1 LETHAL AND SUBLETHAL EFFECTS OF IMIDACLOPRID AND AMITRAZ ON Apis mellifera LINNAEUS (HYMENOPTERA: AP IDAE) LARVAE AND PUPAE By PATRICIA L. TOTH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009
2 2009 Patricia L. Toth
3 To my parents
4 ACKNOWLEDGMENTS I thank m y advisor, Dr. Jamie Ellis, and the other members of my supervisory committee, Dr. Mike Scharf and Dr. Glenn Ha ll, for mentoring me. I thank my parents for all of their moral support. I thank Hannah OMalley, Mike OMalle y, Meredith Cenzer, Renee Cole, and Sparky Vilsaint for technical support. Thanks go to the graduate student s in the Honey Bee Research and Extension Laboratory for all of their help, s upport, and humor: Jason Graham, Eddie Atkinson, and Anthony Vaudo. Thanks go to Debbie Hall for all of her assistance throughout my entire program. I thank those at the Florida Department of Agriculture and Consumer Services, Division of Plant Industry, especially Jerry Hayes and Tom Dowda, for their generous support and encouragement.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4 LIST OF TABLES ...........................................................................................................................7 LIST OF FIGURES .........................................................................................................................9 LIST OF ABBREVIATIONS ........................................................................................................ 10 ABSTRACT ...................................................................................................................... .............11 CHAP TER 1 OVERVIEW OF CURRENT HONEY BEE ASSOCIATED PROBLEM S .......................... 13 Economic Importance of Honey Bees ....................................................................................13 Current Declines in Pollinator Populations ............................................................................14 Introduction to Research .........................................................................................................22 2 TOXICITY AND SUBLETHAL EFFECTS OF IMIDACLOPRID AND AMITRAZ ON HONEY BEE LARVAE AND PUPAE ..........................................................................27 Materials and Methods ...........................................................................................................30 General Rearing Procedure ..............................................................................................30 Pesticide Application .......................................................................................................31 Statistical Analysis .......................................................................................................... 33 Results .....................................................................................................................................34 Imidacloprid .................................................................................................................. ..34 Amitraz ....................................................................................................................... .....36 Discussion .................................................................................................................... ...........38 Imidacloprid .................................................................................................................. ..38 Amitraz ....................................................................................................................... .....41 Summary ....................................................................................................................... ...42 3 SUMMARY ....................................................................................................................... .....55 APPENDIX A METHOD FOR TOPICAL APPLICATION OF SUBC HRONIC IMIDACLOPRID AND AMITRAZ DOSES TO HONEY BEE LARVAE ........................................................ 64 B METHOD FOR TOPICAL APPLICATION OF ACUT E IMIDACLOPRID DOSES TO HONEY BEE LA RVAE ......................................................................................................... 71
6 C METHOD FOR TESTING THE ATTRACTI ON OF FOUNDRESS VAR ROA MITES TO IMIDACLOPRID AND AMITRAZ TREATED LARVAE ............................................ 80 D HYGIENIC BEHAVIOR OF ADULT HONEY BEES T OWARD LARVAE TREATED WITH DIFF ERENT SOLVENTS ....................................................................... 86 LIST OF REFERENCES ...............................................................................................................88 BIOGRAPHICAL SKETCH .........................................................................................................97
7 LIST OF TABLES Table page 1-1 Crops known to be pollinated by honey bees ....................................................................24 1-2 Crops most likely to be treated wi th im idacloprid in California in 2006 .......................... 25 2-1 Percentage of larvae that defecated, pupal surv ival, and adult bee emergence rates for larvae fed a diet containing imidacloprid ........................................................................... 45 2-2 Mean time to larval defecation in hours for larvae fed diet containing im idacloprid ........ 46 2-3 Mean larval weight at defecation for larvae fed a diet containing im idacloprid ............... 48 2-4 Mean time to adult emergence, adult bee weight, and adult bee head weight for larvae fed a diet containing imidacloprid.. ......................................................................... 50 2-5 Percentage of larvae that defecated, pupal surv ival, and adult bee emergence for larvae fed a diet containing amitraz. .................................................................................. 51 2-6 Mean time to larval defecation (h) for larvae fed a diet containing am itraz. ..................... 52 2-7 Mean larval weight at de fecation, time to adult bee em ergence, adult bee weight, and adult bee head weight for larvae fed a diet containing amitraz. ........................................ 53 3-1 Published contact and oral LD50 values for im idacloprid tested on adult honey bees ..... 61 A-1 Percentage of larvae that defecated, pupal surv ival, and adult bee emergence for larvae treated topically with imidacloprid for 4 days ........................................................ 66 A-2 Mean time to larval defecation, larval weight at d efecation, time to adult bee emergence, and adult bee head weight for larvae treated topically with imidacloprid for 4 days............................................................................................................................67 A-3 Percentage of larvae that defecated, pupal surv ival, and adult bee emergence for larvae treated topically w ith amitraz for 4 days. ................................................................ 68 A-4 Mean time to larval defecation, larval weight at d efecation, and adult bee head weight for larvae treated topica lly with amitraz for 4 days. ..............................................69 A-5 Mean time to adult emergence in hours for larvae treated topically with amitraz for 4 days .......................................................................................................................... ..........70 B-1 Percentage of larvae that defecated fo r larv ae treated topically with imidacloprid once during their development........................................................................................... 73 B-2 Percentage of pupal survival for pupae tr eated topically as larvae with im idacloprid once during their development........................................................................................... 74
8 B-3 Percentage of adult bee emergence for a dult bees treated as larvae topically with imidacloprid once during their development ..................................................................... 75 B-4 Mean time to larval defecation for larv ae trea ted topically with imidacloprid once during their development.. .................................................................................................76 B-5 Mean larval weight at de f ecation for larvae treated t opically with imidacloprid once during their development ................................................................................................... 77 B-6 Mean time to adult emergence for larv ae treated topically wi th im idacloprid once during their development.. .................................................................................................78 B-7 Mean adult bee head weight for adult bees trea ted topically as larvae with imidacloprid once during their development.. ................................................................... 79 C-1 Mean number of varroa mites found alive in capped cells containing either im idacloprid or amitraz fed larvae. .................................................................................... 83 D-1 Percentage of larvae abor ted according to solvent tr eatm ent applied topically. ................ 87
9 LIST OF FIGURES Figure page 1-1 Number of honey bee colonies in the U.S. ........................................................................ 26 2-1 Laboratory bioassay: experimental timeline according to larv al development in hours. ........................................................................................................................ ..........54 3-1 Hypothetical model depicting the range of im idacloprid doses tested (5-80 ppb) and the range of doses likely to be tested in an LD50 test (ppb-ppt). ...................................... 62 3-2 Hypothetical model illustrating a typical dose-res ponse curve.. ....................................... 63 C-1 Lifecycle of the varroa mite .............................................................................................. .84 C-2 Field bioassay: experimental timeline acco rd ing to larval development in hours (0528 h, x-axis). ....................................................................................................................85
10 LIST OF ABBREVIATIONS ANOVA Analysis of variance NOED No observable effect dose. The highe st xenobiotic dose administered that does not produce an adverse affect on that organism. Sublethal A xenobiotic administered at a particular dose that does not kill a significant proportion of the experimental population. Subchronic A technique of administering a xenobiotic fo r 2 or more consecutive days, but not for the entire life of the test organism. This differs from chronic application, where a xenobiotic is ap plied to the organism throughout the organisms life. Xenobiotic A chemical found in an organism th at is considered foreign and harmful to that organism.
11 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science LETHAL AND SUBLETHAL EFFECTS OF IMIDACLOPRID AND AMITRAZ ON Apis mellifera LINNAEUS (HYMENOPTERA: AP IDAE) LARVAE AND PUPAE By Patricia L. Toth May 2009 Chair: James D. Ellis Major: Entomology and Nematology The population of feral and managed honey b ee colonies in the United States has declined steadily since the mid 1950s. These de clines have been attributed to a number of causes including, but not limited to, pesticide ex posure, parasite vectored pathogens, arthropod bee pests, and various stressors. Investigators agree that bee declines more likely are due to a mixture of two or more of these causes, but few investig ations have been conducted on interactions between these factor s. Researchers have focused on testing the effects of various xenobiotics on honey bee workers, queens, and drone s. Effects of these xenobiotics on honey bee brood have not been well-investi gated. Herein, I fed bee larvae w ith various concentrations of imidacloprid and amitraz in an attempt to document lethal and sublethal effects of these chemicals on bee brood. Results indicated that larvae fed imidacloprid at 5, 10, 40, and 80 ppb or amitraz at 25, 50, 100, 200, and 400 ppb were signifi cantly less likely to survive to adulthood. Larvae fed 25 and 400 ppb of amitraz were less likely to survive to the prepupal defecation stage. In general, treated bee mortality was greater dur ing pupal development rather than during larval instar stages. I found no effects of imidacloprid or amitraz on larval time to defecation, larval weight at defecation, time to adult emergence, a dult bee weight, or adult bee head weight. As a result of this investigation, several methods are discussed concer ning topical treatment vs. oral
12 applications of imidacloprid and amitraz to developing larvae. In an attempt to better understand interactions among stressors, I wanted to develo p a method for testing th e susceptibility after pesticide exposure of treated larvae to depredation caused by the ectoparasite Varroa destructor (varroa mite). Results presented in this thesis provide important insights into the sublethal effects of chemicals on developing honey bee brood. This understanding ultimately can lead to better honey bee management and agricultural practices important for sustai nable apiculture.
13 CHAPTER 1 OVERVIEW OF CURRENT HONEY BEE ASSOCIATED PROBLEMS Economic Importance of Honey Bees The European honey bee, Apis mellifera L., is an econom ically important insect not only for its production of honey and beeswax, but also because of its role as a pollinator. Pollination is important for the production of agricultural co mmodities such as food (grain, fruits, nuts, vegetables or domesticated animals), fuel, and drugs (Berenbaum et al. 2007). Honey bees are good pollinators because they are generalists, pol linating a wide variety of agricultural crops (Table 1-1) and native plants over a large fo raging area throughout the y ear (Proctor et al. 1996, Delaplane and Mayer 2000). Additionally, pollination by honey bees often results in an increase in quantity and quality of crop yi eld (Delaplane and Mayer 2000). The production of important crops that rely on honey bee pollination has increased steadily in recent years in response to increased huma n demand for these crops (National Agricultural Statistics Service (NASS) 19762008). This increased demand fo r food due to the exponential growth in the population of hu mans has increased the amount of managed honey bee colonies needed for crop pollination (Morse and Calder one 2000). In 2007, there were an estimated 2.44 million honey bee colonies in the U.S.1 (NASS 1976-2008). Southwick and Southwick (1992) estimated the annual benefit of honey bee pollinati on to agriculture at be tween 1.6 and 8.3 billion USD. In 2000, the value attributed to honey bee crop pollination in the United States alone was estimated at 14.6 billion USD, a 57% increase from 1992 (Morse and Calderone 2000). The most recent study suggesting the benefits of insect pollinators in genera l on a worldwide scale estimated their value at 217 b illion USD (Gallai et al.). 1 Beekeepers with fewer than 5 colonies or colonies that did not produce honey were not counted in this survey.
14 Current Declines in Po llinator Pop ulations Pollination is important for plant reproduction and for maintaining pl ant genetic diversity (Delaplane and Mayer 2000). Due to the importance of pollinators, many researchers have investigated pollinator population dynamics. Native pol linators such as birds, bats, beetles, flies, and solitary bees have decreased in abundance for reasons including habitat fragmentation/destruction (Ste phen 1955, Aizen and Feinsinger 1994) and the introduction of pesticides and parasites (Kevan 1975, Berenbaum et al. 2007). Monitoring programs led by the U.S. Department of Agricultures National Ag ricultural Statistics Service (NASS) have documented the decline in managed honey bee colonies since 1947, making them the most significant example of pollinator decline in North America (Berenbaum et al. 2007, NASS 19762008) (Figure 1-1). Factors believed to contribut e to the decline of managed honey bee populations include the introduction of parasitic mites, mite resistan ce to acaricides, introducti on of Africanized bees, pathogens, effects of pathogen-ta rgeted antibiotics, pesticides, and many others (Johansen and Mayer 1990, Morse and Flottum 1997, Wilson et al. 1997). A sharp decline in managed colonies occurring from 1984-1996 corresponds to the introduction of two parasitic mite species, the tracheal mite, Acarapis woodi (Rennie), and the varroa mite, Varroa destructor Anderson and Trueman (DeJong 1997, Wilson et al. 1997). There ar e no established reasons for the managed bee declines occurring from 1947-1972. However, th is decline corresponds with widespread, commercial use of first generation neurotoxic insecticides such as carbaryl, parathion, malathion, and diazinon (Johansen and Mayer 1990), among others. Reasons for the most recent decline (from 2005 until the present) of managed honey bee colonies, referred to as colony collapse disorder or CCD, remain unknown. Beekeepers estimate the total colony losses due to CCD to be somewhere between 35% and 100% (Eccleston
15 2007, Stokstad 2007). Although the cause of CCD remains under investigation, researchers typically suggest the following, among others as candidates: (1) pathogens (known or undiscovered), (2) pesticide use (bot h inside and outside of colonies ), (3) stress (such as colony migration/transportation, overcrowding, and poor nutrition), (4) climate change, (5) lack of genetic diversity, or (6) poor nutrition (Eccleston 2007, Ellis 2007, Johnson 2007, Embrey 2008). Of particular interest to me is how pesticide use might affect immature honey bee development. Pesticides used to suppress pest insect popula tions can affect non-targ et/beneficial insects, including pollinators. The yearly estimated cost of pollination losses due to pesticide exposure is $210 million USD (Pimentel 2005). To limit toxicity of pesticides to pollinators, regulations have been encouraged to avoid certain pestic ide use during crop bloom or where bees are known to forage (Morse and Flottum 1997). Additional measures were taken to protect pollinators by passing of the Federal Insecticide, Fungicide, and Rodenticide Act, which requires regi stered pesticides to be tested to determine the lethal dose (LD50) or c oncentration (LC50) on non-target insects (Desneux et al. 2007). However, legislation typically fails to consider sublethal effects th at could be delayed, indirect, difficult to detect/quantify, or easily overlooked (Desneux et al. 2007) Parameters that have been shown to be influenced by sublethal doses of pesticides include developmental rate, metabolic activity, adult longev ity, immunity, fecundity, and be havior (Desne ux et al. 2007). To date, investigations measuring sublethal e ffects of pesticides on honey bees have been quantified either biochemically or behaviora lly. Bioassays performed using the proboscis extension response (PER) (Bitterman et al. 1983) typically are used to study sublethal effects of pesticides on honey bee behavior. Using this method, fipronil, deltamethrin, endosulfan, and prochloraz have been shown to decrease learning performance in bees (Decourtye et al. 2005).
16 Additional sublethal effect s after treatment of adult bees w ith deltamethrin (700 ng/bee) include a decrease in food consumption, irregular spaced pupae in the comb, and a decrease in forager activity (Decourtye et al. 2004). Investigations quantifying the effects of pesticides on bee brood are scarce, most likely because of difficulties associated with rearing larvae in vitro and/or difficulties controlling variability in field colonies. One investigation in which researchers were able to successfully rear bee larvae in vitro concluded that those larvae treated with dimethoate and carbofuran developed faster than those not treated, ye t they failed to spin silk, sugge sting that the larvae would not pupate (Davis et al. 2000). It is important to focus on the sublethal eff ects of pesticides on bees, because bees are more likely to be exposed to low doses of pestic ides from the environment rather than higher, acute doses. Sublethal do ses of pesticides can be acquired by honey bees in the field as they forage for nectar, pollen (Chau zat et al. 2006), and perhaps wa ter (Wauchope 1978). The theory of flower constancy states that honey bees ar e more likely to forage on the same species of flowers on which they had foraged in the past (F ree 1963). Therefore, bees foraging on pesticidetreated plants will continue to be exposed to those plants over time. Pesticides can contaminate nectar and pollen when directly spraye d on or near flowers frequented by bees, or when plants themselves are treated with systemic pesticides. Systemic pesticides are absorbed by a plant and then tr ansported throughout the plants vascular system. They are especially useful against sucking pest s (Bennett 1957, Sicbaldi et al. 1997). Chemical residues from systemic pesticides are found not only on treated plants but also on neighboring vegetation due to pesticide dr ift even when non-honey bee pollin ated crops are treated. For example, investigators found that fields sown with imidacloprid-dipped seeds resulted in residue
17 on bordering grass and flowers up to 4 days after sowing. Contamination occurs during the sowing process when imidaclopri d residues escape thr ough the fan drain of the planting drill (Greatti et al. 2003, Greatti et al. 2006). Additional problems occur with pesticides that have long residual times. Residual time is the amount of time a pesticide can be found on a treated surface before it degrades. Residual times have been measured for a variety of pe sticides and can last from 2 hours to 7 days (Johansen and Mayer 1990, Wang et al. 2008). Pestic ide applicators who do not follow the label instructions and spray blooming crops risk an additional exposure to honey bee colonies. Once honey bee foragers contact xenobiotic cont aminants in the environment, they have the potential to return the material to their hive. Once in the hive, it is distributed to nurse bees and either stored or distributed to larvae as food (Johansen and Mayer 1990). For example, the highest levels of two pesticides, carbofuran a nd dimethoate, have been found in the adult honey bee crop, but smaller concentratio ns were also present in the hypopharyngeal gland (important in the production of worker and roya l jelly) (Davis and Shuel 1988). The contents of the crop are shared between forager and nurse bees, facilitating the transfer of contaminated substances throughout the hive (Davis 1989). Once pesticides have been introduced into the hive they can be stored and deposited in a numb er of potential products such as honey, pollen, and beeswax. For example, chemicals that are hydroph ilic (water soluble-such as systemic pesticides) are more likely to be dissolved in honey (Wallner 1999). One such systemic pesticide is imidacloprid, a nicotinoid. It was chosen for testing here because of its mode of action, likelihood to be found in flowers due to its systemic properties, and its worldwide popularity. There are 584 register ed products in the Un ited States containing imidacloprid as an active ingredient, and these pr oducts are used on a variety of crops (Table 1-
18 2), structures, and landscapes for insect pest cont rol (Kegley et al. 2008). Imidacloprid is applied primarily as a seed dressing or soil treatment but can be applied as a foliar treatment. Currently, imidacloprid has been banned in France where it was thought to have caused massive bee declines (Stokstad 2007). To da te, it is registered for use in Cameroon, Madagascar, South Africa, Tanzania, Australia, India, New Zeal and, the Philippines, Denmark, Finland, Germany, Hungary, Netherlands, Portugal, United Kingdom, Ca nada, and the United States (Kegley et al. 2008). Imidacloprid functions by targeting the nico tinic acetylcholine receptor (nAChR) and acts as an agonist, resulting in an influx of sodium ions leading to neuroexcitation (Yu 2008). However, long term effects of imidacloprid on in sects include primarily neuroinhibition (Nauen et al. 2001, Scharf 2003). Imidaclopri d is considered highly toxic to honey bees regardless of bee race (Nauen et al. 2001). Using or al toxicity studies on adult b ees, investigators have shown 1750% mortality at doses greater than 3.1 ng AI ( active ingredient) per b ee and a contact LD50 of 42 to 104 ng AI per bee (Nauen et al. 2001). Ev en after imidacloprid is metabolized, its metabolites, such as olefin and 5-hydroxyimid acloprid, have toxic effects on honey bees comparable to those of the pare nt compound (Suchail et al. 2001). Imidacloprid should be considered for possible sublethal effects on be neficial insects due to its persistence in plants long after initial app lication. Out of eleven different pesticides tested, imidacloprid had the longest residual time with 80% mortality for a parasitic wasp 7 days after treatment (Wang et al. 2008). Applications to ci trus trees resulted in xylem fluid containing imidacloprid concentrations of 10g/L at 24 weeks post treatment (Castle et al. 2005). A soil halflife, which is the time it takes for half of the applied amount of a pesticide to degrade, is estimated at 8-48 days (Yu 2008). However, there was a mean estimate of 6 g of imidacloprid
19 per kg of soil 1 to 2 years after soils were plan ted with maize, wheat, sunflower, or rape seeds dressed with imidacloprid (Bonmatin et al. 2003). Imidacloprid has been detected in sunflowers and corn at 1-11 g of imidacloprid/kg of pollen (1-11 ppb) and 1-3 g/kg (1-3 ppb), respectively (Bonmatin et al. 2003). Concentrations of imidacloprid found in pollen samples collected from honey bee colonies were reported to be between 2-6 g/kg (2-6 ppb; Chauzat et al. 2006). Acute sublethal doses of imidacloprid (100, 500, and 1000 ppb) have been found to disorient foragers and prevent/delay them from returning to the hive (Bortolotti et al. 2003). Bees treated with 100 and 500 ppb of imidacloprid expe rienced decreased mobility for up to one hour after a single dose (Medrzycki et al. 2003). Decourtye et al. ( 2004) reported that imidacloprid levels as low as 10 ng/bee resulted in decrease d food consumption by foragers, decreased forager activity, reduced brood area, and reduced honey and pollen stores. These researchers also recorded irregular capping of brood cells by worker bees treated with imidacloprid. In addition to being exposed to pesticides in agricultural and other settings, honey bees can be exposed to pesticides when beekeepers de liberately administer chemical treatments in colonies to control any number of hive pests. The primary example concerns beekeeper use of acaricides for control of the ecotoparasite Varroa destructor Anderson and Trueman (varroa). Treatment for varroa is important to beekeepers because varroa are considered the number one cause of honey bee colony deaths worldwide (DeJong 1997). The varroas natural host is the Asian honey bee, Apis cerana Fabr, but the importation of managed A. mellifera colonies into Asia resulted in a host switch by varroa (DeJong 1997); varroa now successfully parasitizes A. mellifera Unlike for A. mellifera varroa are not lethal to A. cerana colonies due to bee cannibalism and removal of infested brood (Rat h and Drescher 1990, Rath 1999), encapsulation
20 of infested brood (Koeniger 1987), grooming be havior (Rath 1999), and colony level absconding (Woyke 1976) in A. cerana colonies. Varroa infestations cause colony death in A. mellifera but the exact mechanism eliciting collapse is unknown (DeJong 1997). Varroa surviv e by feeding on adult and pupal hemolymph and reproducing inside capped brood. Even though varroa are large parasites (1.1 mm long) in relation to the size of their host, they seem to cause little damage to the bee exoskeleton when feeding (DeJong 1997). However, colony fitness could be limited due to reduced weight gain in worker (De Jong 1982) and drone pupae (Duay et al. 2003). Additionally, it is possible that varroa transmit honey bee viruses such as Kasmir bee virus (Shen et al. 2005) and deformed wing virus (Bowen-Walker et al. 1999), among others. Bee parasitization by the mite has been shown to lower the honey bee immune respons e by down-regulating immune related genes in moderately mite-infested bees (Gregory 2005) an d possibly increasing the replication of varroa transmitted viruses (Yang and Cox-Foster 2005) resulting in the death of the colony. Varroa control in bee colonies is attempted using a lim ited number of different types of acaricides that typically are applied inside managed bee colonies. Many acaricides are lipophilic a nd are distributed easily throu ghout the hive because they adhere to the cuticular wax laye r of the honey bee exoskeleton. Th eir lipophilic properties make acaricides more likely to accumulate in beeswa x as is the case for fluvalinate (Wallner 1999). Despite this, acaricides have been found in honey, as has been shown for coumaphos, an organophosphate (Rial-Otero et al. 2007) Pettis et al. (2004) investig ated the sublet hal effects of coumaphos on bees. They found that female b ee larvae grafted into queen rearing cells constructed of coumaphos-impregnated beeswax we re rejected by nurse bees resulting in 93% larval mortality. Of the surviving queens, thos e reared in coumaphos treated wax weighed less
21 than queens reared in coumaphos free wax (Pet tis et al. 2004). The number of commercially available products for use against varroa is lim ited. Hence, varroa quickly developed resistance to most of the acaricides used agai nst it in the U.S. (Oldroyd 2007). Amitraz [1, 3-di-(2, 4-dimethylphenylimino )-2-methyl-2-azapropane], a formamidine insecticide, is a chemical used widely by beekeep ers in the U.S. and Europe as an acaricide. I chose it for inclusion in this study because of the lack of research on its effects on honey bees. The lack of information may be because amitraz is not registered currently for use against varroa in the U.S. Unlike other acaricides, amitraz is show n to be unstable in honey and in beeswax, but instead it is broken down quickly into its metabo lites (Lodesani et al. 1992, Martel et al. 2007, Rial-Otero et al. 2007). Within days, amitraz ap plied to plates contai ning beeswax was quickly degraded into its metabolites, one of whic h is 2-4-dimethyl-aniline (Wallner 1999). Currently, there are 8 registered products cont aining amitraz in the U.S., which mainly are intended for use to control ticks, but can be used on organophosphate or carbamate-resistant insects such as the pea psy lla and cotton bollworm (Kegley et al. 2008, Yu 2008). Amitraz is registered as an acaricide in most of Europe, but information is lacking on how it may affect honey bees negatively. Concerning its modus operandi amitraz acts on octopamine receptors as an agonist and can cause a varied degree of behavioral effects including redu ced feeding and neurological effects such as neural excita tion/inhibition (Matsumura 1975, Yu 2008). The half-life of amitraz in soil is approximately 2 days, which is short co mpared to other pesticides such as DDT (3-10 yrs), fipronil (18-308 days), and imidacloprid (8-48 days) (Yu 2008). Compared to flumethrin and fluvalinate (bot h are acaricides), amitrazapplied topically is considered to be moderately toxic to adult honey bees with an LD50 of 2.55 g amitraz/mL of
22 acetone (255 ppb, Santiago et al. 2000). Santiago et al. (2000) described amitraz as being twelve times more potent as an acaricide (LD50: 1.7 pg of amitraz per mite) than insecticide (LD50: 2.55 g of amitraz per bee) since va rroa could be killed at much lower doses relative to bees. However, Santiago and colleagues did not invest igate the potential sublet hal effects of amitraz on adult bees or bee brood. There is a minimal amount of research that has been performed on toxicity and sublethal effects of amitraz on honey bee colonies, but such research is needed. Because amitraz is perceived by th e beekeeping industry in the U.S. to be safe for bees, many beekeepers have misused it and/or overdosed thei r colonies. Consequently, it is vital that the effects of amitraz on bees be identified. Introduction to Research The purpose of this study was to investigat e how sublethal doses of imidacloprid and amitraz affect immature bee development. Deve loping bee larvae were treated with various concentrations of one of the tw o test pesticides and then measured to quantify lethal and sublethal effects on larvae and puape (Chapter 2). I first had to en sure that my pesticides were not acutely lethal to bees by comparing treated a nd untreated larvae for the number of larvae that defecated, pupal survival (the number of individuals that survived pupal development of those that defecated as larvae), and adult bee emergence (total number of bees to eclose from those that were grafted as larvae). To measure sublethal effects, I looked for differences between treated and untreated larvae with regard to time to larval defecation (measured from egg hatch to defecation), larval weight at defecation, time to adult emergence (measured from egg hatch to adult bee emergence), adult bee weig ht, and adult bee head weight. I also compared topical vs. oral administrati on of imidacloprid to bee larvae as well as attempted to develop a method for testing the su sceptibility of treated larvae to depredation caused by varroa. My research methods for topica lly applying pesticides and my reasons for
23 choosing to administer the xenobiotic orally rath er than topically are de scribed in Appendices A and B. I introduced pesticide treated larvae in to varroa infested colonies to determine the invasion rate of mites into cel ls containing pesticide-treated larvae (Appendix C). Results from my research suggest possible ways that certain pesticides could be affectin g beneficial insects negatively without having direct effects (immediate, increased mortality; Chapter 3). This new understanding can lead to better management prac tices important for sustainable apiculture.
24 Table 1-1. Crops known to be pollinated by honey bees Agave, alfalfa, allspice, almond, apple, aspa ragus, avocado, lima bean, fava bean, beet, blueberry, Brussels sprout, ca bbage, canola, cantaloupe, carrot, cherry, clove, clover, cotton, cranberry, cucumber, guava, kiwifruit, lettuce, macadamia, onion, passion fruit, peach, pear, plum, raspberry, soybean, squash, strawberry, sunflower, tomatoa, watermelon Proctor et al. 1996, Caron 1999, Delaplane and Mayer 2000, Caron 2001, Berenbaum et al. 2007 a limited to greenhouse production
25 Table 1-2. Crops most likely to be treated with imidacloprid in California in 2006 Kegley et al. 2008 Apples, artichokes, bell peppers, blueberries, bok choy, broccoli, broccoli raab, brussel sprouts, cabbage, cantaloupe, cauliflower, celery, chicory, Chinese cabbage, cilantro, citrus, cotton, cucumbers, endive, grapefruit, grapes, head lettuce, kale, leaf lettuce, lemons, melons, oranges, pears, pecans, pomegranates, potatoe s, pumpkins, spinach, strawberries, succulent beans, sugarbeets, tangelos, tangerines, tomatoes, and watermelons
26 Figure 1-1. Number of honey b ee colonies in the U.S. Numb ers are based on the number of honey producing colonies per year (NASS 1976-2008, http://www.nass.usda.gov ).
27 CHAPTER 2 TOXICITY AND SUBLETHAL EFFECTS OF IMIDACLOPRID AND AMITRAZ ON HONE Y BEE LARVAE AND PUPAE Honey bees are an important bi ological indicator species (Porri ni et al. 2003), with a given colonys health potentially reflecting the health of its surrounding environment. Bees are known to forage up to 4.8 km for a total of 45 km2 around their hives (Winston 1987). While foraging, bees can contact various agrochemicals such as in secticides, fungicides, and herbicides. In some instances, the sudden death of bee colonies has been attributed to pesticide use, and could indicate misuse/overuse of pes ticides in a local area, non-point or point source environmental pesticide contamination, or any number of ot her environmental issues (Johansen and Mayer 1990). Consequently, it is important to understand how pesticides affect honey bee health in order to understand possible connections between pesticide exposure and honey bee health. Despite the importance of understanding pest icide effects on bees, toxicity studies typically focus on acute mortality and behavioral /physiological effects of pesticides on adult honey bees. Pesticide effects on immature honey bees (larva e/pupae) remain understudied (Emmett and Archer 1980, Davis 1989, Davis et al 2000, Aupinel et al. 2007). Therefore, I investigated 2 pesticides whose effect s on honey bee brood are not well understood. The first pesticide I tested was imidaclopr id, a neonicotinoid that affects nicotinic acetylcholine receptors in the insect nervous syst em (Nauen et al. 2001). Imidacloprid acts in the early stages of poisoning by agonizing acetylchol ine receptors, then in later stages by mild antagonism (Nauen et al. 2001). Imidacloprid, by volume, is one of the most abundant insecticides used globall y. It is used to protect crops fr om sucking pests such as aphids, mealybugs, leafhoppers, whiteflies, and thrips (Elbert et al. 1991, Cox 2001). Imidacloprids systemic properties allow it to be absorbed by plants and transporte d throughout the plants vascular tissue, phloem and xylem (Bennett 195 7, Sicbaldi et al. 1997). An example of
28 imidacloprid use is for control of the Asian citrus psyllid ( Diaphorina citri Kuwayama, Homoptera: Psyllidae), which is responsible for vectoring citrus greening disease (GraftonCardwell et al. 2006). Honey bee foragers that vi sit citrus blooms have the potential to be exposed to imidacloprid. In France, imidacloprid was t hought to be linked to massive bee declines occurring from 1997-2000, which led to the ban of the pest icide in 1999 (Oldroyd 2007, Stokstad 2007). Researchers have not found direct evidence linking imidacloprid to honey bee declines in France. The second pesticide I tested was amitraz. Amitraz is an octopamine agonist that stimulates intracellular second messenger si gnaling (Evans and Gee 1980, Hollingworth and Murdock 1980) within invertebrates. Typically, amitraz is used as an acaricide, but it can be used as an insecticide in other pest management progr ams (Yu 2008). Amitraz is of particular interest because it is used by beekeepers in Europe to co ntrol varroa (Marchetti et al. 1984). The harmful effects of amitraz on bees often are overlooked in the U.S. possibly because amitraz is not registered here for use within bee hives. There is a lack of evidence suggesting that amitraz persists in the beeswax of treated hives (Wallner 1999, Martel et al. 2007, Rial-Otero et al. 2007); however, its metabolite, 2, 4dimethylphenylformamide (DMPF), has been det ected in commercial and recycled beeswax from Spain and France with amounts ranging fr om 0.5 to 35 mg/kg, or 500 to 35000 ppb (Korta et al. 2003). Additional research focusing on the detection of DMPF within bee colonies is underway (Frazier et al. 2008). Amitraz LD50s ha ve been determined only for adult bees (Santiago et al. 2000). Consequently, there is a ne ed to understand the direct effects of amitraz on bee brood as no such information exists.
29 I present data here resulting fr om my investigations into le thal and sublethal effects of imidacloprid and amitraz on developing honey bee brood when pesticides are delivered subchronically. Chronic effects re sult from longer term, slower exposure to small doses of pesticides over an organisms entire lifespan, while subchronic effects represent exposure only during part of an organisms life. Both exposure types are biologically relevant for honey bees because adult and larval bees are known to be exposed to pesticides (Davis 1989, Davis et al. 2000, Chauzat et al. 2006). Even though doses en countered in the field are small-6 ppb for imidacloprid found in pollen collected from foraging bees returning to the hive (Chauzat et al. 2006)chronic effects can occur at doses 1/60th of those eliciting acute effects (Decourtye et al. 2005). I hypothesize that doses required to elicit leth al effects on bee brood are the same as those that lethally affect adult bees. Therefore, the doses I chose to test, 5-80 ppb for imidacloprid and 50-400 ppb for amitraz, are below the range of acut e, lethal doses (LD50s) for imidacloprid and amitraz of 192 and 255 ppb (respectively) found in prev ious studies on adult bees (Santiago et al. 2000, Fischer and Chalmers 2007). For imidacloprid, I also chose doses that are relatively small (5-20 ppb) to represent levels (2 -16 ppb) that bees may encounter in their environment (Chauzat et al. 2006, Krischik et al. 2007). For amitraz, I included doses above the known LD50 for adult bees as well as lower doses so that I could te st concentrations believe d to be sublethal. I hypothesized that bee brood would be as vulnerable to imidaclopr id and amitraz as adult bees, since both adult and immature bees contain nico tinic acetylcholine and octopamine receptors, the targets of imidacloprid and amitraz re spectively (Graham 1992, Chapman 1998). I determined mortality by comparing data between chemically treated and untreated/acetone treated (control) larvae. I measured a number of variables to identify potential
30 sublethal effects of these xenobiotics on developing brood. I selected these variables because they were relatively easy to quantify in the laboratory without interrupting or preventing the individual bees development. Materials and Methods General Rearing Procedure I collected h oney bee larvae from 8 commerci al honey bee colonies maintained at the University of Floridas bee biology unit (Gai nesville, FL; N 29 37.632' W 082 21.402'). To produce the test larvae, I confined the queen in each colony to a newly-drawn comb (no stored honey, pollen, or brood) using 2 di fferent sized zinc queen excluder cages (96 106 21 mm or 165 130 24 mm, l w h) at time t = -12 h. I then returned the caged queen and frame to the center of the brood nest (Boot and Calis 1991). During this time, worker bees were able to access and tend the queen. After 24 h, t = 12 h (Peng et al. 1992, Aupinel et al. 2005), I removed the queen from the cage and replaced the cage on the comb as before but this time for 108 h (from t = 0) to allow the eggs to hatch and larvae to re ach an appropriate age for grafti ng. During this time, worker bees were able to access the comb to feed the devel oping larvae. At 108 h, I removed the test frames (now containing 36 12 h old larvae) from the colonies and took them to the laboratory (Figure 2-1). Each queen was confined to a single frame. As such, I grafted 42 larvae from a single frame (1 queen source) into a 96-well treated ti ssue culture plate (7 9 mm, BD Biosciences, Durham, NC), placing 1 larva per well. These tissu e culture plates were used because they are the approximate size of a worker honey bee cell (5.2 mm diameter, Winston 1987). I repeated this procedure for each queen source (n = 8 queens 42 larvae), using separate well plates for each queens offspring. Prior to grafting the larvae in to plates, I pipetted 20 L of larval diet into
31 the bottom of each cell. The diet had a pH that ranged from 4.0-4.5 and consisted of 50% royal jelly (Glory Bee Foods, Eugene, OR), 6% D-gluc ose (Fischer Chemical, Fair Lawn, NJ), 6% Dfructose (Fischer Chemical, Fair Lawn, NJ), 37% double distilled water, and 1% yeast extract (Bacto, Sparks, MD) by volume (Vandenberg and Shimanuki 1987). Prior to adding the diet to each cell, I pre-warmed it to 35oC in an incubator (Percival Sc ientific Inc, Perry, IA). Each subsequent day, I transfe rred larvae to a clean culture plate provisioned with fresh diet. The amount of artificial diet provided to ea ch larva depended on the larvas age. I fed larvae 20 L of diet at hours 108 and 132, 30 L on hour 156, 40 L on hour 180, and 50 L on hour 204 and thereafter (Aupinel et al. 2007). At 204 h post oviposition (larvae are 132 12 h old), I transferred the larvae to a 48-we ll plate (Becton Dickinson Labwar e, Franklin Lakes, NJ, wells were 13 17 mm) because the growing larvae we re too large to handle delicately in a 96-well plate. Throughout the study, trays containing larvae were incubated in the dark at 35oC and ~96% RH (Vandenberg and Shimanuki 1987). Honey bee larvae defecate only once after th ey finish feeding just prior to pupation (Winston 1987). The feces appear as orange/brown streaks in the brood food. Once the larvae defecated (216-240 h from t = 0, Figure 2-1), I move d them individually into clean tissue culture plates, covered them with the plate lid, and allo wed them to pupate. Pupal rearing plates were incubated in the dark at 35oC and ~70% RH (Vandenberg and Shimanuki 1987) until each immature bee reached adulthood or died. Pesticide Application I initia lly was interested in investigati ng the effects of imidacloprid and amitraz on developing bees if a xenobiotic was ingested rather than received topically. I conducted
32 preliminary experiments using both topical and oral applic ations (Appendices A and B); however, after reviewing the preliminary data, I elected to apply the acaricide to the diet. I diluted imidacloprid and amitraz in an acetone solvent and then mixed the stock solutions into the larval diet each day the diet was administered. This was done to limit potential amitraz degradation in the acidic diet (pH betw een 4.0-4.5). I created diets with 5, 10, 20, 40, or 80 ppb for imidacloprid and 25, 50, 100, 200, or 400 ppb for amitraz (ppb per the amount of food given to the larvae). I provided one of the re sulting diets to each larva 4 times (at 132, 156, 180, and 204 h, Figure 2-1) during development. For bot h pesticides, I included 2 controls, one of acetone and a second with no acetone or pesticide application. For imidacloprid, n = 8 queens 7 treatments (5 chemical doses and 2 controls ) 6 larvae/treatment, while for amitraz, n = 7 queens 7 treatments (5 chemical doses and 2 controls) 6 larvae/treat ment. Offspring from one queen source did not contribute an acetone treat ment for either pesticide, which reduced the acetone sample size to 6. Mortality was assed by measuring (1) number of larvae that defecated (expressed relative to the total number of treated la rvae), (2) pupal survival (expressed relative to the total number of larvae that defecated), and (3) adult bee emer gence (expressed relative to the total number of larvae treated). To measure sublethal effects, I measured the (4) time it took each larvae to defecate starting from oviposition (time to larval defecation), (5) larval we ight immediately after defecation (larval weight at def ecation), (6) time it took larvae to emerge as adults starting from oviposition (time to adult emergence), (7) amount each adult bee weighed just after emerging (adult bee weight) and (8) amount each adult bee head weighed immediately after bee emergence (adult bee head weight). I used an electronic, top-loading-bench scale with a draftshield to measure bee weights (Mettler-Toledo Internatio nal Inc, Columbus, OH). To determine larval
33 mortality, I viewed each larva daily under a dissec ting microscope (Fisher Stereomaster, Fisher Scientific, Pittsburgh, PA) to look for movement If no movement was observed, I touched the larvae lightly with a grafting tool to encouragement movement. A larva was considered dead if it did not respond to 2 touches. Statistical Analysis Data for percentage of larvae th at defecated (imidacloprid and amitraz), pupal survival (amitraz), and adult emergence (amitraz) were analyzed using the Kruskal-Wallis test for nonparametric data (Zar 1996) because the data were not normally distributed. Pupal survival (imidacloprid) and adult emergence (imidacloprid) data were normal and analyzed using a 1-way ANOVA with treatment serving as a main effect. Normality was assessed for all data using a Shapiro-Wilk test. Proportion data shown to be normally distributed were transformed prior to analyses using an arcsine square root transf ormation. However, the pr oportion data reported herein are the untransformed means. Where n ecessary, means were compared using Fishers LSD tests. Homogeneity among variances was determined with Levenes test. In order to compare mortality between larvae treated with any level of pesticide or no pesticide, I pooled the treatment data within a pesticide group. For example, to create the pesticide treatment group for imid acloprid, I pooled the data from larvae treated with all levels of imidacloprid (all ppb con centrations) and compared it to p ooled data from larvae not treated with imidacloprid (untreated + acetone trea ted controls) using a 1-way ANOVA. Within treatment (pesticide or control), pooled means were grouped according to the stage (larva or pupa) in which the bee brood died in order to determine if pesticide-induced mortality was higher in the larval or pupal stage. I analyzed the data using a 1-way ANOVA. Data for time to larval defecation, larval we ight at defecation, time to adult emergence, adult bee weight, and adult bee head weight we re analyzed for both pe sticides using a 2-way
34 ANOVA recognizing treatment and queen source as main effects and treatment queen source as the interaction term. The main effects were tested against tr eatment queen source. I found a significant interaction for time to larval defecati on (imidacloprid and amitraz) and larval weight at defecation (imidacloprid). Consequently, I an alyzed these variables by queen source. Where necessary, means were compared using Fisher s LSD tests and differences accepted at 0.05. All statistical tests were conducted using SAS (SAS Institute, 2008). Results Imidacloprid Im idacloprid mortality data are reported in Table 2-1. Treatment did not affect the percentage of larvae that defecated ( 2 = 3.2; df = 6; P = 0.78) or pupal survival ( F = 10.55; df = 6, 47; P = 0.08). There was, however, a treatment e ffect on adult bee emergence (F = 2.80; df = 6, 47; P = 0.02). Larvae fed a diet containing 5, 10, 40, or 80 ppb imidacloprid had lower rates of adult emergence when compared to control and acetone treated larvae. Data were grouped within treatment type (untreated = control + acetone data; imidacloprid treated = 5 + 10 + 20 + 40 + 80 ppb data) according to bee developmental stage (larva or pupa) to compare mortality between larvae and pupae within imidacloprid treated and untreated groups. All data presen ted here and throughout are summari zed as mean std.error (n). Mortality rates were statistically similar ( F = 3.20; df = 1, 26; P = 0.08) for untreated pupae (24.5 6.9 (14) % mortality) and untreated larvae (8.6 2.4 (14) % mortality). In contrast, pupae developing from imidacloprid treated larvae (55.2 4.8 (40) % mortality) experienced significantly higher mortality than did imidaclopr id treated larvae (16.4 3.1 (40) % mortality) ( F = 42.65; 1, 78; P < 0.01). Mortality among imidacloprid treated and untreated larvae was similar ( F = 1.31; 1, 52; P = 0.26). In contrast, pupae developing from imidacloprid treated
35 larvae experienced significantly higher mortality than those in the untreated group ( F = 11.64; 1, 52; P < 0.01). I did not find sublethal effects of imidaclopr id on developing bees for time to larval defecation (Table 2-2), larval weight at defecat ion (Table 2-3), time to adult emergence, adult bee weight, or adult bee head weight (Table 2-4). There was a significant interaction between treatment and queen source for m ean time to larval defecation ( F = 1.44; df = 40, 216; P = 0.05), thus, I analyzed treatment by queen source for th is variable (Table 2-2). Treatments differed significantly only among offspring from queen sources 106 and 112; however, no clear trends were apparent. There was a significant interaction between treatme nt and queen source on larval weight at defecation ( F = 2.01; df = 40, 215; P < 0.01), thus, I analyzed treatm ent by queen source for this variable. There were no signi ficant treatment effects for queen sources 104, 110, 106, 101, 105, and 112 (Table 2-3). There were treatment e ffects for queen sources 108 and 115, but again, no clear trends were apparent. There was no significant treatment queen source interaction ( F = 1.42; df = 33, 95; P = 0.09) or treatment effect ( F = 0.99; df = 6, 95; P = 0.43) on the mean time to adult emergence (Table 2-4). There was a queen source effect ( F = 8.03; df = 7, 95; P < 0.01) on this variable. Offspring from queen sources 104 (511.2 2.4 (13) h), 108 (501.6 2.4 (22) h), 110 (504.0 2.4 (10) h), and 115 (508.8 2.4 (2 2) h) took longer to develop than those from other queen sources (482.4 2.4 (21) 496.8 2.4 (26) h). There was no significant treatment queen source interaction ( F = 0.47; df = 25, 76; P = 0.98) or treatment effect ( F = 1.21; df = 6, 76; P = 0.31) on adult bee weight. There was a significant queen source effect on adult bee weight ( F = 6.07; df = 5, 76; P < 0.01). Adult
36 progeny from queen source 105 (87 2 (20) mg) we ighed significantly less than progeny from the other queen sources (95 3 (9) 104 2 (25) mg). Furthermore, there was no significant treatment queen source interaction ( F = 0.84; df = 33, 94; P = 0.70), treatment effect ( F = 0.88; df = 6, 94; P = 0.51), or queen source effect ( F = 1.51; df = 7, 94; P = 0.17) on mean adult bee head weight (Table 2-4). Amitraz Am itraz mortality data are reported in Table 2-5. Treatment of larvae with amitraz had an effect on the percentage of larvae that defecated ( 2 = 20.12; df = 6; P < 0.01), pupal survival (2 = 17.76; df = 6; P < 0.01), and adult bee emergence ( 2 = 20.85; df = 6; P < 0.01). Larvae treated with 25 and 400 ppb of amitraz were less likely to defecate than were the untreated and acetone controls. Larvae fed 50, 200, and 400 ppb were less likely to pupate than were untreated and acetone controls. Larvae treated w ith any of the tested amitraz c oncentrations (25-400 ppb) were less likely to emerge as adults (Table 2-5). To compare mortality between larvae and pupae within amitraz treated and untreated groups, data were further grouped within treatmen t type (untreated = co ntrol + acetone data; amitraz treated = 25 + 50 + 100 + 200 + 400 ppb da ta) according to bee development stage (larva or pupa). Pupae not treated as larvae experienced significantly higher mortality (38.7 8.2 (13) % mortality) than untreated larvae (5.1 2.2 (13) % mortality) ( F = 15.96; df = 1, 24; P < 0.01). Likewise, pupae treated as larvae with amitr az experienced significantly higher mortality (70.3 4.5% (35) mortality) than treated larvae (25.4 4.3 (35) % mortality) ( F = 40.49; df = 1, 68; P < 0.01). Treated larvae had significantly higher mortality than untreated larvae ( F = 8.84; df = 1, 46; P < 0.01) and pupae treated as larvae had si gnificantly higher mortality than pupae not treated as larvae ( F = 9.35; df = 1, 46; P < 0.01).
37 I did not find sublethal effects of amitraz on developing bees for time to larval defection (Table 2-6), larval weight at defecation, time to adult emergence, adult bee weight, and adult bee head weight (Table 2-7). There was a significan t interaction between queen source and treatment on time to larval defecation ( F = 1.52; df = 34, 175; P < 0.04), so I analyzed the data by queen source. Despite this, no treatment related trends within queen source were apparent (Table 2-6). There was no significant queen s ource treatment interaction ( F = 1.28; df = 34, 171; P = 0.15) or treatment effect ( F = 1.72; df = 6, 171; P = 0.12) on larval weight at defecation, while there was an effect of queen source on this variable ( F = 11.54; df = 6, 171; P < 0.01). Progeny from queen source 104 (112 3 mg) weighed less at defecation than larvae from other queen sources (135 3 to 151 4 mg). In contrast progeny from queen source 108 (151 4 mg) weighed more than larvae from other queen sources (112 3 143 3 mg) (Table 2-7). There was no significant queen s ource treatment interaction ( F = 0.89; df = 25, 52; P = 0.62) or treatment effect ( F = 1.01; df = 6, 52; P = 0.43) on time to adult emergence (Table 2-7). I did find that offspring from queen source 110 (504.0 4.8 (6) h) took longer to reach adulthood than offspring from all other queen sources (482.4 2.4 501.6 4.8) ( F = 3.02; df = 6, 52; P 0.01). There was no significant interaction ( F = 0.33; df = 19, 43; P = 0.99) or treatment effect ( F = 0.17; df = 6, 43; P = 0.98) on adult bee weight. However, qu een source did affect this variable with offspring from queen 106 (109 3 (20) mg) weighing more than offspring from all other queens (95 1 99 2 mg) ( F = 2.91; df = 4, 43; P < 0.03). Furthermore, I did not find a significant queen source treatment interaction ( F = 0.36; df = 25, 51; P = 0.99), treatment effect ( F = 0.45; df = 6, 51; P = 0.84) or queen source effect (F = 0.78; df = 6, 51; P = 0.59) on adult bee head weight (Table 2-7).
38 Discussion In general, control (untreated + acetone treated) larvae in the imidacloprid and amitraz tests had reduced adult emergence rates (imidacl oprid test, 70%; amitraz test, 59%) compared to worker emergence rates typically observed under natural hive condi tions (90-97%, Winston et al. 1981). My control results are simila r to those of Aupinel et al. (2005) who showed that 90% of control larvae survived to defecate and 66% surv ived to adulthood. The relatively low number of adult bees to emerge under control conditions may be an indication of the amount of stress encountered by the immature bees during the artifi cial rearing procedure. This stress may have led to an enhanced effect, infl uencing the toxicity of the pest icide during pupal development. If stress caused enhanced effects of pesticides on bee development, then bees under more natural, unstressed conditions may experience less mortality at the tested pesticide doses. My results may suggest greater pesticide sus ceptibility in stressed bees. Imidacloprid Collectively, the data suggest that im idacloprid may have a delayed, lethal effect on honey bee pupae when larvae are fed low doses of imidacloprid under artificial c onditions. In general, the only measurable effect was a reduced number of treated larvae that were able to develop into adults, which was lower in larvae treated with 5, 10, 40, and 80 ppb imidacloprid than in control larvae. The data presented here suggest that imidacloprid doses of 5, 10, 40, & 80 ppb, which are lower than the adu lt bee LD50 of 192 ppb (Fischer and Chalmers 2007) are lethal to brood. It generally is believed that the larval and pupal stages of insects are more susceptible to xenobiotics (Yu 2008). Eggs and pupae are contai ned in a protective casing and do not feed (Gullan and Cranston 2005), which likely result s in reduced xenobiotic exposure. In my investigation, the larvae were fed the xenobiotic, but the effect s were delayed until the pupal stage.
39 The results for imidacloprid were not dose depe ndent, meaning mortality did not increase as the tested pesticide dose increased. For exampl e, I did not see a signif icant reduction in adult emergence at 20 ppb, but did for all other doses (5, 10, 40, and 80 ppb). Such an atypical distribution of mortality data are not reported in the literature because most data for determining LD50s are summarized graphically or only reported after prob it analysis (Landis and Yu 1995, Yu 2008). Therefore, small discrepancies in doseresponse data typically are not usually reported in the literature. Data further from the LD50 tend to wobble about the mean. Even though my data did not suggest dose dependency, this is not unexpected since greater variation occurs in data far from the LD50 (Suchail et al. 2000, Bra ndt et al. 2002). To determine dose dependency, I should test more imidacloprid doses to include a wider range and analyze the resulting data using probit analysis (Yu 2008). The pooled data suggest that imidacloprid, though fed to developing larvae, was more lethal several days later to developing pupae. There are no known reasons for this delayed toxicity to occur, but bee larvae store their waste until defecation at which time metamorphosis begins (Winston 1987, Chapman 1998). It is possible that imidacloprid is more toxic during this metabolically active period because at this tim e, larval tissues are being broken down and reabsorbed (Winston 1987). The change from prepupa to pupae and then to adult is also an active period of hormone signaling. During pupal developmen t, juvenile hormone is absent for the first time, which signals a change from an immature stage to the imago form (Chapman 1998, Gullan and Cranston 2005). At this time, there is also activity for ecdysone, eclosion hormone, and bursicon, which are all hormones that are associated with molting (Chapman 1998). Detoxification enzymes such as cytochrome P-450 monooxygenases (r eview by Feyereisen 1999, 2005) could be shunted to other metabo lically important functions preventing
40 detoxification and clearance of xenobiotics stored within larvae. This could be particularly important with respect to honey bees because of their lack of xenobiotic detoxifying enzymes, such as cytochrome P-450 monooxygenases, when compared to other ins ects (Claudianos et al. 2006). Imidacloprid did not elicit measurable sublet hal or acutely lethal effects on developing honey bees at the doses I teste d. I did not find that imidacloprid affected the rate at which immature bees develop even though it has been s hown to affect the develo pmental rates of other insects. Abbott et al. (2008) found effects of imidacloprid on Osmia lignaria larvae, a solitary bee. They concluded that acute -sublethal doses (30 and 300 ppb) of imidacloprid mixed with pollen increased female larval developmental tim e (Abbott et al. 2008). To gain a more accurate assessment of the effects of imidacloprid on larv al and pupal development rates, I should have measured larval mortality several times throughout each day. Other sublethal effects could result from larvae eating food containing pesticides. For example, imidacloprid did reduce the number of larvae that reached adulthood, but some larvae still successfully pupated. Consequently, a number of larvae survived the tr eatment. It is possible that sublethal effects could manifest themselves in the surviving adults. These effects could include: disorientation, decrea sed learning ability, decrease d longevity, decreased sperm viability, among other possi bilities (Bitterman et al. 1983, Bortolot ti et al. 2003, Medrzycki et al. 2003, Decourtye et al. 2004, Decourtye et al. 2005, Burley 2007), suggesting that the surviving adults could have been compromi sed in ways that were not considered in the current research. Bee larvae in this study were affected by imid acloprid at doses that could be encountered in the environment. For example, Chauzat et al. (2006) found imidacloprid to be the most frequent pesticide found in pollen brought back to the nest by forager bees with concentrations
41 ranging from 1.1 5.7 g/kg (ppb). Therefore, th e effects I found could occur in managed bee colonies. Results of this investigati on suggest that risk assessment procedures for registering systemic pesticides be modified to include lethal effects on honey bee brood. Findings of the current study suggest that lethal levels of these xenobiotics may be lower than those lethal to adult bees. Therefore, current procedures may be underestimating pesticide toxicity to bees. Warnings have been placed on imidacloprid limiti ng its use during periods of flower bloom. As stated earlier, imidacloprids long residual time in soil (Bonma tin et al. 2003) makes a short restriction of the pesticide non-beneficial be cause plants have been shown to acquire imidacloprid from the soil up to 2 years after treatment. Management practices limiting the use of systemic pesticides within areas important to bee pollination could be adopted. This remains a difficult task due to the large foraging range of honey bees (Winston 1987) and our heavy reliance on pesticides (annual global pesticide sales accumulate to ~33.6 billion USD; (Yu 2008). Amitraz As noted with im idacloprid, toxicity data collected on amitraz-fed larvae indicated that amitraz present in brood food from 25 to 400 ppb re duced larval survival to adulthood. Santiago et al. (2000) found that the LD50 for amitraz on adults bees is around 255 ppb. I did not test this level specifically, but I did br acket this dose by testing 200 and 400 ppb. I did not find amitraz effects on larval mortality at 200 ppb but did at 400 ppb; a lethal concentration similar to that observed previously for adu lts (Santiago et al. 2000). Amitraz showed a more dose-dependent response compared to imidacloprid data for the percentage of treated larvae to emerge as adults. This may have occurred since I tested a larger range of doses for amitraz than I did for imidacloprid.
42 When I pooled the amitraz treatments together according to developmental stage (larvae vs. pupae), I found that pupae fed amitraz as larvae had higher mortality than those fed as larvae on food containing amitraz. However, there was also an increase in mortality in the control groups, suggesting that variables other than th e xenobiotic influenced pupal mortality. Since mortality increased in both untreated and trea ted groups, I specifically cannot conclude that mortality of amitraz-fed larvae is more likely to occur during pupation rather than while feeding in the larval stage. Amitraz did not elicit measurable sublethal eff ects on immature bees in this investigation. However, amitraz may affect developing bees in ways not measured here. For instance, foramidines have been shown to cause d ecreased fecundity in the tobacco budworm, Heliothis virescens (F.); fall armyworm, Spodoptera frugiperda (J. E. Smith); southern armyworm, S. eriddania (Cramer); and pink bollworm, Pectinophora gossypiella (Saunders) (Wolfenbarger et al. 1974) as well as hyperexcitat ion and sterilization in tick s (Hollingworth 1976). I did not examine these variables. It is interesting to not e that amitraz did not affect larval weight at defecation despite the fact that it has been show n to be a feeding deterrent in mites (Gladney et al. 1974) and lepidopteran larv ae (Doane and Dunbar 1973). Even though varroa mites can be lethal to bee colonies (DeJong 1997) and difficult to control due to acaricide resistan ce (Hillesheim et al. 1996, Spreafic o et al. 2001); my data do not suggest that amitraz is an appropr iate acaricide for varroa control. Summary Im idacloprid and amitraz were lethal to immatu re bees, limiting the number of bees that emerged as adults. Unlike for imidacloprid treated larvae, mortality during the pupal stage was dose-dependent for amitraz treated larvae. Imidacloprid and amitraz were lethal to immature bees, lowering the percentage of adult bees that emerged. Unlike for pupae mortality for larvae
43 treated with imidacloprid, pupae mortality for la rvae treated with amitraz appeared to be dosedependent. This difference could be explained by the wider range of amitraz doses I chose to test. A lack of dose dependency for imidacloprid may be explained by the idea that mortality responses at doses relatively cl ose together tend to "wobble" about the mean, especially at the areas on the extremes of the LD50 curve. No acutely lethal effects were observed for imidacloprid treated larvae. However, I did find an increase in larval mo rtality occurring for amitraz treate d larvae, but mortality did not occur at all tested doses. Several reasons could lead to a difference in larval mortality. First, both chemicals have different physico-chemical properties; for instance, amitraz is lipophilic while imidacloprid is hydrophilic (Hollingworth 1976, Elbe rt et al. 1991). Secondly, imidacloprid and amitraz act on different receptors in the insect nervous system: acetylcholine and octopamine respectively (Yu 2008). Finally, I te sted higher amitraz doses than imidacloprid ones in order to remain close to the adult LD50 values. For imidacloprid, toxicity appeared to be delayed, mostly occurring during pupal development. The same could not be concluded for amitraz because there were acutely lethal effects on larvae. Delayed toxicity for amitraz could not be concluded because both untreated and treated larvae displayed increases in pupae mortality over that of larval mortality. This research could have been improved by m odifying methods to increase the percentage of adult bees to emerge under control, untreat ed conditions. This could be achieved by placing the prepupae laterally, and abdom en-first into smaller cells, thereby mimicking their natural orientation in the hive. Another modification I would have chosen to make would be to record the stage of larval or pupal de velopment the treated bees reac hed prior to dying. Davis et al. (2000) showed that dimethoate a nd carbofuran treated larvae faile d to spin silk, suggesting that
44 larvae would fail to molt from the prepupal to th e pupal stage. In my study, this clearly was not the case as some larvae treated with either pesticide completed the prepupal stage. Overall, the xenobiotics tested here appear to be detrimental to honey bee brood. Based on my findings I recommend that risk assessment te sts on honey bees be changed to include effects on larval mortality and delayed effects on pupae when fed as larvae.
45 Table 2-1. Percentage of larvae that defecated, pupal survival, and adult bee emergence rates for larvae fed a diet containing imidacloprid. Treatment Larvae that defecateda (%) Pupal survival b (%) Adult bee emergencec (%) Theoretical survival na na 93.5 2.2 Untreated 91.2 3.3 (8)a76.7 10 (8)a 70 9.2 (8)a Acetone 91.7 3.7 (6)a73.9 10 (6)a 69.4 11.7 (6)a 5 80 6.8 (8)a50.6 8.5 (8)a 38.3 5.2 (8)b 10 88.3 7.4 (8)a37.9 8.3 (8)a 32.9 8.4 (8)b 20 83.7 7.6 (8)a51.7 12.2 (8)a46.7 12 (8)ab 40 81.7 7.3 (8)a43.5 11.5 (8)a 37.5 11.8 (8)b 80 84.2 6.9 (8)a40.4 13.7 (8)a 30 9.4 (8)b Theoretical survival is the pr edicted number of eggs to su rvive to adulthood under natural conditions within the hive (Win ston et al. 1981). Imidacloprid c oncentrations are shown in ppb. Data are mean std. error (n) where n = number of queen sources whose offspring were tested. Columnar means followed by the same letter are not different at 0.05. Means were compared using Fishers LSD tests. a % of larvae that defecated = (the number of larvae that defecated/the total number of treat ed larvae) 100. b % pupal survival = (the number of adult bees that emerged/the number of larvae that defecated) 100. c % of adult bee emergence = (the number of adult bees that em erged/the total number of treated larvae) 100.
46 Table 2-2. Mean time to larval defecation in ho urs for larvae fed diet containing imidacloprid. Queen source Treatment 101 104 105 106 Untreated 244.8 9.6 (5)a240 (6)216 (5)a 228 4.8 (6)a Acetone 230.4 4.8 (5)a na 220.8 4.8 (5)a 240 (6)ac 5 232.8 4.8 (6)a 240 (5) 235.2 4.8 (5)a 240 (6)a 10 235.2 4.8 (6)a 240 (5) 232.8 7.2 (6)a 240 (6)a 20 232.8 4.8 (4)a 240 (6) 228 7.2 (4)a 235.2 7.2 (6)a 40 240 7.2 (5)a 240 (4) 235.2 4.8 (5)a 254.4 4.8 (5)b 80 235.2 4.8 (5)a 240 (6) 230.4 4.8 (5)a244.8 4.8 (6)bc ANOVA F = 0.64; df = 6, 29; P = 0.70 F = na; df = 5, 26; P = na F = 1.65; df = 6, 28; P = 0.17 F = 3.23; df = 6, 34; P < 0.01 Data are separated for treatment by queen source. Imidacloprid c oncentrations are shown in ppb. Da ta are mean std. error (n) where n = number of queen offspring receiving a given treatment. Columnar means followed by the same letter are not different at 0.05. Means were compared using Fishers LSD tests.
47 Table 2-2. Continued Treatment 108 110 112 115 Untreated 240 (6)a 220.8 4.8 (4)a 230.4 9.6 (5)a 240 (5)a Acetone 240 (5)a na 235.2 4.8 (6)a 240 (5)a 5 240 (4)a 228 12 (2)a 240 (5)a 249.6 4.8 (5)a 10 240 9.6 (5)a 228 12 (2)a 240 (6)a 264 9.6 (6)a 20 244.8 4.8 (6)a 228 12 (2)a 244.8 4.8 (6)ac 244.8 4.8 (5)a 40 240 (6)a 240 (2)a 235.2 4.8 (6)a 244.8 4.8 (4)a 80 240 (5)a 228 12 (2)a256.8 4.8 (6)bc 244.8 4.8 (5)a ANOVA F = 1.90; df = 6, 30; P = 0.11 F = 0.44; df = 5, 8; P = 0.81 F = 3.08; df = 6, 33; P < 0.02 F = 1.95; df = 6, 28; P = 0.11
48 Table 2-3. Mean larval weight at defecation for larvae fed a diet containing imidacloprid. Queen source Treatment 101 104 105 106 untreated 131 4 (5)a124 3 (6)a116 3 (5)a133 6 (6)a acetone 133 8 (5)a na 116 7 (5)a 140 5 (6)a 5 142 10 (6)a 118 4 (5)a 123 2 (5)a 137 3 (6)a 10 152 7 (6)a 119 6 (5)a 129 5 (6)a 129 4 (6)a 20 154 11 (4)a 99 9 (6)a 137 11 (4)a 139 10 (6)a 40 150 13 (5)a 100 17 (4)a 138 4 (5)a 140 12 (5)a 80 153 16 (5)a 100 15 (6)a 131 9 (5)a 147 4 (6)a ANOVA F = 0.82; df = 6, 29; P = 0.56 F = 1.35; df = 5, 26; P = 0.28 F = 2.09; df = 6, 28; P = 0.09 F = 0.67; df = 6, 34; P = 0.68 Data are separated for treatment by queen s ource. Defecation weights are given in millig rams. Imidacloprid concentrations are s hown in ppb. Data are mean std. error (n) where n = number of queen offspring receiving a given trea tment. Columnar means followed by the same letter are not different at 0.05. Means were compared us ing Fishers LSD tests.
49 Table 2-3. Continued Treatment 108 110 112 115 Untreated 138 10 (6)ac 166 4 (4)a 126 3 (5)a 123 4 (5)a Acetone 147 4 (5)ac na 129 4 (6)ab 152 8 (5)bc 5 164 4 (4)bc 148 7 (2)a 145 6 (5)bc 142 4 (5)ac 10 150 13 (5)ac 116 2 (2)a 155 4 (6)cd 147 10 (5)bc 20 160 6 (6)b 118 63 (2)a 167 8 (6)de 148 7 (5)bc 40 180 6 (6)b 154 4 (2)a 163 9 (6)ce 158 3 (4)bc 80 189 2 (5)d 163 2 (2)a 152 2 (6)ce 147 8 (5)bc ANOVA F = 5.58; df = 6, 30; P < 0.01 F = 1.07; df = 5, 8; P = 0.45 F = 6.46; df = 6, 33; P 0.01 F = 2.40; df = 6, 27; P = 0.05
50 Table 2-4. Mean time to adult emergence, adult bee weight, and adult bee head weight for larvae fed a diet containing imidacloprid. Treatment Time to adult emergence (h) Adult bee weighta (mg) Adult head weighta (mg) Untreated 494.4 2.4 (32)a 94 2 (20)a11.6 0.2 (32)a Acetone 494.4 2.4 (24)a 97 2 (23)a11.4 0.3 (23)a 5 499.2 4.8 (18)a 94 2 (14)a11.1 0.3 (18)a 10 499.2 4.8 (15)a 101 4 (12)a 12 0.4 (15)a 20 501.6 2.4 (22)a 101 3 (19)a11.7 0.3 (22)a 40 496.8 2.4 (17)a 99 3 (16)a11.6 0.3 (17)a 80 506.4 4.8 (14)a 101 1 (9)a 11.4 0.3 (14)a Imidacloprid concentrations are shown in ppb. Data are mean std. error (n) where n = number of individuals from all queen sources. Column ar means followed by the same letter are not different at 0.05. aMeasurements were taken just after adult bees emerged.
51 Table 2-5. Percentage of larvae that defecate d, pupal survival, and adult bee emergence for larvae fed a diet containing amitraz. Treatment Larvae that defecateda (%) Pupal survival b (%) Adult bee emergencec (%) Theoretical survival na na 93.5 2.2 Untreated 95.2 3.1 (7)a 60.5 10 (7)a 59 10.6 (7)a Acetone 97.2 2.8 (6)a 63.3 14.4 (6)a 61.1 14 (6)a 25 66.7 11.5 (7)bc47.5 11.9 (6)ab 30.9 8.5 (7)b 50 81.4 7.2 (7)ab 25.5 7.8 (7)bc 20 4.9 (7)b 100 81.4 11.3 (7)ab45.2 10.9 (7)ab 30.5 5.7 (7)b 200 80 7.1 (7)ab28.6 5.4 (7)bc 24.3 4.9 (7)b 400 56.2 7.2 (7)c 9.5 9.5 (7)bc 4.8 4.8 (7)c Theoretical survival is the pr edicted number of eggs to su rvive to adulthood under natural conditions within the hive (Winston et al. 1981 ). Amitraz concentrations are in ppb. Data are mean std. error (n) where n = number of queen sources used to produce offspring for each replicate. Columnar means followed by th e same letter are not different at 0.05. Means were compared using Fishers LSD tests. a % of larvae that defecated = (the number of larvae that defecated/the total number of treated larvae) 100. b % pupal survival = (the number of adult bees that emerged/the number of larvae that defecated) 100. c % of adult bee emergence = (the number of adult bees that emerged/th e total number of treated larvae) 100.
52 Table 2-6. Mean time to larval defecation (h ) for larvae fed a diet containing amitraz. Queen source Treatment 101 104 105 106 Untreated 240. (6)a 216 (5)a232.8 4.8 (6)a 240 (6)a Acetone 232.8 4.8 (4)a 216 (6)a235.2 4.8 (6)a 247.2 4.8 (6)ac 25 240 (5)a 220.8 4.8 (5)a240 7.2 (5)a 256.8 4.8 (4)bcd 50 240 (4)a 216 (5)a235.2 7.2 (6)a 240 (3)a 100 240 4.8 (6)a 216 (6)a228 4.8 (6)a 264 (5)bd 200 230.4 4.8 (5)a 216 (6)a240 7.2 (5)a 259.2 4.8 (5)b 400 232.8 12 (4)a 216 (3)a240 14.4 (4)a 264 (4)b ANOVA F = 0.56; df = 6, 27; P = 0.76 F = 1.04; df = 6, 29; P = 0.42 F = 0.43; df = 6, 31; P = 0.85 F = 7.96; df = 6, 26; P < 0.01 Treatment 108 110 112 Untreated 254.4 4.8 (5)a 10.2 4.8 (5)a 240 (6)a Acetone 247.2 4.8 (6)a na 232.8 4.8 (6)a 25 240 (4)a na 230.4 9.6 (5)a 50 252 4.8 (6)a 247.2 7.2 (3)a232.8 4.8 (6)a 100 256.8 4.8 (4)a 240 (1)a225.6 4.8 (5)a 200 256.8 7.2 (3)a 264 14.4 (3)a235.2 7.2 (6)a 400 247.2 7.2 (3)a 288 (1)a223.2 7.2 (3)a ANOVA F = 1.02; df = 6, 24; P = 0.43 F = 2.17; df = 4, 8; P = 0.16 F = 0.73; df = 6, 30; P = 0.63 Data are separated for treatment by queen source Amitraz concentrations are shown in ppb. Data are mean std. error (n) where n = number of queen source offspring receiving a given treatment. Columnar means followed by th e same letter are not different at 0.05. Means were compared using Fishers LSD tests.
53 Table 2-7. Mean larval weight at defecation, time to adult emerge nce, adult bee weight, and adult bee head weight for larvae fed a diet containing amitraz. Treatment Larval weight at defecation (mg) Time to adult emergence (h) Adult bee weighta (mg) Adult head weighta (mg) Untreated 132 3 (39)a 492 2.4 (24)a102 3 (18)a 11.5 0.3 (24)a Acetone 139 3 (34)a 492 2.4 (2 1)a101 2 (20)a 12.1 0.3 (21)a 25 139 5 (27)a 492 4.8 (13)a99 2 (11)a 11.9 0.3 (13)a 50 135 5 (33)a 480 (8)a96 4 (7)a 11.5 0.4 (8)a 100 141 5 (33)a 499.2 4.8 (12)a105 5 (9)a 11.6 0.4 (11)a 200 143 5 (33)a 492 4.8 (10)a100 4 (7)a 11.7 0.3 (10)a 400 140 6 (19)a 504 24.0 (2)a 98 (1)a 13.0 0.8 (2)a Amitraz concentrations are shown in ppb. Data ar e mean std. error (n) where n = number of individuals pooled from all queen sources. Columnar means followe d by the same letter are not different at 0.05. a Measurements were taken just after adult bees emerged.
54 024487296120144168192216240264288312336360384408432456480504528 egg laying egg development larval development pesticide treatment 1 pesticide treatment 2 pesticide treatment 3 pesticide treatment 4 larval defecation prepupal/pupal development adult emergence Figure 2-1. Laboratory bioassay: experimental timeline according to larval development in hours (0-528 h, x-axis). Pesticide treatment refers to the 4 time periods each larva was treated (a t 132, 156, 180, and 204 h) with their corresponding pesticide treatment (imidacloprid or amitraz). The timeline does not illustrate that egg lying began at t = -12 h.
55 CHAPTER 3 SUMMARY The research presented herein was conducted to gain insight into the effects of pesticides on imm ature stages of the honey bee, an importa nt beneficial insect. Honey bee survival is important not only for the production of certain commodities such as honey and beeswax, but also for purposes of crop pollination. It has been estimated that one-third of the worlds food production is dependent on bee pol lination (Pimentel 2005). Theref ore, any decline in honey bee populations potentially could a ffect our global food supply. Beginning in 1950s, investigators began to notice a decline in managed honey bee populations (NASS 1976-2008) Parasitic mites were a major c ontributor to past bee population declines (Morse and Flottum 1997). Reasons for the most recent declines are unknown, but some researchers have suggested pesticides as possible causes, including neonicotinoids and acaricides, (Stokstad 2007). Imidacloprid is a neonicotinoid and plant systemic insecticide that has been found in the pollen and nectar of plants visited by honey bees (Bonmatin et al. 2003, Chauzat et al. 2006, Greatti et al. 2006). Imidacloprid movement into the hive via foraging bees is the proposed mechanism by which honey bee larvae are exposed to plant systemic pesticides (Villa et al. 2000). To date, there has been no evidence to suggest th at imidacloprid is responsible for the widespread losses of bees. However, it is im portant to understand the potential environmental and non-target effects of imidacloprid because it is, by volume, one of the most abundant pesticides used in the wo rld (Kegley et al. 2008). Previous research into the e ffects of pesticides on honey bees has focused primarily on adult bees (Bitterman et al. 1983, Bortolotti et al. 2003, Decourtye et al. 2004, Decourtye et al. 2005), leading investigators to sugge st that quantities of imidaclopr id found to be present in the
56 environment are not lethal to ad ult bees (Schmuck et al. 2001, Maus et al. 2003, Stadler et al. 2003). Despite known effects of imidacloprid, there is a lack of research concerning its effects on honey bee brood. The results from my research (C hapter 2) suggest that small quantities of imidacloprid in the ppb range are in fact lethal to bee brood. These levels are lower than reported LD50s for adult worker bees (Suchail et al. 2000, Nauen et al. 2001, Schmuck et al. 2001, Suchail et al. 2001). Interestingly, the effects of imidacloprid on honey bee brood (presented in Chapter 2) were not dose-dependent at the doses I tested. In my investigation, imidacloprid killed developing brood at 5, 10, 40, and 80 ppb, but not at 20 ppb. Desp ite a lack of clear dose-dependent trends within the range I tested, the pe rcentage of adult emergence at 20 ppb was 34% lower than that of the controls. An increase in adult emergence at 20 ppb compared to 5 and 10 ppb does not insinuate a lack of dose-dependence. Mortality data typically are not reported in the literature, but are analyzed graphically using prob it analysis procedure that cove rs a wider range of doses to produce the LD50 values that reported more co mmonly (Yu 2008). The range of doses I tested for imidacloprid in Chapter 2 (5-80 ppb) may seem large but is not compared to ranges likely to be tested to determine dose dependency, which may cover a range from ppb to ppt (parts per thousand, Figure 3-1). Natural variation in bee re sponse to ppb doses close to one another (such as the ones tested in Chapter 2) are common and usually are overl ooked, especially if they occur on the extreme ends of the dose-response curve (i.e. 0-10%, 90-100%, Figure 3-2). In a toxicity experiment verifying LD50s for imidacloprid on adult honey bees, Suchail et. al. (2000) plotted the log dose (ng/bee) against the percentage of bee mortality. Their graphs illustrate a wobble, or non-linear response, for percentage of bee mort ality within a small rang e of doses, especially
57 on the lower and upper extremes of the response cu rve. However, across all tested doses, their graph depicts a generalized dose-dependent resp onse. Had I wanted to determine an LD50 for imidacloprid, it would be necessary to test doses ranges that in cluded doses > 192 ppb (Fischer and Chalmers 2007) and < 5 ppb. These doses probably would have bracketed the NOED (no observable effect dose). Once test ed, this should lead to a be tter understanding of LD50 values for honey bee brood, which, in turn, can be used for risk assessment. Even though I believe the results from my laborat ory investigations were reliable, I hesitate to extrapolate my findings too far with regard to imidacloprid effects on brood in field colonies. For example, the in vitro rearing program clearly hindered be e development, decreasing adult emergence to 77% in the controls. As such, the rearing program itself could be a stressor that may have influenced the observed imidacloprid eff ects in this study. Consequently, it is possible that imidacloprid may not produce similar effect s in field situations where brood rearing is optimized and background mortality is low (<5%; Winston et al. 1981) However, if my results are an accurate indication of imidacloprid eff ects under natural, stressed hive conditions, then larval exposure to imidacloprid at levels as low as 5 ppb could result in a 45% reduction in the amount of brood in exposed colonies and up to 57% at 80 ppb. Data for amitraz are equally interesting. Amitraz is not known to persist in treated honey bee colonies longer than 9 months post treatment (Lodesani et al. 1992, Korta et al. 2001) and was recorded to have degraded significantly with in days after being app lied to plates containing beeswax (Wallner 1999). Amitraz is thought to break down rapidly in acidic solutions such as honey, which is why it does not persist in hives. An amitraz metabolite, 2,4dimethylphenylformamide (DMPF), has been detected at high levels in beeswax, but research is still in progress to identify DMPF persistence within bee colonies (Korta et al. 2003, Frazier et
58 al. 2008). Although, amitraz may be broken down quickly; its DMPF metabolite (or other unknown metabolites) could have si milar effects on brood development. Future research should consist of measuring bee mortality when the be es are exposed to DMPF and amitraz stored in honey for several days or more. In my research, I found that amitraz is lethal to honey bee brood (larvae and pupae) at doses lower than those found to be lethal to adults (Santiago et al 2000). Consequently, beekeepers should not use this product as an acaric ide. I believe that amitraz is a poor choice for varroa control based on 3 reasons. First, amitr az metabolites can be hazardous to mammals including humans (Hollingworth 1976, Yu 2008). Secondly, varroa have deve loped resistance to amitraz (Spreafico et al. 2001). Finally, my data suggest that amitraz could contribute to significant brood/pupa bee losses if concentrations ranging from 25-400 ppb are present in larval food. Larvae exposed to 25 ppb of amitraz in their food could result in an estimated 50% loss of bee brood, whereas larvae exposed to 400 ppb coul d result in an estimated 92% loss of brood. Given more time, I would like to continue my investigations in field colonies. However, my initial field trials did not lead to conclusive results. This was due to a number of problems I encountered while administering pe sticides to larvae in field col onies. I tried to administer the pesticide to larvae in two ways: via treated diet and by topical applica tion. First, larvae died when I tried to administer small amounts of pestic ide treated diet into thei r cells. I believe that the larvae drowned because the applied volume wa s too large (this method was not discussed in length). The biology of honey bees makes it difficult to deliver th e correct amount of artificial diet to individual larva because bee larvae naturally are provisioned progressively (continuous small doses of food) by nurse workers (Winston 1987).
59 Therefore, I subsequently tried to apply th e pesticide to each larva topically with the pesticide administered onto the larval cuticle in 1 L of acetone (Appendix C). Unfortunately, I believe that the worker bees displayed hygienic behavior (Arathi et al. 2000) and removed the treated larvae because of an effect the so lvent had on larval physiology. One possibility explaining the hygienic removal of the treated larvae could be that the application procedure stressed the treated larvae which, in turn, were removed by the worker bees. If pesticides are found in brood food and adult bees respond to the pe sticide by aborting the brood, this could lead to significant honey bee population d eclines. Colony loss due to this probably is unrelated to CCD, because most colonies e xperiencing CCD have significan t quantities of brood at colony death (Eccleston 2007, Ellis 2007, Johnson 2007, Embrey 2008). Had the worker bees not aborted the larvae, topically applying the pesticides may have been the best method of pesticide delivery in the field. However, as seen for imidacloprid, contact LD50s typically are higher than oral LD50s (Table 3-1). It is well known that the insect cuticle is a formidable barrier to some toxins (Yu 2008). LD50s determined through contact, rather than by ingestion in the diet, would best estimate toxicities of xenobiotics that tend to accumulate in beeswax, such as acaricides (Wallner 1999). For future studies, it will be important to de termine a method for assessing pesticide effects on brood in vivo This should help illumin ate how environmental xenobi otics affect the entire colony and whether such chemicals can be linked to colony declines. Additional research should focus on the long term, sublethal effects of pe sticides on bees. For example, a number of pesticide treated bees survived to adulthood in my study. It would be interesting to determine if these bees functioned normally after emergence.
60 Overall, my research indicates that acute su blethal pesticide doses for honey bee larvae can cause delayed mortality in honey bee pupae at con centrations lower than those shown to affect adult bees. A significant portion of the mortality experienced by the developing bees in Chapter 2 was manifested in the pupal stage, which is co unterintuitive because it was the larvae that were fed a pesticide-containing diet Based on my findings, I recommend that agrochemical safety registration studies be expanded to include immature bee toxicity assays. It is important that pupae be tested as well since in my study, pupae a ppeared to be more vulnerable to pesticides acquired during larval stages than the larvae themselves.
61 Table 3-1. Published contact and oral LD50 values for imidacloprid tested on adult honey bees. Method Contact LD50 Or al LD50/LC50 Source Acute 14 ng/bee and 24 ng/bee5 ng/bee (Suchail et al. 2000) Acute 49-102 ng/bee 41-81 ng/bee (Nauen et al. 2001) Acute na 3.7-40.9 ng/bee (Schmuck et al. 2001) Acute na 40-60 ng/bee (Suchail et al. 2001) Chronic na (0.01. 0.1, and 1 ppb) (Suchail et al. 2001) Acute 17.9 ng/bee na (Iwasa et al. 2004) Each dose is categorized as either acute or ch ronic indicating the number of times each dose was administered to an adult bee.
62 Figure 3-1. Hypothetical model depi cting the range of imidacloprid doses tested in Chapter 2 (triangles, 5-80 ppb) and the range of doses likely to be tested in a full LD50 doseresponse bioassay (plus signs, ppb-ppt). 20 40 60 80 100 ppb ppm ppt Pesticide Dose + + + + + + + +% Mortality 20 40 60 80 100 ppb ppm ppt Pesticide Dose + + + + + + + + 20 40 60 80 100 20 40 60 80 100 ppb ppm ppt Pesticide Dose ppb ppm ppt Pesticide Dose + + + + + + + + + + + + + + + + + + + + + + +% Mortality
63 Figure 3-2. Hypothetical model illustrating a t ypical dose-response curve. Squares indicate mortality values close to the LD50. Larger variations or wobble are likely to occur at mortalities farther from the LD50 value (circles). Log (Dose)% ResponseLD50 Log (Dose)% ResponseLD50
64 APPENDIX A METHOD FOR TOPICAL APPLICATION OF SUBC HRONIC IMIDACLOPRID AND AMITRAZ DOSES TO HONEY BEE LARVAE I investigated a method for topically applyi ng imidacloprid and amitraz to developing bee larvae. I diluted both pesticides (imidacloprid and amitraz) in acetone (Fischer Chemical, Fair Lawn, NJ) to create 5 treatment stock solutions pe r pesticide. I stored th e stock solutions at -20oC in 4 mL-brown amber vials (FischerBrand, Fische r Scientific). For imid acloprid (Chem Service, West Chester, PA), I created 2.5, 5 10, 20, a nd 40 ppb treatment stock solutions while for amitraz (Chem Service, West Chester, PA), I created 15.62, 31.25, 62.5, 125, and 250 ppb treatment stock solutions. For both pesticides, I included two controls: (1) a 1 l drop of acetone and (2) no pesticide/acetone application. Following the general rearing procedure outlined in Chapter 2, I collected same-aged larvae from 5 different queen sources. Once the larvae were 156 h old, I applied a 1 l drop of stock solution to the top of each larva using a Hamilton micro-applicator (Hamilton Company, Reno, NV). I repeated this procedure for 6 larv ae 7 treatments 5 queen sources. For all treatments (5 levels of pesticide and 2 cont rols), I applied the dos es once at hours 156, 180, 204, and 228. All developing larvae were maintained in the dark at 35oC and ~96% RH (Vandenberg and Shimanuki 1987). Both imidacloprid (Zheng a nd Liu 1999) and amitraz (Hollingworth 1976) are stable at 35oC. I determined if doses of imidacloprid and amitraz were lethal to immature bees by comparing mortality data between chemically tr eated and control larvae. Mortality variables included number of larvae that defecated, pupal surv ival (number of larvae to emerge as adults from those that defecated), and adult emergence (total number of bees to reach adulthood from those that were grafted). I measured a number of variables to identify potential sublethal effects of these chemicals on developing brood. These meas urements included time to larval defecation, larval weight at defecation, time to adult emer gence, adult bee weight and adult bee head weight. Imidacloprid Data for percent of larvae that de fecated, pupal survival, and adult bee emergence are recorded in Table A-1. I did not find significant differences between treatments for percentage of larvae to defecate ( 2 = 3.36; df = 6; P = 0.76), pupal survival ( 2 = 2.93; df = 6; P = 0.82), or adult bee emergence ( 2 = 4.09; df = 6; P = 0.66). Sublethal data are re ported in Table A-2. There was no queen treatment interaction ( F = 0.92; df = 24, 107; P = 0.58) or treatment effect ( F = 1.52; df = 6, 107; P = 0.18) for time to larval defecation. I did find a significant queen source effect on time to larval defecation (F = 42.20; df = 4, 107; P < 0.01) in which larvae from queen sources 112 and 103 (259.2 2.4 (27) and 266.4 2.4 (29) h, respectively) defecated later than larvae from queen sources 104 (223.2 2.4 (24) h), 110 (230.2 2.6 (27) h), and 114 (235.2 2.4 (35) h).There was no queen treatment interaction ( F = 0.71; df = 24, 107; P = 0.83) or treatment effect ( F = 0.37; df = 6, 107; P = 0.90) on larval weight at defecation. There was a significant effect of queen source on larval weight at defecation ( F = 4.65; df = 4, 107; P < 0.01). Larvae from queens 103 (114.6 3.8 (30) mg) and 114 (114.7 3.1 (35) mg) were significan tly heavier upon defecation than larvae from queen sources 110 (97.4 2.7 (27) mg) and 104 (103.7 2.5 (24) mg). I did not find a queen treatment interaction ( F = 1.61; df = 6, 6; P = 0.29), treatment effect ( F = 1.08; df = 6, 6; P = 0.46), or queen source effect ( F = 3.77; df = 4, 6; P = 0.07) for time to adult emergence. Finally, there was no queen treatment interaction ( F = 2.56; df = 6, 6; P = 0.14) or treatment effect ( F = 2.77; df = 6, 6; P = 0.12) for adult bee head weight. Emerging adults from queen 112 had the
65 heaviest adult head weights (13.9 1.6 (3) mg) compared to head weights from emerging adults from all queen sources 103 (10.6 0.6 (6) mg), 104 (11.7 1.0 (2) mg), 110 (10.1 0.4 (2) mg), and 114 (10.5 0.2 (10) mg) ( F = 7.22; df = 4, 6; P < 0.02). Amitraz For amitraz treated larvae I found no significant differences between treatments for percent of larvae that defecated ( 2 = 9.55; df = 6; P = 0.14), pupal survival ( 2 = 2.19; df = 6; P = 0.90), or adult bee emergence ( 2 = 3.00; df = 6; P = 0.81) (table A-3). Data for sublethal effects on time to larval defeca tion, larval weight at defecatio n, time to adult emergence, and adult bee head weights are recorded in Tables A-4 and A-5. I did not find a queen treatment interaction ( F = 0.97; df = 24, 97; P = 0.51) or treatment effect ( F = 0.98; df = 6, 97; P = 0.44) for time to larval defecation. I did find a significant effect of queen source (F = 46.46; df = 4, 97; P < 0.01). Larvae from queen sources 103 (261.6 2.4 (23) h) and 112 (259.2 2.4 (28) h) took significantly longer to defecate than larvae fr om queen sources 104 (228.0 2.4 (24) h), 110 (228.0 2.4 (25) h), and 114 (9.6 0.1 (32) mg). I did not find a queen treatment interaction ( F = 1.05; df = 24, 98; P = 0.41), treatment effect ( F = 0.19; df = 6, 98; P = 0.98), or queen source effect (F = 0.19; df = 4, 98; P = 0.43) for larval defecation we ight. There was a significant queen treatment interaction ( F = 5.47; df = 6, 7; P = 0.02) for time to adult emergence. I analyzed time to larval emergence individually according to queen, but I could not detect any trends within the data because of the small sample sizes (Table A-5).There was no queen treatment interaction ( F = 0.40; df = 6, 7; P = 0.86), treatment effect ( F = 1.05, df = 6, 7; P = 0.47), or queen source effect ( F = 0.15; df = 4, 7; P = 0.96) for adult bee head weight. The results from this investigation were not incorporated into the main body of my thesis because I felt the data produced during this test were unreliable. I be lieve the data to be unreliable for 4 reasons. First, the control larv ae experienced a high mortality overall and results for untreated and acetone treated larvae were not c onsistent. Second, I believe that a considerable amount of the mortality larvae experienced while developing was caused by the method of applying the various treatments with the micro applicator. After some treatment applications, the larvae appeared to have been punctured by the a pplicator. Third, a large number of bees died while pupating. I believe this occurred due to my use of salt solutions to control the relative humidity. Once I switched to using a temperature and humidity controlled in cubator, I had better success at rearing larvae. Finally, hydrophilic materials such as imidacloprid are not likely to easily cross the insect cuticle s uggesting that the LD50 values fo r topically applied substances will be larger than orally applied ones (Yu 2008). Initially, I chose to test topical treatment a pplications because this would have been the easiest method to replicate in the field (Appe ndix C). The inability to transfer laboratory experiments into the field is often termed the lab-to-field dilemma due to the difficulties that surround the process (Landis and Yu 1995). Howe ver, after I discovered the difficulties associa ted with topically applyi ng pesticides, I decided to try applying the pesticides at larger doses once during larval development (Appendix B) rather than at smaller doses multiple times.
66 Table A-1. Percentage of larvae that defecate d, pupal survival, and adult bee emergence for larvae treated topically with imidacloprid for 4 days. Treatment Larvae that defecateda (%) Pupal survival b (%) Adult bee emergencec (%) Untreated 67 9.4 (5)a29.3 14.4 (5)a 23.3 11.3 (5)a Acetone 67.3 9.3 (5)a6.7 6.7 (5)a 3.3 3.3 (5)a 2.5 69.3 5.9 (5)a19 9.3 (5)a 14 6.3 (5)a 5 73.3 13.5 (5)a26.7 13.5 (5)a 13.3 6.2 (5)a 10 76.7 4.1 (5)a12 4.9 (5)a 10 4.1 (5)a 20 76.7 11.3 (5)a10.7 6.9 (5)a 10 6.7 (5)a 40 62 3.3 (5)a10 6.1 (5)a 6.7 4.1 (5)a Imidacloprid concentrations are in ppb. Data are mean std. error (n) where n = number of queen sources used. Columnar means followe d by the same letter are not different at 0.05. a % of larval that defecated = (the number of larvae that defecated/the total number of treated larvae) 100. b % pupal survival = (the number of adult bees that emerged/the number of larvae that defecated) 100. c % adult bee emergence = (the numbe r of adult bees that emerged/the total number of treat ed larvae) 100.
67 Table A-2. Mean time to larval defecation, larv al weight at defecation, time to adult bee emergence, and adult bee head weight for larvae treated topically with imidacloprid for 4 days. Treatment Time to larval defecation (h) Larval weight at defecation (mg) Time to adult bee emergence (h) Adult bee head weight (mg) Untreated 247.2 4.7 (18)a 111.7 3.3 (18)a508.8 4.1 (6)a 10.3 0.4 (6)a Acetone 244.8 6.2 (20)a 107.5 5.3 (19)a 480 (1)a 10.6 (1)a 2.5 247.2 4.6 (20)a 107.6 4.8 (20)a508.8 11.5 (4)a 10.8 1.1 (4)a 5 240 4 (22)a 104.9 3.3 (22)a508.8 6 (4)a 12.9 1.4 (4)a 10 247.2 5 (23)a 110.3 3.2 (23)a520.8 7.9 (3)a 10.8 0.3 (3)a 20 240 4.8 (22)a 111.6 5 (22)a520.8 7.9 (3)a 10.6 0.6 (3)a 40 237.6 5.6 (17)a 105.9 4.4 (18)a528 24 (2)a 11.6 0.9 (2)a Imidacloprid treatments are in ppb. Data are mean std. error (n) where n = number of individuals used, pooled from all queen sources. Columnar means followed by the same letter are not different at 0.05.
68 Table A-3. Percentage of larvae that defecate d, pupal survival, and adult bee emergence for larvae treated topically w ith amitraz for 4 days. Treatment Larvae that defecateda (%) Pupal survival b (%) Adult bee emergencec (%) Untreated 80 9.7 (5)a17 7.7 (5)a 14 6.3 (5)a Acetone 58.7 3.7 (5)a18.3 13 (5)a 10 6.7 (5)a 15.62 75.3 7.3 (5)a29 9.5 (5)a 23.3 8.5 (5)a 31.25 52 5.6 (5)a20 13.3 (5)a 10 6.7 (5)a 62.5 68.7 13 (5)a14 9.8 (5)a 11.3 7.8 (5)a 125 70 9.7 (5)a13.3 8.2 (5)a 7.3 4.5 (5)a 250 78 7 (5)a11.7 7.3 (5)a 8 4.9 (5)a Amitraz concentrations are in ppb. Data are mean std. error (n) where n = number of queen sources used. Columnar means followed by the same letter are not different at 0.05. a % of larvae that defecated = (the number of larvae that defecated/the total numbe r of treated larvae) 100. b % of pupal survival = (the nu mber of adult bees that emer ged/the number of larvae that defecated) 100 c % of adult bee emergence = (the number of adult bees that emerged/the total number of treated larvae) 100.
69 Table A-4. Mean time to larval defecation, larval weight at def ecation, and adult bee head weight for larvae treated topically with amitraz for 4 days. Treatment Time to larval defecation (h) Larval weight at defecation (mg) Adult bee head weight (mg) Untreated 242.4 103.2 (22)a105.5 3.1 (22)a9.7 1 (4)a Acetone 244.8 115.2 (17)a105.7 5.1 (17)a9.3 1 (3)a 15.62 244.8 120 (22)a105.5 3.7 (22)a11 0.5 (7)a 31.25 237.6 120 (15)a103.7 3.3 (15)a11.1 0.2 (3)a 62.5 240 115.2 (19)a103.7 4 (19)a10.1 0.7 (3)a 125 240 110.4 (19)a100.7 3.7 (20)a12.1 0.6 (2)a 250 240 103.2 (18)a104.8 3.7 (18)a8.6 0.1 (2)a Amitraz treatments are in ppb. Data are mean st d. error (n) where n = nu mber of individuals used, pooled from all queen sources. Column ar means followed by the same letter are not different at 0.05.
70 Table A-5. Mean time to adult emergence in hours for larvae treated topically with amitraz for 4 days. Queen source Treatment 103 104 110 112 114 Untreated 528 (1) 528 (1)ana na 516 12 (2)a Acetone 600 (1) na na na 492 12 (2)a 15.62 na 492 12 (2)a480 (1)a528 (1)a 504 (3)a 31.25 na na 492 (2)a528 (1)a na 62.5 na 480 (1)ana 552 (2)b na 125 528 (1) na 504 (1)ana na 250 na 528 (1)a504 (1)ana na ANOVA F = na; df = 0, 0; P = na F = 2.33; df = 3, 1; P = 0.44 F = 0.47; df = 3, 1; P = 0.76 F = infty; df = 2, 1; P < 0.01 F = 2.00; df = 2, 4; P = 0.25 Amitraz treatments are in ppb. Data are mean st d. error (n) where n = nu mber of individuals used, pooled from all queen sources. Column ar means followed by the same letter are not different at 0.05.
71 APPENDIX B METHOD FOR TOPICAL APPLICATION OF ACUT E IMIDACLOPRID DOSES TO HONEY BEE LARVAE The purpose of this investigation was to de velop a method for test ing the effects of topically applying acute doses of imidacloprid to developing b ee larvae. In appendix A, my findings suggested that larvae ca n be damaged by applying treatm ents using a micro applicator. As such, I hypothesize that larv ae treated only once during thei r development would be more likely to survive treatment since their expos ure to the applicator would be reduced. Consequently, I chose to use only one pesticide (imidacloprid) in the cu rrent study to determine if I could improve larval surv ivability. To accomplish this, I di luted imidacloprid in acetone (Fischer Chemical, Fair Lawn, NJ) to create 5 treatment stock solutions (5, 50, 100, 250, and 500 ppb imidacloprid/1 l drop) and stored the solutions at -20oC in 4 mL-brown amber vials (FischerBrand, Fischer Scientific). I increase d the doses in this st udy over those used in Appendix A because I treated each larva only once. Following the general rearing procedure outlined in Chapter 2, I collected same-aged larvae from 3 different queen s ources and grafted them into 96-well plates. When the larvae were 156 h old, I applied 1 l of one imidacloprid st ock solution onto the larvas exposed surface using a Hamilton micro-applicator (Hamilton Co mpany, Reno, NV). I repeated this procedure for 6 larvae per stock solution for each queens o ffspring. Furthermore, I included 2 controls for each queen, the first with a 1 l drop of acetone and the second with no acetone or pesticide application. This also was repeated for 6 larvae per queen. To see if larvae of varying ages were more susceptible to topical imidacloprid applications, I repeated the entire procedure on larvae from ages 156, 180, 204, and 228 h. All developing larvae were maintained in the dark at 35oC and ~96% RH (Vandenberg and Shimanuki 1987). I determined if doses of imidacloprid were lethal by comparing mo rtality data between chemically treated and control larvae. Mortal ity variables included number of larvae that defecated, pupal survival (number of larvae to emer ge as an adult from those that defecated), and adult emergence (total number of bees to re ach adulthood from those that were grafted). I measured a number of variables to identify potential sublethal effects of imidacloprid on developing brood. These measurements included tim e to larval defecation, larval defecation weight, time to adult emergence, adult bee weight, and adult bee head weight. Data for percent of larvae that defecate d, pupal survival, and adult bee emergence are recorded in Tables B-1, B-2, & B-3. There was no significant day treatment interaction ( F = 0.96; df = 18, 56; P = 0.51) or treatment effect ( F = 0.47; df = 6, 56; P = 0.83) on percent of larvae that defecated. There was a significant da y effect in which larvae treated only at 84 h (79.4% 3.4) had lower percent larval defecation than larvae treated only at 132 h (94.4% 2.3) ( F = 3.66; df = 3, 56; P = 0.02). Larvae treated at 108 h and 156 h were not different from those treated at 84 h and 132 h. There was no si gnificant day treatment interaction ( F = 0.48; df = 18, 56; P = 0.96), day effect ( F = 1.26; df = 6, 56; P = 0.29), or treatment effect ( F = 0.11; df = 3, 56; P = 0.95) on pupation success. There was no significant day treatment interaction ( F = 0.45; df = 18, 56; P = 0.97), day effect (F = 1.08; df = 6, 56; P = 0.38), or treatment effect ( F = 0.24; df = 3, 56; P = 0.87) on adult bee emergence. Of the response variables tested, there were no significant sublethal effects of imidacloprid on treated larvae (Table B-4, B-5, B-6, & B-7). I did not find a day treatment interaction ( F = 0.73; df = 18, 307; P = 0.78), treatment effect ( F = 0.28; df = 6, 307; P = 0.95) or day effect ( F = 0.97; df = 3, 307; P = 0.41) for time to larval defecation. There was no day treatment
72 interaction ( F = 1.04; df = 18, 304; P = 0.41), treatment effect ( F = 1.45; df = 6, 304; P = 0.21), or day effect (F = 0.98; df = 3, 304; P = 0.42) on larval weight at defecation. There was no day treatment interaction ( F = 0.94; df = 17, 62; P = 0.54), treatment effect ( F = 0.43; df = 6, 62; P = 0.86), or day effect (F = 0.39; df = 3, 62; P = 0.76) for time to adult emergence. There was no queen treatment interaction ( F = 1.54; df = 17, 62; P = 0.11), treatment effect ( F = 1.33; df = 6, 62; P = 0.25), or day effect (F = 0.68; df = 3, 62; P = 0.57) for adult bee head weight. Results from this investigation were similar to those in Appendix A in that these data do not indicate the presence/absence of effects asso ciated with treating developing larvae once with acute doses of imidacloprid. The primary reason fo r this is that there was high mortality among control larvae, indicating that the rearing pr ocedure had not been perfected. I did not see differences between treated and unt reated larvae. This was likely caused by the fact that some chemicals and solvents, especially hydrophilic ones do not easily penetrate the insect cuticle (Yu 2008). I believe that this can be corrected by feed ing larvae diet containing pesticides rather than topically applying the pesticide. Secondly, I will no longer use salt solutions to control humidity in the test chambers because I believe that bette r results can be achieved using a temperature and humidity controll ed incubator.
73 Table B-1. Percentage of larvae that defecated for larvae treated topically with imidacloprid once during their development. Larvae that defecateda (%) Treatment 84 108 132 156 Untreated 66.7 (3)a 87.8 6.2 (3)a91.7 8.3 (3)a 93.3 6.7 (3)a Acetone 72.2 14.7 (3)a 83.3 9.6 (3)a100 (3)a 93.3 6.7 (3)a 5 83.3 9.6 (3)a 100 (3)a94.4 5.5 (3)a 88.9 11.1 (3)a 50 94.4 5.5 (3)a69.4 10 (3)a 88.9 11.1 (3)a 83.3 9.6 (3)a 100 72.2 5.5 (3)a77.8 11.1 (3)a100 (3)a 85.0 7.6 (3)a 250 83.3 9.6 (3)a 94.4 5.5 (3)a91.7 8.3 (3)a 83.3 16.7 (3)a 500 83.3 9.6 (3)a 94.4 5.5 (3)a94.4 5.5 (3)a 83.3 9.6 (3)a Imidacloprid concentrations are in ppb and were administered to larvae either at hours 84, 108, 132, or 156 h. Data are mean std. error (n) wh ere n = number of queen sources. Columnar means followed by the same letter are not different at 0.05. a % of larvae that defecated = (the number of larvae that defecated/the total number of treated larvae) 100.
74 Table B-2. Percentage of pupal survival for pupae treated topically as la rvae with imidacloprid once during their development. Pupal survivala (%) Treatment 84 108 132 156 Untreated 33.3 8.3 (3)a32.8 4.3 (3)a50.0 25.4 (3)a 38.9 20 (3)a Acetone 11.1 11.1 (3)a33.3 33.3 (3)a33.3 33.3 (3)a 13.3 13.3 (3)a 5 21.7 11.7 (3)a27.8 20 (3)a28.9 19.7 (3)a 38.9 13.9 (3)a 50 43.3 23.3 (3)a55.5 29.4 (3)a16.7 16.7 (3)a 31.7 9.3 (3)a 100 6.7 6.7 (3)a 0 (3)a6.7 6.7 (3)a 30.5 19.4 (3)a 250 20.0 20 (3)a27.8 14.7 (3)a48.9 24.7 (3)a 33.3 16.7 (3)a 500 57.8 21.2 (3)a36.7 18.5 (3)a24.4 12.4 (3)a 48.3 25.9 (3)a Imidacloprid concentrations are in ppb and were administered to larvae either at hours 84, 108, 132, or 156 h. Data are mean std. error (n) wh ere n = number of queen sources. Columnar means followed by the same letter are not different at 0.05. a % pupal survival = (the number of adult bees that emerged/the num ber of larvae that defecated) 100.
75 Table B-3. Percentage of adult bee emergence fo r adult bees treated as larvae topically with imidacloprid once during their development. Adult bee emergencea (%) Treatment 84 108 132 156 Untreated 22.2 5.5 (3) a28.9 4.4 (3) a47.2 26.5 (3) a 38.9 20 (3) a Acetone 11.1 11.1 (3) a33.3 33.3 (3) a33.3 33.3 (3) a 13.3 13.3 (3) a 5 16.7 9.6 (3) a27.8 20.0 (3) a27.8 20.3 (3) a 36.1 15.5 (3) a 50 38.9 20.0 (3) a33.3 16.7 (3) a11.1 11.1 (3) a 25 4.8 (3) a 100 5.5 5.5 (3) a 0 (3) a6.7 6.7 (3) a 23.3 14.5 (3) a 250 16.7 16.7 (3) a27.8 14.7 (3) a48.9 24.7 (3) a 33.3 16.7 (3) a 500 44.4 11.1 (3) a33.3 16.7 (3) a22.2 11.1 (3) a 36.1 15.5 (3) a Imidacloprid concentrations are in ppb and were administered to larvae either at hours 84, 108, 132, or 156 h. Data are mean std. error (n) wh ere n = number of queen sources. Columnar means followed by the same letter are not different at 0.05. a % of adult bee emergence = (the number of adult bees that emerged/th e total number of treated larvae) 100.
76 Table B-4. Mean time to larval defecation for larvae treated topically with imidacloprid once during their development. Time to larval defecation (h) Treatment 84 108 132 156 Untreated 10.1 0.2 (10)a9.9 0.1 (12)a 9.8 0.4 (10)a 10.2 0.2 (13)a Acetone 10.4 0.1 (11)a9.9 0.2 (12)a 9.9 0.3 (13)a 9.8 0.2 (12)a 5 10.5 0.3 (12)a10.1 0.2 (15)a10 0.1 (14)a 9.7 0.2 (12)a 50 10 (14)a9.9 0.2 (9)a10.3 0.5 (12)a 9.7 0.1 (11)a 100 10.3 0.3 (11)a10 0.2 (12)a10.1 0.4 (13)a 9.8 0.2 (12)a 250 9.9 0.1 (12)a9.8 0.1 (14)a 9.7 0.1 (11)a 10.1 0.3 (7)a 500 9.9 0.1 (13)a10.1 0.1 (13)a10.1 0.4 (14)a 10.2 0.3 (11)a Imidacloprid concentrations are in ppb and were administered to larvae either at hours 84, 108, 132, or 156 h. Data are mean std. error (n) wher e n = number of individuals from all queen sources. Columnar means followed by th e same letter are not different at 0.05.
77 Table B-5. Mean larval weight at defecation for larvae treated topically with imidacloprid once during their development. Larval weight of defecation (mg) Treatment 84 108 132 156 Untreated 117.8 9.9 (10)a124.4 5.5 (12)a117.2 8.3 (10)a 123.6 6.6 (13)a Acetone 125.3 10.6 (11)a122 6.9 (12)a113.2 5.8 (13)a 116.8 6.6 (12)a 5 134.4 6.4 (11)a121.9 5.4 (15)a120.6 4.5 (14)a 110.4 5.4 (12)a 50 136.1 2.8 (14)a127.8 5.4 (9)a125.3 9.1 (11)a 120.4 5.6 (11)a 100 123.2 6.2 (11)a118.5 6 (12)a123.5 6.5 (12)a 127.7 6.2 (12)a 250 129.9 5.5 (12)a113.9 6.1 (14)a122.5 4 (11)a 137.6 9.3 (7)a 500 124.8 6.7 (13)a127.6 6.6 (13)a133.9 5.2 (14)a 134 9 (11)a Imidacloprid concentrations are in ppb and were administered to larvae either at hours 84, 108, 132, or 156 h. Data are mean std. error (n) wher e n = number of individuals pooled from all queen sources. Columnar means followed by the same letter are not different at 0.05.
78 Table B-6. Mean time to adult emergence for larvae treated topically with imidacloprid once during their development. Time to adult emergence (h) Treatment 84 108 132 156 Untreated 504 13.8 (3)a504 (4)a488 8 (3)a 499.2 4.8 (5)a Acetone 504 (2)a488 8 (3)a492 12.0 (2)a 504 (2)a 5 504 (3)a496. 8 (3)a504 13.8 (3)a 498 6 (4)a 50 500.6 3.4 (7)a488 8 (3)a504 (1)a 488 8 (3)a 100 504 (1)ana 504 (1)a 512 8 (3)a 250 496 8 (3)a492 6.9 (4)a500 4 (6)a 504 13.8 (3)a 500 488. 5 (6)a504 (4)a496 8 (3)a 498 6 (4)a Imidacloprid concentrations are in ppb and were administered to larvae either at hours 84, 108, 132, or 156 h. Data are mean std. error (n) wher e n = number of individuals from all queen sources. Columnar means followed by th e same letter are not different at 0.05.
79 Table B-7. Mean adult bee head weight for a dult bees treated topically as larvae with imidacloprid once during their development. Adult bee head weight (mg) Treatment 84 108 132 156 Untreated 12 1.3 (3)a10.2 0.1 (4)a 10.3 0.7 (3)a 11.5 0.9 (5)a Acetone 12.3 0.4 (2)a10.4 0.6 (3)a 10.3 1.3 (2)a 9.9 0.7 (2)a 5 11.7 0.1 (3)a12.1 0.4 (3)a1.2 0.5 (3)a 10.7 0.2 (4)a 50 11.3 0.4 (7)a11.5 0.5 (3)a11.4 (1)a 9.6 0.8 (3)a 100 13.1 (1)ana 12.9 (1)a 11.9 0.4 (3)a 250 19 0.6 (3)a9.4 0.7 (4)a 11.9 0.5 (6)a 11.4 1.3 (3)a 500 10.4 0.5 (6)a10.6 0.4 (4)a11.1 1 (3)a 11.5 0.5 (4)a Imidacloprid concentrations are in ppb and were administered to larvae either at hours 84, 108, 132, or 156 h. Data are mean std. error (n) wher e n = number of individuals pooled from all queen sources. Columnar means followed by the same letter are not different at 0.05.
80 APPENDIX C METHOD FOR TESTING THE ATTRACTION OF FOUNDRESS VAR ROA MITES TO IMIDACLOPRID AND AMIT RAZ TREATED LARVAE Investigators have recognized many factors that could contribute to a decline in the overall health of honey bees. Such fact ors include pesticides, parasites, pathogens, Africanized bees, and other stressors such as transporting colonies for pollination services (Johansen and Mayer 1990, Morse and Flottum 1997). Unfortunately, most studies on bee decline focus on a single stressor in an effort to reduce experimental variation. This can lead one to igno re/overlook the possibility of interactions between multiple st ressors affecting honey bee health. In this study, I attempted to develop a met hod to investigate an interaction between 2 previously mentioned stressors: pesticides and parasites. More specifically, I measured how imidacloprid and amitraz affected larval honey bee attraction to varroa mites ( Varroa destructor Anderson and Truemann), recognizin g that pesticides and varroa may interact synergistically to kill larvae. The data I present in Chapter 2 provi des the framework for the research presented in this chapter. In Chapter 2, I disc ussed how larval development is affected when larvae feed on a diet containing pesticides. I disc overed that if larvae were trea ted with imidacloprid (5-80 ppb) or amitraz (50-400 ppb), they were less likely to survive to adult hood. In addition to the lethal and sublethal effects described in Chapter 2, pest icides may affect other aspects of honey bee larval life history including their susceptibility to parasites. I chose to use varroa mites in this investigation because they are considered the leading cause of honey bee colony deaths worldwide (DeJong 1997, Berenbaum et al. 2007, Bo ard 2008) and a possible cause of current global declines in managed bee colonies (Stokstad 2007). Varroa are a useful model for studying synergis tic interactions between pesticides and bee pests on bee larvae because varroa life history overlaps with the honey bee brood cycle (Figure 3-1). Varroa reproduce only in brood cells, invadi ng the cells 15 20 hours pr ior to the brood cell being capped by worker bees. Biolog ical factors affecting varroa decision to invade brood cells have been studied (Fuchs 1990, Boot et al. 1994, Bienefeld et al. 1998, Beetsma et al. 1999, Piccirillo and De Jong 2004). However, there is a modicum of data on the effects of nonbiological factors (such as larval pesticide stress) on mite invasi on into brood cells. In one such investigation, Ellis and Delaplane (2001) did no t find that colony treatment with Fumidil B (fumagillin) and/or Terramycin (oxytetracycline) influenced varroa invasion into brood cells. However, no similar studies exis t on other chemical agents. I chose to investigate imidacloprid and amitraz because both are associated closely with honey bee coloniesimidacloprid due to its systemic properties and widespread use (Elbert et al. 1991, Bonmatin et al. 2003, Chauzat et al. 2006, Yu 2008) and amitraz because it has been known to be applied directly into managed colonies by beekeepers in an e ffort to control varroa. Same-aged larvae were collected following the methods outlined in Chapter 2. To produce the larvae, a queen was confin ed to a newly-drawn frame ( no stored honey, pollen, or brood) using 1 of 2 different sized zinc queen ex cluder cages (96 106 21 mm; 165 130 24 mm) at time t = -12 h. The caged queen and frame were returned to the center of the brood nest (Boot and Calis 1991). After 24 h, t = +12 h (Peng et al. 1992, Aupinel et al. 2005), the queen was removed from the cage and the cage secured to the comb as before but this time for 132 h (from t = 0) to allow the eggs to hatch and larvae to grow large enough to treat (Figure 3-2). Worker bees exclusively were able to access the comb in order to feed the new larvae hatching from the eggs.
81 I treated the larvae with one of 2 different pesticides (amitraz or imidacloprid) at one of 5 different doses per pesticide. The tested imidacloprid doses were 100, 200, 400, 800, and 1600 ppb and amitraz doses were 500, 1000, 2000, 4000, and 8000 ppb. These doses were obtained by determining the specific quantity of pesticide received by a larva in 20 L of diet during the oral toxicity test outlined in Chapter 2. For example, th e lowest dose of amitraz mixed into larval diet was 25 ppb (Chapter 2). The lowest pesticide concentration in th e current experiment was equal to 500 ppb. I calculated this by determining that 25 ppb of amitraz in 20 L of larval diet is equivalent to 500 ppb in a 1 L-drop. All pesticid e doses were diluted in acetone to achieve the desired concentration. I included tw o control treatments for both pesticides, with one series of larvae receiving acetone only and the second seri es receiving nothing. I chose to use acetone as the pesticide solvent because it resulted in fewer larvae being removed from the combs by the bees when compared to bee removal of larvae tr eated with methanol or ethanol (Appendix D). At 132 h, the caged frames (now containing 60 12 h old larvae) were removed from the colonies and taken to the lab to be treated. At the laboratory, I placed a sheet of transparency paper (Office Depot Inc, Delray Beach, FL) over the frame to cover the group of similar-aged larvae. I then taped the sheet down on 2 sides to hold it in place on the frame. Using a permanent marker, I marked 14 diagonal rows of 10 larvae per row on the transparency paper. This permitted me to use the paper as a map to find my treated larvae on subsequent days. After marking the rows on the pa per, I labeled each row 1-14 and then randomly assigned each row a treatment (5 doses of imidacloprid + 5 doses of amitraz + 4 controls = 14 treatments). I applied a given treatment to10 larvae in 1 L droplets using a Ham ilton micro-applicator (Hamilton Company, Reno, NV). To prevent the larvae from be ing damaged during treatment applications (see Appendix A & B), I did not touch the applicat or to the larvae and instead sprayed the droplet into the ce ll. Once treated, the larvae were re turned to their original colony until the following days treatmen t. I treated all larv ae consecutively for 3 days at 132 h, 156 h, and 180 h post oviposition and repeated this proce dure for larvae from 9 queens. After the third treatment application, th e larvae were returned to their or iginal colonies. Varroa foundresses enter brood cells immediately before the cells are capped (Morse a nd Flottum 1997). Because cells containing larvae are capped by worker b ees at ~204 h post ovi position (Winston 1987), I left treated larvae in the te st colonies for 228 253 h. Following this period, I removed the treated comb sections from their respective hi ves, uncapped all remain ing treated brood cells, removed the prepupae/pupae within, and counted the number of foundress va rroa mites per cell. To confirm varroa presence in test coloni es, I estimated varroa populations in the colonies using varroa sticky-screens placed on top of the bottom board of the hive (Devlin 2001). I left the screens in the colonies for 2 days, after which I removed the screens and counted the varroa. This permitted me to determine if the test colonies were hosting similar varroa populations. The effect of larval treatment history on the number of foundress mites per cell was analyzed using a 1-way ANOVA recognizing treatment as the main effect. All statistical tests were conducted using SAS (SAS Institute, 2008). I had a total of 2 rows for both controls (untreated and acetone) for both imidacloprid and amitraz, each row containing 10 larvae. Since both pesticide treatments were on the same fram e, controls for both pesticides were pooled. Although I treated 10 larvae per treatment ( 12 treatments) for each of 9 queens, I discovered that adding acetone to brood cells often caused the adu lt bees to abort the treated brood, regardless of the presence of pesticide (A ppendix D). As such, I pooled like treatm ent data from each queen source to increase the sample size for the analysis. In summary: n = 9
82 queens 12 treatments (5 imidacloprid, 5 amitraz and 2 controls) # larvae that were not aborted in each treatment. Overall, I treated 1260 larvae with various doses of imidacloprid, amitraz, acetone, or nothing. Despite the volume of larvae treated, varroa mite invasion into brood cells was unaffected by the chemotherapeutic history of the host larvae ( F = 0.18; df = 11, 211; P = 0.99; Table 3-1). There were no dead varroa present in any of the brood cells. Overall mite abundance was low (imidacloprid = 12; amitraz = 20 mites) This was calculated by summing the total number of mites found in ever y cell across all treatments. Based on the method I used, I do not think te sting varroa attraction to larvae would be achieved easily in the field. Furthermore, there is no conclusive evidence that suggests treatment of larvae with amitraz or imidacloprid influences varroa mite decision to invade worker brood cells. It is difficult to state with any certaint y that imidacloprid and amitraz had no effect on varroa choice because overall, both pesticide gro ups contained low mite abundance. On average, tested colonies had 46.8 16.0 (7) varroa/2 day stic ky screen, mean std. error (n). The lack of dead mites in the cells suggests th at even though the bee larvae were fed an insectic ide/acaricide, lethal properties from the pesticide were not transferred through the pupaes hemolymph and fed on by mites resulting in mite deat h. The data also suggest that mites did not die as a result of contacting pesticide-laden food even though varroa foundresses are known to bury themselves in brood food until the cell is capped. Difficulties in this experiment resulted from problems associated with administering pesticides to brood under field c onditions. Past investigators us ed solitary bees, which mass provision their larvae, to successfully administer pesticide doses to individual larva because each larva was enclosed inside its own cell (Abbott et al. 2008). Consequently, the amount of pesticide consumed by the larva could be controlle d. The difficulty with honey bees is that larvae are tended by worker bees who remove acetone-treat ed larvae from the hive. Why this occurred remains unknown. Instead, future experiments testing my hypothesi s could remain in the laboratory allowing the investigator to control for va riables such as larval abortion. One way to test varroa attraction to pesticide treated brood in th e lab would be to capture volatiles from treated brood and present them to varroa in controlled choice tests. A seco nd method would be to re ar and treat larvae in a laboratory as before and then release varroa onto the well plates, permitting them to invade cells at will. Such tests are not easily replicated in field colonies, but their results could indicate whether or not the chemotherapeutic history of larvae affects varroa inva sion into brood cells.
83 Table C-1. Mean number of varroa found alive in ca pped cells containing either imidacloprid or amitraz fed larvae. Treatment Mite abundance Untreated 0.14 0.06 (73)a Acetone 0.09 0.05 (33)a Imidacloprid 100 0.07 0.07 (15)a Imidacloprid 200 0.06 0.06 (16)a Imidacloprid 400 0.21 0.11 (14)a Imidacloprid 800 0.07 0.07 (14)a Imidacloprid 1600 0.10 0.07 (19)a Amitraz 500 0.12 0.08 (16)a Amitraz 1000 0.12 0.12 (17)a Amitraz 2000 0.13 0.07 (23)a Amitraz 4000 0.12 0.08 (16)a Amitraz 8000 0.14 0.10 (14)a Prepupae/pupae were removed from capped bro od cells between 228-253 h and the number of varroa foundresses counted. Imid acloprid and amitraz doses are given in ppb/1 L drop. Data are mean std. error (n) where n = number of ca pped cells pooled across all queen sources. Columnar means followed by the same letter are not different at 0.05.
84 Figure C-1. Lifecycle of the varroa mite (m odified from DeJong 1997) (1) A phoretic varroa foundress attaches to an adult bee from which she feeds on hemolymph. (2) The foundress enters a brood cell prior to cell capping (~5.5 day old larvae). (3) Mite submerges itself in the brood food at the botto m of the cell. (4) M ite feeds on prepupae hemolymph. (5) The mite foundress begins ovipositing 60 hours after the cell is capped and lays subsequent eggs every 30 minutes. The first egg laid is a male and all others are female. (6) Approximately 1 to 6 eggs develop to adults. (7) The males take 5-6 days to mature while female mites take 7-8 days to mature. (8) Female mites mate with the single male mite within the cell. (9) Adult female mites leave the cell attached to the adult bee. The male and any immature female mites are left in the cell to die. (10) Female mites can be transferred between adult bees. DeJong, D. 1997. Mites: varroa and other parasites of brood, pp. 284, figure 14.1. In R. A. Morse and K. Flottum (eds.), Honey bee pests, predators, and diseases, 3rd ed. A.I. Root Company, Medina, OH.
85 024487296120144168192216240264288312336360384408432456480504528 egg laying egg development larval development pesticide treatment 1 pesticide treatment 2 pesticide treatment 3 larval defecation varroa invade cell prepupal/pupal development adult emergence Figure C-2. Field bioassay: experi mental timeline according to larval development in hours (0-528 h, x-axis). Pesticide treatme nt refers to the 3 time periods each larva was treated (at 132, 156, and 180 h) with their corre sponding pesticide treatment (imidacloprid or amitraz). The timeline dose not illustrate that egg lying began at t = -12 h.
86 APPENDIX D HYGIENIC BEHAVIOR OF ADULT HONEY B EES T OWARD LARVAE TREATED WITH DIFFERENT SOLVENTS In the experiments presented in this th esis (Chapter 2, Appendices A, B, & C), imidacloprid and amitraz treatments were prepared by diluting the pesticides in an acetone solvent. Although acetone worked as a solvent in chapter 2, the in vitro study, there were problems with its use in Appendix C, the field st udy. When applying the tw o pesticides mixed in acetone to larvae in bee colonies, the worker bees regularly responded by removing (aborting) the treated larvae (hygienic behavior). In my ge neral observation, I noticed that the bees were removing the acetone-only treated larvae but not the control (untreated) larvae. Consequently, I decided to investigate the effects of other solvents on bee removal of treated larvae in an effort to find a more suitable solvent to us e in the field study. I limited my investigations to testing three other solvents: methanol, ethanol and water. I did not expect water to be a good solvent for amitraz because amitraz is lypophili c (Hollingworth 1976) but maybe for imidacloprid since it is more hydrophilic (Elbert et al. 1991). To test the effects of each solvent on bee re moval of treated larvae, I treated larvae that weighed approximately 6 mg with one of 5 solv ents: acetone, methanol, ethanol, water, or no solvent. I used a micro-applicator (Hamilton Co mpany, Reno, NV) to treat larvae with 1 L of test solvent: I treated 10 larvae per treatment. All treatments were on a single frame for each queen. After treatment, I return ed the treated frame to its original colony. After 24 h, I brought the treated frame into the laboratory to count the nu mber of larva present fo r each treatment. If a larva was missing it was assumed to have been a borted by worker bees. I analyzed the number of larvae bees aborted with treatmen t as my main effect in a Kruskal-Wallis test for non-parametric data. Data was analyzed in SAS (SAS Institute, 2008). Results from this investigation did not show that solvent choice had any effect on the number of larvae aborted ( 2 = 6.31; df = 4; P = 0.18). Numerically, water caused the lowest abortion rate of all of the solvents (Table D-1) However, the failure of amitraz to dissolve in water only makes it an appropriate solvent for imidacloprid. Methanol, et hanol, and acetone all induced a larger number of aborted larvae. Even though there were no stat istically significant differences between solvents det ected in this investig ation, the large number of larvae aborted from the field experiment (Appendix C) suggests th at applying pesticides topically to larvae inside honey bee colonies is not an e ffective method for testing pesticides.
87 Table D-1. Percentage of larvae aborted according to solven t treatment applied topically. Abortion was determined by the number of em pty cells that contained treated larvae 24 h earlier. Solvent Aborted larvae Untreated 9.3 9.3 (5)a Acetone 37.8 18.8 (5)a Ethanol 51.5 16.8 (5)a Methanol 25 16.1 (5)a Water 6.4 6.4 (5)a Data are mean std. error (n) where n = num ber of individuals from all queen sources. Columnar means followed by the same letter are not different at 0.05. Means were compared using Fishers LSD tests.
88 LIST OF REFERENCES Abbott, V. A., J. L. Nadeau, H. A. Higo, and M. L. Winston. 2008. Lethal and sublethal effects of imidacloprid on Osmia lignaria and clothianidin on Megachile rotundata (Hymenoptera: Megachilidae). J. Econ. Entomol. 101: 784-796. Aizen, M. A., and P. Feinsinger. 1994. Habitat fragmentation, native insect pollinators, and feral honey bees in Argentine 'Ch aco Serrano'. Ecol. Appl. 4: 378-392. Arathi, H. S., I. Burns, and M. Spivak. 2000. Ethology of hygienic beha vior in the honey bee Apis mellifera L. (Hymenoptera: Apidae): behavior al repertoire of hygienic bees. Ethology 106: 365-379. Aupinel, P., D. Fortini, B. Michaud, F. Marolleau, J.-N. Tasei, and J.-F. Odoux. 2007. Toxicity of dimethoate and fenoxycarb to honey bee brood ( Apis mellifera ), using a new in vitro standardized feeding method. Pest Manag. Sci. 63: 1090-1094. Aupinel, P., D. Fortini, H. Dufour, J.-N. Ta sei, B. Michaud, J.-F. Odoux, and M.-H. PhamDelegue. 2005. Improvement of artificial feeding in a standard in vitro method for rearing Apis mellifera larvae. Bull. Insectology 58: 107-111 ( http://www.bulleti nofinsectology.org, Dec 2008). Beetsma, J., W. J. Boot, and J. Calis. 1999. Invasion behavior of Varroa jacobsoni O ud.: from bees into brood cells. Apidologie 30: 125-140. Bennett, S. H. 1957. The behavior of systemic insect icides applied to plants. Annu. Rev. Entomol. 2: 279-296. Berenbaum, M., P. Bernhardt, S. Buchmann, N. W. Calderone, P. Goldstein, D. W. Inouye, P. Kevan, C. Kremen, R. A. Medellin, T. Ri cketts, G. E. Robinson, A. A. Snow, S. M. Swinton, L. B. Thein, and F. C. Thompson. 2007. Status of pollinators in North America. The National Academies Press, Washington, D.C. Bienefeld, K., M. Haberl, and J. Radtke. 1998. Does the genotype of honeybee brood influence the attractiveness for Varroa jacobsoni and/or the reproduction of this parasite? Hereditas 129: 125-129. Bitterman, M. E., R. Menzel, A. Fietz, and S. Schafer. 1983. Classical conditioning of proboscis extension in honeybees (Apis mellifera ). J. Comp. Psychol. 97: 107-119. Bonmatin, J. M., I. Moineau, R. Charvet, C. Fleche, M. E. Colin, and E. R. Bengsch. 2003. A LC/APCIMS/MS method for analysis of imidacloprid in soils, in plants, and in pollens. Anal. Chem. 75: 2027-2033. Boot, W. J., and J. N. M. Calis. 1991. A method to obtain dated brood in honeybee colonies. Bee World 72: 19-21.
89 Boot, W. J., D. J. A. Sisselaar, J. N. M. Calis, and J. Beetsma. 1994. Factors affecting invasion of Varroa jacobsoni (Acari: Varroidae) into honeybee, Apis mellifera (Hymenoptera: Apidae), brood cells Bull. Entomol. Res. 84: 3-10. Bortolotti, L., R. Montanari, J. Marcelino, P. Medrzycki, S. Maini, and C. Porrini. 2003. Effects of sublethal imidaclopr id doses on the homing rate and foraging activity of honey bees. Bull. Insectology 56: 63-67. (http://www.bulletinofinsectology.org Dec 2008). Bow en-Walker, P. L., S. J. Martin, and A. Gunn. 1999. The transmission of deformed wing virus between honeybees ( Apis mellifera L.) by the ectoparasitic mite Varroa jacobsoni Oud. J. Invertebr. Pathol. 73: 101-106. Brandt, A., M. E. Scharf, J. H. F. Pedra, G. Holmes, A. Dean, M. Kreitman, and B. R. Pittendrigh. 2002. Differential expression and induction of two Drosophila cytochrome P450 genes near the Rst(2)DDT locu s. Insect Mol. Biol. 11: 337-341. Burley, L. M. 2007. The effects of miticides on the reproductive physiology of honey bee ( Apis mellifera L.) queens and drones. M.S. thesis, Vi rginia Polytechnic Institute and State Univeristy, Blacksburg. Caron, D. M. 1999. Honey bee biology and beekeeping. Wicwas Press, Cheshire, CT. Caron, D. M. 2001. Neotropical pollinati on: crisis, crossroads, and conservation, pp. 113-156. In C. S. Stubbs and F. A. Drummond (eds.), Bees and crop pollination-crisis, crossroads, conservation. Entomological Society of America, Lanham, MD. Castle, S. J., F. J. Byrne, J. L. Bi, and N. C. Toscano. 2005. Spatial and temporal distribution of imidacloprid and thiamethoxa m in citrus and impact on Homalodisca coagulata populations. Pest Manag. Sci. 61: 75-84. Chapman, R. F. 1998. The insects: structure and functi on. Cambridge University Press, New York, NY. Chauzat, M. P., J. P. Faucon, A. C. Martel J. Lachaize, N. Cougoule, and M. Aubert. 2006. A survey of pesticide residues in pollen load s collected by honey bees in France. J. Econ. Entomol. 99: 253-262. Claudianos, C., H. Ranson, R. M. Johnson, S. Biswas, M. A. Schuler, M. R. Berenbaum, R. Feyereisen, and J. G. Oakeshott. 2006. A deficit of detoxifica tion enzymes: pesticide sensitivity and environmental response in the honeybee. Insect Mol. Biol. 15: 615-636. Cox, C. 2001. Imidacloprid. J. Pestic. Reform 21: 7. Davis, A. R. 1989. The study of insecticide poisoni ng of honeybee brood. Bee World 70: 163174.
90 Davis, A. R., and R. W. Shuel. 1988. Distribution of 14C-labelled carbofuran and dimethoate in royal jelly, queen larvae and nurse honeybees. Apidologie 19: 37-50. Davis, A. R., K. R. Solomon, and R. W. Shuel. 2000. Laboratory studies of honeybee larval growth and development as affected by systemic insecticides at adult-sublethal levels. J. Apic. Res. 27: 146. Decourtye, A., J. Devillers, S. Cluzeau, M. Charreton, and M. H. Pham-Delcgue. 2004. Effects of imidacloprid and deltamethrin on associative learning in honeybees under semi-field and laboratory conditions. Ecotoxicol. Environ. Saf. 57: 410-419. Decourtye, A., J. Devillers, E. Genecque, K. L. Menach, H. Budzinski, S. Cluzeau, and M. H. Pham-Delgue. 2005. Comparative sublethal toxicity of nine pesticides on olfactory learning performances of the honeybee Apis mellifera Arch. Environ. Contam. Toxicol. 48: 242-250. DeJong, D. 1997. Mites: varroa and other parasites of brood, pp. 281-327. In R. A. Morse and K. Flottum (eds.), Honey bee pests, predators, and diseases, 3rd ed. A.I. Root Company, Medina, OH. Delaplane, K. S., and D. F. Mayer. 2000. Crop pollination by bees. CABI Publishing, New York, NY Desneux, N., A. Decourtye, and J.-M. Delpuech. 2007. The sublethal effects of pesticides on beneficial arthropods. Annu. Rev. Entomol. 52: 81-106. Devlin, S. M. 2001. Comparative analyses of sampling methods for varroa mites ( Varroa destructor Anderson and Trueman) on honey bees ( Apis meiliferra L.). M.S. thesis, Simon Fraser University, Burnaby. Doane, C. C., and D. M. Dunbar. 1973. Field evaluation of insectic ides against the gypsy moth and elm spanworm and repellent action of chlordimeform. J. Econ. Entomol. 66: 11871189. Duay, P., D. De Jong, and W. Engels. 2003. Weight loss in drone pupae ( Apis mellifera ) multiply infested by Varroa destructor mites. Apidologie 34: 61-65. Eccleston, C. H. 2007. The case of the disappearing honeybees: An environmental harbinger? Environ. Qual. Manag. 17: 11. Elbert, A., B. Becker, J. Hartwig, and C. Erdelen. 1991. Imidacloprida new systemic insecticide. Pflanzenschu tz-Nachrichten 44: 113-135.
91 Ellis, J. 2007. Colony Collapse Disorder (CCD) in honey bees. Entomology and Nematology Department, Florida Cooperative Extension Se rvice, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL, Publication ENY-150. ( http://edis.ifas.ufl.edu Dec 2008). Ellis, J. D., and K. S. Delaplane. 2001. A scientific note on Apis mellifera brood attractiveness to Varroa destructor as affected by the chemotherapeutic history of the brood. Apidologie 32: 449-450. Embrey, M. 2008. A follow-up on colony collapse disorder. The IR Project 39: 4. ( http://www.ir4.rutgers.edu Dec 2008). Emmett, B. J., and B. M. Archer. 1980. The toxicity of diflubenzuron to honey bee ( Apis mellifera L.) colonies in apple orch ards. Plant Pathol. 29: 177-183. Evans, P. D., and J. D. Gee. 1980. Action of formamidine pesticides on octopamine receptors. Nature 287: 60-62. Feyereisen, R. 1999. Insect P450 enzymes. Annu. Rev. Entomol. 44: 507-533. Feyereisen, R., 2005. Insect cytochrome P450, pp. 1-77. In L. I. Gilbert, K. Iatrou, and S.S. Gill (eds.), Comprehensive molecula r insect science: biochemi stry and molecular biology, vol. 4. Elsevier, Oxford, UK. Fischer, D. L., and A. Chalmers. 2007. Neonicotinoid insecticides and honey bees: technical answers to FAQs, pp. 14. Bayer CropScience. Frazier, M., C. Mullin, J. Frazier, and S. Ashcraft. 2008. What have pesticides got to do with it? American Bee Journal 148: 521-523. Free, J. B. 1963. The flower constancy of honeybees. J. Anim. Ecol. 32: 119-131. Fuchs, S. 1990. Preference for drone brood cells by Varroa jacobsoni Oud: in colonies of Apis mellifera carnica. Apidologie 66. Gallai, N., J.-M. Salles, J. Se ttele, and B. E. Vaissire. Economic valuation of the vulnerability of world agriculture confronted with pollinat or decline. Ecol. Econ. (in press, corrected proof). Gladney, W. J., S. E. Ernst, and R. O. Drummond. 1974. Chlordimeform: a detachmentstimulating chemical for three-host ticks. J. Med. Entomol. 11: 569-72. Grafton-Cardwell, E. E., K. E. Godfrey, M. E. Michaels, C. C. Childers, and P. A. Stansly. 2006. Asian citrus psyllid. University of Califor nia, Division of Agriculture and Natural Resources, Oakland, CA, Publication 8205. (http://anrcatalog.ucdavis.edu Dec 2008).
92 Graham, J. M. e. 1992. The hive and the honey bee. Dadant & Sons, Hamilton, IL. Greatti, M., A. G. Sabatini, R. Barbattini, S. Rossi, and A. Stravisi. 2003. Risk of environmental contamination by the active ingr edient imidacloprid used for corn seed dressing: preliminary results. Bull. Insectology 56: 69-72. ( http://www.bulleti nofinsectology.org, Dec 2008). Greatti, M., R. Barbattini, A. Stravi si, A. G. Sabatini, and S. Rossi. 2006. Presence of the ai im idacloprid on vegetation near corn fields sown with Gaucho dressed seeds. Bull. Insectology 59: 99-103. ( http://www.bulleti nofinsectology.org, Dec 2008). Gullan, P. J., and P. S. Cranston. 2005. The Insects: an outline of entomology. Blackwell Publishing Ltd, Malden, MA. Hillesheim, E., W. Ritter, and D. Bassand. 1996. First data on resistance m echanisms of Varroa jacobsoni (Oud.) against tau-fluvalinate. Exp. Appl. Acarol. 20: 283-296. Hollingworth, R. M. 1976. Chemistry, biological activity, and uses of formamidine pesticides. Environ. Health Perspect. 14: 57-69. Hollingworth, R. M., and L. L. Murdock. 1980. Formamidine pesticides: octopamine-like actions in a firefly. Science 208: 74-76. Iwasa, T., N. Motoyama, J. T. Ambrose, and R. M. Roe. 2004. Mechanism for the differential toxicity of neonicotinoid insecticides in the honey bee, Apis mellifera Crop Prot. 23: 371-378. Johansen, C. A., and D. F. Mayer. 1990. Pollinator protection: a bee & pesticide handbook. Wicwas Press, Cheshire, CT. Johnson, R. 2007. Recent honey bee colony declines, CRS Report for Congress, order code RL33938. ( http://www.fas.org, Dec 2008). Kegley, S. E., B. R. Hill, S. Orme, and A. H. Choi. 2008. PAN Pesticid e Database, Pesticide Action Network, San Francisco, CA. ( http://www.pesticideinfo.org Dec 2008). Kevan, P. G. 1975. Forest app lication of the insecticide fenitrothion and its effect on wild bee pollinators (Hymenoptera: Apoidea) of lo wbush blueberries (Vaccinium spp.) in Southern New Brunswick, Canada. Biol. Conservat. 7: 301-309. Koeniger, N. 1987. Die stliche honigbiene und ihre milbe Varroa jacobsoni Imkerfreund 42: 303-306. Korta, E., A. Bakkali, L. A. Berrueta, B. Gallo, F. Vicente, and S. Bogdanov. 2003. Determination of amitraz and other acaricide residues in beeswax. Anal. Chim. Acta 475: 97-103.
93 Korta, E., A. Bakkali, L. A. Berrueta, B. Gallo, F. Vicente, V. Kilchenmann, and S. Bogdanov. 2001. Study of acaricide stability in h oney: characterization of amitraz degradation products in honey and bees wax. J. Agric. Food Chem. 49: 5835-5842. Krischik, V. A., A. L. Landmar k, and G. E. Heimpel. 2007. Soil-applied imidacloprid is translocated to nectar and kills nectar-feeding Anagyrus pseudococci (Girault) (Hymenoptera: Encyrtidae). E nviron. Entomol. 36: 1238-1245. Landis, W. G., and M.-H. Yu. 1995. Introduction to environmental toxicology: impacts of chemicals upon ecological systems. Lewis Publishers, Boca Raton, FL. Lodesani, M., A. Pellacani, S. Bergomi, E. Carpana, T. Rabitti, and P. Lasagni. 1992. Residue determination for some products used against varroa infestation in bees. Apidologie 23: 257-272. Marchetti, S., R. Barbattini, and M. DAgaro. 1984. Comparative effectiveness of treatments used to control Varroa jacobsoni Oud. Apidologie 15: 363-377. Martel, A. C., S. Zeggane, C. Aurires, P. Drajnudel, J. P. Faucon, and M. Aubert. 2007. Acaricide residues in honey and wax after treatment of honey bee colonies with Apivar or Asuntol 50. Apidologie 38: 534-544. Matsumura, F. 1975. Toxicology of insecticides. Plenum Press, New York, NY. Maus, C., G. Cure, and R. Schmuck. 2003. Safety of imidacloprid seed dressings to honey bees: a comprehensive overview and compilatio n of the current state of knowledge. Bull. Insectology 56: 51-58. ( http://www.bulleti nofinsectology.org Dec 2008). Medrz ycki, P., R. Montanari, L. Bortolotti, A. G. Sabatini, S. Maini, and C. Porrini. 2003. Effects of imidacloprid administered in sub-lethal doses on honey bee behavior: laboratory tests. Bull. Insectology 56: 59-62. ( http://www.bulletinofinsectology.org Dec 2008). Morse, R. A., and N. W. Calderone. 2000. The value of honey bees as pollinators of U.S. crops in 2000. Bee Culture 128: 1-15. Morse, R. A., and K. Flottum. 1997. Honey bee pests, predators, and diseases. A.I. Root Company, Medina, OH. (NASS) National Agricultural Statistics Service. 1976-2008. Honey. United States Department of Agriculture, Washington, D.C. ( http://www.nass.usda.gov Dec 2008). Nauen, R., U. Ebbinghaus-Kintscher, and R. Schmuck. 2001. Toxicity and nicotinic acety lcholine receptor interaction of imidacloprid and its metabolites in Apis mellifera (Hymenoptera: Apidae). Pest Manag. Sci. 57: 577-586.
94 Oldroyd, B. P. 2007. What's killing American hone y bees? PLoS Biology 5: 1195-1199. ( http://biology.plosjournals.org Dec 2008). Peng, Y. S. C., E. Mussen, A. Fong, M. A. Montague, and T. Tyler. 1992. Effects of chlortetracy cline on honey-bee worker larvae re ared in vitro. J. Invertebr. Pathol. 60: 127-133. Pettis, J. S., A. M. Collins, R. Wilbanks, and M. F. Feldlaufer. 2004. Effects of coumaphos on queen rearing in the honey bee, Apis mellifera Apidologie 35: 605-610. Piccirillo, G. A., and D. De Jong. 2004. Old honey bee brood combs are more infested by the mite Varroa destructor than are new brood combs. Apidologie 35: 359-364. Pimentel, D. 2005. Environmental and economic costs of th e application of pe sticides primarily in the United States. Environ. Dev. Sustain. 7: 229-252. Porrini, C., A. G. Sabatini, S. Girotti, F. Fini L. Monaco, G. Celli, L. Bortolotti, and S. Ghini. 2003. The death of honey bees and enviro nmental pollution by pesticides: the honey bees as biological indica tors. Bull. Insectology 56: 147-152. ( http://www.bulleti nofinsectology.org, Dec 2008). Proctor, M., P. Yeo, and A. Lack. 1996. The natural his tory of pollination. Timber Press, Portland, OR. Rath, W. 1999. Co-adaptation of Apis cerana Fabr. and Varroa jacobsoni Oud. Apidologie 30: 97-110. Rath, W., and W. Drescher. 1990. Response of Apis cerana Fabr. towards brood infested with Varroa jacobsoni Oud. and infestation rate of col onies in Thailand. Apidologie 21: 311321. Rial-Otero, R., E. M. Gaspar, I. Moura, and J. L. Capelo. 2007. Gas chromatography mass spectrometry determination of acaricides from honey after a new fast ultrasonic-based solid phase micro-extraction sample treatment. Talanta 71: 1906-1914. Santiago, G. P., G. Otero-Colina, D. M. Sanc hez, M. E. R. Guzman, and R. Vandame. 2000. Comparing effects of three acaricides on Varroa jacobsoni (Acari: Varroidae) and Apis mellifera (Hymenoptera: Apidae) using two applica tion techniques. Fla. Entomologist 83: 468-476. SAS Institute. 2008. PROC user's manual, version 8 ed. SAS Institute, Cary, NC. Scharf, M. E. 2003. Neurological effects of insecticid es, pp. 395-399. In D. Pimentel (ed.), Encyclopedia of Pest Management, vol 2. CRC Press, Boca Raton, FL.
95 Schmuck, R., R. Schoning, A. Stork, and O. Schramel. 2001. Risk posed to honeybees ( Apis mellifera L. Hymenoptera) by an imidacloprid seed dressing of sunflowers. Pest Manag. Sci. 57: 225-238. Shen, M., L. Cui, N. Ostiguy, and D. Cox-Foster. 2005. Intricate transmission routes and interactions between picorna-like viruses (Kashmir bee virus and sacbrood virus) with the honeybee host and the parasitic varro a mite. J. Gen. Virol. 86: 2281-2289. Sicbaldi, F., G. A. Sacchi, M. Trevisan, and A. A. M. Del. 1997. Root uptake and xylem translocation of pesticides from differen t chemical classes. Pestic. Sci 50: 111-119. Southwick, E. E., and L. Southwick, Jr. 1992. Estimating the economic value of honey bees (Hymenoptera: Apidae) as agricultural pollinat ors in the United States. J. Econ. Entomol. 85: 621-633. Spreafico, M., F. R. Eordegh, I. Bernardinelli, and M. Colombo. 2001. First detection of strains of Varroa destructor resistant to coumaphos: results of laboratory tests and field trials. Apidologie 32: 49-56. Stadler, T., D. M. Gines, and M. Buteler. 2003. Long-term toxicity assessment of imidacloprid to evaluate side effects on honey bees exposed to treated sunflower in Argentina. Bull. Insectology 56: 77-81. ( http://www.bulleti nofinsectology.org Dec 2008). Stephen, W. P. 1955. Alfalfa pollination in M anitoba. J. Econ. Entomol. 48: 543-548. Stokstad, E. 2007. Entomology: The case of the empty hives. Science 316: 970. Suchail, S., D. Guez, and L. P. Belzunces. 2000. Characteristics of imidacloprid toxicity in two Apis mellifera subspecies. Environ. Toxicol. Chem. 19: 1901-1905. Suchail, S., D. Guez, and L. P. Belzunces. 2001. Discrepancy between acute and chronic toxicity induced by imidacloprid and its metabolites in Apis mellifera. Environ. Toxicol. Chem. 20: 2482-2486. Vandenberg, J. D., and H. Shimanuki. 1987. Technique for rearing worker honeybees in the laboratory. J. Apic. Res. 26: 90-97. Villa, S., M. Vighi, A. Fini zio, and G. Bolchi Serini. 2000. Risk assessment for honeybees from pesticide-exposed pollen. Ecotoxicology 9: 287-297. Wallner, K. 1999. Varroacides and their residues in bee products. Apidologie 30: 235-248. Wang, H. Y., Y. Yang, J. Y. Su, J. L. Shen, C. F. Gao, and Y. C. Zhu. 2008. Assessment of the impact of insecticides on Anagrus nilaparvatae (Pang et Wang) (Hymenoptera: Mymanidae), an egg parasitoid of the rice planthopper, Nilaparvata lugens (Hemiptera: Delphacidae). Crop Prot. 27: 514-522.
96 Wauchope, R. D. 1978. The Pesticide content of surface wate r draining from agricultural fields: a review. J. Environ. Qual. 7: 459. Wilson, W. T., J. S. Pettis, C. E. Henderson, and R. A. Morse. 1997. Tracheal mites, pp. 253278. In R. A. Morse and K. Flottum (eds.), Honey bee pests, predators, and diseases, 3rd ed. A.I. Root Company, Medina, OH. Winston, M. L. 1987. The biology of the honey bee. Harvard University Press, Cambridge, MA. Winston, M. L., J. A. Dropkin, and O. R. Taylor. 1981. Demography and life history characteristics of two honey bee races ( Apis mellifera ). Oecologia 48: 407-413. Wolfenbarger, D., E. Cantu, P. D. Lingren, and A. R. Guerra. 1974. Activity of chlordimeform-HCl and chlo rdimeform against arthropods attacking cotton. J. Econ. Entomol. 67: 445. Woyke, J. 1976. Brood-rearing efficiency and absconding in Indian honeybees. J. Apic. Res. 15: 133. Yang, X., and D. L. Cox-Foster. 2005. Impact of an ectoparasite on the immunity and pathology of an invertebrate: evidence for host immunosuppression and viral amplification. Proc. Natl. Acad. Sci. USA 102: 7470-7475. Yu, S. J. 2008. The toxicology and biochemistry of insecticides. CRC Press, Boca Raton, FL. Zar, J. H. 1996. Biostatistical analysis. Prenti ce Hall, Upper Saddle River, NJ.
97 BIOGRAPHICAL SKETCH I am a masters student in the Univers ity of Floridas Entomology and Nematology program. I received my bachelors degree at West Virginia University with a major in biology and minor in geology. Afterwards, I was awarde d an internship with the Conservancy of Southwest Florida where I studied sea turtle bi ology. I then became employed with the United States Geological Survey, Biological Resource Division, as an en tomology research assistant. While at the USDA, I participated in a number of research projects involved in the control of invasive pests and conservation of native fauna. In the future, I pl an to pursue a Ph.D. in ecology.