Temperature Dependent Development, Host Range, and Distribution of Cricotopus lebetis (Diptera

MISSING IMAGE

Material Information

Title:
Temperature Dependent Development, Host Range, and Distribution of Cricotopus lebetis (Diptera Chironomidae), a Natural Ememy of Hydrilla verticillata (Hydrocharitaceae) in Florida
Physical Description:
1 online resource (121 p.)
Language:
english
Creator:
Stratman, Karen N
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Entomology and Nematology
Committee Chair:
Overholt, William A
Committee Co-Chair:
Cuda, James P
Committee Members:
Netherland, Michael D
Wilson, Patrick C

Subjects

Subjects / Keywords:
biocontrol -- chironomidae -- cricotopus -- hydrilla -- lebetis
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:
A chironomid midge, Cricotopus lebetis Sublette (Diptera: Chironomidae), was discovered feeding on hydrilla in Crystal River, Citrus Co., Florida in 1992, and may be a recent introduction into Florida. Larvae of the midge mine the apical meristems of hydrilla, causing basal branching and stunting of the plant. We investigated the distribution, temperature-dependent development and host range of the midge. The midge was found in a four of six Florida water bodies surveyed, but it was rarely abundant. The relationship of temperature to larval-pupal development revealed that midge survival was highest at temperatures between 20 and 30°C, and the developmental rate increased with increasing temperature. Results of laboratory host range studies showed that the fundamental host range of C. lebetis included not only hydrilla, but a variety of aquatic plants in several different families, suggesting that this insect may not be a hydrilla specialist. Dual-choice tests with adult females demonstrated that C. lebetis exhibited a preference for certain host plants, and that adults are responsible for choosing suitable sites for larval development. Results from the survey work indicated that C. lebetis is present in several different water bodies throughout Florida, but the factors responsible for its current distribution remain unknown. The results obtained in this thesis provide a better understanding of the abiotic and biotic factors influencing the biology of C. lebetis and its potential use as an augmentative biological control agent. This information will be used to determine how C. lebetis can be exploited in developing long-term management strategies for hydrilla in Florida.
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 Karen N Stratman.
Thesis:
Thesis (M.S.)--University of Florida, 2012.
Local:
Adviser: Overholt, William A.
Local:
Co-adviser: Cuda, James P.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044711:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 TEMPERATURE DEPENDEN T DEVELOPMENT, HOST RANGE, AND DISTRIBUTION OF CRICOTOPUS LEBETIS (DIPTERA: CHIRONOMID AE), A NATURAL EMEMY OF HYDRILLA VERTICILLAT A (HYDROCHARITACEAE) I N FLORIDA By KAREN STRATMAN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Karen Stratman

PAGE 3

3 I would like to thank all who encouraged my intellectual curiosities, especially my advis ors Bill Overholt and Jim Cuda This is also dedicated to my mother and father who have supported my academic interests all throughout my collegiate career. Without these people this milestone would not have been possible

PAGE 4

4 ACKNOWLEDGMENTS I thank my committee Chair William Overholt, co Chair James Cuda and supervisory committee members P. Chris Wilson and Michael Netherland for their mentoring and assistance throughout this project. I also would like to acknowledge the rest of the staff working in the lab (University of Florida, Indian River Research a nd Education Center ( IRREC ) ) as well as county extension agents who have been very Water Management District), Raymond Hix (Florida A&M University), Stacia Hetrick, and Dean Jones (Osceola County Extension) for assisting during sampling for survey work. Finally, I would like to thank my parents who have supported my education and encouraged me to pursue a degree in science. This project was supported by the Hydrilla Integrated Pest Management Risk Avoidance and Mitigation Project ( IPM RAMP ) grant and the Florida Fish and Wildlife Conservation Commission (FWC )

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 LITERATURE REVIEW ................................ ................................ .......................... 13 Introduction ................................ ................................ ................................ ............. 13 Background ................................ ................................ ................................ ............. 14 Biological Control ................................ ................................ ............................. 14 Hydrilla verticillata ................................ ................................ ............................ 16 Description ................................ ................................ ................................ 16 Distribution ................................ ................................ ................................ 16 Introduction of h ydrilla into the United States ................................ ............. 17 Neg ative e ffects ................................ ................................ ......................... 17 Beneficial e ffects ................................ ................................ ........................ 18 Hydrilla C ontrol ................................ ................................ ................................ ....... 19 Mechanical and Chemical Control ................................ ................................ .... 20 Biological Control ................................ ................................ ............................. 21 Cricotopus lebetis The Hydrilla Tip Miner ................................ ............................ 24 Goals and Hypotheses ................................ ................................ ............................ 28 2 TEMPERATURE DEPENDENT DEVELOPMENT, COLD TOLERANCE, AND POTENTIAL DISTRIBUTION OF CRICOTO PUS LEBETIS (DIPTERA: CHIRONOMIDAE), A TIP MINER OF HYDRILLA VERTICILLATA (HYDROCHARITACEAE) ................................ ................................ ....................... 34 Introduction ................................ ................................ ................................ ............. 34 Materials and Methods ................................ ................................ ............................ 36 Source and Culturing of H. verticillata and C. lebetis ................................ ....... 36 Survival and Developmental Time ................................ ................................ .... 37 Cold Tolerance ................................ ................................ ................................ 37 Developmental Rate and Degree Day Requirement ................................ ........ 38 Linear Developmental Rate Model ................................ ................................ ... 38 Nonlinear Developmental Rate Model ................................ .............................. 38 Weather Data from Florida ................................ ................................ ............... 39 Calculation of Degree days and Number of Generations for Geographic In formation System (GIS) Analysis ................................ ............................... 39

PAGE 6

6 Generation of GIS Map for Prediction of C. lebetis Generations in Florida ...... 40 Climatic Suitability Mapping ................................ ................................ .............. 40 Results ................................ ................................ ................................ .................... 41 Survival and Developmental Time ................................ ................................ .... 41 Cold Tolerance ................................ ................................ ................................ 41 GIS Mapping of C. lebetis Generations in Florida ................................ ............ 42 Climatic Suitability Mapping ................................ ................................ .............. 42 Discussion ................................ ................................ ................................ .............. 42 3 HOST RANGE OF CR ICOTOPUS LEBETIS (DIPTERA: CHIRONOMIDAE), A TIP MINER OF HYDRILLA VERTICILLATA (HYDROCHARITACEAE) .................. 55 Introduction ................................ ................................ ................................ ............. 55 Materials and Methods ................................ ................................ ............................ 58 Source and Culturing of H. verticillata and C. lebetis ................................ ....... 58 No Choice Larval Development Tests ................................ .............................. 58 Paired Choice Test ................................ ................................ ........................... 59 Paired Choice Olfactometer Test ................................ ................................ ..... 60 Host Finding Behavioral Test ................................ ................................ ........... 60 Paired Choice Adult Oviposition ................................ ................................ ....... 61 Data Analysis ................................ ................................ ................................ ... 61 Results ................................ ................................ ................................ .................... 62 No Choice Larval Development ................................ ................................ ........ 62 Paired Choice Test ................................ ................................ ........................... 63 Paired Choice Olfactometer Test ................................ ................................ ..... 63 Host Finding Behavior Test ................................ ................................ .............. 64 Paired Choice Adult Oviposition ................................ ................................ ....... 65 Discussion ................................ ................................ ................................ .............. 65 4 THE INFLUENCE OF WATER QUALITY ON CRICOTOPUS LEBETIS (DIPTERA: CHIRONOMIDAE): A HERBIVORE OF HYDRILLA VERTICILLATA (HYDROCHARITACEAE) ................................ ................................ ....................... 85 Introduction ................................ ................................ ................................ ............. 85 Materials and Methods ................................ ................................ ............................ 87 Chironomid Diversity ................................ ................................ ........................ 87 Water Quality ................................ ................................ ................................ .... 88 Pesticide Analysis ................................ ................................ ............................ 88 Fipronil D ose Response ................................ ................................ ................... 89 Data Analysis ................................ ................................ ................................ ... 90 Results ................................ ................................ ................................ .................... 91 Chironomid Diversity ................................ ................................ ........................ 91 Water Quality ................................ ................................ ................................ .... 92 Pesticide Analysis ................................ ................................ ............................ 93 Fipronil Dose Response ................................ ................................ ................... 93 Effect of Alkalinity and Hardness on C. lebetis Larval Development ................ 93 Discussion ................................ ................................ ................................ .............. 93

PAGE 7

7 5 SUMMARY OF FINDINGS ON T HERMAL REQUIREMENTS, HOST RANGE AND DISTRIBUTION OF CRICOTOPUS LEBETIS (DIPTERA: CHIRONOMIDAE), A NATURAL ENEMY OF HYDRILLA VERTICILLATA (HYDROCHARITACEAE) ................................ ................................ ..................... 104 Introduction ................................ ................................ ................................ ........... 104 Recommendation s ................................ ................................ ................................ 10 4 Use of C. lebetis in Florida ................................ ................................ ............. 104 Water t emperature ................................ ................................ ................... 105 Water q uality ................................ ................................ ............................ 105 Potential effe cts on native f lora ................................ ................................ 106 Integrated pest m anagement (IPM) ................................ ......................... 107 Use of C. lebetis Outside of Florida ................................ ................................ 108 Concluding Remarks ................................ ................................ ............................. 108 LIST OF REFERENCES ................................ ................................ ............................. 110 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 121

PAGE 8

8 LIST OF TABLES Table page 3 1 List of plants tested in no choice larval development ................................ ......... 71 4 1 Chironomid species richness at each location fro m each sampling period ......... 97 4 2 Chironomid diversity at each location for each sampling period ......................... 97 4 3 Chironomid species sampled from all locations during survey work ................... 98 4 4 Water quality variables at field sites fro m January 2011 May 2012 .................... 99

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Cricotopus lebetis ................................ ................................ ............................... 30 1 2 Cricotopus lebetis adult struc tures. ................................ ................................ ..... 31 1 3 Cricotopus lebetis larval feeding damage in hydrilla ................................ ........... 32 1 4 Hydrilla collected from Lake Rowell exhibiting highly branched stems caused by larval feeding damage ................................ ................................ ................... 33 2 1 Percent survival of C. lebetis larvae at various temperatures during the tem perature dependent study ................................ ................................ ............ 48 2 2 Developmental rate of C. lebetis larvae exposed to various temperatures during temperature dependent d evelopment ................................ ..................... 49 2 3 Brire 1 model estimating upper and lower tempera ture thresholds .................. 50 2 4 Larval survival of C. lebetis at diff erent exposure times at 5C and 7.5C .......... 51 2 5 Map showing the isothermal lines (LT 50 and LT 90 (lethal time) ) at 5 and 7.5C for C. lebetis ................................ ................................ ................................ ....... 52 2 6 Geographical information system ma p showing the predicted number of generations of C. lebetis in Florida per year. ................................ ...................... 53 2 7 Model prediction of climate suitability for C. lebetis using known sampling locations and climate records ................................ ................................ ............. 54 3 1 Aquatic olfactom eter setup used in the dual choice test ................................ ..... 72 3 2 Ex ample of movement analysis for c orrel ated r a ndom w alk (CRW) ................... 73 3 3 Survival of C. lebetis larvae on various aquatic plants un der no choice conditions ................................ ................................ ................................ .......... 74 3 4 Developmental rate of C. lebetis larvae on various aquatic plants under no choice conditions ................................ ................................ ................................ 75 3 5 Percent survival and development rate of C. lebetis larvae on host plants tested ................................ ................................ ................................ .................. 76 3 6 Percent of plant tips infested with C. lebetis larvae under paired choice conditions ................................ ................................ ................................ ........... 77

PAGE 10

10 3 7 Da mage scores to plants by C. lebetis larvae under paired choice conditio ns. ................................ ................................ ................................ ......... 78 3 8 Results from the paired choice olfactometer t e st ................................ ................ 79 3 9 Relationship between mean squared displacement of C. lebetis larvae and t ime in Petri dish areas ................................ ................................ ....................... 80 3 10 Time spent in each quadrant of Petri dis h ................................ .......................... 81 3 11 Trace patterns of larval movement in a Petri dish from hos t finding behavior test ................................ ................................ ................................ ...................... 82 3 12 Total number of egg masses laid in each paired choice adult oviposition trial. .. 83 3 12 Total n umber of egg masses laid in each paired choice adult oviposition trial. .. 84 4 1 Sampling locations for survey data in Florida wher e hydrilla for chironomid diversity study and water quality variables were collected ................................ 100 4 2 Index of prevalence of chironomids r ecovered from all sampling locations from January 2011 May 2012. ................................ ................................ .......... 101 4 3 Fipronil laboratory bioassay with C. lebetis larvae ................................ ........... 102 4 4 Response of C. lebetis to different levels of water alkalinity/hardness ............. 103

PAGE 11

11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science TEMPERATURE DEPENDENT DEVELOPMENT, HOST RANGE AND DISTRIBUTION OF CRICOTOPUS LEBETIS (DIPTERA: CHIRONOMIDAE) A NATURAL EMEMY OF HYDRILLA VERTICILLATA (HYDROCHARITACEAE) IN FLORIDA By Karen Stratman August 2012 Chair: William A. Overholt Co Chair: James P. Cuda Major: Entomology and Nematology A chironomid midge, Cricotopus lebetis Sublette (Diptera: Chironomidae), was discovered feeding on hydrilla in Crystal River, Citrus Co., Florida in 1992 and may be a recent introduction into Florida. Larvae of the midge mine th e apical meristems of hydrilla, causing basal branch ing and stunting of the plant. We investigate d the distribution, temperature dependent development and host range of the midge. The midge was found in a four of six Florida water bodies surveyed but it was rarely abundant. The relationship of temperature to larval pupal development revealed that midge survival was highest at temperatures between 20 and 30C, and the developmental rate increased with increasing temperature. Results of laboratory host rang e studies showed that the fundamental host range of C. lebetis included not only hydrilla but a variety of aquatic plants in several different families, suggesting that this insect m ay not be a hydrilla specialist. D ual choice tests with adult females dem onstrated that C. lebetis exhibited a preference for certain host plants, and that adults are responsible for choosing suitable sites for larval development. Results from

PAGE 12

12 the survey work indicated that C. lebetis is present in several different water bodie s throughout Florida, but the factors responsible for it s current distribution remain unknown. The results obtained in this thesis provide a better understanding of the abiotic and biotic factors influencing the biology of C. lebetis and its potential use as an augment at ive biological control agent. This information will be used to determine how C. lebetis can be exploited in developing long term management strategies for hydrilla in Florida.

PAGE 13

13 CHAPTER 1 LITERATURE REVIEW Introduction Invasive species h ave the capability to disrupt natural ecosystems ( Dukes and Mooney, 2004 ). Florida, with a unique island type biogeography, is particularly vulnerable to invasion by non indigenous plant species. Florida is similar to an island because it is surrounded on three sides by water and the fourth side by a freeze line. Islands are known to have a depauperate flora and fauna, and thus have less biotic resistance to invasion by exotic organisms (Simberloff et al., 1997). It is estimated that 1,400 exotic plant spec ies have been naturalized in south Florida alone. Approximately 70 of these introduced species have become problematic and require extensive management (Rodgers et al., 2011). Invasive species pose ecological threats, and costs associated with invasive wee ds in the USA are estimated to be $35 billion annually, with $110 million attributed to aquatic weeds (Pimentel et al., 2005). Management of invasive plants relies to a large extent on the use of chemical herbicides, which can impact hydrologic systems, re creation, crop irrigation, drinking water, aquatic organisms, and aquaculture (USGS, 2006). Herbicides can be non selective, cause reoccurring costs to managers, and the use of herbicides often lacks public support. Several common herbicides have been dete cted in sediments, and fish and mollusk populations in both urban and agri cultural streams (USGS, 2006). Moreover, several plant populations have become resistant to chemical herbicides (Holt and Lebaron, 1990; Powles and Yu, 2010). Biological control may provide an alternative to herbicide application, and can be used as part of an Integrated Pest Management (IPM) plan. Classical biological control

PAGE 14

14 is based on the premise that plant populations are suppressed in their native ranges by the action of natural enemies ( Williams, 1954 ; Keane and Crawley, 2002). When a plant is moved to a new area of the world, it escapes regulation by specialized natural enemies and is able to reach higher densities than found in its aboriginal home ( Wolfe, 2002 ; Torc hin et al., 2003 ). The process of biological control involves searching for host specific natural enemies in the native range of the weed species, and releasing the natural enemies in the invasive range of the weed. When successful, biological control is an economica lly efficient and sustainable method to manage invasive plants. Biological control often is used as a component in an IPM program along with mec hanical and chemical methods. Hydrilla verticillata (L.f. Royle) has been a problematic aquatic weed since its introduction to the United States in the 1950s (Schmitz et al., 1991), and has been the focus of biological control exploration since the 1970s (Balciunas and Minno, 1985; Buckingham, 1994). Augmentative biological control is mass rearing and releasing ins ects to reduce pest populations. This project focus ed specifically on a tip mining midge Cricotopus lebetis Sublette (Diptera: Chironomidae) as a potential augmentative biological control agent of hydrilla. Background Biological Control Invasive species p ose a threat to the biodiversity of ecosystems. Invasive plants compete directly with native vegetation and cause problems further up the food chain to both invertebrates and vertebrates (Dukes and Mooney, 2004). Classical biological control is regarded as a practical and affordable way to manage invasive weed species in order to reduce the risks that unwanted species pose to natural ecosystems. Other

PAGE 15

15 benefits of biological control are that it can be effective on its own, or can be used in conjunction with other control meth ods. Biological control may be particularly appropriate for sensitive or conservation areas where mechanical / chemical controls are not permissible or are ineffective (Julien et al., 2007). When proper research methodologies are followed, biological control can be successful and classical biological control is considered the most successful method of biological control Of the biological control agents released, 60% have established, and 33% of those have resulted in some degree of control (McFayden, 1998). K nowledge about a biological control agent can make the difference between successful and unsuccessful establishment in a new environment. The biology and life cycle of the candidate agent must be fully understood before its potential a s a biological control agent can be evaluated. Some core questions that must be considered with any weed biological control agent are: 1) Is the species host specific; 2) Will the agent reach densities high enough to have a significant impact on the target plant; 3) Will the agent be able to perform over a wide range of environmental conditions; 4) How will the success of the biological control agent be measured (Julien et al., 2007). Knowing the answers to these questions will increase the success rate of biological control. Because of increasing concern about the risks associated with biological control, careful risk assessment is essential for obtaining approval for release of a new biological control agent. The historical record indicates that very few b iological control agents have had adverse effects, such as negative impacts to human health, non target effects to native species, enhancing pest species, and becoming pests themselves (Howarth, 1991; Louda et al., 2003). Risk assessment of biological cont rol agents is

PAGE 16

16 difficult because of the complexity of predicting commun ity and ecosystem wide impacts. Developing cost benefit and risk assessments for every biological control agent would minimize these deleterious effects and improve the overall efficacy of classical biological control (Simberloff and Stiling, 1996). Proper assessment of host range is an essential component in biological control programs. Host range testing reveals if a potential biological control agent is sufficiently specialized on the target plant to receive approval for release. Host range evaluat ion involves determining the fundamental host range of the organism and predicting its field host specificity (Sheppard et al., 2005). Hydrilla verticillata Description Hydrilla is a submersed rooted aquatic weed in the family Hydrocharitaceae. Hydrilla c an be either monoecious or dioecious. Stems are long and slender with some branching. The leaves are small (20 mm long and 4 mm wide), lanceolate, and occur in whorls of 3 8. The midrib is distinct and sometimes bears small spines. Male flowers are solitar y and released underwater as buds, which float and open at the surface of the water. Female flowers are inconspicuous with three transparent petals occasionally with reddish streaks (Cook and Lnd, 1982). Distribution Hydrilla is widely distributed and oc curs in Europe, Asia, Australia, New Zealand, Pacific islands, Africa, South America, and North America. Hydrilla occurs in temperate areas, but thrives in tropical regions (Langeland, 1996).

PAGE 17

17 Introductio n of h ydrilla into the United States There have been two separate introductions of hydrilla into the USA, with a dioecious form found in the southeast and parts of California, and a monoecious form found in the northeast, California and Washington ( Steward et al., 1984 ; Madeira et al., 2000 ). Only the dioec ious female biotype is known to occur i n Florida. The dioecious form was imported into the USA in the late 1950s from Sri Lanka through the aquarium trade and rapidly spread throughout Florida during the 1960s and the rest of the southeastern USA in the 1970s (Schmitz, 1991). The pathway of introduction of the monoecious form into the northeastern USA is unknown. The USA dioecious population is most closely related to a population found in Bangalore, India whereas the USA monoecious population is closely related to plants from Seoul, Korea (Madeira et al., 1997). A major concern is that hydrilla will continue to spread in the northern USA and cause problems similar to those it has created in the southern USA. Ten years after into Florida, it was established in major water bodies of all drainage basins in Florida (Langeland, 1996). Outside of Florida, hydrilla has spread along the Gulf Coast, extending up the Atlantic Coast to Maryland and Delaware, and into the western states of California, Arizona, and Washington (Madeira et al., 2000). Negative e ffects There are contradictory research findings about the impact of hydrilla on native communities (Haller and Sutton, 1975; Hoyer et al., 2008). Haller and Sutton (1975) found tha t hydrilla displaces nati ve vegetation such as Vallisneria neotropicalis Marie Vict), whereas Hoyer et al. (2008) indicated that it has no significant impact on native plant and animal diversity, species richness, and abundance because it has occupied a va cant niche (Hoyer et al., 2008) Hydrilla has the ability to grow in shallow waters to

PAGE 18

18 depths of 20 feet or greater. It also grows rapidly and can effectively compete for sunlight. Once it reaches the surface, hydrilla branches profusely producing a dense mat that shades out native vegetation (Langeland, 1996). In highly infested areas, hydrilla can create dense mats of vegetation impeding boat traffic, and it often becomes entangled in boat propellers, thus facilitating its movement to new locations withi n water bodies, and to new water bodies. Hydrilla can produce a large quantity of propagules and meristems, which gives it a competitive advantage over native plants. The dense mats of hydrilla support filamentous algae and small invertebrates, which furth er limit light penetration (Haller and Sutton, 1975). Although hydrilla may provide beneficial effects for a select number of aquatic species, deleterious effects may occur to native flora. In drainage canals, hydrilla can greatly reduce flow and cause flo oding and canal damage. Negative effects associated with hydrilla invasion include interference with recreational and commercial use of lakes, tourism and sportfishing, reduction of real estate values, reduction of flow in drainage canals, and clogging of intake pipes (Schmitz et al., 1991 ; Langeland, 1996 ). Beneficial e ffects its management. Hydrilla provides shelter, breeding and oviposition sites for many aquatic organis ms. Hydrilla attracts many small invertebrates, but relatively few of these species use living hydrilla tissue as a food source (Balciunas and Minno, 1985). Waterfowl, most commonly coots ( Fulica americana Gmelin) and ringneck ducks ( Aythya collaris Donova n), consume the stems leaves and tubers of hydrilla as a food source (Esler, 1989). Studies have shown that hydrilla supported a high diversity of duck species, and increases in hydrilla were correlated with increases in waterfowl

PAGE 19

19 populations (Esler, 1990 ). Waterfowl enthusiasts argue that protection of hydrilla is important to the maintenance of healthy waterfowl populations (Johnson and Montalbano, 1984). Largemouth bass anglers also argue that there is a positive relationship between aquatic vegetation and favorable bass fishing. Dense stands of aquatic vegetation can provide habitats for invertebrates, which are consumed by sportfish (Moxley and Langford, 1982). Among anglers, largemouth bass fishermen are most opposed to vegetation management, and som e would prefer to have more aquatic vegetation (Slipke et al., 1998). Highly dense patches of vegetation may negatively impact bass populations, but maintaining moderate levels of vegetation can prove to be beneficial (Brown and Maceina, 2002). Largemouth bass anglers often view aquatic vegetation as beneficial to good quality fishing regardless of the extent of the coverage ( Wilde et al., 1992 ; Slipke et al., 1998). The everglade snail kite ( Rostrhamus sociabilis plumbeus ) is a federally listed endangered species. Snail kites utilize the exotic apple snail ( Pomacea insularum ) as a food source. Foraging snail kites require snails within 6 inches of the surface, and snails nested in the top of hydrilla plants can be easily captured. The exotic apple snail ha s invaded Lake Toho, and snail kite populations have increased since its introduction. Since the snail kite has benefited from exotic apple snails nested in hydrilla, treatments of hydrilla in Lake Toho have been reduced (FWS, 2010). d negative effects give rise to conflicts of interest in creating management solutions for this invasive weed. Finding balance between management solutions that seek to greatly reduce the abundance of hydrilla, and solutions that seek

PAGE 20

20 to manage hydrilla in ways to benefit waterfowl populations is a difficult task. Those who favor little management of hydrilla to protect waterfowl habitat see the current hydrilla management measures as potentially harmful (Johnson and Montalbano 1987). Control strategies that allow maintenance of some hydrilla, while reducing its adverse effects, are needed. Biological control may prove to be a solution acceptable to both sides, because biological control does not seek complete eradication of a weed species, but rather a reduction in density. Hydrilla Control Mechanical and Chemical Control Due to the diversity of water uses in Florida, effective control of hydrilla is difficult to achieve because of a very limited number of environmentally soun d options (Hoyer et al., 2005). Efforts to control hydrilla rely primarily on the application of synthetic herbicides, the most common of which used to be fluridone. Fluridone received approval by the EPA in 1986 (Arias et al., 2005) and until 2004 was use d to control hydrilla with little effect on native vegetation (Doong et al., 1993). Fluridone is a phytoene desaturase (PDS) inhibitor, and under high light intensity causes bleaching of the green photosynthetic tissues (Chamovitz et al., 1993 ; Boger and S andman, 1998 ). With millions of dollars spent annually on herbicide applications, this may not be the most sustainable or economically efficient manner to control hydrilla. Recently, fluridone resistance has been documented at several locations in central Florida (Michel et al., 2004). Typically, resistance does not occur in plants that do not reproduce sexually (Maxwell and Mortimor, 1994) because recombination of alleles and selection of herbicide resistant genes occurs more frequently in sexually repro ducing plants All known hydrilla in Florida is dioecious female, and reproduces asexually by

PAGE 21

21 subterranean tubers, turions, root crowns, and stem pieces. A mutation in the PDS gene and subsequent selection for resistance is thought to be responsible for th e observed herbicide resistance (Chamovitz et al., 1993; Michel et al., 2004). Endoreduplication, which causes variable ploidy levels in hydrilla, can result in gene duplication. The lity to adapt to its environment and may have contributed to the rapid development of herbicide resistance (Puri et al., 2007). Fluridone resistance has resulted in the inability to control large infestations of hydrilla with the herbicide (Michel et al., 2004). Increased amounts of fluridone must be applied in order to manage resistant populations. Alternative herbicides such as endothall, diquat, and chelated copper have replaced fluridone. Endothall is now the most commonly used herbicide to control hyd rilla; however, it is reported to provide only 4 8 months of control compared to 1 2 years of control with fluridone (Hoyer et al., 2005). In order to be effective, endothall must be applied in high concentrations, and requires a long exposure time (Nether land et al., 1991). Heavy reliance on endothall for hydrilla control is cause for concern because over time resistance to endothall may also develop and increased tolerance to endothall has been recorded in Lake Maitland. (Mike Netherland, personal commun ication). Moreover, fluridone replacement herbicides are more expensive and treatment can range from $700 1600 per ha, compared to treatment costs ranging from $125 600 per ha for fluridone (Arias et al., 2005). Biological Control Biological control of hy drilla has been investigated since the 1970s. Seve ral natural enemies have been identified, a few have been released, but none have been

PAGE 22

22 able to provide long term control. Exploration for natural enemies has been conducted in Asia, Australia, and Africa ( B uckingham, 1994; Balciunas et al., 2003; Overholt and Cuda, 2005). Insects from these locations were brought back to Florida for additional research, and four were approved for release, including, two Bagous spp. weevils and two ephydrid flies in the genus Hydrellia, (Buckingham, 1988 ; ; Buckingham and Bennett, 2001). Bagous affinis Hustache (Coleoptera: Curculionidae) is a weevil from India that was first introduced into F lorida in 1987 (Buckingham, 1988 ). During the dry season, lar vae feed on exposed vegetation during low water conditions resulting in reduced sprouting rates of tubers (Buckingham 1988; Godfrey and Anderson, 1994). The larvae burrow into the soil and feed on below ground tubers, and in doing so, they prevent the tube rs from sprouting. Due to their unique adaptation to seasonal flooding and drought cycles in their native range, which do not occur in Florida, permanent populations of B. affinis failed to establish (Godfrey et al., 1994). Bagous hydrillae that feeds on hydrilla stems. Adults feed on expo sed stems and leaves whereas larvae feed and develop only inside submersed stems. During larval development, they fragment the stem and larvae c ontinue to feed on the stem until they reach maturity (Buckingham and Balciunas, 1994). Bagous hydrillae was first introduced in Florida in 1991 but failed to establish (Balciunas and Purcell, 1991). Flooding and drought are thought to be major abiotic mor tality factors of this biological control agent (Buckingham and Balciunas, 1994).

PAGE 23

23 Of the four insects approved for release, only one ephydrid fly in the genus Hydrellia established in Florida, namely Hydrellia pakistanae Deonier (Diptera: Ephydridae). Anot her ephydrid, Hydrellia balciunasi Bock (Diptera: Ephydridae), established in Texas, but failed to establish in Florida (Center et al., 1997; Grodowitz et al., 1997). In a laboratory setting, Hydrellia pakistanae consumed 60 70% of hydrilla apical leaves. Field studies revealed that H. pakistanae populations never reached high enough densities to cause the level of damage observed in the laboratory, and the highest whorl damage in the field was 15%. Field studies also revealed that 100% larval mortality occ urred at 36 C, which greatly compromised the effectiveness during warm summer months (Cuda et al., 2008; Wheeler and Center, 2001). An adventive Asian insect, Parapoynx diminutalis, Snellen (Lepidoptera: Pyralidae), was discovered feeding on hydrilla in Fort Lauderdale, Florida in 1976 (Del Fosse et al., 1976). The larvae make cases out of their host plants, attach themselves to the leaf surface and feed on the leaves and stems of hydrilla (Buckingham and Bennett, 1996). Host range testing of this aquatic moth revealed that it fed on a variety of aquatic plants, and therefore was not evaluated further for its potential for augmentative biological control (Buckingham and Bennett, 2001). The moth can still be found in Florida, although i t was never purposely released. In addition to insects, grass carp ( Ctenopharyngodon idella Val.) have been examined as augmentative biological control agents of hydrilla. Although not host specific, grass carp preferentially feed on hydrilla stems and leaves (Sutton and Vandi ver, 1986). Small grass carp can consume quantities as great as their body weight per day, and 8 10% of body weight of larger fish is due to hydrilla consumption

PAGE 24

24 (Shireman and Maceina, 1981). Grass carp are exotic, and after their introduction into the Uni ted States, they spread faster than any other introduced fish, and in many cases are considered a nuisance species (Guillory and Gasaway, 1978). Triploid carp, which are functionally sterile, were produced in order to preclude reproduction of individuals r eleased for vegetation control. Stocking water bodies with sterile grass carp can result in the removal of all aquatic vegetation, even in large systems (Klussmann et al., 1988). When properly stocked, grass carp can provide long term continuous control of aquatic vegetation, but the probability of predicting vegetation changes following a grass carp release is very low due to dynamic processes (Ppalov, 2006). Although various biological control measures have been investigated, these efforts have met litt le success in the field, with the possible exception of sterile grass carp. Fluridone resistance has created a renewed interest in searching for new biological control agents of hydrilla (Overholt and Cuda 2005 ; Cuda et al., 2008 ). Chemical, physical, and mechanical methods for controlling hydrilla provide only short term control and are expensive. The discovery of a viable biological control agent may be the key to successful long term, environmentally sustainable management of hydrilla populations. Cricot opus lebetis The Hydrilla Tip Miner Chironomidae is a dipteran family of non biting midges. They have a cosmopolitan distribution, and inhabit almost any habitat that is aquatic or wet (Wirth, 1949). The majority of chironomids are microphagous; feeding on algae, animals and detritus, (Oliver, 1971). There are two types of larvae, free living and sedentary. The sedentary types construct and pupate in protective cases. When the pupa is ready to emerge as an adult, it rises to the surface of the water and a dult eclosion occurs. This is a critical

PAGE 25

25 adults live for a short period of time and have reduced non feeding mouthpart s. The main function of the adult life is mating and reproduction. Water quality is an important variable in determining midge densities and distributions. Due to their sensitivity to water quality, chironomid communities have been used as indicators of water quality and pollution (Saether, 1979). Chiro nomids also make good water quality indicators because of their widespread distribution in water bodies of varying quality, and their habitat preferences (Paine and Gaufin, 1956). Runoff from surrounding land uses can contain harmful pesticides that impact water quality. Chironomids have been shown to be very sensitive to the presence of pyrethroid pesticides (Anderson, 1989). Given that certain midge species are sensitive to water quality, this factor should be taken into consideration when developing biol ogical contro l programs that rely on midges. A chironomid midge was first discovered in 1992 in Kings Bay attacking the stems of hydrilla (Cuda et al., 2002). The hydrilla was stunted and did not reach the surface as is typical of hydrilla. The origins of this i nsect are unknown. I t could have been introduced with hydrilla, or could be a native species that expanded its host range to utilize hydrilla. The larvae, which can easily be recognized by the diagnostic blue band around the second and third thoracic segments, was identified as Cricotopus lebetis Sublette (Epler et al., 2000) (Fig. 1 1) The following description of adults is provided by Epler et al. (2000). brown markings; these colors fade to pale brown/stra mineous wit h dark brown to brownish markin gs in alcohol preserved material. In alcohol preserved material, brown to dark brown on antennae, head, throacic vittae (vittae sometimes jointed posteriorly by diffuse brown area), scutellum, postnotum, median ane pisterum II and approximate ventral half of epimeron. Wings clear with light brown veins;

PAGE 26

26 halteres pale. Legs with fore and hind coxae light, mid coxa brown; all trochanters light; fore femur light brown basally, much darker in apical 1 / 3 to ; mid and hin d femora basally light with brown apical 1 / 3 to ; tabiae with brown basal and apical bands, fore tibia slightly darker in middle than mid and hind tabiae; fore tarsi brown, mid and hind tarsi light brown to stramineous. (Fig. 1 Larvae mine into the a pical stem of hydrilla and feed on meristematic tissues. In preparation for pupation, the larvae mine into the stem, and eventually cause abscission of the tip (Cuda et al., 2002) (Fig. 1 3). Cricotopus lebetis has only been found attacking hydrilla, excep t for one specimen collected from a Potamogeton s p. (Dana Denson, personal communication) and therefore may have a limited host range but testing is needed to confirm this hypothesis Cricotopus lebetis may have value as an augmentative biological control agent because it prevents hydrilla from reaching the flow, and decrease the likelihood of hydrilla becoming entangled in boat propellers and thereby spreading to new l ocations (Fig. 1 4). This type of damage has the potential to increase light penetration, thereby increasing the overall biodiversity of the communi ty. Unlike most other chironomid species, C. lebetis feeds on living plant tissues, as opposed to detritus a nd other substrates. Feeding on living plant tissue is rare, but has been previously observed in Cricotopus myriophylli Oliver, a congener that feeds on the invasive aquatic plant Eurasian milfoil ( Myriophyllum spicatum L.). Studies indicated that C. myrio phylli had a narrow host range and only fed on two milfoil species (Macrae et al., 1990). Since one species of Cricotopus exhibited a degree of host specificity, the host range of C. lebetis may also be restricted to one or a few species. The first record of this insect was in 1957 in Louisiana (Sublette, 1964), but hydrilla was not observed in Louisiana until the 1970s (Cook and Lnd, 1982), which suggests that either hydrilla

PAGE 27

27 arrived in Louisiana earlier than thought, or the midge has hosts other than hy drilla. Water quality variables also may play a key role in determining the distribution of the midge. Conducting surveys in lakes across Florida, and obtaining water quality parameters may help explain the presence or absence of the midge in certain water bodies. There are various factors affecting the success of an introduction of a new organism. One of the most important factors is temperature. Understanding the thermal requirements of an insect can help in understanding the potential areas for establish ment of C. lebetis. Among the outcomes of the temperature dependent development studies are degree day requirements, which are the degree days that are required to complete one generation, and the establishment of upper and lower temperature thresholds for development. Integration of temperature developmental data with Geographical Information Systems (GIS) interpolation functions allows for production of maps showing the number of generations of C. lebetis may experience in Florida. Cold tolerance studies reveal whether or not an organism is tolerant to cold temperatures, and using this data isothermal lines can be generated to estimate the limits of northern distributions. With the exception of basic information on its biology (Cuda et al., 2002) and impa ct on hydrilla ( Schmid, et al., 2010; Cuda et al., 2011), there is little information responses to variation in water quality. Characterization of these factors may help determine whether th is midge has the potential to be a successful biological control agent of hydrilla. Once research has been conducted, an IPM program for hydrilla can be developed,

PAGE 28

28 incorporating augmentative release of C. lebetis if appropriate. If the midge is shown to be highly host specific, then it may be possible to receive approval for release of the midge in states where it does not occur. Goals and Hypotheses The overall goal of this study was to better understand the biology, host specificity, and habitat requireme nts of C. lebetis. A series of experiments was conducted in the laboratory, greenhouse, and under field conditions to assess the following hypotheses: Hypothesis 1: Water temperature is a determining factor for survival and development of C. lebetis. Objec tive 1: Determine the influence of temperature on developmental rate and survival of C. lebetis Objective 2: Generate a map predicting the number of generations/year of C. lebetis across Florida. Objective 3: Predict the USA distribution of the midge based on its physiological tolerance to cold. Hypothesis 2: Cricotopus lebetis is a preferential tip miner of Hydrilla verticillata. Objective 1: Determine the host range of C. lebetis under no choice and choice conditions Hypothesis 3: Water quality (pestici des, pH, and water hardness and alkalinity) affects the distribution of C. lebetis in Florida. Objective 1: Estimate chironomid diversity in hydrilla tips in Florida lakes and correlate water quality to the presence or absence of C. lebetis (Fig. 1 5).

PAGE 29

29 Obj ective 2: Based on results of field surveys, conduct laboratory studies to determine the influence of selected water quality parameters on C. lebetis survival and development.

PAGE 30

30 A B C D Figure 1 1. Cricotopus lebetis. A) egg mass B) larva C) pupa D) adult. Photo credit: Jerry F. Butler, University of Florida

PAGE 31

31 Figure 1 2. Cricotopus lebetis adult structures. 1) Male fore, mid and hind legs, 2) Male abdomen 3) Hypopygium 4 5) Inferior olsella variation in Florida material 6) Inferior volsella 7) Variation of gonostylus de to angle observation 8) Female genitalia, ventral 9) Female genetalia, lateral 10) Female coxosternapodeme. Source: Epler et al. 2000.

PAGE 32

32 Figure 1 3. Cricotopus lebetis larval feeding damage in hydrilla

PAGE 33

33 Figure 1 4. Hydrilla collected from Lake Rowell exhibiting highly branched stems caused by larval feeding damage, September 2010

PAGE 34

34 CHAPTER 2 TEMPERATURE DEPENDENT DEVELOPMEN T, COLD TOLERANCE, A ND POTENTIAL DISTRIBUTI ON OF CRICOTOPUS LEBETI S (DIPTERA: CHIRONOMIDAE), A TIP MINER OF HYDRILLA VERTICILLAT A (HYDROCHARITACEAE) Introduction Hydrilla verticillata (L.f. Royle) is a rooted submersed aquatic macrophyte that can be found throughout Florida, and other parts of the USA. There have been at least two separate introductions of hydrilla into the USA, with a dioecious form found predominately in the southeast, and a monoecious form found predominately in the northeast (Madeira, et al., 2000). The dioecious form of hydrilla was imported into the USA in the late 1950s from Sri Lanka through the aquarium trade. It rapidly spread throughout Florida during the 1960s and into the rest of the southeastern USA in the 1970s (Schmitz, 1991). The pathway of introduction of the monoecious form into the nort heastern USA is unknown. Hydrilla has been shown to displace native vegetation such as V allisneria neotropicalis (Haller and Sutton, 1975). Hydrilla grows rapidly and can eff ectively compete for sunlight. Once it reaches the surface, dioecious hydrilla br anches profusely producing a dense mat of vegetation, thus shading out native flora (Langeland, 1996). Additional negative impacts include interference with recreational and commercial use of lakes, tourism, and sportfishing, reduction of real estate value s, reduction of flow in drainage canals, and clogging of intake pipe s (Schmitz et al., 1991; Langeland, 1996 ). In highly infested areas, hydrilla creates dense mats of vegetation impeding boat traffic, and often becomes entangled in boat propellers, thus f acilitating its movement to new locations within water bo dies, and to new water bodies.

PAGE 35

35 Effective control of hydrilla is difficult to achieve because of a very limited number of environmentally sound options (Hoyer et al., 2005). Efforts to control hydrill a had relied primarily on the application of synthetic herbicides, specifically fluridone. Recently, fluridone resistance has been documented in several locations in central Florida and has resulted in the use of alternative herbicides such as endothall an d acetolactate synthase ( ALS ) inhibitors. Therefore, new management approaches to control hydrilla populations are being investigated (Cuda and Gillett Kaufman, 2011) Biological control is one possible management approach, used either alone or integrated with other tactics. In 1992, the chironomid midge Cricotopus lebetis Sublette (Diptera: Chironomidae) was discovered damaging apical meristems of hydrilla in Crystal Ri ver Florida (Cuda et al., 2002). Although not certain, the midge was thought to be an adventive species, as it was not discovered in the USA until 1957 in Louisiana (Sublette 1964). Cuda et al. (2002) speculated that the midge might have been introduced i nto the USA as a contaminant of hydrilla imported through the aquarium trade. Hydrilla damaged by the midge at Crystal River was atypically stunted and did not reach the surface. In preparation for pupation, the larvae mine into the stem, and eventually th is causes abscission of the tip (Cuda et al., 2002). This type of damage has the potential to satisfy both those who desire hydrilla stands (e.g. bass fishermen, duck hunters), and those who want to control it. Preventing hydrilla from reaching the water s urface would improve boat navigation and water flow, and decrease the likelihood of hydrilla becoming entangled in boat propellers and thereby spreading to new locations. Despite its potential as a biological control agent of hydrilla, additional research to supplement the work comp leted by Epler et al. (2000), Cuda et al. (2002 2011) Schmid

PAGE 36

36 et al. (2010) and Cuda and Gillett Kaufman (2011) is needed to address general biology and impact of C. lebetis on hydrilla. The purpose of this study was to determi ne the influence of temperature on survival and developmental rate and to use this information to generate maps predicting the average number of generations of the midge in Florida, and the predicted isothermal lines for the north ern extent of the distribu tion. Materials and Methods Source and Culturing of H. verticillata and C. lebetis Hydrilla was collected from Lake Tohopekaliga (Toho), Osceola Co., FL (28.2 N, 81.4 W), and C. lebetis was collected from Lake Rowell, Bradford Co., FL (29.9 N, 82.1 W). Both cultures were maintained at the Biological Control Research and Containment Laboratory (BCRCL), Fort Pierce, FL. Cricotopus lebetis was reared by placing hydrilla tips in a large aer ated container within a cage constructed from PVC tubing covered with a fine mesh cloth. Containers were filled with well water, C. lebetis egg masses were placed in the containers and adults that emerged were collected using a mouth aspirator. Adults were transferred to a 250 ml separatory funnel that had approximately 15 ml of well water as described by Cuda et al. (2002). Females oviposit on the water surface and egg masses were collected by opening the sto pcock on the separatory funnel. Hydrilla was pro pagated from stems collected at Lake Toho and placed into 10.16 cm pots containing a layer of potting soil covered by sand. The pots were placed into 378 liter tanks in a greenhouse and covered with 60% shade cloth.

PAGE 37

37 Survival and Developmental Time Temper ature dependent development of C. lebetis was investigated in environmental chambers (Model No. E36L, Percival Scientific, Inc., Perry, IA) maintained at 10 constant temperatures (10, 15, 20, 22, 25, 27, 30, 32, 35, 36 C 1C). Photoperiod was kept constan t at 14:10 (L:D). Healthy, undamaged plant tips, 4 6 cm in length, were placed individually in 35 ml test tubes filled with well water as described by Cuda et al. (2002). Each test tube was placed in a rack that held 40 tube s Two newly hatched larvae were transferred to each plant tip using a pipette. Once the larvae were introduced into the tubes, a cap with ventilation holes was placed on each tube. Tips were checked daily to ensure that they were fully submerged in order to prevent larval desiccation. T ubes with tips that were completely destroyed received replacement tips to allow complete development to adulthood Approximately one week after the larvae were introduced, the test tubes were checked daily for adult emergence. The number of days to complete development was recorded in order to calculate the development rate. Cold Tolerance Cold tolerance studies were conducted using 2 nd 4 th instar larvae. Four insects were placed inside a 35 ml vial containing two hydrilla tips and well water. Inse cts were acclimated from 20 C to the final temperature in intervals of 5 C every two hours. The larvae were exposed to three constant temperatures (5 C, 7.5 C, 10 C) for 0.5, 1, 2, 4, 8, 16, and 32 days. After each exposure time, insects were placed at roo m temperature and survival was assessed by observing for movement once the water reached room temperature. The effect of temperature and exposure times on midge survival was analyzed using logistic regression (SAS Institute, 2008). The LT 50 and LT 90 (letha l time)

PAGE 38

38 at 5 C and 7.5 C were used to predict isothermal lines delineating regions favorable for C. lebetis establishment based on historical weather data. Following methods outl ined by Diaz et al. ( 2008 ) a model was created in NAPPFAST (Borchert and Mag arey, 2007), a database of daily weather information from stations across North America, to record the number of days at or below 5 o C and 7.5 o C. Probability maps were generated using the last 10 years of weather data to examine the frequency of occurrenc e of reaching the LT 50 and LT 90 The maps were imported into ArcGIS 9.0 and a line indicating a frequency of occurrence of at least 5 out of the last 10 years was created. Developmental Rate and Degree Day Requirement Survival at different temperatures w as analyzed with analysis of variance using the general linear model procedure and means were separated using Student Newman Keuls test (PROC GLM; SAS Institute 2008). Linear Developmental Rate Model Developmental rate at different temperatures was analyze d using linear regression. The linear portion (15 35C) of the developmental rate curve [R(T) = a + bT)] was modeled using least squares regression in Excel (Microsoft, Redmond, WA), where T = temperature, a = intercept, and b = slope. The base temperatur e threshold was estimated by the intersection of the regression line and the x axis (R(T) = 0). Degree days were calculated as the inverse slope of the fitted regression line. Nonlinear Developmental Rate Model The nonlinear relationship between developmen t rate R(T) and temperature T was analyzed using the Brire 1 model which allows estimation of the upper and lower developmental thresholds (Brire et al., 1999). The model is defined as R(T) = a T (T

PAGE 39

39 T 0 ) (T L T) 1/2 where R = developmental rate, T = tem perature, T 0 = base temperature threshold, T L = lethal temperature, and a = empirical constant. T 0 and T L were initially set to 6 and 36 C, respectively, and the equation was then solved iteratively. Weather Data from Florida Daily minimum and maximum tem peratures from Florida were obtained from 91 weather stations through the Applied Climate Information System (Climate Information for Management and Operational Decisions [CLIMOD], Southeast Regional Climate Center; http://acis.sercc.com ). Daily minimum and maximum temperatures were averaged from the last 5 11 years depending the availability of data, which provided 365 values for maximum and minimum temperatures for each station. When there were no missing data point s the maximum period of weather data was from 1 January 2002 to 1 January 2012. The minimum period of data available varied, as data points were inconsistently missing from the series. Calculation of Degree days and Number of Generations for Geographic Inf ormation System (GIS) Analysis The DegDay program version 1.01, which is an Excel (Microsoft, Redmond, WA) application developed by University of California Davis (http;//biomet.ucdavis.edu) was used to calculate accumulated degree days for C. lebetis Thi s application uses the upper and lower temperature threshold for an organism, and daily average of minimum and maximum temperatures to calculate accumulated degree days by using a single sine method. (Baskerville and Emin, 1969). The lower and upper tempera ture thresholds were estimated from the Briere 1 nonlinear model as 9.52 and 36C, respectively. The linear regression model was used to calculate the degree days (K) for C. lebetis [R(T) =

PAGE 40

40 a bT] as K = 1/b (Campbell et al., 1974). The prediction of the nu mber of generation per year was calculated by dividing the cumulative degr ee days per station by K. Generation of GIS Map for Prediction of C. lebetis Generations in Florida Weather station name, latitude, longitude, and number of C. lebetis generations pe r year were inserted in Microsoft Excel and imported into ArcGIS 9.0 (ESRI Inc., Redlands, CA). The imported file was converted to a shape file using the ADD X Y DATA function followed by the selection of the State Plane Projection. A shapefile of the bord er of Florida was obtained from the AWhere Continental database (AWHERE, Inc., Denver, CO) and used to deli neate the range of predictions. The Geostatistical Analysis function in ArcGIS (ESRI, Inc.) was used to generate prediction grids of C. lebetis gener ations across Florida. Values at un sampled locations were predicted by interpolation of values at sample locations. The inverse distance weighted (IDW) deterministic method was used, where predictions are made by mathematical formulas that generate weight ed averages of nearby known values. The IDW model gives more influence to points that are closer than to ones that are farther away. The parameters used in the IDW analysis were as follows: 1. The number of stations used for interpolation was set to a maximum of 15 and minimum of 10. 2. The Power Optimization option was selected generating a Power value of p = 2. This weights weather station values proportional to the inverse distance raised to the power of p. 3. The search neighborhood shape was circular because t here were no directional influences on the weighting of number of generations per station. Ellipse parameters were set to: angle, 0 major and minor semiaxis, 1020596. Climatic Suitability Mapping Geographic coordinates from sample locations of C. lebetis were obtained from voucher specimens, literature and known field collection sites, including my own data

PAGE 41

41 and data provided by Dana Denson (Reddy Creek Improvement District), Doug Strom (Water and Air Associates, Inc.) and Robert Rutter (Florida Department of Environmental Protection, Punta Gorda Branch Office). The predicted distribution of C. lebetis in North America was generated from known records and selected climatic variables using the BIOCLIM model (Hijmans et al. 2012), in the freeware program DIVA GIS. Climatic variables included in the model were maximum temperature in warmest month, minimum temperature in coldest month, and annual mean temperature. Results Survival and Developmental Time Larval survival varied with temperature. Larvae could not c omplete development at low and high temperature extremes (10 and 36C). Only a single individual was able to complete development at 35C. Survival to adulthood was highest at temperatures between 20 and 30C (Fig. 2 1). The development rate increased with increasing temperature, u ntil reaching 32C (Fig. 2 2). Degree day requirements (K) were calculated to be 495.29. The Brire 1 model estimated the lower and upper developmental thresholds at 9.52 and 36C, respectively (Fig. 2 3). These values were very similar to those found in the laboratory tests. The model showed that the rate of development increased with temperature until the curve reached an optimum at 30C, and then decreased rapidly as the temperature approached the upper developmental threshold (Fig. 2 3). Cold Tolerance Larval survival when exposed to 5C for a prolonged period exceeded 4 days in most cases. Only 50% and 10% of insects were able to survive for 8 and 16 days, respectively. No insects were able to survive after 32 days exposure to 5C (Fig. 2 4)

PAGE 42

42 but larvae were able to survive past 16 days10C. The isothermal lines showed that at both 5 and 7.5C, 50% of C. lebetis individuals will experience mortality just north of Florida, in northern Louisiana and Georgia, and in north Texas. Th e lines also predicted survival in California and southwestern Arizona. The lines indicating 90% mortality occur further north into southeast South Carolina, extending through the southeast and stopping in the panhandle of Texas, and then starting again in central Arizona to north California (Fig. 2 5). GIS Mapping of C. lebetis Generations in Florida Based on degree day requirements, C. lebetis is predicted to complete several generations per year in Florida, ranging from 6.8 to 11.7, with most generation s in the southern portion of the state, and fewest in the panhandle (Fig. 2 6). Florida counties located south of Palm Beach County had the highest number of generations ranging from 10.2 to 11.7. Counties in the middle portion are predicted to support 8.2 to 10.2 generations per year. Climatic Suitability Mapping The predicted distribution indicated that the climate is suitable for the establishment of C. lebetis throughout much of the southeastern United States. The highest suitability occurred in Florid a, and southern Louisiana. The full extent of the prediction ranged from southern South Carolina to central and east Texas. The map also indicated that areas in California and Arizona were suitable for establishment (Fig. 2 7). Discussion Temperature is an important factor that determines community structure and insect distribution. Understanding the thermal requirements for C. lebetis is an

PAGE 43

43 important step for predicting where the insect can establish. As expected, the predicted number of generations per y ear of C. lebetis increased as latitude decreased. CLIMOD data only includes air temperatures, but was used to generate the map because comprehensive water temperature data for Florida water bodies are unavailable. Water temperature data collected from sam pling water bodies throughout Florida in 2011 2012 indicate that temperatures were below the upper developmental threshold (Chapter 2). Cricotopus lebetis develops in the tips of hydrilla, which grow near the water surface during summer months. As water su rface temperature is correlated with air temperature (McCombie, 1959), using air temperature data for predictive modeli ng was a reasonable assumption. Knowing the number of generations that the midge can complete in a year is important, as it will influence population growth in the field. Further research will be required to validate the prediction of number of generations per year and to correlate that with population dynamics in the field. Host plant quality also can affect the develo pmental rate of insects. It has been shown that insects developing on suboptimal hosts typically have longer developmental times, and can be more vulnerable to predation due to longer exposure to predators (Hggstrm and Larsson, 1995). For example, in Cha pter 3 I show that the developmental time of C. lebetis varies de pending on host plant species. Within a species, host plants can vary due to genotype, ontogeny, phenology, and environmental conditions. Van et al. (1977) demonstrated that the growth habit s of hydrilla vary depending on water depth and light intensity, and these physiological changes to hydrilla could conceivably influence performance of the midge. Thus, the actual number

PAGE 44

44 of C. lebetis generations per year may be lower than what was predict ed due to changes in host plant quality. Field sampling has shown that C. lebetis occur rence is sporadic (Chapter 4). In summer months, the water temperature in Florida lakes can range from 30 35C (Beaver, et al., 1981) which is approaching the upper deve lopmental thresh old. Water temperatures in hydrilla mats near the water surface tend to be higher than water temperatures at a one meter depth in the hydrilla mat, or in open water (Bowes, et al., 1979; Cuda et al. 2008 ). In some portions of Florida, wate r temperatures in hydrilla mats may be too high for C. lebetis development, and result in high mortality. Vegetation mats reaching 45C have been reported (Wheeler and Center, 2001), which is well above the upper lethal threshold of C. lebetis Wheeler and Center (2001) partially attributed the poor performance of the ephydrid fly, Hydrellia pakistanae to high temperatures in hydrilla mats. Likewise, C. lebetis may suffer high mortality during the peak of the summer months, resulting in local extirpation. Thus, temperature extremes could play a critical role in determining the persistence of C. lebetis in Florida water bodies. If permanent populations of C. lebetis cannot be sustained in water bodies experiencing temperature extremes, then restocking effort s would have to be made periodically. The BIOCLIM model predicted that suitable locations for the establishment of C. lebetis exist throughout the southeastern USA. However, the data available to generate this map was limited, as extensive field sampling for C. lebetis has not been conducted. Although data was limited, this map resembles the distribution based on the isothermal cold tolerance lines, suggesting that the model prediction may be reasonable. Further

PAGE 45

45 sampling, both within and outside the predic ted distribution, would be useful for The results from the cold tolerance study indicate that C. lebetis can tolerate exposure to 5C for 0.5 4 days. High mortality did not occur until exposure time exceeded 8 days. This may indicate that C. lebetis has a broader lower temperature threshold compared to the upper threshold. At high temperatures approaching the upper developmental threshold, nearly 100% mortality occurred. High rates of survival were re corded at 30C, but there was a steep decline in survival once temperatures reached 35C. Wider ranges of lower temperature thresholds are common among insects (Bayoh and Lindsay 2004). The lower developmental threshold, the temperature at which the devel opmental rate is estimated to be zero, was 9.52C. Water temperature data are useful to determine sites where C. lebetis can survive and establish. Water temperatures during the winter months throughout Florida range from 8 15C (Beaver et al., 1981), an d these temperatures are mostly within the range for prolonged periods of time will be unsuitable for establishment of C. lebetis. The isothermal lines support this hypo thesis, as 50 and 90% mortality was predicted just north of Florida and north of southeastern South Carolina respectively. The isothermal lines also show that establishment may be possible in California and in southern Arizona, but no records of C. lebetis exist in those states. In North America, Cricotopus lebetis has been found only in Louisiana and Florida (Cuda et al., 2002), although it would seem likely that it also occurs in Mississippi and Alabama as well. The insect could be native to North America and was able to expand its host range to include

PAGE 46

46 2002). The overwintering strategy of C. lebetis is unknown. Chironomids have been shown to exhibit a variety of methods t o deal with cold temperatures including supercooling, freeze tolerance, and dispersal to new areas (Danks, 2007). It is also unknown what stage of this insect overwinters, although nearly all chironomids overwinter as larvae (Danks, 1971). Chironomids have been shown to survive short periods of extreme cold temperature by freeze tolerance (Bouchard et al., 2006). If C. lebetis exhibits any of these characteristics, then the predicted range may extend further north than what was shown in the prediction maps. Further studies investigating the freeze tolerance of this insect would aid in determining whether or not C. lebetis possesses overwintering adaptations. The global distribution and native range of this insect are unknown. Cuda et al. (2002) speculate th at the midge may have been introduced with hydrilla through the aquarium plant trade. If so, this does not provide much help in narrowing the source population since hydrilla is so widely distributed in the Old World. In order to determine the origin of th e midge, foreign exploration should be conducted to search for the midge in locations where hydrilla occurs. Then genetic work should be conducted on different geographical populations to determine the geographic origin. Temperature dependent development a nd cold tolerance studies provide basic information that can be used to develop or improve rearing methods, and to predict field colonization and establishment. Temperature experiments revealed that optimal temperature conditions for C. lebetis are from 20 30C, and explain why this insect is

PAGE 47

47 widely established throughout Florida. Ongoing studies of population dynamics and field impact studies will reveal the potential value of C. lebetis as an augmentative biological control agent.

PAGE 48

48 Figure 2 1. Percent survival of C. lebetis larvae at various temperatures during the temperature dependent study. Letter groupings represent statistically different means. 0 10 20 30 40 50 60 70 80 90 100 10 15 20 22 25 28 30 32 35 36 % Survival Temperature C a a a a a b b c d d

PAGE 49

49 Figure 2 2. Developmental rate of C. lebetis larvae exposed to various temperatures during temperature dependent development. Dots are observed values the line is the expected linear regression. R = rate; T = temperature. 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 10 15 20 25 30 35 40 Developmental rate (1/days) Temperature C R(T) = 0.0037 + 0.002 (T) R 2 = 0.721

PAGE 50

50 Figure 2 3. Brire 1 model estimating upper and lower temperature thresholds. Dots represent observe d values. Upper threshold: 36 C, Lower threshold: 9.52C 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 8 12 16 20 24 28 32 36 40 Developmental rate (1/days) Temperature C

PAGE 51

51 Figure 2 4. Larval survival of C. lebetis at different exposure times at A) 5C B) 7.5C. Single dots are observed values and lines are expected value of the logistic regression. Maximum survival occurred at 7.5 C at exposure time of 16 days. 0 20 40 60 80 100 0 5 10 15 20 25 30 35 40 % Larval survival A 0 20 40 60 80 100 0 5 10 15 20 25 30 35 40 45 Exposure time (days) B

PAGE 52

52 Figure 2 5. Map sho wing the isothermal lines (LT 50 and LT 90 ) at 5 and 7.5C for C. lebetis Lethal times at each exposure indicate after how many days in which 50 and 90% of the population will experience mortality. Maximum survival occurred at 18 days and 5C.

PAGE 53

53 Figure 2 6. Geographical information system map showing the predicted number of generations of C. lebetis in Florida per year.

PAGE 54

54 Figure 2 7. Model prediction of climate suitability for C. lebetis using known sampling locations and climate records Purple dots represent locations where C. lebetis has been recovered from this study, published literature, and voucher specimens. Map indicates suitability percentiles for establishment locations for C. lebetis

PAGE 55

55 CHAPTER 3 HOST RANGE OF CRICOTOPUS LEBETIS (DIPTERA: CHIRONOMID AE), A TIP MINER OF HYDRILLA VERTICILLAT A (HYDROCHARITACEAE) Introduction Hydrilla verticillata (L.f. Royle) is a submersed aquatic weed that has been the target of biological control programs since the 1970s. Hydrilla was impor ted into the United States through the aquarium trade in the late 1950s, and rapidly spread throughout Florida and the rest of the southeastern USA within 20 years (Schmitz et al., 1991). Shortly after its introduction, hydrilla was present in all major wa ter bodies of all drainage basins in Florida (Langeland, 1996). Hydrilla causes many negative impacts including displacement of native vegetation (Haller and Sutton, 1975), impediment of boat traffic, recreational and commercial losses, clogging of intake pipes and canals, and reductions in tourism and real estate values (Langeland, 1996; Schmitz et al., 1991). Managing hydrilla is both time consuming as expensive, and there are few effective management options (Hoyer et al., 2005). The long term use of the herbicide fluridone has resulted in the selection of several populations of fluridone resistant hydrilla in Florida, and has limited management options further, resulting in the inability to control large infestations (Michel et al., 2004). Biological control is one possible management approach, either alone or integrated with other tactics. There are three main types of biological control; classical, augmentation, and conservation (Cuda et al., 2008). Classical biological control is based o n the premise that plant populations are suppressed in their native ranges by the action of natural e nemies ( Williams, 1954 ; Keane and Crawley, 2002 ). When a plant is moved to a new area of the world, it escapes regulation by specialized natural enemies an d is able to reach higher densities than occur in its aboriginal home ( Wolfe,

PAGE 56

56 2002; Torchin et al., 2003 ). The process of biological control involves searching for host specific natural enemies in the native range of the weed species, and releasing the nat ural enemies in the invasive range of the weed. Classical biological control has been the most successful method of biological control. Augmentative biological control involves the release of natural enemies to supplement natural populations, and relies on mass rearing and continual release efforts. This method is used when natural enemies are unlikely to reach high enough densities on their own to achieve control of the pest species. Conservation biological control involves identifying factors that limit t he effectiveness of a natural enemy, and modifying them to increase the effectiveness of the beneficial species. This can involve either reducing the factors that limit effectiveness or providing resources that the natural e nemy needs in the environment. Host specificity testing is an essential component of any biological control progr am (Sheppard et al., 2005) and is used to determine whether a potential agent is suitable to be released in the field. The centrifugal phylogenetic method developed by Wapsh ere (1974) is followed to define the host range of an insect. Predicting non target effects of a potential agent is critical to assess ecological risks associated with releasing biological control agents (Louda and Arnett, 2000). Biological control has adv antages over oth er management options including diminished health concerns because there is less use of herbicides, (Pimental and Andow, 1984; Pimental et al., 1984). Moreover, decreased herbicide use will result in decreased selection pressure for plants to develop resistance to herbicides (Holt and Hochberg, 1997), and costs over time are less. An advantage of augmentative biological control over classical biological control is that release permits are not required to make field releases in areas where th e agent already occurs.

PAGE 57

57 Various insects from around the globe have been investigated as potential biological control agents of hydrilla, and a four have been released (Buckingham, 1988 ; 1; Overholt and Cuda, 2005). Unfortunately, none of the insects released have had significant impacts in the field (Buckingham and Balcionas, 1994; Wheeler and Center, 2001), and efforts to identify classical biological control agents are still ongoing. In 1992, the chironomid midge Cricotopus lebetis Sublette (Diptera: Chironomidae) was discovered damaging apical meristems of hydrilla in Crystal River, Florida (Cuda et al., 2002). Although not certain, the midge was thought to be an adventive species, as i t was not discovered in the USA until after the introduction of hydrilla (Epler et al., 2000). Moreover C. lebetis has only been collected from hydrilla, except for one specimen recovered from a Potamogeton s p. (Dana Denson, personal communication), sugge sting that it may be monophagous. The feeding habits and behavior of the insect also suggest that it may be specific to hydrilla, as it utilizes the apical stems to complete development (Cuda et al., 2002). This insect may have use as an augmentative biolo gical control agent since it prevents hydrilla from reaching the water surface, and therefore diminishes many of the negative effects of the weed (Cuda et al., 2011). Because C. lebetis already occurs in Florida waterways, permits are not required to relea se it within the state, but host specificity testing is required if the insect were to be released in other areas of the USA, or in other areas of the world, where it does not occur. The purpose of this study was to investigate the host range of C. lebetis and use this information to gauge its potential as a biological control agent of hydrilla.

PAGE 58

58 Materials and Methods Source and Culturing of H. verticillata and C. lebetis Hydrilla was collected from Lake Tohopekaliga (Toho) Osceola Co., FL (28.2 N, 81.4 W ), and C. lebetis was collected from Lake Rowell, Bradford Co., FL (29.9 N, 82.1 W). Both cultures were maintained at the Biological Control Research and Containment Laboratory (BCRCL), Fort Pierce, FL. Cricotopus lebetis was reared by placing hydrilla t ips in a large aerated container within a mesh cage constructed from PVC pipes. Containers were filled with well water, and C. lebetis masses were placed in the containers. Emergent adults were collected using a mouth aspirator, and transferred to a 250mL separatory funnel with approximately 15 mL of well water. Females oviposit on the water surface, and egg masses were collected by opening the stopcock on the separatory funnel (Cuda et al., 2002). Hydrilla was propagated from stems collected at the field s ite and placed into 10.16 cm pots containing a layer of potting soil covered by sand. The pots were placed into 378 liter tanks and covered with 60% shade cloth inside a greenhouse. No Choice Larval Development T ests Healthy, undamaged plant tips, 4 6 cm in length, were placed individually in 35 ml test tubes filled with well water. Each test included three non target plants and hydrilla as the control (Table 3 1). There were 10 test tubes per plant species and the tubes were placed randomly in a rack that held 40 tubes (Cuda et al., 2002). The experiment was replicated three times. Each tip was exposed to two newly hatched C. lebetis larvae by using a pipette to transfer the larvae to each test tube. Once the larvae were introduced into the tubes, a cap wi th ventilation holes was placed on each tube. All racks were placed in an environmental growth chamber maintained at 25 C and 14:10

PAGE 59

5 9 (L:D) photoperiod (Model No. E36L, Percival Scientific, Inc., Perry, IA) Tips were checked daily to ensure that they were fully submerged in order to prevent larval desiccation. Tubes with tips that were completely destroyed received replacement tips to allow complete development to adulthood Approximately one week after the larvae were introduced, the test tubes were checke d daily for adult emergence. Plant species that supported development to the pupal stage were considered to be suitable hosts. The number of days to develop was recorded in order to calculate the development rate on each plant species. Paired Choice T est P lant tips of hydrilla and other species that supported complete development ( Elodea canadensis Michx. Egeria densa Planch. and Najas guadalupensis (Spreng.) Magnus. ) were used in a dual choice experiment. Plant tips were placed in a container that was di vided into two sections using wire screen mesh with holes approximately 1 cm 2 Each side of the container had 40 plant tips, one side with hydrilla and the other side with a selected test plant. The container was placed into a small mesh cage, and aerated with an aquarium pump placed in the center of the container. In total, 100 neonates were placed in the center of the container and released. Larvae were left in the container to develop and after 10 days and plant tips were dissected under a microscope to determine the presence or absence of larvae. Damage to tips was rated on a scale of 0 5 with 0 no damage, 1 minimal damage, not visible to naked eye, 2 light damage 10 20%, 3 moderate damage 20 50%, 4 heavy damage >50%, 5 tip abscission. This process was replicated 3 times with each plant species.

PAGE 60

60 Paired Choice Olfactometer Test An aquatic olfactometer was constructed using a Y tube and a low flow peristaltic pump (Model No. 13 876 1, Fisher Scientific, Waltham, MA). The Y tube was glass, with an inside diameter of 4mm having two arms and a stem 5.3 cm and 5.5 cm in length, respectively. Distilled wat er was pumped at a rate of 0.575 ml/ min into each arm of the Y tube, and was measured by collecting water from each arm in a graduated cylinder f or 2 0 minutes and dividing the volume of water from each arm by 10. A small plant tip was placed in one arm of the Y tube and the other arm remained empty (Fig. 3 1). The Y tube was placed on a flat translucent surface that was lighted from the back. A cam era attached to a microscope was used to project the image onto a computer screen, and screen recording software was used to record all trials. A neonate was released into the stem of the Y tube and given a maximum of 10 minutes to move towards one of the olfactometer arms. Larvae that had not moved into one of the arms after 10 minutes were recorded as no response. A larva was recorded as having made a decision once it entered an arm of the Y tube. This process was replicated 10 times using hydrilla, E. ca nadensis and distilled water only. Host Finding Behavioral Test A small Petri dish, 3.5 cm in diameter, was divided into quadrants using an indelible marker, and a small circle was drawn in the middle of the dish to serve as the starting point. Two holes were drilled into opposite ends of the Pet ri dish and plastic tubing, 0.20 cm in inside diameter was inserted into each hole. Distilled water was pumped through the Petri dish using a peristaltic pump at a rate of 0 .25 ml /min to create a constant flow. A s ingle plant tip of either hydrilla or E. canadensis was placed in quadrant 4 so that the water current was flowing across the plant tip to the other side of

PAGE 61

61 the dish. A pipette was used to place a neonate into the center of the dish, and the larva was give n 60 minutes to locate the plant tip. Video imagery was captured through a dissecting microscope and recorded on a computer. The path of the larva was traced on an acrylic sheet, and the time spent in each quadrant was recorded. This process was replicated 10 times using hydrilla, E. canadensis, and a control with no plant material. The position of the insect was also determined every three minutes to examine their movement pathways (Fig. 3 2). Paired Choice Adult Oviposition A plastic divider 2.6 cm. in he ight was glued to the base of a cubic cage (29.8 cm on each side) to divide the base of the container in two. The cages were translucent plastic with mesh screen on three sides with a plastic top and bottom. Each side of the container was filled with 300 m L of distilled water, and 20 plant tips ( H. verticillata, N. guadalupensis, or E. canadensis) approximately 5 8 cm in length were placed into one side of the cage. The experiment also was conducted using pieces of artificial aquarium plants resembling hydr illa ( Walmart, Bentonville, AR ). Four C. lebetis pairs were released into the cage, where they had equal access to each container. Adults were left in the cages for 48 hours, after which the containers were examined for the presence of egg masses. The numb er and location of egg masses were recorded. Data Analysis Means from the no choice larval development, the paired choice test, and the time spent in each quadrant from the host finding behavior test were compared with analysis of variance and means separated with Student Newman Keuls test when ANOVAs were significant (SAS Institute, 2008). The relationship between survival and development rate from the no choice larval test was analyzed in Excel (Microsoft,

PAGE 62

62 Redmond, WA) using a least squared regressi on [R(S) = a + bS)], where T was the survival rate, R was the developmental rate, a was the intercept, and b was the slope, using all data points except zero values. Data from the aquatic olfactometer bioassay were analyzed using a G test of independence ( Sokal and Rohlf, 1995). The movement patterns of the host finding behavior test were examined by fitting the data to a correlated random walk model (CRW) as described in Brouwers and Newton (2009). The equation was as follows: R 2 n = n L 2 + (2 L 2 1 ) ( c / (1 c ) ) ( n (1 c ( n 1)/2 ) / (1 c )) where L 1 = m ean move length (here in cm ); L 2 = mean s quared move length (here in cm 2 ); n = number of consecutive moves; c = mean cosine of the turn angle. The mean square distances travelled in each phase (every 3 minutes) wer e compared to model predicted values with linear regression and the regression slopes were tested for equality to one (PROC REG, SAS Institute 2008). The results from the paired choice adult oviposition tested were analyzed using a G test of independence ( Sokal and Rohlf, 1995). Results No Choice Larval D evelopment Larvae were able to complete development on the majority of plants tested (Fig. 3 3). The plants that supported the best development were in the same family as hydrilla (Hydrocharitaceae) or the closely related family, Najadaceae (Fig. 3 4). High survival rates on some of the plants indicate that they may be better hosts for C. lebetis than dioecious hydrilla. Survival was higher on mon o ecious hydrilla (100%) and E. canadensis (96.7%) than dioecious hydrilla (56.6%), Developmental rate also varied between hosts, and was higher on E. canadensis (13.9 days) than either of the hydrilla typ es. Interestingly, development was fast on Vallisneria americana Michx. (15.2 days), a member of the Hydrocharitaceae family, althoug h it was a very poor host for survival

PAGE 63

63 (6.7%). Plants in families more distantly related to hydrilla, such as Potamogetonaceae, Ceratophyllaceae, and Cyperaceae, were generally poorer hosts than the more closely rela ted plants. Survival on Chara vulgaris L., a green algae only distantly related to vascular plants, was 43.3%. Utricularia macrorhiza Leconte, a carnivorous bladderwort, supported larval feeding, but was the only plant tested that did not allow complete de velopment to adulthood. Dioecious hydrilla, the target plant, supported moderate performance of C. lebetis in comparison to the other plants tested with an average survival rate of 56.6% and a development time of 19.73 days. There was a significant positiv e linear relationship between survival and development time (F = 12.92, df = 39, P < 0.001) (Fig. 3 5). Paired Choice T est The proportion of plant tips infested with larvae was higher for E. canadensis (76.4%) compared to hydrilla (26.8%) (F = 15.46, df = 5, P = 0.017) (Fig. 3 6). Moreover, damage to E. canadensis was higher than damage to hydrilla with an average score of 3.77 for E. canadensis compared to 1.88 for hydrilla (F = 13.73, df = 5, P = 0.021). When hydrilla was compared with N. guadalupensis t he percent of plants with larvae was significantly higher in N. guadalupensis (F = 6.59, df = 5, P = 0.043) but there was no difference in the damage score (F = 2.89, df = 5, P = 0.14) (Fig. 3 7). Hydrilla and E. densa did not differ in either the damage score (F = 0.02, df = 5, P = 0.88) or the percent of p lant tips infested with larvae (F = 0.50, df = 5, P = 0.52). Paired Choice Olfactometer Test For the majority of replications, larvae did not respond. In the test with hydrilla versus distilled water, one larva entered the olfactometer arm with distilled water. Larvae did not respond in all the other replications with hydrilla. In the test with E.

PAGE 64

64 canadensis versus distilled water, only one larva responded and chose E. canadensis (Fig. 3 8). Host Find ing Behavior Test Regressions of observed mean square distance traversed on the CRW model predicted values were significant for hydrilla, E. canadensis, and distilled water only (hydrilla: P = 0.004, E. canadensis: P = 0.0005, control : P = 0.007) and the s lopes of the regressions were not different from one for E. canadensis (F = 3.28, df = 18, P = 0.08) and the control (no plant material) ( F = 0.53, df = 18, P = 0.48 ) This suggests that in arenas with E. canadensis or only water, the movement pattern conf ormed to a theoretical model of correlated random walk, in which the direction of movement is correlated with the direction taken in the previous step (i.e., organisms tend to continue moving in t he same direction) (Fig. 3 9). However, with hydrilla, the s lope of the regression of observed values on model predicted values w as less than one ( F = 4.9, df = 18, P = 0.04), indicating that movement did not conform to the CRW model. The time s spent in the four quadrants of the Petri dish w ere not different for hydrilla, E. canadensis or the control, but there appeared to be a slight trend towards spending more time in quadrant 4, which had the plant piece, or nothing in the case of the control (Fig. 3 10). The trace patterns indicate that larvae may swim ver y close to the host plant, but not locate the host (Fig. 3 11). In one replication with a hydrilla tip, the larvae was able to locate the plant in 10.4 minutes, but in the remainder of replications, the larvae had not settled on the host plant after 60 min utes. Larvae did not locate the plant tip in the trials with E. canadensis.

PAGE 65

65 Paired Choice Adult Oviposition Females preferred to lay eggs on or near hydrilla when given a choice between hydrilla and artificial hydrilla (G adj = 13.01, df = 1 P < 0.001 ), an d hydrilla and distilled water (G adj = 16.76, df = 1 P < 0.001 ) (Fig. 3 11). There was no difference when given a choice between hydrilla and N. guadalupensis (G adj = 0.517, df = 1 P > 0.3 ). Females preferred to lay eggs on E. canadensis over hydrilla (G adj = 5.1, df = 1 P < 0.05 ). There was also preference to lay eggs on artificial hydrilla versus distilled water (G adj = 4.69, df = 1 P < 0.05 ) (Fig. 3 12). Discussion Host range testing is an essential component of biological control programs. Accurate information about the host range helps to quantify potential risks to native and economically important plant species (Briese, 2005). Hydrilla is a highly damaging wit h the use of herbicides. ( Langeland, 1996 ) The presence of fluridone resistance throughout the state has increased concerns about the availability of too ls to manage hydrilla ( Hoyer et al., 2005 ). Biological control may be an appropriate management strateg y if effective host specific agents can be identified. Determining the fundamental host range of C. lebetis is a first step toward assessing its potential as a biological control agent. The observed mining activity (Cuda et al., 2002) was an indication tha t C. lebetis may be host specific to hydrilla, as leaf mining insects tend to have narrow host ranges (Hespenheide, 1991). Therefore, host range testing was pursued to test this hypothesis. The feeding and oviposition preferences were analyzed and behavior al tests were conducted in order to understand the host range of this insect.

PAGE 66

66 The results of the no choice larval developmental study showed that C. lebetis has a broad host range and can feed on a variety of host plants belonging to several different fam ilies. Moreover, C. lebetis was able to complete development in a taxonomically diverse group of plants including a non vascular plant, C. vulgaris although survival tended to be highest on plants in the family Hydrocharitaceae, and in the closely related Najadaceae. However, in contrast to the hypothesized narrow host range, these findings suggest that C. lebetis is a generalist This information provides new insight into the biology of C. lebetis and its potential as an augment at ive biological control ag ent. Interestingly, two test plants E. canadensis and monoecious hydrilla, were more suitable for the development of C. lebetis than dioecious hydrilla. These plant species occur further north in the USA than Florida. Elodea canadensis ranges from Quebec south to North Carolina, Alabama and Arkansas, and westward to Manitoba, British Columbia, Colorado and California (Nichols and Shaw, 1986). It was reported in Jackson County in the panhandle of Florida in 1937; however, recent surveys s uggest that it may no longer occur there (Raymond Hix, personal communication). Monoecious hydrilla occurs in northeastern USA and the Atlantic coastal area as far south as South Carolina (Madeira, et al., 2000) and originated from Korea (Madeira et al., 1 997; Steward et al., 1984). There are concerns that monoecious hydrilla may cause problems as severe in the northern USA as dioecious hydrilla causes in the south (Langeland, 1996). If field tests reveal that C. lebetis prefers hydrilla in the field, it ma y still be worthwhile exploring the use of C. lebetis to control monoecious hydrilla

PAGE 67

67 populations, but use of the midge may be restricted to the southeast based on temperature studies (Chapter 1). The results of the paired choice tests confirm the findings of the no choice larval development test. In the no choice larval development test, E. canadensis was one of the best hosts, and that finding is supported by the paired choice test where the damage score and percent of in fested tips of E. canadensis were significantly higher than those of hydrilla. There were no differences between hydrilla and E. densa, and this is consistent with the no choice tests because the performance on tho se two plants was very similar. The olfact ometer and host finding behavior tests were conducted to investigate the foraging behavior of C. lebetis The low response rate, and random movement in locating a host plant, may be a reflection of the generalist feeding niche of C. lebetis The host findi ng behavior test confirmed the results of the olfactometer test, as the larvae were unable to locate the host in all but one trial. The arena of the host finding host; under normal co nditions in water bodies, the insect would be exposed to light only from above, whereas the Petri dish arena was lighted from below. Lighting from the underside, which find ing behavior. In addition, the paired choice larval host finding experiment was conducted over a much longer period than the Petri dish experiment (10 days compared to 60 minutes), which could have influenced results. Moreover, C. lebetis may be a nocturna l forager (J. P. Cuda, personal communication), and thus may not have

PAGE 68

68 displayed typical foraging behavior under lighted conditions, regardless of whether the light source was above or below the insect. Cricotopus lebetis seems to lack the ability to locate host plants using olfactory cues. Although there appeared to be a trend towards spending more time in the quadrant where the host plant was located, the same trend was seen when no plant material was in the Petri dish. This suggests that the insect may ha ve exhibited a tendency to move upstream against the water flow as quadrant 4 was whe re the water entered the dish. The pathway the insect followed during the 60 minute appeared to be random based on the fit of the fit of the data to the CRW model. The obs erved and predicted values were not different with E. canadensis or the control, indicating that larval movement did not appear to be directed. Interestingly, when hydrilla was in the arena, movement did no t conform to the model. Examination of the data su ggests that when hydrilla was present, displa cement was less than predicted. Accord ing to Brouwers and Newton (2009 ), when observed values fall below model predicted values (i.e., when the slope < 1), more random movement is suggested with less directional persistence. In some cases, the insects were observed swimming in very close proximity to the host plant, but they did not make contact. Therefore, it is hypothesized that C. lebetis larvae must make physical contact with a plant to be able to determine i f the plant is a suitable host. It has been shown that many generalist insects lack receptors to detect acute olfactory signals (Schoonhoven et al., 2005); whereas specialist herbivores often have receptors to detect host specific plant volatiles (Anderson et al., 1995; Hansson et

PAGE 69

69 al., 199 9; and Stranden et al., 2003). Cricotopus lebetis may not be able to detect plant derived chemicals that have diffused into the water. The oviposition test suggests that females are responsible for choosing an adequate sit e for larval development and survival. Females overwhelmingly choose to oviposit in the vicinity of plant material, whether it was a real or artificial. This finding suggests that females may respond to both visual and olfactory cues. Egg masses were laid on artificial plants when it was the only material available, but when given a choice between real and artificial plants, the real plants were chosen. Studies have shown that the visual cue of light polarization affects oviposition behavior of some chirono mid species ( Lerner et al., 2008). The initial step in selecting oviposition sites could be based on visual cues, which stimulate landing in the vicinity of plant material. Once females land, additional cues could indicate whether or not that plant is a su itable host. Moreover, the gelatinous egg masses of chironomid species are moved by water currents, and adhere to substrates (Williams, 1982). Therefore, the survival strategy of C. lebetis appears to be to oviposit egg masses in areas with plants that can support larval development. In this strategy, larvae do not search for a host, which is consistent with the findings of the larval host finding experiments. Field abundances may be low because of biotic (predation, competition) and abiotic (weather, pest i cides, water quality) factors. Laboratory bioassays to delineate the fundamental host range of C. lebetis suggest that this insect is a generalist feeder of aquatic plants. However, the ecological host range may differ from the fundamental host range (Haye et al., 2005). A next step in assessing the potential value of C. lebetis as an augmentative biological control in

PAGE 70

70 Florida (or a classical biological control agent for release in other states or countries) would be to evalua te the host range in the field. A potential shortfall of laboratory based host range tests is that they are unnatural in comparison to nature. If field choice tests reveal that C. lebetis prefers hydrilla over other aquatic plants, C. lebetis may have use as an augmentat ive biological control agent. With the exception of one specimen recovered from a Potamogeton s p. (Dana Denson, personal communication), all field records to date of C. lebetis that are associated with host plants have been from hydrilla. This could be due to a sampling bias, or reflect an actual field preference for hydrilla. If the latter is true, then releases of C. lebetis may result in minim al damage to non target plants. Moreover, C. lebetis does not kill hydrilla plants, but rather stunts their verti cal growth and results in profuse branching below the surface (Cuda et al., 2011). The effects of C. lebetis on other plant species has not been investigated, but are likely sub lethal. Impact studies to determine the densities at which this insect can cre ate a significant reduction in hydrilla biomass need to be conducted. Since C. lebetis is already found widely in Florida and does not appear to have much impact on hydrilla, insect densities may have to be significantly augmented to achieve acceptable lev els of control.

PAGE 71

71 Table 3 1 List of plants tested in no choice larval development Family Species Origin Common Name Hydrocharitaceae Elodea canadensis Michx. Native Canadian Waterweed Egeria densa Planch. Exotic Brazilian elodea Vallisneria americana Michx. Native American Eelgrass, tapegrass Hydrilla verticillata (monoecious) (L.f. Royle) Exotic Hydrilla Najadaceae Najas guadalupensis (Spreng.) Magnus Native Southern Naiad Potamogetonaceae Potamogeton illinoensis Morong Native Illinois Pondweed Ceratophyllaceae Ceratophyllum demersum L. Native Coontail Alismataceae Sagittaria kurziana Glck Native Strap leaf Sagittaria Cyperaceae Eleocharis baldwinii (Torr.) Chapm Native Road grass Lentibulariaceae Utricularia macrorhiza Leconte Native Bladderwort Characeae Chara vulgaris L. Native Muskgrass

PAGE 72

72 Figure 3 1. Aquatic olfactometer setup used in the dual choice test

PAGE 73

73 Figure 3 2. Example of movement analysis for c orr elated random w alk ( CRW ) model Insect was given 60 minutes to locate plant tip in quadrant 4. Arrows indicate each step insect made every 3 minutes. Water flow

PAGE 74

74 Figure 3 3. Survival of C. lebetis larvae on various aquatic plants under no choice conditions Survival occurred on 3 non native plant s. Survival was higher on monoecious hydrilla, E. canadensis, than on dioecious hydril l a 0 10 20 30 40 50 60 70 80 90 100 Hydrilla verticillata (monoecious) Elodea canadensis Najas guadalupensis Egeria densa Hydrilla verticillata Chara vulgaris Potamogeton illinoensis Ceratophyllum demersum Vallisneria americana Sagittaria kurziana Eleocharis baldwinii Utricularia macrorhiza % Survival d d d cd cd bcd b ab ab a a e

PAGE 75

75 Figure 3 4. Developmental rate of C. lebetis larvae on various aquatic plants under no choice conditions Development was faster on E. canadens is and V. americana than on dioecious hydrilla. 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 Hydrilla verticillata (monoecious) Elodea canadensis Najas guadalupensis Egeria densa Hydrilla verticillata Chara vulgaris Potamogeton illinoensis Ceratophyllum demersum Vallisneria americana Sagittaria kurziana Eleocharis baldwinii Utricularia macrorhiza Developmental rate (1/days) a bc bc cd cd e e de ab e e f

PAGE 76

76 Figure 3 5. Percent s urvival and development rate of C. lebetis larvae on host plants tested. As developmental rate increased survival on host plant increased. Developmental rate was slowest on C. vulgaris, P. illin oensis, and S. kurziana and highest on E. canadensis. 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0 10 20 30 40 50 60 70 80 90 100 Developmental rate (1/days) % Survival R(S) = 0.153 + 13.483 (S) R 2 = 0.248

PAGE 77

77 Figure 3 6. Percent of plant tips infested with C. lebetis larvae under paired choice conditions 0 10 20 30 40 50 Egeria densa Hydrilla verticillata a a 0 10 20 30 40 50 60 70 80 Najas guadalupensis Hydrilla verticillata % Plants with larva b a 0 20 40 60 80 100 Elodea canadensis Hydrilla verticillata b a

PAGE 78

78 Figure 3 7. Damage scores to plants by C. lebetis larvae under paired choice conditions 0 = no damage;1 = minimal damage not visible to naked eye; 2 = light damage 10 20%; 3 = moderate damage 20 50% ;4 = significant damage >50%; 5 = tip abscission. 0 1 2 3 4 5 Najas guadalupensis Hydrilla verticillata a a 0 1 2 3 4 5 Elodea canadensis Hydrilla verticillata Damage Score a b 0 1 2 3 4 5 Egeria densa Hydrilla verticillata a a

PAGE 79

79 Figure 3 8. Results from the paired choice olfactometer test 5% of larvae associated with E. canadensis and distilled H 2 O respectively, 90% no response, and 0% associated with hydrilla. No Response Distilled H 2 O E. canadensis

PAGE 80

80 Figure 3 9. Relationship between mean squared displacement of C. lebetis larvae and time in Petri dish areas with A) H. verticillata B) E. canadensis C) no plant material. CRW stands for correlated random walk. Significant P values obtained for all models show that observed values correlate to CRW, showing that movement patterns are random. A B C

PAGE 81

81 Figure 3 10. Time spent in each quadrant of Petri dish in response to pieces of A) H. verticillata B) E. canadensis or C) no plant material in quadrant four 0 5 10 15 20 25 30 35 40 One Two Three Four a b ab a 0 5 10 15 20 25 30 35 One Two Three Four Time (minutes) a a a a 0 5 10 15 20 25 30 One Two Three Four Quadrant b ab b a A B C

PAGE 82

82 Figure 3 11. Trace patterns of larval movement in a Petri dish from host finding behavior test with A ) H. verticillata B) E. canadensis. Neonate was given 60 minutes to locate host plant in quadrant 4. A B

PAGE 83

83 Figure 3 12. Total number of egg masses laid in each paired choice adult oviposition trial 0 3 6 9 12 15 18 Hydrilla Distilled water a b 0 3 6 9 12 15 Artificial Hydrilla Distilled water Egg masses a b 0 3 6 9 12 15 Hydrilla Artificial Hydrilla a b

PAGE 84

84 Figure 3 12. Total number of egg masses laid in each paired choice adult oviposition trial 0 2 4 6 8 10 12 14 16 Hydrilla verticillata Elodea canadensis b a 0 2 4 6 8 10 12 Hydrilla verticillata Najas guadalupensis a a

PAGE 85

85 CHAPTER 4 THE INFLUENCE OF WAT ER QUALITY ON CRICOTOPUS LEBETIS (DIPTERA: CHIRONOMIDAE): A HER BIVORE OF HYDRILLA VERTICILLAT A (HYDROCHARITACEAE) Introduction Hydrilla verticillata (L.f. Royle) is a submersed aquatic weed that exhibits fast growth rate and is capable of forming dense monocultures. Hydrilla causes negative impacts including displacement of native vegetation (Haller and Sutton, 1975), impediment of boat tr affic, recreational and commercial losses, clogging intake pipes and canals, and reductions in tourism and real estate values ( Schmitz et al., 1991 ; Langeland, 1996 ). Hydrilla is difficult to manage, because few sound options are available (Hoyer et al., 2005). Biological control has been investigated, but has had little success (Buckingham and Balciunas, 1994; Forno and Julien 2000 ; Gurr and Wratten, 2000; Wheeler a nd Center, 2001 ). Due to the recent development of resistance to the herbicide fluridone in some Florida hydrilla populations (Michel et al., 2004), biological control is once again being explored as a possible management tool. The family Chironomidae is c omprised of non biting midges. They have a cosmopolitan distribution, and occupy almost any habitat that i s aquatic or wet (Wirth, 1949). Water quality is an important variable that determines midge densities and distributions. Due to their sensitivity to water quality, chironomid communities have been used as indicators of water quality and pollution (Saether, 1979). Moreover, they are widely distributed in water bodies of varying quality (Paine and Gaufin, 1956). Although, some Cricotopus spp. have been s hown to be highly tolerant to pollutants (Boesel,

PAGE 86

86 1983). Given that certain midge species can be sensitive to water quality, water quality should be taken into consideration when developing biological control programs. In 1992, the chironomid midge Cricoto pus lebetis Sublette (Diptera: Chironomidae) was discovered in Crystal River Florida attacking the apical meristems of the invasive weed Hydrilla verticillata (L.f. Royle) (Cuda et al., 2002). Cricotopus lebetis may have potential as an augmentative biolog ical control agent of hydrilla. This insect damages the plant by mining into the apical meristem while preparing for pupation, which causes tip abscission (Cuda et al., 2002). Studies have shown that larval feeding of C. lebetis is capable of causing a sig nificant reduction in biomass of hydrilla (Cuda et al., 2011). Although hydrilla is present in many Florida water bodies, the distribution of the midge is largely unknown. During a recent survey in Florida waterways (Fall 2010), C. lebetis was only found i n Lake Rowell, Bradford, Co. (J. P. Cuda, unpublished data) urban, residential, and natural areas. Land use activities can influence water quality, and affect chironomid d istribution and abundance. Water quality variables such as temperature, pH, dissolved oxygen, hardness, alkalinity, and presence of pesticides may be key factors that determine the distribution of this midge. Pyrethroid pesticides are used extensively in a griculture and chironomids have been shown to be sensitive to their presence in water bodies (Anderson, 1989). Fipronil, p henylpyraloze class insecticide, is used extensively to control terrestrial insect populations, and due to runoff, has been documente d in Florida water bodies (Harmon Fetcha et al., 2005). This chemical is highly toxic to fish and aquatic invertebrates ( Schlenk et al., 2001; Key et al., 2003; Stehr et al., 2006), and could influence

PAGE 87

87 distributions of C. lebetis and other chironomids. The presence of fipronil, even in extremely low concentrations, could be lethal to chironomid populations. Therefore, st udies were conducted to monitor pyrethroid and fipronil concentrations in select locations in Florida to examine possible correlations with the distribution of C. lebetis The goals of this research were to assess water quality parameters several water bodies containing hydrilla, and correlate water quality parameters to the presence or absence of C. lebetis and other chironomid species. Mate rials and Methods Chironomid Diversity Six Florida water bodies were sampled quarterly from January 2011 to May 2012 to determine chironomid diversity in hydrilla tips. Locations were in the northern (Lake Rowell, Wacissa Springs), central (Lake Tohopekali ga (Toho), Bulldozer Canal), and southern (Lake Istokpoga, Lake Okeechobee) portions of the state (Fig. 4 1). A four pronged steel hook attached to a rope was thrown into the water and dragged along the hydrosoil to collect hydrilla. Hydrilla was placed in large re closable bags partially filled with water from the corresponding sample site to prevent desiccation of hydrilla and associated midges. All bags with hydrilla were placed in a cooler for transport to the laboratory. Several liters of water were co llected from each sampling site so that the hydrilla could be maintained in the same water quality f or midge rearing. In total, 300 hydrilla tips approximately 5 8 cm in length were randomly selected from each sample and placed in containers with water collected from the corresponding water body. The containers were placed in emergence cages and aerated. Emergence cages were monitored daily for midge emergence for 14 days after the sample dat e. Emerged adults were placed in vials containing 95% ethanol and separated by species. Individuals that

PAGE 88

88 could not be identified to species were sent to John H. Epler for authoritativ e identification Species richness of each sample was determined as the n umber of reared midge species. Diversity was calculated using the Shannon W i e ner Index ( Shannon and Weaver, 1949 ) Water Quality Water quality variables, including temperature, dissolved oxygen, conductivity, and pH, were taken in the field from each locat ion during every sampling occasion using an YSI 556 data logger ( YSI, Inc., Yellow Springs, OH ). Water samples were collected from the field, placed in 250 ml ste r ile plastic Nalgene containers, and then placed on ice. Alkalinity and water hardness were de termined in the laboratory following EPA method 310.1 ( alkalinity ) and EPA method 130.2 ( hardness ) (EPA, 1971; EPA 1978). Pesticide Analysis Presence of pyrethroid pesticides was determined by deploying semi p ermeable membrane devices (SPMD Environmental Sampling Technologies Inc. St. Joseph MO ) at two locations (Lake Okeechobee, Lake Rowell) in the fall of 2011. These two locations were selected based on two factors; 1) C. lebetis was abundant in Lake Rowell in August 2010, whereas it was not found in Lake Okeechobee on any sampling date and 2) Lake Rowell is primarily surrounded by natural areas whereas Lake Okeechobee is surrounded mostly by agricultural and urban areas. Three SPMDs spiked with Permeability Reference Compound (PRC ) were d eployed at both locations. The canisters containing the SPMDs were not opened until they were ready for deployment. During deployment, the boat motor was shut off when in close proximity to the deployment site to avoid contamination with hydrocarbons. Thre e SPMDs were loaded into a single canister provided by the Environ mental Sampling Technologies Inc

PAGE 89

89 (EST), then deployed into the water. A cinder block was used as an anchor, and the SPMD was suspended in the water column using a 1 gallon plastic milk jug filled with Styrofoam. The SMPDs were submersed in the water column at a depth of approximate ly 1 m for 1 month. SPMDs were then collected from the field, placed back into the shipping canisters, held on ice, and shipped overnight to EST for extraction and dialysis. Extracts were analyzed for pyrethroid pesticides using a gas chromatograph equipped with electron capture detectors (Wu et al., 2010). The presence of fipronil in Lake Okeecho bee was determined in May 2012. Four one liter water samples were coll ected in glass bottles, filtered, and analyzed for the presence of fipronil using gas chromatography (Wu et al., 2010). Fipronil Dose Response A preliminary range finding study was conducted to obtain initial information on the sensitivity of C. lebetis to fipronil. Insects were exposed to six concentrations of fipronil mixed with well water (0.0, 0.02, 0.2, 2.0, 20.0, 200.0 and 2000.0 g/L ) in test tubes (35 ml) with 20 ml of the mixture. A single, undamaged, healthy hydrilla tip 3 5 cm in length was placed in each tube, along with one 8 day old larva. Test tubes were placed in a rack in an environmental growth chamber at 25C and 14:10 (L:D) photoperiod (Model No. E36L, Percival Scientific, Inc., Perry, IA) Larval survival was assessed every 24 hours for 9 6 hours. A second definitive was conducted using a narrower range of fipronil concentrations (0.0, 0.5, 2.0, 5.0, 10.0, 15.0, and 20.0 g/L ). Concentrations during the definitive test were confirmed by GC ECD using the method of Wu et al. (2010). Effect of Alkalinity and Hardness on C. lebetis Larval Development Nanopure water was used to make three different solutions of water hardness and alkalinity; soft (alkalinity: 12, hardness: 12), moderately hard (alkalinity: 65,

PAGE 90

90 hardness: 90) and very hard (alkali nity: 235, hardness: 300). Each water solution was placed in a 38 liter aquarium and aerated. Hydrilla was propagated from stems collected in Lake Toho and placed into 10.16 cm pots containing a layer of potting soil covered by sand. Hydrilla was grown in each aquarium for a minimum of 3 weeks prior to insect exposure. Tips 4 6 cm in length from each aquarium were harvested and placed individually in 35 ml test tubes filled with water from the corresponding aquarium. Control treatments used hydrilla grown i n well water and test tubes filled with well water. Two neonates of C. lebetis larvae were transferred to each test tube using a pipette and then each tube was stoppered with a cap with ventilation holes. All racks were placed in an environmental growth c hamber maintained at 25 C and 14:10 (L:D) photoperiod (Model No. E36L, Percival Scientific, Inc., Perry, IA) Tips were checked daily to ensure that they were fully submerged in order to prevent larval desiccation. Tubes with tips that were completely des troyed received replacement tips to allow midges to complete development to adulthood Starting approximately one week after the larvae were introduced, the test tubes were checked daily for adult emergence. Tips that supported development to the pupal sta ge were considered as suitable for complete development. The number of days to develop to pupation was recorded. Data Analysis Survival and developmental time under different water quality conditions were analyzed with ANOVA using the general linear model procedure, and means were separated with Student Newman Keuls procedure (PROC GLM; SAS Institute 2008). The Shannon W i e ner Index was calculated for each sampling occasion and for total diversity at each sampling location. The equation was as follows:

PAGE 91

91 Where H = the Shannon W i e ner diversity index; p i = abundance of species i The frequency of occurrence (incidence) and abundance were calculated for each chironomid to obtain an index of prevalence ( Ip) (Zhou et al. 2003) Incidence was determined by dividing the number of locations where the chironomid was found on each sampling date by the total number of locations sampled. Abundance was determined for each sampling occasion by summing the total number of individuals of ea ch species found and dividing by the number of plant tips sampled on that sampling occasion. The Ip was calculated for each chironomid species as the product of chironomid abundance and incidence as follows: Where Fo = the number of loca tions a species was found, N 1 = the number of locations sampled, S = the sum of the abundance values for a species and N 2 = number of tips collected (Zhou et al. 2003). Confidence limits (95%) were calculated for each Ip value following methods provided b y Buonaccorsi and Leibhold (1988), and means compared by examining overlap in confidence limits (84%) as described by Payton et al. (2003). Data from the fipronil studies were analyzed using Proc Probit in SAS (SAS Institute, 2008) to generate a logistic r egression of the data. Results Chironomid Diversity Species richness ranged from 8 15 taxa per location and a total of 18 species were found associ ated with hydrilla (Table 4 1). Bulldozer Canal had the highest

PAGE 92

92 species richness and Wacissa Springs had the lowest. Species diver sity, measured by the Shannon W i e ner index ranged from 1.29 2.27 per location (Table 4 2), with Bulldozer Canal having the highest diversity and Lake Okeechobee the lowest. Cricotopus lebetis was found at four of the six sampling loc ations (Lake Toho, Lake Rowell, Lake Istopkoga, Bulldozer Canal), but at low densities except for Lake Rowell in the fall of 2010. Total numbers of C. lebetis collected at each site were Lake Rowell: 27, Lake Toho: 4, Lake Istopkoga: 12, Bulldozer Canal: 3 (Table 4 3). The prevalence index was highest for Tanytarsini spp (0.0076) and second highest for Cricotopus sylvestris (0.0042) a congener of C. lebetis, that does not feed on living plant tissues The rarest species based on the index of prevalence we re Nilobezzia schwarzii (0.000003) and Cricotopus politus (0.000003) (Figure 4 1). The prevalence of Cricotopus lebetis was intermediate with an index of 0.0023. Water Quality Water quality measurements at field sites indicated that there were little differences in water quality among water bodies sampled (Table 4 4). The dissolved oxygen content in Bulldozer Canal was very low for the majority of the year (<1 mg/L), except in the winter when an unsea sonal storm event occurred. Other sites had dissolve d oxygen concentrations ranging from 2.60 7.57 mg/L. The temperatures at each location for each sample date were within the thermal limits of C. lebetis. For all sampling locations the temperature ranged from 14.00 28.31C. The pH for all sampling locati on and dates ranged from 6.34 8.63. Conductivity ra nged from 0.196 40.9 (mS/cm). Bulldozer canal and Wacissa Springs were the only sample sites with water hardness and alkalinity levels above 100 mg/L CaCO 3 Lake Istopkoga had alkalinity

PAGE 93

93 levels ranging from 16 38 mg/L CaCO 3 34 mg/L CaCO 3 and hardness ranged from 33.0 42.5 mg/L CaCO 3 Lake Okeechobee had moderate alkalinity and hardness ranging from 40 98 and 40 91 mg/L CaCO 3 respectively. Lake Toho and Istopkoga had lower hardness ranging from 27 37 and 22 42.5 mg/L CaCO 3 respectively. Pesticide Analysis No evidence of pyrethroid pesticides was found in Lake Okeechobee or Lake Rowell from deploying SPMD units. One water sample collected in Lake Okeechobee was c onfirmed to have 15 ng/L of fipronil. Fipronil Dose Response The results from the range finding test were used to determine the range of concentration s to use in the definitive test. For the definitive study, 100% mortality occu rred at 20.0, 15.0, and 10.0 g/L of fipronil. No mortalit y occurred in the control, and 9 0% 60%, and 4 0% mortality occurred at 5.0, 2.0, and 0.5 g/L respectively. The estimate of the LC 50 (lethal concentration) using this data was 0.91 g/L and the LC 90 was 4.52 g/L (Fig. 4 2). Effect of Alkalinity and Hardness on C. lebetis Larval Development There were no significant differences in devel opment time (F = 3.18, df = 3, P = .0848) or survival (F = 0.89, df = 3, P = 0.49) of larvae reared in the different alkalinity/h ardness water quality treatments (Fig. 4 3). Discussion Water quality is an important attribute of aquatic habitats that may influence the distribution and abundance of chironomid communities (Saether et al., 1979). In Florida, land surrounding water bodi es is used in a variety of different ways that may impact

PAGE 94

94 water quality. Water bodies adjacent to agricultural land may receive runoff containing pesticides and other chemicals. Chironomids are sensitive to many chemicals and may be negatively impacted by runoff (Gresens et al., 2007) but some chironmids are highly tolerant to pollutants (Boesel, 1984) SPMDs are capable of detecting pyrethroid pesticides over time, but may not detect other pesticides. Therefore, pesticide classes that were not found in th e SPMD samples may have been present and influenced the distribution and abundance of C. lebetis and other chironomids. The water quality data collected from the field, coupled with the laboratory study on larval development in water with different levels of alkalinity/hardness, suggest that these variables may not have a major impact on survival and distribution of C. lebe tis. Cricotopus lebetis was found at Bulldozer Canal, which had high alkalinity and hardness and low dissolved oxygen, and also was found at Lake Istopkoga, which had low alkalinity and hardness, and average dissolved oxygen. Laboratory testing, which reve aled that larval survival and development did not vary in response to differences in alkalinity and hardness, is consistent with the field data. Conductivity at Bulldozer Canal was higher than at Lake Istopkoga, and C. lebetis was recovered from both these locations. Lake Okeechobee, which is surrounded largely by agriculture (Flaig and Havens, 1995), had a Shannon W i e ner Index of 1.29, which was low compared to the other sampled water bodies. Fipronil was detected in Lake Okeechobee, but at a lower concen tration than what was shown to be lethal in the laboratory. Additional sampling would likely reveal temporal fluctuations, and if fipronil reaches lethal concentrations in the field. Low concentrations of fipronil have been reported to be toxic to

PAGE 95

95 macroinv ertebrates and grass shrimp were reported to have a LC 50 of 0.32 /L (Key et al., 2003), and thus, may affect the distribution of chironomids. The presence of fipronil could also affect colonization of C. lebetis in Lake Okeechobee if augmentative releases were to be made. Reducing agricultural runoff and implementing best management practices could improve water quality in Lake Okeechobee and the macroinvertebrate communities within. The other location that C. lebetis was not recovered was Wacissa Springs, perhaps because of cooler water temperatures. Wacissa Springs has a constant temperature throughout the year of 21C (Raymond Hix, personal communication). This is within the thermal limits of C. lebetis, but is near the lower end of the optimal t emperature rang e for development (Chapter 2). Low water temperature would slow the negatively affect po pulation growth and abundance. Moreover, Wacissa Springs was less frequent ly sampled than other water bodies, and therefore the probability of recovering C. lebetis even if was as abundant as in the water bodies where it was found, would be lower. The sporadic recovery of C. lebetis over different sampling periods suggests tha t this insect does not exhibit seasonality. It was found during all four sampling periods (winter, spring, summer, fall), albeit not within the same water body. In Lake Toho, its seasonal incidence was highest as it was found in the winter, spring, and fal l. The seasonal abundance of the midge likely responds to changes in environmental conditions, such as temperature and food availability, which vary throughout the year. The abundance of Cricotopus lebetis was highest in Lake Rowell in the fall of 2010, bu t

PAGE 96

96 surprisingly, it was not recovered from Lake Rowell on later sampling dates. The reasons for the high abundance of C. lebetis in the fall of 2010 in Lake Rowell are unknown, but could be a combination of biotic, abiotic and anthropogenic factors. Overall water quality does not seem to be as important in determining the distribution of C. lebetis as was hypothesized. The midge was found in four widely distributed water bodies in Florida, with different water quality characteristics. Thus, the midge does n ot appear to be highly sensitive to variation in the water quality parameters that were measured. Pesticides, however, may play a role, based t he laboratory study on the sensitivity of the midge to fipronil Additional sampling may reveal a correlation bet ween the presence of pesticides and chironomid diversity.

PAGE 97

97 Table 4 1.Chironomid species richness at each location from each sampling period. Means and standard deviations were calculated when there was more than one sample/location/sampling period Plan t material from Lake Rowell and Wacissa Springs was sampled separately and shipped to the facility in Fort Pierce, and these locations were not sampled during each sampling event. Winter Spring Summer Fall Total Lake Toho 4 2.83 5.5 3.54 2 6 15 Lake Istokpoga 3 1.41 3.5 0.71 5 5 12 Lake Rowell -2 4 6 9 Lake Okeechobee 3 2.83 2.5 2.12 -3 8 Bulldozer Canal 4 7 2 2 13 Wacissa Springs -7 7 -8 Table 4 2 Chironomid diversity at each location for each sampling period. Means and standard deviations were calculated when there was more than one sample/water body/sampling period Plant material from Lake Rowell and Wacissa Springs was sampled separately and s hipped to the facility in Fort Pierce, and these locations were not sampled during each sampling event. Winter Spring Summer Fall Total Lake Toho 1.55 1.37 0.52 1.89 1.54 2.27 Lake Istokpoga 1.29 0.81 0.37 1.23 1.14 1.89 Lake Rowell -0.67 0.64 1.43 1.42 Lake Okeechobee 1.28 1.52 -1.05 1.29 Bulldozer Canal 1.27 1.65 0.15 0.64 0.55 2.11 Wacissa Springs -1.79 1.48 -1.88

PAGE 98

98 Table 4 3. Chironomid species sampled from all locations during survey work Water body Species Lake Toho Lake Istokpoga Lake Okeechobee Bulldozer Canal Lake Rowell Wacissa Springs Ablabesmyia rhamphe 0 1 0 1 1 0 Apedilum elachistus 3 1 2 36 0 0 Bezzia glabra 2 0 0 0 2 0 Chironomini sp p 3 27 2 1 8 8 Cricotopus bicintus 1 0 3 5 0 0 Cricotopus lebetis 9 12 0 3 27 0 Cricotopus politus 0 0 0 0 0 1 Cricotopus sylvestris 2 3 53 25 30 5 Dicrotendipes sp p 26 11 7 20 3 12 Glyptotendipes sp p 11 0 0 6 0 0 Larsia decolorata 7 0 2 3 2 0 Nanacladius alternantherae 1 2 0 0 0 0 Nilobezzia schwarzii 0 1 0 0 0 0 Parachironomous hazelriggi 32 17 15 9 0 8 Pentaneura inconspicua 5 0 0 0 0 5 Pseudochironomous richardsoni 5 1 0 5 0 0 Tanytarsus buckleyi 10 3 0 7 9 17 Tanytarsini sp p 17 2 2 6 94 17

PAGE 99

99 Table 4 4. Water quality variables at field si tes from January 2011 May 2012. Means and standard deviations were calculated when there were multiple samples per location and season Plant material from Lake Rowell and Wacissa Springs was sampled separately and shipped to the facility in Fort Pierce, and these locations were not sampled during each sampling event Water Body Season Temperature (C) pH Dissolved Oxygen (mg/L) Conductivity (mS/cm) Alkalinity (mg/L) CaCO 3 Hardness (mg/L) CaCO 3 Lake Okeechobee Winter 18.25 2.51 6.75 0.46 5.62 2.18 2.25 90 45.26 61 15.56 Spring 24.92 0.84 8.18 0.04 5.55 1.14 3.36 4.46 86 39.6 74 24 Summer 23.68 6.92 2.6 7.08 98 91 Fall 20.45 6.52 4.42 2.73 40 40 Lake Istopkoga Winter 18.57 1.03 7.23 0.42 6.41 1.03 1.26 26 8.49 28.5 4.95 Spring 26.34 1.03 8.63 6.69 0.71 1.35 1.63 24 2.83 42.5 2.12 Summer 27.27 8.88 6 3.07 38 37 Fall 21.84 6.74 6.74 1.32 16 22 Lake Rowell Winter 15.6 7.23 6.44 2.84 32 39.5 0.71 Spring 23.8 ---31 4.24 42.5 2.12 Summer 29.8 ---34 33 Fall -----39 Lake Toho Winter 16.9 1.18 6.24 1.03 4.98 1.28 1.52 41 1.41 20.5 13.44 Spring 26.13 2.34 7.48 0.08 6.74 0.71 1.46 1.81 42 11.31 35 4.24 Summer 28.31 7.22 5.98 2.77 52 37 Fall 22.7 7.33 5.31 1.8 34 27 Bulldozer Canal Winter 14 7.12 4.04 40.9 156 144 Spring 25.33 1.07 7.13 0.06 1.43 0.78 6.62 8.12 155 15.56 152 4.24 Summer 27.51 7.01 0.93 11.79 124 125 Fall 20.55 6.34 0.55 1.72 42 27 Wacissa Springs Winter 21 ---65 128 Spring 21 ---81 154 Summer 21 ---111 208 Fall 21 -----

PAGE 100

100 Figure 4 1. Sampling locations for survey data in Florida where hydrilla for chironomid diversity study and water quality variables were collected. Northern locations in red (Wacissa Springs, Lake Rowell), central locations in yellow (Lake Toho p ekaliga, Bulldozer Canal), southern locations in white (Lake Istokpoga, Lake Okeechobee) Source: Google Earth.

PAGE 101

101 Figure 4 2 Index of prevalence of chironomids recovered from all sampling locations from January 2011 May 201 2 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 Index of prevalence a b c d d e f f g g h h hi i ij j k k

PAGE 102

102 Figure 4 3 Fipronil laboratory bioassay with C. lebetis larvae A) range finding test B) dose response test Showing lethal concentration (LC ) at which 50 and 90% population mortality are predicted to occur. LC 50 LC 90 0 10 20 30 40 50 60 70 80 90 100 0.02 0.2 2 20 200 2000 % Mortality Concentration (ug/L) LC 90 LC 50 0 10 20 30 40 50 60 70 80 90 100 0.02 0.2 2 20 % Mortality Concentration (ug/L) A B

PAGE 103

103 Figure 4 4 Response of C. lebetis to different levels of water alkalinity/har dness A) developmental rate B) p ercent s urvival 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 Control Very Soft Moderately Hard Very Hard Developmental rate (1/days) Water Quality a a a a 0 20 40 60 80 100 Control Very Soft Moderately Hard Very Hard % Survival Water Quality a a a a A B

PAGE 104

104 CHAPTER 5 SUMMARY OF FINDI NGS ON THERMAL REQUI REMENTS, HOST RANGE AND DISTRIBUTION OF CRICOTOPUS LEBETIS (DIPTERA: CHIRONOMID AE), A NATURAL ENEMY OF HYDRILLA VERTICILLAT A (HYDROCHARITACEAE) Introduction Hydrilla verticillata (L.f. Royle) is one of the most devastating aquatic invasive plants introduced into the USA. Since its introduction, costly programs have been enacted to control hydrilla, but management remains difficult (Hoyer et al., 2005). The rec ent advent of fluridone resistance has raised awareness of the aggressiveness of this plant if left unmanaged (Michel et al., 2004). The hydrilla miner, Cricotopus lebetis Sublette (Diptera: Chironomidae) was discovered in 1992 in Crystal River Florida at tacking the apical meristems of hydrilla (Cuda et al., 2002). The discovery of the insect stimulated interest in assessing its potential as a biological control agent. The thermal requirements, fundamental host range, and host finding behavior were studied Hydrilla surveys in water bodies throughout Florida were conducted to analyze water quality variables, and to search for the presence of pesticides that could impact chironomid distributions and densities. The results obtained from these studies will be used to evaluate the potential of the hydrilla miner for augmentative biological control of hydrilla in Florida, and for release as a classical biological control agent in areas where it does not occur. Recommendations Use of C. lebetis in Florida Cricoto pus lebetis may have value as an augmentative biological control agent of hydrilla. Since the midge occurs in Florida, no permits would be required to release C. lebetis within the state. Based on the results of field surveys, temperature dependent

PAGE 105

105 develop ment studies, climatic mapping, host range testing and water quality studies, the midge ca n establish throughout Florida. However, there is some evidence that pesticide pollutants in the water may influence its occurrence, but further study and sampling is nee ded to confirm this statement. Following are some factors that should be considered for augmentative use of C. lebetis : W ater t emperature The thermal limits of C. lebetis are broad, as development occurred at temperatures ranging from 15 35C. The idea l temperature range was 20 30C, and this is within the normal limits for temperatures in Florida water bodies. Based on climatic suitability mapping and isothermal lines, C. lebetis is predicted to establish throughout Florida. Colder temperatures in the northern part of the state may decrease the number of generations per year. The cold tolerance studies show that C. lebetis does not tolerate cold temperatures for extended periods of time. Although it is unknown if this insect exhibits diapause, cold tolerance is likely to increase if it does, as has been shown for other insects (Pullin, 1996). Temperatures in vegetation mats during the summer months can be significantly higher than temperatures within the water column (Wheeler and Center, 2001; Mike N etherland, personal communication), and exceed the upper developmental threshold of C. lebetis It is unknown which portion of the water column C. lebetis prefers, but if this insect colonizes vegetation below the surface mats, it should be able to escape temperature extremes. Water q uality Many chironomid species are sensitive to water quality and pesticide contamination (Madden et al., 1992). In the present study, water quality parameters such as pH, conductivity, dissolved oxygen, alkalinity, and hardness did not appear to

PAGE 106

106 influence the di stribution of C. lebetis and for hardness and alkalinity, this was supported by the laboratory study; Lake Okeechobee exhibited low species richness, and C. lebetis was never recovered at this location. Although C. lebetis was never recovered from this la ke, this does not mean that is it is not present. Analysis of water samples collected from Lake Okeechobee confirmed the presence of fipronil. The results from the fipronil dose response test indicated that C. lebetis was sensitive to this pesticide, even in low concentrations. Fipronil contamination in Florida water bodies could play a role in the distribution and potential establishment of C. lebetis and other chironomid species. Other water quality variables that were not assessed in this study could als o be influential in shaping chironomid communities. Further studies and sampling are needed to fully understand the distributional patterns of C. lebetis throughout Florida. Potential effects on native f lora A surprising finding of this research was the d iscovery that C. lebetis is polyphagous. The host range study tested a subset of the aquatic plants that are found in Florida water bodies, and C. lebetis is probably able to feed and develop on additional plants that were not tested. An ideal biological c ontrol agent would attack only the target species, and this is not the case with C. lebetis. Monoecious hydrilla and E lodea canadensis were shown to be more suitable hosts than dioecious hydrilla. Most of the plants tested are native to Florida, and C. leb etis performed well on the native naiad, N ajas guadalupensis. This naiad, along with other native aquatic vegetation, is commonly displaced by hydrilla (Langeland, 1996 ), and releasing C. lebetis in locations where N. guadalupensis is present may cause the plant additional stress. Cricotopus lebetis also attacked other important native aquatic flora, and although survival was not

PAGE 107

107 high on these plants, C. lebetis could at tack those plants in the field. However, the feeding habits of C. lebetis show that it d oes not kill hydrilla, but rather it changes plant architecture (Cuda et al., 2011), and the damage could be similar on other host plants. Thus, even if C. lebetis were to attack native plants, the damage inflicted may be insignificant. Therefore, proper r isk assessment is necessary before making field releases. Given the results of the laboratory host range testing, field host range studies should be conducted to determine the difference between the ecological host range and the fundamental host range. Stu dies with Parapoynx diminutalis Snellen showed that this insect had a broad host range in the laboratory (Buckingham and Bennett, 1989) but in the field it attacked fewer plants ( Buckingham, 1994 ). If C. lebetis exhibits a preference for hydrilla in the f ield, or has limited dispersal capabilities, it may be possible to limit non target effects. However, based only on laboratory studies, it appears that the wide host range of C. lebetis may limit its value as a biological control agent. Integ rated pest m an agement (IPM) Cricotopus lebetis may not have significant impact on hydrilla populations when used alone, but if used in combination with the fungus Mycoleptodiscus terrestris (Mt) or the herbicide imazamox, there could be synergistic effects on hyd rilla m anagement (Cuda and Gill ett Kaufman 2011) Using all three control tactics could provide a sustainable control method for hydrilla because control does not rely on one management option. Multiple management options can reduce the dependence on herbicides, and reduce the risk of herbicide resistance When in high concentrations, Mt is lethal to hydrilla (Shearer and Nelson, 2009), but if applied at lower doses, Mt weakens the plant. Currently Mt can only be applied for experimental use but if proven

PAGE 108

108 to be effective for management of hydrilla standa rds could be developed Low doses of Mt could be used in tandem with C. lebetis to induce multiple stresses to hydrilla, and decrease the probability of the development of resistance to Mt. Establishing an IPM program integrating C. lebetis, Mt, and imazam ox could be beneficial and cost effective for various reasons: 1) mass releasing insects can require costly rearing operations and continual releases, which could be reduced through integration; 2) herbicides are costly and time consuming to apply and ther e is a risk of hydrilla developing resistance to additional herbicides. Integration with Mt and or C. lebetis could lower the amount of herbicides applied, and thereby reduce the costs and the risk of herbicide resistance. Use of C. lebetis O utside of Flo rida Introducing an insect into areas where it is not known to occur is inherently more risky, and more complicated from a regulatory point of view, than releasing it in area where it already occurs. The broad host range of C. lebetis makes it very unlikel y that it would be approved for release in USA states, or other areas of the world, where it is not known to occur. Furthermore, environmental conditions at northern USA locations may be unsuitable for permanent establishment of C. lebetis as predicted by cold tolerance studies and species distribution modeling If the insect is successful at reducing populations of hydrilla in Florida without evidence of non target effects, then consideration could be given to releasing it in locations where it has not be en found. Concluding Remarks This study provides valuable insight on the thermal requirements, host range and distribution of a recently discovered insect herbivore of hydrilla. Based on previous knowledge of the insect, the expectation was that C. lebetis would be host specific to hydrilla. However, laboratory based larval host finding experiments, coupled with

PAGE 109

109 observations of ovipositional behavior, suggest that the midge is an opportunistic forager, and feeds on most aquatic plants that it may enc ounter. Field studies should be conducted to confirm the broad the host range of C. lebetis Hydrilla continues to be one of the most troublesome weeds plaguing Florida waterways. Therefore, the costs and benefits of releasing C. lebetis should be determin ed in order to assess its use as an augmentative biological control agent. Exploration for new natural enemies of hydrilla in its native range should continue.

PAGE 110

110 LIST OF REFERENCES Anderson, P., B.S. Hansson, and J. Lofquist. 1995. Plant volatile receptor ne urons on the antennae of female and male Spodoptera littoralis. Physiol Entomol. 20: 189 198. Anderson, R. L. 1989. Toxicity of synthetic pyrethroids to freshwater invertebrates. Environ Toxicol and Chem. 8: 403 410. Arias, R. A., M.D. Netherland, B.E. S cheffler, A. Puri, and F. E. Dayan. 2005. Molecular evolution of herbicide resistance to phytoene desaturase inhibitors in Hydrilla verticillata and its potential use to generate herbicide resistant crops. Pest Manag Sci. 61: 258 268. Balciunas, J.K. and M.C. Minno, 1985. Insects damaging hydrilla in the USA. J Aquat Plant Manage. 23: 77 83. Balciunas, J.K. and M.F. Purcell 1991. Distribution and biology of a new Bagous weevil (Coleoptera: Curculionidae) which feeds on the aquatic weed Hydrilla verticillata. J Aust Entomol Soc. 30: 333 338. Baskerville, G.L, and P. Emin. 1969. Rapid estimation of heat accumulation from maximum and minimum temperatures. Ecology. 50: 514 517. Bayoh, M.N., and S.W. Lindsay. 2004. Temperature related duration of aquatic states of the Afrotropical malaria vector mosquito Anopheles gambiae in the laboratory. Med Vet Entomol. 18: 174 179. Beaver, J.R., T.L. Crisman, and J.S. Bays. 1981. Thermal regimes of Florida Lakes. Hydrobiologia 83: 267 273. Boesel, M.W., 1983. A review of the genus Cricotopus in Ohio, with a key to adult of species of the northeastern United States (Diptera: Chironomidae). Ohio J Sci. 83: 74 90. Boger, P., and G. Sandman. 1998. Carotenoid biosynthesis inhibit or herbicides mode of action and resistance mechanisms. Pesticide Outlook. 9: 29 35. Borchert, D., and R. Magarey. 2007. User manual for NAPPFAST. ( http://www.nappfast.org/usermanual /nappfast manual.pdf ) Bouchard, R.W. Jr., M.A. Carrillo, S.A. Kells, and L.C Ferrington Jr. 2006. Freeze tolerance of larvae of the winter active Diamesa mendotae (Diptera: Chironomidae): a contrast of adult strategy for survival at low temperatures. Hy drobiologia. 568: 403 416.

PAGE 111

111 Bowes, G, A.S. Holaday, and W.T. Haller. 1979. Seasonal variation in the biomass, tuber density, and photosynthetic metabolism of hydrilla in three Florida lakes. J Aquat Plant Manage. 17: 61 65. Brire, J. F., P Pracos, A.Y. L e Roux, and J.S. Pierre. 1999. A novel model of temperature dependent development for arthropods. Environ Entomol. 28: 22 29. Briese, D.T. 2005. Translating host specificity test results into the real world: the need to harmonize the yin and yang of curre nt testing procedures. Biol Control. 35: 208 214. Brown, S.J., and M.J. Maceina. 2002. The influence of desparate levels of submersed aquatic vegetation on largemouth bass population characteristics in a Georgia reservoir. J Aquat Plant Manage. 40: 28 35 Brouwers, N.C, and A.C. Newton. 2009. Movement analyses of wood cricket ( Nemobius sylvestris) (Orthoptera: Gryllidae). B Entomol Res. 100: 623 634. Buckingham, G.R. 1988. Reunion in Florida hydrilla, a weevil, and a fly. Aquatics. 10: 19 25. Buckingham, G.R., 1994. Biological control of aquatic weeds, pp. 413 480. In D. Rosen, F.D. Bennett, and J.L. Capinera (eds.). Pest management in the subtropics: biological control a Florida perspective, Intercept, Andover, UK. Buckingham, G.R. and J.K. Balciunas. 1994. Biological studies of Bagous hydrillae. Aquatic plant control research program. Technical Report A 94 6. Buckingham, G.R., and C. A. Bennett. 1996. Laboratory biology of an immigrant Asian moth, Parapoynx diminut alis (Lepidoptera: Pyral idae), on Hydrilla verticillata (Hydrocharitaceae). Fla Entomol. 79: 353 363. Buckingham, G.R., and C. Bennett. 2001. Life History and laboratory host range tests of Parapoynx seminealis (Walker) (Crambidae: Nymphulinae) in Florida U.S.A. J Lepid Soc. 55: 111 118. Buckingham, G.R., E.A. Okrah, and M.C. Thomas. 1989. Laboratory host range tests with Hydrellia pakistanae (Diptera: Ephydridae), an agent for biological control of Hydrilla verticillata (Hydrocharitaceae). Environ Entom ol. 18: 164 171. Campbell, A., B.D. Frazier, N. Gilbert, A.P. Gutierrez, and M. Mackauer. 1974. Temperature requirements of some aphids and their parasites. J Appl Ecol. 11: 431 438.

PAGE 112

112 Center, T.D., M.J. Grodowitz, A.F. Cofrancesco, G. Jubinsky, E. Snodd y, and J.E. Freedman. 1997. Establishment of Hydrellia pakistanae (Diptera: Ephydridae) for the biological control of the submersed aquatic plant Hydrilla verticillata (Hydrocharitaceae) in the southeastern United States. Biol Control. 8: 65 73. Chamovitz D., G. Sandmann, and J. Hirschberg. 1993. Molecular and biochemical characterization of herbicide resistant mutants of cyanobacteria reveals that phytoene desaturation is a rate limiting step in caro tenoid biosy nthesis. J Biol Chem. 268: 17348 17353. [CL IMOD] Climate Information for Management and Operational Decisions. 2012. Southeast Regional Climate Center ( http://acis.sercc.com ) Cook, C. D.K., and R. L nd. 1982. A revision of the genus hydrilla (Hydrocharitaceae ). Aquat Bot. 13: 485 504. Cuda, J.P., B.R. Coon, Y.M. Dao, and T.D. Center. 2002. Biology and laboratory rearing of Cricotopus lebetis (Diptera: Chironomidae), a natural enemy of the aquatic weed hydrilla (Hydrocharitaceae). Annu Entomol Soc America. 95: 587 596. Cuda, J.P., R. Charudattan, M.J. Grodowitz, R.M. Newman, J.F. Shearer, M.L. Tamayo, and B. Villegas. 2008. Recent advances in biological control of submersed aquatic weeds. J Aquat Plant Manage. 46: 15 32. Cuda. J.P., B.R. Coon, Y .M. Dao, and T.D. Center. 2011. Effect of an herbivorous stem mining midge on the growth of hydrilla. J Aquat Plant Manage. 49: 83 89. Cuda J.P and J.L. Gillett Kaufman. 2011. New hydrilla project: RAMP UP! Aquatics 33: 9 Danks, H.V. 1971. Overwintering of some north temperate and arctic Chironomidae: I. The winter environment. Can Entomol. 103: 589 604. Danks, H.V. 2007. How aquatic insects live in cold climates. Can Entomol. 139: 443 471. Del Fosse, E.S., and B.D. Perkins. 1976. A new U S record for Parapoynx diminutalis (Lepidoptera: Pyralidae), a possible biological control agent for Hydrilla verticillata. Fla Entomol. 59: 19 20. Diaz, R., W.A. Overholt, J.P. Cuda, P.D. Pratt, and A. Fox. 2008. Temperature dependent development, surviv al, and potential distribution of Ischnodemus variegatus (Hemiptera: Blissidae), a herbivore of west indian marsh grass. Ann Entomol Soc Am. 101: 604 612.

PAGE 113

113 Doong, R.L., G.E. MacDonald, and D.G. Shilling. 1993. Effect of fluridone on chlorophyll, carotenoid s and anthocyanin content of hydrilla. J Aquat Plant Manage. 31: 55 59. Dukes, J.S., and H.A Mooney. 2004. Disruption of ecosystem processes in western North America by invasive species. Rev Chil De Historia Natural. 77: 411 437. EPA Method 130.2 Approv ed for NPDES (Issued 1971). EPA Method 310.1 Approved for NPDES (Issued 1978). Epler, J.H., J.P. Cuda, and T.D. Center. 2000. Redescription of Cricotopus lebetis (Diptera: Chironomidae), a potential biocontrol agent of the aquatic weed hydrilla (Hydroch aritaceae). Fla Entomol. 83: 171 180. Esler, D. 1989. Assessment of American coot herbivory of hydrilla. J Wildl Manage. 53: 1147 1149. Esler, D. 1990. Avian community responses to hydrilla invasion. Wilson Bull. 102: 427 440. Flaig, E.G., and K.E. Have ns. 1995 Historical trends in the Lake Okeechobee ecosystem I. Land use and nutrient loading. Arch Hydrobiol. 107: 1 24. Florida Fish and Wildlife Conservation Commission, U.S. Fish and Wildlife Service. 2010. Lake Toho hydrilla control treatment plan for 2010/2011. ( http://myfwc.com/media/1205516/TohoHydrillaKite.pdf ) Godfrey, K.E., L.W.J. Anderson, S.D. Perry, and N. Dechoretz. 1994. Overwintering and establishment potential of Bagou s affinis (Coleoptera: Curculionidae) on Hydrilla verticillata (Hydrocharitaceae) in northern California. Fla Entomol. 77: 221 230. Godfrey, K.E. and L.W.J. Anderson. 1994. Feeding by Bagous affinis (Coleoptera: Curculionidae ) inhibits germination of hydrilla tubers. Fla Entomol. 77: 480 488. Greathead, D.J. 1973. Progress in the biological control of Lantana camara in East Africa and discussion of problems raised by the unexpected reaction of some of the more promising inset s to Seasamum indicum pp. 89 92. In P.H. Dunn (ed.), Proc eedings, 2 nd Int Symp. Biol Contr Weeds Commonwealth Inst of Biological Control October 1971, Commonwealth Agriculture Bureaux Gresens, S.E., K.T. Belt, J.A. Tang, D.C. Gwinn, and P.A. Banks. 20 07. Temporal and spatial responses of Chironomidae (Diptera) and other benthic invertebrates to urban stormwater runoff. Hydrobiologia. 575: 173 190.

PAGE 114

114 Grodowitz, M.J., T.D. Center, A.F. Cofransesco, and J.E. Freeman. 1997. Release and establishment of Hydr ellia balciunasi (Diptera: Ephydridae) for the biological control of the submersed aquatic plant Hydrilla verticillata (Hydrocharitaceae) in the United States. Biol Control. 9: 15 23. Guillory, V., and R.D. Gasaway. 1978. Zoogeography of the grass carp in the United States. T Am Fish Soc. 107: 105 112. Gurr, G. and S. Wratten. 2000. Biological control: measures of success. Kluwer Academic Publishers, Dordrecht, The Netherlands. Haller, W.T., and D.L. Sutton. 1975. Commu nity structure and competition between hydrilla and vallisneria. Hyacinth Cont. J., 13: 48 50. Hggstrm, H., and S Larsson. 1995. Slow larval growth on a suboptimal willow results in high predation mortality in the leaf beetle Galerucella lineola. Oecolo gia. 104: 308 315. Hansson, B.S. M. C. Larsson, and W. L. Leal, 1999 Green leaf volatile detecting olfactory receptor neurons display very high sensitivity and specificity in a scarab beetle. Physiol Entomol. 24: 121 126. Harmon Fetcho, J.A., C.J. Hapem an, L.L. McConnell, T.L. Potter, C.P. Rice, A.M. Sadeghi, R.D. Smith, K. Bialek, K.A. Sefton, B.A. Schaffer, and R. Curry. 2005. Pesticide occurrence in selected south Florida canals and Biscayne Bay during high agricultural activity. J Agric Food Chem. 53 : 6040 6048. Haye, B., H. Goulet, P.G. Mason, and U. Kuhlmann. 2005. Does fundamental host range match ecological host range? A retrospective case study of a Lygus plant bug parasitoid. Biol Control. 35: 55 67. Hespenheide, H. A. 1991. Bionomic s of leaf mining insects. Annu Rev Entomol 36: 535 560 Hijmans, R.J., L. Guarino, P. Mathur. 2012. DIVA GIS, version 7.5. A geographic information system for the management and analysis of genetic resources data. Manual. International Potato Center and Internation al Plant Genetic Resources Institute, Lima, Peru. Holt, R. D., and M.E. Hochberg. 1997. When is biological control evolutionary stable (or is it)? Ecology. 78: 1673 1683. Holt, J. S. and H. M. Lebaron. 1990. Significance and distribution of herbicide resistance. Weed Technol. 4: 141 149

PAGE 115

115 Howarth, F.G. 1991. Environmental impacts of classical biologic al control. Annu Rev Entomol. 36: 485 509. Hoyer, M.V., M. Woods Jackson, M.S. Allen, and D.E. Canfield Jr. 2 008. Lack of exotic hydrilla infestation effects on plant, fish and aquatic bird community measures. Lake Reserv Manage. 24: 331 338. Hoyer, M.V., M.D. Netherland, M.S. Allen, and D.E. Canfield, Jr. 2005. Hydrilla management in Florida: A summary and disc ussion of issues identified by professionals with future management recommendations. Final Document, 14 June 2005. 68 pp. Johnson F.A. and F. Montalbano III. 1984. Selection of plant communities by wintering waterfowl on Lake Okeechobee, Florida. J Wildl Manage. 48: 174 178. Johnson, F.A., and F. Montalbano III. 1987. Considering waterfowl habitat in hydrilla control policies. Wildl Soc Bull. 15: 446 469. Julien, M.H., J.K. Scott, W. Orapa, and Q. Paynter. 2007. History, opportunities and challenges fo r biological control in Australia, New Zealand and the pacific islands. Crop Prot. 26: 255 265. Keane, R.M., and M.J. Crawley. 2002. Exotic plant invasions and the enemy release hypothesis. Trends Ecol Evol. 17: 164 170. Key, P.B., K.W. Chung, A.D. Opatk iewicz, E.F. Wirth, and M.H. Fulton. 2003. Toxicity of the insecticides fipronil and endosulfan to selected life stages of the grass shrimp ( Palaemonetes pugio) Bull Environ Contam Toxicol. 70: 533 540. Klussmann, W.G., R.L. Noble, R.D. Martyn, W.J. Clark R.K. Betsill, P.W. Bettoli, M.F. Cichra, and J.M. Campbell. 1988. Control of aquatic macrophytes by grass carp in Lake Conroe, Texas, and the effect on the reservoir ecosystem. Texas Agriculture Experiment Station MP 1664, College Station, Texas. Langel and, K.A. 1996. Hydrilla verticillata (L .F.) Royle (Hydrocharitaceae), The perfect aquatic weed. Castanea. 61: 293 304. Lerner A., N. Meltser, N. Sapir, C. Erlick, N. Shashar, and M. Broza. 2008. Reflected polarization guides chironomid females to oviposition sites. J Exp Biol. 211: 3536 3543. Louda, S.M., and A.E. Arnett, 2000. Predicting non target ecological effects of biological control agents: evidence from Rhinocyllus conicus, pp. 551 567. I n N.R. Spencer (eds.) Proceedings Symposium: International biological control of weeds. 4 14 July 1999. Monta na State University, Bozeman, MT.

PAGE 116

116 Louda, S.M., R.W. Pemberton, M.T. Johnson, and P.A. Follett. 2003. Non target effects iological control? Retrospective analyses to reduce risk associated with biocontrol introductions. Annu Rev Entomol. 48: 365 396. Macrae, I.V., N.N. Winchester, and R.A. Ring. 1990. Feeding activity and host preference of the milfoil midge, Cricotopus myr iophylli Oliver Diptera: Chironomidae). J Aquat Plant Manage. 28: 89 92. Madden, C.P., P.J. Suter, B.C. Nicholson, and A.D. Austin. 1992. Deformities in chironomid larvae as indicators of pollution (pesticide) stress. Physiol Ecol. 26: 551 557. Madeira, P.T., T.K. Van, K.K. Steward, and R.J. Schnell. 1997. Random amplified polymorphic DNA analysis of the phenetic relationships among world wide accessions of Hydrilla verticillata. Aquat Bot. 59: 217 236. Madeira, P.T., C.C. Jacono, and T.K. Van. 2000. Monitoring hydrilla using two RAPD procedures and the nonindigenous aquatic species database. J Aquat Plant Manage. 38: 33 40. Maxwell, B.D., and A.M. Mortimer. 1994. Selecting for herbicide resistance, pp. 1 25 I n S.B. Powles and J.A.M. Holtum, ( ed s. ), Herbicide Resistance in Plants, Biology and Biochemistry. Lewis Publishing, Boca Ranton FL McFayden, R.E.C. 1998. Biological control of weeds. Annu Rev Entomol. 43: 369 393. McCombie, A. M.,1959. Some relations between air temperatures and the sur face water temperature of lakes. Limnol and Oceanogr 4: 525 258. Michel, A.,R.S. Arias, B.E. Scheffler, S.O. Dukes, M.D. Netherland, and F.E. Dayan. 2004. Somatic mutation mediated evolution of herbicide resistance in the nonindigenous invasive plant hyd rilla ( Hydrilla verticillata). Molecular Ecol. 13: 3229 3237. Moxley, D.J., and F.H. Langford. 1982. Beneficial effects of hydrilla on two eutrophic lakes in central Florida. Proc Annu Conf. 36: 280 286. Netherland, M.D., W.R. Green, and K.D. Getsinger. 1991. Endothall concentration and exposure time relationships for the control of Eurasian watermilfoil and hydrilla. J Aquat Plant Manage. 29: 61 67. Nichols, S.A., and B. H. Shaw. 1986. Ecological life histories of the three aquatic nuisance plants, My riophyllum spicatum, Potamogeton crispus and Elodea canadensis. Hydrobiologia. 131: 3 21.

PAGE 117

117 Two Indian Bagous weevils (Cole op tera, Curculionidae), tuber feeders of Hydrilla verticillata (Hydrocharitaceae) one potential biocontrol agent in Florida. Fla Entomol. 72: 462 469. Oliver, D.R. 1971. Life history of the Chironomidae. Annu Rev Entomol. 16: 211 230. Overholt, W.A. and J.P. Cuda. 2005. University of Florida scientists explore Afric a for natural enemies of hydrilla. Aquatics. 27: 18 20. Paine, G.A. and A.R. Gaufin, 1956. Diptera as indicators of pollution in a Midwestern stream. Ohio J Sci. 56: 291 304. Payton, M.E., M.H. Greenstone, and N. Schenker. 2003. Overlapping confidence in tervals or standard error intervals: What do they mean in terms of statistical significance? J Insect Sci. 3: 1 6. Pimental, D., and D. Andow. 1984. Pest management and pesticide impacts. Insect Sci Appl. 5: 141 149. Pimental D., C. Glenister, S. Fast, a nd D. Gallahan. 1984. Environmental risks of biological pest controls. Oikos. 42: 283 290. Pimentel, D., R. Zuniga, and D. Morrison. 2005. Update on the environmental and economic costs associated with alien invasive species in the United States. Ecol Eco n. 52: 273 288. Ppalov, I. 2006. A review of grass carp use for aquatic weed control and its impact on water bodies. J Aquat Plant Manage. 44: 1 12. Powles, S.B., and Q. Yu. 2010. Evolution in action: plants resistant to herbicides. Ann RevPlant Biol. 61: 317 347. Pullin, A.S. 1996. Physiological relationships between insect diapause and cold tolerance: coincidence or coevolution? European Journal of Entomology, 93, 121_129. Puri, A, G.E. MacDonald, W.T. Haller. 2007. Ploidy variations in fluridone s usceptible and resistant hydrilla ( Hydrilla verticillata) biotypes. Weed Sci. 55: 578 583. Rodgers, L. M. Bodle, D. Black, and F. Laroche. 2011. Chapter 9: Status of the nonindigenous species in the south Florida environment. 2011 South Florida Report. V olume I: the South Florida Environment Saether, O.A. 1979. Chironomid communities as water quality indicators. Holarctic Ecol. 2: 65 74.

PAGE 118

118 SAS Institute. 2008. SAS/STAT Schlenk, D., D.B. Huggett, J. Allgood E. Bennett, J. Rimoldi, A.B. Beeler, D. Block, A.W. Holder, R. Hovinga, and P. Bedient. 2001. Toxicity of fipronil and its degradation products to Procambarus sp.: field and laboratory studies. Arch Environ Contam Toxicol. 41:325 332. Schmid, T. A., J.P Cuda, G.E. Macdonald, and JL. Gillmore. 2010. Performance of two established control agents on hydrilla genotypes susceptible and res is tant to fluridone herbicide. J. Aquat Plant Manage. 48: 102 105. Schmitz, D.C., B.V. Nelson, L.E. Nall, and J.D. Schardt. 1991. Exotic aquatic plants in Florida: a historical perspective and review of the present aquatic plant regulation program. Proceedings of the Sym on Exotic Pest Plants. 91: 303 326. Schoonhoven, L.M., J. J.A. van Loon, and M. Dicke. 2005. Insec t plant biology: Host plant selection: How to find a host plant. 2 nd ed. Oxford University Press, New York, NY. Shannon, C.E. and W. Weaver. 1949. The mathematical theory of communications. The University of Illinois Press, Urbana, IL. Shearer, J. F., an d L. S. Nelson. 2009. Combining ALS inhibiting herbicides with the fungal pathogen Mycoleptodiscus terrestris for control of hydrilla APCRP Technical Notes Collection. ERDC/TN APCRP CC 11. Vicksburg, MS: U.S. Army Engineer Research and Development Cent er. Sheppard, A.W., R.D. van Klinken, and T.A. Heard. 2005. Scientific advances in the analysis of direct risks of weed biocontrol control agents to nontarget plants. Biol Control. 35: 215 226. Shireman, J.V. and M.J. Maceina. 1981. The utilization of gr ass carp, Ctenopharyngodon idella Val., for hydrilla control in Lake Baldwin, Florida. J Fish Biol. 19: 629 636. Simberloff, D., D.C Schmitz, and T.C. Brown (eds). 1997. Strangers in paradise: impact and management of non indigenous species in Florida. I sland Press. Washington, D.C. Simberloff, D., and P. Stiling. 1996. Risks of species introduced for biological control. Biol Conserv. 78: 185 192. Slipke, J.W., M.J. Maceina, and J.M. Grizzle. 1998. Analysis of the recreational fishery and angler attitudes toward hydrilla in Lake Semino le, a southeastern reservoir. J Aquat Plant Manage. 36: 101 107.

PAGE 119

119 Sokal, R.R, and F. J. Rohlf. 1995. Biometry: Estimation and hypothesis testing. 3 rd ed. W H Freeman, New York, NY. Stehr, C.M., T.L. Linbo, J.P. Incardona, and N.L. Scholz. 2006. The developmental neurotoxicity of fipronil: notochord degeneration and locomotor defects in zebrafish embryos and larvae. Toxicol Sci. 92: 270 278. Steward, K.K., T.K. Van, V. Carter, and A.H. Pieterse. 1984. Hydrilla invades Washington D.C. and the Potomac. Amer J Bot. 7:162 163. Stranden, M., I. Liblikas, W.A. Koenig, T.J. Almaas, A.K. Borg Karlson, and H. Mustaparta. 2003. Germacrene D receptor neurons in three species of h eli o thine moths: structure activity relationships. J Comp physiol A. 189: 563 577. Sublette, J. E. 1964. Chironomidae (Diptera) of Louisiana I. Systematics and immature stages of some lentic chironomids of west central Louisiana. Tulane Stud Zool 11: 109 150. Sutton, D.L., and V.V. Vandiver. 1986. Grass carp a fish for biological management of hydrilla and other aquatic weeds in Florida. University of Florida, Agricultural Experiment Station Bulletin 867, Gainesville FL Torchin, M.E., K.D. Lafferty, A.P. Dobson, V.J. McKenzie, and A.M Kuris. 2003. Introduced species and their missing parasites. Nature. 421: 628 630. [USGS] U.S. Geological Survey. 2006. water, 1992 2001. Circular 1291. Van, T.K., W.T. H aller, G. Bowes, and L.A. Garrard. 1977. Light quality on growth and chlorophyll composition in hydrilla. J Aquat Plant Manage. 15: 29 31. Wapshere, A.J. 1974. A strategy for evaluating the safety of organisms for biological weed control. Ann Appl Biol. 7 7: 201 211. Wheeler, G.S., and T.D. Center. 2001. Impact of the biological control agent Hydrellia pakistanae (Diptera: Ephydridae) on the submersed aquatic weed Hydrilla verticillata (Hydrocharitaceae). Biol Cont. 21: 168 181. Wilde, G.R., R.K. Riechers, and J. Johnson. 1992. Angler attitudes toward control of freshwater vegetation. J Aquat Manage. 30: 77 79. Williams, C.J. 1982. The drift of some chironomid egg masses (Diptera: Chironomidae) Freshwater Biol. 12: 573 578.

PAGE 120

120 Williams, J.R. 1954. T he biological control of weeds pp. 95 98. I n : Report of the Sixth Commonwealth Entomological Congress, London, UK. Wirth, W.W. 1949. A revision of the Clunionine midges with descriptions of a new genus and four new species. Univ Calif Publ Entomol. 8: 151 182. Wolfe, L.M. 2002. Why alien invaders succeed: support for the escape from enemy hypothesis. Am Nat. 160: 705 711. Wu, J., J. Lu, C. Wilson, and H. Lu. 2010. Effective liquid extraction method for analysis of pyrethroid and phenylpyrazole pestici des in emulsion prone surface water samples. J of Chromatogr A. 1217: 6327 6333. Zhou, G. W.A. Overholt, and S.W. Kimani Njogu. 2003. Species richness and parasitism in an assemblage of parasitoids attacking maize stem borers in co astal Kenya. Ecol Entomo l. 28: 109 118.

PAGE 121

121 BIOGRAPHICAL SKETCH Karen Stratman was born and raised in Cincinnati Ohio. Upon high school graduation in 2006, she attended Clemson University in S outh Carolina for her Bachelor of Science degree in environmental and natural r esou rces. Her first experience working with invasive plants was in the summer of 2009 at Congaree National Park in Hopkins South Carolina. There she worked to control and manage a variety of invasive plants. Since then she has been interested in researching ways to effectively manage invasive species. In 2010, she obtained her Bachelor of Science and began a Master of Science of e ntomology at the University of Florida. There she studied biological control of invasive weeds, and her research focused on resear ching Cricotopus lebetis, a potential biological control agent of hydrilla. She is a current member of the Entomological Society of America, Florida Exotic Pest Plant Council, and the Florida Aquatic Plant Management Society. Her plans for the future inclu de continued work and research with invasive plants and developing effective management plans.