ENVIRONMENTAL CONCENTRATIONS, FATE, AND RISK ASSESSMENT OF INSECTICIDES USED FOR ADULT MOSQUITO MANAGEMENT by Jerome Joseph Schleier III A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Land Resources and Environmental Sciences MONTANA STATE UNIVERSITY Bozeman, Montana November 2008 ©COPYRIGHT by Jerome Joseph Schleier III 2008 All Rights Reserved ii APPROVAL of a thesis submitted by Jerome Joseph Schleier III This thesis has been read by each member of the thesis committee and has been found to be satisfactory regarding content, english usage, format, citation, bibliographic style, and consistency, and is ready for submission to the Division of Graduate Education. Dr. Robert K. D. Peterson Approved for the Department of Land Resources and Environmental Sciences Dr. Bruce D. Maxwell Approved for the Division of Graduate Education Dr. Carl A. Fox iii STATEMENT OF PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirements for a master’s degree at Montana State University, I agree that the Library shall make it available to borrowers under rules of the Library. If I have indicated my intention to copyright this thesis by including a copyright notice page, copying is allowable only for scholarly purposes, consistent with “fair use” as prescribed in the U.S. Copyright Law. Requests for permission for extended quotation from or reproduction of this thesis in whole or in parts may be granted only by the copyright holder. Jerome Joseph Schleier III November 2008 iv ACKNOWLEDGEMENTS I want to thank my advisor, Dr. Robert Peterson for giving me an opportunity to obtain my masters and for his mentoring and guidance during my studies. I want to thank my other committee members, Dr. David Weaver and Dr. Kevin O’Neill, for their guidance and support. Thanks to team CoBRA, specifically Ryan Davis, Marjolein Schat, Leslie Shama, Courtney Pariera Dinkins, and Loren Barber for their valuable friendship, support, and guidance. I want to thank Dr. Paula Macedo for helping to write the grant that has funded my research. I would like to thank my parents, Susan and Jerome Schleier Jr., and my grandparents, Pete and Geri Kolleas, for all their support that they have given me throughout my life and helping me get to where I am today. The Department of Land Resources and Environmental Sciences at Montana State University has also been a great support and has provided a wonderful learning environment. I want to also thank those who generously provided field and technical support: Dr. Paula Macedo, Gary Goodman, Dave Brown, the staff of Sacramento-Yolo Mosquito and Vector Control District, Dave Whitesell, ADAPCO Vector Control Services, Janice Stroud, Dennis Candito, Ryan Arkoudas, Mike Mazzarelli, Peter Connelly, Dr. Bill Feiler, John Paige, Heidi Hickes, Angela Schaner, Jona Verreth, Dr. Graham White, Dr. Stanton Cope, and Bill Svoboda. I would also like to thank Lee Beloit and Randall Olsen for allowing me to use their land for these experiments. My research was supported by grants and support from the U.S. Armed Forces Pest Management Board’s Deployed War Fighter Protection Program and the California Mosquito and Vector Control Research Foundation. v TABLE OF CONTENTS 1. INTRODUCTION ......................................................................................................... 1 Integrated Pest Management Tactics ............................................................................. 3 Ultra-Low-Volume Insecticide Applications ................................................................. 3 Insecticides ................................................................................................................ 4 Environmental Fate of Insecticides Used in Mosquito Management ............................ 6 Effects of Ultra-Low-Volume Applications on Non-Target Organisms ..................... 11 Risk Assessment .......................................................................................................... 12 Previous Risk Assessments, Reports, and Biomonitoring Studies .............................. 13 Objectives .................................................................................................................... 15 2. ENVIRONMENTAL CONCENTRATIONS AND FATE OF NALED AND PERMETHRIN USED FOR ADULT MOSQUITO MANAGEMENT ..................... 18 Abstract ........................................................................................................................ 18 Introduction .................................................................................................................. 19 Materials and Methods ................................................................................................. 22 Statistical Analysis .................................................................................................. 26 Non-detectable Concentrations ............................................................................... 27 Results .......................................................................................................................... 27 Permethrin ............................................................................................................... 27 Naled ....................................................................................................................... 28 Discussion .................................................................................................................... 29 3. A PROBABILISTIC RISK ASSESSMENT OF NALED AND PERMETHRIN USING ACTUAL ENVIRONMENTAL CONCENTRATIONS ............................... 44 Abstract ........................................................................................................................ 44 Introduction .................................................................................................................. 45 Materials and Methods ................................................................................................. 48 Problem Formulation .............................................................................................. 48 Hazard Identification .............................................................................................. 49 Toxicity and Dose-response Relationships ............................................................. 49 Environmental Concentrations................................................................................ 49 Acute Exposure ....................................................................................................... 50 Risk Characterization .............................................................................................. 54 Probabilistic Analysis ............................................................................................. 54 Results .......................................................................................................................... 55 Discussion .................................................................................................................... 57 vi TABLE OF CONTENTS – CONTINUED 4. TOXICITY AND RISK ASSESSMENT OF PERMETHRIN AND NALED FOR NON-TARGET TERRESTRIAL INSECTS ............................................................... 68 Abstract ........................................................................................................................ 68 Introduction .................................................................................................................. 69 Materials and Methods ................................................................................................. 71 Laboratory Bioassay ............................................................................................... 71 Data Analysis .......................................................................................................... 73 Field Bioassays ....................................................................................................... 74 Non-Target Insect Risk Assessment ....................................................................... 75 Results .......................................................................................................................... 75 Toxicity Testing ...................................................................................................... 75 Field Bioassay ......................................................................................................... 76 Risk Assessment ..................................................................................................... 76 Discussion .................................................................................................................... 77 5. ENVIRONMENTAL CONCENTRATIONS, FATE, AND RISK ASSESSMENT OF PYRETHRINS AND PIPERONYL BUTOXIDE AFTER AERIAL ULTRA-LOWVOLUME APPLICATIONS FOR ADULT MOSQUITO MANAGEMENT............ 81 Abstract ........................................................................................................................ 81 Introduction .................................................................................................................. 82 Materials and Methods ................................................................................................. 85 Mosquito Bioassay .................................................................................................. 88 Results .......................................................................................................................... 88 Discussion .................................................................................................................... 91 6. CONCLUSION .......................................................................................................... 101 REFERENCES CITED ................................................................................................... 105 vii LIST OF TABLES Table Page 1. Mean permethrin deposition at 25, 50, and 75 m from the spray source, control, and spike concentrations 1, 12, and 24 h after application in 2007 and 2008 ......................................................................................... 32 2. Mean naled deposition at ground level 25, 50, and 75 m from the spray source, control, and spike concentrations 1 and 12 h after application in 2007 and 2008 ......................................................................................... 33 3. Mean naled deposition at 1.25 m above the ground 25, 50, and 75 m from the spray source, control, and spike concentrations 1 and 12 h after application in 2007 and 2008 ......................................................................................... 33 4. Custom input distributions for deposition onto surfaces and aerial concentrations for permethrin as measured in Chapter 2.............................................. 61 5. Custom input distributions for deposition onto surfaces and aerial concentrations for naled as measured in Chapter 2 ...................................................... 61 6. Assumptions for body weight, respiration rate, and frequency of hand-to-mouth activity for each group assessed....................................................... 62 7. Acute total potential exposure (PE) means, 50th, 90th, and 95th percentile confidence intervals for each group and chemical assessed ........................ 63 8. Acute risk quotient (RQ) means and 50th, 90th, and 95th percentile confidence intervals for each group and chemical assessed ......................................... 63 9. Sensitivity analysis (percent contribution of the input variable to the exposure variance) of uncertain factors for permethrin and naled.......................... 64 10. The 24 h lethal concentrations that kill 50% of a population (LC50) for house crickets treated with Permanone, Permanone and piperonyl butoxide (PBO), technical permethrin, technical permethrin and PBO, and naled ..................................................................................................................... 80 viii LIST OF TABLES – CONTINUED Table Page 11. Risk quotients (deposition concentration / LC50) for the actual environmental concentration (AEC) and estimated environmental concentrations from ISCST3for Permanone, Permanone + piperonyl butoxide (PBO), technical grade permethrin, technical grade permethrin + PBO, and naled at 24 h ......................................................................................................... 80 12. Analytical detection limits for water and deposition pads for each compound ...................................................................................................... 95 13. Mean piperonyl butoxide ground deposition concentrations 1, 12, 24, and 36 h after application in Princeton and Colusa, CA ............................. 95 14. Mean piperonyl butoxide water concentrations 1 h before the spray event and 1, 12, 24, and 36 h after the spray event from irrigation ditches located in Princeton and static ponds in Colusa, CA ...................... 95 15. Mean percent mortality for Culex pipiens and Culex tarsalis in bioassay cages in Princeton and Williams, CA ...................................................... 96 16. Mean percent mortality for Culex pipiens and Culex tarsalis in bioassay cages in Colusa and Williams, CA........................................................... 96 ix LIST OF FIGURES Figure Page 1. Site layout with wind direction, driving direction and locations where spray began and ended for the application of naled in 2007......................................... 34 2. Site layout with wind direction, driving direction and locations where spray began and ended for the application of permethrin in 2007 ................................ 35 3. Site layout with wind direction, driving direction and locations where spray began and ended for the application of naled in 2008......................................... 36 4. Site layout with wind direction, driving direction and locations where spray began and ended for the application of permethrin in 2008 ................................ 37 5. Ultra-low-volume machine spraying naled ................................................................... 38 6. Sampler layout for naled and permethrin applications with deposition samplers 25, 50 and 75 m and air sampler at 25 m from the spray source. .................. 39 7. Mean concentrations of permethrin in 2007 25, 50, and 75 m from the spray source 1, 12, and 24 h after application ......................................................... 40 8. Mean concentrations of permethrin in 2008 25, 50, and 75 m from the spray source 1, 12, and 24 h after application ......................................................... 41 9. Mean concentrations of naled at ground level and 1.25 m above the ground in 2007 25, 50, and 75 m from the spray source 1 and 12 h after application ............................................................................................................ 42 10. Mean concentrations of naled at ground level and 1.25 m above the ground in 2008 25, 50, and 75 m from the spray source 1 and 12 h after application .......................................................................................................... 43 11. Exposure routes for all groups assessed...................................................................... 65 12. Toddler risk quotients (RQ) for permethrin from Peterson et al. (2006) using estimated environmental concentrations, 95th percentile confidence interval RQ of Schleier III et al. (2008c) using estimated environmental concentrations, and the 95th percentile confidence interval RQ of the current study using environmental concentrations .................................................................. 66 x LIST OF FIGUES – CONTINUED Figure Page 13. Toddler risk quotients (RQ) for naled from Peterson et al. (2006) using estimated environmental concentrations and the 95th percentile confidence interval RQ of the current study using environmental concentrations ...................................................................................... 67 14. Locations where deposition and water samples were taken in Princeton, CA ........... 97 15. Locations where depositionand water samples were taken in Colusa, CA................. 98 16. Change in piperonyl butoxide concentrations on deposition pads from 1 to 36 h after the spray with exponential decay curve for Princeton and Colusa, CA .......................................................................................... 99 17. Change in piperonyl butoxide concentrations in water from 1 to 36 h after the spray with exponential decay curve for Princeton and Colusa, CA. .. 100 xi ABSTRACT One of the most effective ways of managing adult mosquitoes that vector human and animal pathogens is the use of ultra-low-volume (ULV) insecticides. Due to a lack of studies examining the environmental fate of ULV insecticides and because previous risk assessments have shown that environmental concentrations of insecticides contributed the largest amount of variance to the estimated total exposure, I measured deposition onto surfaces and air concentrations of permethrin and naled. I also conducted risk assessments for human and other non-target organisms using the values I measured. Deposition concentrations of permethrin and naled generally decreased as distance from the spray source increased. Overall, approximately 3.2% of the permethrin and 15% of the naled applied deposited on the ground within 75 m from the spray source 1 h after application. Concentrations of permethrin and naled 12 h after application were not significantly different than concentrations 1 h after application. The results of my probabilistic human-health risk assessment using actual environmental concentrations showed that previous risk assessments overestimated risks. Thus they were conservative in protecting human health. The non-target risk assessment and field bioassay using the house cricket, Acheta domesticus (L.), as a surrogate for medium- to large-bodied ground dwelling insects showed that ULV applications most likely would not result in impacts on populations. I also measured actual environmental concentrations of pyrethrins and piperonyl butoxide (PBO) after aerial ULV applications. Pyrethrins were not detected in the water or on deposition pads. However PBO was detected in the water and on deposition samples, but concentrations rapidly decreased to background levels by 36 h after application. The estimated risks of pyrethrins and PBO to aquatic surrogates were lower than those estimated by previous ecological risk assessments. My study is the first to relate actual environmental concentrations of ULV insecticides to estimates of humanhealth risks. Results of my environmental fate studies, human-health and non-target risk assessments, and the current weight of scientific evidence, demonstrate that the risks to humans and the environment after ULV applications of insecticides most likely are below regulatory levels of concern. 1 CHAPTER 1 INTRODUCTION Mosquitoes vector many important human and animal pathogens that cause disease in humans and animals. Some of these diseases include: western, eastern, and Venezuelan equine encephalitis, St. Louis and Japanese encephalitis, malaria, human filariasis, yellow fever, Rift Valley fever, and dengue (Harwood and James 1979). According to the World Health Organization (WHO) annually there are at least 300 million acute cases of malaria each year globally (WHO 1998). In Africa alone, malaria costs $12 billion every year (WHO 1998). Currently in the United States, the mosquitoborne pathogen garnering much attention is West Nile virus (WNV). Since WNV was introduced to New York City in 1999, it has subsequently spread through the Americas, creating human and animal health risks. Currently, WNV is the major cause of encephalitis in North America (Davis et al. 2008). West Nile virus is a Flavivirus that is mosquito-borne and transmitted in cycles between vertebrates (especially birds) and mosquitoes (Dauphin et al. 2004). The WNV strains which were circulating in New York showed a 99.7% homology with a strain that was isolated from Israel in 1998 (Lanciotti et al. 1999). A high degree of similarity between the strains of WNV circulating during 1999 indicates that a single strain of the virus was introduced to the United States (U.S.) (Lanciotti et al. 1999). The main vectors of WNV in North America are Culex mosquitoes and the main amplifying hosts are birds (Dauphin et al. 2004). Birds amplify the virus by building up 2 high viremia levels so that transmission back to mosquitoes can take place during the next blood meal. Mosquito species that are able to replicate the virus and transport the virus to the salivary glands are considered to be competent vectors. Horses and humans can be adversely affected by WNV, but are considered to be dead-end hosts for the virus (Dauphin et al. 2004). Dead-end hosts are infected individuals that amplify a low amount of viremia for a short duration, which means that they cannot pass the virus back to a mosquito (Bunning et al. 2002). However, nonviremic transmission, which is the transmission of a pathogen from vector to vector without the need for propagation in the host, has been documented in mice (Higgs et al. 2005). WNV is not only transmitted by mosquitoes, but has been documented in people who have received blood donations and organ transplants from an infected person (Harrington et al. 2003; Iwamoto et al. 2003). The virus has not progressed southward at a rapid pace in North America as would be expected if migratory birds were transporting the virus, but has moved in all directions which matches year-round resident bird populations that disperse around the same region (Rappole and Hubalek 2003). However, in Europe there is evidence that the virus may be dispersed by migratory birds (Rappole et al. 2000). In 1998, white storks (Ciconia ciconia (L.)) migrating from their breeding grounds in Europe to Israel were found to be carrying four different WNV isolates (Malkinson et al. 2002). WNV is thought to most significantly affect the avian family Corvidae (crows, magpies, jays, and ravens). Finding dead corvids is often an indication that WNV is present in the population of mosquitoes and precedes human WNV cases (Eidson et al. 2001). 3 Integrated Pest Management Tactics Reducing contact between infected mosquitoes and humans and animals is the only way to reduce infection rates, morbidity, and mortality due to mosquito-borne pathogens like WNV (Marfin and Gubler 2001). Mosquito control districts use surveillance to determine population sizes, and when populations exceed a threshold the decision to treat is made (CDC 2003). Additionally mosquito control practitioners use the media and education to communicate the risks of mosquito-borne pathogens to the public. There are several different tactics in an integrated pest management framework, which mosquito management practitioners can utilize. These tactics include personal protective measures, biological control with bacterial agents or fish, growth regulators, transgenic mosquitoes, and ultra-low-volume (ULV) insecticide spraying. However, there are risks to humans and non-target organisms associated with each tactic (Atkinson et al. 2001; Peterson et al. 2006; Davis et al. 2007; Macedo et al. 2007; Peterson and Sing 2007; Antwi et al. 2008; Schleier III et al. 2008b; Schleier III et al. 2008c; Schleier III et al. 2008d). Ultra-Low-Volume Insecticide Applications Application methods for ULV aerosols have been the world-wide standard in mosquito control for more than 30 years (Mount et al. 1996; Mount 1998). Ultra-lowvolume is the minimum effective volume of insecticide that is used as a space spray for adult mosquitoes. Small droplets from 5 to 25 μm are the optimum size to impinge on 4 and knock-down flying adult mosquitoes (Weidhaas et al. 1970; Lofgren et al. 1973; Mount 1998). There are many advantages to using ULV as opposed to non-thermal or thermal aerosol insecticides which were traditionally diluted in fuel oil (Mount 1998). The advantages of ULV aerosols include lower cost, elimination of fuel oil, increased effective payload, rapid application, increased safety by the elimination of the dense fuel oil and pesticide fog, lowered environmental impact, and lower non-target impact (Mount 1998). Insecticides The insecticides currently registered by the U.S. Environmental Protection Agency (USEPA) for mosquito management are malathion (O,O-dimethyl dithiophosphate of diethyl mercaptosuccinate) and naled (1,2-dibromo-2, 2-dichloroethyl dimethyl phosphate) (organophosphate), pyrethrins ((Z)-(S)-2-methyl-4-oxo-3-(penta2,4-dienyl)cyclopent-2-enyl (1R,3R)-2,2dimethyl-3-(2-methylprop-1enyl)cyclopropanecarboxylate) which is a mixture of cinerin I and II, jasmolin I and II, and pyrethrins I and II, and d-phenothrin (3-phenoxybenzyl (1R)-cis, transchrysanthemate), permethrin ((3-phenoxyphenyl)methyl 3-(2,2-dichloroethenyl)-2,2dimethylcyclopropane carboxylate), and resmethrin (5-benzyl-3-furylmethyl (1RS,3RS;1RS,3SR)-2,2-dimethyl-3-(2-methylprop-1-enyl) Cyclopropanecarboxylate) (pyrethroid), along with a synergist piperonyl butoxide (5-[2-(2butoxyethoxy)ethoxymethyl]-6-propyl-1,3-benzodioxole; PBO) which is present in many of the pyrethroid and pyrethrins formulations. 5 Naled, like other organophosphate insecticides, causes toxicity by inhibiting acetylcholin esterase, which inhibits neurological and neuromuscular functions and can also overwhelm the nervous system causing dizziness, nausea, and confusion (USEPA 2002b). Naled has negative effects on non-target organisms like honey bees (Apis mellifera) as well. Honey yield decreased significantly after aerial spraying of Dibrom® concentrate at a rate of 49.25 mL/ha when compared to control hives (Zhong et al. 2003). Average deposition of naled at ground level around the bee hives ranged from 378 to 4,135 μg/m2 (Zhong et al. 2003). However, when naled is applied during the night corresponding to when a mosquito abatement spray would take place, it had no significant effect on honey bees (Caron 1979). Photolytic transformation of naled leads to oxidation and degradation products. Naled is degraded in the atmosphere by reaction with photochemically produced hydroxyl radicals and by hydrolysis and biodegradation (Chevron 1975; USEPA 1999). Significantly more naled was recovered from dark chambers than from direct sunlight with the half-life of naled in direct sunlight being 1.37 h (Tietze et al. 1996b). The primary breakdown product of naled is dichlorvos, which is also a registered pesticide (USEPA 1999). Pyrethroid insecticides are synthetic derivatives of pyrethrin insecticides, which are toxins isolated from the flowers of Chrysanthemum species. Permethrin is a type I pyrethroid insecticide, which acts much like DDT (dichlorodiphenyltrichloroethane) and causes persistent activation of the sodium channels on the neuron, causing uncontrollable muscle firing. 6 Permethrin residues on cotton leaves from ULV spraying of cotton fields after one day caused 87% mortality of honey bees, but after 10 days only caused 6.1% of bee mortality in cage experiments (Waller et al. 1988). Ten days after the ULV spraying, only 2.4% of the initial deposition was present. Permethrin is stable at pHs of 5 and 7 and does not undergo hydrolosis, but at pH 9 it does undergo hydrolysis (USEPA 2005c). Permethrin also breaks down by photolysis and microbial activity (USEPA 2005a). Permethrin has a high Koc value, indicating that it binds strongly to soil and organics. Pyrethrins are natural insecticides isolated from the flowers of the Chrysanthemum species. Pyrethrins work by the same mode of action as pyrethroids. The major route of degradation for pyrethrins in the environment is photolysis, but they also undergo hydrolysis at a pH of 9 (USEPA 2006d). Pyrethrins, like permethrin, also absorb strongly to soil and organics. Environmental Fate of Insecticides Used in Mosquito Management Tucker et al. (1987) examined the effects of 91% malathion and 85% naled sprayed via aerial and truck-mounted ULV on copepods and juvenile fish in coastal areas. They measured the amount of insecticide deposited on boat docks to determine the amount deposited on the surface of the water. In general, as with other studies, the amount of insecticide that was deposited on the ground was 1.4 to 22.3% of the insecticide sprayed. The concentrations of malathion and naled on the ground after truckmounted ULV were 0.0793 and 0.0573 μg/cm2 12 min after application, respectively. 7 Water samples were taken from the test sites after application, and the amount of naled and malathion was 20 μg/L and 5 μg/L, respectively. Levels of naled from truckmounted ULV peaked at 0.71 μg/L about 15 minutes after spraying, but were not detectable after 9 h. There was no significant mortality of copepods and juvenile fish after truck-mounted ULV applications malathion or naled. For aerial applications Tucker et al. (1987) observed concentrations of malathion and naled in the water at 5 and 20.15 μg/L, respectively. The amount of malathion deposited on the ground 24 to 36 min after aerial ULV were 0.381 and 0.492 μg/cm2, respectively. Hennessey et al. (1992) measured terrestrial concentrations of naled at 1.5 and 6 h post aerial ULV applications within and out of a theoretical spray zone. The researchers measured concentrations within the spray zone at 15, 30, and 90 m from the spray path. Concentrations ranged from 0.011 to 0.005 μg/cm2 1.5 h post application, and at 6 h post application concentrations ranged from 0.007 to 0.003 μg/cm2. Outside of the theoretical spray zone, they measured concentrations at 15, 30, 90, 150, and 750 m downwind, and observed concentrations ranging from non-detectable (ND) to 0.009 μg/cm2 1.5 h post application, and concentrations ranging from 0.004 to 0.0005 μg/cm2 6 h post application. Moore et al. (1993) determined the amount of malathion deposited on human targets, and at ground level at different distances. There were no significant concentration differences found in the ground-level patches at 15.2, 30.4, and 91.2 m downwind of the spray head when sprayed at 58.5 g/ha. Actual malathion deposition on filter paper ranged from 0.0059 to 0.075 μg/cm2 which is 1 to 14% of the insecticide 8 sprayed, assuming an even amount of coverage. The amount deposited on the masks of stationary human subjects was not significantly different from a jogger running in the spray path. The amount deposited on the backs of the stationary human targets was significantly less than the front of the targets. Knepper et al. (1996) quantified the amount of malathion and permethrin deposition on sod grass after ULV spraying in a suburban neighborhood in Michigan. Blocks of sod were placed at 7.6, 15.2, 30.4, and 91.4 m from the edge of the road. Malathion was detected on 19 of 32 blocks and permethrin was detected on 20 of 32 blocks. The detection of both insecticides was greatest 15 minutes post treatment at a distance of 7.6 m. The mass of malathion deposited ranged from 0 to 9.22 μg/cm2. The mass of permethrin deposited ranged from 0 to 14.3 μg/cm2. Regression analyses showed that residues at 7.6 m declined logarithmically as a function of time. Detection of the insecticides also declined as a negative exponential function of distance from the road. Tietze et al. (1994) investigated deposition patterns of malathion at 5, 25, 100, and 500 m from the spray truck. The mass of malathion deposited decreased with distance from the spray head. The amount of malathion deposited at 500 m was significantly less than at 5, 25, and 100 m. The amount measured on the ground ranged from 0.3 to 3.8% of the insecticide sprayed. Average deposition at the distances listed above were 0.022, 0.017, 0.016, and 0.002 μg/cm2, respectively. Based on laboratory derived terminal-fall velocities and swing-slide analysis, the distance the droplets traveled horizontally exceeded the targets. 9 Tietze et al. (1996a) investigated deposition rates of truck-mounted ULV malathion around houses, and assessed the impact on non-target species. The house cricket (Acheta domesticus (L.)) also was evaluated as an indicator to determine spray distribution. Long-term deposition of malathion was not determined because samples were collected only at about 30 to 55 minutes after application. The mean deposition of malathion ranged from 0.03 to 0.089 μg/cm2 with deposition decreasing as distance from the spray head increased. Malathion deposition did not differ significantly between the front yard, side of house, behind the house, and the back yard. Cricket mortality ranged from 5.4 to 8.9 % and also decreased as distance from the spray head increased. Stepwise regression indicated that the crickets were a good indicator of spray deposition. Amounts that were collected in the peridomestic environment were greater than those collected in the open field study done by Tietze et al. (1994). Discrepancies may be with the equipment, because two different sprayers were used or possibly because the spray truck had to slow down because of traffic. Pierce et al. (2005) investigated deposition rates and water concentrations of permethrin sprayed via truck-mounted ULV and water concentrations of naled sprayed via aerial ULV in the Florida Keys National Marine Sanctuary. Long-term deposition of permethrin was not determined because samples were collected only at 2 to 4 h after application; however, water concentrations were measured 12 h after application. Permethrin applications were made at 8:00 P.M., while naled applications were made at 7:00 A.M. Deposition concentrations of permethrin ranged from 0.00005 to 0.005 μg/cm2 and water concentrations ranged from 5.1 to 9.1 μg/L. Naled water 10 concentrations ranged from 0.1 to 0.6 μg/L 14 h after application. Distances from the spray source were not given. There are few data on actual environmental concentrations of aerial or truckmounted ULV applications of pyrethrins and PBO. Jensen et al. (1999) found nondetectable (ND) concentrations of pyrethrins and permethrin in water sample from wetlands before and after truck-mounted ULV. Weston et al. (2006) examined water concentrations of pyrethrins and PBO 10 h after the second aerial ULV application of mosquito insecticides and 34 h after the third aerial ULV application of mosquito insecticides over Sacramento, CA. They observed concentrations of 0.44 to 3.92 μg/L of PBO, but did not detect pyrethrins. Lothrop et al. (2007) measured concentrations of pyrethrins and PBO directly under the airplane through to 300 m from the flight path, to maximize efficacy while minimizing evaporation of the insecticide in a desert environment. They observed concentrations of pyrethrins ranging from ND to 0.0791 μg/cm2, and for PBO they observed concentrations ranging from ND to 1.07 μg/cm2. The researchers observed very low mortality of caged mosquitoes with an average mortality ranging from only 1.5 to 12% 1 h after application. Concentrations of PBO is Suffolk county, New York were detected in 33.3% of water samples taken with concentrations ranging from ND to 59.8 μg/L and resmethrin was detected in 11.1% of water samples taken with concentrations ranging from ND to 0.293 μg/L. Concentrations of resmethrin and PBO become ND by four days after helicopter or truck-mounted ULV applications (Abbene et al. 2005). Zulkosky et al. (2005) measured concentrations of resmethrin ranging from ND to 0.98 μg/L and 11 concentrations of PBO ranging from ND to 15 μg/L 1 h after truck-mounted ULV application. D-phenothrin was ND in all samples taken during the study (Zulkosky et al. 2005). The environmental fate studies that have been conducted to date have found that the amount of malathion and permethrin deposition and droplet abundance decreased with distance from the from the point of origin for truck-mounted ULV. Malathion, permethrin, pyrethrins, PBO, and naled all rapidly degrade in the environment with residues becoming ND 9 to 72 h post application. The studies were pseudo-replicated since the treatments were not applied separately through space, time, distance, or as combinations of the three. Also many of these studies did not obtain samples before the application of ULV insecticides to determine levels of these insecticides already present in the environment. Effects of Ultra-Low-Volume Applications on Non-Target Organisms Researchers in New York provided evidence that the 1999 lobster die-off in Long Island Sound was most likely not caused by the use of ULV resmethrin and d-phenothrin insecticides in response to the introduction of WNV (Landeck-Miller et al. 2005; Pearce and Balcom 2005; Levin et al. 2007). Jensen et al. (1999) showed that the use of truckmounted ULV above wetlands had no significant impact on aquatic macroinvertbrates and Gambusia affinis (Baird and Girard), but did have a significant impact on flying insects. However, flying insect abundance recovered by 48 h after application. Milam et al. (2000) found less than 10% mortality for Pimephales promelas Rafinesque and 12 Daphnia pulex Leydig after truck-mounted and aerial ULV of Biomist®, but Ceriodaphnia dubia J. Richard had an average mortality of 47%. Davis and Peterson (2008) demonstrated that there was little if any significant impact on aquatic and terrestrial invertebrates after single and multiple applications of either permethrin or d-phenothrin by truck-mounted ULV. Boyce et al. (2007) demonstrated that after aerial ULV with pyrethrins and PBO, there was no impact on large-bodied insects within the spray zone, however they did observe an impact on smaller bodied insects. Hill et al. (1971) found that aerial applications of ULV malathion significantly decreased populations of Hymenoptera, however these populations rapidly re-established. Aerial applications of ULV malathion did not significantly affect Hemiptera, Coleoptera, and Diptera (excluding Culicidae) populations (Hill et al. 1971). Risk Assessment Human-health and ecological risks can be described in quantitative terms as a function of toxicity and exposure (NRC 1983). Toxicity is determined by a series of studies, and from those no-observable-adverse-effect-levels (NOAEL) and LC50s are chosen and represent acute and chronic toxic effects. Exposure is contacting, ingesting, or inhaling the chemical of interest. Risk assessment typically utilizes a tiered approach extending from deterministic models (Tier 1), which are based on extremely conservative assumptions, to field evaluation and probabilistic models (Tier 4), which use refined assumptions (SETAC 1994). Risk assessments use conservative assumptions in lowertier assessments which represent overestimates of toxicity and exposure, with resulting 13 deterministic and quantitative risk values typically being conservative and erring on the side of safety. Tier 4 risk assessments use refinements like experimentally derived environmental concentrations. In environmental and human health risk assessments of pesticides, risk quotients (RQ) is a method to quantitatively express risk (Peterson 2006). RQs are calculated by dividing the potential exposure (PE) by its toxic endpoint value depending on weather it is oral, inhalation, or dermal. RQs are compared to a level of concern (LOC) which is set by USEPA to determine if regulatory action is needed. If an RQ breaches a regulatory LOC at a lower tier, then risk managers decide to either restrict the product use or progress to higher tier risk assessments or to field-verified models (USEPA 2006e). Probabilistic risk assessments differ from deterministic by sampling values from the distributions of exposures and biological parameters. Probabilistic risk assessments using Monte Carlo simulation have one or more parameters that are repeatedly sampled on a random basis from their respective distributions and propagated into an equation to generate probability of the exposures occurring. Previous Risk Assessments, Reports, and Biomonitoring Studies Since WNV was introduced into the United States in 1999, more areas of the country have been experiencing large-scale insecticide applications for mosquito-borne diseases like WNV. With the majority (>60%) of the American public not concerned about the threat of WNV (Ho et al. 2007), there has been greater public attention to the human-health and environmental risks associated with ULV insecticide applications 14 (Thier 2001; Roche 2002; Peterson et al. 2006; Reisen and Brault 2007). In response to concerns about the safety of ULV insecticides, tier I/II (reasonable worst case) risk assessments have been performed to quantify estimates of risk. Peterson et al. (2006) performed a deterministic human-health risk assessment for acute and sub-chronic exposures to six mosquito insecticide active ingredients, and the synergist PBO, after ground-applied ULV applications. They found that risks to humans acutely and subchronically from the insecticides most likely would result in negligible risk. Schleier III et al. (2008c) performed a probabilistic two-dimensional risk assessment of the same insecticides and populations subgroups of Peterson et al. (2006). The probabilistic analysis showed that Peterson et al. (2006) over estimated risk by about 10-fold. The results of Schleier III et al. (2008c) supported the findings of Peterson et al. (2006) that the risks to humans from mosquito adulticides is most likely negligible. The sensitivity analysis performed by Schleier III et al. (2008c) showed that air concentrations and dermal exposure contributed the most to the model output variance, which contributes to a large uncertainty surrounding the estimates of risk. Davis et al. (2007) conducted a deterministic ecological risk assessment and examined the same mosquito insecticides and synergist as Peterson et al. (2006) using a tier I/II assessment and found that the risks to mammals, birds, and aquatic vertebrates and invertebrates most likely are negligible after truck-mounted ULV applications. Schleier III et al. (2008b) examined deterministically and probabilistically the six mosquito insecticides and the synergist as well, and found similar results, demonstrating that the equine risks from truck-mounted ULV are very low. The probabilistic analysis 15 of Schleier III et al. (2008b) demonstrated that the deterministic analysis was sufficiently conservative, with deterministic exposures between the 85th to 95th percentile of exposures. Macedo et al. (2007) determined that the risks to military personnel exposed to truck-mounted ULV permethrin, resmethrin, phenothrin, or PBO are most likely negligible. Carr Jr. et al. (2006) results showed that the use of aerially applied ULV resmethrin above agricultural fields as a result of a public health emergency would most likely result in negligible human dietary risk. Currier et al. (2005) found no statistical differences in naled, permethrin, and dphenothrin urinary metabolites in humans from areas that were treated with truckmounted ULV and non-treated areas at 0.04, 0.002, and 0.0036 lbs/acre, respectfully. Kutz and Strassman (1977) and Duprey et al. (2008) demonstrated that aerial spraying of naled does not result in increased levels of naled urinary metabolites in humans. Both studies found that there were detectable levels of organophosphate metabolites before and after the aerial spraying in about the same proportion. Epidemiological studies, reports, and regulatory assessments have concluded that risks to humans and non-target organisms from exposure to mosquito insecticides most likely are negligible (Karpati et al. 2004; NYCDOH 2005; O'Sullivan et al. 2005; Suffolk-County 2006). Objectives Even though previous risk assessments conducted to date have not revealed unacceptable exposures, it is important to conduct a field assessment because there have 16 been few fate studies conducted and also to ensure appropriate conservatism in the risk assessments. All of the previous risk assessments used models (AgDRIFT, AERMOD, ISCST3, and PRIZM-EXAMS) to generate the environmental concentrations (Peterson et al. 2006; Davis et al. 2007; Macedo et al. 2007; Schleier III et al. 2008b; Schleier III et al. 2008c). The estimates by the models contributed to the largest amount of uncertainty to the risk assessments (Schleier III et al. 2008c). Because of ongoing concerns by the public about the safety of insecticides used for the control of adult mosquitoes the aim of the present study was reduce the uncertainty associated with the fate of ULV insecticides. The objectives of my study were to quantify the amount of naled and permethrin that is deposited at 25, 50, and 75 m from the spray head at ground level and 1.25 m above the ground. Deposition samples were taken at 1 and 12 h for naled and permethrin after each spray event at each distance, as well as a 24-h sample for permethrin at 25 m at ground level and at 25, 50, and 75 m at 1.25 m above the ground. Also, the quantification of breathable insecticide particulates at 1.25 m above ground level at 25 m from the spray source at 1 and 12 h was taken for both permethrin and naled. In addition, I measured terrestrial surface residues and surface water concentrations of pyrethrins and PBO in an urban setting after a typical aerial ULV application. To accomplish this objective, I examined the concentrations of insecticide present before and after a single ULV application, and related those to refining risk assessments for ULV insecticides. The amount of both insecticides deposited and inhaled will be used to refine the human-health risk assessment of Peterson et al. (2006) and the probabilistic human-health 17 risk assessment of Schleier III et al. (2008c). Previous risk assessments have demonstrated that deposition and inhalation of insecticides were contributing the most to the overall exposure estimate and the model output variance (Peterson et al. 2006; Davis et al. 2007; Schleier III et al. 2008b; Schleier III et al. 2008c). 18 CHAPTER 2 ENVIRONMENTAL CONCENTRATIONS AND FATE OF NALED AND PERMETHRIN USED FOR ADULT MOSQUITO MANAGEMENT Abstract One of the most effective ways of managing adult mosquitoes that vector human and animal pathogens is the use of ultra-low-volume (ULV) insecticides. Because of the lack of environmental fate studies and concerns about the safety of the insecticides used for the management of adult mosquitoes, I conducted an environmental fate study in Cascade and Ulm, Montana, USA, after truck-mounted applications of permethrin ([3phenoxyphenyl]methyl 3-[2,2-dichloroethenyl]-2,2-dimethylcyclopropane carboxylate) and naled (1,2-dibromo-2, 2-dichloroethyl dimethyl phosphate). One hour after application concentrations of permethrin on cotton dosimeters 25, 50, and 75 m from the spray source were 0.002, 0.004, and 0.001 μg/cm2 in 2007, and 0.005, 0.002, and 0.0009 μg/cm2 in 2008, respectively. One hour after application concentrations of naled 25, 50, and 75 m were 0.002, 0.004, and 0.001 μg/cm2 in 2007, and 0.005, 0.002, and 0.0009 μg/cm2 in 2008, respectively. Deposition concentrations 12 h after application were not significantly different than 1 h after application for both permethrin (p = 0.41) and naled (p = 0.51). During 2007 and 2008 permethrin applications, two quantifiable air concentrations of 0.375 μg/m3 and 0.397 μg/m3 were observed 1 h after application. During 2007 and 2008, naled air concentrations ranged from 2.3 to 4 μg/m3 1 h after application. There were no quantifiable air concentrations between 1 and 12 h after 19 application in either 2007 or 2008 for both naled and permethrin. Environmental concentrations observed in this study demonstrate that models used in previous risk assessments were sufficiently conservative (i.e., the models overestimated environmental concentrations). However, the current study demonstrates the inadequacies of models such as AgDrift® and AGDISP, which currently are used by the U.S. Environmental Protection Agency to estimate environmental concentrations of ULV insecticides. Introduction West Nile virus (WNV) has become endemic to North America and disease cases occur throughout the virus transmission season. Since the arrival of WNV, more areas of the country have been experiencing large-scale insecticide applications for mosquitoborne pathogens like WNV. To effectively manage infection rates, morbidity, and mortality due to mosquito-borne pathogens like WNV, there must be a reduction in contact between infected mosquitoes and humans and animals (Marfin and Gubler 2001). One of the most effective ways of managing high densities of adult mosquitoes that vector human and non-human animal pathogens is ultra-low-volume (ULV) aerosol applications of insecticides (Mount et al. 1996; Mount 1998). Ultra-low-volume utilizes small droplets from 5 to 25 μm, which are the optimum size to impinge on and knockdown flying adult mosquitoes (Weidhaas et al. 1970; Lofgren et al. 1973; Haile et al. 1982). With the majority (>60%) of the American public not concerned about the threat of WNV (Ho et al. 2007), there has been greater public attention to the human- 20 health and environmental risks associated with ULV insecticide applications (Thier 2001; Roche 2002; Peterson et al. 2006). In response to these concerns, tier I/II risk assessments have been performed to quantify reasonable worst-case estimates of risk. Peterson et al. (2006) performed a reasonable worst-case human-health risk assessment for six mosquito insecticide active ingredients, including permethrin ([3phenoxyphenyl]methyl 3-[2,2-dichloroethenyl]-2,2-dimethylcyclopropane carboxylate) and naled (1,2-dibromo-2, 2-dichloroethyl dimethyl phosphate), after truck-mounted ULV applications. Schleier III et al. (2008c) performed a probabilistic human-health risk assessment of the same insecticides as Peterson et al. (2006), and supported their findings that the risk after mosquito adulticide treatments most likely would not exceed a regulatory threshold. Davis et al. (2007) and Schleier III et al. (2008b) examined permethrin and naled as well, and found similar results, demonstrating that ecological and equine risks from truck-mounted ULV would most likely not exceed regulatory thresholds. However, in all of these studies naled had the highest estimated risks of any of the insecticides assessed. Previous probabilistic risk assessments of ULV insecticides have determined through sensitivity analysis that estimated air concentrations and ground deposition of the insecticides contribute the largest amount of variance to the exposure (Schleier III et al. 2008a; Schleier III et al. 2008c). Currently there are no publicly available data concerning the terrestrial deposition of naled, and there are only two studies examining the terrestrial deposition of permethrin after truck-mounted ULV applications (Knepper et al. 1996; Pierce et al. 2005). In addition there are no publicly available studies that measure air concentrations of 21 permethrin and naled after truck-mounted ULV applications. Knepper et al. (1996) measured concentrations of permethrin on sod grass ranging from non-detectable (ND) to 0.014 μg/cm2 after truck-mounted ULV spraying in a suburban neighborhood. Pierce et al. (2005) measured concentrations of permethrin ranging from 0.00005 to 0.005 μg/cm2 in the Florida Keys National Marine Sanctuary, 2 to 4 h after application. Jensen et al. (1999) found no detectable concentrations of permethrin in water samples from wetlands before and after truck-mounted ULV. The majority of the studies examining the fate of truck-mounted ULV have been performed with malathion (O,O-dimethyl dithiophosphate of diethyl mercaptosuccinate). These studies have found that 1 to 22.3% of the insecticide sprayed settled onto the ground within 500 m from the spray source, with the amount substantially decreasing by 24 to 36 h after application (Moore et al. 1993; Tietze et al. 1994; Knepper et al. 1996; Tietze et al. 1996a) Although risk assessments have been performed, they have relied on estimates of environmental concentrations of adulticides from models that do not have algorithms for ULV application methods. These models are designed for industrial and agricultural applications, not ULV. In addition, the probabilistic risk assessment of Schleier III et al. (2008c) demonstrated that the estimated deposition of insecticides and air concentration are contributing the largest amount of variance to the potential exposure. Because of the concerns about the safety of adulticides used for the control of adult mosquitoes, and because of the lack of actual environmental concentration data and uncertainties associated with the fate of the insecticides, I conducted an environmental fate study in the summers of 2007 and 2008. The objectives of my study were to characterize terrestrial 22 surface residues and air concentrations of permethrin and naled after truck-mounted ULV applications. Materials and Methods The study occurred in conjunction with the Cascade County Weed and Mosquito Control District outside of Cascade (N47°13.489', W111°42.040'), and Ulm (N47°25.402', W111°29.767'), MT, USA during the summers of 2007 and 2008, respectively. Because of considerations of cost and logistical issues, naled and permethrin were applied on separate evenings. Therefore, four field experiments were conducted, one for naled and one for permethrin each year. The applications took place in open fields with no vegetation taller than 20 cm. Within each study site, surface residue and air concentration samples were taken. Surface residue sample collectors were placed 25, 50, and 75 m, and air concentration samplers were placed 25 m from the spray source. There were three sample replicates with 200 m buffer zones between replicates (Figures 1-4). Insecticide were applied by a truck equipped with a Bison (VecTec Inc., Orlando, FL, USA) ULV generator (Figure 5). Permanone® 10% EC (Bayer Environmental Science, Research Triangle Park, NC, USA) was mixed 1:1 with BVA oil (BVA Inc. Wixom, MI, USA) and was applied at the maximum application rate of 7.85 g/ha with a flow rate of 205 ml/min. Trumpet® EC (AMVAC, Los Angeles, CA, USA) was applied undiluted at the maximum application rate of 22.42 g/ha with a flow rate of 44.36 ml/min. Truck speed was 16.1 km/h and spraying began and ended 100 m on each side 23 of the sample collectors (Figures 1-4). Applications occurred when the prevailing wind was blowing perpendicular to the collection site. Temperature, wind speed, and relative humidity at 1.5 m above ground level were recorded with a Kestrel® 4000 pocket weather tracker (Nielsen-Kellerman, Boothwyn, PA, USA). In 2007, permethrin was applied on August 12 at 2030 h MDT, and naled at 2020 h on August 27. Winds speeds during the permethrin application were out of the northeast at 8 km/h with wind gusts to 17.7 km/h. Wind speeds during the naled application were 2.4 km/h out of the north with wind gusts to 4.8 km/h. At the time of application, average temperature and relative humidity were 22oC and 35% for permethrin and 21oC and 33% for naled. In 2008, permethrin was applied at 2015 h on July 25, and naled at 2000 h on August 12. Winds speeds during the application of permethrin were out of the northeast at 7 km/h with wind gusts to 12.9 km/h. Wind speeds during the application of naled were 8 km/h out of the southwest with wind gusts to 12.9 km/h. At the time of application, average temperature and relative humidity was 27oC and 27.5% for permethrin and 24oC and 23% for naled. Collections of surface residues at ground level and at 1.25 m above the ground were taken on 10-cm X 10-cm (100 cm2) cotton dosimeters pinned to a piece of cardboard (Ross et al. 1990) (Figure 6). The cotton dosimeters at 1.25 m were perpendicular to the ground. Cotton dosimeters were separated by 15 cm at the edges of the patches on the piece of cardboard. The cardboard was covered with plastic wrap to prevent contact between the cardboard and dosimeters. Before each application 24 dosimeters were placed 25 m from the spray line at ground level 1.5 h before spraying and were collected just before spraying began. For permethrin, two dosimeters were pinned on each piece of cardboard at ground level and collected 1 and 12 h after application. At 25 m, a third dosimeter was collected at 24 h. At 1.25 m above the ground, three dosimeters were pinned on each piece of cardboard in the treated and control areas and were collected 1, 12, and 24 h after application. Three blank and three spiked dosimeters were placed in a control (untreated) area, with one of each collected at 1, 12, and 24 h after application. For naled, two dosimeters were pinned on each piece of cardboard at ground level and 1.25 m above ground level, and were collected 1 and 12 h after application. Two blank and two spiked cotton pads were placed in a control (untreated) area for naled and one of each was collected 1 and 12 h after application. The control areas were located where no spraying or drift could occur, but were subject to the same meteorological conditions as the residue samples. The cotton pads were dosed with 0.75 μg of technical grade insecticide, which were used as a positive control. Cotton dosimeters were collected with tweezers. The tweezers were rinsed with pesticide-grade acetone between dosimeters to prevent cross contamination. Individual samples of naled and permethrin were stored in separate 60 ml I-Chem™ glass jars with Teflon® lids (Chase Scientific Glass, Vineland, NJ, USA). Jars were placed in a cooler with dry ice for transport from the field site to the lab. Jars were stored in a freezer at <4oC to prevent degradation of insecticide until analysis (Lewis 1999). 25 Air samples were continuously drawn by an SKC® Model 224-PCXR4KDB universal pump (SKC Inc. Eighty Four, PA, USA) through an SKC sorbent polyurethane foam (PUF; SKC Catalog No. 226-92) cartridge for both naled and permethrin. PUF cartridges were kept in aluminum packing during transport to and from the field. The cartridges were removed from the aluminum package and attached to the pump via a 1.5m piece of flexible plastic tubing. The flow rate was set at 2 L/min in 2007 and 5 L/min in 2008. Cartridges were 1.25 m off the ground and attached to a stake. After the 1 and 12 h sampling periods, PUF cartridges were wrapped in the original aluminum foil and placed in a cooler on dry ice for transport to the lab. Polyurethane foam cartridges were stored in a freezer at <4oC until analysis (Lewis 1999). One field blank and one spiked PUF for each collection time were treated exactly the same as the other cartridges except no air was pulled through it (Lewis 1999). The PUFs were spiked with 0.75 μg of insecticide and were treated like the field blanks. Extraction of all samples occurred within seven days of sampling to prevent degradation (Lewis 1999). PUF cartridges were removed from their glass contains using tweezers and cut into six pieces and placed into a 60 ml I-Chem glass jars with Teflon lids. Dosimeters and PUF cartridges were extracted using 45 ml of high pressure liquid chromatography grade hexane (Fischer Scientific, Pittsburgh, PA, USA). Jars were placed on a shaker table for 2 h. A 13 ml aliquot was concentrated to 1 ml using nitrogen evaporator at 30-35 oC. Extraction efficiency was determined by using methyl chlorpyrifos (O,O-dimethyl O-3,5,6-trichloro-2-pyridyl phosphorothioate) for naled and permethrin. 26 Analysis was done by the Montana State Department of Agriculture’s Chemical Analytical Laboratory in Bozeman, MT, USA. The reported detection limits were 0.03 μg for cis-permethrin and trans-permethrin and 0.0015 μg for naled for both dosimeters and PUFs. Quality assurance measures included the analysis of reagent blanks, matrix blanks (dosimeters and PUFs), duplicates, and spiked samples (laboratory and field). Permethrin analysis was performed on a Agilent 6890 gas chromatograph electron capture detector (Agilent Technologies, Inc., Santa Clara, CA, USA) equipped with a Restek RTX-5 column with Intraguard (30 m X 0.25 mm X 0.25 μm) (Restek U.S., Bellefonte, PA, USA). The temperature program used to separate cis- and transpermethrin started at 60oC, and increased at 25 oC/min to 280 oC, and held there for 4 min. Naled analysis was performed using an Agilent 5973 gas chromatograph mass spectroscopy detector equipped with a Restek RTX-5 column (30 m X 0.25 mm X 0.25 μm). The temperature program for naled started at 80oC, and increased at 20 oC/min to 280 oC, and held there for 2 min. Statistical Analysis BoxCox transformations were performed on the raw data to determine appropriate transformation. I used Statistical Analysis System 9.1 (SAS Institute, Cary, NC, USA) to run repeated measures analysis of variance (α = 0.05; PROC GLM) on log transformed concentrations to determine differences between times, distances, heights, and year sprayed (Helsel 2005). 27 Non-detectable Concentrations For non-detectable (ND) concentrations in the deposition data, I substituted half of the detection limit when the number of NDs was less than 10% of the data points (Lubin et al. 2004). If the number of NDs was greater than 10% of the data they were quantified using the maximum-likelihood estimate (MLE) (Helsel 2005). The MLE was determined using, NDC = exp(μln+σ2ln/2), (1) where NDC is the mean concentration for non-detects, μ is the mean of detected concentrations, and σ2 is the variance of the detected concretions (Helsel 1990; Helsel 2005). Results Permethrin Permethrin was not detected on the control or background dosimeters (Table 1). Recovery of methyl chlorpyrifos ranged from 78 to 110%, and recovery of both field and laboratory spikes for cis- and trans-permethrin ranged from 94 to 130% in both 2007 and 2008. There was no significant difference between concentrations at ground level and at 1.25 m above the ground (F = 0.11, p = 0.74), so data for the two heights were combined for analysis. There was a significant difference between the 2007 and 2008 concentrations (F = 5.95, p = 0.02), and a significant effect of distance from the spray source (F = 16, p < 0.0001; Table 1). Concentrations at 24 h were significantly lower than concentrations measured 1 h (F = 5.72, p = 0.02) after application. However, 28 concentrations at 12 h were not significantly different from the 1 h samples (F = 0.67, p = 0.41). During 2007, there was only one quantifiable air concentration of 0.375 μg/m3 1 h after application. In 2008 one quantifiable concentration of 0.397 μg/m3 and one detection below the limit of quantification measured 1 h after application. There were no quantifiable air concentrations measured between 1 and 12 h after application in either 2007 or 2008. Naled There were no detectable amounts of naled on the control or background dosimeters (Tables 2 and 3). Recovery of methyl chlorpyrifos ranged from 64 to 108%, and recovery of both field and laboratory spikes of naled ranged from 107 to 130% in both 2007 and 2008. There was a significant difference between concentrations at ground level and 1.25 m above the ground (F = 4, p = 0.05). There was a significant difference between the 2007 and 2008 concentrations (F=20.42, p<0.0001). There was a significant effect of distance from the spray source (F = 19.83, p < 0.0001). Concentrations at 1 h were not significantly different than concentrations measured 12 h after application (F=0.45, p=0.51). During 2007, there was only one quantifiable air concentration of 3.91 μg/m3 1 h after application, and in 2008 there were three quantifiable concentrations of 2.3, 2.9, and 4 μg/m3 1 h after application. In 2007 and 2008 there were no detectable air concentrations measured between 1 and 12 h after application. 29 Discussion There were similarities between the permethrin and naled data, with concentrations of neither pesticide decreasing significantly from 1 to 12 h after application. The concentrations measured in 2007 for permethrin and naled were similar to those measured in 2008. This is most likely because the main route of degradation for both permethrin and naled is photolysis (USEPA 2002b, 2006b) and all field application were after 2000 h. There were no detectable air concentrations of permethrin or naled measured between 1 and 12 h after application. These results suggest that the insecticides moved rapidly through the sample collectors and that exposure to ULV insecticides maybe limited to within 1 h after application. The concentrations generally decreased as distance from the spray source increased, which is similar to the finding of Knepper et al. (1996), but in 2007 the concentration of permethrin was greater at 50 than at 25 m. The significant difference between ground level and 1.25 m above the ground for naled, but not permethrin, could be due to the heavier oil used in the formulation of Trumpet. Approximately 4.4% of the permethrin and 28% of the naled sprayed settled onto the ground 25 m from the spray source. Previous studies of truck-mounted ULV applications have found 1 to 22.3% of the insecticide sprayed settled onto the ground (Tucker et al. 1987; Moore et al. 1993; Tietze et al. 1994; Knepper et al. 1996). Knepper et al. (1996) measured approximately 0.003 μg/cm2 of permethrin at 15 min after application 30.5 m from the spray source in a domestic setting. Their results and the 30 results of Pierce et al. (2005) were similar to what I measured 1 h after application at 25 m in 2007 and 2008. Previous risk and regulatory assessments have used models like ISCST3, AERMOD, AgDrift®, and AGDISP to estimate environmental concentrations of insecticides (Mickle et al. 2005; Peterson et al. 2006; Davis et al. 2007; Macedo et al. 2007; Schleier III et al. 2008a; Schleier III et al. 2008b; Schleier III et al. 2008c; USEPA 2008). Using the modeling assumptions of Peterson et al. (2006) the estimated environmental concentrations using ISCST3 at 25 m for permethrin and naled were 31 and 6.4 times more than what was measured in this study. The values predicted by ISCST3 were 1.7 times more than what was found in other studies with permethrin (Moore et al. 1993; Tietze et al. 1994; Knepper et al. 1996). The estimated environmental concentrations using AERMOD at 25 m for permethrin and naled were 21 and 5 times more than the highest concentrations measured in this study. Using the model assumptions of Schleier et al. (2008b) concentrations estimated by AgDrift® at 25 m from the spray source for permethrin and naled were 8 and 12 times lower than what was measured in the field. Mickle et al. (2005) found that AGDISP accurately estimated deposition of malathion sprayed from truck-mounted ULV. The USEPA used AGDISP to estimate environmental concentrations of permethrin based on the model inputs of Mickle et al. (2005) for the reregistration eligibility document of permethrin. Deposition concentrations estimated by AGDISP at 25 m using the assumptions outlined by Mickle et al. (2005), and weather conditions and application rate 31 for the permethrin sprays in 2007 and 2008, were 6 and 4 times less than what was seen in the current study. Current and past environmental fate studies and existing models demonstrate that a separate model needs to be designed which can more accurately predict concentrations of ULV applications. Both ISCST3 and AERMOD were sufficiently conservative models for conducting lower tiered risk assessments (Peterson et al. 2006; Davis et al. 2007; Macedo et al. 2007; Schleier III et al. 2008c); however, the current study shows that AgDrift and AGDISP underestimated environmental concentrations, and thus risks. AgDrift and AGDISP are designed to model high boom, agricultural sprays where the nozzles are facing toward the ground, which is considerably different than ULV applications. During ULV applications, the nozzles are pointed at a 135° angle to the ground and sprayed out so that the resulting spray can drift. The results of my study demonstrate that a more realistic model needs to be developed so that federal, state, and local officials can more accurately estimate the risks of ULV insecticides for use in regulatory documents and communications with the public. A more realistic spatial and temporal model also has advantages for understanding efficacy. The year effect likely indicates that environmental conditions significantly affect the deposition of permethrin and naled. A more extensive study with additional temporal and spatial replicates should be conducted to examine the effect of wind speed, air temperature, humidity, and other factors that influence the behavior of ULV aerosols. 32 Table 1. Mean permethrin deposition 25, 50, and 75 m from the spray source, control, and spike concentrations in μg/cm2 (± standard error) 1, 12, and 24 h after application in 2007 and 2008 Distance 2007 2008 1h 25 m 0.002 ±0.001 0.005 ±0.0007 50 m 0.004 ±0.001 0.002 ±0.001 75 m 0.001 ±0.0006 0.0009 ±0.0002 Control ND* ND Spike 0.0182 0.0035 12 h 25 m 0.002 ±0.001 0.004 ±0.0007 50 m 0.003 ±0.0004 0.001 ±0.0004 75 m 0.0008 ±0.0005 0.0009 ±0.0002 Control ND ND Spike 0.0164 0.0034 24 h 25 m 0.002 ±0.001 0.004 ±0.0008 50 m 0.001 ±0.0004 0.0007 ±0.0003 75 m 0.0002 ±0.0001 0.0004 ±0.0004 Control ND ND Spike 0.017 0.004 *Non-detectable concentrations 33 Table 2. Mean naled deposition at ground level 25, 50, and 75 m from the spray source, control, and spike concentrations in μg/cm2 (± standard error) 1 and 12 h after application in 2007 and 2008 Distance 2007 2008 1h 25 m 0.05 ±0.01 0.02 ±0.003 50 m 0.07 ±0.01 0.006 ±0.002 75 m 0.07 ±0.01 ND Control ND* ND Spike 0.01 7.5 12 h 25 m 0.05 ±0.007 0.02 ±0.002 50 m 0.07 ±0.007 0.007 ±0.003 75 m 0.07 ±0.006 0.0006 ±0.0005 Control ND ND Spike 0.008 11 *Non-detectable concentrations Table 3. Mean naled deposition 1.25 m above the ground 25, 50, and 75 m from the spray source in μg/cm2 (± standard error) 1 and 12 h after application in 2007 and 2008 Distance 2007 2008 1h 25 m 0.01 ±0.002 0.02 ±0.005 50 m 0.007 ±0.001 0.01 ±0.005 75 m 0.005 ±0.004 0.0005 ±0.0005 12 h 25 m 0.01 ±0.001 0.01 ±0.001 50 m 0.005 ±0.0006 0.01 ±0.003 75 m 0.005 ±0.004 0.002 ±0.001 34 Driving Direction Wind Direction Figure 1. Site layout with wind direction, driving direction, and locations where the spray began and ended for the application of naled in 2007. Start and stop indicate where the spraying began and stopped for each of the three replicates. Mid 1, Mid 2, and Mid 3 are located 100 m from the start and stop for each replicate. For numbered locations the first number indicates the replicate and the second number indicates the distance from the sprayer (example 11 is replicate 1 at 25 m from the spray source). 35 Wind Direction Driving Direction Figure 2. Site layout with wind direction, driving direction, and locations where the spray began and ended for the application of permethrin in 2007. Start and stop indicate where the spraying began and stopped for each of the three replicates. Mid 1, Mid 2, and Mid 3 are located 100 m from the start and stop for each replicate. For numbered locations the first number indicates the replicate and the second number indicates the distance from the sprayer (example 11 is replicate 1 at 25 m from the spray source). 36 Wind Direction Driving Direction Figure 3. Site layout with wind direction, driving direction, and locations where the spray began and ended for the application of naled in 2008. Start and stop indicate where the spraying began and stopped for each of the three replicates. Mid 1, Mid 2, and Mid 3 are located 100 m from the start and stop for each replicate. For numbered locations the first number indicates the replicate and the second number indicates the distance from the sprayer (example 11 is replicate 1 at 25 m from the spray source). 37 Wind Direction Driving Direction Figure 4. Site layout with wind direction, driving direction, and locations where the spray began and ended for the application of permethrin in 2008. Start and stop indicate where the spraying began and stopped for each of the three replicates. Mid 1, Mid 2, and Mid 3 are located 100 m from the start and stop for each replicate. For numbered locations the first number indicates the replicate and the second number indicates the distance from the sprayer (example 11 is replicate 1 at 25 m from the spray source). 38 Figure 5. Ultra-low-volume machine in operation 39 1.25 m sampler Air sampler Ground sampler Figure 6. Sampler layout for naled and permethrin applications with deposition samplers 25, 50, and 75 m and air sampler 25 m from the spray source. 40 0.005 Permethrin (μg/cm2) 0.004 0.003 1h 12 h 0.002 24 h 0.001 0 25 m 50 m 75 m Distance Figure 7. Mean concentrations of permethrin (μg/cm2 ± standard error) in 2007 25, 50, and 75 m from the spray source 1, 12, and 24 h after application 41 0.006 Permethrin (μg/cm2) 0.005 0.004 0.003 1h 12 h 24 h 0.002 0.001 0 25 m 50 m 75 m Distance Figure 8. Mean concentrations of permethrin (μg/cm2 ± standard error) in 2008 25, 50, and 75 m from the spray source 1, 12, and 24 h after application 42 0.09 0.08 Naled (μg/cm2) 0.07 0.06 0.05 0.04 Ground Level 0.03 1.25 m 0.02 0.01 0 1h 1h 1h 12 h 12 h 12 h 25 m 50 m 75 m 25 m 50 m 75 m Time (h) and Distance (m) Figure 9. Mean concentrations of naled at ground level and 1.25 m above the ground (μg/cm2 ± standard error) in 2007 25, 50, and 75 m from the spray source 1 and 12 h after application 43 0.03 Naled (μg/cm2) 0.025 0.02 0.015 Ground Level 0.01 1.25 m 0.005 0 1h 1h 1h 12 h 12 h 12 h 25 m 50 m 75 m 25 m 50 m 75 m Time (h) and Distance (m) Figure 10. Mean concentrations of naled at ground level and 1.25 m above the ground (μg/cm2 ± standard error) in 2008 25, 50, and 75 m from the spray source 1 and 12 h after application 44 CHAPTER 3 A PROBABILISTIC RISK ASSESSMENT OF NALED AND PERMETHRIN USING ACTUAL ENVIRONMENTAL CONCENTRATIONS Abstract Ultra-low-volume (ULV) aerosol applications of insecticides are used to manage high densities of adult mosquitoes. Because of the uncertainties associated with environment concentrations of insecticides used for ULV, the objective of my study was to use actual environmental fate data and probabilistic models to evaluate the conservatism and reduce the uncertainty of previous human-health risk assessments of permethrin ([3-phenoxyphenyl]methyl 3-[2,2-dichloroethenyl]-2,2-dimethylcyclopropane carboxylate) and naled (1,2-dibromo-2, 2-dichloroethyl dimethyl phosphate). Toddlers and infants were the highest risk groups while adult males were the lowest risk group assessed in this study. Median total acute exposure for permethrin ranged from 0.0000005 to 0.000003 mg/kg body weight (BW)/day. Median total acute exposures of naled ranged from 0.00005 to 0.00006 mg/kg BW/day. I used the risk quotient (RQ) method, which is calculated by dividing the total potential exposure for each group and chemical by its ingestion toxic endpoint value (RfD). Median RQs for permethrin ranged from 0.000002 to 0.00001 and naled ranged from 0.003 to 0.006 for all groups. Sensitivity analysis demonstrated that deposition of insecticides onto surfaces contributed the largest variance to the exposure. My results show that previous risk assessments used conservative exposure scenarios that overestimated risks, thus being conservative in 45 protecting human health. The current weight of evidence demonstrates that ULV insecticide exposures and risks are most likely below a regulatory level of concern. Introduction Ultra-low-volume (ULV) aerosol applications of insecticides are used to manage high densities of adult mosquitoes. To effectively manage infection rates, morbidity, and mortality due to mosquito-borne pathogens, there must be a reduction in contact between infected mosquitoes and humans and other animals (Marfin and Gubler 2001). One of the most effective ways of managing high densities of adult mosquitoes that vector human and veterinary pathogens is ULV aerosol applications of insecticides (Mount et al. 1996; Mount 1998). Ultra-low-volume utilizes small droplets from 5 to 25 μm, which are the optimum size to impinge on and knock-down flying adult mosquitoes (Weidhaas et al. 1970; Lofgren et al. 1973; Haile et al. 1982). Since West Nile virus (WNV) was introduced into the United States in 1999, more areas of the country have been experiencing large-scale insecticide applications for mosquitoes. With the majority of Americans not concerned about contracting WNV (Ho et al. 2007), there has been greater public attention to the human-health and environmental risks associated with ULV insecticide applications (Thier 2001; Roche 2002; Peterson et al. 2006). In response to concerns about the safety of ULV insecticides, tier I/II (reasonable worst case) risk assessments have been performed to quantify estimates of risk. Peterson et al. (2006) performed a deterministic human-health risk assessment for acute and sub-chronic exposures to six mosquito insecticide active 46 ingredients, and the synergist piperonyl butoxide (5-[2-(2-butoxyethoxy)ethoxymethyl]6-propyl-1,3-benzodioxole; PBO), after ground-applied ULV applications. They found that acute and subchronic risks to humans from the insecticides would not reach regulatory levels of concern. Schleier III et al. (2008c) performed a probabilistic two-dimensional risk assessment of the same insecticides and populations groups of Peterson et al. (2006). The results of Schleier III et al. (2008c) supported the findings of Peterson et al. (2006) that the risks to humans from mosquito adulticides would most likely not exceed regulatory levels of concern. The sensitivity analysis performed by Schleier III et al. (2008c) showed that air concentrations and deposition of insecticides contributed the largest amount of variance to the estimated exposure. Macedo et al. (2007) demonstrated that the acute, sub-chronic, and chronic risks to military personnel exposed to truck-mounted ULV permethrin ([3phenoxyphenyl]methyl 3-[2,2-dichloroethenyl]-2,2-dimethylcyclopropane carboxylate), resmethrin ((5-benzyl-3-furylmethyl (1RS,3RS;1RS,3SR)-2,2-dimethyl-3-(2-methylprop1-enyl) Cyclopropanecarboxylate), d-phenothrin (3-phenoxybenzyl (1R)-cis, transchrysanthemate), or PBO are most likely negligible. Carr Jr. et al. (2006) showed that the use of aerially applied ULV resmethrin above agricultural fields as a result of a public health emergency would most likely result in negligible human dietary risk. Davis et al. (2007) conducted a deterministic tier I/II ecological risk assessment and examined the same mosquito insecticides and synergist as Peterson et al. (2006). They found that the risks to mammals, birds, and aquatic vertebrates and invertebrates 47 most likely would not result in exposures that exceed a level of concern after truckmounted ULV applications. Schleier III et al. (2008b) examined deterministically and probabilistically the six mosquito insecticides and the synergist as well, and found similar results, demonstrating that the equine risks from truck-mounted ULV would not exceed a regulatory threshold. The probabilistic analysis of Schleier III et al. (2008b) demonstrated that the deterministic analysis was sufficiently conservative, with deterministic exposures between the 85th to 95th percentile of exposures. Currier et al. (2005) found no statistical differences in naled (1,2-dibromo-2, 2dichloroethyl dimethyl phosphate), permethrin, and d-phenothrin urinary metabolites in humans from areas that were treated with truck-mounted ULV and non-treated areas. Kutz and Strassman (1977) and Duprey et al. (2008) demonstrated that aerial spraying of naled does not result in increased levels of urinary metabolites. Epidemiological studies, reports, and regulatory assessments have concluded that risks to humans and non-target organisms from exposure to mosquito insecticides most likely are negligible (Karpati et al. 2004; NYCDOH 2005; O'Sullivan et al. 2005; Suffolk-County 2006). Although the deterministic and probabilistic risk assessments conducted to date have not revealed unacceptable exposures, they have relied on models like AERMOD and ISCST3 to estimate environmental concentrations after truck-mounted ULV applications. These models are designed for industrial plumes and agricultural applications, which greatly differ from ULV applications. In addition, the probabilistic risk assessment of Schleier III et al. (2008c) demonstrated that the estimated air concentration and deposition of insecticides are contributing the largest amount of 48 variance to the potential exposure. Because of the uncertainties associated with environmental concentrations of the insecticides, the objective of my study was to use actual environmental concentration data to probabilistically evaluate the conservatism and reduce the uncertainty of previous human-health risk assessments (Peterson et al. 2006; Schleier III et al. 2008c). Materials and Methods Problem Formulation A two-dimensional probabilistic risk assessment of acute human exposure to permethrin and naled used during truck-mounted ULV applications. Sensitivity analysis to estimate the impact of input variables on the overall variation in the output values of the model (Cullen and Frey 1999). I used the two-dimensional analysis to separate uncertainty because of a lack of knowledge about the fate of the insecticides used for ULV, from variability (i.e. body weight, inhalation rate, etc.) (Frey 1992; Burmaster and Wilson 1996; Frey and Rhodes 1996). Acute exposures were defined in this study as single-day exposures after a single insecticide application. Exposures to several population groups to account for age related differences in exposure. Groups included adult males and adult females (18-65 years of age), youth (10-12 years of age), children (5-6 years of age), toddlers (2-3 years of age), and infants (0.5-1.5 years of age). 49 Hazard Identification Permethrin, a pyrethroid insecticide which is a neurotoxin that acts on the sodium channels of mammals, and naled, an organophosphate insecticide which is a neurotoxin that inhibits acetylcholine esterase in mammals. Both compounds are currently registered by the United States Environmental Protection Agency (USEPA) for use in adult mosquito management in the United States. Toxicity and Dose-response Relationships Dose-response information for each compound was reviewed and endpoints were chosen based on acute exposure. The toxicity endpoints used in this assessment were ingestion reference doses (RfD) that are based on the no-observed-adverse-effect-level in mammals with inclusion of appropriate safety factors, which are determined by the USEPA. The acute oral RfD for permethrin and naled are 0.25 and 0.01 mg/kg body weight (BW)/day, respectively (USEPA 2002b, 2006b). Environmental Concentrations I used the environmental fate data for 1.25 m above the ground and air concentrations from Chapter 2 (Tables 4 and 5). Environmental fate data were obtained from field work conducted in 2007 and 2008. Permethrin and naled were applied at the maximum application rates of 7.85 and 22.42 g active ingredient/ha as listed on the labels. Custom distributions for both permethrin and naled deposition data at 1.25 m above the ground were created using concentrations 1 and 12 h after application because 50 these times were not significantly different (Chapter 2). Distributions for deposition onto surfaces were chosen based on the Anderson-Darling goodness of fit test, which for nonnormalized data, weights the differences between two distributions at their tails (Pettitt 1977; Oracle 2007). Custom uniform distributions were created from the minimum and maximum air concentrations measured for both permethrin and naled due to the small sample size for both. Acute Exposure I assumed multi-route exposures immediately after a single-spray event were limited to 24 h. Routes of insecticide exposure to each group were inhalation, dermal, and dietary and non-dietary ingestion (Figure 11). Assumptions of body weight, respiration rate, and frequency of hand-to-mouth activity are presented in Table 6. Because the data from Chapter 2 demonstrate that the inhalation exposure is most likely limited to 1 h, I assumed that each group would be outside when the spray truck passed and the duration of the exposure was for 1 h. Inhalation exposure was estimated by PEInhalation = (AEC * RR * D) / BW, (1) where PEInhalation is potential exposure from inhalation (mg/kg BW), AEC is actual environmental air concentrations from Chapter 2, which was represented with a uniform distribution using the highest and lowest concentrations detected (µg/m3; Tables 4 and 5), RR is respiratory rate for each group (m3/h; Table 6), D is duration of exposure, and BW is body weight for each group (kg; Table 6). Dermal exposure from spray deposition was estimated by 51 PEDermal = (AEC * CF * SA * PSE * AR) / BW, (2) where PEDermal is potential exposure from dermal contact (mg/kg BW), AEC is the estimated environmental concentrations of insecticide that would settle onto surfaces (skin) (μg/cm2; Tables 4 and 5), CF is the conversion from μg/cm2 to mg/m2, SA is the surface area of the group as estimated by equation 3 (m2), AR is dermal absorption rate for permethrin and naled which are 15 and 100% (USEPA 2002b, 2006a, b), PSE is the percent body surface area exposed to the spray, in which I assumed a triangular distribution with the maximum being shorts only (82.3%), minimum being a single layer of clothing with only head, hands, and feet exposed (20%), and the most likely being a person with a tee-shirt and shorts with their head, forearms, hands, lower legs, and feet exposed (38.7%) (USEPA 1997b), and BW is body weight (kg). Surface area for all groups was estimated using SA = 4BW + 7 / BW + 90, (3) where SA is surface area (m2), and BW is body weight (kg) (USEPA 1997b, 2002a). For infants and toddlers, hand-to-mouth exposure from insecticide settling onto their hand was estimated by PEHand-to-mouth skin = [(AEC* CF * AR * HSA) * PC * SEF] / BW, (4) where PEHand-to-mouth skin is potential exposure from hand-to-mouth activity from the insecticide settling on the skin (mg/kg BW), AEC is the actual environmental concentration of insecticide that would settle onto the skin (μg/cm2), CF is the conversion from μg/cm2 to mg/m2, HSA is hand surface area for each group as calculated by equation 5 (m2), PC is percent of the hand contacted with the mouth which I assumed to 52 be 50%, SEF is saliva extraction factor of 50% (USEPA 2005b), and BW is body weight (kg). Child hand surface area (HSA) was estimated by CSA = (SA * PH) / 2, (5) where SA is the surface area as calculated by equation 3, and PH is the percent surface area of two hands. PH for infants had a triangular distribution with a mean of 5.3% and a minimum value of 5.21% and a maximum value of 5.39% and for toddlers I used a triangular distribution with a mean of 5.68% a minimum of 5.57% and a maximum of 5.78% (USEPA 2002a). Although not included in Peterson et al. (2006), but examined by Schleier III et al. (2008c), acute hand-to-mouth exposure from turf dislodgeable residue to toddlers and infants through contact of the hand with insecticide that has settled onto turf. Exposure from hand-to-mouth activity from turf dislodgeable residue was estimated by PEHand-to-mouth turf = [(AEC * CF * CSA * DR) * PC * AR * FA * SEF * D] / BW, (7) where PEHand-to-mouth turf is potential exposure from hand-to-mouth turf dislodgeable residue (mg/kg BW), AEC is the actual environmental concentration of insecticide that settled onto surfaces (μg/cm2), CF is the conversion from μg/cm2 to mg/m2, CSA is child hand surface area as estimated by equation 6 (m2), DR is the dislodgeable residue, which I assumed to be 20% for both chemicals (USEPA 1997a), PC is percent of the hand contacted with the mouth which I assumed to be 50% per event, FA is frequency of handto-mouth activity (events/hour; Table 6), SEF is saliva extraction factor of 50% (USEPA 53 2005b), D is duration of exposure which I assumed to be 6 h, and BW is body weight (kg). For acute ingestion exposure from tomatoes that were exposed to the pesticide, I assumed that all foods containing tomatoes eaten per day were consumed from tomatoes grown in a home garden without being washed. In addition, I assumed there would be no degradation in the preparation process. Acute ingestion was estimated by PEIngestion = [(AEC * CF) * SAT] / BW, (8) where PEIngestion is potential exposure from consuming exposed produce (mg/kg BW), AEC is the actual environmental concentration of insecticide that settles onto surfaces (μg/cm2), CF is the conversion from μg/cm2 to mg/m2, SAT is the surface area of tomatoes consumed as estimated by equation 9 (m2), and BW is body weight (kg). The surface area of tomatoes consumed was estimated by SAT = 69 + [0.57 + (WT * BW)]/1000, (9) where SAT is the surface area of tomatoes consumed (Eifert et al. 2006) (m2), and WT is the average weight of tomatoes consumed per day by each group (g/kg BW), and BW is body weight (kg). The average amount plus the standard error of tomatoes consumed per day by adult males and females, youth, children, toddlers, and infants is 0.804, 0.804, 0.874, 1.19, 1.77, and 1.21 (g/kg BW), respectively (USEPA 1997c). Total acute exposure to active ingredients for each group was estimated by PEacute = PEInhalation + PEDermal + PEHand-to-mouth skin+ PEHand-to-mouth turf + PEIngestion (10) 54 Risk Characterization I used the risk quotient (RQ) method for my risk assessment, which is calculated by dividing the total potential exposure as estimated by equation 10 for each group and chemical by its respective ingestion toxic endpoint value (RfD). The multi-route exposure in my assessment was compared to the ingestion RfD because it provided a conservative endpoint, which is based on the most sensitive NOAEL. Estimated RQs are compared to a RQ level of concern (LOC) which is set by the USEPA or another regulatory agency to determine if regulatory action is needed. The RQ LOC used in the assessment was 1.0. An RQ >1.0 means that the estimated exposure is greater than the relevant RfD. Probabilistic Analysis Probabilistic risk assessments differ from deterministic risk assessments by sampling values from the distributions of environmental concentrations and biological parameters rather than using one point estimate. To perform the probabilistic risk assessment, I used Monte Carlo simulation (Crystal Ball® 7.3; Oracle®, Denver, CO, USA) to estimate the exposures and RQs. Probabilities of occurrence of RQ values were determined by incorporating sampling from the statistical distribution of each input variable used to calculate the RQs. Two-dimensional analyses were performed using 10,000 randomizations of variability and 250 randomizations of uncertainty to calculate the mean and 50th, 90th, and 95th percentile confidence interval for my estimates of risk. Two-dimensional simulation 55 is a way to distinguish, analyze, and visualize the variability and uncertainty associated with a risk assessment (Burmaster and Wilson 1996). Two-dimensional analysis was chosen because it disaggregates and evaluates the consequences of variability and uncertainty, and in addition it is able to simulate how uncertainties may be independent from one individual to another (Frey 1992). For the two-dimensional analysis, body weight, inhalation rate, percent body surface area exposed, percent surface area of two hands, and hand-to-mouth frequency were designated as variable. A small data set was used for deposition onto surfaces and air concentrations, so these parameters were designated as uncertain (Burmaster and Wilson 1996). Respiratory rate, body weight, percent surface area of two hands, air concentrations, and deposition onto surfaces were truncated at zero because it is not possible for these quantities to have negative values. Sensitivity analysis was performed using 20,000 iterations on all parameters to determine percent contribution of the input variable to the output variance of the model for each group and chemical. Results Risk quotients for both naled and permethrin did not exceed the RQ LOC of 1. Toddlers and infants were the highest risk groups and adult males were the lowest risk group assessed in this study (Tables 7 and 8). Median total acute exposure for permethrin ranged from 0.000013 to 0.000095 mg/kg BW/day (Table 7). Median total acute exposures of naled ranged from 0.001 to 0.0023 mg/kg BW/day (Table 7). Median RQs 56 for permethrin ranged from 0.00005 to 0.00038 and naled ranged from 0.11 to 0.23 for all groups (Table 8). Sensitivity analysis for permethrin demonstrated that deposition of insecticides onto surfaces contributed ≥ 89% of the variance to the exposure for both permethrin and naled for all groups assessed (Table 9). Insecticide concentrations in the air contributed about 3% while inhalation rate, percent body area exposed, and percent surface area of two hands contributed ≥ 3.4% of the variance to the exposure for permethrin (Table 9). Percent body surface area exposed contributed about 8.8% of the variance for exposure to naled (Table 9). These values are also highly uncertain because there has only been one study that has measured potential dermal exposure after truck-mounted ULV application (Moore et al. 1993). Additionally, there is uncertainty surrounding air concentrations because of lack of studies. In addition to uncertainties for the environmental fate of the insecticides, there are also toxicological uncertainties. The inhalation LC50 of naled aerosols like those used in ULV have been shown to be up to 20 times more toxic than the ingestion LC50 (Berteau et al. 1977; Berteau and Dean 1978). However, if the RfD was reduced by 20 times, naled exposure would still not exceed the RQ LOC. There are also toxicological uncertainties with respect to piperonyl butoxide (PBO). Piperonyl butoxide is a commonly added synergist in formulations of permethrin, and has been shown to increase the toxicity of pyrethroids, but there is no indication that PBO acts as a synergist in mammals (Paul et al. 2005; Amweg et al. 2006a; USEPA 2006c). If a 10-fold 57 uncertainty factor was applied to the RfD based on the toxicological uncertainties of PBO synergizing with permethrin, the resulting RQs would still not exceed the RQ LOC. Discussion In the present assessment, I estimated the risks from a multi-route exposure after a truck-mounted ULV insecticide application. The RQs at the 95th confidence interval of exposure were similar to those of Schleier III et al. (2008c), who used estimated environmental concentrations rather than AECs. However, the sensitivity analysis of Schleier III et al. (2008c) showed that air concentrations contributed the most variance to the total exposure, while the current sensitivity analysis, using AECs, showed that deposition onto surfaces contributed the most to the exposure variability. The USEPA in its reregistration eligibility decision documents only estimated exposure from inhalation after ULV applications (USEPA 2006b, c, d, e). In my study, inhalation contributed about 23% to the estimated total exposure for adult males and females, youth, and children, but only about 8% to the total exposure of toddlers and infants. Schleier III et al. (2008c) found that inhalation exposure contributed about 60% to the estimated total exposure for adult males and females, youth, and children and 8% to the total exposure of toddlers and infants. Dermal exposure contributed about 41% to the estimated total exposure for adult males and females, youth and children, and about 17% to the total exposure of toddlers and infants. Ingestion exposure contributed about 31% to the estimated total exposure for all groups assessed. The ingestion exposures are most likely very conservative because Carr et al. (2006) showed that the use of ULV 58 resmethrin above agricultural fields would not result in measureable concentrations on agricultural commodities. Turf dislodgeable residue and hand-to-mouth activities contributed 19 and 35% to estimated total exposure of toddlers and infants, respectively. Schleier III et al. (2008c) showed that turf dislodgeable residue contributed about 60% to the total exposure of toddlers and infants. My results support the findings of Schleier III et al. (2008c) that when examining the acute risks after truck-mounted ULV application, dermal, ingestion, hand-to-mouth activities, and inhalation should be considered. Previous studies of truck-mounted ULV applications have found 1 to 22% of the insecticide sprayed during application settled onto the ground, with concentrations decreasing substantially over 36 h (Tucker et al. 1987; Moore et al. 1993; Tietze et al. 1994; Knepper et al. 1996). The environmental concentrations used in the current study show that about 4.4% of the permethrin and 14% of the naled applied settled onto the ground at 25 m from the spray source (Chapter 2). Risk assessments have used ISCST3 and AERMOD to estimate deposition onto surfaces and air concentrations (Peterson et al. 2006; Davis et al. 2007; Macedo et al. 2007; Schleier III et al. 2008c). The estimated environmental concentrations using ISCST3 at 25 m for permethrin were 31 times more and for naled were 6 times more than the data used in the current study (Chapter 2). The air concentrations that were estimated by AERMOD were 205 and 5 times more than the highest permethrin and naled air concentrations used for this study, respectively (Chapter 2). To estimate dermal exposure, Peterson et al. (2006) and Schleier III et al. (2008c) used the default flagger scenario in the USEPA Pesticide Handler Exposure Database 59 (PHED). My results show that when using actual deposition data onto dosimeters, PHED may underestimate the dermal exposure. However, even though dermal exposures may have been underestimated, the RQs were still lower than both Peterson et al. (2006) and Schleier III et al. (2008c), because of the reduction in exposure from other routes. For example, Schleier III et al. (2008c) estimate of permethrin inhalation exposure was 3 times higher than the present study. The deterministic risk estimates of Peterson et al. (2006) were 15 and 2 times greater than my results at the 95th confidence interval of exposure for permethrin and naled, respectively (Figures 12 and 13). The probabilistic risk estimates of Schleier III et al. (2008c) at the 95th confidence interval were 2 times greater than the results of the present study at the 95th confidence interval of exposure for permethrin (Figures 12). My results show that previous risk assessments used conservative exposure scenarios that overestimated risks, thus being conservative in protecting human health. Results of previous biomonitoring studies for permethrin and naled showing no increase in urinary metabolites after ULV applications (Kutz and Strassman 1977; Currier et al. 2005; Duprey et al. 2008), adding support to these findings. The results of the current study using AECs demonstrates that the estimated exposures and risks are lower than previous risk assessments (Peterson et al. 2006; Schleier III et al. 2008c). Sensitivity analysis demonstrates that more data needs to be generated for deposition onto surfaces and actual dermal exposures to further reduce the uncertainties of the risk assessment. In addition, there are uncertainties about air concentrations because of a lack of studies. Air concentrations should be studied further 60 because when using AECs it contributes 23% to the total exposure for adult males and females, youth, and children. The results of the present study demonstrate that the USEPA should consider a multi-route exposure when examining the risks of insecticides used for adult mosquito management, because inhalation only accounts for about 8% of the total exposure for toddlers and infants. Estimates of risks for truck-mounted ULV applications using AECs were lower than previous risk assessments using estimated environmental concentrations. But my exposure scenarios were conservative and most likely overestimated exposure. For example, I assumed that all groups would be outside for six hours following a ULV application. Applications of ULV typically take place at dusk, when mosquitoes are actively seeking a blood-meal. In addition, I assumed that all groups would be respiring at a moderate activity level when the spray truck passed. Most likely the actual environmental exposures would be lower than the current estimates of risk. Models used in past risk assessments and regulatory documents are not accurate in modeling ULV applications; but ISCST3 and AERMOD seem to provide conservative estimates of environmental concentrations. Because, deposition onto surfaces is contributing the most variance to the exposure, a model is needed that can more accurately predict deposition onto surfaces so that more accurate risk assessments can be performed on current and future insecticide active ingredients. However, even with these uncertainties, the current weight of evidence demonstrates that ULV insecticide exposures and risks are most likely below regulatory levels of concern. 61 Table 4. Custom input distributions for deposition onto surfaces and aerial concentrations for permethrin as measured in Chapter 2 Distribution Input Parameter Concentration Units Type Log-normal Mean 0.031 μg/cm2 Deposition onto Surfaces (Truncated) SD* 0.0086 Air Concentration Uniform Minimum Maximum 0.01 0.397 μg/m3 *Standard Deviation Table 5. Custom input distributions for deposition onto surfaces and aerial concentrations for naled as measured in Chapter 2 Distribution Input Parameter Concentration Units Type Minimum 0 Alpha 2.36 Beta Deposition onto Surfaces μg/cm2 (Truncated) Beta 21 Maximum 0.12 Air Concentration Uniform Minimum Maximum 2.3 4 μg/m3 62 Table 6. Assumptions for body weight, respiration rate, and frequency of hand-to-mouth activity for each group assessed Input Group Parameter Values Units Distribution Source Variables Adult Mean 78.65 kg Malesb SD* 13.23 Adult Mean 65.47 kg Femalesc SD 13.77 d Youth Mean 36.16 kg Body Log-normal Portier et SD 7.12 Weight (Truncated) al. (2007) Mean 19.67 kg Childrene SD 2.81 f Toddlers Mean 13.27 kg SD 1.62 g Infants Mean 9.1 kg SD 1.24 Adult Mean 17.53 m3/day Males SD 2.8 Adult Mean 13.78 m3/day Females SD 2.1 Youth Mean 11.3 m3/day Log-normal Brochu et Respiration SD 2.14 Rate (Truncated) al. (2006) Children Mean 7.74 m3/day SD 1.04 Toddlers Mean 5.03 m3/day SD 0.94 Infants Mean 3.72 m3/day SD 0.81 Toddlers Location 5.3 events/h Scale 3.41 Hand-toShape 0.56 Weibull Xue et al. Mouth (Truncated) (2007) Infants Location 14.5 events/h Frequency Scale 15.98 Shape 1.39 a SD= Standard Deviation, b 18-65 years of age, c 18-65 years of age, d 10-12 years of age, e 5-6 years of age, f 2-3 years of age, g 0.5-1.5 years of age 63 Table 7. Acute total potential exposure (PE) means, 50th, 90th, and 95th percentile confidence intervals for each group and chemical assessed Adult Adult Female Youth Children Toddlers Infants Male 18-65 18-65 10-12 0.5-1.5 5-6 years 2-3 years years of years of years of years of of age of age age age age age Chemical PEa Mean Permethrin 50th 90th 95th 3.3E-05 3.0E-05 5.5E-05 6.3E-05 3.5E-05 3.3E-05 5.8E-05 6.8E-05 5.3E-05 4.8E-05 8.3E-05 9.5E-05 7.0E-05 2.5E-05 1.5E-04 2.5E-04 1.4E-04 1.3E-04 1.9E-04 2.2E-04 1.6E-04 1.5E-04 2.1E-04 2.3E-04 Mean 1.2E-03 1.3E-03 1.5E-03 1.8E-03 2.2E-03 50th 1.1E-03 1.2E-03 1.4E-03 1.7E-03 2.1E-03 Naled 90th 1.6E-03 1.8E-03 2.1E-03 2.5E-03 3.0E-03 95th 1.7E-03 1.9E-03 2.2E-03 2.7E-03 3.2E-03 a Total potential acute exposure as estimated by equation 10 (mg/kg BW/day) 2.4E-03 2.3E-03 3.2E-03 3.5E-03 Table 8. Acute risk quotient (RQ) means and 50th, 90th, and 95th percentile confidence intervals for each group and chemical assessed Adult Adult Youth Children Toddlers Infants Male Female Chemical RQa 18-65 18-65 10-12 0.5-1.5 5-6 years 2-3 years years of years of years of years of of age of age age age age age Mean 1.3E-04 1.4E-04 2.1E-04 2.8E-04 5.4E-04 6.2E-04 1.3E-04 1.9E-04 1.0E-04 5.1E-04 6.0E-04 Permethrin 50th 1.2E-04 90th 2.2E-04 2.3E-04 3.3E-04 6.1E-04 7.7E-04 8.4E-04 95th 2.5E-04 2.7E-04 3.8E-04 1.0E-03 8.6E-04 9.1E-04 Mean 50th Naled 90th 95th a Risk Quotient 1.2E-01 1.1E-01 1.6E-01 1.7E-01 1.3E-01 1.2E-01 1.8E-01 1.9E-01 1.5E-01 1.4E-01 2.1E-01 2.2E-01 1.8E-01 1.7E-01 2.5E-01 2.7E-01 2.2E-01 2.1E-01 3.0E-01 3.2E-01 2.4E-01 2.3E-01 3.2E-01 3.5E-01 64 Table 9. Sensitivity analysis (percent contribution of the input variable to the exposure variance) of uncertain factors for permethrin and naled Permethrin Adult Adult Youth Children Toddlers Infants Males Females 18-65 18-65 10-12 0.5-1.5 5-6 years 2-3 years years years of years years of of age of age of age age of age age Deposition onto 91.7a 92.1 90.8 93.4 97.3 98.2 surfaces Air Concentrations 5.3 4.1 4.1 3.5 0.6 0.2 Inhalation Rate 0.3 0.3 0.6 0.2 0.1 0 Body Weight 1.4 2.3 3.4 2.1 0.1 0.3 Percent Body Surface Area 1.3 1.2 1 0.8 0.2 0.1 Exposed Percent Surface Area NA NA NA 0 0 NAb of Two Hands Frequency of HandNA NA NA NA 1.5 1.2 To-Mouth activity Naled Adult Adult Youth Children Toddlers Infants Males Females 18-65 18-65 10-12 0.5-1.5 5-6 years 2-3 years years of years of years years of of age of age age age of age age Deposition onto 88.2 88.6 89.3 90.3 95.3 96.8 surfaces Dermal Absorption 8 7.5 6.1 6.1 3 1.5 Air Concentrations 1.3 1 1.1 1.1 0.4 0.2 Inhalation Rate 0.1 0.1 0.1 0 0.1 0 Body Weight 0.7 1.1 1.1 1.2 0.1 0.4 Percent Body Surface Area 1.7 1.7 1.5 1.3 0.5 0.3 Exposed Percent Surface Area NA NA NA NA 0 0 of Two Hands Frequency of HandNA NA NA NA 0.6 0.8 To-Mouth activity a b Percent contribution of the input variable to the output variance Not applicable exposure for group Inhalation exposure (PEInhalation) Ingestion exposure from consuming tomatoes grown in the home garden (PEIngestion) Hand-to-mouth exposure from insecticide settling onto their hand (PEHand-to-mouth skin) Figure 11. Exposure routes for all groups assessed Hand-to-mouth exposure from turf dislodgeable residue (PEHand-to-mouth turf) 65 Dermal exposure from the insecticide settling on skin (PEDermal) 66 0.0063 0.006 Risk Quotient 0.005 0.004 0.003 0.002 0.0015 0.0009 0.001 0 Peterson et al. (2006) Schleier III et al. (2008) Current Study Figure 12. Toddler risk quotients (RQ) for permethrin from Peterson et al. (2006) using estimated environmental concentrations, 95th percentile confidence interval RQ of Schleier III et al. (2008c) using estimated environmental concentrations, and the 95th percentile confidence interval RQ of the current study using environmental concentrations 67 0.5 0.4726 0.45 0.4 Risk Quotient 0.35 0.32 0.3 0.25 0.2 0.15 0.1 0.05 0 Peterson et al. (2006) Current Study Figure 13. Toddler risk quotients (RQ) for naled from Peterson et al. (2006) using estimated environmental concentrations and the 95th percentile confidence interval RQ of the current study using environmental concentrations 68 Chapter 4 TOXICITY AND RISK ASSESSMENT OF PERMETHRIN AND NALED FOR NONTARGET TERRESTRIAL INSECTS Abstract The house cricket, Acheta domesticus (L.), was used a indicator species for medium- to large-bodied ground dwelling insects for the laboratory derived lethal concentration that kills 50% of a population (LC50), non-target risk assessment, and field bioassay. The LC50 values at 24 h for Permanone®, Permanone + piperonyl butoxide (PBO), technical grade permethrin, and technical grade permethrin + PBO ranged from 0.052 to 0.9 μg/cm2. The synergist ratio was 2.65 for Permanone + PBO and 1.57 for technical grade permethrin + PBO. The toxicity of technical grade permethrin was about 10-fold greater than Permanone, which is the formulated product of permethrin. The 24 h LC50 for naled was 0.44 μg/cm2. To conduct a risk assessment for non-target ground dwelling insects, I used the risk quotient (RQ) method, which is calculated by dividing actual environmental concentration (AEC), or estimated environmental concentration (EEC) for each chemical by its respective LC50. The environmental model ISCST3, which was used in previous ecological risk assessments, was used for to EECs. The risk assessment using EECs resulted in RQs that exceeded levels of concern but when AECs were used RQs did not exceed a regulatory level of concern. Field bioassays using caged crickets showed no significant mortality for permethrin and naled after a single truckmounted ultra-low-volume (ULV) application. The current study demonstrated that a 69 single ULV application of synergized or unsynergized permethrin, or naled most likely would not result in population impacts on medium- to large-bodied insects. Introduction West Nile virus (WNV) has become endemic to North America and disease cases occur throughout the virus transmission season. Since the arrival of WNV, more areas of the country have been experiencing large-scale insecticide applications for mosquitoborne pathogens like WNV. Ultra-low-volume (ULV) aerosol applications of insecticides are used to manage high densities of adult mosquitoes (Mount et al. 1996; Mount 1998). Ultra-low-volume is the minimum effective volume of insecticide that is used as an outdoor space spray for adult mosquitoes. Currently, adult mosquito management utilizes pyrethroids and pyrethrins synergized with piperonyl butoxide (PBO), or organophosphate insecticides (Rose 2001). The insecticides commonly used for mosquito management are highly toxic to certain non-target organisms like invertebrates and aquatic organisms, and there have been concerns about the effect ULV applications on these organisms (Paul et al. 2005; Amweg et al. 2006a; Amweg et al. 2006b; Paul and Simonin 2006; Weston et al. 2006). Pyrethrins, pyrethroids, and organophosphate ULV insecticide sprays have had detrimental effects on non-target organisms like honey bees (Womeldorf et al. 1974; Caron 1979; Pankiw and Jay 1992; Hester et al. 2001; Zhong et al. 2003). Aerial applications of ULV malathion significantly decreased populations of Hymenoptera (Hill 70 et al. 1971). However, the aerial applications did not significantly affect Hemiptera, Coleoptera, and Diptera (excluding Culicidae) populations (Hill et al. 1971). Truck-mounted applications of ULV permethrin had no significant impact on aquatic macroinvertbrates and Gambusia affinis (Baird and Girard) when used above wetlands, but it did have a significant impact on small flying insects (Jensen et al. 1999). Flying insect populations, though, recovered within 48 h after the initial application. Aerial ULV applications with pyrethrins and PBO had no significant impact on caged medium- to large-bodied insects within the spray zone, but researchers observed an impact on smaller-bodied insects (Boyce et al. 2007). Davis et al. (2007) found that the risks to mammals, birds, and aquatic vertebrates and invertebrates most likely would not result in exposures that exceed regulatory levels of concern after truck-mounted ULV applications. Davis and Peterson (2008) demonstrated that there was little to no significant impact on aquatic and terrestrial invertebrates after single and multiple applications of truck-mounted ULV. Lawler et al. (2008) found that the use of ULV pyrethrins synergized with PBO did not cause significant mortality of the aquatic invertebrates, Daphnia magna Straus and Callibaetis californicus Banks. However, truck-mounted ULV applications of malathion (O,O-dimethyl dithiophosphate of diethyl mercaptosuccinate) have been shown to have a significant effect on house crickets, Acheta domesticus (L.). in a peridomestic setting (Tietze et al. 1996a). The mortality of crickets ranged from 12.5 to 48.7%, depending on the location in residential yards. There have been few studies examining the effects truck-mounted ULV applications on non-target ground dwelling organisms. In addition, there are few data on 71 non-target arthropods with respect to toxicity and risks associated with insecticides used for adult mosquito management. Therefore, the objectives of my study were to determine the susceptibility of house crickets to permethrin and naled, and to perform an ecological risk assessment using the house cricket as a surrogate for medium- to large-bodied ground dwelling arthropods. I chose the house cricket because Antwi and Peterson (2008) demonstrated that they were more susceptible than adult convergent lady beetle, Hippodamia convergens (Guérin-Méneville), and larval fall armyworm, Spodoptera frugiperda (J. E. Smith). I compared the laboratory derived lethal concentration that kills 50% of a population (LC50) to concentrations estimated by the model ISCST3 and actual environmental concentrations. In addition to the toxicity and risk assessment, caged crickets were placed in the field to determine if the insecticide applications caused significant mortality. Materials and Methods Laboratory Bioassay Adult house crickets were obtained from Big Apple Herpetological (Hauppauge, NY, USA) for both the laboratory and field experiments. House crickets were chosen as the indicator for the lab and field experiments because of the ease of handling, rearing, and known susceptibility and sensitivity too many insecticides. Additionally, they are considered to be one of the best indicators of environmental pollution, because they are more susceptible than other species of invertebrates (Harris 1966; Bass 1986; Brieger et al. 1992; Tietze et al. 1996a; Hoffmann et al. 2002; Antwi and Peterson 2008). 72 Technical grade permethrin ([3-phenoxyphenyl]methyl 3-[2,2-dichloroethenyl]2,2-dimethylcyclopropane carboxylate; 98% purity) and synergist piperonyl butoxide (PBO; 2-(2-butoxyethoxy)ethyl 6-propylpiperonyl ether; 98.2% purity) were obtained from Sigma-Aldrich (St. Louis, MO, USA), Permanone® 10% emulsified concentrate (EC) (permethrin) was obtained from Bayer Environmental Science (Research Triangle Park, NC, USA), and Trumpet® EC (naled; 1,2-dibromo-2, 2-dichloroethyl dimethyl phosphate) was obtained from the AMVAC Corporation (Los Angeles, CA, USA). Stock solutions for technical grade permethrin, technical grade permethrin + PBO (1:1 technical grade permethrin/PBO), Permanone, Permanone + PBO (1:1 Permanone/PBO), and naled were prepared for each experiment in high pressure liquid chromatography acetone (99.7% purity; EMD Chemicals, Gibbstown, NJ, USA). The LC50 was determined under laboratory conditions using methods similar to those of Snodgrass (1996) and Antwi and Peterson (2008). Acetone solutions of the different active ingredients listed above were applied in 20-ml glass scintillation vials with a total inside surface area of 40.26 cm2 (Thermo Fisher Scientific Inc., Waltham, MA, USA). A 0.5-ml aliquot of test solution was dispensed into individual glass vials. Acetone was used as the control. Vials were placed on hot dog rollers (model HDR-565, The Helman Group, Ltd., Oxnard, CA, USA) and rotated mechanically so that the acetone dried and the insecticide was uniformly coated in the vial. One insect was placed in each vial and covered with a perforated cap, without diet. Treated vials were placed on large plastic trays and left on the laboratory bench (21.28 ± 0.04oC, photoperiod of 16:8 [L:D] h). Mortality was assessed at 24 h, and insects that did not move when stimulated 73 with forceps were considered dead. Eight treatment replicates were used for each concentration, with five experimental replicates for each insecticide treatment. All combinations of permethrin were run at the same time so that differences between the treatments and times could be analyzed. The experimental design for the permethrin experiments was a complete randomized block with 8 vials (individuals) per concentration, 7 different concentrations (blocks), and 4 treatments (technical grade, technical grade + PBO, Permanone, and Permanone + PBO). To establish the concentration-mortality relationships for Permanone and Permanone mixed with PBO, insects were exposed to 12 concentrations: 0, 0.026, 0.078, 0.13, 0.16, 0.19, 0.26, 0.65, 0.78, 1.04, 1.3, and 2.6 μg/cm2. To establish the concentration-mortality relationships for technical grade permethrin and technical grade permethrin mixed with PBO, insects were exposed to 12 concentrations: 0, 0.015, 0.031, 0.037, 0.07, 0.093, 0.12, 0.16, 0.19, 0.23, 0.31, and 0.78 μg/cm2. To establish the concentration-mortality relationships for naled, insects were exposed to 8 concentrations: 0, 0.08, 0.2, 0.24, 0.32, 0.39, 0.79, and 1.59 μg/cm2. Data Analysis Abbott’s formula was used to correct for control mortality, in which insects were exposed to vials treated with acetone only (Abbott 1925; Perry et al. 1998). Synergist ratios were calculated by dividing the unsynergized LC50 by the synergized LC50 (Casabe et al. 1988). Data were analyzed using Statistical Analysis System 9.1 (SAS Institute, Cary, NC, USA) and dosage-mortality regressions were estimated by probit analysis 74 (PROC PROBIT). However, for Permanone the poor fit of the model was accounted for by multiplying the variances by a heterogeneity factor (χ2/k-2), where k is the number of concentrations to account for extra-binomial variations that causes poor fit (Williams 1982; Robertson et al. 2007). Significant differences (P ≤ 0.05) between permethrin LC50 estimates was determined by the LC50 ratio test of Wheeler et al. (2006). Field Bioassays The study occurred in conjunction with the Cascade County Weed and Mosquito Control District outside of Cascade (N47°13.489', W111°42.040'), and Ulm (N47°25.402', W111°29.767'), MT, USA during the summers of 2007 and 2008, respectively. In 2007 only Trumpet was applied, while in 2008 both Permanone and Trumpet were applied. Each insecticidal product was applied once per year in late July or early August. A truck was equipped with a Bison (VecTec Inc., Orlando, FL, USA) ULV generator. Permanone 10% EC was mixed 1:1 with BVA oil (BVA Inc. Wixom, MI, USA) and was applied at the maximum application rate of 7.85 g active ingredient (ai)/ha with a flow rate of 205 ml/min. Trumpet was applied undiluted at the maximum application rate of 22.42 g ai/ha with a flow rate of 44.36 ml/min. Approximately 15 adults were placed in a 25-cm X 25-cm X 8-cm mesh wire screen cages. One cage of crickets was placed on the ground 25, 50, and 75 m from the spray truck. There were three sample replicates with 200 m buffer zones between replicates (Figures 1-4). Three cages of crickets were placed in a control area. Cricket mortality was assessed 1 h after application because low overnight temperatures caused 100% mortality of control crickets. BoxCox transformations were run on mortality to 75 determine the correct transformation. Analysis of variance (α = 0.05) was run on mortalities that were transformed by ln(y + 1) to determine differences between crickets in the treated and control areas. Non-Target Insect Risk Assessment Ecological risk assessment can be described in quantitative terms as a function of toxicity and exposure (NRC 1983; USEPA 1998). The LC50 values at 24 and 48 h were compared with the actual environmental concentrations (AEC) 25, 50, and 75 m from the spray source during 2007 and 2008 (Chapter 2), and the deposition estimated by ISCST3 (EEC) (for modeling assumptions see Peterson et al. 2006). The deposition concentrations of permethrin for ISCST3 and AEC were 0.18 and 0.0024 μg/cm2. The deposition concentrations of naled for ISCST3 and AEC were 0.342 and 0.036 μg/cm2. I used the risk quotient (RQ) method for estimating risk, which is calculated by dividing the deposition rate by the LC50. Estimated RQs are compared to a RQ level of concern (LOC) which is set by the United State Environmental Protection Agency (USEPA) or another regulatory agency to determine if regulatory action is needed. The RQ LOC used for my assessment was 0.5 (USEPA 2006e). An RQ ≥ 0.5 means the estimated exposure is ≥ 50% of the LC50. Results Toxicity Testing The LC50 values for Permanone, Permanone + PBO, technical grade permethrin, technical grade permethrin + PBO, and naled are given in Table 10. There was generally 76 a good fit to the model assumptions. Technical grade permethrin alone and synergized was significantly more toxic than Permanone alone and synergized. The addition of PBO did not significantly increase the toxicity of Permanone or technical grade permethrin. The 24 h LC50 values for Permanone, Permanone + PBO, technical grade permethrin, and technical grade permethrin + PBO ranged from 0.052 to 0.9 μg/cm2 (Table 10). The 24 h LC50 for naled was 0.44 μg/cm2 (Table 10). The synergist ratio of Permanone + PBO and technical grade permethrin + PBO was 2.65 and 1.57, respectively. Field Bioassay There was no significant difference in mortality between crickets in the control and treated areas for the Permanone (F = 0.08, p = 0.78) and naled spray events (F = 1, p = 0.33). During the 2007 naled application, average mortality was 18% in the treated area and 10% in the control area. During the 2008 permethrin application, average mortality was 5% in the treated area and 6% in the control area. During the 2008 naled application, average mortality was 1% in the treated area and 3% in the control area. Risk Assessment Environmental concentrations estimated by ISCST3 resulted in RQs that exceeded the LOC for Permanone + PBO, technical grade permethrin, technical grade permethrin + PBO, and naled, however Permanone alone did not exceed the RQLOC (Table 11). Risk quotients using the AEC did not exceed the RQ LOC, which were about 10- to 100-fold lower than the RQs using EECs (Table 11). 77 Discussion The toxicity of technical grade permethrin was about 10-fold greater than Permanone. Previous studies have shown that the cis-isomer of permethrin is more toxic than the trans-isomer (Alzogaray et al. 1998). However, the increased toxicity of the technical grade cannot be explained by this phenomenon because technical grade permethrin contained 26.4% cis-isomer, while the Permanone formulation contained 35%. Stevens et al. (2002) found that technical grade organophosphates have a higher toxicity than the formulated products. The increased toxicity of technical grade permethrin could be due to a more rapid penetration through the cuticle because high boiling point oils like those in Permanone can slow the penetration of pyrethroid insecticides (Matsumura 1985; Pree et al. 1996). The toxicity of permethrin to house crickets is similar to that of adult tarnished plant bugs, Lygus lineolaris (Palisot de Beauvois), green stink bugs, Acrosternum hilare (Say), and southern green stink bugs, Nezara viridula (L.), using similar methods to the present study (Snodgrass 1996; Snodgrass et al. 2005). The LC50 values for technical grade permethrin + PBO were similar to technical grade resmethrin ((5-benzyl-3furylmethyl (1RS,3RS;1RS,3SR)-2,2-dimethyl-3-(2-methylprop-1-enyl) Cyclopropanecarboxylate), d-phenothrin (3-phenoxybenzyl (1R)-cis, transchrysanthemate) synergized with PBO (Antwi and Peterson 2008). Piperonyl butoxide is commonly used with pyrethroids for mosquito management because it can suppress or delay the onset of resistance to pyrethroids and pyrethrins as well as reduce the concentration needed to have a lethal effect (Farnham 1998; Xu et al. 78 2005). The results of my study showed a synergist ratio of 1.57 to 2.65, which is similar to previous studies (De Vries and Georghiou 1981; Gist and Pless 1985; Pree et al. 1996). When examining the risks of insecticides to non-target arthropods, it is important to consider the AECs of insecticides and the toxicity of the formulated product. In their ecological risk assessment, Davis et al. (2007) used the conservative model ISCST3 to estimate concentrations of mosquito insecticides that would be deposited on the ground. The concentrations estimated by ISCST3 were 75 times greater for permethrin and 10 times greater for naled than the AECs. The RQ’s for synergized Permanone, synergized and unsynergized technical grade permethrin, and naled exceeded the RQ LOC when compared to the estimated environmental concentrations of ISCST3. However, when the AECs were used, it reduced the RQs for all insecticide combinations below the RQ LOC. The field bioassays support the results of the risk assessment using AECs that the exposure most likely would not result in impacts on ground dwelling insects. My estimates of risks for non-target ground dwelling invertebrates using AECs were lower than the risks using EECs. Even though I used AECs to estimate risks, the RQs most likely still overestimate the risks. For example, to derive the LC50 I coated the insecticide in glass vials so that the crickets were in constant contact with the insecticide. However, in the field non-target organisms most likely are not in constant contact with the insecticide. In addition, house crickets have also been shown to be more sensitive to pyrethroid insecticides than smaller non-target organisms like adult convergent lady beetles (Antwi and Peterson 2008). The results of the risk assessment using AECs are supported by the field bioassays and demonstrate that a single ULV application of 79 synergized or unsynergized permethrin or naled most likely will not result in population impacts on medium- to large-bodied insects. The results for permethrin and naled are different from those of Tietze et al. (1996a), who found that ULV applications of malathion caused significant mortality of sentinel crickets. However, the results of my study are supported by Davis and Peterson (2008), Boyce et al. (2007), and Hill et al. (1971), showing little to no impact on medium- to large-bodied common terrestrial arthropods after single and multiple ULV applications. 80 Table 10. The 24 h lethal concentrations that kill 50% of a population (LC50) for house crickets treated with Permanone, Permanone and piperonyl butoxide (PBO), technical permethrin, technical permethrin and PBO, and naled LC50 C. I. (95%)b dfc Treatment Slope ± SEa χ2 P>χ2 2 (μg/cm ) Permanone 2.73 ± 0.37 0.9 0.71 - 1.29 62.74 9 <0.0001 d Permanone + PBO 4.7 ± 0.51 0.34 0.29 - 0.4 5.25 8 0.73 Technical Permethrin 2.91 ± 0.47 0.082 0.06 - 0.1 13.35 9 0.15 Technical Permethrin 3.41 ± 0.52 0.052 0.039 - 0.065 6.15 9 0.72 + PBO Naled 3.53 ± 0.54 0.44 0.36 - 0.56 7.78 5 0.17 a Standard Error b 95% Confidence Interval for the LC50 c Degrees of Freedom d Piperonyl Butoxide Table 11. Risk quotients (deposition concentration / LC50) for the actual environmental concentration (AEC) and estimated environmental concentrations from ISCST3for Permanone, Permanone + piperonyl butoxide (PBO), technical grade permethrin, technical grade permethrin + PBO, and naled at 24 h Chemical AEC ISCST3 Permanone 0.0027 0.2 a Permanone + PBO 0.0071 0.53 Technical Permethrin 0.029 2.19 Technical Permethrin + PBO 0.046 3.46 Naled 0.081 0.78 a Piperonyl Butoxide 81 CHAPTER 5 ENVIRONMENTAL CONCENTRATIONS, FATE, AND RISK ASSESSMENT OF PYRETHRINS AND PIPERONYL BUTOXIDE AFTER AERIAL ULTRA-LOWVOLUME APPLICATIONS FOR ADULT MOSQUITO MANAGEMENT Abstract One of the most effective ways of managing adult mosquitoes that vector human and animals diseases is the use of ultra-low-volume (ULV) insecticides. Because of concerns about the safety of the insecticides used for the management of adult mosquitoes, I conducted an environmental fate and efficacy study in Princeton and Colusa, California, USA, after aerial applications of pyrethrins ((Z)-(S)-2-methyl-4-oxo3-(penta-2,4-dienyl)cyclopent-2-enyl (1R,3R)-2,2dimethyl-3-(2-methylprop-1enyl)cyclopropanecarboxylate) and piperonyl butoxide (5-[2-(2butoxyethoxy)ethoxymethyl]-6-propyl-1,3-benzodioxole; PBO). One hour before application, concentrations of PBO in water were 0.008 and 0.2175μg/L for Princeton and Colusa, respectively. One hour after the spray event in Princeton, the average concentrations of PBO were 0.0125 μg/cm2 on ground deposition pads and 0.1723 μg/L in surface water samples, with concentrations decreasing significantly over time. One hour after the spray event in Colusa, the average concentrations of PBO were 0.0199 μg/cm2 on deposition pads and 1.274 μg/L in surface water samples, with concentrations decreasing significantly over time. Significant time and location effect for both deposition pads and water samples in Princeton and Colusa was observed (p < 0.001 and 82 p = 0.014). One hour after application, mortality of Culex tarsalis Coquillett and Culex pipiens L. in sentinel cages was significantly higher than the control site at both locations (p < 0.001). Risk quotients for aquatic surrogate species in Princeton and Colusa were ≤ 0.002 1 h after application, which did not exceed the Environmental Protection Agency risk quotient level of concern for endangered aquatic organisms of 0.05. My results suggest that the amounts of pyrethrins and PBO deposited on the ground and in water after aerial ULV-insecticide applications are lower than what has been estimated by previous exposure and risk assessments. Introduction One of the most effective ways of managing high densities of adult mosquitoes that vector human and veterinary pathogens is the use of ultra-low-volume (ULV) aerosol applications of insecticides, which have been the world-wide standard for adult mosquito management for more than 30 years (Mount et al. 1996; Mount 1998). Ultra-low-volume is the minimum effective volume of insecticide that is used as a space spray for adult mosquitoes. Small droplets from 5 to 25 μm are the optimum size for ULV. Smaller droplets tend to travel farther and not settle out of the air column as quickly as larger droplets, making ULV an effective measure for the control of adult mosquitoes seeking a blood-meal. Since West Nile virus (WNV) was accidentally introduced into the United States in 1999, more areas of the country have experienced active mosquito control programs, there has been greater public attention to the human-health and environmental risks 83 associated with ULV insecticide applications (Thier 2001; Roche 2002; Reisen and Brault 2007). Aerial applications of pyrethrins ((Z)-(S)-2-methyl-4-oxo-3-(penta-2,4dienyl)cyclopent-2-enyl (1R,3R)-2,2dimethyl-3-(2-methylprop-1enyl)cyclopropanecarboxylate) and piperonyl butoxide (5-[2-(2butoxyethoxy)ethoxymethyl]-6-propyl-1,3-benzodioxole; PBO) over Sacramento, CA during the 2005 were correlated with a reduction in cases of WNV within the treated area (Carney et al. 2008). However, even with evidence demonstrating a reduction in risk of contracting WNV after mosquito control activities, groups are still concerned about the risks of the insecticides. In response to these concerns, tier I/II risk assessments have been performed to quantify reasonable worst case estimates of risk. Peterson et al. (2006) performed a reasonable worst-case human-health risk assessment for six mosquito insecticide active ingredients, including pyrethrins and the synergist piperonyl butoxide (PBO), after ground-applied ULV applications. Schleier III et al. (2008c) performed a probabilistic human-health risk assessment of the same insecticides and synergist as Peterson et al. (2006) and supported their findings of negligible risk after mosquito adulticide treatments. Davis et al. (2007) and Schleier III et al. (2008b) examined pyrethrins and PBO as well, and found similar results, demonstrating that ecological and equine risks from truck-mounted ULV were most likely negligible. Carr Jr. et al. (2006) showed that the use of aerially applied ULV resmethrin above agricultural fields as a result of a public health emergency would result in negligible human dietary risk. Other biomonitoring and epidemiological studies, reports, and regulatory assessments also have concluded that risks to humans and other non-target organisms from exposure to 84 mosquito insecticides most likely are negligible (Karpati et al. 2004; Currier et al. 2005; NYCDOH 2005; O'Sullivan et al. 2005; Suffolk County 2006; Duprey et al. 2008). Although risk assessments have been performed, they have relied on estimates of environmental concentrations of adulticides from models that do not have algorithms for ULV-type application methods. There are few data on actual environmental concentrations of ULV insecticides after application. Jensen et al. (1999) found nondetectable (ND) concentrations of pyrethrins and permethrin in water sample from wetlands before and after truck-mounted ULV. Concentrations of PBO is Suffolk County, New York were detected in 33% of water samples taken with concentrations ranging from ND to 59.8 μg/L and resmethrin was detected in 11% of water samples taken with concentrations ranging from ND to 0.293 μg/L. Concentrations became ND by four days after helicopter or truck-mounted ULV applications for both PBO and resmethrin (Abbene et al. 2005). Weston et al. (2006) examined water concentrations of pyrethrins and PBO 10 h after the second aerial ULV application of mosquito adulticides and 34 h after the third aerial ULV application of mosquito adulticides over Sacramento, CA. They observed concentrations of 0.44 to 3.92 μg/L of PBO, but did not detect pyrethrins, most likely because pyrethrins rapidly degrade in the environment. Lothrop et al. (2007) measured concentrations of pyrethrins and PBO directly under the aircraft and out to 300 m from the flight path, to maximize efficacy while minimizing evaporation of the insecticide in a desert environment. They observed concentrations of pyrethrins ranging from ND to 0.0791 μg/cm2, and for PBO they observed concentrations ranging from ND to 1.07 μg/cm2. 85 Because of the concerns about the safety of adulticides used for the control of adult mosquitoes, and because of the lack of actual environmental concentration data, I conducted an environmental fate and efficacy study. The objectives of the study were to characterize terrestrial surface residues and surface water concentrations of pyrethrins and PBO in an urban setting after a typical aerial application. To accomplish my objective, I examined the concentrations of insecticide present before and after a single ULV application, and related those to refining risk assessments for ULV insecticides. Materials and Methods The insecticide Evergreen® EC 60-6 (60% PBO and 6% pyrethrins; McLaughlin Gormley King MGK®, Golden Valley, MN, USA) was applied in Princeton (39°24'17.58"N, 122°0'38.58"W) and Colusa (39°12'23.76"N, 122°0'24.44"W), California, USA on June 26 and 28, 2007, respectively. Application over Princeton was at 2022 and Colusa was at 2052 PDT. Undiluted Evergreen EC 60-6 was applied aerially at a rate of 2.8 g/ha pyrethrins and 28 g/ha PBO using a fixed wing Piper Aztec aircraft by Adapco Vector Control Services, in conjunction with Sacramento-Yolo Mosquito and Vector Control District. The release height of insecticide from the aircraft for Princeton and Colusa was 61 and 91 m above the ground, respectively. Temperature, wind speed, and relative humidity at ground level were recorded with a Kestrel® 4000 pocket weather tracker (Nielsen-Kellerman, Boothwyn, PA, USA). At each location (Colusa and Princeton), I used four sampling sites (Figures 14 and 15), with two sub-samples placed 1 m apart at each sampling site. Samples were 86 taken from both surface water and ground deposition at four time intervals (1, 12, 24, and 36 h) following the spray event. Ground deposition samples and water were collected 1 h before spraying to determine background levels of pyrethrins and PBO. At each spray location, there were 76 samples collected with four sites, two sub-samples, two media (surface water and ground deposition), four times (1, 12, 24, and 36 h post-application), four background samples, one spike (positive control), and one negative control. Collection of surface residues at ground level was on 10-cm x 10-cm (100 cm2) cotton dosimeters pinned to a cardboard backing. The cardboard backing was covered with plastic wrap to prevent contact between the cardboard and cotton pads. Four cotton pads were pinned to each cardboard backing before the spray event at each location. One pad was collected from each sub-sample at each of the times listed above. Four blank and four spiked cotton pads were placed in a control (untreated) area. Controls for Princeton and Colusa were located 4 and 6 km, respectively, upwind of the spray zones. In Princeton and Colusa, deposition pads were spiked with 0.133 and 0.325 μg/cm2 of pyrethrins and 0.07 and 0.023 μg/cm2 of PBO, respectively. Spikes and controls were set out when the aircraft finished spraying and one pad from each was collected at the times listed above. Cotton pads were placed in open areas where minimal vegetation occurred to represent a worst-case assessment of ground deposition. Cotton pads were collected with tweezers which were rinsed with high pressure liquid chromatography grade acetone between pads to prevent cross contamination. Individual samples were stored in separate 120-ml I-Chem™ glass jars with Teflon® lids (Chase 87 Scientific Glass, Vineland, NJ, USA). After collection, jars were immediately placed on dry ice for transport to Caltest Analytical Laboratory (Napa, CA, USA) for analysis. Water samples were collected from flowing irrigation ditches in Princeton and static golf course ponds in Colusa, which were the only available water bodies in each town. Water was taken approximately 2 to 6 cm under water in 1-L amber I-Chem glass jars with Teflon lids and immediately placed on dry ice and transferred to liquid ice after returning to the Sacramento-Yolo Mosquito and Vector Control District Laboratory. The analytical laboratory analyzed the six active chemical components of pyrethrins (cinerin I and II, jasmolin I and II, and pyrethrins I and II) as well as PBO. Both cotton pads and water were extracted with dichloromethane. Extraction was done using U.S. Environmental Protection Agency (USEPA) SW846 method 3550B (USEPA 1996b) for deposition pads and 3510C (USEPA 1996a) for water. Analysis was done using a gas chromatography mass spectrometry (GC/MS) with USEPA SW846 method 8270 (USEPA 1996c). I used Statistical Analysis System 9.1 (SAS Institute, Cary, NC, USA) to run repeated measures analysis of variance (α = 0.05) using a mixed model on log transformed concentrations to determine differences between times and treatments, and performed regression analysis of chemical components with respect to time using SigmaPlot® 8 (SPSS, Chicago, IL, USA) (Helsel 2005). For ND concentrations, I substituted half of the detection limit when the NDs were less than 10% of the PBO data (Lubin et al. 2004). 88 Mosquito Bioassay Disposable bioassay cylindrical cardboard cages 15-cm diameter and 4.5-cm deep were used with 14 x 18 polyester mesh screens covering the vertical circular surfaces of the cage. A hole was placed in the side of the cardboard where cotton pads moistened with sugar water were used (Townzen and Natvig 1973). One cage containing approximately 25 laboratory-reared adult Culex tarsalis Coquillett and one cage containing approximately 25 laboratory-reared adult Culex pipiens L. was placed at each of nine sites within the spray zone. Eighteen additional cages (one of each mosquito species per site) were placed on nine sites outside of the spray zone in Williams, California, USA, to serve as controls during each spray event. Cages were placed vertically at 1 m with a screened surface positioned to face the predominant wind direction at a 45o angle (Bunner et al. 1989). Mosquito mortality was evaluated at the time of placement and at 1 h after the spray event. Cages were then brought to the Sacramento-Yolo Mosquito and Vector Control District Laboratory and re-evaluated at 2, 12, and 24 h after the spray event. I ran mixed model repeated measures analysis of variance (α = 0.05) on log transformed mortality percentages to determine significant differences between times, treatments, locations, and species. Results The analytical detection limits for both water and deposition pads are presented in Table 12. The average ambient air temperature at ground level at the time of the spray was 31oC at Princeton and 24oC at Colusa. Average wind speed for Princeton at ground 89 level was 8.6 km/h with a peak wind gust of 15.8 km/h out of the southeast and a relative humidity of 35.5%. Average wind speed for Colusa was 2.7 km/h with a peak wind gust of 9.3 km/h from the west southwest and a relative humidity of 55.9%. At the release height in Princeton and Colusa, winds were variable at 11-19 km/h and 22 km/h, respectively. There was a significant time and location (Princeton vs. Colusa) effect for deposition pads (F = 13.7, 15.35, p < 0.001). In Princeton, pyrethrins were not detected on the deposition pads (Table 13). Pyrethrins and PBO were not detected on any of the background samples taken 1 h before the spray event. One hour after the spray event, the average concentration of PBO was 0.0125 μg/cm2 (Table 13). On the spiked pads, pyrethrins were ND after 12 h, with PBO decreasing with time (Table 13). There were no detectable amounts of any compound on the control pads. There was a significant exponential decay relationship for the deposition pads (F = 7.23, p = 0.018, r2 = 0.341, y = 0.0138e-0.0437x) (Figure 16). Pyrethrins were not detected on any deposition pads in Colusa, nor were pyrethrins detected on any of the background samples taken 1 h before the spray event (Table 13). However, PBO was detected in one background sample at a concentration of 0.031 μg/cm2. On the spiked pads, pyrethrins were ND on all pads, with PBO decreasing with time (Table 13). There were no detectable amounts of any compound on the control pads. One hour after the spray event, average concentrations of PBO were 0.02 μg/cm2 but concentrations decreased 65% to 0.007 μg/cm2 36-h after the spray (Table 13). There 90 was a significant exponential decay relationship for the deposition pads (F = 14.66, p = 0.002, r2 = 0.51, y = 0.0218e-0.0352x) (Figure 16). Pyrethrins were not detected in any water samples in Princeton before or after the spray event (Table 14). There was background PBO in the water at an average concentration of 0.008 μg/L (Table 14). One hour after application the average concentration was 0.172 μg/L (Table 14). Concentrations decreased approximately 77% between 1 and 12 h (Figure 17, Table 14). Concentrations 1 h after the spray event were significantly higher than the background samples taken 1 h before the spray event (F = 28.25, p = 0.0001). There was no significant difference between background samples and concentrations at 36 h after application (F = 0.43, p = 0.52). There was a significant exponential decay relationship for the water samples (F = 21.37, p = 0.0005, r2 = 0.622, y = 0.5661e-1.2039x) (Figure 17). In Colusa, the background concentrations of PBO in water averaged 0.218 μg/L (Table 14). One hour after application, the average concentration was 1.274 μg/L, with the average 36 h concentrations dropping below the background concentration (Table 14, Figure 17). Concentrations 1 h after the spray event were significantly higher than the background samples taken 1 h before the spray event (F = 26.25, p = 0.0001). There was no significant difference between background samples and concentrations at 24 h after application (F = 2.71, p = 0.12). There was a significant exponential decay relationship for the water samples (F = 11.53, p = 0.004, r2 = 0.452, y = 1.3446e-0.0478x) (Figure 17). There was a significant time and location (Princeton vs. Colusa) effect for water samples (F = 16.85, 233.5, p < 0.0001). 91 Pyrethrins were not detected in the background water samples or 1, 12, and 24 h after application; however, 36 h after application there were detectable amounts of jasmolin II in five of the eight sub-samples (Table 14). The concentration of jasmolin II for five of the eight sub-samples was 0.213 μg/L. The demonstrated rapid decline of pyrethrins observed during this study and the magnitude of the measured concentrations make it unlikely to be due to the aerial application. Sentinel cage bioassay data in Princeton showed that mortality 1 h after application was 40% for C. tarsalis and 22% for C. pipiens (Table 15), and Colusa mortality 1 h after application was 73% for C. tarsalis and 77% for C. pipiens (Table 16). At 24 h after application Princeton showed a mortality rate of 86% for C. tarsalis and 88% for C. pipiens (Table 15), Colusa showed 96% mortality for both C. tarsalis and C. pipiens (Table 16). Control mortality at 1, 2, 12, and 24 h after application for both Princeton and Colusa did not exceed 5% (Table 15 and 16). Mortality 1, 2, 12, and 24 h after the applications in Princeton and Colusa for both C. tarsalis and C. pipiens was significantly different than mortality in the control town of Williams (F = 1400, p < 0.0001). There also was a significant time and location (Princeton vs. Colusa) effect (F = 11.93, p < 0.0001 and F = 6.17, p = 0.014, respectively). There was no significant difference in mortality between the two mosquito species (F = 0.88, p = 0.348). Discussion There were similarities between the ground and water deposition data from Princeton and Colusa, with concentrations peaking 1 h after application and decreasing 92 significantly over time in an exponential decay pattern. Pyrethrins were not detected on the deposition pads in either Princeton or Colusa. The current study found ND concentrations of pyrethrins and lower concentrations of PBO with higher mosquito mortality than the values reported by Lothrop et al. (2007). They measured concentrations and efficacy directly under the aircraft and at distances out to 300 m, while I measured concentrations within the spray zone that were 100 m to over 300 m from the aircraft spray path. Lothrop et al. (2007) observed an average mortality ranging from only 1.5 to 12% 1 h after application. Their observed 2-h depositions of PBO were 17 and nine times greater than what I observed in Princeton and Colusa, respectively. However, Lothrop et al. (2007) did not report the analytical detection limits, and obtained samples at only one time after application. The significant location effect in water may be explained by the two different water sources sampled in each town. The results of the present study support the findings of Weston et al. (2006) of no detectable amounts of pyrethrins in the water after a spray event. In my study and in that of Weston et al. (2006), there were detectable levels of PBO in all of the background samples. Piperonyl butoxide concentrations at Princeton 1 h after application were lower than any concentration that Weston et al. (2006) reported in their study. In the Colusa static ponds, 1-h concentrations were lower than four of the seven sites sampled by Weston et al. (2006), which were sampled from flowing creeks. After 12 h in the Colusa ponds, concentrations were less than what Weston et al. (2006) reported for 34 h after a third application by aircraft. Concentrations from Princeton 12 h after application were 11-fold lower than the lowest concentration Weston et al. (2006) 93 reported for 10 h after application. T At 36 h after application, concentrations at my sites were lower than the lowest concentration reported by Weston et al. (2006) by 82-fold in Princeton and13-fold in Colusa. The amount of pyrethrins and PBO deposited on the ground after aerial ULV application was much lower than what was estimated by previous risk assessments using modeled environmental concentrations after truck-mounted ULV. The amount of PBO deposited on the ground was 47- to 76-fold less than that estimated by Peterson et al. (2006) and Davis et al. (2007) using the tier-1 model, ISCST3 (USEPA 1995). Amounts of PBO were two- to three-fold less than what Schleier III et al. (2008b) estimated using the model, AgDrift® (Teske et al. 2002). This indicates that exposures and concomitant human-health and ecological risks from aerial applications most likely are much lower than ground applications of pyrethrins and PBO. The insecticides used for adult mosquito management are most toxic to aquatic invertebrates and vertebrates. Using the same toxic endpoints and surrogate species as Davis et al. (2007), the risk quotients (RQ) (environmental concentration/toxic endpoint) for PBO samples taken 1 h before the application for Daphnia magna, rainbow trout (Oncorhynchus mykiss Walbaum), and bluegill sunfish (Lepomis macrochirus Rafinesque) in Princeton were <0.0001, and in Colusa 0.0004, <0.0001, and 0.0001, respectively. Risk quotients for PBO 1 h after application in Princeton were 0.0003 for D. magna, <0.0001 for rainbow trout, and <0.0001 for bluegill sunfish; in Colusa RQ’s were 0.0002 for D. magna, 0.0003 for rainbow trout, and 0.0007 for bluegill sunfish. The 94 increase in RQ’s for PBO were well below the USEPA RQ level of concern of 0.05 for endangered aquatic organisms. My results and those of previous studies suggest that the risk of PBO and pyrethrins applied aerially for mosquito control is negligible for aquatic vertebrates and invertebrates. Furthermore, my results suggest that levels of PBO in water return to baseline levels by 36 h after application. However, because PBO is a synergist for pyrethroid insecticides, there is a potential for it to synergize with pyrethroids already present in the aquatic sediment and water column (Paul et al. 2005; Amweg et al. 2006a; Weston et al. 2006). Weston et al. (2005) found seven different pyrethroid insecticides in creek sediments in Sacramento, California, USA most likely occurring from residential use, and not from agricultural or mosquito control practices. Pyrethrins are unstable compounds, especially in sunlight, and even with spikes above the detection limits of the analytical laboratory, little was recovered from the Princeton spikes and there were ND levels in Colusa. If PBO is used as a surrogate for potential concentrations of pyrethrins, we would have seen 0.002 and 0.001 μg/cm2 1 h after application in Princeton and Colusa, respectively. These values are 103- and 205.8fold less than the concentrations modeled in ISCST3 and six- and 12-fold smaller than the concentrations modeled in AgDrift (Peterson et al. 2006; Davis et al. 2007; Schleier III et al. 2008b). My results show the amount of pyrethrins and PBO deposited on the ground and in the water after aerial ULV application is lower than what was estimated by previous risk assessments using modeled environmental concentrations after truck-mounted ULV. 95 These concentrations decreased with time, and in the case of water were not significantly different from background levels by 24 to 36 h after the application. Future research should be directed towards developing a model that can better predict both truck-mounted and aerial ULV-insecticide applications for adult mosquito management. Table 12. Analytical detection limits for water and deposition pads for each compound Compound Water (μg/L) Pad (μg/cm2) Cinerin I 0.05 0.02 Cinerin II 0.05 0.02 Jasmolin I 0.05 0.02 Jasmolin II 0.05 0.02 Piperonyl Butoxide 0.001 0.0007 Pyrethrin I 0.1 0.05 Pyrethrin II 0.5 0.2 Table 13. Mean piperonyl butoxide ground deposition concentrations in μg/cm2 (± standard error) at 1, 12, 24, and 36 h after application in Princeton and Colusa, CA Sample Time Princeton Colusa 1h 0.0125 ±0.004 0.0199 ±0.003 12 h 0.0105 ±0.006 0.0182 ±0.002 24 h 0.0028 ±0.0009 0.0055 ±0.0007 36 h 0.003 ±0.001 0.0072 ±0.0009 Control 1 h ND ND Spike 1 h 0.11 0.056 Control 12 h ND ND Spike 12 h 0.19 0.02 Control 24 h ND ND Spike 24 h 0.012 0.0095 Control 36 h ND ND Spike 36 h 0.025 0.011 ND= Non-detectable concentrations 96 Table 14. Mean piperonyl butoxide water concentrations 1 h before the spray event (background) and 1, 12, 24, and 36 h after the spray event in μg/L (± standard error) from irrigation ditches located in Princeton and static ponds in Colusa, CA Sample Time Princeton Colusa Background 0.008 ±0.002 0.218 ±0.012 1h 0.172a ±0.036 1.274a ±0.328 12 h 0.039b ±0.012 0.791 ±0.067 b 24 h 0.029 ±0.014 0.366b ±0.046 36 h 0.012b ±0.002 0.073b ±0.009 a significantly different than 1 h before application b significantly different than 1 h after application Table 15. Mean percent mortality (± standard error) for Culex pipiens and Culex tarsalis in bioassay cages in Princeton (treated) and Williams, CA (control). Princeton Williams Time (h) C. pipiens C. tarsalis 40b ±11.89 1 22.45a ±7.19 2 55.26a ±14.79 57.25b ±13.86 a 12 78.37 ±9.18 76.67b ±8.18 24 85.63a ±6.95 87.88b ±6.8 a significantly different than C. pipiens control b significantly different than C. tarsalis control C. pipiens C. tarsalis 0 0 1.18 ±0.78 4.5 ±1.33 0.34 ±0.34 0.34 ±0.34 0.34 ±0.34 0.80 ±0.53 Table 16. Mean percent mortality (± standard error) for Culex pipiens and Culex tarsalis in bioassay cages in Colusa (treated) and Williams, CA (control). Colusa Williams Time (h) C. pipiens C. tarsalis C. pipiens C. tarsalis a b 1 72.63 ±12.86 0 0 77.16 ±9.54 a b 2 81.31 ±8.42 83.57 ±7.34 1.30 ±0.67 0 12 91.31b ±7.3 1.30 ±0.67 0 93.78a ±3.49 a b 24 95.5 ±2.7 95.47 ±4.53 1.30 ±0.67 0 a significantly different than C. pipiens control b significantly different than C. tarsalis control 97 Figure 14. Locations where deposition (patch) and water (water) samples were taken in Princeton, CA 98 Figure 15. Locations where deposition (patch) and water (water) samples were taken in Colusa, CA 99 0.018 a 0.016 0.014 0.012 0.010 0.008 0.006 PBO (μg/cm2) 0.004 0.002 0.000 0.025 b 0.020 0.015 0.010 0.005 0.000 0 10 20 30 Time (h) Figure 16. Change in piperonyl butoxide concentrations on deposition pads from 1 to 36 h after the spray with exponential decay curve for Princeton (a) (F = 7.23, p = 0.018, r2 = 0.341, y = 0.0138e-0.0437x) and Colusa, CA (b) (F = 14.66, p = 0.002, r2 = 0.51, y = 0.0218e-0.0352x). 100 0.25 a 0.20 0.15 0.10 PBO (μg/L) 0.05 0.00 1.8 b 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 10 20 30 Time (h) Figure 17. Change in piperonyl butoxide concentrations in water from 1 to 36 h after the spray with exponential decay curve for Princeton (a) (F = 21.37, p = 0.0005, r2 = 0.622, y = 0.5661e-1.2039x) and Colusa, CA (b) (F = 11.53, p = 0.004, r2 = 0.452, y = 1.3446e0.0478x ). 101 CHAPTER 6 CONCLUSION I measured deposition of the insecticides permethrin ([3-phenoxyphenyl]methyl 3-[2,2-dichloroethenyl]-2,2-dimethylcyclopropane carboxylate) and naled (1,2-dibromo2, 2-dichloroethyl dimethyl phosphate) at ground level and 1.25 m above the ground and air concentrations after truck-mounted ultra-low-volume (ULV) applications. Concentrations measured in the field showed that the insecticides moved rapidly through the sample collectors because there was no significant difference between samples taken 1 and 12 h after application and no air concentrations were measured after the 1-h sampling interval. There was a significant year effect, indicating that environmental conditions significantly influenced the deposition of the insecticides. Generally, concentrations decreased as distance from the spray source increased. One hour after application approximately 4.4, 3.9, and 1.3% of the permethrin sprayed settled onto the ground 25, 50, and 75 m from spray source, respectively. One hour after application approximately 28, 16, and 14.9% of the naled sprayed settled onto the ground 25, 50, and 75 m from spray source, respectively. Overall, approximately 3.2 and 15% of the permethrin and naled sprayed deposited on the ground within 75 m from the spray source 1 h after application. I conducted a probabilistic human-health risk assessment using the actual environmental concentrations of permethrin and naled I measured from the field experiments. My human-health risk assessment was approximately 5 to 400 times lower 102 than previous risk assessments. Previous risk assessments used conservative exposure scenarios that overestimated risks, thus they were conservative in protecting human health, which is also supported by biomonitoring studies for permethrin and naled exposure after ULV applications. Previous human-health risk assessments used ISCST3, AERMOD, and the United States Environmental Protection Agency (USEPA) Pesticide Handler Exposure Database (PHED). My results suggest that both ISCST3 and AERMOD are sufficiently conservative models for conducting lower tiered risk assessments, but PHED may underestimate dermal exposures. Although dermal exposures may have been underestimated, the estimated risks were still lower than previous risk assessments because the use of the actual environmental concentrations from this study reduced the exposure from other routes. However, my results suggest that AGDISP, which was used by the USEPA in the registration of d-phenothrin (3-phenoxybenzyl (1R)-cis, transchrysanthemate) for mosquito management, underestimated environmental concentrations of permethrin and naled, and thus risks. I used the house cricket, Acheta domesticus (L.), as a indicator species for largeto medium-bodied ground dwelling invertebrates. I used the house cricket to determine the laboratory derived lethal concentration that kills 50% of a population (LC50) for synergized and unsynergized permethrin for the formulated product used in the field and technical grade permethrin and the formulated product of naled. The deterministic nontarget risk assessment for medium- to large-bodied ground dwelling insects using estimated environmental concentrations resulted in risks that exceeded levels of concern 103 for most active ingredient combinations assessed. However, actual environmental concentrations of permethrin and naled did not result in exposures that exceeded regulatory levels of concern. The field bioassay experiment showed no significant mortality of caged crickets exposed to permethrin and naled applied by truck-mounted ULV. The results of the risk assessment using actual environmental concentrations was supported by the field bioassays demonstrating that a single ULV application of synergized or unsynergized permethrin or naled most likely will not result in population impacts on medium- to large-bodied insects. In addition, to measuring actual environmental concentrations and conducting human-health and ecological risk assessments for truck-mounted ULV, I also conducted an environmental fate, efficacy, and ecological risk assessment for aerial ULV applications of pyrethrins ((Z)-(S)-2-methyl-4-oxo-3-(penta-2,4-dienyl)cyclopent-2-enyl (1R,3R)-2,2dimethyl-3-(2-methylprop-1-enyl)cyclopropanecarboxylate) and piperonyl butoxide (5-[2-(2-butoxyethoxy)ethoxymethyl]-6-propyl-1,3-benzodioxole; PBO). Pyrethrins were not detected in the water or on deposition pads. Piperonyl butoxide was detected in water and deposition samples, but concentrations rapidly decreased to background levels by 36 h after application. The estimated risks for aquatic surrogates such as Daphnia magna and rainbow trout, Oncorhynchus mykiss, from aerial ULV applications of pyrethrins and PBO were lower than previous ecological risk assessments. In addition, the aerial ULV applications caused significant mortality of Culex tarsalis and Culex pipiens in sentinel cages. 104 My results for both truck-mounted and aerial ULV applications reveal that models currently used to estimate environmental concentrations of insecticides applied via ULV are inadequate for accurately estimating environmental concentrations. Because of lack of studies and uncertainties about the fate of insecticides during ULV applications, a model needs to be developed specifically for ULV applications. A more extensive study with additional temporal and spatial data points should be conducted to examine the effect of wind speed, air temperature, humidity, and other factors that influence the behavior of ULV aerosols. However, even with the current understanding of the fate of ULV insecticides, estimates of risks for truck-mounted and aerial ULV applications using actual environmental concentrations were lower than previous risk assessments using estimated environmental concentrations. Results of my environmental fate studies, human-health and non-target risk assessments, and the current weight of scientific evidence demonstrates that the risks to humans and other non-target organisms after ULV applications of insecticides most likely are below regulatory levels of concern. 105 REFERENCES CITED Abbene, I.J., S.C. Fisher, and S.A. Terracciano (U.S. Geological Survey). 2005. Concentrations of insecticides in selected surface water bodies in Suffolk County, New York, before and after mosquito spraying, 2002-04. Open-File Report 20051384. New York, NY. 1-14. Abbott, W.S. 1925. A method of computing the effectiveness of an insecticide. Journal of Economic Entomology 18: 265-267. Alzogaray, R.A., M.I. Picollo, and E.N. Zerba. 1998. Independent and joint action of cisand trans-permethrin in Triatoma infestans (Hemiptera: Reduviidae). Archives of Insect Biochemistry and Physiology 37: 225–230. Amweg, E.L., D.P. Weston, C.S. Johnson, J. You, and M.J. Lydy. 2006a. Effect of piperonyl butoxide on permethrin toxicity in the amphipod Hyalella azteca. Environmental Toxicology and Chemistry 25: 1817-1825. Amweg, E.L., D.P. Weston, J. You, and M.J. Lydy. 2006b. Pyrethroid insecticides and sediment toxicity in urban creeks from California and Tennessee. Environmental Science and Technology 40: 1700-1706. Antwi, F., L.M. Shama, and R.K.D. Peterson. 2008. Risk assessments for the insect repellents DEET and picaridin. Regulatory Toxicology and Pharmacology 51: 3136. Antwi, F.B., and R.K.D. Peterson. 2008. Toxicity to non-target insects after exposure to δ-phenothrin and resmethrin. Pest Management Science in press. Atkinson, P.W., A.C. Pinkerton, and D.A. O'Brochta. 2001. Genetic transformation systems in insects. Annual Review of Entomology 46: 317-346. Bass, E.L. 1986. Use of the cricket Acheta domestica L. as a bioassay organism for the toxic extract from Gonyaulax monilata (dinophyceae). Journal of Phycology 22: 546-548. Berteau, P.A., and W.A. Dean. 1978. A comparison of oral and inhalation toxicities of four insecticides to mice and rats. Bulletin of Environmental Contamination and Toxicology 19: 113-120. Berteau, P.E., W.A. Deen, and R.L. Dimmick. 1977. Effect of particle size on the inhalation toxicity of naled aerosols. Toxicology and Applied Pharmacology 41: 183. 106 Boyce, W.M., S.P. Lawler, J.M. Schultz, M. S.J., L.S. Kimsey, M.K. Niemela, C.F. Nielsen, and W.K. Reisen. 2007. Nontarget effects of the mosquito adulticide pyrethrin applied aerially during a West Nile virus outbreak in an urban California environment. Journal of the American Mosquito Control Association 23: 335-339. Brieger, G., J.R. Wells, and R.D. Hunter. 1992. Plant and animal species composition and heavy metal content in fly ash ecosystems. Water Air and Soil Pollution 63: 87103. Brochu, P., J.F. Ducre-Robitaille, and J. Brodeur. 2006. Physiological daily inhalation rates for free-living individuals aged 2.6 months to 96 years based on doubly labeled water measurements: comparison with time-activity-ventilation and metabolic energy conversion estimates. Human and Ecological Risk Assessment 12: 736-761. Bunner, B.L., F.G. Posa, S.E. Dobson, F.H. Broski, and L.R. Boobar. 1989. Aerosol penetration relative to sentinel cage configuration and orientation. Journal of the American Mosquito Control Association 5: 547-551. Bunning, M.L., R.A. Bowen, C.B. Cropp, K.G. Sullivan, B.S. Davis, N. Komar, M.S. Godsey, D. Baker, D.L. Hettler, D.A. Holmes, B.J. Biggerstaff, and C.J. Mitchell. 2002. Experimental infection of horses with West Nile virus. Emerging Infectious Diseases 8: 380-386. Burmaster, D.E., and A.M. Wilson. 1996. An introduction to second-order random variables in human health risk assessments. Human and Ecological Risk Assessment 2: 892-919. Carney, R.M., S. Husted, C. Jean, C. Glaser, and V. Kramer. 2008. Efficacy of aerial spraying of mosquito adulticide in reducing incidence of West Nile virus, California, 2005. Emerging Infectious Diseases 14: 747-754. Caron, D.M. 1979. Effects of some ULV mosquito abatement insecticides on honey bees. Journal of Economic Entomology 72: 148-151. Carr, W.C., P. Iyer, and D.W. Gammon. 2006. A dietary risk assessment of the pyrethroid insecticide resmethrin associated with its use for West Nile virus mosquito vector control in California. The Scientific World Journal 6: 279-290. Casabe, N., F. Melgar, E.J. Wood, and E.N. Zerba. 1988. Insecticidal activity of pyrethroids against Triatoma infestans. Insect Science and its Application 9: 233236. CDC (U.S. Department of Health and Human Services, Public Health Service ,Centers for Disease Control and Prevention National Center for Infectious Diseases, Division of Vector-Borne Infectious Diseases). 2003. Epidemic/epizootic West 107 Nile virus in the United States: guidelines for surveillance, prevention, and control, 3rd revision. Fort Collins, CO. Chevron (Chevron Chemical Company). 1975. What you need to know about Dibrom 14®. San Francisco. 1-24. Cullen, A.C., and H.C. Frey. 1999. Probabilistic techniques in exposure assessment. Plenum Press, New York, N.Y. Currier, M., M. McNeill, D. Campbell, N. Newton, J.S. Marr, E. Perry, S.W. Berg, D.B. Barr, G.E. Luber, S.M. Kieszak, H.S. Rogers, S.C. Backer, M.G. Belson, C. Rubin, E. Azziz-Baumgartner, and Z.H. Duprey. 2005. Human exposure to mosquito-control pesticides---Mississippi, North Carolina, and Virginia, 2002 and 2003. MMWR Morbidity and Mortality Weekly Report 54: 529-532. Dauphin, G., S. Zientara, H. Zeller, and B. Murgue. 2004. West Nile: worldwide current situation in animals and humans. Comparative Immunology Microbiology & Infectious Diseases 27: 343-355. Davis, L.E., J.D. Beckham, and K.L. Tyler. 2008. North American encephalitic arboviruses. Neurologic Clinics 26: 727-757. Davis, R.S., and R.K.D. Peterson. 2008. Effects of single and multiple applications of mosquito insecticides on nontarget arthropods. Journal of the American Mosquito Control Association 24: 270-280. Davis, R.S., R.K.D. Peterson, and P.A. Macedo. 2007. An ecological risk assessment for insecticides used in adult mosquito management. Integrated Environmental Assessment and Management 3: 373-382. De Vries, D.H., and G.P. Georghiou. 1981. Absence of enhanced detoxification of permethrin in pyrethroid resistant house flies. Pesticide Biochemistry and Physiology 15: 242-252. Duprey, Z., S. Rivers, G. Luber, A. Becker, C. Blackmore, D. Barr, G. Weerasekera, S. Kieszak, W.D. Flanders, and C. Rubin. 2008. Community aerial mosquito control and naled exposure. Journal of the American Mosquito Control Association 24: 42-46. Eidson, M., L. Kramer, W. Stone, Y. Hagiwara, and K. Schmit. 2001. Dead bird surveillance as an early warning system for West Wile virus. Emerging Infectious Diseases 7: 631-635. Eifert, J.D., G.C. Sanglay, D.J. Lee, S.S. Sumner, and M.D. Pierson. 2006. Prediction of raw produce surface area from weight measurement. Journal of Food Engineering 74: 552-556. 108 Farnham, A.W. 1998. The mode of action of piperonyl butoxide with reference to studying pesticide resistance. In: D. G. Jones (ed) Piperonyl butoxide: the insecticide synergist. Academic Press, London Frey, H.C. 1992. Quantitative analysis of uncertainty and variability in environmental policy making, AAAS/EPA Environmental Science and Engineering Fellow. American Association for the Advancement of Science, Washington D.C. Frey, H.C., and D.S. Rhodes. 1996. Characterizing, simulating, and analyzing variability and uncertainty: An illustration of methods using an air toxics emissions example. Human and Ecological Risk Assessment 2: 762-797. Gist, G.L., and C.D. Pless. 1985. Synergistic activity of piperonyl butoxide with nine synthetic pyrethroids against the fall armyworm, Spodoptera frugiperda. Florida Entomologist 68: 316-319. Haile, D.G., G.A. Mount, and N.W. Pierce. 1982. Effect of droplet size of malathion aerosols on kill of caged adult mosquitoes. Mosquito News 42: 576-582. Harrington, T., M.J. Kuehnert, H. Kamel, R.S. Lanciotti, S. Hand, M. Currier, M.E. Chamberland, L.R. Petersen, and A.A. Marfin. 2003. West Nile virus infection transmitted by blood transfusion. Transfusion 43: 1018-1022. Harris, C.R. 1966. Influence of soil type on activity of insecticides in soil. Journal of Economic Entomology 59: 1221-1225. Harwood, R.F., and M.T. James. 1979. Entomology in human and animal health. 7th ed. Macmillan Publishing Co., Inc, New York. Helsel, D.R. 1990. Less than obvious: statistical treatment of data below the detection limit. Environmental Science and Technology 24: 1766-1774. Helsel, D.R. 2005. Computing Summary Statistics. Nondetects and Data Analysis. pp 55-79. John Wiley & Sons, Hoboken, NJ, USA Hennessey, M.K., H.N. Nigg, and D.H. Habeck. 1992. Mosquito (Diptera, Culicidae) adulticide drift into wildlife refuges of the Florida Keys. Environmental Entomology 21: 714-721. Hester, P.G., K.R. Shaffer, N.S. Tietze, H. Zhong, and J. Griggs Jr. 2001. Efficacy of ground applied ultra low volume malathion on honey bee survival and productivity in open and forest areas. Journal of the American Mosquito Control Association 17: 2-7. 109 Higgs, S., B.S. Schneider, D.L. Vanlandingham, K.A. Klingler, and E.A. Gould. 2005. Nonviremic transmission of West Nile virus. Proceedings of the National Academy of Sciences of the United States of America 102: 8871-8874. Hill, E.F., D.A. Eliason, and Kilpatri.Jw. 1971. Effects of ultra-low volume applications of malathion in Hale County, Texas III. Effect on nontarget animals. Journal of Medical Entomology 8: 173-179. Ho, S.S., D. Brossard, and D.A. Scheufele. 2007. The Polls - Trends - Public reactions to global health threats and infectious diseases. Public Opinion Quarterly 71: 671692. Hoffmann, B.D., L.M. Lowe, and A.D. Griffiths. 2002. Reduction in cricket (Orthoptera : Ensifera) populations along a gradient of sulphur dioxide from mining emissions in northern Australia. Australian Journal of Entomology 41: 182-186. Iwamoto, M., D.B. Jernigan, A. Guasch, M.J. Trepka, C.G. Blackmore, W.C. Hellinger, S.M. Pham, S. Zaki, R.S. Lanciotti, S.E. Lance-Parker, C.A. DiazGranados, A.G. Winquist, C.A. Perlino, S. Wiersma, K.L. Hillyer, J.L. Goodman, A.A. Marfin, M.E. Chamberland, L.R. Petersen, P. Blake, W. Bower, L. Dowdy, J. Fleming, J. Guarner, J. Jimenez, M. Kuehnert, F. Leguen, T. Luu, S. Mallon, R. Moseley, A. Nejman, P. Page, L. Pealer, X.S. Qi, E. Rico, J. Roehrig, P. Rollin, M. Salameh, W.J. Shieh, P. Tso, and D. Withum. 2003. Transmission of West Nile virus from an organ donor to four transplant recipients. New England Journal of Medicine 348: 2196-2203. Jensen, T., S.P. Lawler, and D.A. Dritz. 1999. Effects of ultra-low volume pyrethrin, malathion, and permethrin on nontarget invertebrates, sentinel mosquitoes, and mosquitofish in seasonally impounded wetlands. Journal of the American Mosquito Control Association 15: 330-338. Karpati, A.M., M.C. Perrin, T. Matte, J. Leighton, J. Schwartz, and R.G. Barr. 2004. Pesticide spraying for West Nile virus control and emergency department asthma visits in New York City, 2000. Environmental health perspectives 112: 11831187. Knepper, R.G., E.D. Walker, S.A. Wagner, M.A. Kamrin, and M.J. Zabik. 1996. Deposition of malathion and permethrin on sod grass after single, ultra-low volume applications in a suburban neighborhood in Michigan. Journal of the American Mosquito Control Association 12: 45-51. Kutz, F.W., and S.C. Strassman. 1977. Human urinary metabolites of organophosphate insecticides following mosquito adulticiding. Mosquito News 37: 211-218. Lanciotti, R.S., G.D. Ebel, V. Deubel, A.J. Kerst, S. Murri, R. Meyer, M. Bowen, N. McKinney, W.E. Morrill, M.B. Crabtree, L.D. Kramer, and J.T. Roehrig. 1999. 110 Origin of the West Nile Virus responsible for and outbreak of encephalitis in the northeastern United States. Science 286: 2333-2337. Landeck-Miller, R.E., J.R. Wands, K.N. Chytalo, and R.A. D'Amico. 2005. Application of water quality modeling technology to investigate the mortality of lobsters (Homarus americanus) in western Long Island Sound during the summer of 1999. Journal of Shellfish Research 24: 859-864. Lawler, S.P., D.A. Dritz, C.S. Johnson, and M. Wolder. 2008. Does synergized pyrethrin applied over wetlands for mosquito control affect Daphnia magna zooplankton or Callibaetis californicus mayflies? Pest Management Science 64: 843-847. Levin, M., B. Brownawell, and S. De Guise. 2007. Sumithrin immunotoxicity in the American lobster (Homarus americanus) upon experimental exposure. Journal of Shellfish Research 26: 1161-1164. Lewis, R.G. (U.S. Environmental Protection Agency). 1999. Compendium method TO10A: Determination of pesticides and polychlorinated biphenyls in ambient air using low volume polyurethane foam (PUF) sampling followed by gas chromatographic/multi-detector detection (GC/MD). Cincinnati, OH. Lofgren, C.S., D.W. Anthony, and G.A. Mount. 1973. Size of aerosol droplets impinging on mosquitoes as determined with a scanning electron microscope. Journal of Economic Entomology 66: 1085-1088. Lothrop, H.D., H.Z. Huang, B.B. Lothrop, S. Gee, D.E. Gomsi, and W.K. Reisen. 2007. Deposition of pyrethrins and piperonyl butoxide following aerial ultra-low volume applications in the Coachella Valley, California. Journal of the American Mosquito Control Association 23: 213-219. Lubin, J.H., J.S. Colt, D. Camann, S. Davis, J.R. Cerhan, R.K. Severson, L. Bernstein, and P. Hartge. 2004. Epidemiologic evaluation of measurement data in the presence of detection limits. Environmental Health Perspectives 112: 1691-1696. Macedo, P.A., R.K.D. Peterson, and R.S. Davis. 2007. Risk assessments for exposure of deployed military personnel to insecticides and personal protective measures used for disease-vector management. Journal of Toxicology and Environmental Health, Part A 70: 1758–1771. Malkinson, M., C. Banet, Y. Weisman, S. Pokamunski, R. King, M.-T. Drouet, and V. Deubel. 2002. Introduction of West Nile virus in the Middle East by migrating white storks. Emerging Infectious Diseases 8: 392-397. Marfin, A.A., and D.J. Gubler. 2001. West Nile encephalitis: an emerging disease in the United States. Clinical Infectious Diseases 33: 1713-1719. 111 Matsumura, F. 1985. Toxicology of insecticides. Plenum Press, New York, NY. Mickle, R.E., O. Samuel, L. St. Laurent, P. Dumas, and G. Rousseau (Institut national de sante publique du Quebec (INSPQ)). 2005. Direct comparison of deposit from aerial and ground ULV applications of malathion with AGDISP predictions. REMSpC Report 2005-02. Milam, C.D., J.L. Farris, and J.D. Wilhide. 2000. Evaluating mosquito control pesticides for effect on target and nontarget organisms. Archives of Environmental Contamination and Toxicology 39: 324-328. Moore, J.C., J.C. Dukes, J.R. Clark, J. Malone, C.F. Hallmon, and P.G. Hester. 1993. Downwind drift and deposition of malathion on human targets from ground ultralow volume mosquito sprays. Journal of the American Mosquito Control Association 9: 138-142. Mount, G.A. 1998. A critical review of ultra low-volume aerosols of insecticide applied with vehicle mounted generators for adult mosquito control. Journal of the American Mosquito Control Association 14: 305-334. Mount, G.A., T.L. Biery, and D.G. Haile. 1996. A review of ultralow-volume aerial sprays of insecticide for mosquito control. Journal of the American Mosquito Control Association 12: 601-618. NRC. 1983. Risk assessment in the Federal government: managing the process. National Research Council, National Academy Press, Washington, D.C. NYCDOH (New York City Department of Health). 2005. Adult mosquito control programs: environmental impact statement (EIS). New York, NY, USA. O'Sullivan, B.C.Y., J. Lafleur, K. Fridal, S. Hormozdi, S. Schwartz, M. Belt, and M. Finkel. 2005. The effect of pesticide spraying on the rate and severity of ED asthma. American Journal of Emergency Medicine 23: 463-467. Oracle. 2007. Crystal Ball® 7.3: user manual. Oracle USA, Inc., Redwood City, CA. Pankiw, T., and S.C. Jay. 1992. Aerially applied ultra-low volume malathion effects on caged honey bees (Hymenoptera: Apidae), caged mosquitoes (Diptera: Culicidae), and malathion residues. Journal of Economic Entomology 85: 687-691. Paul, E.A., and H.A. Simonin. 2006. Toxicity of three mosquito insecticides to crayfish. Bulletin of Environmental Contamination and Toxicology 76: 614-621. Paul, E.A., H.A. Simonin, and T.M. Tomajer. 2005. A comparison of the toxicity of synergized and technical formulations of permethrin, sumithrin, and resmethrin to trout. Archives of Environmental Contamination and Toxicology 48: 251–259. 112 Pearce, J., and N. Balcom. 2005. The 1999 Long Island Sound lobster mortality event: findings of the comprehensive research initiative. Journal of Shellfish Research 24: 691-697. Perry, A.S., I. Yamamoto, I. Ishaaya, and R.Y. Perry 1998. Evaluation of toxicity in insects. Insecticides in agriculture and environment: retrospects and prospects. pp 14-15. Springer-Verlag, Berlin, Germany Peterson, R.K.D. 2006. Comparing ecological risks of pesticides: the utility of a risk quotient ranking approach across refinements of exposure. Pest Management Science 62: 46-56. Peterson, R.K.D., P.A. Macedo, and R.S. Davis. 2006. A human-health risk assessment for West Nile virus and insecticides used in mosquito management. Environmental Health Perspectives 114: 366-372. Peterson, R.K.D., and S.E. Sing. 2007. Releasing biocontrol agents: risk assessment and overdue reform. IOBC-NRS Newsletter 29: 1 and 3. Pettitt, A.N. 1977. Testing the normality of several independent samples using the Anderson-Darling statistic. Applied Statistics 26: 156-161. Pierce, R.H., M.S. Henry, T.C. Blum, and E.M. Mueller. 2005. Aerial and tidal transport of mosquito control pesticides into the Florida Keys National Marine Sanctuary. Revista De Biologia Tropical 53: 117-125. Portier, K., J.K. Tolson, and S.M. Roberts. 2007. Body weight distributions for risk assessment. Risk Analysis 27: 11-26. Pree, D.J., A.B. Stevenson, and E.S. Barszcz. 1996. Toxicity of pyrethroid insecticides to carrot weevils: enhancement by synergists and oils. Journal of Economic Entomology 89: 1254-1261. Rappole, J.H., S.R. Derrickson, and Z. Hubalek. 2000. Migratory birds and spread of West Nile virus in the western hemisphere. Emerging Infectious Diseases 6: 319328. Rappole, J.H., and Z. Hubalek. 2003. Migratory birds and West Nile virus. Journal of Applied Microbiology 94: 47-58. Reisen, W., and A.C. Brault. 2007. West Nile virus in North America: perspectives on epidemiology and intervention. Pest Management Science 63: 641-646. Robertson, J.L., R.M. Russell, H.K. Preisler, and N.E. Savin. 2007. Bioassays with arthropods. 2nd ed. CRC Press, Boca Raton. 113 Roche, J.P. 2002. Print media coverage of risk-risk tradeoffs associated with West Nile encephalitis and pesticide spraying. Journal of Urban Health: Bulletin of the New York Academy of Medicine 79: 482-490. Rose, R.I. 2001. Pesticides and public health: integrated methods of mosquito management. Emerging Infectious Diseases 7: 17-23. Ross, J., T. Thongsinthusak, H.R. Fong, S. Margetich, and R. Krieger. 1990. Measuring potential dermal transfer of surface pesticide residue generated from indoor fogger use: an interm report. Chemosphere 20: 349-360. Schleier III, J.J., R.S. Davis, L.M. Barber, P.A. Macedo, and R.K.D. Peterson. 2008a. A probabilistic risk assessment for deployed military personnel after the implementation of the “Leishmaniasis Control Program” at Tallil Air Base, Iraq. Journal of Medical Entomology in review. Schleier III, J.J., R.S. Davis, L.M. Shama, P.A. Macedo, and R.K.D. Peterson. 2008b. Equine risk assessment for insecticides used in adult mosquito management. Human and Ecological Risk Assessment 14: 392-407. Schleier III, J.J., P.A. Macedo, R.S. Davis, L.M. Shama, and R.K.D. Peterson. 2008c. A two-dimensional probabilistic acute human-health risk assessment of insecticide exposure after adult mosquito management. Stochastic Environmental Research and Risk Assessment (in press): DOI: 10.1007/s00477-00008-00227-00475. Schleier III, J.J., S.E. Sing, and R.K.D. Peterson. 2008d. Regional ecological risk assessment for the introduction of Gambusia affinis (western mosquitofish) into Montana watersheds. Biological Invasions 10: 1277-1287. SETAC (Society for Environmental Toxicology and Chemistry and SETAC Foundation for Environmental Education). 1994. Aquatic Dialogue Group: Pesticide risk assessment and mitigation. Pensacola, FL. 220. Snodgrass, G.L. 1996. Glass-vial bioassay to estimate insecticide resistance in adult tarnished plant bugs (Heteroptera: Miridae). Journal of Economic Entomology 89: 1053-1059. Snodgrass, G.L., J.J. Adamczyk, and J. Gore. 2005. Toxicity of insecticides in a glassvial bioassay to adult brown, green, and southern green stink bugs (Heteroptera : Pentatomidae). Journal of Economic Entomology 98: 177-181. Stevens, M.M., A. Ali, S. Helliwell, L.J. Schiller, and S. Hansen. 2002. Comparison of two bioassay techniques for assessing the acute toxicity of pesticides to chironomid larvae (Diptera : Chironomidae). Journal of the American Mosquito Control Association 18: 119-125. 114 Suffolk-County (Suffolk County Department of Public Works and Department of Health Services). 2006. Draft generic environmental impact statement. Date Accessed: 1 March. <http://www.suffolkmosquitocontrolplan.org/tasks/task14.html>. Suffolk County (Suffolk County Department of Public Works and Department of Health Services). 2006. Draft generic environmental impact statement. Suffolk County, NY, USA. Teske, M.E., S.L. Bird, D.M. Esterly, T.B. Curbishley, S.L. Ray, and S.G. Perry. 2002. AgDRIFT®: A model for estimating near-field spray drift from aerial applications. Environmental Toxicology and Chemistry 21: 659-671. Thier, A. 2001. Balancing the risks: vector control and pesticide use in response to emerging illness. Journal of Urban Health: Bulletin of the New York Academy of Medicine 78: 372-381. Tietze, N.S., P.G. Hester, and K.R. Shaffer. 1994. Mass recovery of malathion in simulated open field mosquito adulticide tests. Archives of Environmental Contamination and Toxicology 26: 473-477. Tietze, N.S., P.G. Hester, K.R. Shaffer, and F.T. Wakefield. 1996a. Peridomestic deposition of ultra-low volume malathion applied as a mosquito adulticide. Bulletin of Environmental Contamination and Toxicology 56: 210-218. Tietze, N.S., K.R. Shaffer, and P.G. Hester. 1996b. Half-life of naled under three test scenarios. Journal of the American Mosquito Control Association 12: 251-254. Townzen, K.R., and H.L. Natvig. 1973. A disposable adult mosquito bioassay cage. Mosquito News 33: 113-114. Tucker, J.W., C.Q. Thompson, T.C. Wang, and R.A. Lenahan. 1987. Toxicity of organophosphorous insecticides to estuarine copepods and young fish after field applications. Journal of the Florida Anti-Mosquito Association 58: 1-6. USEPA (U.S. Environmental Protection Agency). 1995. ISCST3. Date Accessed: <http://www.epa.gov/scram001/tt22.htm#screen>. USEPA (U.S. Environmental Protection Agency). 1996a. Method 3510C: Separatory funnel liquid-liquid extraction. Washington, DC. 1-8. USEPA (U.S. Environmental Protection Agency). 1996b. Method 3550B: Ultrasonic extraction. Washington, DC. 1-14. USEPA (U.S. Environmental Protection Agency). 1996c. Method 8270C: Semivolatile organic compounds by gas chromatography/mass spectrometry (GC/MS). Washington, DC. 1-54. 115 USEPA (U.S. Environmental Protection Agency). 1997a. Draft: Standard operating procedures (SOPs) for residential exposure assessments. Date Accessed: Nov. 14, 2007. <http://www.epa.gov/pesticides/trac/science/trac6a05.pdf>. USEPA (U.S. Environmental Protection Agency). 1997b. Exposure Factors Handbook. Vol I. General factors. EPA/600/P-95/002Fa. Washington, DC. USEPA (U.S. Environmental Protection Agency). 1997c. Exposure factors handbook. Vol II. Food ingestion factors. EPA/600/P-95/002Fb. Washington, DC. USEPA (United States Environmental Protection Agency). 1998. Guidelines for Ecological Risk Assessment. EPA/630/R-95/002F. Washington D.C., USA. 1116. USEPA (U.S. Environmental Protection Agency). 1999. Memorandum from S. Hummel, Health Effects Division to T. Myers, Special Review and Reregistration Division and K. Monk Special Review and Reregistration Division. Naled. Revised HED risk assessment for RED. PC Code 034401. DP Barcode D260129. Washington D.C. . 1-100. USEPA (U.S. Environmental Protection Agency). 2002a. Child-Specific Exposure Factors Handbook (Interim Report). EPA-600-P-00-002B. Washington, D.C. 1448. USEPA (U.S. Environmental Protection Agency). 2002b. Interim reregistration eligibility decision for naled. Case No. 0092. Washington D.C. 1-148. USEPA (U.S. Environmental Protection Agency). 2005a. Memorandum from M. Rexrode and J.L. Meléndez, Environmental Fate and Effects Division, to M.T. Shamim, Environmental Fate and Effects Division. Re: EFED Revised Risk Assessment for the Reregistration Eligibility Decision on Permethrin After Error Corrections Comments from the Registrant, Phase I. PC Code 109701, DP Barcode D313975. Washington D.C. USEPA (U.S. Environmental Protection Agency). 2005b. Memorandum from S.L. Kinard, Health Effects Division, to T. Moriarty, Special Review and Reregistration Division. Malathion: updated revised human health risk assessment for the reregistration eligibility decision document (RED). PC Code: 057701, DP Barcode: D321543. Washington D.C. USEPA (U.S. Environmental Protection Agency). 2005c. Memorandum from S.L. Kinard, Y. Yang and S. Ary, Health Effects Division, to J. Guerry, Special Review and Reregistration Division. Re: Permethrin. HED Chapter of the Reregistration Eligibility Decision Document (RED). . PC Code 109701, Case No. 52645-53-1, DP Barcode D319234. Washington D.C. 116 USEPA (U.S. Environmental Protection Agency). 2006a. Interim Reregistration Eligibility Decision for Naled Case No. 0092. Washington D.C. 1-130. USEPA (U.S. Environmental Protection Agency). 2006b. Reregistration eligibility decision (RED) for permethrin. Case No. 2510. Washington D.C. 1-95. USEPA (U.S. Environmental Protection Agency). 2006c. Reregistration eligibility decision for piperonyl butoxide (PBO). Case No. 2525. Washington D.C. 1-111. USEPA (U.S. Environmental Protection Agency). 2006d. Reregistration eligibility decision for pyrethrins. Case No. 2580. Washington D.C. 1-108. USEPA (United States Environmental Protection Agency). 2006e. Technical overview of ecological risk assessment. Date Accessed: 1 November. <http://www.epa.gov/oppefed1/ecorisk_ders/toera_risk.htm>. USEPA (U.S. Environmental Protection Agency). 2008. Memorandum from B. Daiss, Reregistration Branch 4, to J. Howenstine, Reregistration Branch 1 Special Review and Registration Division, and A. Sibold, Insecticide Branch Registration Division. Re: d-Phenothrin (Sumithrin®) Reregistration Eligibility Decision (RED) Document PC Code No. 069005; DP Barcode No. 326933. Washington D.C. Waller, G.D., B.J. Estesen, N.A. Buck, K.S. Taylor, and L.A. Crowder. 1988. Residual life and toxicity to honey bees (Hymenoptera: Apidae) of selected pyrethroid formulations applied to cotton in Arizona. Journal of Economic Entomology 81: 1022-1026. Weidhaas, D.E., M.C. Bowman, G.A. Mount, C.S. Lofgren, and H.R. Ford. 1970. Relationship of minimum lethal dose to optimum size of droplets of insecticides for mosquito control. Mosquito News 30: 195-200. Weston, D.P., E.L. Amweg, A. Mekebri, R.S. Ogle, and M.J. Lydy. 2006. Aquatic effects of aerial spraying for mosquito control over an urban area. Environmental Science and Technology 40: 5817-5822. Weston, D.P., R.W. Holmes, J. You, and M.J. Lydy. 2005. Aquatic toxicity due to residential use of pyrethroid insecticides. Environmental Science and Technology 39: 9778-9784. Wheeler, M.W., R.M. Park, and A.J. Bailer. 2006. Comparing median lethal concentration values using confidence interval overlap or ratio tests. Environmental Toxicology and Chemistry 25: 1441-1444. WHO (World Health Organization). 1998. Malaria in Africa. Date Accessed: 17 November 2006. 117 Williams, D.A. 1982. Extra-binomial variation in logistic linear models. Applied Statistics 31: 144-148. Womeldorf, D.J., E.L. Atkins, and P.A. Gillies. 1974. Honey bee hazards associated with some mosquito abatement aerial spray applications. California Vector Views 21: 51-55. Xu, Q., H.Q. Liu, L. Zhang, and N.N. Liu. 2005. Resistance in the mosquito, Culex quinquefasciatus, and possible mechanisms for resistance. Pest Management Science 61: 1096-1102. Xue, J., V. Zartarian, J. Moya, N. Freeman, P. Beamer, K. Black, N. Tulve, and S. Shalat. 2007. A meta-analysis of children’s hand-to-mouth frequency data for estimating nondietary ingestion exposure. Risk Analysis 27: 411-420. Zhong, H.E., M. Latham, P.G. Hester, R.L. Frommer, and C. Brock. 2003. Impact of naled on honey bee survival and productivity: aerial ULV application using a flat fan nozzle system. Archives of Environmental Contamination and Toxicology 45: 216-220. Zulkosky, A.M., J.P. Ruggieri, S.A. Terracciano, B.J. Brownawell, and A.E. McElroy. 2005. Acute toxicity of resmethrin, malathion and methoprene to larval and juvenile American lobsters (Homarus americanus) and analysis of pesticide levels in surface waters after ScourgeTM, AnvilTM and AltosidTM application. Journal of Shellfish Research 24: 795-804.