarring (Dev 1964, Chandrasekaran 2005), and prolonged high -density infestations leave plants looking scorched, or result in the death of the plant (Mound and Palmer 1981). Damaged plants are considered unattra ctive, and do not sell. Several authors have explored the damage caused by and preferences of S. dorsalis on different cultivars of rose. Gahukar (2003) tested the cultivars Gladiator, Folklore, and Landora while Onkarappa and Mallik (1998) worked with Redgold, and Duraimurugan and Jagadish (2004a, b) worked with Local Red. Recentl y researchers have documented the impact and performance of this thrips on the Knock Out Rose in Florida and Texas (Palmer and Vea 2008). Only Gahukar (2003) compared the effect of cultivar on thrips populations. He showed that while there was a high degree of variability from year to year and among cultivars, plots of

PAGE 95

95 Gladiator had the highest infestation rate, the highest abundance of thrips in blooms, and greater numb er s of thrips for all sizes of flower. Gahukar also sho wed that the color of blooms had an effect on population distribution, with red flowers holding more thrips than orange and yellow flowers. The Knock Out Rose Rosa Radrazz (Radler 2001) has becom e a popular landscape choice in Florida due to its abundant flowers, low maintenance requirements, and strong resistance to black spot, Diplocarpon rosae in Floridas humid environment (Brown 2007). Unfortunately, the Knock Out Rose has also proven to be an important ornamental host to the invasive thrips, S. dorsalis (Silagyi and Dixon 2006). This has led to a change in the ornamental landscape as fewer nurseries in south Florida offer and fewer landscapers recommend the previously popular rose to homeo wners out of concern for the damage caused by the thrips (personal observation). The Conard -Pyle company recently developed a new cultivar of Knock Out Rose the Sunny Knock Out Rosa Radsunny (PP#18562). It is reputed to have similar foliage and di sease resistance to the original Knock Out Rose Rosa Radrazz (PP#11836), but produces bright yellow blooms that fade to white as they age, and the petioles release a strong sweetbriar fragrance. However, the susceptibility of this new cultivar to S. dorsalis has not yet been determined. There is concern that the common nursery practice of adding nutrients to promote or push the growth of ornamental plants may be aggravating populations of S. dorsalis Plants provided with higher nitrogen content a re frequently more attractive to pest insects (Slansky and Rodriquez 1987). Heavily fertilized plants have more free amino acids in their tissues, allowing for the rapid growth of the plant and the populations of pest insects feeding upon them (Mattson

PAGE 96

96 1980). Fertilized plants may provide additional attractive cues to pest populations by increasing plant height or the size and number of blooms produced. This trend holds true for the Thysanoptera, with elevated concentrations of fertilizer resulting in a hi gher growth rate of Frankliniella occidentalis (Chau et al. 2005). This process will be mediated by the specialized secondary compounds produced by resistant cultivars (Schuch et al. 1998) or by particular host species (Reitz 2002). With higher fertilizer rates, Duraimurugan and Jagadish (2004b) observed larger populations of S. dorsalis on rose. Varghese and Giraddi (2005) recommended halving the amount of fertilizer used on chili in conjunction with azadirachtin application when controlling for S. dorsali s As this thrips may thrive in both agricultural and ornamental landscapes, both grower and homeowner must simultaneously apply a comprehensive integrated pest management program that considers both of their unique environments to prevent the establishment of reservoir populations or the development of resistance to commercial pesticides. Additional knowledge and awareness of the specific impact of this thrips upon additional varieties of landscape roses popular in Florida would be valuable in constructing techniques for integrated pest management. The purpose of these experiments was to evaluate the effect of two recommended fertilizer rates and cultivar on the population density of thrips and host plant damage. Materials and Methods Fertilizer Rate wit h Radrazz T he first experiment (May to August), examined two recommended fertilizer rates on Radrazz using eighteen Radrazz rose s of approximately one year in age. These were grown in 11.36 liter containers in a medium consisting of 50% Canadian spha gnum peat moss, 25% processed pine bark, and a mixture of perlite and vermiculite (Fafard 3B Mix, Conrad Fafard Inc., Agawam, Massachusetts). They were placed in an outdoor area at the University of Florida

PAGE 97

97 Tropical Research and Education Center ( TREC ) cam pus in Homestead, Florida. These plants had established infestations derived from naturally occurring populations of S. dorsalis Six plants were then randomly assigned to one of three treatments. The high fertilizer treatment was recommended by local gr owers to accelerate plant growth, which were fertilized with a n application of 7.92 grams of liquid fertilizer (Peters Professional, Scotts Company, Marysville, Ohio) (N H3: NO3: (NH2)2CO: P2O5: K2O = 4:6:10:20:20) every other week, and 13.0 grams of pelle t fertilizer (NH3: NO3: P2O5: K2O = 15:9:12) (Osmocote Plus, Scotts Company, Marysville, Ohio) once every other month. The maintenance fertilizer treatment was recommended by local growers to maintain plants, which were fertilized with 3.96 grams of liquid fertilizer once per month, and 7.5 grams of pellet fertilizer once every other month. The control plants received only water when other plants received their application of liquid fertilizer, and no solid fertilizer amendments. All plants were then randomly placed in an uneven four by five grid that separated each plant from their neighbors by a third of a meter. Fertilizer Rate with Radsunny and Radrazz Cultivars The second experiment (August to October) compared the effect of fertilizer rate on Radrazz and Radsunny cultivars. T wenty -four Radrazz and Radsunny bare root rose plants provided by the ConardPyle Company were raised in a thrips -free greenhouse in 11.36 liter containers using identical media to the first experiment for three mont hs, and then transferred to an outdoor growing area at the TREC campus. These plants were exposed to Radrazz roses previously infested (three S. dorsalis /cm2 in flowers) with naturally established populations of S. dorsalis by placing them half a meter d ownwind of an infested plot. Four replicates containing one of each treatment and cultivar were randomly placed within a six by eight grid that separated each plant from their neighbors by a third of a meter. Eight plants of

PAGE 98

98 each cultivar were randomly ass igned to one of three fertilizer treatments identical to the rates used in the previous experiment. Susceptibility of Six Cultivars The third experiment (September to November) compared populations of S. dorsalis on the top five selling cultivars of rose in Florida, and an additional Knock Out cultivar. Five plants from each of six cultivars of rose in 11.36 liter containers were donated by Nelsons Florida Roses in Apopka, Florida. These cultivars includ ed Angel Face, Don Juan, Radcon (PP# 15070), Pink Summer Snow St Patrick (PP #9591), and Sun Flare. They were transported to the TREC on September 9th, 2008 and exposed to naturally occurring populations of thrips. F ive replicates with one plant from each of six cultivars were placed in a random locati on within a five by six grid that separated each plant from their neighbors by a third of a meter. All plants in all experiments remained exposed to ambient environmental conditions throughout the experiments, and were irrigated with 93 15 mL using an overhead system at six in the morning and two in the afternoon, according to local water restrictions and extension office recommendations. T he total number of buds and flowers was counted on each plant biweekly to develop a measure of the total floral outp ut of the plant. A damage estimate similar to that described by Krishna Kumar et al (1996) for pepper was developed during preliminary observations and used to rate the damage to plants each week. Plants were rated from zero, with no symptoms, to five, at which point the plant was defoliated and near death. For the first two experiments, mean thrips population estimates were assessed by randomly sampling a bud or flower (as available) from each plant every other week at three in the afternoon, as per Durai murugan and Jagadish (2004a). In the third experiment, one flower, one leaf, and one bud was sampled from each plant at approximately three in the afternoon once every other week in order to determine the distribution of the pest within the plant as per

PAGE 99

99 On karappa and Mallik (1998). For consistency, only mature leaves, buds, and flowers of a similar age were sampled. All samples were removed with pruning shears and sealed immediately in small plastic containers with two drops of 95% ethanol. Samples that cou ld not immediately be processed were stored at 6 C. T he total number and life stage s of thrips in a sample was recorded by washing the samples with 75% ethanol and pouring the runoff through a 25 m mesh as per Seal and Baranowski (1993), and then examining the collected debris under a dissection sco pe at a minimum of 12x magnification. The Funderburk et al. (2007) key to species of ornamental thrips was used as a primary aid in identification. Scirtothrips dorsalis could rapidly be differentiated from other thrips present in the sample by their relat ive size and other distinguishing characters, including antennal, pronotal, and abdominal characters. P lant samples were then dried overnight and the area of each dried sample was determined using a leaf area meter (LI 3000, LiCor, Lincoln, Nebraska) to de termine the density of thrips populations in each sample (thrips/cm2) Biweekly sampling for the first experiment was conducted from 29 May 2008 to 7 August 2008. For the second experiment, floral samples were collected bi weekly from 1 September 2008 to 22 October 2008. In the third experiment, s amples were collected on the day of arrival in Homestead, and then biweekly thereafter from 11 September 2008 to 11 November 2008. Voucher specimens were collected and sent to the Florida Department of Plant Industr y for preservation and to reconfirm species identification throughout the experiment. The total number, proportion at each life stage, density of thrips, the size and number of flowers, and the recorded damage by cultivar and fertilizer treatment were com pared among treatments. The total number of thrips and thrips densities were transformed by applying the square root to homogenize variance before analysis. Treatments were compared by ANOVA

PAGE 100

100 using JMP statistical software (SAS Institute 2007). Means from t he first two experiments were separated using pairwise t -test comparisons, and by Tukeys HSD in the third experiment. A two -way ANOVA was used to test for interactions between cultivar and fertilizer rate in the second experiment. Least square means were used to fit the general linear model and determine the strength of relationships between various factors in each experiment. Results Damage to plants during these experiments did not exceed a rating of two, where the new growth was wrinkled, and recent flush had scarring along the base of veins. Most sepals showed signs of scarring, and the outer petals of flowers and most leaves showed some stippling. Roses with a damage rating of one had damage on some sepals, but only a few petals and leaves showed any s tippling. Fertilizer Rate with Radrazz In the first experiment, there were no significant differences among treatments for the total number of thrips ( F = 0.48; df = 2; 108, p = 0.6124) (Figure 4 1 ), thrips density ( F = 0.50; df = 2, 108; p = 0.6062) (Figure 4 2 ), damage ratings ( F = 1.62; df = 2, 108; p = 0.2036), or the mean area of sampled flowers ( F = 0.52; df = 2, 108; p = 0.5989) for any and all evaluation dates, but mean values for all variables tested were greater for high and normal fertilizer rates than the controls, and total number of flowers and buds produced was significantly higher for the high and maintenance rates compared to the controls ( F = 4.28; df = 2, 108; p = 0.0163). Fertilizer Rate with Radsunny and Radrazz Cultivars In th e second experiment, no significant interaction was detected between cultivar and fertilizer rates for any of the tested variables, so treatment data could be compared independently by cultivar and fertilizer rate for analysis No significant difference wa s observed between Radrazz and Radsunny cultivar s for thrips abundance for any life stage or sex, thrips

PAGE 101

101 densities, damage rating, floral area, or the total number of flowers and buds produced ( p > 0.24). A signi ficant difference was observed among fe rtilizer treatments for the total number of thrips ( F = 3.46; df = 2, 232; p = 0.0330) ( Figure 4 3 ), for the total number of nymphs ( F = 3.99; df = 2, 232; p = 0.0197) and adults ( F = 3.24; df = 2, 232; p = 0.0411), for damage rating ( F = 5.07; df = 2, 232; p = 0.0070) ( Figure 4 4 ), but not for thrips densities (thrips/ cm2) (p = 0.1505). For all variables, high and normal rates of fertilizer had higher mean values than the unfertilized control. Further analysis by week showed that these significant and incr easing differences had occurred in the total number of thrips after five ( F = 3.32; df = 2, 44; p = 0.0455) and seven weeks ( F = 4.53; df = 2; 44 p = 0.0162) ( Figure 4 5 ) after treatments had begun. There were no significant weekly differences between trea tments in mean thrips density (thrips/ cm2) ( Figure 4 6 ). A significant difference was not observed among fertilizer treatments for the total number of flowers and buds ( F = 2.38; df = 2, 232; p = 0.0951) or for the mean area of flowers produced (F = 2.98; df = 2, 150; p = 0.0538). However, mean values for the high and normal fertilizer treatments did tend to be greater than the mean values for the controls. During the experiment, no plant received a damage rating greater than two (moderate damage) in any g iven week. Due to the limited range of damage expressed, the predictive relationship of damage ratings to thrips populations could not be described. However, damage appeared to be more predictive of the total number of thrips than thrips density, and the s cale was better at predicting the more abundant nymphs than adult populations. Susceptibility of Six Cultivars In the third experiment, preliminary sampling of the roses revealed a large number of Frankliniella bispinosa (Morgan) (Thysanoptera: Thripidae ). Throughout the entire period of the

PAGE 102

102 experiment, F. bispinosa accounted for a quarter of all thrips sampled, but the mean number and percentage of F. bispinosa nymphs and adults collected on plants steadily fell while the number and percentage of S. dors alis captures increased ( Figure 4 7 ). No significant interaction was observed between cultivar and time for thrips abundance, density, or damage. Thrips density was observed to increase on all hosts for the first five weeks, and then stayed statistically constant ( Figure 4 8 ). After five weeks, S. dorsalis were detected on all plants for all cultivars of rose tested. Cultivar ( F = 3.55; df = 5; p = 0.0037) and plant organ ( F = 96.09; df = 2; p < 0.0001) had a significant effect upon thrips densities, but no interaction was detected between cultivar and plant organ ( F = 1.81; df = 10; p = 0.0562). The significant difference in thrips density between cultivars ( F = 2.23; df = 5, 519; p = 0.0496) ( Figure 4 9 ) suggested three tiers of hosts. Radcon had the highest thrips density at 1.64 0.31 thrips/ cm2 (MSE); followed by Don Juan, Pink Summer Snow, and St. Patrick with approximately 1.10 0.27 thrips/ cm2; and Angel Face and Sun Flare with approximately 0.59 0.16 thrips/ cm2 per weekly sample. P lant organ s sampled also showed significant differences ( F = 89.14; df = 5, 519; p < 0.0001) ( Figure 4 10), with buds showing 0.56 more thrips/ cm2 than petals or leaves. While no plant was observed with a damage rating greater than one during the experime ntal period, a significant difference was detected in mean damage ratings for each cultivar ( F = 5.42; df = 5, 519; p < 0.0001) ( Figure 4 11). Discussion The results of the first and second experiment comparing different rates of fertilizer appear to conf lict. In the first experiment, fertilizer rate had no significant effect upon either thrips populations or floral traits, while in the second experiment fertilizer rate had a significant effect

PAGE 103

103 upon thrips populations. The low er number of samples per treat ment in the first experiment certainly reduced the power of the test, but the data was suggestive of trends observed in the second experiment. T here are other possible explanations: in t he first experiment fertilizer treatments were applied to plants with an established infestation of S. dorsalis and in the second experiment, fertilizer treatments were applied to thrips -free plants without an established infestation. Established populations that have not yet reached capacity may be less susceptible to ass ortment and dispersal pressures, and less indicative of thrips host choices. While the Radsunny and Radrazz cultivars of Knock Out Rose can easily be distinguished by their different foliar appearance, the obvious difference in petal color, and by a very distinct floral scent, populations of S. dorsalis seem to be evenly distributed throughout both hosts without regard to cultivar. This might appear surprising, as this thrips has been captured in greater frequencies on yellow (Tsuchiya et al. 1995, Chu et al. 2006) or yellow and green (Tsuchiya et al. 1995) sticky traps than those of other colors. However, Rani and Sridhar (2003) and Gahukar (2003), recovered more adults of S. dorsalis on red or orange rose petals than yellow ones in a small arena choice test. While these results may appear to conflict, it is possible that S. dorsalis uses different cues and selection criteria to choose a host at different spatial scales. Rani and Sridhar, and Gahukars choicetests occurred in small arenas within the lab, while Tsuchiya et al. and Chu et al. performed their choice tests in the field. Comparing those observations to the results of the second experiment suggests that at a range of less than one meter, S. dorsalis does not appear to choose between the tw o Knock Out roses based on criteria that are obvious to human observers, such as color or scent. Other factors may be more important to S. dorsalis local dispersal and distribution.

PAGE 104

104 This experiment does reveal that some choice of host is occurring, and hints at some of those hidden criteria, as there is an uneven distribution of S. dorsalis among fertilizer treatments. Populations of S. dorsalis were significantly more abundant on hosts that have been treated with fertilizer than those that were untreated. Interestingly, thrips density between the three fertilizer rates remained relatively constant, suggesting that populations of this thrips have distributed themselves evenly across the plants in order to exploit resources with minimum competition. It is possible that with higher pest pressure and a higher total population of thrips, crowding would begin to occur. High nutrient plants would show significantly greater densities as a larger proportion of free nutrients should still be available when compare d to unfertilized hosts. However, it is also possible that there are physical upper limits on the number of thrips that might exploit a host simultaneously. Each thrips destroys mesophyll cells when it feeds, and leaves a patch of scarred and cellulosic t issue behind. Therefore higher densities may well exceed the ability of the afflicted plant to replace damaged tissue. The relationship between thrips density and observed feeding damage might have been better described by a different damage rating scale t han the one used. The limited range of damage observed made it difficult to draw strong conclusions about the relationship between thrips populations and damage. With this in mind, the scale was predictive of the number of nymphs and adults detected in flowers, but could not account for a large proportion of the variation present in populations. Perhaps most surprising was that while the highest fertilizer rate tested did produce significantly more and larger blooms than the controls, the higher rate did n ot produce significantly more or larger blooms than the maintenance treatment of fertilizer. This suggests that the maintenance level provides sufficient nutrients for near -maximum floral growth, with

PAGE 105

105 diminishing returns in additional flowers grown with ne arly twice the investment of fertilizer. It is possible that the higher rate of fertilizer allowed for a greater proportion of vegetative growth, a variable not tested in this experiment. However, flowers produced by the high nutrient plants also showed si gnificantly higher damage, and analysis suggests that this is due in part to the larger populations of thrips identified on the plant. Varghese and Giraddis (2005) recommendations to halve the fertilizer used on chili can probably be used as a control mec hanism to limit populations and damage caused by S. dorsalis in the Knock Out and other cultivars of rose. When comparing absolute counts of thrips on the other cultivars tested, St. Patrick, Don Juan, and Radcon had the largest mean populations of S. dorsalis but when looking at thrips density per square centimeter, the smaller flowers of Radcon and Pink Summer Snow were not statistically different than those of the larger St. Patrick and Don Juan roses This is similar to the findings of Gahukar (2003), in that large open flowers can play host to larger absolute numbers of thrips, but smaller and more compact flowers may have a relatively greater number of thrips per unit area. This suggests that these hosts provide similar environments f or S. dorsalis and that vegetative size should determine total thrips population. Factors endogenous to the cultivar of rose still appear to play a role, as the Angel Face cultivar had a lower mean density of thrips and experienced less damage relative to thrips density than the other cultivars tested. This may represent host resistance preventing growth, or a thrips preference base d on the structural, chemical composition of a particular cultivar. It should be noted that while thrips densities are valuable for comparing different cultivars of rose or different species of plant the total number of thrips may be important to questions of population expansion and control. Larger populations are better able to persist through

PAGE 106

106 catastrophic ecological events, and increase pressure on surrounding environments (Memmott et al. 1997, Fagan et al. 2002), especially if the larger roses have a greater overall floral area, or produce a larger number of buds. S. dorsalis was found in significantly higher densities on t he rose buds than on rose flowers or on leaves. Many Thripidae appear to prefer the upper part of the plant canopy and the outer extremities of their hosts (Reitz 2002, Hansen et al. 2003), and removing growth terminals in citrus resulted in increased migr ation of S. dorsalis populations away from treated plants (Shibao et al. 1993, Shibao 1997). It has been suggested that thrips prefer these areas because nutrient flow in plants is towards these areas of new growth (Lewis 1997). This is not inconsistent wi th prior research regarding this thrips on Capsicum annuum L. ( chili pepper ) (Seal et al. 2006), Ricinus communus L. ( castor ) (Raizada 1965), or Camellia spp. L. ( tea ) (Dev 1964). Females of S. dorsalis prefer to lay their eggs inside of young leaves and b uds at the apical meristem of plants, but as populations increase, will lay their eggs within the surface of mature leaves (Dev 1964, Raizada 1965). After hatching, nymphs will migrate from older leaves to the newer growth at terminals (Onkarappa and Malli k 1998). While thrips were not present in high densities or abundances on the leaves, the nymphs left evidence of their presence through their distinctive feeding damage. More importantly, S. dorsalis was present on leaves in densities that were not signif icantly different than those within the petals of flowers. This should not be surprising, as the petals of flowers are merely developmentally modified leaves distributed within a much smaller volume. This may help to explain a discrepancy in the literature where this thrips is sometimes referred to as a flower thrips, and in other references is referred to as an exclusively foliar thrips. S. dorsalis may thrive equally well on both substrates, and exogenous environmental factors such as heavy

PAGE 107

107 precipitation, predation, or competition may determine the distribution of populations within the plant. T here was no interaction between plant organ and cultivar, further evidence suggesting that this thrips will exploit all roses in a similar fashion, even as they af flict some varieties in greater densities than others. This was not necessarily the expected result, as Shibao et al. (1990) observed that while overall populations of S. dorsalis were similar between two varieties of grape, there were still internal distr ibutional differences. This implies that the method by which this thrips exploits these plant resources is fairly similar, in spite of the roses morphological (and presumably chemical and nutrient) diversity. Of the cultivars tested, Radcon, Don Juan and the Sun Flare experienced the worst damage relative to their thrips abundance and density. Of the three, the Sun Flare had the highest susceptibility to damage by this thrips, with the highest damage relative to the number and density of thrips present. However, of the cultivars tested, the Radcon consistently scored higher damage ratings, abundances, and densities than most of the other five cultivars tested. This continues to lend credence to the initial speculation by extension agents that th e Knock Out variety of rose may be an important host to consider when surveying for this thrips. While damage ratings never exceeded a value of two prior research ( personal observations ) suggests that this may not be unusual for this thrips behavior d uring the fall season that this experiment was conducted in south Florida. It is possible that damage and thrips density would increase on all cultivars during seasons that are more inclement to population growth of S. dorsalis thereby aggravating the pla nts with more pest pressure and increasing the observable differences between cultivar responses to the pest. Furthermore, it is possible that the presence of multiple cultivars may not describe the performance of this thrips when only one of

PAGE 108

108 these hosts i s present. It is possible that in a no -choice experiment, some cultivars would actually show a higher degree of damage or thrips abundance and density. Acknowledgements The researchers would like to thank the Conard-Pyle company for providing the Knock Ou t roses used in this study, and to thank Nelsons Roses of Florida for donating the other cultivars of rose tested.

PAGE 109

109 Figure 4 1 Weekly m ean total number of thrip s plus mean standard error for three fertilizer rates (no significant differences).

PAGE 110

110 Figure 4 2 Weekly m ean thrips density ( S. dorsalis /cm2) plus mean standard error for three fertilizer rates (no significant difference)

PAGE 111

111 Figure 4 3 Mean total number of Scirtothrips dorsalis per floral sample for each fertilizer rate plus mean standard error (different letters among fertilizer treatments are significantly different by muliple pairwise t -test comparisons at a = 0.05).

PAGE 112

112 Figure 4 4 Mean damage rating recorded for each fertilizer treatment plus mean standard error (different letters among fertilizer treatments are significantly different by multiple pairwise t -test comparisons at a = 0.05).

PAGE 113

113 Figure 4 5 Weekly m ean total number of thrip s plus mean standard error for three fertilizer rates (* = high and maintenance rates are signi ficantly different from control by multiple pairwise t -test comparisons at a = 0.05).

PAGE 114

114 Figure 4 6 Weekly m ean thrips density ( S. dorsalis /cm2) plus mean standard error for three fertilizer rates (no significant differences detected using multiple pairw ise t -t ests, with p > 0.25).

PAGE 115

115 Figure 4 7 Mean weekly t otal number of thrips for F. bispinosa and S. dorsalis plus mean standard error over time.

PAGE 116

116 Figure 4 8 Weekly m ean thrips density ( S. dorsalis /cm2) plus mean standard error for each cultivar f rom mid -September to mid November.

PAGE 117

117 Figure 4 9 Mean thrips density ( S. dorsalis /cm2) plus mean standard error on each cultivar (cultivars connected by the same letter are not significantly different by TukeyKramer HSD at a = 0.05).

PAGE 118

118 Figure 4 10. Me an thrips density ( S. dorsalis /cm2) plus mean standard error on different plant organs sampled ( organs connected by the same letter are not significantly different by Tukey -Kramer HSD at a = 0.05).

PAGE 119

119 Figure 4 11. Mean damage rating plus mean standard err or for each cultivar over nine weeks of exposure (cultivars connected by the same letter are not significantly different by Tukey -Kramer HSD at a = 0.05).

PAGE 120

120 CHAPTER 5 CONCLUSIONS Scirtothrips dorsalis is an opportunistic generalist herbivore with parthenogenetic reproductive ability. These adaptations have allowed the insect to expand its range globally, and are important considerations when exploring the insects potential impact upon the Florida landscape. The broad polyphagy of the organism permits its e xploitation of a diverse group of potential hosts within its operational environment, but while this thrips is capable of subsisting upon many hosts, it is better able to complete development on some hosts than others under particular environmental conditi ons. This is most obvious when comparing the annual performance of this thrips upon rose to its performance upon buttonwood or schefflera. While all life stages were identified on the two alternative hosts, suggesting that reproduction was occurring, popul ations were significantly higher on rose. More importantly, the population demographics of S. dorsalis on rose showed a much smaller proportion of adults, suggesting a higher rate of growth. These factors may be intrinsic to the plant, and I had hoped tha t by comparing different cultivars of rose, I might be able to tease out some of these factors. While there were some standouts in their responses to the thrips, most cultivars were infested in a similar fashion and suffered similar damage from the thrips. More importantly, the thrips appeared to infest all roses in a similar fashion. Total thrips numbers were much higher in flowers than on leaves, but the density of thrips exploiting these plant resources was very similar within the plant as well as betwee n plants, with an overall trend moving the thrips towards areas of peak growth on the plant. As an opportunistic generalist, it will feed wherever it is most advantageous, or wherever it is not specifically excluded. The tightly packed foliage of flowers m ay serve as a refuge, protecting the thrips from exposure to the elements and predators.

PAGE 121

121 It is also important not to forget that while the alternate hosts did not develop large populations, S. dorsalis were still a large proportion of the thrips fauna obs erved throughout this project. T his implies that S. dorsalis has probably become established in the south Florida landscape, and with a large number of available ornamental, weed, and agricultural hosts to subsist upon, this pest will probably continue to persist in small population reservoirs somewhere in the environment no matter what control mechanisms are applied. These reservoir populations will gradually recolonize lost territory from unmanaged areas, and a realistic management program should emphasi ze proactive measures that prevent or limit populations during crucial periods of peak growth. These measures may include factors not tested in these experiments, such as the efficacy of various insecticides as well as the development of longterm biologic al control, and should also consider the appropriate environmental conditions for the expansion of this thrips. Modifying the nutritional content of these hosts will also play a significant role in the performance of this thrips. Fertilizing plant hosts a ppears to either make them more attractive to S. dorsalis or provides more nutrients permitting a greater rate of population increase on a host. While higher rates of fertilization did not result in significantly larger numbers or sizes of flowers, there were trends suggesting that over time it would result in significantly larger populations of thrips. This factor and the practice of nutrient pushing may explain some of sporadic explosions of previously unidentified infestations of S. dorsalis at nurse ries. Small populations appear to cause relatively little damage, and may go unobserved but when nursery growers fertilize, these populations can increase at a higher rate, thereby creating more observable damage.

PAGE 122

122 Environmental factors extrinsic to the plant also play an important role in defining populations of S. dorsalis As observed by other researchers, populations show a positive relationship with increasing temperature, and months with greater accumulated degree days show larger thrips populations on all hosts tested. Increasing humidity and rainfall appear to be negatively correlated with this thrips population growth a trend common to the Thysanoptera as a whole. Periods of peak population growth also appear to be periods of peak population d ispersal. Thrips population density remained similar within a plot, suggesting that the thrips is capable of detecting population size through unknown means, possibly through volatile chemical communication or the number of interactions on a plant. When th ese cues increase in frequency, the thrips may disperse to reduce competition for resources. Thrips dispersal flight appears to be defined by environmental factors similar to those that regulate population growth. As the temperature rises and thrips accumu late degree hours, the number captured while dispersing increases until the majority of thrips who disperse will have flown. Humid conditions reduce these flights. Based on these results, we should anticipate larger populations of this thrips towards late spring and mid fall in Florida, when our climate is at its warmest and driest. A period of growth and dispersal should also follow any unusually long droughts or months with unusually high minimum temperatures. Growers should prepare themselves accordingly, with regular trapping to observe population levels in the field, and then defend their crops with the appropriate control measures.

PAGE 123

123 LIST OF REFERENCES Abro, G. H., T. S. Syed, U. M. A, and M. S. Zhang. 2004. Effect of application of a plant growth regu lator and micronutrients on insect pest infestation and yield co mponents of cotton. Journal of Entomology 1: 12 16. Arthurs, S., C. L. McKenzie, J. Chen, M. Dogramaci, M. Brennan, K. Houben, and L. Osborne. 2009. Evaluation of Neoseiulus cucumeris and Am blyseius swirskii (Acari: Phytoseiidae) as biological control agents of chilli thrips, Scirtothrips dorsalis (Thysanoptera: Thripidae) on pepper. Biological Control 49: 91 96. Ananthakrishnan, T. 1993. Bionomics of Thrips. Annual Reviews of Entomology 38: 71 92. Asaf Ali, K., E. Abraham, S. Thirumurthi, and T. Subramaniam. 1973. Con trol of the scab thrips ( Scirtothrips dorsalis ) infesting grapevine ( Vitis vinifera ). South Indian Horticulture 21: 113 114. Bailey, S. F. 1940. The distribution of injurio us thrips in the United States. Journal of Economic Entomology 33: 133 136. Bansiddhi, K., and S. Poonchaisri. 1995. Thrips of vegetables in Thailand, pp. 109 110. In B. L. Parker, M. Skinner and T. Lewis [eds.], Thrips Biology and Management. Plenum P ress, New York, NY. Bisevac, L. 1997. A New Method for Mounting Thrips (Thysanoptera) on Slides Australian Journal of Entomology 36: 220. Blum, M. S. 1991. Chemical ecology of the Thysanoptera, pp. 95 108. In B. L. Parker, M. Skinner and T. Lewis [eds.], Towards Understanding Thysanoptera. Proceedings of the International Conference on Thrips. USDA Technical Report NE 147, Radnor, PA. Bournier, J. P. 1999. Deux Thysanoptres, nouveaux dprdateurs du cotonnier en Cte d'Ivoire. Annales de la Socit Entom ologique de France 35: 275 281. Brown, S. P. 2007. Growing roses in Florida, pp. 1 4, EDIS. IFAS, Gainesville, Florida. Casey, C., J. Newman, K. Robb, S. Tjosvold, H. Peterson, and M. Parrella. 2007. IPM Program successful in California greenhouse cut ro ses. California Agriculture 61: 71 78. Chandrasekaran, M. 2005. Symptomatology of chili leaf curl. Journal of ecotoxicology and environmental monitoring 15: 377 380. Chang, N. T. 1995. Major pest thrips in Taiwan, pp. 105 108. In B. L. Parker, M. Ski nner and T. Lewis [eds.], Thrips Biology and Management. Plenum Press, New York, NY. Chatzivassiliou, E. K., I. Livieratos, G. Jenser, and N. I. Katis. 2000. Ornamental plants and thrips populations associated with Tomato Spotted Wilt Virus in Greece. Phytoparasitica 28: 1 8.

PAGE 124

124 Chau, A., K. M. Heinz, and F. T. D. Jr. 2005. Influences of fertilization on population abundance, distribution, and control of Frankliniella occidentalis on chrysanthemum. Entomologia Experimentalis et Applicata 117: 27 39. Chen, C. C., and R. J. Chiu. 1996. A Tospovirus infecting peanut in Taiwan. Acta Horticulturae. 431: 57 67. Chiemsombat, P., O. Gajanandana, N. Warin, R. Hongprayoon, A. Bhunchoth, and P. Pongsapich. 2008. Biological and molecular characterization of tospovi ruses in Thailand. Archives of Virology 153: 571 577. Chin, D., H. Brown, M. Neal, and B. M. Thistleton. 2007. Mango flower insects and their effect on fruit quality in the Northern Territory, pp. 25 33. In P. Industries [ed.]. Chu, C. -C., M. A. Ciompe rlik, N. -t. Chang, M. L. Richards, and T. J. Henneberry. 2006. Developing and evaluating traps for monitoring Scirtothrips dorsalis (Thysanoptera: Thripidae). Florida Entomologist 89: 47 55. Collingham, E. 2006. Curry: A Tale of Cooks and Conquerors. Oxf ord University Press. Cornell, H. V., and B. A. Hawkins. 1994. Patterns of parasitoid accumulation, pp. 77 89. In Hawkins and Sheehan [eds.], Parasitoid Community Ecology. Oxford University Press, New York, NY. Crawley, M. J. 1986. The population biology of invaders. Philosophical transactions of the Royal Society of London 314: 711 731. Delphia, C. M., M. C. Mescher, and C. M. D. Moraes. 2007. Induction of plant volatiles by herbivores with different feeding habits and the effects of induced defenses on host -plant selection by thrips. Journal of Chemical Ecology 33: 997 1012. Dev, H. N. 1964. Preliminary studies on the biology of the Assam thrips, Scirtothrips dorsalis Hood, on tea. Indian Journal of Entomology 26: 184 194. Diffie, S., G. B. Edwards and L. A. Mound. 2008. Thysanoptera of Southeastern U.S.A.: A checklist for Florida and Georgia. Zootaxa 1787: 45 62. D Primary results of study on morphological, ecological, biological characteristics of Scirtothrips dorsalis Hood on orange and mandarin trees in Cao Phong state farm, Hoa Binh province. TC Nng nghi 33. D uraimurugan, P., and A. Jagadish. 2002. Seasonal incidence and effect of weather parameters on the population dynamics of chili thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) on Rose., pp. 180 183. In B. S. Babu, K. S. Varaprasad, K. Anitha R. D. V. J. Prasada Rao, S. K. Chakrabarty and P. S. Chandurkar [eds.], Resources management in plant protection during twenty first century. Plant Protection Association of India, Hyderabad, India.

PAGE 125

125 Duraimurugan, P., and A. Jagadish. 2004a. Evaluation of suitable technique and determination of appropriate stage for sampling thrips in rose flowers. Food, Agriculture & Environment 2: 187 189. Duraimurugan, P., and A. Jagadish. 2004b. Nonchemical management of Scirtothrips dorsalis Hood damaging rose flow ers. Journal of Applied Zoological Researches 15: 163 165. Ekvised, S., S. Jogloy, C. Akkasaeng, M. Keerati kasikorn, T. Kesmala, I. Buddhasimma, and A. Patanothai. 2006. Field evaluation of screening procedures for thrips resistance in peanut. Asian Jou rnal of Plant Science 5: 838 846. Ellington, C. P. 1980. Wing mechanics and take off preparation of thrips (Thysanoptera). Journal of Experimental Biology 85: 129 136. EPPO. 1997. Scirtothrips dorsalis pp. 1425. In I. M. Smith, D. G. McNamara, P. R. S cott and M. Holderness [eds.], Quarantine Pests for Europe, 2nd Edition. CABI/EPPO, Wallingford. EPPO. 2003. First report of Scirtothrips dorsalis in Israel Bulletin OEPP/EPPO Bulletin 32: 1 15. EPPO. 2005. EPPO Standards Diagnostic protocols for regul ated pests Scirtothrips aurantii, Scirtothrips citri, Scirtothrips dorsalis Bulletin OEPP/EPPO Bulletin 35: 353 356. Fagan, W. F., M. A. Lewis, and P. van den Driessche. 2002. Invasion theory and biological control. Ecology Letters 5: 148 157. Fery, R L., and J. M. Schalk. 1991. Resistance in pepper ( Capsicum annuum L.) to western flower thrips [ Frankliniella occidentalis (Pergande)]. HortScience 26: 1073 1074. Fournier, F., G. Bovin, and R. K. Stewart. 1995. Effect of Thrips tabaci (Thysanoptera: T hripidae) on yellow onion yields and economic thresholds for its management. Journal of Economic Entomology 88: 1401 1407. Funderburk, J. E., J. Stavisky, and S. Olson. 2000. Predation of Frankliniella occidentalis (Pergande) in field peppers by Orius in sidiosus (Say). Environmental Entomology 29: 376 382. Funderburk, J. E., S. Diffie, J. Sharma, A. Hodges, and L. Osborne. 2007. Thrips of ornamentals in the southeastern US, pp. 1 10, EDIS. IFAS, Gainesville, Florida. Gahukar, R. 2003. Factors influenc ing thrips abundance and distribution on rose flowers in central India. Journal of Entomology Research 27: 271 279. Gentile, A. G., and S. F. Bailey. 1968. A revision of the genus Thrips Linnaeus in the New World, with a catalogue of world species (Thysa noptera: Thripidae). University of California Press, Berkely, CA.

PAGE 126

126 George, S., and R. S. Giraddi. 2007. Management of chili ( Capsicum annuum L.) thrips and mites using organics. Karnataka Journal of Agricultural Science 20: 537 540. Gilbert, M. J. 1986. F irst African record of Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) a potential pest of citrus and other crops in southern Africa. Journal of the Entomological Society of Southern Africa 49: 159 161. Glick, P. A. 1939. The distribution of insects spiders, and mites in the air, pp. 1 150. In U. S. D. o. Agriculture [ed.]. USDA Technical Bulletin. Grimaldi, D., A. Shmakov, and N. Fraser. 2004. Mesozoic thrips, and early evolution of the Order Thysanoptera (Insecta). Journal of Paleontology 78: 941 952. Groves, R. L., J. F. Walgenbach, J. W. Moyer, and G. G. Kennedy. 2003. Seasonal dispersal patterns of Frankliniella fusca (Thysanoptera: Thripidae) and Tomato Spotted Wilt Virus occurrence in central and eastern North Carolina. Journal of Economic Entomology 96: 1 11. Hansen, E. A., J. E. Funderburk, S. R. Reitz, S. Ramachandran, J. E. Eger, and H. J. McAuslane. 2003. Within plant distribution of Frankliniella species (Thysanoptera: Thripidae) and Orius insidiosus (Heteroptera: Anthocoridae) in fi eld pepper. Environmental Entomology 32: 1035 1044. Heming, B. S. 1993. Structure, function, ontogeny, and evolution of feedng in thrips (Thysanoptera). In C. W. Shaefer and R. A. B. Leschen [eds.], Functional Morphology of Insect Feeding. Entomological Society of America, Lanham, Maryland. Hoddle, M. S., and L. A. Mound. 2003. The genus Scirtothips in Australia (Insecta, Thysanoptera, Thripidae). Zootaxa 268: 1 40. Hoddle, M. S., S. Nakahara, and P. P. A. 2002. Foreign exploration for Scirtothrips pers eae Nakahara (Thysanoptera: Thripidae) and associated natural enemies on avocado ( Persea americana Miller.) Biological Control 24: 251 265. Hoddle, M. S., J. M. Heraty, P. F. Rugman -Jones, L. A. Mound, and R. Stouthamer. 2008. Relationships among species of Scirtothrips (Thysanoptera: Thripidae, Thripinae) using molecular and morphological data. Annals of the Entomological Society of America 101: 491 500. Hood, J. D. 1919a. On some new Thysanoptera from southern India. Insecutor Inscitiae Menstruus 7: 90 103. Hood, J. D. 1919b. Two new genera and thirteen new species of Australian Thysanoptera. Proceedings of the Biological Society of Washington 32: 75 92. Horber, E. 1980. Type and classification of resistance, pp. 15 21. In F. G. Maxwell and J. P. R [eds.], Breeding plants resistant to insects. Wiley, New York, New York.

PAGE 127

127 Howard, D. F., M. S. Blum, and H. M. Fales. 1983. Defense in thrips: forbidding fruitiness of a lactone. Science 220: 335 336. Inc., S. I. 2007. JMP Statistics and Graphics Guide computer program, version 7. By Inc., S. I., Cary, NC. Irwin, M. E. 1991. Agroecological niches and thrips (Thysanoptera: Thripidae) dynamics, pp. 133 147. In B. L. Parker, M. Skinner and T. Lewis [eds.], Towards Understanding Thysanoptera. Proceedings of the International Conference on Thrips. USDA Technical Report NE 147, Radnor, PA. Jan de Kogel, W., and E. H. Koschier. 2002. Thrips responses to plant odors, pp. 189 190. In L. A. Mound and R. Marullo [eds.], Thrips and Tospoviruses: Proceedings of t he 7th International Symposium on Thysanoptera. CSIRO Entomology, Reggio Calabria, Italy. Jerado, A. 2006. Floriculture and Nursery Crops Outlook, pp. 25. In E. R. Service [ed.]. Jones, D. R. 2005. Plant viruses transmitted by thrips. European Journal of P lant Pathology 113: 119 157. Kajita, Y. 1986. Predation by Amblyseius spp. (Acarina: Phytoseiidae) and Orius sp. (Hemiptera: Anthocoridae) on Thrips palmi Karny (Thysanoptera: Thripidae). Applied Entomology and Zoology 21: 482 484. Kannan, M., S. Utham asamy, and S. Mohan. 2004. Impact of insecticides on sucking pests and natural enemy complex of transgenic cotton Research Communications 86: 726 729. Kirk, W. D. J. 1995. Feeding behavior and nutritional requirements, pp. 21 29. In B. L. Parker, M. Sk inner and T. Lewis [eds.], Thrips Biology and Management. Plenum Press, New York, NY. Kirk, W. D. J., and I. L. Terry. 2003. The spread of the western flower thrips Frankliniella occidentalis (Pergande). Agricultural and Forest Entomology 5: 301 310. Kla ssen, W., D. R. Seal, M. A. Ciomperlik, and D. A. Fieslemnann. 2008. The chili thrips, Scirtothrips dorsalis : current status in the greater Caribbean region. Proceedings of the Caribbean Food Crops Society 44. Kolb, T. E., L. H. McCormick, and D. L. Shumwa y. 1991. Physiological responses of pear thrips -damaged sugar maples to light and water stress. Tree Physiology 9: 401 413. Krishna Kumar, N. K., M. Aradya, A. A. Deshpande, N. Anand, and P. R. Ramachandar. 1996. Initial screening of chili and sweet pepp er germplasm for resistance to chili thrips, Scirtothrips dorsalis Hood. Euphytica 89: 319 324. Kulkarni, G. S. 1922. The Murda disease of chili ( Capsicum ). Agricultural Journal of India 22: 51 54.

PAGE 128

128 Lee, G. -S., J.-H. Lee, S. -H. Kang, and K. -S. Woo. 2001. Thrips species (Thysanoptera: Thripidae) in winter season and their vernal activities on Jeju Island, Korea. Journal of Asia -Pacific Entomology 4: 115 122. Lewis, T. 1973. Thrips. Their biology, ecology and economic importance. Academic Press, Lond on, GB. Lewis, T. 1981. Pest monitoring to aid insecticide use. Philosophical Transactions of the Royal Society: Biological Sciences 295: 153 162. Lewis, T. 1997. Thrips as crop pests. CAB International, Oxon, GB. Lewis, T., and L. R. Taylor. 1964. Diurn al periodicity of flight by insects. Transactions of the Royal Entomological Society of London 116: 393 469. Li, J.-M., G. W. Liu, and R. B. Peng. 2004. Bionomics and control of the yellow tea thrips, Scirothrips dorsalis infesting litchi. Entomological Knowledge 41: 172 173. Lingeri, M. S., J. S. Awaknavar, S. Lingappa, and K. A. Kulkarni. 1998. Seasonal occurrence of chili mites ( Polyphagotarsonemus latus [Banks]) and Thrips ( Scirtothrips d orsalis Hood). Karnataka Journal of Agricultural Science 11: 3 80 385. Loomans, A. J. 2003. Parasitoids as biological control agents of thrips pests, pp. 208, Entomologie. Wageningen Universiteit, Wageningen. Loomans, A. J., M. T. Murai, and I. D. Greene. 1997. Interactions with hymenopterous parasitoids and parasit ic nematodes. In T. Lewis [ed.], Thrips as crop pests. CAB International, Wallingford. Ludwig, S. 2008. Chili thrips: A new pest in Texas. In http://chillithrips.tamu.edu/ [ed.]. Texas AgriLife Extension Service, College Station. Masui, S. 2007a. Timing an d distance of dispersal by flight of adult yellow tea thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae). Japanese Journal of Applied Entomology and Zoology 51: 137 140. Masui, S. 2007b. Synchronism of immigration of adult yellow tea thrips, Sc irtothrips dorsalis Hood (Thysanoptera: Thripidae) to citrus orchards with reference to their occurrence on surrounding host plants. Applied Entomology and Zoology 42: 517 523. Mattson, W. J. 1980. Herbivory in relation to plant nitrogen content. Annual Ecology and Systematics 11: 119 161. Mattson, W. J., and R. A. Haack. 1987a. The role of drought stress in provoking outbreaks of phytophagous insects, pp. 365 394. In P. Barbosa and J. C. Schultz [eds.], Insect Outbreaks. Academic Press Inc, San Diego CA.

PAGE 129

129 Mattson, W. J., and R. A. Haack. 1987b. The Role of drought in outbreaks of plant -eating insects. BioScience 37: 110 118. McMurtry, J. A., and B. A. Croft. 1997. Life -styles of phytoseiid mites and their roles in biological control. Annual Review o f Entomology 42: 291 321. Meissner, H., A. Lemay, D. Borchert, B. Nietschke, A. Neeley, R. Magarey, M. Ciomperlik, C. Brodel, and T. Dobbs. 2005. Evaluation of possible pathways of introduction for Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) fro m the Caribbean into the continental United States, pp. 125. In USDA [ed.]. Plant Epidemiology and Risk Assessment Laboratory, Raleigh, North Carolina. Memmott, J., S. V. Fowler, and R. L. Hill. 1997. The effect of release size on the probability of establ ishment of biological control agents: gorse thrips ( Sericothrips staphylinus ) released against gorse ( Ulex europaeus ) in New Zealand. Biocontrol Science and Technology 8: 103 115. Mikunthan, G., and M. Manjunatha. 2008. Impact of habitat manipulation on mycopathogen, Fusarium semitectum to control Scirtothrips dorsalis and Polyphagotarsonemus latus of chili. BioControl 53: 403412. Mochizuki, M., and K. Honma. 2001. Damage control of yellow tea thrips, Scirtothrips dorsalis Hood by mulching near ultra -vio let radiation reflective film on young tea fields. Chagyo kenkyu hokoku (Tea research) 91: 13 19. Moritz, G., D. C. Morris, and L. A. Mound. 2001. Thrips ID: an interactive identification and information system (CD), Pest thrips of the world. CSIRO Publis hing, Collingwood, Australia. Moritz, G., L. A. Mound, D. C. Morris, and A. Goldarazena. 2004. Visual and molecular identification of pest thrips: an identification and information system using molecular and microscopical methods (CD), Pest thrips of the w orld. Lucid CSIRO publishing, Collingwood, Australia. Morse, J. G., and M. S. Hoddle. 2006. Invasion biology of thrips. Annual Review of Entomology 51: 67 89. Mound, L. A. 1998. Thysanoptera: an identification guide. CAB International, Oxon, New York. Mo und, L. A. 2002. So many thrips so few tospoviruses?, pp. 15 18. In L. A. Mound and R. Marullo [eds.], Thrips and Tospoviruses: Proceedings of the 7th International Symposium on Thysanoptera. CSIRO Entomology, Reggio Calabria, Italy. Mound, L. A. 2005. Thysanoptera: diversity and interactions. Annual Review of Entomology 50: 247 269.

PAGE 130

130 Mound, L. A., and J. M. Palmer. 1981. Identification, distribution and host -plants of the pest species of Scirtothrips (Thysanoptera: Thripidae). Bulletin of Entomologica l Research 71: 467 479. Mound, L. A., and R. zur Strassen. 2001. The genus Scirtothrips in Mexico: a critique of the review by Johansen and Mojica Guzman (1998). Folia Entomologica Mexico 40: 133 142. Mukhopadhyay, A., S. Sannigrahi, and G. Biswas. 1997. Colonization and utilization of young tea plants by some insects and mites in Darjeeling foothills. Ecology and evolution of plant -feeding insects in natural and man -made environments.: 59 69. Nagata, T., A. K. Inoue -Nagata, J. C. van Lenteren, R. Goldb ach, and D. Peters. 2002. Factors determining vector competence and specificity for transmission of Tomato spotted wilt virus. Journal of General Virology 83: 663 671. Nair, V. V., P. Raghunath, and A. Viskalakshmi. 1991. Control of the thrips Scirtothrips dorsalis on rose. Entomon 16: 327 329. Nathan, R. 2001. The challenges of studying dispersal. Trends in Ecology and Evolution 16: 481 483. Nathan, R., G. Perry, J. T. Cronin, A. E. Strand, and M. L. Cain. 2003. Methods for estimating long -distance d ispersal. Oikos 103: 261 273. Nault, B. A., J. Speese III, D. Jolly, and R. L. Groves. 2003. Seasonal patterns of adult thrips dispersal and implications for management in eastern Virginia tomato fields. Crop Protection 22: 505 512. Nietschke, B. S., D M. Borchert, R. D. Magarey, and M. A. Ciomperlik. 2008. Climatological potential for Scirtothrips dorsalis (Thysanoptera: Thripidae) establishment in the United States. Florida Entomologist 91: 79 86. Okada, T., and I. Kudo. 1982a. Overwintering sites and stages of Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) in Tea fields. Japanese Journal of Applied Entomology and Zoology 26: 177 182. Okada, T., and I. Kudo. 1982b. Relative abundance and phenology of Thysanoptera in a Tea field. Japanese Jou rnal of Applied Entomology and Zoology 26. Okada, T., M. Otaishi, and T. Kaneko. 1981. Diurnal change in activity of adults of Scirtothrips dorsalis Hood. Japanese Journal of Applied Entomology and Zoology 48: 44 46. Onkarappa, S., and B. Mallik. 1998. D istribution and management of Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) on rose, pp. 165 167. In P. P. Reddy, N. K. Kumar and A. Verghese [eds.], Advances in IPM for horticultural crops. Indian Institute of Horticultural Research, Bangalore, I ndia.

PAGE 131

131 Palmer, C., and E. Vea. 2008. IR4 Ornamental Horticulture Program Thrips Efficacy: Chili Thrips ( Scirtothrips dorsalis ), Gladiolus Thrips ( Thrips simplex ), Weeping Fig Thrips (Gynaikothrips uzeli ), Western Flower Thrips ( Frankliniella occidentalis ), pp. 76, Ornamental Summary Reports. Rutgers University, Princeton, New Jersey. Panickar, B. K., and J. R. Patel. 2001. Population dynamics of different species of thrips on chili, cotton, and pigeonpea. Indian Journal of Entomology 63: 170 175. Paul, S. and A. Konar. 2005. Ecological study on the occurrence of chili thrips in plains of West Bengal. Indian Journal of Ecology 32: 102 103. Pearsall, I. A. 2002. Daily flight activity of the western flower thrips (Thysan., Thripidae) in nectarine orchards in British Colombia, Canada. Journal of Applied Entomology 126: 293 302. Pearsall, I. A., and J. A. Myers. 2000a. Evaluation of sampling methodology for determining the phenology, relative density, and dispersion of western flower thrips (Thysanoptera: T hripidae) in Nectarine Orchards. Journal of Economic Entomology 93: 494 502. Pearsall, I. A., and J. A. Myers. 2000b. Population dynamics of western flower thrips (Thysanoptera: Thripidae) in nectarine orchards in British Columbia. Journal of Economic En tomology 93: 264 275. Peng, R. K., K. Christian, and K. Gibb. 2004. Implementing ant technology in commercial cashew plantations and continuation of transplanted green ant colony monitoring, pp. 1 42. Rural Industries Research and Development Corporati on, Canberra, Australia. Pramanick, C., and M. D. Mohasin. 2004. Relationships between plant morphological characters and incidence of chili thrips Scirtothrips dorsalis Hood in District Nadia of West Bengal. Environment and Ecology 22: 880 883. Radler, W. J. 2001. Shrub rose plant named 'RADrazz', pp. 1 3. In U. S. P. Office [ed.], Plant Variety Database. Conard -Pyle Incorporated, USA. Raizada, U. 1965. Life history of Scirtothrips dorsalis Hood with detailed external morphology of the immature stage. Bulletin of Entomology 6: 30 49. Raizada, U. 1976. A preliminary report on the fungi infesting thrips (Thysanoptera: Insecta). Entomon 1: 155 157. Ramakrishna Ayyar, T. V. 1932. Bionomics of some thrips injurious to cultivated plants in South India. Agriculture and Livestock of India 2: 391 403. Rani, B. J., and V. Sridhar. 2003. Screening of polyhouse grown rose varieties for resistance to thrips, Scirtothrips dorsalis Hood. Journal of Ornamental Horticulture (New Series) 6: 165 171.

PAGE 132

132 Reddy, D. V. R ., A. S. Ratna, M. R. Sudarshana, R. Poul, and I. K. Kumar. 1992. Serological relationships and purification of bud necrosis virus, a Tospovirus occurring in peanut ( Arachis hypogaea L.) in India. Annals of Applied Biology 120: 279 286. Reddy, D. V. R., M. R. Sudarshana, A. S. Ratna, A. S. Reddy, P. W. Amin, I. K. Kumar, and A. K. Murthy. 1991. The occurrence of yellow spot virus, a member of tomato spotted wilt group on peanut ( Arachis hypogaea L.) in India, pp. 77 88. In H. T. Hsu and R. H. Lawson [ed s.], Virus Thrips -Plant Interactions of TSWV. Proceedings of the USDA Agricultural Research Service. Reitz, S. R. 2002. Seasonal and within -plant distribution of Frankliniella thrips (Thysanoptera: Thripidae) in north Florida tomatoes. Florida Entomologist 85: 431439. Reitz, S. R., J. E. Funderburk, and S. M. Waring. 2006. Differential predation by the generalist predator Orius insidiosus on congeneric species of thrips that vary in size and behavior. Entomologia Experimentalis et Applicata 119: 179 188. Rhainds, M., and L. Shipp. 2003. Dispersal of adult western flower thrips (Thysanoptera: Thripidae) on chrysanthemum plants: impact of feeding induced senescence of inflorescences. Environmental Entomology 32: 1056 1065. Rhainds, M., L. Shipp, L. Woodro w, and D. Anderson. 2005. Density, dispersal, and feeding impact of western flower thrips (Thysanoptera: Thripidae) on flowering chrysanthemum at different spatial scales. Ecological Entomology 30: 96 104 Sakai, S. 2001. Thrips pollination of androdioec ious Castilla elastica (Moraceae) in a seasonal tropical forest. American Journal of Botany 88: 1527 1534 Sakurai, T., T. Inoue, and T. Murai. 2002. Intraspecific variation in transmission of TSWV by Frankliniella occidentalis results from distinct viru s accumulation, pp. 51 57. In L. A. Mound and R. Marullo [eds.], Thrips and Tospoviruses: Proceedings of the 7th International Symposium on Thysanoptera. CSIRO Entomology, Reggio Calabria, Italy. Saxena, P., M. R. Vijayaraghavan, R. K. Sarbhoy, and U. Raizada. 1996. Pollination and gene flow in chillies with Scirtothrips dorsalis as pollen vectors. Phytomorphology 46: 317 327. Schuch, U. K., R. A. Redak, and J. A. Bethke. 1998. Cultivar, fertilizer, and irrigation affect vegetative growth and susceptibi lity of chrysanthemum to western flower thrips. Journal of the American Society for Horticultural Science 123: 727 733. Schultz, J. C. 1989. Potential causes of the pear thrips outbreak in sugar maple, pp. 113 129. In B. L. Parker, M. Skinner and T. Le wis [eds.], Towards Understanding Thysanoptera. Proceedings of the International Conference on Thrips. USDA Technical Report NE 147, Radnor, PA.

PAGE 133

133 Seal, D. R., and R. M. Baranowski. 1993. Effectiveness of different insecticides for the control of melon thrip s, Thrips palmi Karny (Thysanoptera: Thripidae) affecting vegetables in south Florida. Procedings of the Florida State Horticultural Society 105: 315 319. Seal, D. R., and M. Ciomperlik. 2004. Scirtothrips dorsalis Hood on St. Lucia and St. Vincent: Report of an investigation, 14 23 January 2004, pp. 19. USDA -APHIS -PPQ. Seal, D. R., M. A. Ciomperlik, M. L. Richards, and W. Klassen. 2006a. Distribution of Chili thrips, Scirtothrips dorsalis (Thysanoptera: Thripidae) in pepper fields and pepper plants on St. Vincent. Florida Entomologist 89: 311 320. Seal, D. R., M. A. Ciomperlik, M. L. Richards, and W. Klassen. 2006b. Comparative effectiveness of chemical insecticides against the chili thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae), on pep per and their compatibility with natural enemies. Crop Protection 25: 949 955. Shibao, M. 1996a. Effects of temperature on development of the chili thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae), on grape. Applied Entomology and Zoology 31: 81 86. Shibao, M. 1996b. Damage analysis of chillie thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) on grape. Japanese Journal of Applied Entomology and Zoology 40: 293 297. Shibao, M. 1996c. Effects of a vinyl film cover on the populatio n density of the chillie thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae), on grape. Applied Entomology and Zoology 31: 174 177. Shibao, M. 1997a. Effects of Resource Availability on population density of the chillie thrips, Scirtothrips dors alis Hood (Thysanoptera: Thripidae), on grape. Applied Entomology and Zoology 32: 413 415. Shibao, M. 1997b. Effects of insecticide application on population density of the chillie thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae), on grape. A pplied Entomology and Zoology 32: 512 514. Shibao, M., F. Tanaka, and F. Nakasuji. 1990. Seasonal changes and infestation sites of the chillie thrips, Scirtothrips dorsalis (Thysanoptera: Thripidae) on grapes. Japanese Journal of Applied Entomology and Z oology 34: 145 152. Shibao, M., A. Hosomi, and H. Tanaka. 2000. Seasonal fluctuation in percentage parasitism of the yellow tea thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) by an egg parasitoid of Megaphragma (Hymenoptera: Trichogrammatid ae) on grapes. Entomological Science 3: 611 614. Shibao, M., F. Tanaka, K. Fujisaki, and F. Nakasuji. 1993. Effects of lateral shoot cutting on population density of the chillie Thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) on Grape. Appli ed Entomology and Zoology 28: 35 41.

PAGE 134

134 Shibao, M., S. Ehara, A. Hosomi, and H. Tanaka. 2004. Seasonal fluctuation in population density of phytoseiid mites and the yellow tea thrips, Scirtothrips dorsalis Hood (Thysanoptera: Thripidae) on grape, and predat ion of the thrips by Euseius sojaensis (Ehara) (Acari: Phytoseiidae). Applied Entomology and Zoology 39: 727 730. Shibao, M., S. Ehara, A. Hosomi, and H. Tanaka. 2006. Effect of insecticide application on the population density of yellow tea thrips, Scir tothrips dorsalis Hood (Thysanoptera: Thripidae) and Euseius sojaensis (Ehara) (Acari: Phytoseiidae) on grapes. Japanese Journal of Applied Entomology and Zoology 50: 247252. Silagyi, A. J., and W. N. Dixon. 2006. Assessment of chili thrips, Scirtothrips dorsalis Hood, pp. 9. Division of Plant Industry, Gainesville, Florida. Sims, K., J. Funderburk, and D. Boucias. 2005. Host -parasite biology of Thripinema fuscum (Tylenchida: Allantonematidae) and Frankliniella fusca (Thysanoptera: Thripidae). Journal of N ematology 37: 4 11. Singh, H. 2005. Thrips incidence and necrosis disease in sunflower, Helianthus annuus L. Journal of Oilseeds Research 22: 90 92. Skarlinsky, T. 2004. Identification Aid for Scirtothrips dorsalis Hood, 1919, pp. 6. In USDA [ed.]. u npublished. Skarlinsky, T. L. 2003. Survey of St. Vincent pepper fields for Scirtothrips dorsalis Hood, pp. 5. USDA -APHIS -PPQ, Miami, Florida. Slansky, J. F., and J. G. Rodriquez. 1987. Nutritional ecology of insects, mites, spiders and related invertebrat es. In J. F. Slansky and J. G. Rodriquez [eds.], Nutritional Ecology. John Wiley and Sons, New York. Takanashi, M., K. Takagi, and Y. Hirose. 1996. Egg parasitism by Megaphragma spp. of Scirtothrips dorsalis and some techniques for determining it. Plant Pr otection Bulletin of Japan 50: 61 64. Tatara, A. 1994. Effect of temperature and host plant on the development, fertility, and longevity of Scirtothrips dorsalis Hood (Thysanoptera: Thripidae). Applied Entomology and Zoology 29: 31 37. Tatara, A., and K. Furuhashi. 1992. Analytical study on damage to satsuma mandarin fruit by Scirtothrips dorsalis Hood (Thysanoptera : Thripidae), with particular reference to pest density. Japanese Journal of Applied Entomology and Zoology 36: 217 223. Taylor, L. R. 1974. Insect Migration, Flight Periodicity and the Boundary Layer The Journal of Animal Ecology 43: 225 238. Tschuch, G., P. Lindemann, and G. Moritz. 2002. Chemical defence in thrips, pp. 277 278. In L. A. Mound and R. Marullo [eds.], Thrips and Tospovi ruses: Proceedings of the 7th International Symposium on Thysanoptera. CSIRO Entomology, Reggio Calabria, Italy.

PAGE 135

135 Tsuchiya, M., S. Masui, and N. Kuboyama. 1995. Color attraction of yellow tea thrips (Scirtothrips dorsalis Hood). Japanese Journal of Applied Entomology and Zoology 39: 299303. Varadharajan, S., and R. Veeravel. 1995. Population dynamics of chili thrips Scirtothrips dorsalis Hood in Annamalainagar. Indian Journal of Ecology 22: 27 30. Varghese, T. S., and R. S. Giraddi. 2005. Integration of neem cake in the plant protection schedule for thrips and mite management in chili (cv. Byadagi). Karnataka Journal of Agricultural Science 18: 154 156. Venette, R. C., and E. E. Davis. 2004. Chili thrips/ yellow tea thrips, Scirtothrips dorsalis Hood (Th ysanoptera: Thripidae) Mini Pest Risk Assessment. University of Minnesota, Saint Paul, MN. Wheatley, A. R. D., J. A. Wightman, J. H. Williams, and S. J. Wheatley. 1989. The influence of drought stress on the distribution of insect on four groundnut genotyp es grown near Hyderabad, India. Bulletin of Entomological Research 79: 567 578. Whitfield, A. E., D. E. Ullman, and T. L. German. 2005. Tospovirus thrips interactions. Annual Review of Phytopathology 43: 549 589. Zaman, Z., and B. Maiti. 1994. Insects and mites infesting seedlings of mango in West Bengal. Environment and Ecology 12: 734 736.

PAGE 136

136 BIOGRAPHICAL SKETCH While born in Houston, Texas in 1977, Andrew Derksen was raised a citizen of the world sometime between the Permian and the late Mesozoic. He earned his undergraduate in biology and anthropology from Grinnell College in 2000. He misses the mountains, and would like to be working with molecular paleontology but finds that life is full of unexpected turns and it is best to be prepared no matt er where your path in life takes you. He received his Masters in Entomology from the University of Florida in the summer of 2009, and works for the state of Florida as a Pest Survery Specialist for the Cooperative Agricultural Pest Survey Program (CAPS) in Miami.