RESPONSE OF Aedes albopictus (DIPTERA: CULICIDAE) TO TRAPS, ATTRACTANTS, AND ADULTICIDE S IN NORTH CENTRAL FLORIDA By DAVID FRANKLIN HOEL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005
Copyright 2005 by David Franklin Hoel
To my wife, Joyce; my son, Michael; my daughter, Caroline; and my mother
iv ACKNOWLEDGMENTS I greatly appreciate the Naval Medica l and Education and Training Commands Duty Under Instruction program, for giving me this unique opportunity to pursue a Doctor of Philosophy degree in entomology. This has been my most rewarding educational experience yet. Th e staff at the United States Department of Agriculture Animal Research Service (USDA ARS), and the Department of Entomology and Nematology provided outstanding support in bo th personnel and material. They provided for all my research needs and made feel welcome and a part of their team. My supervisory committee is an exceptional group of scientists. Dr. Daniel Kline helped me immensely and guided me soundl y through all my research. He provided separate office space, vehicles, materials, lab use, and guidance for all areas of my research as needed and sometimes on very s hort notice. I cant thank him enough for all he has done. In addition, he always made time to talk about anything I brought to him regardless of subject. He made me feel li ke a member of his lab. I have the same feelings for Dr. Jerry Butler who served as a mentor, teacher, and friend during my stay at the University of Florida. In addition, I thank him for his sacr ifice of joining my committee even as his retirement was approach ing and better things lay ahead for him than laboring over one more graduate student. He and his wife Mar ilyn have been fun to visit with and always friendly. I gratefully acknowledge Dr. Sandy Allans support in all things pertaining to the USDA ARS facilities that I used un der her supervision. She was a tremendous help and very forgiving when my little subjects escaped and tormented both
v her and the work staff in the mosquito-reari ng facility. Many thanks go to Dr. Steve Valles and Dr. Jack Petersen, both toxicologist s, for their guidance and the use of their equipment in my resistance studies. Many others in the Department of En tomology and Nematology and at the USDA ARS deserve special mention. I thank my gra duate advisor (Dr. Don Hall) for letting me use his property for my studies and for his overall friendliness and kindness to me while I was there. Dr. Grover Smart, who preceded Dr Hall, was also kind and helpful. I extend sincere thanks to one of the most helpful and best administrators Ive ever met: Mrs. Debbie Hall, graduate staff of the Ento mology and Nematology Department at the University of Florida (UF). She helped me quickly through my administrative headaches and was also a good friend. Two of my prof essors deserve special thanks here: Dr. Pauline Lawrence and Dr. Simon Yu taught ex cellent classes, always had time for questions and visits, and gui ded me through the difficult subjects of insect physiology and toxicology, respectively. Dr. Gene Gerber g took the time to befriend me, share his rich knowledge and stories of Army ento mology, and introduced me to many of his professional associates in the vect or and pest control industries. I am indebted to a number of people at USDA ARS Gainesville. Dr. Klines laboratory crew (Joyce Urban and Aaron Lloyd) were a trem endous help and among my best friends while I was in Gain esville. They helped me in most aspects of my research, providing support with material, large outdoor cage use, and administrative functions. Dr. Uli Bernier provided lab space for my resistance studies and listened patiently and sympathetically as I whined about Gator foot ball losses to Florid a State University.
vi Thanks go to Dr. Jerry Hogsette for the use of his property in my research, and to Genie White for help with the SAS program. Mosquito control collaborators for my susceptibility study included Ms. Marah Clark of the City of Jacks onville; Mr. Pat Morgan of I ndian River Mosquito Control District; Mr. Billy Kelner, Citrus County Mos quito Control District; Ms. Jodi Avila, UF graduate student working in Quincy; and Bill Johnson and Julie Player of Escambia County Mosquito and Rodent Management Divi sion. I give heartfel t thanks to all of them for their help. I was able to return to gr aduate school partly because of the encouragement and support of Commander Michael O. Mann and Captain Jim Need, bot h excellent Navy entomologists who are now retired (and Florida Gators too!!). I owe them both a special debt of gratitude for making this opportunity possible, but for helping me toward my career as a Navy entomologist, and for being 2 of the best Commanding Officers Ive had since Ive been in the Navy. Special thanks are in order for Dr. J im Olson of Texas A&M University who started me along the path of medical entomol ogy and has been my most important mentor for the last 20 years. May God bless hi m for his patience, friendship, and support. Looking back over it all, I think he was the be st of the best and I will always remember all that he did for me. My parents, Patricia and Fr ank Hoel, always encouraged me to work hard and to excel in my educational endeavors. I love them both and think of them everyday. Their investment in time and love pa id big dividends in my life.
vii Most of all, I thank my wife Joyce for her never-ending support and love for me during this very busy and trying period of my life. She has been a wonderful mother to our 2 angels, Michael and Caroline, and kept our lives sane and in order while I was away from home with my work and studies.
viii TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................xii LIST OF FIGURES.........................................................................................................xiv ABSTRACT.....................................................................................................................xv i CHAPTER 1 LITERATURE REVIEW OF MOSQUITO TRAPS, ATTRACTANTS, AND ADULTICIDES USED TO CONTROL Aedes albopictus ..........................................1 Introduction to Aedes albopictus ..................................................................................1 Ecology of Aedes albopictus .................................................................................2 Distribution............................................................................................................3 Significance of Aedes albopictus in Florida..........................................................5 Medical Significance.............................................................................................6 Literature Review of Mosquito Attractants..................................................................9 Classification of Mo squito Attractants..................................................................9 Host-Seeking Activity of Mosquitoes.................................................................10 Visual Attractants of Mosquitoes........................................................................11 Chemical Attractants of Mosquitoes...................................................................18 Physical Attractants of Mosquitoes.....................................................................21 Introduction to Surveillance and Resi dential Traps Used for Mosquito Surveillance and Control.........................................................................................24 Carbon Dioxide-Supplemented Traps.................................................................25 Carbon Dioxide-Generating (Propane) Traps.....................................................30 Introduction to Aedes albopictus Adulticide Sus ceptibility Review..........................32 Research Objectives....................................................................................................34 2 Aedes albopictus RESPONSE TO ADULT MOSQUITO TRAPS IN LARGECAGE TRIALS..........................................................................................................36 Introduction.................................................................................................................36 Materials and Methods...............................................................................................39 Large Outdoor-Screened Cages...........................................................................39 Mosquitoes..........................................................................................................40
ix Description of Traps Tested................................................................................41 Surveillance Traps...............................................................................................42 Residential Traps.................................................................................................43 Statistical Analysis..............................................................................................47 Results........................................................................................................................ .47 Discussion...................................................................................................................49 3 FIELD EVALUATION OF CARBON DIOXIDE, 1-OCTEN-3-OL, AND LACTIC ACID-BAITED MOSQUITO MAGNET PRO TRAPS AS ATTRACTANTS FOR Aedes albopictus IN NORTH CENTRAL FLORIDA.........62 Introduction.................................................................................................................62 Materials and Methods...............................................................................................64 Trap Placement and Rotation..............................................................................64 Attractants............................................................................................................65 Statistical Analysis..............................................................................................66 Results........................................................................................................................ .67 Aedes albopictus ..................................................................................................67 Other Mosquito Species......................................................................................68 Discussion...................................................................................................................69 Aedes albopictus ..................................................................................................69 Other Mosquito Species......................................................................................75 4 RESPONSE OF Aedes albopictus TO SIX TRAPS IN SUBURBAN SETTINGS IN NORTH CENTRAL FLORIDA............................................................................83 Introduction.................................................................................................................83 Materials and Methods...............................................................................................84 Site Selection and Trapping Scheme...................................................................84 Traps....................................................................................................................85 Statistical Analysis..............................................................................................86 Results........................................................................................................................ .87 Aedes albopictus ..................................................................................................87 Other Mosquito Species......................................................................................88 Discussion...................................................................................................................91 Aedes albopictus ..................................................................................................91 Other Mosquito Species......................................................................................95 5 SUSCEPTIBILITY OF Aedes albopictus TO FIVE COMMONLY USED ADULTICIDES IN FLORIDA................................................................................107 Introduction...............................................................................................................107 Materials and Methods.............................................................................................111 Collection Sites..................................................................................................111 Egg Collection Apparatus..................................................................................112 Adult Rearing....................................................................................................112 Insecticide..........................................................................................................113
x Bioassay Test Procedure....................................................................................114 Analysis of Data................................................................................................117 Results.......................................................................................................................1 18 Discussion.................................................................................................................121 6 LABORATORY RESPONSE OF Aedes albopictus TO LIGHT EMITTING DIODES OF EIGHT DIFFERENT COLO RS AND ULTRAVIOLET LIGHT OF EIGHT DIFFERENT FLICKER FREQUENCIES..................................................144 Introduction...............................................................................................................144 Materials and Methods.............................................................................................147 Visualometer......................................................................................................147 Light Emitting Diodes.......................................................................................148 Artificial Host....................................................................................................149 Mosquitoes........................................................................................................150 Flicker Response Trials.....................................................................................150 Statistical Analysis............................................................................................151 Results.......................................................................................................................1 51 Aedes albopictus Response to Light of Different Color....................................151 Aedes albopictus Response to Flickering Ligh t of Different Frequencies........152 Discussion.................................................................................................................153 Colored Light Preference..................................................................................153 Flickering Light Preference...............................................................................156 7 EVALUTION OF LIGHTAND MO TOR-MODIFIED CENTERS FOR DISEASE CONTROL TRAPS FOR WO ODLAND MOSQUITOES IN NORTH CENTRAL FLORIDA..............................................................................................164 Introduction...............................................................................................................164 Materials and Methods.............................................................................................166 Trap Rotation and Collection............................................................................166 Trap Modification..............................................................................................167 Trial Location....................................................................................................168 Statistical Analysis............................................................................................169 Results.......................................................................................................................1 69 Discussion.................................................................................................................172 Species...............................................................................................................181 8 SURVEILLANCE AND CONTROL OF Aedes albopictus : THE IMPORTANCE OF TRAPS, ATTRACTANTS AND ADULTICIDES............................................183 Introduction...............................................................................................................183 Traps, Trapping, and Attractants..............................................................................184 Pesticide Response....................................................................................................188
xi APPENDIX A LARGE-CAGE Aedes albopictus CAPTURE RESULTS WITH RESIDENTIAL AND SURVEILLANCE MOSQUITO TRAPS.......................................................193 B Aedes albopictus CAPTURE TOTALS IN CDC LIGHT TRAPS AT SIX SITES IN GAINESVILLE, FLORIDA................................................................................196 C PESTICIDE DILUTIONS FOR SUSCEPTIBILITY STUDY................................198 D CIRCUIT DESCRIPTION OF 555 FREQUENCY GENERATORS......................204 E CAPACITANCE IN MICRO FARADS OF TEN DIFFERENT FREQUENCY GENERATING 555 INTEGRATED CIRCUITS....................................................205 LIST OF REFERENCES.................................................................................................206 BIOGRAPHICAL SKETCH...........................................................................................223
xii LIST OF TABLES Table page 2-1 Trap attractant features used in Aedes albopictus large-cage trials..........................57 3-1 Totals, means, and SEM of Aedes albopictus collected from Mosquito Magnet Pro traps over 3 identical trials with 4 treatments....................................................79 3-2 Sex ratios of Aedes albopictus collected from Mosquito Magnet Pro traps over 3 identical trials with 4 treatments. n = 12 periods (48 h)..........................................79 3-3 Treatment sex ratios of Aedes albopictus over 3 trials and 4 treatments with the Mosquito Magnet Pro. n = 12 periods (48 h)..........................................................79 3-4 Mosquito Magnet Pro trap counts pe r attractant treatment (means SEM)............80 3-5 Adult totals of the 5 most abundant mo squito species collected from Mosquito Magnet Pro traps with 4 treatmen ts. n = 12 periods (48 h).....................................81 4-1 Trap features and chemical attractants used in comparison trials with residential and surveillance traps in Gainesville, Florida........................................................102 4-2 Total adult Aedes albopictus caught in 6 traps over 3 trials in suburban neighborhoods in Gainesville, Florida ove r 36 days (n = 18 periods of 48 h).......102 4-3 Sex ratios of Aedes albopictus caught in 6 traps over 3 trials in suburban neighborhoods in Gainesville, Florida ove r 36 days (n = 18 periods of 48 h).......102 4-4 Adult mosquito count per trap................................................................................103 4-5 Trap performance ranking of the most commonly occurring mosquito species in residential settings in Gainesville, Florida.............................................................104 5-1 Baseline insecticide susceptibility bi oassay results for adult females of a colonized USDA ARS strain of Aedes albopictus n = 150..................................134 5-2 Insecticide susceptibility results fo r Inverness, Citrus County, Florida and USDA ARS colony populations of adult female Aedes albopictus .......................135 5-3 Insecticide susceptibility results for Quincy, Gadsden County, Florida and USDA ARS colony populations of adult female Aedes albopictus .......................136
xiii 5-4 Insecticide susceptibility results for Vero Beach, Indian River County, Florida and USDA ARS colony populations of adult female Aedes albopictus .................137 5-5 Insecticide susceptibility results fo r Pensacola, Escambia County, Florida and USDA ARS colony populations of adult female Aedes albopictus .......................138 5-6 Insecticide susceptibility results for Jacksonville, Duva l County, Florida and USDA ARS colony populations of adult female Aedes albopictus .......................139 5-7 Insecticide susceptibility results fo r Gainesville, Alachua County, Florida and USDA ARS colony populations of adult female Aedes albopictus .......................140 6-1 Average number of b ite-sec for 8 h exposure of Ae. albopictus to artificial host illuminated by light of different colors..................................................................159 6-2 Average number of b ite-sec for 8 h exposure of Ae. albopictus to artificial host illuminated by flickering light of different frequencies.........................................159 7-1 Power consumption of standard and m odified CDC light traps with effective operating days produced from 6 V, 12Ah rechargeable gel cell batteries............178 7-2 Trial 1 results of modi fied light and motor CDC li ght trap counts with 500 mL/min CO2 (means SEM) at the Horse Teaching Unit.....................................179 7-3 Trial 2 results of modi fied light and motor CDC li ght trap counts with 500 mL/min CO2 (means SEM) at the Horse Teaching Unit.....................................180 7-4 Trial 3 results of modi fied light and motor CDC li ght trap counts with 500 mL/min CO2 (means SEM) at Austin Cary Memorial Forest.............................181 A-1 Trial counts, means, and treatments (trap type) of Ae. albopictus in large-cage trials at USDA ARS Gainesville, Florida...............................................................193 B-1 Gainesville Ae. albopictus counts from 6 light traps in Gainesville, Florida.........196 E-1. Capacitance of 10 different fr equency-generating capacitors..................................205
xiv LIST OF FIGURES Figure page 2-1 Large outdoor screened cages used in trap efficacy trials, USDA ARS Gainesville................................................................................................................58 2-2 Traps tested in large-cage efficacy trials with Aedes albopictus ..............................59 2-3 Large-cage Aedes albopictus trap capture means in re sidential and surveillance traps.......................................................................................................................... 60 2-4 Large-cage trap capture and biting mean s (total catch or bites/number of trials) of Ae. albopictus .......................................................................................................61 3-1 Mosquito Magnet Pro used in suburban trials to collect adult Aedes albopictus .....81 3-2 Capture totals by treatment over 3 trials with the Mosquito Magnet Pro trap for the most common mosquitoes collected fr om 4 suburban sites in Gainesville, Florida......................................................................................................................82 4-1 Relative percent trap capture of th e 9 most commonly occurring mosquito species in suburban neighborhoods in Gainesville, Florida...................................105 5-1 Aedes albopictus egg collection sites, nor th and central Florida...........................141 5-2 Aedes albopictus egg collection apparatus.............................................................142 5-3 Partitioned box holding insecticidecoated 20 mL scintillation vials....................143 6-1. Visualometer used in color preference tests............................................................160 6-2 Diagram of a 555 integrated circuit frequency generator.......................................161 6-3 Duration of feeding (sec) over an 8 h period (mean SQRT ( n + 1) SEM) for Aedes albopictus on artificial host i lluminated with light of different colors........162 6-4 Duration of feeding (sec) over an 8 h period (mean SQRT ( n + 1) SEM) for Aedes albopictus on artificial host i lluminated with ultraviolet (380 nm) light emitting diodes of different frequencies.................................................................163 7-1 Light emitting diode-modi fied CDC light traps.....................................................182
xv 7-2 Wiring schematic of light emitti ng diode-modified CDC light traps.....................182 B-1 August and September 2003 Aedes albopictus trap totals for each of 6 CDC light traps set in residential neighbor hoods in Gainesville, Florida...............................197
xvi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy RESPONSE OF Aedes albopictus (DIPTERA: CULICIDAE) TO TRAPS, ATTRACTANTS, AND ADULTICIDE S IN NORTH CENTRAL FLORIDA By David Franklin Hoel August 2005 Chair: Daniel L. Kline Co chair: Jerry F. Butler Major Department: Entomology and Nematology We examined the response of Aedes albopictus (Skuse) to traps, attractants and adulticides in North Central Flor ida. Residential traps performe d as well as or better than standard surveillance traps in large-cage trials. Mosquito Magnet (MM) Pro, MM Liberty, MM-X, and Fay-Prince traps were the most effectiv e. Two of 3 octenol-baited traps caught more Ae. albopictus than similar unbaited traps in large-cage trials. Capture rates of the Wilton trap were signifi cantly improved by decreasing recommended operational height, from 3 ft to 20 in. Counterflow geometry traps (MM Pro, MM Liberty, MM-X) were highly eff ective against this species ( outperforming adhesive traps, sound traps and light traps). In field trials octenol baits were slightly repellent to Ae. albopictus compared with unbaited (cont rol) traps. Lactic acid bait proved to be more attractive than control or oc tenol-baited traps. Octenol + lactic acid combined was superior to all other treatments and perfor med significantly better than octenol alone. Other mosquitoes including Culex Ochlerotatus and Psorophora species responded to
xvii octenol-baited or control traps more favorably th an to lactic acid or octenol + lactic acid baited traps. In suburban field trials, re sidential counterfl ow geometry traps optimally baited with octenol + lactic acid collected significantly more Ae. albopictus than did surveillance traps including CDC light traps and traps designed specifically to capture Aedes mosquitoes (Fay-Prince and Wilton tr aps). Laboratory tests showed that ultraviolet light was more attractive to Ae. albopictus than white, violet, blue, green, orange, red, infrared, or no light. No preference was obs erved for ultraviolet light flickered at 8 different fr equencies (10-, 30-, 40-, 60-, 120-, 150-, 200and 500 Hz). Six field populations of Ae. albopictus from central and north Florida were susceptible to 5 adulticides (malathion and naled, organophosphates, and resmethrin, d-phenothrin, and permethrin, pyr ethroids) commonly used by vector control agencies. Permethrin proved to be most toxic to this species. Light traps were modified with small mo tors and blue light emitting diodes to conserve battery power and extend operational use by 3 x or 4 x depending on motor/light combination. No preferen ce was observed for traps equipped with incandescent or blue light for 18 species analyzed. More mosquitoes were collected from standard motor than small motor traps, but differences were not significant. Species composition remained fairly constant between all traps.
1 CHAPTER 1 LITERATURE REVIEW OF MOSQUITO TRAPS, ATTRACTANTS, AND ADULTICIDES USED TO CONTROL Aedes albopictus Introduction to Aedes albopictus Aedes albopictus (Skuse) is currently one of the most common and troublesome suburban mosquitoes, occurring from the Atla ntic seaboard throughout the south central and Midwest United States. This has only recent ly become the case, as this exotic pest was absent from North America until the mi d 1980s. Its incredibly rapid advance throughout much of the eastern half of the Un ited States is well documented and it is now firmly established in 25 states (Moore 1999). Several concerns are associated with the establishment of Ae. albopictus in the United States among mosquito control and public health agencies: first, it has become one of the most common nuisance mosquito es occurring in urban settings and is especially associated with household envir onments; second, it is extremely difficult to control using standard mosqu ito control practices; and third, it is a known vector of several arthropod-borne (arbovira l) diseases, some that occu r in the United States and some that could soon appear as a result of accidental or intentional introduction. A review of the ecology of this mosquito is wa rranted by its recent establishment in North America, its colonization of such a large ge ographical area of the United States, and the problems associated with its presence.
2 Ecology of Aedes albopictus Aedes albopictus was first described by Skuse (1896) in Bombay, India as Culex albopictus It is a member of the subgenus Stegomyia group Scutellaris characterized as small black mosquitoes with white or silver scales on the legs, thorax, and head. These mosquitoes breed readily in natural and ar tificial containers but not in ground pools (Watson 1967). Eggs are laid singularly a nd are spaced evenly about the substrate on which they are cemented (Estrada-Franco and Craig 1995). They are laid just above the water line in situations where water is present or in the humid recesses of flood-prone natural and artificial contai ners. Blood-engorged females are capable of producing about 40 to 90 eggs per blood meal during the first gonotrophic cycle and fewer eggs after later blood meals (Gubler 1970). Larvae are comm only found in tree holes, rock pools, and water-holding plants such as bromeliads, ba mboo stumps, and coconut shells and husks. Artificial containers often used for breeding include discarde d tires, clogged rain gutters, water-collection barrels, cisterns, tin cans, birdbaths, a nd almost any other type of manmade product capable of holding rainwater. Aedes albopictus is primarily a daytime biter (E strada-Franco and Craig, 1995). Its diurnal biting behavior is us ually bimodal with peak activ ity occurring in mid-morning and late afternoon hours (Ho et al. 1973). A lthough similar in app earance and ecology to the yellow fever mosquito, Aedes aegypti (L.), it is not as st rongly anthropophagic and has been described as opportunistic biter that feeds on a wide range of mammals and birds (Savage et al. 1993, Nie bylski et al. 1994). Its prop ensity to feed on humans coupled with the ability to v ector certain arboviruses adds the qualifier dis ease vector in addition to nuisance mosqu ito. An aggressive biter, Ae. albopictus is usually one of the first mosquito species to attempt to feed when present in the field (personal
3 observations of author from Florida, North Carolina, Texas and Hawaii). It typically bites on the lower extremities with the lower le gs and ankles being favored sites (Watson 1967), however, it will also readily bite about the head, neck, and arms when convenient (Shirai et al. 2002). Aedes albopictus is stealthy, shying away from the front of its target in preference for the hind or blind side. Its silent flight and pa inless bite enhance its ability to feed and depart before being noticed. Aedes albopictus tends to avoid direct sunlight and thus is often associated with field-forest fringe ar eas in rural environments (Hawle y 1988). Adult flight range is limited (rarely more than 200 meters from site of emergence) and is often near the ground (Bonnet and Worchester 1946). Adults are no t seen flying in strong winds (personal observation). Distribution Aedes albopictus is believed to have originated in the tropical forests of Southeast Asia. It commonly occurs in Vietnam, Th ailand, Japan, China, Korea, and many Pacific and Indian Ocean islands (Hawley 1988). It is well suited to both tropical and temperate climates, ranging to 36o N in Japan and 42o N in North America (Estrada-Franco and Craig 1995). Because it often ovi posits in artificial c ontainers and natural sites, its range has expanded dramatically worl dwide since the end of World War II. Global shipping of tires, especially from Asian nations to the rest of the world, is credited for later establishment of Ae. albopictus in parts of Africa; Europe; and North, Central, and South America (Reiter 1998). It was recently intercep ted in Darwin, Australia, but is not yet known to be established there (Lamche and Whelan 2003). Aedes albopictus was introduced into Hawaii sometime before 1902 (Usinger 1944). By 1902 it was reportedly very numerous and conspicuous (Perkins 1913). It
4 rapidly established itself th roughout the island chain and is now abundant on the islands of Oahu, Kauai, Maui, Molokai and Hawaii (K enneth Hall, Hawaii State Department of Health, Vector Control Branch, Director, personal communication). The first established population of Ae. albopictus discovered in the continental United States was in Houston, Texas in 1985 (Sprenger and Wuithiranyagool 1986) although a single specimen was cau ght in a light trap 2 years earlier at a tire dump in Memphis, Tennessee (Reiter and Darsie 1984). Shipments of military supplies (tires) from Asia were responsible for 2 earlier introductions of Ae. albopictus into the United States, but apparently it did not become established on either occasion (Eads 1972, Pratt et al. 1946). Since its introduction to the continental United States, Ae. albopictus spread rapidly across much the southeastern and central portions of the country. In 1999 it was reportedly established in 911 counties in 25 states, although the Centers for Disease Control and Prevention (CDC) has reports of its presence from 919 counties in 26 states (Moore 1999). Apparently, Ae. albopictus was unable to remain established in several of the more northern counties in whic h it was found. As of December 2004, Ae albopictus had expanded its range to 1,035 counties in 32 states (Janet McAllister, CDC, personal communication). Georgia was the first state to report Ae. albopictus established in every county; Tennessee, North Carolina, South Carolina, De laware, and Florida have since followed. Aedes albopictus is established as far north as New Jersey, continuing westward to Chicago, Illinois. An average isotherm of -5oC appears to limit northern expansion (Rodhain and Rosen 1997). Its range extends so uthward along the Atlantic seaboard into
5 the Florida Keys, and westward along the Gulf Coast into Texas. Westward expansion is limited to west Texas up through eastern Nebraska Thus it occurs in all of the south and much of the Midwest United States. At the ti me of this writing, it has been discovered in Orange County, California and ther e is concern that it might already be established there (Linthicum et al. 2003). It only took Ae. albopictus 8 years to colonize all 67 Florida counties. The first reported infestation was discovered in 1986 at a tire repository in Jacksonville, Duval County (Peacock et al. 1988). By 1992, it was reported from southern Lee County (Hornby and Miller 1994) and soon after fr om all counties (OMeara et al. 1995a). Significance of Aedes albopictus in Florida The establishment of Ae. albopictus in Florida is important to vector control and public health officials for several reasons. First, Ae. albopictus has displaced Ae. aegypti in many places where it has become establishe d in the continental United States and Hawaii. Aedes aegypti the yellow fever mosquito, is the most important worldwide vector of urban dengue fever, a rapi dly emerging disease of the tropics ( www.cdc.gov/ncidod/dvbid/dengue/index.htm ). This is mostly due to its synantrophic lifestyle and anthropophagic feed ing preference (as compared to Ae. albopictus ). The implication associated with this replacement is that the severity of a dengue outbreak in Florida or other states where Ae. albopictus is present could be somewhat lessened should this disease soon emerge here, as Ae. albopictus is seemingly of lesser importance in dengue transmission than Ae. aegypti (Gubler 1997). Second, Ae. albopictus is now probably the most abundant nui sance mosquito associated with household dwellings in the southeast United States (M oore and Mitchell 1997). It ha s a propensity to breed and feed in close association with humans and can quickly build to high numbers in suburban
6 settings. Third, unlike most mos quitoes, it is a day biter. Th is makes control of adults by conventional methods very difficult. Typical ly, vector control pe rsonnel concentrate the bulk of their control efforts on killing adult mosquitoes. This is accomplished by using ultra low volume (ULV) pesticide applicati on in which pesticide is sheared by a highpressure air stream as it exits the nozzle of the ULV machine. Control is provided by 5 to 20 micron spherical particles of pesticide created at the no zzle head drifting through the air to impinge on flying adults. This is th e easiest, quickest, and most common type of adult mosquito control used in the United States today. Ultra low volume pesticide application is most effectiv e at night when atmospheric conditions favor the lateral dispersion of pesticide drift, the inversion condition. Unfortunately, ULV mosquito control is not very effective during day light hours because of lapse atmospheric conditions in which air moves vertically as a result of solar heating of the earth (Sutton 1953). Because of this atmospheric phe nomenon, ULV adulticiding is typically performed at night. As a result of its diurnal flight and biting habits, Ae. albopictus is more difficult to control using ULV technol ogy than most other commonly occurring mosquitoes that are active at night. Medical Significance Aedes albopictus has been shown to be naturally or experimentally infected with at least 23 arboviruses (Moore and Mitchell 1997). The most important disease agent vectored is dengue fever. In the United States, field-caught Ae. albopictus have been found naturally infected with 4 arboviruses of public health concern: Cache Valley virus (Mitchell et al. 1998), eastern equine ence phalitis (Mitchell et al. 1992, CDC 2001), West Nile virus (Holick et al. 2002), and La Crosse encephalitis virus (Gerhardt et al. 2001).
7 Although Ae. albopictus is considered a secondary vector of dengue to Ae. aegypti it nevertheless has been incriminated in several dengue epidemics where Ae. aegypti was lacking at outbreak foci in Hawaii (K enneth Hall, personal communication), the Seychelles Islands (Calisher et al. 1981), Japan (Sabin 1952), a nd China (Qiu et al. 1981). Epidemics of dengue fever vectored primarily by Ae. albopictus may behave differently than epidemics borne by Ae. aegypti Dengue epidemics vectored by Ae. aegypti have been much more explosive with many cases of ten occurring within the first few months of the outbreak as was the case during the 1943-44 outbreak in Honolulu (Gilbertson 1945). Aedes albopictus was suspected of transm itting dengue in the 2001-2002 Hawaiian outbreak because of the absence of Ae. aegypti at the time of the outbreak in areas where it occurred. The la tter outbreak resulted in far fewer cases than the 19431944 outbreak (200 vs. 1,500, respectively) and occu rred over a longer period of time. While socioeconomic conditions (lifestyle and home improvements; air conditioning, television, screened windows) undoubtedly f actor into the disparity between the 2 Hawaiian outbreaks, Ae. aegypti had largely vanished from Hawaii at the time of the last outbreak as opposed to the 1943-1944 ep idemic (Kenneth Hall, personal communication). A second factor that might lessen the severity of an Ae. albopictus vectored dengue epidemic is its feeding preferences compared to those of Ae. aegypti Aedes albopictus feeds on a wide variety of animal s (Niebylski et al. 1994), whereas Ae. aegypti prefer humans (Christophers 1960, Harringt on et al. 2001). This factor lessens the threat of a vector-borne disease outbreak in humans compared to a vector that has a feeding preference for humans. Regardless, Ae. albopictus is essentially a competent vector of dengue viruses, being highly suscep tible to oral infecti on (Rosen et al. 1983)
8 and capable of transovarial transmi ssion (Shroyer 1990), thus functioning as a maintenance vector. Other viruses discovered in American field-caught Ae. albopictus include Potosi virus (Harrison et al. 1995, Mitchell 1995) Tensaw virus (Mitchell et al. 1992), and Keystone virus (Nayar et al. 2001). Experimentally, Ae. albopictus has been shown to be a competent vector of many disease agents of public health concern, most of which occur overseas. Included are all 4 dengue viruses (Rosen et al. 1985), Japanese encephalitis virus (Mitchell 1995), eastern equine encephalitis virus (Mitchell et al. 1992 ), Venezuelan equine encephalitis virus, (Turell and Beaman 1992), chikungunya vi rus (Yamanishi 1999), Ross River virus (Mitchell and Gubler 1987), and Rift Valley fever virus (Turell et al. 1988). The potential for Ae. albopictus to acquire and transmit these pathogens has led to concern among public health and vector control practi tioners that another exotic disease agent introduction into the United States such as the case of West Nile virus in New York in 1999 could cause a further burden to human health and agricu lture in the United States. In summary, Ae. albopictus is an exotic mosquito that has rapidly colonized much of the southern and Midwest United States si nce its introduction 20 years ago. It is a serious nuisance pest and a secondary disease vector that has largely displaced the more important disease vector Ae. aegypti Aedes albopictus like its closely related Ae. aegypti is a difficult mosquito to control by conventional ULV methods. Several questions arise as our unwelcome guest con tinues to consolidate its hold on the eastern half of our country: In addition to being a severe nuisance, coul d it serve as a dengue vector should this disease agent be introduced into the southern U.S.? Aedes albopictus recently vectored dengue virus in Hawaii and possi bly in South Texas (Rawlings et al. 1996).
9 Are commonly used adulticides effective in killing this mosquito? Is it developing resistance to these insecticides? Adultici des are usually the first line of defense in attempts to achieve a rapid reduc tion of pest mosquito populations. Are any of the many new residential mo squito traps effective at catching Ae. albopictus ? How effective are older survei llance traps in comparison to new technology traps? Might some of these tr aps be used in conjunction with other control efforts to reduce populations of biting adult Ae. albopictus ? Literature Review of Mosquito Attractants Classification of Mosquito Attractants Adult female mosquitoes use visual cues in their quest to locate mates, oviposition sites, food sources (carbohydrat es), resting sites, overwinte ring sites, and blood meal hosts (Allan et al. 1987). In addition, they use chemical and physical cues in hostseeking activity (Laarman 1955). Host-see king visual cues include size, shape, movement, contrasting color, and light. Chemical cues involve carbon dioxide (CO2) and other odors that are produced by vertebrates and act as attractants. Physical cues include radiant and convective heat, moisture, sound, and surface structure (Laarman 1955). Much research has been done to identify these attractants, find which attractants work best for specific mosquito species, and incorporate them into mosquito traps for surveillance and control (Ser vice 1993, Mboera and Takken 1997). One purpose of this study was to determine which residentia l traps performed best at capturing Ae. albopictus Visual and physical cues characteris tic of a trap might be useful in luring Ae. albopictus to the trap. We reviewed those visual and physical characteristics in general for all mosquitoes and then specifically for Ae. albopictus We also reviewed chemical attractants. Residential traps supplemented w ith lures were assessed as to their lures effectiveness in large-cage trials and then added during field tr ials if they proved beneficial in capturing Ae. albopictus The more effective traps may prove useful in
10 supplementing traditional control methods (pes ticide application) in an integrated program aimed at reducing nuisance populations of Ae. albopictus (Kline 1994a, Mboera and Takken 1997). Host-Seeking Activity of Mosquitoes Host-seeking behavior of adult female mo squitoes is linked to the physiological requirements of ovarian development. With few exceptions, mosquitoes are obligatory blood feeders. Many species of mosquito can produce a first batch of eggs without a blood meal if larvae acquire enough protein fr om their environmen t while feeding, but this batch is often much smaller than an egg batch produced from a blood meal. This condition is termed autogeny and is believed to be a mechanism to enhance survival in situations in which hosts are scarce (Cleme nts 1992). Apart from autogeny, mosquitoes must locate a host to obta in blood for egg production. Three behavioral steps ar e involved in host location by hematophagous insects: appetitive searching, activation and orient ation, and attraction (Sutcliffe 1987). Mosquitoes engaged in appetitive searching ar e hungry and seek a hos t to feed on. Their flight is non-oriented as they search for visu al or chemical cues (or both). Upon receipt of a host cue, the insect is activated and switches to orient ed flight toward the host. Diurnally active mosquitoes may rely more on visual stimuli (color and movement) during activation for host location than nocturnally feeding mos quito species (Allan et al. 1987). Although it is unknown exactly how hematophagous insects follow an odor source to its origin, hypotheses include the ability to follow odor plumes of increasing concentration (Sutton 1953), de tection of pulse frequency of a plume (Wright 1958), and discernment of plume boundaries (Farkas and Sh orey 1972). In the la st stage (attraction), the insect makes a final decision of whether to feed in the immediate presence of the host,
11 often while in contact with th e host. Heat, water vapor, visu al attraction, degree of host specificity, gustatory preferen ce, and blood hunger influence this final decision to take a blood meal (Lehane 1991). Physical cues of h eat and moisture come into play only in the attraction stage when the insect is in or nearly in direct contact with the host. Visual Attractants of Mosquitoes Extensive field studies over the last 70 years sought to determine the visual qualities attractive to host-seeking mosquitoes. Many of these studies used various types and combinations of mechanical trapping de vices to uncover key physical and chemical attributes effective in targeting many species of mosquitoes (Service 1993). Visual and chemical components are among the most importa nt trap attributes in attracting hostseeking mosquitoes. Visual qualities of host shape and size, motion, patterning, trap color, color contrast, light color, and light in tensity serve as visual stimuli (Bidlingmayer 1994). Host shape and size often play an important role in attraction for many species of mosquitoes. Bidlingmayer and Hem (1980) found that large unpainted plywood-covered suction traps presenting large conspicuous s ilhouettes were visible to many nocturnally active mosquitoes for distances up to 19 m. When capture rates of covered traps were compared to uncovered (inconspicuous) traps, larger catches were made in covered traps for 12 of 13 species collected. Gillie s and Wilkes (1974) found that African Culex mosquitoes were visually attracted to larg e ramp traps over smaller suction traps while the reverse was true for most other genera of mosquitoes. Browne and Bennett (1981) showed trap-shape preferences among Canadi an woodland mosquitoes using 2 traps of equal surface area but different shape. They found that Ochlerotatus cantator (Coquillett) and Mansonia perturbans (Walker) preferred cube traps to pyramid traps by
12 a ratio of 2:1, whereas Oc. punctor (Kirby) preferred pyramid traps to cube traps (p < 0.01) in all cases. Paul Choate (Departm ent of Entomology, University of Florida, Gainesville, FL) suggested that the size of a host played an important role in host selection among mosquito species based on blood meal analysis. He believed that host selection was driven by a mosquito species preference for a specific quantity of expired CO2, which is a function of host size (Kline 1994a). Movement has been shown to increase the attractancy of living and inanimate objects to host-seeking female mosquitoes. Moving targets were twice as attractive to diurnally active Aedes mosquitoes in comparison to stationary targets of similar size under laboratory conditions (Brown 1956). Wood and Wright (1968) found a small, positive effect on capture of Ae. aegypti on similar rotating vs. stationary targets in laboratory investigations. When 2 mice enclos ed in transparent airtight containers (1 anesthetized, 1 moving) were exposed to hungry Ae. aegypti the moving mouse attracted 3.7 x more females than the anesthetized mouse (Sippell and Brown 1953). Gillett (1972) credited movement as the primary a ttractant to field workers attacked by biting Aedes mosquitoes. In Japan, Kusakabe and Ik eshoji (1990) noticed increased catches of Ae. albopictus in traps operating in the vicinity of field personnel compared to unattended traps. They attributed increased trap capture of Ae. albopictus to the mosquitos attraction to worker movement. Recently, an experimental moving-ta rget trap was field tested at an Ae. albopictus infested site in Texas and found to be efficient in collecting it (Dennett et al. 2004). Thus, it appears that the visual aspects of size, shape, and movement of a potential food target in fluence host choice among host-seeking Aedes mosquitoes.
13 Transmitted light has long been known to attract host-seeking mosquitoes (Headlee 1932). Many studies show species-specific pref erences to light of varying wavelength in both laboratory and field settings. Hematopha gous insects are generally sensitive to UV light (350 nm to 400 nm) and less responsive to the red end of the color spectrum (above 650 nm). Most hematophagous insects show pe ak sensitivities to color of 355 nm (UV) and 450 nm to 550 nm (blue-green), while some higher Diptera (stable fl ies, tsetse flies) respond to the red/orange portion of the spectrum of about 620 nm (Lehane 1991). In Florida, Ali et al. (1989) field-tested 6 different co lored light bulbs (100 W) in New Jersey light traps (NJLT) to determine attractancy by species. Culex and Psorophora mosquitoes made up most of the catc h and blue and white light was more attractive than yellow, red, orange, or green lig ht. Burkett et al. (1998) field tested CDC light traps set with red, orange, yellow, green, infrared, and blue light emitting diodes (LEDs) and incandescent (white) light for Florida woodland mosquitoes. They found significant differences in color preference to 8 species belonging to Ochlerotatus, Anopheles, Culiseta, Culex, Psorophora and Uranotaenia mostly for blue, green, and white light. The effect of white and blue light was particularly evident on trap capture of Ps. columbiae (Dyar and Knab), Cs. melanura (Coquillett), An. crucians Wiedemann, and Ur. sapphirina (Osten Sacken). In laboratory tests using filtered light of 50 nm wavelength intervals (350 nm to 700 nm), Burkett (1998) showed that Ae. albopictus responded most frequently to 600 nm (yelloworange) and 500 nm (green) light. In his study, Ae. aegypti responded most frequently to 600 nm (orange) and white light. Electroretinograph (ERG ) examination of Ae. aegypti revealed bimodal peak sensitivities at 323 nm to345 nm (UV) and 523 nm (green) li ght with a detection range from 323 nm
14 to 621 nm (Muir et al. 1992). Li ttle information exists as to Ae. albopictuss preference, if any, for light in the field. Herbert et al. (1972) collected significantly more Ae. albopictus in Vietnam using unlit, CO2-baited CDC light traps than using lit and CO2baited CDC light traps. Surveillan ce with CDC light traps set in Ae. albopictus -infested regions of northern Thailand s howed that incandescent light wa s not particularly effective in attracting it (Thurm an and Thurman 1955). Light intensity may play a role in mosqu ito attraction. Headlee (1937) stated that light of attractive frequencies could be made repellent to mosquitoes above a certain point of intensity, although he offered no c onvincing data to suppor t this view. New Jersey light traps fitted with 100 W lamps a ttracted more mosquitoes than NJLTs fitted with standard 25 W lamps (Headlee 1932). Br eyev (1963) found that traps fitted with 2 109 W lamps caught 3.5 x more mosquitoes than a simila r trap fitted with 1 25 W lamp. Ali et al. (1989) found no significant differences in the attractiveness of colored lamps of 25-, 40-, and 100 W among 17 species of mosquito es collected in Florida; the number of Aedes mosquitoes collected was insufficient fo r analysis. Barr et al. (1963) found a positive correlation with increasing light intens ity to catches of California rice field mosquitoes testing traps set with 25-, 50-, 75-, and 100 W lamps of similar color. Ochlerotatus melanimon (Dyar) was predominant and incr easingly caught at higher light intensity. Breyev (1963) caught significantly more Ae. vexans (Meigen) with 220 W lamp traps as opposed to similar traps set w ith 109 W lamps. It appears that woodland Aedes mosquitoes prefer higher intensity li ght in the range of 25 W to 220 W. Aedes albopictus and Ae. aegypti seem little attracted to light of intensities commonly used in adult traps (25 W to 100 W). This may be expected of diurnal species (Thurman and
15 Thurman 1955), but laboratory trials indicate that Ae. albopictus prefer yellow-orange and green light to blue, UV, red, and infrared light (Burkett 1998). Little research has been conducted into the attractiveness of flickering light to mosquitoes. Mosquitoes might find flickering light more attractive than steady light due to extra contrast against the environment. The rate of light flicker is measured in cycles per second, or Hertz (Hz). Flicker fusion fr equency is defined as the frequency of flickering light at which the eye is no longer able to distinguish f licker and is a function of the recovery time of photoreceptors; it is a function of the eye s ability to discern movement in the environmen t. (Lehane 1991). Flicker ra tes above the flicker fusion frequency of a particular organism give th e appearance of steady, non-flickering light to the organism. In general, flicker fusion fr equency in man is between 20 and 30 flashes per second (20 Hz to 30 Hz), 200 Hz to 300 Hz in fast-flying diurnal flies, and 10 Hz to 40 Hz in slow-flying nocturnal insect s (Mazokhin-Porshnyakov 1969). Two studies using flickering light as a mosquito attractan t showed no advantage in capture rates over traps with non-flickering lamps (Vavra et al. 1974, Lang 1984). Both studies used low flicker rates (2 Hz and 1 Hz to 60 Hz, resp ectively). More study is needed on this subject, especially on the higher frequencie s of 500 Hz to 600 Hz at which mosquito flight occurs (Dr. Jerry Bu tler, personal communication). Reflected light as (opposed to transmitted light) is generated from sunlight or artificial light, and imparts color to the objec t from which it is reflected. Diurnally active mosquitoes are more likely to have better de veloped color vision than nocturnal species that probably rely more on intensity contrast than color for host location (Allan et al. 1987). Brett (1938) showed color preference for black and red in Ae. aegypti and stated
16 that this mosquito has colour vision and a colour preference. Gjullin (1947) believed that mosquito response to color was a result of their attraction to spectral reflectance rather than color discrimina tion. Either way, the colors we perceive show varying degrees of attractiveness to different species of mosquitoes. Most field research shows that most mosquitoes are attracted to darker, less reflective colors than to lighter, more reflective colors. Brown (1954) found that Canadian woodland Aedes mosquitoes were more attracted to traps set with darker colored cloth (black, blue, and red) than to traps set with lighter colored cloth (green, white, and yellow). Attractive ness varied inversely with reflectivity or bright ness within a range of colored surfaces (475 nm to 625 nm) (blue to orange). However, red cloths were sometimes highly attractive to a few Aedes species. Red color is also used to lure Anopheles mosquitoes to box shelters used in surveillance programs (G oodwin 1942, Breeland 1972). Aedes aegypti is known to prefer black, shiny su rfaces to other colors (Brett 1938, Brown 1956, Gilbert and Gouck 1957). This finding led to the development of the Wilton mosquito trap (Wilton and Kloter 1985) that uses a black shiny cylinder as an attractant for Ae. aegypti These same qualities apparently attract Ae. albopictus ; shiny black plastic cups serve well as oviposition containers for this mosquito (personal observation). Black color has been incorporated into traps targeting Ae. albopictus and Ae. aegypti The duplex cone trap, a suction trap uses a glossy black inner cone to visually attract Ae. albopictus In a Louisiana study, the dup lex cone trap was preferred by Ae. albopictus over CDC light traps, gravid trap s, animal-baited (hamster) traps, malaise, and Trinidad traps. The duplex c one trap caught as many adults as did human-
17 bait collections (Freier and Francy 1991). The black in terior cone color and CO2 were the only attractant factors the trap used. Color contrast especially between wh ite and black, has a noticeable effect on mosquito attraction. Sippell and Brown (1953) showed that Ae. aegypti were increasingly attracted to cubes painted with alternating square check s of black and white, as the size of the checks decreased. The same experiment showed that cubes painted with alternating stripes of black and white became more attractive to Ae. aegypti as the ratio of black and white stripes increased (i.e., as the size of the stripes decr eased). Brown (1956) likewise reported that contour (black-white ch ecked interface) raised the attractiveness of a target by 1.6 x over a surface of uniform color to woodland Aedes mosquitoes. He said the flicker effect created by these contrasting colors was the factor responsible for attraction. This study also showed that solid black panels were 5 x more attractive to Ae. aegypti and 4 x to 10 x more attractive to field Aedes mosquitoes than solid white panels. Haufe (1964) found that black and white-color ed traps performed better than woodlandcolored traps but did not say which species we re responsive, or what type of trap was used. The Fay-Prince sucti on trap was designed to catch Ae. aegypti based on color contrast (black suction tube and panel contra st with a white cover) (Fay and Prince 1970). It caught significantly more Ae. aegypti than other portable trap s using black color and light as attractants showing that co lor contrast was more attractive to Ae. aegypti than light or solid color (Klote r et al. 1983). Research on Ae. albopictuss attraction to color showed that black, reflective surfaces are more attractive than dull black or white surfaces (Kusakabe and Ikeshoji 1989). Traps using color contrast, especially black and
18 white, need further investigation to compare their Ae. albopictus attraction to that of noncontrasting traps. Chemical Attractants of Mosquitoes Hematophagous insects are known to be attr acted to certain biologically derived waste secretions and odors commonly found in vertebrate urine, breath, and skin emanations. These chemical compounds (odors) activate compound-specific receptors that, in turn, stimulate host-seeking be havior in hungry insects (Lehane 1991). Compounds known to attract mo squitoes include carbon CO2 (Rudolfs 1922), 1-octen-3ol (octenol) (Takken and Kline 1989, Kline et al. 1991a, Kline et al 1991b), lactic acid (Kline et al. 1990), acetone (Ber nier et al. 2003), phenols (K line et al. 1990), and some amino acids, especially lysine and alanin e (Brown and Carmichael 1961, Roessler and Brown 1964). Of these, only 3 are sold for use with mosquito traps: CO2, octenol, and lactic acid. Carbon dioxide is one of many products of vertebrate resp iration. It is well known for its usefulness as a mosquito attrac tant (Rudolfs 1922, Gillies 1980, Mboera and Takken 1997) and is the most commonly used and easily obtainable attractant for hematophagous insects. Recently, Shone et al. (2003) caught significantly more Ae. albopictus in CO2-baited traps than in unbaited traps. Herbert et al. (1972) obtained similar results with Ae. albopictus comparing CO2-baited and unbaited CDC light traps. One-octen-3-ol, or octenol, a component of ox breath, was initially discovered to be a potent attractant for several sp ecies of tsetse fly (Hall et al. 1984, Vale and Hall 1985). Shortly afterwards, it was shown to be an at tractant for several important genera of mosquitoes including several Aedes species (Kline et al. 1991b, Kline 1994b). In general, octenol used alone has been a good at tractant for only a few mosquito species
19 but apparently produces an additive or synergistic response in some Aedes, Anopheles, Coquillettidia, Psorophora and Mansonia mosquitoes in the presence of CO2. In North America, Shone et al. (2003) found that Fay-Prince traps baited with octenol + CO2 were significantly more attractive to Ae. albopictus than traps baited with octenol alone were. No significant difference in catch was noted between CO2-baited and octenol + CO2baited Fay-Prince traps. In contrast, octenol + CO2-baited CDC light traps caught fewer Ae. albopictus than those traps set with CO2 alone in northern Thailand (Vythilingam et al. 1992). Lower capture rates of Ae. albopictus were also seen in trials using CO2generating Mosquito Magnet Pro traps in Hawaii (Sean Bedard, American Biophysics Corporation, personal communica tion). Further investigation is needed to determine the effect of CO2-baited traps with and without octenol on capture rates of Ae. albopictus Lactic acid (L (+)-stereoisomer of lactic acid ) is an end product of glycolysis and is found in muscle, in blood, and on skin of ve rtebrates (Mahler and Cordes 1971). It has been found to be an attractant for a few mosquito species, mostly for Aedes and Anopheles (Kline et al. 1990, Murphy et al. 2001, Bernier et al. 2003). It is only produced in the L(+) form in vertebrates and it is 5 x more attractive to Ae. aegypti than its D-lactic acid isomer (Acree et al. 1968). In South Carolina, more Ae. vexans and Oc. sollicitans (Walker) were caught using only minute quantities of lactic acid in CO2baited, unlit NJLTs than were caught with sim ilar traps lacking lact ic acid (Stryker and Young 1970). Lactic acid-baited suction trap s at Edgewood Arsenal, Maryland largely decreased the catch of Cx. quinquefasciatus Say Cx. restuans Theobald, and Mn. perturbans (Walker) (USAEHA 1970) compared to lactic acid-free traps. Aedes aegypti is known to be attracted to lac tic acid at low levels (Bernier et al. 2003). Two classes of
20 grooved peg sensilla on the antennae of Ae. aegypti have been located that are sensitive to lactic acid (Davis and Sokolove 1976). One sensilla class is lactic acid-excited and the other is lactic acid-inhibited. Interaction between these 2 cl asses of sensilla affects the degree of attraction of Ae. aegypti to lactic acid sources. Low concentrations (~2 L) attract Ae. aegypti (Bernier et al. 2003) whereas hi gher concentrations repel them. Water-diluted lactic acid applied to huma n subjects was increasingly repellent to Ae. albopictus at concentrations from 1 to 10,000 parts per million (ppm) (Shirai et al. 2001). Research is needed to determine the response of Ae. albopictus to commercially packaged lactic acid lures in the field. To date, only octenol and l actic acid baits have been developed into commercial attractants for trap use with residential mo squito traps. They come in prepackaged, disposable plastic containers for use w ith American Biophysics Corporation products (Bio Sensory also manufactures bait cartr idges for the Coleman Company, Blue Rhino, and others). The effect of octenol on many mosquito species has been published, however, few studies have investigated Ae. albopictus response to it. Initial observations and field reports indicate the Ae. albopictus may be somewhat repelled by octenol and attracted to lactic acid. Furthe r investigation is needed to ascertain the re sponse of this mosquito to these attractants in the field. Carbon dioxide is a potent attr actant for most mosquito species (Gillies 1980). It has traditionally been supplied to traps in the form of dry ice and compressed gas, typically at release rates between 50 mL/min to 1000 mL/min (Kline et al. 1991b, Kline 1994b). It is an attractant at these rates but becomes somewh at repellent at increasingly higher release rates ( 2000 mL/min) (Kline and Mann 1998). Recently, new technology
21 has been incorporated in to traps that produce CO2, heat, and moisture from propane. These stand-alone residential traps typically operate for several weeks on standard 20 lb (9 kg) propane tanks compared to traditiona l surveillance traps th at require daily or weekly replenishment of CO2 for effective operation. Th e author has noted initial positive response of Ae. albopictus to these traps in Hawaii, but few efficacy trials have been published comparing supplemented CO2 traps to stand-along CO2 traps in capturing Ae. albopictus Physical Attractants of Mosquitoes Physical attractants of mo squitoes include radiant a nd convective heat, moisture, sound, and surface structure (Laarman 1955). Hematophagous insects encounter these qualities as they make their final approach to a target during the at traction stage of hostseeking behavior (Dodd and Burgess 1995). Radiant and convective heat has been ex tensively investigated as mosquito attractants. Radiant heat is transferred through space withou t heating the space itself but heats objects in which it comes into contac t. Surface heat of a body that raises the temperature of the surrounding medium (air) is convective heat. Howlett (1910) first reported that convective heat was more attractiv e to mosquitoes than radiated heat. He placed a glass tube filled with warm water next to a gauze bag filled with Ae. scutellaris and observed their reaction. The tube was rela tively non-attractive when held parallel to the side of the bag. At an equal distance away but beneath the bag, the tube became attractive to mosquitoes and they attempted to feed. Peterson and Brown (1951) demonstrated the same phenomena using heated billiard balls as a heat source. Aedes aegypti attempted to feed on heated balls (110oF) placed at the bottom of a cage. Feeding activity stopped with the insertion of an airtight
22 window of crystalline thalli um bromoiodide between the ball and mosquitoes. The crystalline window allowed passa ge of radiant heat but bloc ked convective heat. Robots heated to 98oF attracted 3 x more woodland Aedes mosquitoes than robots heated to 50oF to 65oF (Brown 1951), and heat added to trap s increased the capture of salt marsh Oc. taeniorhynchus (Wiedemann) in Florida (Kline and Lemire 1995). Kusakabe and Ikeshoji (1990) caught significantly more female Ae. albopictus in traps to which a heat plate (36oC to 40oC) was added compared to traps used at ambient temperatures. Targets heated within the range of body temp eratures are more attractive to Ae. aegypti (and probably Ae. albopictus ) than targets at ambient temperat ures provided that there is at least several degrees (Celsi us) difference between the tw o (Peterson and Brown 1951). Moisture is considered by many authors to be the most important factor influencing the attraction of mosquitoes to an obj ect (Brown 1951, Brown et al. 1951, Brown 1956). As an attractant, moisture is greatly enhanced when blended with CO2 and heat, which mimics vertebrate breath (Wood and Wright 1968, Khan et al. 1966). Moist surfaces are often more attractive to Aedes mosquitoes than dry surfaces (Brown 1951), and high moisture content (relative humid ity) is generally beneficial for mosquito longevity. Moist air has been found to be 3 x to 5 x more attractive than dry air to Ae. aegypti in laboratory studies (Brown et al. 1951), a nd Brown (1951) found that moisture was the most important factor in attracting woodland Aedes mosquitoes to a target when ambient temperatures were above 60oF. It appears that moistu re-producing traps would be advantageous for trapping Aedes mosquitoes compared with traps that produce no moisture, but comparison trials with both trap types are needed to determine their efficacy in collecting Ae. albopictus
23 Sounds are periodic mechanical vibrations within bodies of gas, liquid, or elastic solids measured in terms of frequency and intensity (Clements 1999). In mosquitoes, sound waves strike sound-sensitive areas of the antennae (chordotonal organs) causing vibration of receptors that in turn stimu late electrical firing of sensory neurons. Chordotonal organs are mechanotransducers com posed of sensory units called scolopidia. Chordotonal organs in culicids are housed with in the Johnstons organ on the pedicel of the antenna and in male mosquitoes each Johnstons organ contains approximately 7,000 scolopidia. Stimulation of the Johnstons organ enables mosquitoes to determine the direction of incoming sound energy for mate location (Clements 1999). Male and female culicids are capable of discerning a range of sound frequencies and respond to frequencies of 250 Hz to 1,500 Hz (Kahn et al. 1945). Most mosquitoes produce tonal emanations that are species un ique and closely related species such as Ae. aegypti and Ae. albopictus can sometimes be distinguished by flight tone alone (Kahn 1945). Aedes albopictus male and female wingbeat fr equencies average 724 Hz to 772 Hz and 524 Hz to 542 Hz, respectively (Kanda et al. 1987, Brogdon 1994). Several investigators have used sound fr equencies within the range of mosquito flight to trap mosquitoes. Ikeshoji an d Ogawa (1988) used 400 Hz sound to trap Ae. albopictus in a forest. They caught more female Ae. albopictus using a 400 Hz (female wingbeat frequency) than a 900 Hz (male wi ngbeat frequency) sound trap and reported a dramatic decrease in field populations of Ae. albopictus after only 8 days of trapping. Kanda et al. (1987) found 480 Hz sound attractive to Ae. albopictus males, a frequency close to female wingbeat frequency (524 Hz). Aedes albopictus males were collected in significantly larger numbers in sound-baited bl ack traps than lactic acidor heat-baited
24 black traps (Kusakabe and Ikeshoji 1990) Ikeshoji and Yap (1990) reduced Ae. albopictus male and female field populations by 81% and 76%, respectively, over a 2week trapping period on a 2 ha Malaysian pi neapple farm using in secticide-tr eated sound traps of 400 Hz. This study demonstrated the possibility of using sound traps to decimate Ae. albopictus populations in a limited area. Presen tly, the only commercially available mosquito sound traps mimic animal heartbeat. Further investigation is needed for both animaland frequency-specific sound traps to assess their effectiveness in collecting Ae. albopictus Several authors have review ed surface structure (textu re) in relation to oviposition preferences of Aedes mosquitoes (Wilton 1968, Russo 1978, Thirapatsakun et al. 1981). In general, rough surfaces are favored among ovipositing Aedes mosquitoes to smooth surfaces. No published reports were found of trap surface texture as it pertains to mosquito attraction. Introduction to Surveillance and Residential Traps Used for Mosquito Surveillance and Control Trapping systems used for mosquito survei llance and control ha s been extensively reviewed by Service (1993). Service esse ntially split adult mosquito traps into 3 categories: non-attractant traps, animal-baited traps, and attrac tant traps. Non-attractant traps sample adult mosquito populations in a non-biased fashion; mosquitoes captured are not necessarily seeking a blood meal or targ eting the trap, but are inadvertently caught while flying. Non-attractant traps include Malaise, cartop, ramp, and rotary traps. Animal-baited traps come in a variety of sh apes and sizes depending on the animal used as bait. They can be extremely effective in luring mosquitoes due to the animals production of kairomones (CO2, lactic acid, acetone, phenols, and/or octenol), visual
25 qualities (size, shape, and movement) and phys ical attractants (texture, heat and water vapor). Animal-baited traps are usually im practical for homeowners and vector control agencies due to food, shelter, and maintenance requirements. Attracta nt traps use all or some combination of chemical-, visual(color color contrast, and/or light) and physical features (sound, heat, and moistu re) to attract mosquitoes. These cues were discussed in depth in the previous section. Many of thes e traps are available to the general public although some effective experimental traps ha ve never been marketed, (i.e., the duplex cone trap, Freier and Francy 1991). Since the last edi tion of Services book (1993), stand-alone CO2-generating propane traps have b een developed primarily for use on homeowner and business properties. A review of residential and surveillance mosquito traps follows. Carbon Dioxide-Supplemented Traps Carbon dioxide-supplemented mosquito traps include CDC type traps, 1 counterflow geometry trap (CFG), the omni -directional Fay-Prince trap, the Wilton trap, adhesive traps, and sound traps. Manuf acturers recommend that these traps be supplemented with CO2, as they do not generate this critically important attractant (Rudolfs 1922, Reeves and Hammon 1942, Brown et al. 1951, Laarman 1955). CDC light trap. A prototype CDC miniature light trap was developed by Sudia and Chamberlain (1962). There are many varieties of the CDC light tra p, all of which are small, lightweight, and use lig ht to attract mosquitoes. Along with the bulky and heavy New Jersey Light trap that uses a 25 W lamp as its sole attr actant, CDC light traps are the mainstay of mosquito survei llance in the United States (K line 1999). The CDC light trap (model 512, John W. Hock Company, Gainesville FL) is one of the most popular CDC light traps owing to its rugge dness and effectiveness in colle cting mosquitoes. It is
26 battery powered which allows for sampling in remote locations apart from alternating current (AC) electricity sources (required to operate the NJLT). This trap is usually supplemented with dry ice or compressed CO2 gas to greatly enhance mosquito capture (McNelly 1995). In Thailand, Mi ller et al. (1969) caught 100 x and 30 x more mosquitoes ( Aedes Mansonia and Culex species) in lit + CO2-baited CDC traps compared to unlit + CO2-baited-, and unbaited traps, respec tively. In Vietnam, Herbert et al. (1972) trapped significantly more Ae. albopictus in unlit + CO2-baited CDC light traps than in lit + CO2-baited and unlit + unbaited CDC light traps. In Louisiana, Freier and Francy (1991) caught fewer Ae. albopictus in lit + CO2-baited CDC light traps than in all but 1 of 6 other surveillance traps in a co mparison study. Kloter et al. (1983) caught more male Ae. aegypti in lit + CO2-baited CDC light traps than in unlit + CO2-baited CDC light traps. Female Ae. aegypti responded to lit and unlit traps in approximately equal numbers. Neither species responded well to CDC light traps, but further investigation is warranted with respect to Ae. albopictus Mosquito magnet-X (MM-X). Counterflow geometry (CFG) traps differ from other suction traps in that mos quitoes are caught in an updraft air current at the bottom of the trap as opposed to a downdr aft air current at the top of the trap, typical of most suction traps. In addition, a countercurrent (downdraft) is generated separately below the intake tube distributing a pl ume of attractant beneath the trap. The MM-X trap uses 2 fans to create to opposing ai r current; it is the only non-CO2-generating CFG trap. The MM-X is used with either dry ice or compressed CO2 gas and can be supplemented with octenol, Lurex (lactic acid bait, American Biophysics Corporation, N. Kingstown, RI) or both. Due to the recent development of this trap, published efficacy data are scant. In
27 Florida field tests, Kline (1999) caught 7.8 x more mosquitoes in the MM-X than in an ABC professional trap (CDC-type trap). In South Korea, 2 MM-X traps, both baited with CO2 and 1 baited with octenol, outperformed 3 other traps (CDC light trap, Mosquito Magnet Pro, and Shannon trap) baited w ith various combinations of light, CO2 and octenol (Burkett et al. 2001). In larg e-cage trials, Kline (2002) caught more Ae. aegypti in the MM-X than in either of 2 propane-pow ered CFG traps or the ABC trap. Overall, the MM-X has fared well in comparison tests ag ainst other trap types for several genera of mosquitoes (Mboera et al. 2000, Burkett et al. 2001, Kline 2002), but studies detailing its efficacy in trapping Ae. albopictus are lacking. Fay-Prince trap. The Fay-Prince trap was design specifically to capture Ae. aegypti. The omni-directional Fay-Prince trap is a modification of the original singledirectional Fay-Prince trap (Fay and Prince 1970). This suction trap makes use of 2 visual features to attract Ae. aegypti : contrasting black and white color and black panels. It is supplemented with dry ice or compressed CO2 to enhance capture of Ae. aegypti and Ae. albopictus In Louisiana, the Fay-Prince trap caught significantly more Ae. aegypti with or without CO2 than black cylinder (Wilton) traps or unlit and lit CDC light traps with different colored lights (Kloter et al. 1983). Schoele r et al. (2004) compared the Fay-Prince-, ABC-, and the Wilton trap capture against human-landing rates for Ae. aeg ypti in Peru. The Fay-Prince trap caught significantly less Ae. aegypti than the other traps, however, with the ex ception of human-landing c ounts, all collections of Ae. aegypti were relatively small (< 100) over the cour se of the study. Shone et al. (2003) caught significantly more Ae. albopictus with octenol + CO2-baited Fay-Prince traps than in unbaited Fay-Prince traps. No significant diffe rence was seen between trap captures with
28 and without octenol provided the trap was baited with CO2. Fay-Prince traps collected significantly more Ae. albopictus and Ae. aegypti than duplex cone, CDC light-, and bidirectional Fay-Prince traps in north Florida (Jensen et al. 1994). Originally designed to collect Ae. aegypti the Fay-Prince trap has proven effective in capturing Ae. albopictus Comparison trials are needed to further asse ss this trap against newer stand-alone CFG models. Wilton trap. Black cylinder suction traps we re found to be highly attractive to adult male and female Ae. aegypti (Fay 1968). Further refine ments produced the CDC Wilton trap that is effective in collecting Ae. aegypti and Culex quinquefasciatus (Wilton and Kloter 1985). It is a simple, compact, bl ack cylinder trap visu ally attractive to Aedes mosquitoes. Like other mosquito traps, it works best when supplemented with CO2. Few comparison studies have incorporated the W ilton trap; those mentioned above (Kloter et al. 1983, Schoeler et al. 2004) produced mixed results. No comparison studies were found that assessed the efficacy of Wilton traps in collecting Ae. albopictus Results of multiple-trap field studies incorporating the Wilton trap would provide useful data to decision makers involved in Ae. albopictus surveillance and control efforts. Mosquito Deleto. The Mosquito Deleto model 2200 is a non-suction adhesive trap that catalytically c onverts propane into CO2. The glossy black color of the adhesive strip and large size of this ground-mounted unit make it an excellent visual target. The trap can be supplemented with an octenol cartr idge provided by the manufacturer (The Coleman Company, Wichita, KS). The Mos quito Deleto was outperformed by 5 of 6 traps in a multiple comparison field trial in north Florida (Smith et al. 2003). The trap is relatively new to the commercial market and fu rther research is needed at ascertain its
29 effective in capturing Ae. albopictus Unpublished data from USDA ARS indicates that this trap has produced vari able results in capturing Ae. aegypti Bugjammer biting insect trap. This trap relies primarily on sound to attract hostseeking insects. A recorded dog heartbeat is provided by a speaker embedded in the head of the trap. Visual attracta nts include an alternating black and white stand and a glossy white adhesive strip. Sound traps typically use adhesive strips or non-adhesive, insecticide-treated surf aces to trap or kill host-seeki ng insects (Kanda et al. 1990, Ikeshoji and Yap 1990). Few researchers have field-test ed sound traps; the traps tested used high frequency buzzing, not animal heartbeat mimi cs to attract mos quitoes (Ikeshoji 1986, Kanda et al. 1987, Ikeshoji and Ogawa 1988). Experimental sound traps broadcast intermittent sound at 100 decibels between 370 Hz to 900 Hz to attract male and female mosquitoes (Ikeshoji et al. 1985, Ikeshoji 1986) Wingbeat frequencies of mosquitoes typically range from 300 Hz to 800 Hz (Moore et al. 1986) and impart the characteristic buzzing sound associated with many species of mosquitoes. Ikeshoji and Yap (1990) used 400 Hz insecticide-trea ted sound traps to reduce an Indonesian population of Ae. albopictus females and males by 80.9% and 75.6%, respectively. This study indicated that sound trapping may be effective in contro lling some species of mosquitoes. Traps emitting heartbeat sound as opposed to buzzing of high frequency acoustic systems need to be investigated for effectiveness in lu ring and attracting mos quitoes. A Bugjammer analog, the Sonic Web ( http://www.bugjammer.com/reports/cmave/tracey.htm ), was effective in trapping stable flies ( Stomoxys calcitrans ) but made no mention of mosquito capture. Testing is needed to assess the e ffectiveness of this type of sound trap in collecting Ae. albopictus
30 Carbon Dioxide-Generating (Propane) Traps Mosquito traps that use propane to generate power and produce CO2 have recently been developed and are now available for purchase to the general public (Kline 2002). Most of these traps are stand-alone devices meant to rid homeowners properties of nuisance levels of mosquitoes. They operate on standard 20 lb propane barbeque tanks. Several models have provided protection (s ignificant reduction in nuisance populations of biting insects) for up to 1 acre around the trap (Daniel Kline, USDA ARS, personal communication). These traps offer some crit ical advantages over traditional trapping systems: long term operation (3 weeks of uninterrupted operation) battery-free operation, no need for daily replenishment of dry i ce or use with cumbersome compressed gas cylinders, and the ability to operate in remote locations independent of electricity. Carbon dioxide-generating propane traps ar e unique in their operation. Propane is burned until a catalyst is sufficien tly heated, at which time th e flame extinguishes itself. The heated catalyst continuously co nverts propane into water vapor, CO2 and heat. Heated water vapor and CO2 are exhausted to the outside of the trap, providing a plume of mosquito attractants. Catalytic heat is provided to a second device adjacent to the catalyst, the Thermo Electric Module (TEM). A temperature gradient across the TEM (between the heated side and the ambient temp erature side) generates electricity used to power 2 fans. One fan provides the exhaust pl ume of attractants, the other fan provides a counterflow updraft used to trap host-seeking insects. Mosquito Magnet Pro. The Mosquito Ma gnet Pro (MM Pro) is a stand-alone propane trap that can be supplemented with octe nol or lactic acid bait cartridges. This trap is very effective in trapping mosquitoes and Culicoides biting midges (Daniel Kline, personal communication). In South Kor ea, Burkett et al. (2001) caught more Anopheles
31 Culex and Aedes mosquitoes in a MM Pro than in CD C light traps, ABC Pro traps, or NJLT light traps. The MM Pro was outpe rformed by MM-X and Shannon traps. In a similar Korean study, Burkett et al. (2002) caught more Culex tritaeniorhynchus Giles and Ae. vexans in an octenol-baited MM Pro than in all other traps which included a Shannon trap, ABC Pro, miniature black light trap, and NJLT. In large-cage trials, the MM Pro caught more mosquitoes than the ABC Pro trap but fewer mosquitoes than the MM-X trap (Kline 2002). In Ha waii, the MM Pro trapped more Ae. albopictus than CDC light traps (personal ob servation). To date, no studies have been published comparing the efficacy of the MM Pro to other trap types with respect to collecting Ae. albopictus Mosquito Magnet Liberty. The Mosquito Ma gnet Liberty (MM Liberty) is similar to the MM Pro in that it uses counte rflow technology and propane-generated CO2 to collect biting insects. The MM Liberty is smaller, more co mpact, and weighs less than the MM Pro. It relies on AC electricity for power. It too can be supplemented with octenol or lactic acid bait car tridges. The MM Liberty outp erformed 5 other residential traps in a north Florida comp arison trial (Smith et al. 2002) (Only 1 trap outperformed the MM Liberty, itself a propane burner). Th e MM Liberty produced superior mosquito catches in a comparison study using 7 traps and caught 3 x as many mosquitoes than the second best trap (Smith and Walsh 2003). Order of effectiveness was MM Liberty > SonicWeb > MM Defender > Lentek trap > Mo squito Deleto > Ecotrap. At a Houston, Texas tire repository, the MM Liberty caught more Ae. albopictus than 6 other traps (Dennett et al. 2004). Order of effectiv eness was MM Liberty > Fay-Prince trap > moving target trap (experimenta l) > Dragonfly > CDC light tr ap (unlit) > CDC light trap (lit) > Mosquito Deleto. The MM Libert y produced excellent results collecting Ae.
32 albopictus in most comparison trials, warranting further comparison trials with other residential and surveillance traps. Introduction to Aedes albopictus Adulticide Susceptibility Review Application of ultra low volum e (ULV) mosquito adulticides is the one of the most common and often the cheapest, quickest, and most effective method used by public health and vector control or ganizations to achieve a rapid reduction in nuisance and disease-carrying mosquitoes ( http://vector.ifas.ufl.edu/ ). Adulticides by definition are those chemical pesticides designed for us e against adult hematophagous insects, as opposed to larvacides that are used against th e larval stage of immature insects in an attempt to kill them before they reach the adult biting stage. The number of adulticides available to vect or control agencies for control of adult mosquitoes is in decline. Reasons incl ude extremely high pesticide research and development and re-registration costs for a restricted mark et (cost ineffective), the general publics concern over health and environmental hazards associated with broadcast insecticide applications, oppositi on to pesticide use by many environmental groups, and escalating pesticide litigation that results in increasing re gulatory restriction by governmental agencies (Fehrenbach 1990, Kline 1994a, Rose 2001). Thus, it is essential to monitor the pesticide susceptib ility status of key nuisance and vector mosquito species so timely cha nges in control strategies can be implemented to delay or prevent the development of resistance and pr eserve the usefulness of a limited number of public health pesticides. Aedes albopictus has been present in the United St ates for only 20 years and reports of its susceptibility to adu lticides commonly used in the United States are lacking. Initial insecticide susceptibility tests on Ae. albopictus were conducted in Texas, where it
33 first found established (Khoo et al. 1988, Robe rt and Olson 1989, Sames et al. 1996). Using the topical application me thod, Khoo et al. (1988) found that Ae. albopictus was resistant to malathion but susceptible to resmethrin (Scourge). Robert and Olson (1989) tested malathion, naled, bendiocarb, and resmethrin on Ae. albopictus using the coatedvial assay technique (Plapp 1971). Aedes albopictus adults were susceptible to bendiocarb and resmethrin but tolerant to malathion and naled. Using the coated-vial technique, Sames et al. (1996) found adult female Ae. albopictus collected from south Texas susceptible to malathion, chlor pyrifos, resmethrin, and permethrin. Aedes albopictus has been established in Fl orida since 1986 (OMeara et al. 1995a). Despite the existence of a tremendous amount of data on the susceptibility status of many Florida mosquito species (Breaud 1993) only 2 papers address the susceptibility status of Ae. albopictus in this state, both were larval as says. Ali et al. (1995) tested 10 insecticides on a la boratory strain of Ae. albopictus larvae. The study was useful in determining baseline le thal concentration (LC)50 and LC95 larvacide levels and relative toxicities between insecticides, but no comp arisons were made against field-collected larvae. Malathion was significantly less leth al than all other or ganophosphate (OP) and pyrethroid insecticides. Liu et al. (2004) tested 9 insecticid es against 4 field populations of Ae. albopictus larvae from Alabama and Florida a nd a susceptible laboratory colony. Larvae were susceptible to all insecticides except deltamethrin and chlorpyrifos, to which low levels of resistance was detected. It appears that Ae. albopictus in Texas and Florida have some level of tolerance to malathion a nd possibly deltamethrin, but further testing is needed to determine the extent of tolerance or resistance, if any, in field populations of
34 this pest in Florida. The efficacy of register ed mosquito adulticides in controlling Florida populations of Ae. albopictus is currently unknown. Research Objectives Aedes aegypti may well be the most extensively researched mosquito in the world (Christophers 1960, Clements 1992). It was th e first mosquito discovered to transmit a disease of human importance (yellow fever in 1901), and found to be established in most tropical and semitropical regions of the world. Because of this, extensive research over the past century has sought to determine its bionomics, feeding preferences, vector competency, distribution, seasonality, ovipositi on preferences, pesticide response, and the effects of source reduction, attr actants, repellents, and trap s in efforts to control it (Christophers 1960, Service 1993). The closely related and medically important Ae. albopictus has greatly expanded its range since the end of WW II and has become newly established in many regions of the world to include North America, Africa, and Europe (Watson 1967, Hawley 1988, Estrada-Franco and Craig 1995, Ali and Na yar 1997, Reiter 1998, CDC 2001). Research into the many aspects of its bionomics, v ector competency, and control is only now beginning to approach the amount of effort already expended on Ae. aegypti The goals of our research were to answer some important questions concerning the recent establishment of Ae. albopictus into Florida: Which of several commonly used residentia l mosquito traps are best at trapping this mosquito? How to they compare against surveillance traps specifically designed to capture Ae. aegypti ? Do they impact the biting rates of Ae. albopictus in their vicinity? Do commercial mosquito baits (octenol a nd lactic acid) enhance trap capture of Ae. albopictus in the presence of carbon dioxide (a universal mo squito attractant)? Are they better used alone or in combination?
35 Are any transmitted light colors more attract ive to this mosquito than others? Is flickering light more attractive than stea dy light? If so, at what frequency? Is Ae. albopictus developing resistance to any of 5 adulticides routinely used in Florida to control biting adult female mosquitoes? Finally, CDC light traps (model 512) were modified with power-preserving LEDs and small motors in an attempt to enha nce capture of woodland mosquitoes and extend battery life. Are these battery-life preserving modified traps as efficient as standard traps in collecting woodland mosquitoes?
36 CHAPTER 2 Aedes albopictus RESPONSE TO ADULT MOSQUITO TRAPS IN LARGE-CAGE TRIALS Introduction Florida contains a mix of climatological and physical characteristics shared by few other states: subtropical and mild temperate climates, relatively warm winters (January mean temperatures of 59oF and 68oF, Orlando and Miami, respectively), high annual average rainfall (54.1 in statewide) and high relative humidity rates ( http://water.dnr.state .sc.us/climate/sercc/climatein fo/monthly/monthly_seasonal.html ). Florida contains large expans es of wetlands and swamps, a nd is characterized by flat topography (Black 1993). These conditions provide mosquitoes with an ideal environment in which to live and breed. Historically, much of Florida remained uninhabitable in part due to severe nuisan ce populations of mosquitoes. Outbreaks of malaria, yellow fever, encephalitis, and de ngue further exasperated settlement efforts (Florida Coordinating Council on Mosquito Co ntrol 1998). Although the first permanent European settlement in North America was es tablished by the Spanish near present-day St. Augustine in 1564, Florida did not acquire a million residents until 1920. During that year, the neighboring states of South Carolin a, Georgia, and Alabama had populations of 1.6 million, 2.9 million and 2.3 milli on, respectively. Mosquitoes and other biting flies probably had as much influence as did climate in the lack of settlement and development of Florida during this period.
37 Technological advances in chemistry and mosquito control techniques during the 1930s and 1940s led to wide-scale control of nuisance populations of mosquitoes. In 1939 Dr. Paul Muller discovered the insecticidal properties of DDT (dichlorodiphenyltrichloroethan e), launching an era of intens ive research into pesticide development. This insecticide was so succe ssful in controlling mo squitoes and houseflies that Muller was awarded the Nobel Peace Prize for his discovery in 1948. During this time, DDT began to be used extensively in Fl orida for mosquito control. Concurrently, vector control agencies in the state were in creasing in number and effectiveness. The combined effect was to make large areas of Florida inhabitable. Today, Florida is the fourth most populous state in the union and one of the fastest growing, with a population approaching the combined total of Georgia, South Carolina and Alabama (U.S. Census Bureau, www.census.gov ). Heavy reliance on insecticides as a primar y means of mosquito control led to the development of many problems over the next 4 decades. Resistance and ensuing control failures among key nuisance mosquito species to several insecticide classes began developing in the 1950s and was widespr ead by the 1980s (Breaud 1993). Rachael Carsons Silent Spring (1962) alerted the public to some negative aspects associated with widespread broadcast applica tion of insecticides and herb icides and in general cast pesticides (and pesticide applicator s) in a bad light. Public co ncern turned to fear in some quarters and led to the formation of environmen tal groups intolerant of most, if not all, pesticide use. The costs associated with bringing a ne w pesticide to market has increased tremendously in the past 3 decades and is now approximately 60 million dollars per
38 compound ( http://www.epa.gov/pesticides ). Federal regulations impacting the cost of pesticide research and development include th e Safe Drinking Water Act, the Clean Air Act, the Clean Water Act, Department of Tr ansportation regulations, Occupational Safety and Health Administration regul ations, and many state regula tions. The additional cost of compliance with so many new regulations and increased litigati on brought about as a result of violations of some of these same regulations has forced many effective products off the market and led to a reduction of ne w products entering the market. Increased litigation has also curtailed certain mosquito control practices such as wetland ditching and placed restrictions on others (e.g., pest icide use next to or over bodies of water) (Florida Coordinating Council on Mosquito Cont rol 1998). It was under these conditions and during this time that the first comme rcial mosquito traps were developed. Since first introduced to the public in the late 1990s, a large variety of mosquito control traps have been developed for the co mmercial market. Manufacturers claim that their traps will keep populations of mosqu itoes below nuisance levels and protect as much as 1.25 acres of land per trap. On-site mosquito trapping offers homeowners many advantages over traditional ch emical control methods: a redu ction of biting insects (i.e., reduction of insect bites) on their immediate premises, insect control without exposure to pesticides, peace of mind that children and pets are not exposed to pesticides during ongoing control operations, continuous contro l of biting insects as opposed to periodic control offered by local mos quito control agencies, and the potential for reduced exposure to disease-carrying inse cts. In addition, many of these traps are easy to operate and work independently for 2 to 3 weeks at a time.
39 Residential and surveillance mosquito tr aps were evaluated in large outdoorscreened cage trials to determine their efficacy in trapping Ae. albopictus Several important parameters can be manipulated in large-cage trials including species composition, species age and numbers rele ased (Kline 1999). Control of these parameters enhances reproducibility and provi des for quantification between comparative tests. Additionally, assessment of a speci es performance under natural meteorological conditions is obtained through the use these cages (McDonald et al. 1978). Eight different trap models suppl emented with combinations of CO2 and octenol were tested to determine their efficacy in capturing Ae. albopictus Recommended trap height placement was modified in several trials to determine if height had any significance on capture rates. Octenol was test ed with several traps to assess effect on capture rates. Our test results were used to narrow trap select ion in field comparison trials (Chapter 4). Materials and Methods Large Outdoor-Screened Cages The large outdoor-screened cages used in these trials are located at the USDA ARS Gainesville, FL research facili ty. Cages are constructed of framed aluminum covered in black nylon screening similar to structures used to protect outdoor swimming pools from biting insects Two identical cages were used in our study, both are 9.2 m wide x 18.3 m long x 4.9 m high, peaking at 6.1 m along the le ngth of the cage ceiling (Kline 1999) (Fig. 2-1). A shelter (3 m wide x 3 m long x 2.5 m high) at the center of the cage provided trap protection from rain and direct sunlight. Landscaping was provided to resemble backya rd environments typical of southern U.S. households. The following shrubs and grasses are maintained in both cages: Azalea
40 ( Rhodendron hybrid ), Rhaphiolepis Indian Hawthorn ( Raphiolepis indica ), Ligustrum waxleaf ( Ligustrum lucidum ), Ilex dwarf burford holly ( Ilex cornuta "Burfordii"), Viburnum odortissuiumum Liriope giant ( Liriope muscari ), perennial ( Crossandra spp ), Golden dewdrop ( Duranta repens) False heather ( Cuphea hyssopifolia ), and Ilex schillings yaupon holly ( Ilex vomitoria Schilling Dwarf) (Ruide Xue, Anastasia Mosquito Control District, Di rector, personal communication). Trials were conducted during the warmer months of the year, from 19 May to 29 September 2003 and from 7 May to 17 July 2004. Traps were set between 0800 and 1200 and collected 24 h later. Meteorologi cal data (temperature, wind speed, and precipitation) was gathered from the Gainesville Regional Airport via the National Oceanic and Atmospheric Administrations National Climatic Data Center website ( http://www.ncdc.noaa. gov/servlets/ULCD ). The airport is approximately 7 miles from the test site. Trial dates in which the average hourly wind speed exceeded 10 miles per hour (mph) or temperatures dropped below 60oF (15.6oC) were excluded from analysis. Mosquitoes Colonized Ae. albopictus established from wild Ga inesville popul ations and maintained in the insect rearing facility at USDA ARS Gainesville were used in all trials (Ruide Xue, personal communi cation). Adults were aspirated from breeding cages, inactivated on a ch ill table at 42oF (5.6oC) for 15 min, sexed, counted, and placed into screened paper cups (Solo Cup Company, Urba na, IL). Mosquitoes were marked with fluorescent powders (BioQuip Products Inc., Rancho Dominguez, CA) by adding approximately 0.025 g of powder to each of 4 cups. Marking powder color was changed at every test to allow for accurate determinati on of release date. This prevented inclusion
41 of previous releases with ongoing trial counts. Female mosquitoes were equally distributed into cups and held at ro om temperature until they became active. Initially, 1,000 3to 6-day old adult females were released at every trial into the 4 corners of the cage. Midway through the fi rst summer, the release rate was reduced to 500 females per trial. An inability to m eet production demands in a timely fashion necessitated this change. Subs equently, landing rates and trap collections of early highrate releases (1,000 females) were halved to standardize results with the 500 female-pertrial release rates. Impact of trap efficacy against biting mos quitoes was determined within the cage at the completion of each trial. Biting adult females were collected with a mechanical flashlight aspirator (Hausherrs Machine Work s, Toms River, NJ) in collection tubes as they landed. Collecting was conducted for 3-mi n periods in all 4 co rners of the cage (6 min at each end) and for 6 min next to traps (for a total of 18 min). Separate collecting tubes were used at each collection site. Tr ap and landing collections were placed into a freezer to kill mosquitoes, which were la ter counted under UV light (fluorescent dye color-identified for correct count by day). Description of Traps Tested All traps were set according to manufacturer s directions. Two traps, the Centers for Disease Control and Prevention (CDC) light trap and the Wilton tra p, were also tested at reduced heights. The Mosquito Deleto 2200 was tested without octenol although it can be supplemented with an octenol cartridge Mosquito Magnet Pro, MM Liberty, and Mosquito Magnet-X (MM-X) traps were test ed with and without octenol. Carbon dioxide was provided to those tr aps not generating their own CO2. Flow rate was set at 500 mL/min for these traps.
42 Surveillance Traps CDC light trap. The CDC light tr ap (model 512, John W. Hock Company, Gainesville, FL) (Fig. 2-2a) us ed a 6 V DC motor and 4-blade fan to draw flying insects through an 8.5 cm diameter clear plastic cy linder body. A 36 cm diameter beveled edge aluminum lid was set approximately 3 cm above the cylinder body creating a downdraft air current. Nets or quart jars were att ached to the bottom of the trap to collect mosquitoes. The trap was set 150 cm (5 ft) above ground and used a 6.3V incandescent lamp (CM-47) as an attractant. The CDC li ght trap is compact, lightweight, portable and routinely used for mosquito surveillance in lo cations lacking AC elect ricity. The trap can be supplied with CO2 from dry ice or compressed gas to enhance mosquito capture. In these trials, CO2 was provided from a 9 kg compressed gas cylinder. A flow rate of 500 mL/min was achieved by using a 15-psi single-stage regula tor equipped with microregulators and an inline filter (Flowset 1, Clarke Mosquito Cont rol, Roselle, IL). Carbon dioxide was delivered to the trap th rough a 2 m long, 6.4 mm outer diameter clear plastic Tygon tubing (Saint-Gobain Performance Plastic, Akron, OH). Power was provided by a 6 V, 12 ampere-hour (A-h), rechargeable gel cell battery (Battery Wholesale Distributors, Georgetown, TX). Unless othe rwise noted, all CO2 equipment and batteries were identical and provided by these sources. Fay-Prince trap. The omni-directional FayPrince trap (hereafter, Fay-Prince trap) (model 112, John W. Hock Company, Gainesvi lle, FL) is a downdr aft suction trap designed specifically to capture Ae. aegypti (Fay and Prince 1970) (Fig. 2-2b). It captures flying insects from all directions around its perimeter and is an improvement over the original design that trapped from only 1 direction and th e later bi-directional trap. The Fay-Prince trap consists of 4 40.5 cm x 17.5 cm sheet metal arms set at 90o
43 angles to each other with a fa n at the center of the arms. Th e 4-blade fan is set in an 8.5 cm diameter black plastic cylinder to which a collecting net is attached. The fan is covered by a 40 cm2 rain shield set 10 cm above 4 vertical metal plates (27.5 cm x 6.5 cm) equally spaced from the center of the fan. The Fay-Prince trap makes use of contrasting black and white panels that se rve as a visual attr actant and can be supplemented with CO2 to enhance capture. It is set w ith the top of the cylinder 90 cm (3 ft) above ground and is powered by a 6 V rechar geable gel-cell batter y. It is bulky and heavier (2.7 kg) than most ot her portable mosquito traps. Wilton trap. The CDC Wilton trap (model 1912, John W. Hock Company, Gainesville, FL) is a downdraft suct ion trap designed to capture Ae. aegypti and Culex quinquefasciatus (Wilton and Kloter 1985) (Fig. 2-2c). It consists of a single 14.5 cm long x 8.5 cm diameter black cylinder that serves as a visual attractan t. Suction through the top of the trap is provided by a 4-blade fa n driven by a 6 V direct current (DC) motor. A rechargeable 6 V gel-cell battery powered the system. A white plastic collection cup with a stainless steel screen bottom and remova ble wire funnel is set inside the cylinder between the opening of the trap and the fa n; mosquitoes are trapped before passing through the fan. The wire funnel prevents mosq uito escape. The Wilton trap is set 90 cm (3 ft) above ground from its opening and is supplemented with CO2 to enhance capture. The trap is small, lightweight, and portable. Residential Traps Bugjammer biting insect trap. The Bugj ammer Home and Garden Unit (Applica Consumer Products, Inc., Miami Lakes, FL) (F ig. 2-2d) is an acous tic trap designed to catch hematophagous insects. The Bugjamme r is an adhesive, non-suction trap, which uses an embedded speaker to emit the record ing of a dog heartbeat and color contrast
44 (black and white) to attract mosquitoes to a white adhesive pad for capture. The adhesive pad surrounds a plastic resonato r cap that houses the speaker and protects it from rainfall. The Home and Garden unit is powered by 120V AC electricity whereas the Professional Unit relies on 4 D cell batteries to provid e 6 V DC power for operation. The batterypowered unit is supplemented with CO2 accessory equipment whereas the AC unit is not. The ground-mounted trap is mushroom shaped and stands 39 cm tall with a 10.5 cm base and a 17 cm diameter acoustic head that provides a trapping surface area of approximately 930 cm2. It is top-heavy and has a tende ncy to tip over in high winds, but offers a large visual target to mosquitoes. The Home and Garden Unit was used it our trials (without CO2, as intended by the manufacturer). Mosquito Deleto. The Mosquito Dele to 2200 System (The Coleman Company, Wichita, KS) is a non-suction adhesive trap that catalytically converts propane into CO2 at the rate of 170 mL to 200 mL/min to lure mo squitoes to the trap (F ig. 2-2e). As such, it is a stand-alone unit. The glossy black adhe sive strip serves as a visual attractant for mosquitoes and the trap can be supplemented with an octenol cartridge. The adhesive strip provides a surface area of approximately 750 cm2. A standard 9 kg barbecue grill propane tank is mounted directly beneath th e adhesive strip. The unit is ground-mounted and stands approximately 100 cm tall. The Mosquito Deleto 2200 is mounted on a metal frame equipped with wheels for movement. It is bulky and heavy (21.5 kg) when equipped with a 9 kg propane tank, but easy to move. Mosquito Magnet Pro. The Mosquito Magnet Pro Trap (MM Pro) (American Biophysics Corporation (ABC), North Kingsto wn, RI) (Fig. 2-2f) is a counterflow geometry (CFG) trap that catalyt ically converts propane into CO2, water vapor, and heat.
45 Counterflow geometry traps use opposing air curr ents to trap flying insects. An exhaust current of attractants is emitted below and away from an adjacent vacuum current that sucks mosquitoes and other flying insects into a trap chamber. The MM Pro uses propane to create attract ants and power the unit. Propane is burned until a catalyst is sufficien tly heated, at which time th e flame extinguishes itself. The heated catalyst continuously co nverts propane into water vapor, CO2, and heat. Heated water vapor and CO2 are exhausted to the outside of the trap, providing a plume of mosquito attractants. Catalytic heat is pr ovided to a second device adjacent to it, the Thermo Electric Module (TEM). A temperat ure gradient across the TEM (between the heated side and ambient temperature side) ge nerates electricity used to power 2 fans. One fan provides the exhaust plume of attrac tants, the other fan provides a counterflow updraft used to trap biting insects. Carbon dioxide is produ ced at the rate of approximately 520 mL/min (Karen McKenzie ABC, personal communication). Water vapor, CO2, and heat are known to be highly attractive to Ae. aegypti (Peterson and Brown 1951, Sipple and Brown 1953). The MM Pro is a stand-alone unit. The collection net is protected by a PVC shell m ounted on a black metal stand and base. The trap stands 100 cm high and is supported by an 84 cm x 56 cm stainless steel base equipped with wheels and a storage slot for a 9 kg propane tank. The unit is heavy (32 kg), bulky, and somewhat difficult to move over long distances, and should be shut down before being moved to prevent malfunctio ning of the TEM. The intake tube opening stands 52 cm above ground. Unlike CDC light trap s, the intake tube is oriented down and mosquitoes are drawn up into th e tube instead of down an in take tube (cylinder). The smaller exhaust tube protrudes 10 cm below th e center of the flared intake tube and
46 releases the plume at this point. The black e xhaust tube serves as a visual attractant and color-contrasts with the light gray or white intake tube. In addition to visual, heat, moisture, and CO2 attractants, the Mosquito Magnet Pro can be supplemented with an octenol or Lurex (lactic acid) cartridge, or both. Mosquito Magnet Liberty. The Mosquito Magnet Liberty (ABC, North Kingstown, RI) (Fig. 2-2g) is similar to the Mosquito Magn et Pro in form and function, but is smaller and powered by 120V AC electricity. A 12 V battery powers the unit in lieu of AC electricity. Any interruption to AC power shuts the unit off. In our study, the MM Liberty was powered by a 12 V rechargeable gel battery (SeaGel Deep Cycle Gel 31, West Marine, Watsonville, CA) during field trials for this reason. Water vapor, heat, and CO2 are provided by combustion of propane as described above and the unit can be supplemented with octenol, Lurex or both. Carbon dioxide is produ ced at the rate of 420 mL/min (Karen McKenzie, ABC, pers onal communication). This wheel-mounted ground unit is lighter (14.5 kg), more compact, and easier to transpor t than the MM Pro. It stands 84 cm high with the intake tube 54.5 cm above ground. The exhaust tube is seated in similar fashion as the MM Pro a nd is black. The head houses the motor and collection net and is supported by 2 black steel tubes (3 cm diameter) that may serve as additional visual attractants (personal obs ervation). Like the MM Pro, the MM Liberty offers a large visual target to host-seeking mosquitoes. Mosquito Magnet-X. The Mosquito Ma gnet-X (MM-X) (ABC, North Kingstown, RI) (Fig. 2-2h) is a CFG trap that produces suction and exhaust air currents in similar fashion to the MM Pro and MM Liberty. It requires an independent power and CO2 source as it does not use propane to generate these products. The MM-X is moderately
47 bulky (56 cm x 23 cm) but very lightwei ght (2.7 kg). It consists of an 80 mm intake fan, an oval-shaped clear PVC tr apping container (shell), a 40 mm exhaust fan, and exhaust and intake tubes. Screening inside the PV C shell allows air movement and prevents insect escape. Mosquitoes collected in this trap are well preserved and easy to identify. Contrasting black exhaust and white inta ke tubes provide vi sual attraction for mosquitoes. The unit is equipped with 3.2 mm x 6.4 mm ID/OD flexible vinyl tubing with quick connect Luer fittings that connect the end of the li ne with the trap head. An inline filter and flow control orifice provided CO2 at a rate of 500 mL/min from a compressed gas cylinder equipped with a 15-psi single-stage regulator. The unit can also be supplemented with dry ice as an alternate source of CO2. A 12 V rechargeable battery provided power. The MM-X can be supplem ented with octenol, Lurex or both. A description of attractant combina tions used in our trials is give n for each trap (Table 2-1). Trap results for all trials are included in Appendix A, Large-cage Trap Results. Statistical Analysis Traps were randomly assigned to either cage until an equal number of trials were completed in each cage. Six or 7 trials were run for each treatment. Data were transformed by SQRT ( n + 1) and subjected to GLM (for analysis of variance) (SAS Institute 2001). Comparisons of means were made using the Ryan-Einot-Gabriel-Welsh (REGW) multiple range test. Significance was set at 0.05. When necessary, comparisons between 2 individual treatm ents were conducted using 1-tailed t -tests ( = 0.05). Results Highly significant differences were found between trap capture means (F = 4.87, df = 12, p < 0.001) (Fig. 2-3). The MM Pro + octenol caught more adult female Ae.
48 albopictus than any other trap combination. Comp arable results were obtained with the Fay-Prince trap, the MM-X trap, and the MM Liberty + octenol; all 4 traps achieved trap means of over 100 Ae. albopictus In REGW multiple comparisons, the MM Pro + octenol trapped significantly more mosquitoes than the Wilton trap set at standard height, Bugjammer, and Mosquito Deleto traps (Fig. 23). Insignificant but noticeable trap mean differences existed between the MM Pro + octenol and remaining traps. All traps captured significantly more Ae. albopictus (p = 0.05) than the Bugjammer or the Mosquito Deleto (with the exception of the Wilton trap, standard height). Capture means of the 2 most successful tr aps were approximately equal (MM Pro + octenol and the Fay-Prince trap, 115 .3 and 113.5 .9, re spectively). Residential traps that produced their own CO2 outperformed surveillance traps supplemented with CO2 with the exception of the Mosquito De leto. Wilton and CDC light traps, both surveillance traps, collected mid-range means from 62.07 to 51.17. Wilton and CDC light traps were each tested with 2 tr eatments: trapping at the manufacturers recommended height and trapping at experiment al heights of 15 in to 20 in above ground. The CDC light trap performed equally well at both heights (standard height mean of 51.17, low height mean of 53.5), but low height Wilton traps collected 2.43 x as many Ae. albopictus as did Wilton traps set at recommended heights. Octenol-baited traps, with the exception of the MM-X, outperformed like traps not baited with octenol. The MM Pro + oc tenol trap caught significantly more Ae. albopictus (63.4%, t = -2.27, df = 10, = p = 0.046) than did th e octenol-free MM Pro. Trap means were 28.1% higher in octenol-b aited MM Liberty traps compar ed to unbaited traps, but
49 not significant. Capture means between octenol-baited and oc tenol-free MM-X traps were within 13% of each other. An assessment was made of the impact of trapping on biting activity. Biting collections were made at equal time intervals (3 min) in each corner of the cage and for 6 min next to the trap. Biting sums for all tr ials were compared against corresponding trap sums. Pearsons correlation coefficient was run comparing trap means to biting means. Data from these trials are shown in Fig. 24. A correlation coefficient of r = -0.13 (p = 0.26) was obtained by this method. This resu lt demonstrated no correlation between trap capture and biting activity (a high invers e correlation approaching -1.0 would have demonstrated trap effectivene ss in reducing biting activity). Discussion It is well known that Aedes mosquitoes (i.e., Ae. aegypti and Ae. albopictus ) do not respond well to standard mosqu ito surveillance traps such as the New Jersey light trap or CDC light trap (Thurman and Thurman 1955, Christophers 1960, Herbert et al. 1972, Service 1993, Jensen et al. 1994). In response to this problem, surveillance traps using black color and color contrast as attractants were developed to co llect these medically important pests (Sipple and Brown 1953, Fay 1968, Fay and Prince 1970, Wilton and Klotter 1985, Freier and Francy 1991). Th e Fay-Prince and Wilton trap, which take advantage of black and/or contrasting colors have compared favorably against standard surveillance traps in multi-trap comparison studi es (Klotter et al. 1983, Jensen et al. 1994, Burkett et al. 2004, Dennett et al. 2004) and are currently the traps of choice for surveillance efforts targeting adult Ae. albopictus Within the past 10 years, many new residen tial mosquito traps have been developed for homeowner and business premises use. In itial reports indicate that some of them
50 work very well in trapping Ae. albopictus (Sean Bedard, ABC, personal information). In addition, many of these traps can be supplemente d with octenol and/or lactic acid baits, known mosquito attractants (Acree et al. 1968, Kline et al. 1990). Surveillance and residential traps were tested in large-cage trials in an attempt to determine which would best serve as surveillance and/or control de vices in suburban settings. Trap features contributing to success are discussed below and listed in Table 2-1. Significant differences (F = 4.87, df = 12, p < 0.001) were seen between capture means from 8 traps comprising 13 treatments. Traps broke out into 3 groups based on performance: high, average, and low. Trap means categories were derived from REGW test results. High performance traps incl uded those with capture means above 90 and included the MM Pro + octenol (115), the Fay-Prince trap (113.5), the MM-X (107.8), the MM Liberty + octenol (103.6), and the MM-X + octenol (93.8). Average performance traps collection means ranged fr om 30 to 89 and included the MM Liberty (74.4), the MM Pro (63.8), the low height Wilto n trap (62.1), the low height CDC light trap (53.5), and the CDC light trap (51.1). Low performance trap means of less than 30 included the Wilton trap (25.6), the Bugjamme r (2.8), and the Mosquito Deleto model 2200 (0.8). High performance traps have several charac teristics in common. First, none of them use incandescent light as an attr actant (MM Pro, MM Liberty, MM-X, Fay-Prince trap). In published trap comparison studi es, unlit traps outperformed lit traps in collecting Ae. albopictus (Herbert et al. 1972, Dennett et al 2004 and agree with results of our large-cage studies. Second, the MM Pro, MM Libert y, and MM-X traps are all CFG updraft traps. Other traps are downdraft traps or rely on adhesive paper to trap
51 mosquitoes (an exception to this observation was the high performance Fay Prince trap, a downdraft trap). K line (1999) trapped 1.7 x more Oc. taeniorhynchus in updraft traps (MM-X) than in downdraft traps (ABC Pro) in large-cage trials. Updraft CFG traps produce a CO2 plume near the trap intake and mos quitoes are trapped as they approach and enter the plume. Do wndraft traps blow the CO2 plume below and away from the trap intake. These trap operational differences may play an important role in capture totals. Third, some CFG traps produce heat ( MM Pro, MM Liberty), a known mosquito attractant (Howlett 1910, Peterson and Brow n 1951) that can significantly increase mosquito capture in the presence of CO2 (Kline and Lemire 1995). Fourth, water vapor is produced in propane-burning traps (MM Pr o, MM Liberty). Moisture in the presence of heat increases the attr activeness of targets to Ae. aegypti in laboratory settings (Brown et al. 1951) and to Aedes mosquitoes in the field (provi ded air temperature is above 60o F) (Brown 1951). Fifth, all 3 traps use contrast ing black and white colo r as an attractant at the trap intake. This color scheme is a proven attractant for Aedes mosquitoes (Brett 1938, Sipple and Brown 1953, Haufe 1964). Average performance trap capture means ranged from 30 to 89 Ae. albopictus These included the unbaited (no octenol) MM Liberty (74.4), unbaite d MM Pro (64.8), low height Wilton trap (62.1) low height CDC light trap (53.5), and CDC light trap (51.1). Of these traps, CFG traps (MM Li berty, MM Pro, MM-X) outperformed standard downdraft surveillance traps (w ith the exception of the FayPrince trap), although capture mean differences were not significant. The CDC light trap, Fay-Prince trap, and Wilton trap are hung from a line and are capable of some movement, both rotationally and laterally, due to wind disturbance. Movement is a proven attractant for Ae. aegypti
52 (Sipple and Brown 1953, Brown 1956, Kusakabe and Ikeshoji 1990), however, we do not know the effect of incidental movement in attracting host-seeking mosquitoes. An experimental movement trap showed satisfa ctory results in a mu lti-trap comparison study targeting Ae. albopictus but was outperformed by severa l stationary residential and surveillance traps (D ennett et al. 2004). Both MM Pro and MM Liberty traps produced better results baited with octenol. The octenol-baited MM Pro caught more (1.77 x ) Ae. albopictus than unbaited traps and octenol-baited MM Liberty traps caught 28.1% more Ae. albopictus than unbaited traps. Octenol is known to increase trap capture for many Aedes ( Ochlerotatus ) mosquitoes in the presence of CO2 (Kline et al. 1991a, Kline et al. 1991b, Kline 1994b, Kline and Mann 1998). In contrast to our large-cage resu lts, unpublished field results (Sean Bedard, ABC, personal communi cation) indicate that Ae. albopictus is slightly repelled by octenol in the presence of CO2. Comparison trials between octe nol-baited and co ntrol (octenolfree) MM Pros (Chapter 4) show ed that octenol had a small, but insignificant, depressive effect on Ae. albopictus collections. Shone et al. (2003) found no differences in Ae. albopictus catches between octenol-baited and unba ited Fay-Prince traps (in the presence of CO2). It appears that octenol-baited traps targeting other mosquito species will not adversely affect Ae. albopictus collections, but traps set specifically for Ae. albopictus should probably avoid octenol baits, as octenol can depress trap counts. Trap height may be an important to Ae. albopictus capture. Two traps, the CDC light trap and the Wilton tra p, were tested at recommende d heights (5 ft and 3 ft, respectively) and at lower heights with tr ap intake set 15 in to 20 in above ground. Aedes albopictus flies close to the ground and prefers to bite on the lower extremities, such as
53 around the ankle and lower legs (Bonnet and Worcester 1946 Watson 1967, Shirai et al. 2002). No significant difference was seen be tween CDC light trap means at different heights, but a significant difference ( t = 2.31, df, = 10 p = 0.04) was seen in Wilton trap capture means (trap means of 25.6 at standard height and 62.1 at low height). The CDC light trap is generally not very effective in collecting Ae. albopictus whereas the Wilton trap, using highly attractive black, was desi gned specifically for these mosquitoes. Wilton traps set close to the ground takes a dvantage of 2 behavioral attributes of Ae. albopictus : near-ground level resting and biting preference, and the fact that this mosquito is a weak flier which quickly settles on ground or lowlying vegetation under mild wind conditions (Bonnet and Worcester 1946, Hawley 1988). The results of these trials indicate that a further investigati on into trap height placement is warranted, especially with some of the newer residential models that are highly effective in trapping Ae. albopictus We suspect that decreas ing the intake height of several models by 50%, especially the MM Pro and MM Libert y, would enhance capture rates. Low performance traps (with means < 30) include the Wilton trap set at the recommended height of 3 ft (25.6), the Bugj ammer Home and Garden Unit (2.8) and the Mosquito Deleto 2200 (0.8). It is apparent fr om Table 2-1 that as the number of trap features decrease, so does trap performance (the Fay-Prince trap is an exception to the rule). The small visual target Wilton traps offer mosquitoes may be one reason it did not perform as well as the CDC li ght trap. Although painted an attractive black, this trap lacks color contrast typical of most other trap s. However, Wilton traps set at low heights (15 in to 20 in) outperformed all low performa nce traps. Thus, trap height setting is
54 probably an important factor with respect to Wilton trap performance in targeting Ae. albopictus Sound traps offer an interesting alternative to more traditional traps that use other key attractants. Most work with sound traps has occurred in those pa rts of Asia to which Ae. albopictus is native. Experimental sound trap s mimic wingbeat frequencies within the range of adult mosquitoes 300 Hz to 800 Hz (Moore et al. 1986) producing a buzzing sound similar to that of flying adults. Ikes hoji and Yap (1990) used insecticide-treated sound traps to reduce Ae. albopictus populations in a Malaysian village by almost 90%. Interestingly, adult males were attracted to the traps as well and a 75% reduction was achieved. Kanda et al. (1987) used dry icebaited sound traps set at 350 Hz and 480 Hz. The 350 Hz trap was effective in capturing Mansonia mosquitoes and at 480 Hz larger numbers of Ae. albopictus males were caught. Culex mosquitoes have also been lured to sound traps operated at about 400 Hz and 100 de cibels (dB) (Ikeshoji et al. 1985, Ikeshoji and Ogawa 1988). Most sound traps use adhesive surfaces to capture mosquitoes (Kanda et al. 1987), others use insecticide-coated surfac es (Ikeshoji and Yap 1990). The Bugjammer Lawn and Garden Unit uses sound to attract flying insect to an adhesive paper. As tested, it was not supplemented with CO2 (per manufacturers direction). Capture rates were so low that trapping was discontinued after 3 trials. The traps adhesive paper remained sticky th roughout all trials. Th e Bugjammer relies on glossy black and white contrasting colors as an additional attractant. Sound emitted from the trap mimics dog heartbeat, not the wingbeat frequency of experime ntal sound traps. This trap has apparently ach ieved good results in trapping Ae. aegypti when
55 supplemented with CO2 (Dr. Kline, personal comm unication), but operating on sound alone, it was ineffective in trapping Ae. albopictus The Mosquito Deleto, the only other adhesive trap used in our study, produces heat and CO2 from propane combustion. Glossy black adhesive paper provides an additional attractant. These traps produce CO2 at a rate of 170 mL/min to 200 mL/min. This trap was ineffective in collecting Ae. albopictus and tests were disconti nued after 3 trials. The adhesive surface remained tacky during all tr ials and it was assumed that adhesiveness was not a factor. Low performance traps share 2 common feat ures. They produce little or no CO2 as compared with other traps. The Bugjam mer Home and Garden Unit was the only unit tested without CO2 (per instructions). The Mosquito Deleto produced CO2 at rates less than half that of other traps. Most researchers agree that CO2 is the most important chemical attractant to mosquitoes and that wi thout it, trap capture rates plummet (Lehane 1991, Service 1993, Mboera and Takken 1997). Th e second feature of these 2 traps is that they rely on adhesive surfaces to capture insects. Such surfaces apparently do not work well with Ae. albopictus although good results have been obtained with Ae. aegypti and Psorophora columbiae (Dyar and Knab) ( www.bugjammer.com ). Residential CFG traps outperformed downdr aft surveillance traps in collecting Ae. albopictus The lone exception was the Fay-Prince trap, which performed nearly as well as the top-performing MM Pro + octenol. The addition of octenol to traps increased trap capture in 2 of 3 cases, 1 significantly (MM Pr o). However, field studies (Shone et al. 2003, Sean Bedard, ABC, personal communicati on, field trial results of Chapter 3) indicate that octenol-baited traps lightly depress Ae. albopictus capture rates, although
56 not significantly. Trap height can impact capture rates. Th e Wilton trap caught 3 x as many Ae. albopictus set 15 in to 20 in above ground as opposed to traps set at recommended heights, but this trend was not seen with CDC light traps. Adhesive traps do not appear to work very well in trapping Ae. albopictus (Mosquito Deleto, Bugjammer biting insect trap ). An adhesive tape experimental sound trap mimicking the wingbeat-frequency bu zz of mosquito flight was effective in collecting Ae. albopictus (when supplemented with CO2) (Ikeshoji and Yap 1990), but the Bugjammer biting insect trap, mimicki ng dog heartbeat, did not work well (although it was tested without CO2). Our studies indicate that resi dential CFG traps can perfor m as well, if not better, than standard surveillance traps in collecting Ae. albopictus With the ex ception of the Fay-Prince trap, CFG residentia l traps could effectively replace surveillance traps in Ae. albopictus surveillance programs. Benefits of pr opane-powered residential traps include increased catches, long-term operation (3 week s vs. 1 or 2 days), stand-alone operation, and monetary savings associated with reduced trap attendance. Trap features highly attractive to Ae. albopictus include CO2, contrasting color, black, heat, and water vapor. Design of new traps targeting Ae. albopictus should incorporate these features.
57Table 2-1. Trap attracta nt features used in Aedes albopictus large-cage trials. Trap Performance CO2 Octenol2 Vapor3Heat Sound Light Contrast4 Black5 Height6 CFG7 Sum Performance average Pro + O High Yes Yes Yes Yes No No Yes Yes No Yes 7 Fay-P. High Yes1 No No No No No Yes Yes No No 3 MM-X High Yes1 No No No No No Yes Yes No Yes 4 Lib + O High Yes Yes Yes Yes No No Yes Yes No Yes 7 MM-X + O High Yes1 Yes No No No No Yes Yes No Yes 5 5.2 Liberty Average Yes No Yes Y es No No Yes Yes No Yes 6 Pro Average Yes No Yes Yes No No Yes Yes No Yes 6 Wilt Low Average Yes1 No No No No No No Yes Yes No 3 CDC Low Average Yes1 No No No No Yes No Yes Yes No 4 CDC Average Yes1 No No No No Yes No Yes No No 3 4.4 Wilton Low Yes1 No No No No No No Yes No No 2 Bugjammer Low No No No No Yes No Yes Yes No No 3 Deleto Low Yes No No Yes No No No Yes No No 3 2.7 1Source of CO2, gas cylinder (500 mL/min). 2Octenol cartridges from Amer ican Biophysics Corporation. 3Trap produces water vapor. 4Trap has contrasting black and white colors. 5Trap hung close to ground vs manufacturers recomme ndation (15 in to 20 in off ground). 6Low height setting advantageous for trapping Ae. albopictus 7CFG = counter flow geometry.
58 Figure 2-1. Large outdoor screened cages used in trap efficacy trials, USDA ARS Gainesville.
59 A B C D E F Figure 2-2. Traps tested in la rge-cage efficacy trials with Aedes albopictus : A) CDC model 512 trap. B) omni-directional Fay-Prince trap. C) Wilton trap. D) Bugjammer Home and Garden Unit. E) Mosquito Deleto 2200 System. F) Mosquito Magnet Pro trap. G) Mosquito Magnet Liberty trap. H) Mosquito Magnet-X (MM-X).
60 G H Figure 2-2. Continued a ab ab ab ab c c bc ab ab ab ab ab 0 20 40 60 80 100 120 140Pro ProO Lib LibO Wil WilL CDC CDCL Fay MMX MMXO Bug DelTrapMean Aedes albopictus capture Figure 2-3. Large-cage Aedes albopictus trap capture means in residential and surveillance traps. Multiple comp arisons (Ryan-Einot-Gabriel-Welsh multiple range test) were performed after SQRT (n + 1) transformation. Means within each treatment having th e same letter are not significantly different ( =0.05, n=6 or 7 trap days). Pro = MM Pro, Lib = MM Liberty, Wil = Wilton trap, Bug = Bugjammer, O = octenol baited, L = low height trap.
61 0.00 20.00 40.00 60.00 80.00 100.00 120.00Activit y Pro ProO Lib LibO Wilt WiltL CDC CDC-L Fay MMX MMXO Bug Deleto Trap mean Bite mean Figure 2-4. Large-cage trap capture and biting means (tot al catch or bites/number of trials) of Ae. albopictus. Pearsons correlation coe fficient of r = -0.13 was achieved for all trials (n = 6 or 7). Pro = MM Pro, Lib = MM Liberty, Wilt = Wilton trap, Bug = Bugjammer, O = oc tenol baited, L = low height trap.
62 CHAPTER 3 FIELD EVALUATION OF CARBON DIOX IDE, 1-OCTEN-3-OL, AND LACTIC ACID-BAITED MOSQUITO MAGNET PRO TRAPS AS ATTRACTANTS FOR Aedes albopictus IN NORTH CENTRAL FLORIDA Introduction Numerous chemical compounds have been sc reened as attractants for biting adult female mosquitoes, but few stand as strong potential candidates base d on laboratory and field test results. Among the more successful compounds are CO2 (Rudolfs 1922, Gillies 1980, Mboera and Takken 1997), 1octen-3-ol (octenol) (Kline et al. 1991a, Kline et al. 1991b, Kline 1994b, Kline and Mann 1998), lactic acid (Acree et al. 1968, Kline et al. 1990, Bernier et al. 2003), phenols (Kline et al. 1990), butanone (K line et al. 1990), acetic acid (Vale and Hall 1985) and several amino acids (Brown and Carmichael 1961, Roessler and Brown 1964). They are derived, in part, from the physiological processes of respiration, perspiration, and waste elimination in mammals, birds, and reptiles. Attractants commercially developed for use w ith mosquito traps in clude carbon dioxide, octenol, and lactic acid. Carbon dioxide is considered a universa l attractant for hematophagous insects, especially mosquitoes (Kline 1994a). It was one of the first compounds shown to attract mosquitoes (Rudolfs 1922) and is used extens ively to boost capture rates in many field studies (Gillies 1980). The volatile compound octenol is a component of breath in ruminants (oxen and cattle) (H all et al. 1984) and is also produced by invertebrates, fungi, and some plants such as clover a nd alfalfa (Kline 1994b). Octenol was first recognized as an attractant after it was discovered to grea tly increase trap capture of
63 tsetse flies in East Africa (V ale and Hall 1985). Earl y field tests identifi ed octenol as an attractant alone or in combination with CO2 for certain mosquito species of Ochlerotatus, Aedes, Anopheles, Psorophora, Coquillettidia, Mansonia, and Wyeomyia (Takken and Kline 1989, Kline et al. 1991a Kline 1994b). Lactic acid, a component of sweat, increased trap capture rates of Ochlerotatus, Anopheles, and Culex mosquitoes when blended with CO2 (Stryker and Young 1970, Kline et al. 1990). Octenol and Lurex (L(+)-lactic acid) are commercially av ailable as mosquito trap supplements (American Biophysics Corporation (ABC), North Kingstown, RI). Data are lacking as to Ae. albopictuss response to these compounds in the field. Two published octenol studies targeting Ae. albopictus in the United States exis t. Shone et al. (2003) used Fay-Prince traps supplemented with CO2 and octenol to capture Ae. albopictus in Maryland. They found that traps baited with CO2 alone or in combination with octenol attracted significantly more Ae. albopictus than traps lacking CO2 or baited only with octenol. No significant diffe rence in trap capture was seen between octenol + CO2and CO2-baited traps. Dennett et al (2004) obtained superior re sults from an octenol-baited MM Liberty trap as opposed to ot her unbaited traps in collecting Ae. albopictus from a tire repository in Housto n, Texas. Conversely, an octe nol-baited Mosquito Deleto caught the smallest number of Ae. albopictus in that study. Preliminary field observations indicate that octenol may have a small depressive effect on capture rates of Ae. albopictus (Sean Bedard, ABC, personal communication), however, this is not supported by the above-mentioned tests. No published data exists as to the effec tiveness of lactic acid (Lurex) as an attractant for Ae. albopictus. Lurex became commercially available for the first time in
64 2004. Earlier studies with Aedes mosquitoes and lactic acid involved Ae. aegypti, not Ae. albopictus. The attractancy or repellency of lactic acid to Ae. aegypti is concentration dependent. Two types of l actic acid-sensitive receptor s exist on antennal grooved-peg sensilla of this mosquito, 1 type shows an increase in spike frequency in the presence of lactic acid, the other type s hows a decrease in spike frequenc y at the same concentrations (Davis and Sokolove 1976). Slow release rates of lactic acid from Lurex cartridges are believed to enhance capture of Ae. albopictus in CO2-baited traps (Alan Grant, ABC, personal communication). We attempted to de termine the efficacy of traps baited with octenol, Lurex, octenol + Lurex, or neither in the presence of CO2 in capturing Ae. albopictus. The effects of these treatments on ot her mosquito species trapped during our study are included. Materials and Methods Trap Placement and Rotation Three field trials were conduc ted using the Mosquito Magn et Pro (MM Pro) trap in 4 separate suburban neighborhoods in Gainesvi lle, Florida. Trapping occurred from 1321 Aug., 25 Aug.-2 Sept., and 9-17 Sept. 2003. Four test locations were selected based on homeowner complaints of nuisance populatio ns of biting mosquitoes. Two of the 4 sites were owned by professional ento mologists knowledgeable in mosquito identification and aware of nuisance populations of Ae. albopictus on their properties; initial surveys showed the presence of Ae. albopictus at all test locations. All sites consisted of a mix of pine and hardwood trees with minor amounts of undergrowth. One site was planted extensively in Neoregelra (red finger nail), Bilbergia pyramidalis, and Bilbergia spp. bromeliads, a second site had lesser numbers of Neoregelra bromeliads. Tank bromeliads are excellent breeding sources for Ae. albopictus (OMeara et al.
65 1995b). Traps were placed in shaded areas under trees or just inside a tree line next to open spaces. The Mosquito Magnet Pro trap was chosen based on positive coll ection results with Ae. albopictus in previous field inves tigations and large-cage trials (Chapter 2). The MM Pro uses counterflow geometry technology (CFG) and produces warm, moist air, and CO2 at the rate of approximately 520 mL/min (Karen McKenzie, ABC, personal communication). This combination of attracta nts mimics animal breath and has made the MM Pro a very effective mosquito trap. Traps were set between 0800 and 1200 and collected approximately 48 h later (1 trapping period). Aedes albopictus is diurnally active with a bimodal feeding habit during daylight hours (Watson 1967, Hawley 1988). The 48 h trapping interval insured uninterrupted trapping through 1 continuous 24 h period to take advant age of this feeding rhythm. Trap rotation took from 3 to 4 h be tween sites and was randomized. The entire trap was moved as opposed to switching attr actants between stati onary traps. This eliminated the possibility of residual odors biasing trap performa nce across treatments (octenol leaves a noticeable odor on equipmen t it has been in contact with for several days). Preventive maintenance was performed at the start of each trial by clearing the combustion chamber with compressed CO2 gas per recommendation of the MM Pro operations manual and by physically removing debris from the inside of the trap collecting tube and net chamber with a moist sponge. Attractants Treatments included an u nbaited trap (control, CO2 only), a trap baited with an octenol cartridge only, a trap baited with a Lurex cartridge only, and a trap baited with octenol + Lurex cartridges. Octenol a nd Lurex cartridges are manufactured by ABC
66 and are designed for use with the Mosquito Magnet Pro, Liberty, and MM-X traps. Lurex cartridges contain 4.88 g lactic acid embedded into a 13.8 g clear gelatin matrix sealed in a plastic package, which slowly rele ases lactic acid over 3 weeks at an average of 0.23 g/day ( http://www.mosquitomagnet.com ). This low lactic acid release rate is apparently below the repellency threshold of Ae. albopictus (Davis and Sokolove 1976). Octenol cartridges slowly release 1.66 g oc tenol from a microporous polyethylene block over 3 weeks at 80oF (26.7oC). The block is encased in a porous plastic package. All cartridges were replaced at the beginni ng of each trial (after 8 days of use). Trap efficacy was assessed with biting counts of adult female Ae. albopictus on most collection days. Biting mosquitoes we re collected for 3 min before traps were rotated or removed from the fi eld. Mosquitoes were coll ected as they landed using a hand-held flashlight aspirator equipped w ith a collecting tube (Hausherrs Machine Works, Toms River, N.J.). All biting collectio ns were made within a 1-acre radius of the trap (within 36 m of the trap). Mo squitoes were anesthetized with CO2 and transferred to labeled paper cups (Solo Cup Company, Urba na, IL) for later identification to species using the keys of Darsie and Morris (2000). Statistical Analysis A 4 x 4 Latin square design was used for each of the 3 trials. A requirement of the Latin square design is that each treatmen t combination be set at each location per collection period without repe ating that treatment combination. This removed differences among rows (days) and columns (t rap locations) from the experimental error for a more accurate analysis of treatment co mbinations on mosquito collections. Trap rotation was randomized between trials. Aedes albopictus collections were analyzed for treatment, site, and period (= 48 h) effect using a 3-way ANOVA (SAS Institute, 2001).
67 The Ryan-Einot-Gabriel-Welsh (REGW) multip le range test was used to determine significant difference between treatments ( = 0.05). Other mosquito species were then analyzed for treatment, site, and peri od effects using a 3-way ANOVA and multiple comparisons with the REGW multiple range test ( = 0.05). All capture data were transformed with log10 (n + 1) prior to analysis. Results Aedes albopictus Twelve trap periods (48 h each) over 3 trials resulted in a catch of 6,787 mosquitoes. Aedes albopictus comprised 19.5% of the total catch (1,321 adults). Table 3-1 shows total adult Ae. albopictus capture by treatment and trial for each test period. Significant differences were observed in Ae. albopictus collection totals between treatments (F = 3.44, df = 3, p = 0.029) and co llection sites (F = 5.69, df = 3, p = 0.003). The sum response of Ae. albopictus to treatment was octenol + Lurex + > Lurex > control > octenol. Octenol + Lurex-b aited traps caught significantly more Ae. albopictus than did octenol-baited traps and noti ceably more than either Lurex-baited or control traps (Table 3-1). The collection site planted extensively in tank bromeliads produced 40.2% (531) of all Ae. albopictus adults collected among the 4 sites. A review of trap data over 3 trials indi cates that very similar sex ratios were trapped across all treatments dur ing our study (Table 3-2). Overall, an average of 2.74 females were caught for every male during the study period. Aedes albopictus adults collected from the first 2 trials produced near ly identical female: ma le ratios of 2.61:1 and 2.64:1 (trials 1 and 2, respectively). On ly during trial 3 did this ratio increase to 3.21:1. The author has no Ae. aegypti trapping experience to compare with these ratios, however, extensive trapping and surveillance with CDC light traps in Texas and at 2 rural
68 sites in Alachua County, Florida typically produc e few, if any, male mosquitoes of other species common to the southern United States A detailed look into sex ratios under treatment for each trial indicat e that that octenol, Lurex, and octenol + Lurex had similar collection ratios ranging from 2.21:1 to 2.80:1 (female: male). The exception was the control, which had the highest fema le: male ratio (4.02:1) (Table 3-3). A highly significant difference (F = 12.29, df = 2, p < 0.0001) was seen in the number of Ae. albopictus adults captured between trials (T able 3-1). Trial 1 totaled 741 adults, trial 2 totaled 277 adults, and trial 3 totaled 303 adults. The trial 1 total was 62.1% greater than trial 2 and 59.1% greater than trial 3. These data indicate a possible suppressive effect of the MM Pro trapping of Ae. albopictus populations over time. Aedes albopictus biting activity during this these trials remained stable. Three-minute biting collection totals of Ae. albopictus from trap sites on most collection days during these trials were 65, 81, and 61 for trials 1, 2, and 3, respectively. Other Mosquito Species Other mosquito species collected from our 4 suburban sites included Anopheles crucians s.l., Culex erraticus (Dyar and Knab), Cx. nigripalpus Theobald, Ochlerotatus atlanticus (Dyar and Knab), Oc. infirmatus (Dyar and Knab), Oc. triseriatus (Say), Psorophora ferox (von Humboldt), Ps. columbiae, Coquillettidia perturbans (Walker), Wyeomyia mitchellii (Theobald), and other Culex and Psorophora mosquitoes unidentifiable to species. Table 3-4 show s treatment collection means for each species collected from our tests. Significant differences between treatments were found in 4 of the 11 species listed in Table 3-4, including Ae. albopictus. Significantly more Cx. nigripalpus (F = 14.89, df = 3, p = 0.0001) were caught in octenol-baite d and unbaited traps than in Lurex-baited
69 and octenol + Lurex-baited traps. Similar treatment results were seen with Oc. infirmatus (F = 8.41, df = 3, p = 0.0003) and Ps. ferox (F = 10.95, df = 3, p = 0.001). In descending order, Culex nigripalpus, Ae. albopictus, Oc. infirmatus, Ps. ferox, and Cx. erraticus were the 5 most abundant mosquitoes collected (Table 3-5). Trap totals for each species are given in Fig. 3-2. Many Psorophora mosquitoes (188) were excluded from analysis as specimens were damaged to an extent that species identification was not possible. It is likely that most of these were Ps. ferox; few other Psorophora mosquitoes were collected from these sites (Ps. columbiae 24, Ps. howardii Coquillett 17, and Ps. ciliata Fabricius 1). Excluding Ae. albopictus, a total of 27 male mosquitoes were collected during all trials a tiny proportion (0.5%) of the other 5,411 mosquitoes caught in our study. Discussion Aedes albopictus It is believed that Aedes albopictus became established in North America sometime during the early 1980s. Although several accide ntal introductions were previously intercepted in retrograde cargo returning from the Pacific war theater (Pratt et al. 1946) and Vietnam (Eads 1972), this species remained absent from the United States until the 1980s. It was discovered in over 40 locations in and around Houston, Texas in August 1985 (Sprenger and Wuithiranyagool 1986). It no w occurs in 1,035 counties in 32 states, inhabiting southern, Mid-Atla ntic, and Mid-Western states (Janet McAllister, CDC, personal communication). It may have just recently become established in parts of California as well (Linthicum et al. 2003). Aedes albopictus has received a lot of media atte ntion since its arrival, and for several good reasons. First, it is well adapted for breeding in artificial containers in and
70 around households. It can quickly establish large populations and is well suited for colonizing rural, forested, a nd unpopulated areas. Hence, once established in an area, it is extremely difficult to control. Second, it thrives in temperate and tropical climates such as those from which it came Asia (Watson 1967, Hawley 1988), implying that it may not yet be finished with its expansion within the United States. Third, it is a competent vector of many dis ease agents of public health co ncern, mainly dengue viruses (Sabin 1952, Calisher et al. 1981, Qiu et al. 1981), and has already been found infected with eastern equine encephalitis virus (Mitchell et al. 1992), West Nile virus (Holick et al. 2002), and La Crosse ence phalitis virus (Gerhardt et al. 2001) within the United States. Fourth, it is not readily attracted to standard mosquito surveillance light traps used by vector control personnel to m onitor mosquito populations (Thurman and Thurman 1955, Service 1993). Finally, it is extremely difficult to control with conventional adulticides because it is diurnally active. Ultra low volume insecticide applications are most effec tive at dusk and night, when at mospheric inversion conditions prevail. Several residential mosquito traps have produced good initial results in capturing Ae. albopictus (Smith et al. 2002, Smith and Walsh 2003, Dennett et al. 2004). Along with their recent availability to homeowners is the addition of 2 mosquito attractants marketed for these machines, specifically, octeno l and lactic acid baits. We showed that octenol alone had a slightly depres sive, though not significant, effect on Ae. albopictus capture rates compared with unbait ed traps (in the presence of CO2). Except for 2 results (octenol vs. control, trial 2 and octenol vs Lurex, trial 3), octenol-baited traps were outperformed by all other treatments during the entire study (Table 3-1). These data
71 agree with field observations in Hawaii in which octenol-baited MM Pro traps generally collected fewer Ae. albopictus than unbaited traps (Sean Bedard, ABC, personal communication). Dennett et al. ( 2004) trapped significantly more Ae. albopictus from an octenol-baited MM Liberty trap than any of 6 other trap types, but did not use an unbaited MM Liberty for comparison. Few comparison trials have been published comparing Ae. albopictus capture in octenol-baited and unbaited residential traps. No octenol comparison trials w ith the MM Pro were found. Large-cage trials (Chapter 2) produce d opposite results from those noted above; octenol-baited MM Pros caught 1.77 x more Ae. albopictus than did unbaited MM Pros. The reasons for this are unclear, but trap la rge-cage results may have been biased due to cloud cover and relative hu midity on test days. Aedes albopictus is a day-biter and prefers shaded to sunlit areas (Watson 1967). Bright, clear days would have minimized shade in the cages, adversely impacting Ae. albopictus host-seeking activity. Traps set in this field study were always positioned in shad ed places at all collec tion sites. Overall, there was no significant difference between octenol-baited and c ontrol trap means (t = 0.38, df, = 24, p = 0.71). Certainly, more comparison studies are needed between like traps baited with and without octenol. It should be noted that using the highly effective Fay Prince trap, Shone et al. ( 2003) caught similar numbers of Ae. albopictus with octenol-baited and unbaited tr aps in the presence of CO2. Results of our large-cage and field trials indicate that octenol-baited tr aps are just as effective in trapping Ae. albopictus as unbaited traps in the presence of CO2. Lactic acid bait has only recently become available for purchase and use with residential traps. It is a known attractant to Ae. aegypti (Acree et al. 1968), which have
72 lactic acid-sensitive rece ptors on their antennae (Davis and Sokolove 1976). In laboratory olfactometer studies, lactic acid + acetone blends we re as attractive as lactic acid + CO2 blends in attracting this mosquito; it works well at low release rates when combined with certain other mosquito attr actants (Bernier et al. 2003). Few published reports exist concerning the effects of lactic acid on Ae. albopictus. In 1 Japanese field study, lactic acid-baited traps were no more attractive to Ae. albopictus than unbaited traps (Kusakabe and Ikeshoji 1990), although no mention was made of release rates. At high skin surface concentrations (> 41.7 ppm), lactic acid was repellent to Ae. albopictus (Shirai et al. 2001). Apart from these studies nothing else was found in the literature review of lactic acid and Ae. albopictus. In our study, MM Pro traps baited with Lurex collected more Ae. albopictus than did octenol-baited traps or control traps. Lurex trap totals from trials 1 and 2 were superior to octenol and control tr eatments, however, in trial 3, Lurex-baited traps caught less Ae. albopictus than any other treatment. Trial 1 Lurex results were the second best throughout all trials (199 adults). The control treatment consis ted of an unbaited MM Pro tr ap, this trap used only CO2 as an attractant. Although the unbaited trap outperformed the octenol-baited trap, difference in trap capture was minor. Usi ng CDC light traps, Vythilingam et al. (1992) collected twice as many Ae. albopictus in CO2-baited traps than in octenol + CO2-baited traps, although trap totals were small and trap mean differences were not significant. No octenol flow rate was given. In our study, a large gap was noted between control and Lurex capture means, with Lurex-baited traps collecting 25% more Ae. albopictus than control traps. This difference was not significant.
73 Octenol + Lurex-baited traps achieved superior results, capturing almost 2.5 x as many adult Ae. albopictus as octenol-baited traps, 2.1 x more adults than control traps, and 1.6 x more adults than Lurex-baited traps. Our results agree well with previous findings in which blends of 2 or 3 attracta nts were shown to work better than just 1 attractant alone (Gillies 1980, Kline et al. 1990, Lehane 199 1, Bernier et al. 2003). Octenol + Lurex blends had an additive effect over unbaited or octenol-baited traps, in the presence of CO2. Based on our results, surveillan ce or population reduction efforts targeting Ae. albopictus with residential traps would best be served by using traps baited with octenol + Lurex. It is interesting to note that almost identic al sex ratios were obt ained in the first 2 trials of our study (2.61:1 and 2.64:1 female: ma le) (Table 3-2). During trial 3, this ratio increased slightly to 3.2 females per male. These relatively low sex ratios were not seen in any of the other species collected in our study, in fact, of the other 5,411 mosquitoes collected, only 27 were males (0.5%) whereas almost 27% of all Ae. albopictus captured in our study were males. Sex ratios were si milar between treatments as well (Table 3-3), with octenoland octenol + Lurex-baited traps capturing approximately equal female: male ratios of Ae. albopictus (2.8:1 and 2.7:1, respectively) Lurex alone attracted the highest ratio of males a nd control treatments the lowest ratio of males. There may be a couple of reasons for these high male ratios. First, Ae. albopictus is a weak flier with a limited flight rang e (Bonnet and Worcester 1946, Watson 1967). Given that adults rarely travel more than 200 m a day, those adults occurring in and around suburban settings are not likely to tr avel very far from their breeding sites, provided that a blood source is close by (the homeowner, pets or wildlife) and breeding
74 sites remain available during summer and la te fall. These necessities, if met in a homeowners backyard, would keep them in cl ose proximity to residential traps for most of their adult lives, that being no more th an 24 days (Estrada-Franco and Craig 1995). Chances are good that they woul d eventually notice and appro ach a trap. Males, being smaller than females on average (personal obs ervation), are probably weaker fliers and would be more susceptible to the vacuum for ce of CFG traps. Second, males were often seen swarming in the vicinity of the author during biting collection periods. It could be that male Ae. albopictus are attracted to lac tic acid present on human skin (and possibly CO2). It is also possible that males are attr acted to objects that females have approached, being attracted to the sound of their flight (wingbeat frequency) (Kanda et al. 1987, Ikeshoji and Yap 1990, Kusakabe and Ikes hoji 1990). The sum of all 3-min biting collections were 110 female and 93 male Ae. albopictus. Males collected off the author comprised almost 46% of Ae. albopictus landing (biting) rates. With only 1 or 2 exceptions, Ae. albopictus was always the first mosquito to approach for a blood meal. Males of this species never took long to appear after females began to bite. Trap collection totals fell following the firs t trial, from 741 adults to 277 (trial 2) and 303 (trial 3). These results are highly significant (p = 0.0002, trials 1 and 2; p = 0.0005, trials 1 and 3, Tukeys multiple comparis on test). In a separate, concurrent study, the USDA ARSs Mosquito and Fly Research Unit monitored Gainesville mosquito populations with CDC light trap s at 6 locations. City surveillance showed a slight increase in Ae. albopictus populations from August to September; 4 of the 6 trap sites caught more adults in September than in August (Appendix B). A decrease in trap capture occurred at our sites while Gainesvi lle populations were stable over the same
75 time period. This data tends to support the idea th at trapping with at tractant-baited CFG traps on a particular property depressed Ae. albopictus populations at those sites. Biting activity remained fairly constant when taken immediately following a trapping period, from one trial to the next. These data seem to indicate that population suppression of Ae. albopictus with the MM Pro was minimal, if any. However, all properties maintained active breeding sites during the test peri od and approximately 1 week was given between trials to allow for some recovery of Ae. albopictus populations. Biting collections were made within the clai med operational effective distance of the trap (i.e., inside the 1-acre radius surrounding the trap). It is qui te possible that biting adults collected during 3-min collection periods were mo re concentrated in the vicinity of traps as a result of attractant plumes produced by n earby traps. The removal of such a large percentage of males (> 25%) during operati on of these traps would only serve to reduce adult populations over time as part of the breeding population was removed concurrently with females. Our results are encouraging because this mosquito is very difficult to control with adulticides and conventional surveillance traps; backyard populations were reduced through time during late summer months when Ae. albopictus was normally at its greatest prevalence. Other Mosquito Species Culex nigripalpus, Ae. albopictus, Oc. infirmatus, Ps. ferox, and Cx. erraticus were the 5 most abundant mosquitoes collected in our study. Significant di fferences were seen between treatment preferences for 3 of these species: Cx. nigripalpus, Oc. infirmatus, and Ps. ferox. Culex nigripalpus was the most abundant mosquito trapped, collected at almost 3 x the total Ae. albopictus, the second most abundant mosqu ito. Control and octenol-baited
76 traps collected significantly more Cx. nigripalpus than did Lurexor octenol + Lurexbaited traps. The 2 traps baited with Lurex comprised a combined total capture rate of 13.5%, indicating a strong prefer ence for other treatments by this species. Control traps caught 38% more Cx. nigripalpus than octenol-baited traps. In north Florida, Kline and Mann (1998) obser ved a similar depressive effect on Cx. nigripalpus capture rates in octenol-baited CDC light traps. In the Florida Everglades, Kline et al. (1990) trapped approximate ly equal numbers of Cx. nigripalpus in octenol-baited and unbaited CDC light traps. Carbon dioxide was provided to CDC light traps in both studies. Much smaller numbers of Cx. erraticus were collected in our study and results were different from Cx. nigripalpus treatment preferences The majority of Cx. erraticus were caught in octenol-baited traps (67) followed by control traps (34). Lurex-baited and octenol + Lurex-baited traps collected fewer Cx. erraticus (29) than the control trap alone. Treatment means we re not significantly different. Our results are similar to those of Kline et al. (1991b) in which larger numbers of Cx. erraticus were collected in octenol + CO2-baited traps than in CO2-baited traps. Their treatment differences were not significant. Ochlerotatus infirmatus was the third most abundant sp ecies trapped in our study. It was caught in significantly larger numbers in control and octenol-b aited traps than in octenol + Lurex or Lurex-baited traps. Almost 42% of this species was collected in control traps, while another 31% was co llected from octenol-baited traps. No significant difference was noted between contro l and octenol-baited traps collections, but control traps caught 25% more Oc. infirmatus than octenol-baited traps. The 2 Lurex-
77 baited traps collected less than 30% of th e total, indicating th at Lurex might be somewhat repellent to Oc. infirmatus. Kline and Mann (1998) caught Oc. infirmatus in CDC light traps in the following order: octenol + CO2 > CO2 > octenol. These results are slightly different from those obtained in our study, however, their collection means were much smaller than ours (CO2 treatment means of 0.80 and 19.8, respectively). No other published data pertaining to lactic acid or octenolbaited traps and Oc. infirmatus collection was found in the litera ture search. It appears that the addition of octenol to CO2-baited traps would not adversely affect Oc. infirmatus capture rates. Lactic acid baits should be avoided if this species is the target of trapping efforts. Psorophora ferox was the fourth most abundant sp ecies caught (147). This number might have increased dramatically (by as much as 188) had adult specimens been in better condition. The Psorophora species count included those Psorophora adults unidentifiable to species due to destruction of key identification characters (i.e., the hind legs were missing). Analysis of Ps. ferox and Ps. spp. yielded similar results. In both cases, significant differences were seen between treatments (Table 3-4) and the order of treatment effectiveness was control > octeno l > octenol + Lurex > Lurex (Table 35). Lurex-free traps caught significantly more Psorophora mosquitoes than Lurexbaited traps. Contro l traps collected 2.4 x and 2.6 x more Ps. ferox and Psorophora species than octenol-baited traps, respectively. Differences were not significant. Kline (1994b) reported that octenol + CO2 had a synergistic effect on Psorophora capture rates as compared to octenol alone. Kline et al. (1991b) caught 3 x more Ps. columbiae in unlit CDC light traps baited with octenol + CO2 than in CO2-baited traps. Their results were different from our results with Ps. ferox. This discrepancy may well lie in host
78 preference differences between these 2 sp ecies, regardless, trap means were not significantly different between treatments in either study. Based on these data, control treatments (CO2 alone) are prefe rred if targeting Ps. ferox; octenol + CO2 should probably be used if Ps. columbiae is the target. Trap collections for other species fell from trial 1 (2,386) to trial 2 (1,547) and from trial 1 and trial 3 (1,533). Although this decr ease was not significan t, it did represent a reduction of approximately 35% from the firs t trial sum over the following 2 trials. Results indicate that the Mosquito Magne t Pro can adversely impact mosquito populations in suburban settings. Octenol + Lurex-baited MM Pro traps caught significantly less Ae. albopictus in 2 subsequent trapping periods after the fi rst trapping period. This may indicate a reduction in backyard populations however, biting rates at thes e sites remained steady. It appears that humans are still more attractive to Ae. albopictus than well-baited MM Pro traps, however, the reduction in trap tota ls through time was encouraging because Gainesville CDC light trap count s increased during the same tim e frame. The addition of Lurex bait to MM Pro traps enhances capture of this pest. Mosquito Magnet Pros can also be used with no bait or with octenol to target and reduce biting populations of Culex, Ochlerotatus, and Psorophora mosquitoes in these same settings. Further efficacy studies using Lurex-baited tr aps would be useful in dete rmining its impact on suburban and woodland mosquito populations.
79 Table 3-1. Totals, means, and SEM of Aedes albopictus collected from Mosquito Magnet Pro traps over 3 identical trials with 4 treatments. Means followed by the same letter are not significantly different (Ryan-Einot-Gabriel-Welsh Multiple Range Test; p < 0.05). n = 12 periods (48 h). Treatment Trial 1 Trial 2 Trial 3 Total Mean ( SEM) L&O 320 86 117 523 45.58a .83 Lurex 199 93 42 334 27.83ab .45 Control 113 39 99 351 20.92ab .95 Octenol 109 59 45 213 17.75b .44 Sum trial 741a 277b 303b 1,321 Table 3-2. Sex ratios of Aedes albopictus collected from Mosquito Magnet Pro traps over 3 identical trials with 4 tr eatments. n = 12 periods (48 h). Replicate Ae. albopictus Female Male Ratio, female: male Trial 1 741 536 205 2.61 Trial 2 277 201 76 2.64 Trial 3 303 231 72 3.21 Total 1,321 968 353 2.74 Table 3-3. Treatment sex ratios of Aedes albopictus over 3 trials and 4 treatments with the Mosquito Magnet Pro. n = 12 periods (48 h). Treatment Trial 1 Trial 2 Trial 3 Total Treatment ratios Control Female 100 31 70 201 4.02 Male 13 8 29 50 1 Octenol female 77 39 41 157 2.8 Male 32 20 4 56 1 Lurex female 139 66 25 230 2.21 Male 60 27 17 104 1 L&O female 220 65 95 380 2.66 Male 100 21 22 143 1
80Table 3-4. Mosquito Magnet Pro trap count s per attractant treatment (means SEM). Means within each row followed by the same letter are not significantly di fferent (Ryan-Einot-Gabriel-Welsh Multiple Range Test). n = 12 trap periods (48 h). Species Lurex Octenol Control Lurex + Octenol p-value Ae. albopictus2,3 27.8 .5ab 17.8 .4b 20.9 .0ab 43.6 .8a 0.029 An. crucians s.l. 0 a 0.6 .6a 0.3 .3a 0.1 .1a 0.67 Cq. perturbans 0 a 0.8 .3a 0.1 .1a 0.8 .6a Cx. erraticus 0.5 .2a 5.6 .8a 2.8 .1a 1.9 .8a 0.08 Cx. nigripalpus2,3 14.3 .4b 106.8 .7a 173.3 .1a 29.5 .4b 0.0001 Oc. atlanticus1,2 0.4 .2a 1.9 .1a 2.6 .1a 0.5 .2a 0.11 Oc. infirmatus2,3 5.5 .1b 14.8 .0a 19.8 .2a 7.2 .6b 0.0003 Oc. triseriatus2 0.4 .2a 1.1 .5a 1.3 .6a 0.9 .4a 0.48 Ps. columbiae2 0 a 1.2 .7a 0.8 .7a 0.1 .1a 0.12 Ps. ferox2,3 0.5 .3c 3.2 .3ab 7.5 .3a 1.1 .6bc 0.0001 Wy. mitchellii2 0.6 .4a 0.8 .5a 1.9 .4a 0.3 .2a 0.63 Cx. spp.2 1.7 .1a 2.8 .5a 1.9 .6a 1.9 .3a 0.81 Ps. spp.2,3 1.2 .7b 3.5 .3ab 9.3 .4a 1.8 .3b 0.0007 1Adults could not be distinguished from Oc. tormenter. 2Significant position effect (p < 0.05). 3Significant period effect (p < 0.05).
81 Table 3-5. Adult totals of the 5 most abunda nt mosquito species co llected from Mosquito Magnet Pro traps with 4 treatments. n = 12 periods (48 h). Species Lurex Octenol Control L&O1 Total Ae. albopictus 334 213 251 523 1,321 Cx. erraticus 9 67 34 20 130 Cx. nigripalpus 171 1,281 2,080 354 3,886 Oc. infirmatus 66 177 237 86 566 Ps. ferox 6 38 90 13 147 Ps. spp. 14 42 111 21 188 1Lurex + octenol. Figure 3-1. Mosquito Magnet Pro used in suburban trials to collect adult Aedes albopictus. Note tank bromeliads; this site produced large numbers of mosquitoes.
82 ab a a a a a b a ab a b a b bc a ab b b c a0 500 1000 1500 2000 2500Ae. albopictusCx. erraticusCx. nigripalpus Oc. infirmatusPs. feroxCollection totals Control Octenol Lurex + Oc Lurex Figure 3-2. Capture totals by treatment over 3 trials with the Mosquito Magnet Pro trap for the most common mosquitoes collected from 4 suburban sites in Gainesville, Florida. Number of mos quitoes collected within each treatment with the same letter is not significantly different ( = 0.05, Ryan-EinotGabriel-Welsh multiple range test).
83 CHAPTER 4 RESPONSE OF Aedes albopictus TO SIX TRAPS IN SUBURBAN SETTINGS IN NORTH CENTRAL FLORIDA Introduction A variety of residential mos quito traps have been developed and sold in the United States over the last 10 years. These trap s were developed for homeowner and business premises use and claim to provide protecti on from biting insects for up to 1.25 acres, depending upon model. Residential traps are e ngineered to provide visual, chemical, and physical attractants for biting insects and to operate over long periods of time with little or no maintenance. They became available to the general public while Ae. albopictus was expanding its established range to over 1, 000 counties in the United States (Janet McAllister, CDC, personal communication). Few published studies exist assessing mosquito trap efficacy in collecting Ae. albopictus in the United States; they have only just begun due to the rece nt arrival of this mosquito in the United States (Sprenger and Wuithiranyagool 1986). Initial studies focused on experimental mosquito traps (Fre ier and Francy 1991) or surveillance traps routinely used by vector control agencies for surveillance of Ae. aegypti (Jensen et al. 1994, Shone et al. 2003). Aedes aegypti surveillance traps were also effective in capturing Ae. albopictus in large-cage trials at the USDA ARS Gainesville facility (Chapter 2). The goal of our study was to assess the effectiveness of 6 traps (3 residential, 3 surveillance) in collecting Ae. albopictus at suburban settings where these traps would be used by homeowners or v ector control personnel for control or
84 surveillance of this mosquito. We hope th at 1 or more of these traps might prove superior to the others and t hus provide better results for both surveillance and control efforts at a chosen site. Materials and Methods Site Selection and Trapping Scheme Six traps were evaluated fo r their efficacy in trapping Ae. albopictus in north central Florida. Selected traps were culled from an initial group of 8 traps tested in large outdoor cages at USDA ARS Gainesville, Florida (Chapter 2). Three id entical field trials were conducted in 6 separate suburban nei ghborhoods in Gainesville during the summer of 2004 (Fig. 4-1). Selection of the 6 test locations was based on homeowner complaints of severe nuisance populations of biting mosquitoes on their properties. Initial surveys showed the presence of Ae. albopictus at all test locations. All test sites had a mix of pine and hardwood trees with minor amounts of undergrowth, typical of suburban neighborhoods in Gainesville. One site was planted extensively in Neoregelra (red finger nail), Bilbergia pyramidalis, and Bilbergia spp. bromeliads; a second site had lesser numbers of Neoregelra (red finger nail) bromeliads. Another site contained a large number of artificial containers a nd tree holes, ideal breeding sites for Ae. albopictus (Watson 1967). Traps were placed in shaded areas under trees or just inside a tree line next to open spaces. All test locations were separated by a minimum of 1 mile. Trapping occurred from 12-24 July, 2-17 A ug. and 25 Aug.-10 Sept. 2004. Traps were left in place 48 h (1 trapping period) to allow for 1 uninterrupted daylight period as Ae. albopictus most actively feeds in the early mo rning and late afternoon (EstradaFranco and Craig 1995). Collect ion data from 4-6 August (tri al 2) were excluded and rerun the following period due to a trap failure. During trial 3, traps were withdrawn from
85 the field for 24 h on 13 Aug. (Hurricane Charlie) and from 4-8 Sept. (Hurricane Frances). At the end of each trapping period, biting ra tes were obtained for a 3-min period at each site using a hand-held mechan ical aspirator (Hausherrs Machine Works, Toms River, N.J.) approximately 25 ft from the trap (within a 1-acre radius of the trap). Trap captures were lightly anesthetized with CO2, stored in labeled paper cups (Solo Cup Company, Urbana, IL), and frozen for later identificati on to species using the keys of Darsie and Morris (2000). All Anopheles quadrimaculatus Say, An. crucians and Ochlerotatus atlanticus (Dyar and Knab)/Oc. tormentor (Dyar and Knab) were pooled as these mosquitoes were taxonomically indis tinguishable from sibling species. Traps Technical details of all trap s used in our study are includ ed in Chapter 2. Based on those results, the following traps (with brief de scription) were selected for field trials: Centers for Disease Control and Preventi on (CDC) light trap (model 512, John W. Hock Company, Gainesville, FL), a ba ttery-powered, suction surveillance trap (control trap). CDC Wilton trap (model 1912, John W. Ho ck Company, Gainesville, FL), a battery-powered suction surveillance trap. Omni-directional Fay-Prince trap (m odel 112, John W. Hock Company, Gainesville, FL), a battery-power ed suction surveillance trap. Mosquito Magnet Pro Trap (MM Pro) (American Biophysics Corporation (ABC), North Kingstown, RI), a propane-powered counterflow geometry (CFG) residential trap. These traps produce updraft suction and downdr aft exhaust plumes of attractants in close proximity to each other. Mosquito Magnet Liberty (MM Liberty) (ABC), an electricity-powered, updraft CFG residential trap. Mosquito Magnet-X (MM-X) (ABC), a ba ttery-powered, updraft CFG residential trap.
86 CDC light-, Wiltonand Fay-Prin ce traps were provided with CO2 from a 9 kg (20 lb) compressed gas cylinder. A flow rate of 500 mL/min was achieved by using a 15-psi single stage regulator equippe d with microregulators and an inline filter (Clarke Mosquito Control, Roselle, IL). Carbon di oxide was delivered to the trap through 6.4 mm diameter plastic Tygon tubing (Saint-Gobain Perfor mance Plastic, Akron, OH) secured with tape 2 in from the intake (suction) cylinder. The MM-X used similar compressed gas cylinders and regulators but was supplied with an ABC microregulator, inline filter, and flexible vinyl tubing. A ll residential traps (MM Pro, MM Liberty, MMX) were provided with octenol + Lurex car tridges as these traps were engineered for supplementation with attractants (ABC). Re sults from an earlier study (Chapter 3) showed that traps baited with both attractants caught more Ae. albopictus than unbaited, octenol-baited, or Lurex-baited traps. Surveillance traps were baited on withCO2 and hung at heights recommended per manufacturer inst ructions. Trap attr actant qualities and chemical attractants are listed in Table 4-1. Statistical Analysis Traps were randomly rotated between sites in a 6 x 6 Latin square design. Trap, period (48 h), and position effects were evaluated using a 3-way ANOVA (SAS Institute, 2001) for the total number of Ae. albopictus collected. Multiple comparisons were made with the Ryan-Einot-Gabriel-Welsh (REGW) multiple range test to determine significant differences between trap means ( = 0.05). The entire mosquito collection was then analyzed for trap, period, and position effects using a 3-way ANOVA. Multiple comparisons were made using the REGW multiple range test ( = 0.05). All capture data were transformed with log10 (n + 1) prior to analysis.
87 Results Aedes albopictus Eighteen trap periods (48 h each) over 3 tr ials from 6 suburban sites yielded a total of 37,237 mosquitoes. Aedes albopictus comprised 14.2% of the entire catch (5,280 adults) (Table 4-2). Significant difference s between trap means were found (F = 48.01, df = 5, p < 0.0001), site (F = 14.89, df = 5, p < 0.0001), and period (F = 2.24, df = 17, p = 0.0086). Order of trap effectiveness wa s the MM Liberty > MM-X > MM Pro > FayPrince trap > CDC light trap > Wilton trap (Table 4-2). The MM Liberty and MM Pro produced CO2 at rates of 420 mL/min and 520 mL/min, respectively. Other traps were provided with CO2 at the rate of 500 mL/min. Thus, CO2 rates for all traps were approximately equal. The MM Liberty, MM Pr o, and MM-X were bait ed with octenol + Lurex cartridges. The decision to use both attractants in each residential trap was based on superior results achieved in previous attractant comparison trials (Chapter 3). Fay-Prince-, CDC lightand Wilton traps we re not baited with octenol or Lurex attractants, but with CO2 only. Residential traps (MM Liberty, MM-X, MM Pro) accounted for 80.3% of all Ae. albopictus captured. Differences in collection m eans between residential traps were not significant. Residential trap counts were significantly different from surveillance trap counts (t = 14.60, p < 0.0001). Among surveilla nce traps, the FayPrince trap caught significantly more Ae. albopictus than did the Wilton trap (p < 0.0001) and 31.3% more than the CDC light trap (not significant) The CDC light trap caught 26.8% more Ae. albopictus than the Wilton trap, but this difference was not significant. Site proved to be si gnificant with respect Ae. albopictus collections (F = 14.89, df = 5, p =< 0.0001). Over half (53.5%) of all Ae. albopictus trapped in our study were
88 collected from 2 of the 6 sites. One site c ontained a large number of natural (tree holes) and artificial containers and the other site was planted exte nsively in tank bromeliads. Both sites were heavily treed and well shaded. Of the remaining 4 sites, only 1 had small numbers of tank bromeliads and all had few, if any, artificial containers found breeding Ae. albopictus. Thirty-five percent of all Ae. albopictus captured in our study were male. Sex ratios (female: male) were lower and closer in residential traps than in surveillance traps (Table 4-3). Residential trap sex ratios ra nged from approximately 1. 5:1 (MM Pro) to 4:1 (MM-X). Surveillance trap sex ratios were 5.4:1 (Wilton trap), 9.5: 1 (Fay-Prince trap), and 20.7:1 (CDC light trap). Trapping suppressed Ae. albopictus populations over time wi th trapping reductions of approximately 18.5% between trials 1 and 2 and trials 2 and 3. Although these reductions were not significant, they demonstrated a small s uppressive effect in continued trapping at suburban sites. It was not possibl e, nor was it the intent ion, to assess trapping impact on Ae. albopictus populations according to trap type since all trap models were rotated between all sites. However, it is interesting to note that trapping impact was greater at sites using only a residential trap (M M Pro, Chapter 3) than at sites using both residential and surveillance traps. A signifi cant reduction in trap capture occurred in residential traps over time; this was not th e case in our study using both trap types Other Mosquito Species Twenty-seven species of mosquitoes were captured in our st udy, representing 35% of all mosquito species (77) occurring in Florida (Darsie and Morri s 2000). Apart from Ae. albopictus, 31,957 mosquitoes were collected, incl uding 22 males (0. 07% of the total excluding Ae. albopictus males). Other mosquito species trapped included Ae. vexans,
89 Anopheles crucians s.l., An. quadrimaculatus s.l., An. perplexens Ludlow, Coquillettidia perturbans (Walker), Culiseta melanura, Culex erraticus, Cx. nigripalpus, Cx. quinquefasciatus Say, Cx. salinarius Coquillett, Mansonia titillans (Walker), Ochlerotatus atlanticus s.l., Oc. canadensis canadensis (Theobald), Oc. dupreei (Theobald), Oc. fulvus pallens (Wiedemann), Oc. infirmatus, Oc. taeniorhynchus (Wiedemann), Oc. triseriatus, Psorophora ciliata (Fabricius), Ps. columbiae, Ps. cyanescens (Coquillett), Ps. ferox, Ps. howardii (Coquillett), Uranotaenia lowii Theobald, Ur. sapphirina, and Wyeomyia mitchellii. The 9 most prevalent species collected in descending order were Cx. nigripalpus, Ae. albopi ctus, Oc. infirmatus, Ps. ferox, Cx. erraticus, Wy. mitchell ii, Oc. triseriatus, Cq. perturbans, and Oc. atlanticus. Significant differences between trap means we re seen for these 9 species (Table 4-4). Culex nigripalpus and Ae. albopictus were the 2 most abundant species caught in our study, comprising 85.1% of all mo squitoes. Tremendous numbers of Cx. nigripalpus emerged at the beginning of September, 3 weeks after Hurricane Charlie passed through Gainesville (13 August, 2004). During August, Gainesville received approximately twice (14.53 in) its normal amount of rainfall (6. 63 in) (National Ocean ic and Atmospheric Administration). Two sites pr oduced most of these mosquitoes; 1 site was in close proximity to a flood plain and the other was heavily wooded with a plastic childrens swimming pool that held water and leaves, pr oviding an ideal breeding site for this mosquito. This also happened to be the most productive Ae. albopictus site in our study. The MM-X and CDC light traps coll ected the largest numbers of Cx. nigripalpus, 8,603 and 7,511, respectively. Species means (total from all traps/108 trap days) were 244.41 for Cx. nigripalpus and 48.89 for Ae. albopictus.
90 The following mosquito species were caught in large enough numbers for analysis but had collection means of < 27 per trap In descending order they included Oc. infirmatus (26.35), Ps. ferox (10.18), Cx. erraticus (3.14), Wy. mitchellii (2.21), Oc. triseriatus (2.06), Cq. perturbans (1.69), and Oc. atlanticus (1.33). Two surveillance traps (Fay-Princeand CDC light traps) a nd 1 residential trap (MM-X) performed well with these species (Table 4-4). Octenol + Lurex baits may have negatively impacted residential trap collections of Culex and Psorophora mosquitoes, as witnessed in the attractant study (Chapter 3). Relative percent composition of the 6 traps with respect to mosquito species is given in Fig. 4-1. Culex erraticus (F = 14.42, df = 5, p < 0.0001) and Cx. nigripalpus (F = 8.11, df = 5, p < 0.0001) were caught in significantly higher numbers in CDC light traps than in any other trap test ed. Forty five percent of all Cx. erraticus were captured in CDC light traps, almost twice the rate of the next most effective trap, the Fay-Prince trap (23%). Approximately 60% of all Cx. nigripalpus were collected in MM-X and CDC light traps. Ochlerotatus infirmatus and Oc. triseriatus likewise responded well to CDC light traps (40% and 23%, respectively) followed by MM-X and Fay-Prince traps. Over half of all Oc. atlanticus were collected in MM-X traps fo llowed by CDC light traps (23%). Coquillettidia perturbans responded equally to MM-X (38% ) and CDC light traps (36%). Approximately 38% of all Ps. ferox were trapped in CDC light traps, followed by Wilton traps (24%). Wyeomyia mitchellii was attracted to the Wilton trap (43% of the total) in much larger numbers than in the next best traps, CDC lightand Fay-Prince traps (17% and 16%, respectively), although the difference was not significant.
91 Table 4-5 ranks trap performance based on assignment of numerical values for the 9 most abundant mosquito species. A rank of 1 was assigned to a to species/trap combination for best performance with comb inations ranked sequentially through 6. Trap scores were derived from their respec tive ranking for each mosquito species divided by the total number of mosquito species coll ected (9). Low scores indicated superior performance, higher-scored traps performed less well. Trap rankings of our study were as follows: CDC light trap > MM-X > Fay-Pr ince trap > MM Liberty > Wilton trap > MM Pro. Discussion Aedes albopictus Aedes (Stegomyia) mosquitoes are generally not well represented in surveillance efforts that typically make use of light tr aps, even when present in large numbers (Thurman and Thurman 1955, Service 1993). The medical significance of these mosquitoes (Yellow fever and dengue vectors) necessitated the development of effective adult traps capable of providing more accura te information on population densities and seasonal distribution. Two traps were developed specifically for Ae. aegypti: the Wilton trap and Fay-Prince trap. Both used black co lor as an attractant and the Fay-Prince trap incorporated color-contrast as an additiona l attractant (Fay 1968, Fay and Prince 1970, Wilton and Kloter 1985). Field studies have demonstrated that these traps are also effective Ae. albopictus surveillance devices (Jensen et al. 1994, Shone et al. 2003, Dennett et al. 2004). Just over 80% of all Ae. albopictus collected during these trials were captured in residential traps with significantly larger collections than in surveillance traps. Percent composition of trap means and separation of means by REGW multiple comparisons are
92 presented in Fig. 4-1. These highly significan t results indicate that the residential traps are much more effective in trapping Ae. albopictus than surveillance traps in suburban settings. Similar results were obtained by Dennett et al. (200 4) trapping with 3 surveillance and 3 residential trap s at a tire repository in Hous ton, Texas. Two residential traps, the MM Liberty and Dragonfly trap accounted for 63% of all Ae. albopictus collected in that study. The MM Li berty collecting si gnificantly more Ae. albopictus than any other trap. Intere stingly, the MM Liberty used in that study was baited with octenol, which was shown in the attractants st udy (Chapter 3) to lightly depress capture rates in MM Pro traps. Residential trap means (MM Liberty, MM-X and MM Pro) were not significantly different. The MM Liberty caught 25.5% more Ae. albopictus than did the MM Pro and 7.7% more than the MM-X. The MM Liberty was more effective in collecting Ae. albopictus than the MM Pro despite both having inlet tubes approximately equal in height, both producing similar amounts of CO2, and both optimally baited with octenol + Lurex baits. This difference may be due to the presence of 2 black support arms next to the intake tube on the MM Liberty, wh ich are absent on the MM Pro. On many occasions Ae. albopictus was observed swarming next to the support arms in close proximity of the inlet tube (personal observation). The MM-X recently changed status from an e xperimental trap to a residential trap available for purchase from ABC. As noted in Table 4-1, it offers color contrast and black visual attractants as does the MM Li berty and MM Pro. It too was baited with octenol + Lurex. A possible s econdary visual attraction featur e of this trap is lateral movement in wind. Hung from a rope off a support post, winds under 10 mph will
93 produce limited lateral movement, a known stimulus of Ae. aegypti (Sipple and Brown 1953). The MM-X also has CFG updraft t echnology and captured nearly as many Ae. albopictus as did the MM Liberty (1,468 and 1,591, respectively). The MM-X has an important advantage over the other 2 resident ial traps: a large containment shell that affords protection from high velocity air cu rrents produced in ABC residential traps. Captured mosquitoes are often alive and in good physical condition at retrieval. High flow air velocity in the MM Liberty a nd MM Pro desiccate and damage trapped mosquitoes, the containment shell of the MM-X affords protection against damaging air currents. This is important if adults have to be live-captured in disease studies. MM-X collections had to be frozen before identific ation began due to the large number of living specimens at retrieval. During the third field trial, a large cotton ball soaked in 5% sugar water was added to the MM-X in an attempt to enhance survival rates. No noticeable difference in survival rates was obser ved compared with earlier trials. Surveillance traps caught significantly fewer Ae. albopictus than residential traps in our study. The Fay-Prince tr ap captured twice as many Ae. albopictus (473) as the Wilton trap (238) and 31.3% more than the CD C light trap. Jensen et al. (1994) obtained similar field results in north Florida with Fay-Prince traps collecting significantly more Ae. albopictus than CDC light traps. Schoeler et al. (2004) trapped approximately equal numbers of Ae. aegypti in Peru with Fay-Prince and W ilton traps, however, totals were small (21 and 19, respectively). Kloter et al. (1983) caught significantly more Ae. aegypti in Fay-Prince traps compared with black cy linder traps (Wilton prototype trap). Apart from these published reports, nothing was found comparing capture rates between FayPrince and Wilton traps for Ae. albopictus. Results from these 2 studies were similar to
94 our results. Large-cage trials (Chapter 2) mirrored our field results (Fay-Prince trap (113.5) > CDC light trap (51.2) > Wilton trap (25.8)). In both tests, surveillance traps were provided with CO2 at the rate of 500 mL/min and all traps were set at manufacturers recommended he ights. Significantly more Ae. albopictus were collected in low-height Wilton traps (set with the ope ning 15 in to 20 in above ground) than in Wilton traps set at recommended heights (Chapt er 2). Following prot ocol (setting traps per manufacturers recommendations) in our tria ls prevented testing of Wilton traps at modified heights, but it seems that the lower setting would have increased traps collections. Sex ratios (female: male) were much closer in residential traps than in surveillance traps. Males comprised 30.0% of all Ae. albopictus capture in residential traps and 9.4% of all surveillance trap captures. All female: male residential trap ratios were between 1.5:1 and 4:1. Female: male surveillance tr ap ratios ranged between 5:1 and 21:1 (Table 4-3). One possible factor that may have infl uenced these ratios was the use of octenol + Lurex baits with residential traps and but not in surveillance traps. Kusakabe and Ikeshoji (1990) found that lactic acid-baited targets were eq ually attractive to both sexes of Ae. aegypti in the laboratory, but could not reproduce these results with male Ae. albopictus in the field. Lactic acid was reported to increase Ae. albopictus capture in CO2-baited traps in Japan, but sex ratios were not given in the re port (Ikeshoji 1993). Lactic acid receptors have been located on the antennae and maxillary palpi of female Ae. aegypti, but nothing was reported about the presence of these receptors in males (Acree et al. 1968, Davis and Sokolove 1976). In the attr actant bait study (Cha pter 3), lactic acidbaited MM Pro traps caught 57.1% more male Ae. albopictus than traps lacking lactic
95 acid baits, although these results were not significant. Trap totals of male Ae. albopictus were as follows: octenol + Lurex (143) > Lu rex (104) > octenol ( 56) > control (50). It appears that Ae. albopictus males are capable of detecting lactic acid but further fieldtesting is needed and males should be examin ed for the presence of lactic acid-sensitive receptors. Our results demonstrate a clear preference of Ae. albopictus for residential traps over surveillance traps. A review of trap f eatures (Table 4-1) shows several important advantages these traps have that are lacking in surveillance traps: supplemental attractive baits, updraft technology, producti on of heat and water vapor, and color contrast. It was noted that the 2 least effective traps, the CDC light and Wilton trap, rely on incandescent light and solid black color as primary attractan ts. Incandescent light is known not to be particularly attractive to Ae. albopictus (Thurman and Thurman 1955, Service 1993) and the Wilton trap lacked color contrast (black and white) present in the 4 best performing traps (MM Liberty, MM-X, MM Pr o, and Fay-Prince traps). Contrasting black and white is also highly attractive to Ae. aegypti (Sippell and Brown 1953, Christophers 1960). Future Ae. albopictus surveillance and contro l efforts would be better served using these newer, better-equipped residential traps. Other Mosquito Species Of the 27 mosquito species collected in these trials, only 9 species were collected in large enough numbers for meaningf ul analysis. In addition to Ae. albopictus, these included Cq. perturbans, Cx. erraticus, Cx. nigrip alpus, Oc. atlanticus, Oc. infirmatus, Oc. triseriatus, Ps. ferox, and Wy. mitchellii. Significant differences were seen between trap captures for all of these mosquitoes (T able 4-4). Relative percent trap composition of each species is given in Fig. 4-1.
96 More Cx. nigripalpus (26,396) were trapped than a ny other species and comprised 70.9% of the entire collection. Trap means were highest in the MM-X trap (477.9) but the REGW multiple comparison test gave the CDC light trap (mean of 417.3) superior ranking (Table 4-4, Fig. 4-1). This discrepancy is due to ex tremely large count variations in MM-X trap collections (low count of 0, high count of 5,470). Logarithmic transformation of highly variab le count data can result in different rankings between count means and log transformed count means for identical data sets (D r. Littell, Dept. of Statistics, University of Fl orida, personal communication). Above average rainfall brought about by 2 hurricanes during the su mmer of 2004 contributed greatly to Culex production. Carbon dioxide is routinely used to enhance trap capture of Culex mosquitoes (Service 1993). Octenol-baited tr aps have given mixed results in Cx. nigripalpus capture rates (Kline et al. 1990, K line et al. 1991b), and most Culex mosquitoes show little response to it (Kline 1994b, Van Essen et al. 1994). Culex nigripalpus responded well to octenol-baited MM-X traps in our study and well to octenol-free CDC light traps. Perhaps light, contrasting co lor, and octenol contributed to good performance in these traps. Fay-Prince and MM Liberty traps pr oduced favorable results as well (means of 327.4 and 148.8, respectively). It appears that trap visual qualities are an important factor in collecting Cx. nigripalpus as all 4 traps had either li ght or contrasting color as attractive components (Table 4-1). Lesser numbers of Cx. erraticus were trapped (339) but again, the CDC light trap produced the best result (152) followed by th e Fay-Prince trap (80). CDC light trap results were significantly differe nt from other traps, and th e Fay-Prince trap caught more
97 than twice as many Cx. erraticus as the third best trap (MM-X). The lack of lactic acid bait in CDC lightand Fay-Prin ce traps may have contributed to their relatively high capture rates. Lurex-baited MM Pro traps attracted the fewest Cx. erraticus in the attractant study (Chapter 3). Octenol + Lu rex combination counts were also low, however, octenol-baited traps were most attractive. Using CO2-baited, unlit CDC traps in Arkansas rice fields, Kline et al. (1994b) obtained higher mean s in traps to which octenol was added (no significant difference). Li ght traps appear to be a good choice for collecting Cx. erraticus and lactic acid should be avoide d if targeting this mosquito. A small number of Cq. perturbans were collected during these trials (183). The MM-X and CDC light traps caught approximately equal numbers of this mosquito (71 and 66, respectively), and signi ficantly more than the remaining 4 traps (Table 4-1). Despite the small number captured, results indi cate that light traps are a good surveillance tool for this species. Our MM-X result was mirrored by Kline (1999) in which an MM-X caught significantly more Cq. perturbans than an ABC Pro trap (CDC-type light trap). The addition of octenol to the MM-X may have biased capture results relative to unbaited traps (Kline et al. 1994b). In fact, octenol alone has been sh own to be more attractive to this species than traps baited with CO2 alone, a rare occurrence among Florida mosquitoes (Kline et al. 1990). Personal obser vations from the University of Floridas Horse Teaching Unit (Chapter 7) show that CD C light traps are an excellent choice for collecting Cq. perturbans. Campbell (2003) trapped more Cq. perturbans with MM-X and CDC light traps than with MM Pro and MM Liberty traps (no octenol) at the same location.
98 Three species of Ochlerotatus were trapped in significant numbers in our study. They include Oc. infirmatus, Oc. triseriatus, and Oc. atlanticus. Ochlerotatus infirmatus was the third most abundant mosquito caught an d is common throughout most of Florida. Significantly more Oc. infirmatus were caught in CDC light-, Fay-Prince and MM-X traps than the remaining 3 traps (Table 4-4) The CDC light trap accounted for 40.5% of all adults collected (Table 4-1). Few pub lished reports of mosquito trapping and/or attractants include data on Oc. infirmatus (Kline and Mann 1998, Kline 1999). Kline (1999) found no significant difference in capture means between MM-X and ABC Pro light traps, but both means were less than 2. Kline and Mann (1998) used different attractants with CDC light traps and obtained significantly higher means of Oc. infirmatus in octenol + CO2-baited traps than in CDC light traps baited with CO2, butanone, CO2 + butanone, and octenol. Most trap means were relatively small (< 5 for octenol + CO2). We mention here that in our stu dy, baited traps also contained Lurex, shown to reduce capture rates of Oc. infirmatus relative to Lurex -free traps (Chapter 3). The octenol + Lurex-baited MM-X trap collected significantly more Oc. atlanticus than all other traps except the CDC li ght trap, accounting for over half of all adults collected. Ochlerotatus atlanticus was captured in signifi cantly high numbers in octenol + CO2-batied CDC light traps compared with butanone-, CO2 + butanone-, CO2-, or octenol-baited traps (Kline and Mann 1998) Similar results were seen with Oc. triseriatus capture sums; the MM-X and CDC light trap each caught 51 adults from a total of 224 (45.5%). No other reports of multiple trap comparisons involving Oc.
99 atlanticus and Oc. triseriatus were found. Octenol + CO2-baited MM-X traps are recommended for targeting Oc. atlanticus. Psorophora ferox was trapped in significantly higher numbers in CDC lightand Wilton traps than in MM Pro and MM Libert y traps (Table 4-4). The Fay-Prince and MM-X traps produced intermediate results. Octenol, Lurex, and octenol + Lurex blends were shown to be repellent to this species (Chapter 3) a nd like-baited MM Pro, MM Liberty, and MM-X traps collected the smal lest numbers in our study. Almost 40% (Fig. 4-1) of all Ps. ferox were collected in CDC light traps indicating that this species is strongly attracted to light. The Wilt on traps good result in collecting Ps. ferox may be owed to black color mimicking re flected water or tree holes. Psorophora ferox breeds in shaded thickets and water-filled pot holes (Carpenter and LaCasse 1955). Significant differences between collection sites and trap means were obtained for Wy. mitchellii. Only 4 of the 6 test sites produced this species; 1 site with a single specimen. The most productive site was ex tensively planted in tank bromeliads and accounted for 78.6% (202) of all adults. The second and third best sites had lesser number of bromeliads on their properties ( 31 and 16 adults, respectively). Wilton trap results were significantly better than all othe r traps with the exception of the CDC light trap (Table 4-4), and it accounted for 43.5% of the Wy. mitchellii catch (Fig. 4-1). CDC lightand Fay-Prince trap totals were appr oximately equal (44 and 41, respectively). Surveillance traps (Wilton-, CDC light-, a nd Fay-Prince traps) accounted for 76.2% of the total catch. It appears th at this mosquito is highly attr acted to black surfaces that may mimic reflected water. Wyeomyia mitchellii breeds primarily in bromeliads with minor breeding in tree holes and bamboo st umps (Carpenter and LaCasse 1955).
100 It is obvious that traps baited with oc tenol + Lurex were less attractive to Wy. mitchellii than those not so bait ed. Kline et al. (1990) collected twice as many Wy. mitchellii in CO2-baited CDC light tr aps than in octenol + CO2-baited traps. However, slightly more Wy. mitchellii were collected from octenol + CO2-baited CDC light traps than in CO2-baited CDC light traps (Takken and Klin e 1989). It appears that the addition of Lurex to residential traps may have depressed Wy. mitchellii counts Similar results were seen in our attractants study. The order of trap captu re in that study was control > octenol > Lurex > octenol + Lurex. We note that means were small and not statistically different from each other. Residential traps baited with octeno l + Lurex caught significantly more Ae. albopictus than surveillance traps. The excellent results achieved by CFG traps in our study are similar to those achieved in a multi-trap study targeting Ae. albopictus by Dennett et al. (2004). In both trials the MM Liberty attained supe rior results among all traps tested. In our study, MMX and MM Pro trap totals were not significantly different from the MM Liberty or each other. Ease of use, long term operation (3 weeks) and superior results of propane -powered CFG traps (MM Libe rty and MM Pro) make them ideal candidates in long-term surveilla nce or reductions programs targeting Ae. albopictus. Trap rankings (Table 4-5) for all mos quitoes in our study were as follows: CDC light trap > MM-X > Fay-Prince trap > MM Li berty > Wilton trap > MM Pro. Results indicate that light is an important attractan t for most mosquito species collected in our study. Except for Ae. albopictus and Wy. mitchellii, these mosquitoes prefer to feed at night. Incandescent light (from the CDC light trap) is a pparently a good attractant for the
101 majority of those mosquitoes. The MM-X and Fay-Prince traps, both making use of contrasting colors, performed well. Trap s using CFG technology performed well (MM-X and MM Liberty). The MM-X trap, only recently available for purchase, is very useful in collecting and preserving most of the mos quito species encountered in our study. Although traditional surveillance tr aps performed well in trapping Culex, Ochlerotatus, and Psorophora mosquitoes, residential traps also performed well and offer homeowners the advantage of long-term use (3 weeks) with little attenda nce or maintenance required for operation. The primary advantage of re sidential traps is that they produce the CO2, the most effective of mosquito attractants, whereas surveillance traps must be constantly resupplied with CO2 and a power source (battery). These factors favor the newer, residential mosquito traps for homeowner use for on-premises mosquito control. Additionally, these traps can be operated wit hout Lurex and/or octenol baits, possibly increasing their efficiency in collecting those mosquitoes that may be repelled by these attractants.
102 Table 4-1. Trap features and chemical a ttractants used in comparison trials with residential and surveillance traps in Gainesville, Florida. Trap Oct1 Lur2 CO2 3 Contrast4 Black surface Light Updraft Heat Water vapor MM Pro Yes Yes 520 Yes Yes No Yes Yes Yes MM Liberty Yes Yes 420 Yes Yes No Yes Yes Yes MM-X Yes Yes 500 Yes Yes No Yes No No Fay-P No No 500 Yes Yes No No No No Wilton No No 500 No Yes No No No No CDC No No 500 No Yes5 Yes No No No 1Oct = octenol. 2Lur = Lurex. 3CO2 flow rate in mL/min. 4Contrast of white and black colored surfaces. 5Black plastic trap cover vice aluminum cover. Table 4-2. Total adult Aedes albopictus caught in 6 traps over 3 trials in suburban neighborhoods in Gainesvill e, Florida over 36 days (n = 18 periods of 48 h). Number of mosquitoes colle cted within each trap with the same letter is not significantly different ( = 0.05, Ryan-Einot-Gabriel-Welsh multiple range test). Trap Trial 1 Trial 2 Trial 3 Sum Mean REGW test MM Liberty 739 474 378 1,591 88.39 A MM-X 535 520 413 1,468 81.56 A MM Pro 462 408 315 1,185 65.83 A Fay-Prince 130 176 167 473 26.28 B CDC 125 91 109 325 18.06 BC Wilton 141 67 30 238 13.22 C Sum trial 2,132 1,736 1,412 5,280 Table 4-3. Sex ratios of Aedes albopictus caught in 6 traps over 3 trials in suburban neighborhoods in Gainesvill e, Florida over 36 days (n = 18 periods of 48 h). Trap Total Female Male Ratio F:M MM Liberty 1,591 1,075 516 2.08 MM-X 1,468 1,173 295 3.98 MM Pro 1,185 723 462 1.56 Fay-Prince 473 428 45 9.51 CDC 325 310 15 20.67 Wilton 238 201 37 5.43
103Table 4-4. Adult mosquito count per trap (means SEM). Means within each row ha ving the same letter are not significantly different (Ryan-Einot-Gabriel-Welsh Multiple Range Test). n = 18 trap periods (48 h). Species MM Pro MM Liberty MM-X Fay-Prince CDC Trap Wilton Trap p-value Ae. albopictus2,3 65.8 .4a 88.4 .9a 81.6 .9a 26.3 .9b 18.1 .2bc 13.2 c 0.0001 Cq. perturbans2 0.5 .2b 0.4 .2b 3.9 .3a 1.3 .4ab 3.7 .1a 0.3 .1b 0.0001 Cx. erraticus3 0.6 .2c 1.6 .8bc 2.6 .2bc 4.4 .3b 8.6 .8a 1.1 .8c 0.0001 Cx. nigripalpus2,3,4 27.8 .3c 148.8 .9b 477.9 .8b 327.4 .6b 417.3 a 67.2 .6bc 0.0001 Oc. atlanticus1,3 0.06 .06c 0.2 .1bc 4.2 .8a 1.5 .7bc 1.8 .7ab 0.2 .1bc 0.0001 Oc. infirmatus3 6.1 .7b 7.1 .4b 32.8 .8a 35.9 .8a 64 .3a 12.3 .4b 0.0001 Oc. triseriatus2,3 0.8 .3b 1.9 .8ab 2.8 .8a 2.3 .6ab 2.8 .8a 1.7 .5ab 0.03 Ps. ferox2,3 0.8 .3c 2.9 .3bc 8.1 ab 11.1 .6ab 23.6 .3a 14.6 .2a 0.0001 Wy. mitchellii2 0.6 .4b 1.2 .5ab 1.7 .6ab 1.3 .9ab 2.4 .2ab 6.1 .1a 0.003 1Adults could not be distinguished from Oc. tormenter. 2Significant position effect (p < 0.05). 3Significant period effect (p < 0.05). 4One or more trap means and REGW rankings differ due to variability in trap capture. 7
104Table 4-5. Trap performance ranking of the most commonly occurr ing mosquito species in residen tial settings in Gainesville, Fl orida. Rank assigned according to total capture for each species. n = 18 collection periods (48 h). Species MM Pro MM Liberty MM-X Fay-Prince CDC Trap Wilton Trap Ae. albopictus 3 1 2 4 5 6 Cq. perturbans 4 5 1 3 2 6 Cx. erraticus 6 4 3 2 1 5 Cx. nigripalpus 6 4 1 3 2 5 Oc. atlanticus 6 4.5 1 3 2 4.5 Oc. infirmatus 6 5 3 2 1 4 Oc. triseriatus 6 1.5 3 4 1.5 5 Ps. ferox 6 5 4 3 1 2 Wy. mitchellii 6 5 3 4 2 1 Rank average (rank) 5.44 (6) 3.89 (4) 2.33 (2) 3.11 (3) 1.94(1) 4.28 (5)
105 a b c c c a b bc b bc a a bc a a c b c bc bc bc a a b ab b ab b b bc 0 10 20 30 40 50 60 Ae. albopictus Cq. perturbans Cx. erraticusCx. nigripalpus Oc. atlanticusPercent species captured Pro Liberty MMX Wilton CDC Fay Figure 4-1. Relative percent trap capture of the 9 most commonly occurring mosquito species in suburban neighborhoods in Gain esville, Florida. Means within each species group having the same le tter are not significantly different (Ryan-Einot-Gabriel-Welsh Multiple Range Test). = 0.05, n = 18 trap periods (48 h each).
106 b c b b ab ab ab b ab ab a a a a ab b ab ab a a b ab ab a 0 10 20 30 40 50 60Oc. infirmatusOc. triseriatusPs. feroxWy. mitchelliiPercent species captured Pro Liberty MMX Wilton CDC Fay Figure 4-1. Continued
107 CHAPTER 5 SUSCEPTIBILITY OF Aedes albopictus TO FIVE COMMONLY USED ADULTICIDES IN FLORIDA Introduction Organized mosquito control operations ar e provided by 56 of Floridas 67 counties as well as many municipalities, cities, and to wns (Florida Department of Agriculture and Consumer Services, Bureau of Entomology and Pest Control, http://www.flaes.org/aesent/mosquito/mosqcontroldirectory.html ). These agencies often rely on adulticides to provide rapid elimination of biting adult mos quitoes. Although this method of mosquito control is only 1 of several available and the least preferre d, it is nevertheless a common and sometimes routine event. One problem associated with the routine us e of adulticides is the development of resistance or tolerance. Ins ecticide resistance is defined as the ability in a strain of insects to tolerate doses of toxicants whic h would prove lethal to the majority of individuals in a normal population of the sa me species (W.H.O. 1957). Tolerance has been defined as an organisms increased abili ty to metabolize a chemical subsequent to an initial exposure (Hodgson and Levi 2001). Tolerant field populations exhibit higher lethal concentration (LC)50 and LC95 (lethal concentrations necessary to kill 50% and 95 % of test populations, resp ectively) values than sus ceptible populations, but do not exceed values considered to be resistant. The World Health Organization defined tolerant mosquito larval populations as ha ving resistance ratios (RR) less than 10 x and adult RRs less than 4 x (Brown and Pal 1971). Resistan ce ratios are derived by dividing
108 the field colony LC50 by the susceptible colony LC50. The coated vial bioassay technique described in the materials and methods section has been used by the Texas A&M (TAMU) Mosquito Research Laboratory to mo nitor mosquito susceptibility in Texas for 25 years. They consider a mosquito population as being tolerant to an insecticide as its RR approaches 10 x and resistant if the RR exceeds 10 x (J. K. Olson, Texas A&M University, Department of Entomology, pe rsonal communication). Resistance ratios of 10 x have also been used to assess resistance in other insects, such as cockroaches (Valles et al. 1997). Insecticide resistance in Florida mosquito es was first noticed in Brevard County during the summer of 1949, when DDT failed to control Ochlerotatus taeniorhynchus and Oc. sollicitans to the same level that it had the previous 5 years (Deonier and Gilbert 1950). By 1958, most populations of Oc. taeniorhynchus were resistant to DDT to some degree throughout the state (B reaud 1993). Over the next 40 years (1949-1989), various populations of salt marsh Ochlerotatus, Culex quinquefasciatus, and Cx. nigripalpus developed resistance to adultici des containing the active ingredie nts (a.i.) of chlorpyrifos, naled, malathion, and fenthion (Breaud 1993). The state of Florida has been monitoring mosquito insecticide suscep tibility at the John A. Mulrennan, Sr. Public Health Entomology Research and Education Cent er (PHEREC) in Panama City, Florida ( http://www.pherec.org/center.html ) since 1964 (Boike et al. 1989). The goals of PHEREC are to establish baseline susceptibility levels in target organisms, track the occurrence of resistan ce through time and location, and provide control agencies with recommendations to alleviate or reverse resistan ce trends where it o ccurs (Petersen et al.
109 2004). At the present time, th e susceptibility status of Ae. albopictus to adulticides routinely tested at PHEREC is unknown. Aedes albopictus has been present in the United States for only 20 years. During this time, it has colonized 1,035 counties in 32 states (Janet McAllister, CDC, personal communication) and all 67 c ounties of Florida (OMeara et al. 1995). Widespread distribution, severe nuisance, and poten tial disease threats associated with Ae. albopictus make it a key target for vector control agenci es. As a target of control efforts often relying on adulticides, the potential to de velop resistance to these compounds exists. Aedes albopictus has a limited flight range that mi ght result in increased selection pressure as it does not move much more than 200 m its entire lifetime (Bonnet and Worcester 1946). Limited flight range could serve to intensify selection pressure for resistance due to the mosquitos reduced ability to escape treated areas. Likewise, immigration of susceptible breedi ng adults into an area of resi stant adults would not be as great as more mobile species, possibly serv ing to limit the reintr oduction of susceptible genes back into a resistant population. Few public health adulticides are currently available for use against mosquitoes in the United States. Reasons include high cost of bringing new compounds to market ($50100 million), a limited niche market in which to recover costs and earn profits, increased public opposition associated with broadcast pest icide applications, especially from certain environmental organizations, and the threat of litigation due to accidental wildlife or domestic animal poisoning (Kline 1994a, Ro se 2001, Fehrenbach 1990). Thus, with a limited number of public health a dulticides available for use by vector control agencies, it becomes imperative that the effectiveness of th ese insecticides be preserved. If resistance
110 or tolerance to insecticides were detected in a timely fashion, management strategies could be implemented to stall or overcome this problem. Aedes albopictus is a native of the tropical and temperate regions of Asia (Gubler and Kuno 1997) and the American infestation is believed to have originated from Japan (Kambhampati et al. 1991). Because of its rece nt establishment in this county, little is known of the susceptibility status of Ae. albopictus to adulticides commonly used in the United States. A review of the suscep tibility status of Asian populations of Ae. albopictus might provide information of potential inherited insectic ide resistance or inherent insecticide tolerance in this exotic mosquito. Aedes albopictus has demonstrated some DDT resi stance in China through elevated DDT-dehydrogenase production (Neng et al. 1992). Aedes albopictus populations resistant to DDT are also known to exist in India, Thailand, Cambodia, Singapore, and the Philippines (Mouchet 1972). In Japa n, Okinawan popula tions of larval Ae. albopictus were susceptible to 8 OP insecticides including malathion, but 1 strain was resistant to DDT (15 x) (Miyagi et al. 1994). In one of the mo st comprehensive larval susceptibility studies to date, Wesson ( 1990) tested 26 strains of Ae. albopictus larvae to 5 organophosphate (OP) insecticides. Fourteen st ains tested were from America, 5 from Brazil, 5 from Japan, and 2 were from Sout heast Asia. Japanese and American strains showed tolerance to malathion and fenitrothi on, both OPs. The American strains were more tolerant to OP insecticides than thos e from other geographic regions tested, and the Chicago strain showed low-level resistan ce to malathion. Other countries having malathion-resistant populations of Ae. albopictus include Vietnam, Sri Lanka, and Singapore (W.H.O. 1986). In Texas, Ae. albopictus has demonstrated some tolerance to
111 malathion (Robert and Olson 1989, Sames et al. 1996). It remained susceptible to permethrin, resmethrin, chlorpyrifos, be ndiocarb, and naled in both studies. The goal of our study was to determine whether adult Ae. albopictus has developed resistance or tolerance to any of 5 commonly used adulticides in Florida. Materials and Methods Collection Sites Aedes albopictus eggs were collected from 6 sites located in north and central Florida (Fig. 5-1) and bred through the F4 generation to produce enough adults to complete 3 tests of each adulticide for each of the 6 sites (Robert and Olson 1989). Sites were chosen on recommendations from mosquito control personnel in Vero Beach (Indian River County), Inverness (Citru s County), Jacksonville (Duval County), Gainesville (Alachua County), Quincy (G adsden County), and Pensacola (Escambia County). These sites cover a broad geographi cal range of north a nd central Florida and each site generated excessive client complaints for their respective mosquito control district. South Florida co llecting sites were excluded due to the presence of Ae. aegypti in coastal areas and urban areas south of Vero Beach and Tampa. Eggs cannot be visually differentiated between Ae. aegypti and Ae. albopictus (Linley 1989), and separation of adults reared from mixed egg batches would have been excessively time consuming. No Ae. aegypti eggs were collected during our study. However, Oc. triseriatus was frequently reared from egg papers collected from Quincy and Gainesville. On such occasions, F-1 adults were aspirated from their cage, immobilized at 5oC on a chill table, and separated. Except for Ga inesville (Evergreen Cemetery), all sites contained waste tires or retreaded tires for sale.
112 Egg Collection Apparatus Eggs were collected from 16 oz black plastic cups (W&A Cups, Illustrated, Lancaster, TX) lined with 1 piece of thick, coarse # 76 seed germination paper (Anchor Paper Company, St. Paul, MN) cut into 10 cm x 25 cm strips. Collection site and date were recorded in ink on the upper right hand corn er of each paper. Cups were filled half way with 200 mL of tap water. Each c up had a 0.95 cm diameter hole drilled approximately midway to prevent water overflow (from rainfall). Thirty-cm nails were inserted into the ground, 1 per cup, and cups were attached with size 64 rubber bands to prevent tipping. Each site was provided with a rain shelter constructed from two 1.2 m pieces of 2.5 x 30 cm plywood boards nailed adjacen t to each other forming a 0.6 m x 1.2 m rain and sun shield. Shelters were mounted 60 cm above ground on 2.5 x 3.8 cm garden stakes. Five or 6 cups were set under each shelter as described above (Fig. 5-2). Egg papers were collected after a period of 7 to 10 days. They were allowed to dry at room temperature overnight and mailed in Ziploc plastic bags by collaborators from 5 different locations outside of Gainesville. Adult Rearing Field-collected egg papers were placed into 16 cm x 30 cm x 5 cm enamel pans and immersed in tap water. Twenty-four h later, first instar larvae were transferred to 38 cm x 50 cm x 7.6 cm plastic trays (Plastic Former In corporated, Greer, SC). Larvae were fed a 50 mL solution of 3:2 liver: brewers yeast media (20 g/L) every second day until pupae developed (usually in 5 to 6 days). Pupae a nd fourth instar larvae were transferred to 414 mL (14 fl oz) Ziploc plastic bowls and placed into cag es for emergence. Adults and larvae were maintained at 27oC, 70% RH and a photoperiod of LD 14:10 h.
113 Adults were fed 5% sugar solution as a carbohydrate source from disposable 3 oz plastic cups supplied with a cotton ball. Af ter 7 days, F-1 mosquitoes were offered a restrained chicken or human hand and allowe d to feed for 15 min (we found F-1 adults reluctant to feed on chickens on many occasi ons, but readily took to human flesh. This was not usually the case in subsequent genera tions. Approval to feed on the author was granted by the University of Florida Health Science Center Institutional Review Board Agreement IRB # 686-2003). One oviposition c up (as described above) was added to the cage and left in place for 6 or 7 days. Pape rs were removed and allowed to dry overnight in screen-covered plastic containers to prev ent escaped gravid female mosquitoes from ovipositing on them. Dried papers were placed in large Ziploc plastic bags labeled with collection site and stored in airtight plastic co ntainers until ready for us e. In this way, a sufficient number of F-4 eggs were produced to provide the necessary number of adult females for all tests. Insecticide The 5 most commonly used adulticides by Fl orida mosquito control agencies were tested against adult female Ae. albopictus; they contained the active ingredients malathion, naled, permethrin, d-phenothrin, and resmethrin (Ali and Nayar 1997). The 2 OP insecticides tested were Fyfanon ULV (96.5% malathion) (O,O-dimethyl-S-(1,2-di (ethoxycarbonyl)-ethyl) phosphor odithioate) and Dibrom 14 Concentrate (87.4% naled) (dimethyl 1,2-dibromo-2,2-dichloroethyl phosphate) manufactured and provided by Cheminova Inc. (Wayne, NJ) and Amvac Chemical Corp. (Los Angeles, CA), respectively. The 3 pyrethroid in secticides tested were Scourge Insecticide 4 + 12 (4% resmethrin) ((5-phenylmethyl-3 -furanyl) methyl 2,2-dimet hyl-3-(2-methyl-1-propenyl cyclopropane carboxylate)), manufactured by Bayer Environmental Science USA
114 (Montvale, NJ), Biomist 4 + 4 ULV (4% permethrin) (3-phenoxyphenyl) methyl (+/-) cis, trans-3-(2,2-dichlorethenyl)-2,2-dim ethyl cyclopropanecar boxylate, and Anvil 10 + 10 ULV (10% d-phenothrin) (3-phenoxybenzyl-(1R S, 3RS, 1RS, 3SR) 2,2-dimethyl-3(2-methylprop-1-enyl) cyclopropanecarboxylate) both formulated by Clarke Mosquito Control Products, Inc. (Roselle IL). The second number on pyrethroid labels refers to percent of piperonyl butoxide (PBO), a microsomal enzyme i nhibitor that synergizes the pyrethroid. Almost a ll commercially available pyrethro ids are formulated with PBO. Bioassay Test Procedure This procedure is a modification of the in secticide-coated vial technique developed by Plapp (1971) used by mosquito control pr actitioners for 25 years in Texas (J. K. Olson, personal communication). This pr ocedure was modified by using formulated adulticides as a source of a.i. instead of t echnical grade insecticide. The coated-vial technique test assesses the toxicity of vari ous insecticides to adult females of any mosquito species. The insecticide in question is serially diluted with reagent-grade ACS (American Chemical Society) acetone until a de sired range of concentr ations is obtained. Specific amounts of this inse cticide and acetone solution are added to 20 mL glass scintillation vials (Wheaton Science Products, Millville, NJ) and manually rolled until all acetone evaporates depositing an evenly coat ed film of insecticide formulation on the inside surface of the vial. The concentrati on of insecticide per vial is recorded as micrograms a.i. per vial (g/vial). Determination of baseline ranges was obtained empirically by evaluation of susceptible colony populations. Th e range included concentrati ons that killed few if any adults through concentrations that killed mo st or all adults. Adult mosquitoes were exposed to insecticide-treated vials for 24 h af ter which mortality rates were recorded to
115 calculate LC50 and LC95 values. Field samples of the sa me mosquito species were then collected and tested using the same or hi gher concentrations (as necessary) of the insecticide under investigation. Susceptibility of field popul ations was then determined by dividing their LC50 by the colony LC50. This test routinely relies on LC50 values, not LC95 values, to determine susceptibility of test populations. Response at the LC50 level is more reliable than at other LC levels as vari ability in the 95% confid ence interval (CI) is at a minimum (Simon Yu, University of Flor ida Toxicologist, pers onal communication). The derived quotient yields a RR used to de termine whether resist ance or tolerance was present. Resistance ratios approaching 10 we re indicative of an insecticide-tolerant mosquito population while populations with RRs exceeding 10 were considered resistant to the insecticide in ques tion (Sames et al. 1996). Insecticide formulations were serially d iluted with acetone until a desired range of concentrations were obtained for testing. Appendix C shows calculations used to dilute insecticide into end-use concentrations. Te st solutions were prepared after consulting insecticide labels for insecticide strength. Labeled rates of pounds of active ingredient per gallon were converted to milligrams pe r milliliter. An appr opriate quantity of insecticide was pipetted with Oxford BenchmateTM 200 l to 1000 l digital pipettors (Sherwood Medical, St. Louis, MO) into 20 mL glass scinti llation vials to obtain a 1,000 mg or 100 mg base sample of a.i. Acetone was added to bring the total solution volume to 10 mL. The base quantity insecticide was then serially diluted 1:10 with acetone to provide a range of concentrations differing 10-fold in strength. Insecticide-coated vials we re labeled with agent and g/vial with an indelible ink marker. Six or occasionally 7 concentrati ons of a given insec ticide were used in
116 increments of 0.01-, 0.03-, 0.06-, 0.1-, 0.3-, and 0.6 g/vial. The range of concentrations most commonly used were 0.03-, 0.06-, 0.1-, 0.3-, 0.6-, and 1.0 g/vial, but higher and lower ranges were sometimes used for diffe rent insecticides de pending on mortality response. Insecticides in herently more toxic to Ae. albopictus (e.g., permethrin) were tested at lower ranges than other less toxic in secticides (such as d-phenothrin). Higher or lower test rates were always kept at the same interval (e.g., 0.06to 3.0 g/vial or 0.01to 0.06 g/vial, as necessary). When less than 300 l of insecticide solution was needed for a particular vial, an extra amount of acetone was added to bring the total volume of solution to 300 l. This insured similar coating of all inside surfaces of the vial. Five vials were prepared for each concentra tion. Additionally, 5 control vials were produced with acetone alone. Testing 6 concen trations of an inse cticide thus required preparation of 35 vials (30 with insecticide and 5 controls). The vials were hand-rolled under a fume hood until dry, placed back into their original part itioned holding box and the front of the box was labeled with collection site, insecticide, range of concentrations, and date of preparation (Fig 5-3). Vials were left under the fume hood overnight to ensure complete removal of acetone residue. The following day, vials were supplemented with a 0.75-cm2 piece of blotter paper soaked in 5% sugar solution for a source of carbohydrate. Three to 7-day-old adult female Ae. albopictus were collected from cages with a mechanical aspirator (Hausherrs Machine Works, Toms River, NJ) and lightly narcotized with CO2 flowed at the rate of 500 mL/min. Adults were laid on white paper and separated by sex (Fig. 5-3). Five adult females were added to each vial with forceps for a total of 25 adults per concentration, 175 per test (35 vials x 5 mosquitoes each). Care was taken to prevent injury to
117 mosquitoes during transfer by grasping only le gs or wings. A medium sized cotton ball was used to seal the tops of the vials a nd then covered with a damp paper towel to provide adequate humidity for the duration of the trial. Test mosquitoes were placed into a Forma Scientific environmental chamber (T hermo Electron Corp., Franklin, MA) set at 28oC and left overnight. Twenty-four h later, mortality was recorded for all vials. Mosquitoes were counted as dead if f ound lying on their backs and unable to right themselves after several gentle taps on the vial. For every test of F-4 field adults, 1 test with colonized adults was run concurrently with the same insecticide preparation. Gene rally, 2 or 3 field te sts were run with 1 colony test. Baseline colony LC50s of each trial were compared against known colony LC50s as a check against dilution error in vial preparation. Tests in which a colony LC50 exceeded the average baseline colony LC50 by more than100% were rejected. Analysis of Data Mosquito dose-response mortality was dete rmined using PC Probit (Finney 1971). This program provides an analysis of dose re sponse using the common logarithm of dose. Mortality data is calculated into LC intervals of 10 beginning with LC10 and ending with LC90. Each LC output included upper and lower 95% CIs. The slope and intercept of the probability units (probit) analysis curve wa s generated. A Chi-square goodness-of-fit value was calculated by PC Probit to determine the validity of the test. Tests scoring a Chi-Square p-value greater than p > 0.05 were accepted; those scoring less than p > 0.05 were rejected. At a significant level of p > 0.05, all confidence limits were calculated with a t-value of 1.96. The LC95 value and CIs of a given te st, commonly used as a high point in insecticide susceptibility tests, was obtained using the SAS Probit Analysis Program (SAS 2001). PC Probit did not generate these values.
118 Abbotts corrected mortality formula, embedded into an Excel spreadsheet (Microsoft Corporation, Redmond, WA) was us ed to calculate cont rol mortality at the conclusion of each test (Abbott 1925). It was generated by subtracti ng the percentage of living insects in the treatment from the percenta ge of living insects in the control, divided by the percentage of living insects in the control, and multiplying the quotient by 100. Abbotts formula was applied to all tests with any control mortality. Tests with control mortality under 10% were accepted (Simon Yu personal communication). With 25 adult mosquitoes comprising a control sample, contro l mortalities of 3 or more mosquitoes (> 10%) resulted in rejection of the test. As mentioned earlier, 3 field tests were r un concurrently with 1 colony test for each site and insecticide combination. Field sample LC50s were derived from the mean of these tests. Overlap of LC50 95% CIs among all results of a particular test was a criterion used for obtaining means. Field tests in wh ich these CIs did not ove rlap were excluded from analysis (Table 5-2). Results Results of insecticide su sceptibility bioassay tests on colony populations of Ae. albopictus for each insecticide are summari zed in Table 5-1. Slopes, LC50s, and LC95s with corresponding 95% CIs are given. Co lony toxicity comparisons against field populations are given in Tabl es 5-2 through 5-7. The LC50 values in colony adults ranged from a low of 0.037 g/vial for permethrin to a high of 0.358 g /vial for d-phenothrin. Naled (0.084 g/vial), malathion (0.153 g/vial), and resmethrin (0.235 g/vial) were intermediate in toxicity to colony mosquitoes. Table 5-2 gives LC50 and LC95 results for Ae. albopictus collected from Citrus County, Florida. This site was a commercial filling station that sold used tires in
119 Inverness. Tires were stacked in rows on the ground and exposed to rainfall. Over the past 3 years, the site was frequently treated with malathion, permethrin, and d-phenothin by Citrus County Mosquito Control Dist rict (MCD) due to large numbers of Ae. albopictus produced there. The Citrus County ma lathion mean (from 3 field samples) was 0.509 g/vial. The susceptible colony malathion LC50 was 0.153 g/vial giving a RR of 3.33, highest among all pesticides te sted on populations of Citrus County Ae. albopictus. The permethrin RR was 3.02 with a mean LC50 of 0.112 g/vial, the highest mean permethrin LC50 of all 6 populations examined in our study. Naled had a RR of 0.64, d-phenothrin 0.99, and resmethrin 0. 92. Thus, susceptible colony naled, dphenothrin, and resmethrin LC50s were approximatel y equal to field LC50s. Insecticide susceptibility results on fi eld populations collected from Quincy, Gadsden County, Florida are given in Table 53. This collection site produced large numbers of Ae. albopictus; 3-min landing rates often exceeded 100 adults. Due to the ease in catching large numbers of adult female Ae. albopictus, field-collected adults were used to start the test colony (as opposed to initiating the colony from field-collected eggs). Tires sold from this facility were stored in piles outside the work office and exposed to weather. At leas t several hundred tires were on s ite during each of 4 visits over a 1 year period. Gadsden County MC trea ted this site with malathion approximately once per week during the study period. Mala thion had a low RR of 1.07. Naled and permethrin had the highest RRs, 2.32 and 2.30, respectively. Resmethrin and dphenothrin RRs were 0.87 and 1.16 respectiv ely, equally susceptible as the colony populations.
120 Indian River County field samples were colle cted (as eggs) from a tire dump at an inactivated WW II airfie ld within the city limits of Vero Beach. The tire dump has been cleaned up, but scattered tires exist under dense vegetation that provides an ideal breeding habitat for Ae. albopictus. The site was located within half a mile of the Indian River MCD and residential neighborhoods. It was co-located with the Vero Beach recycling centers turn-in point. Indian River MCD provided both ground and aerial applications of ULV adulticid es to control large populations of salt marsh and citrus grove mosquitoes. Permethrin was applie d from ground-mounted machines, naled was applied aerially. Naled had a RR of 2.71, th e highest among insecticides tested on Vero Beach populations (Table 5-4). The permethr in RR was 2.05. D-phenothrin, resmethrin, and malathion had RRs of 1.26, 2.21, and 1.95, respectively. The Pensacola collection site was a wa ste tire dump located in a midtown residential area. An undeveloped 5-acre lot wa s used as an illegal dumpsite for tires and other waste before the property could be secure d with a chain linked fe nce. The site was located next a vehicle salvage lot and bot h sites produced landing rates of 40 to 50 Ae. albopictus per min. The site was a constant source of trouble for Escambia County Mosquito and Rodent Management Divisi on and a permethrin ULV treatment was provided once a week throughout the summer of 2004. Malathion was used at this site before 2004. The mean malathion LC50 in field mosquitoes collected from this site was 0.463 g/vial with a RR of 3.40, the second highest malathion LC50 and RR of all sites sampled (the Citrus County malathion LC50 and RR was higher). Naled had a RR of 2.18 and resmethrin, d-phenothrin, and permet hrin had RRs of 2.14, 1.39, and 1.05, respectively.
121 Three insecticides were tested against Jacksonville, Duval County, populations of Ae. albopictus: malathion, permethrin, and d-phenothr in. A test colony was started from field-collected adults after an initial colony begun from e ggs failed. The site was a heavily forested salvage yard in a rural area east of town Jacksonville MC routinely surveyed and treated the site due to homeowne r complaints in the vicinity of the site. Most tires in the yard were removed by the city but many water-holding containers remained and landing rates in late summer were as high as 20 adult females per min. Ground ULV applications at th e site included permethrin and resmethrin. Naled was applied aerially. Resistance ratios for permet hrin, d-phenothrin and malathion were 1.78, 1.92, and 1.19, respectively (Table 5-6). Lethal concentration 50s for permethrin were 0.063, 0.064, and 0.070 g/vial with a mean of 0.066 (RR 1.78), well below levels considered resistant (RR of 10 x). The final site surveyed in our study wa s Evergreen Cemetery in Gainesville, Alachua County, Florida. This is one of the older, larger cemeteries in Gainesville and often visited due to its hist orical significance and aesthe tic surrounding. The City of Gainesville MC (public works mosquito co ntrol) had received enough complaints of nuisance populations of biting mosquitoes (mostly Ae. albopictus, personal observation) from visitors that routine weekly applicati ons of malathion were initiated during summer months. Mean LC50s and (RRs) of Ae. albopictus to malathion, permethrin, and dphenothrin were 0.235(1.54), 0.031(0. 84), and 0.555 g/vial (1.55), respectively (Table 5-7). Discussion The coated-vial bioassay test was origina lly developed by F. W. Plapp to assess insecticide resistance in Heliothis virescens, the tobacco budworm, in the Midwest
122 United States (Plapp 1971). It was later modi fied to monitor insecticide resistance in mosquitoes as a service to state vector control agencies provided by the Texas A&M Mosquito Research Laboratory, and has been in place since the early 1980s (J. K. Olson, personal communication). The dos e-response bioassay test, su ch as the one used here, allowed for computation of a large range of lethal doses, their confidence intervals, and slopes. Slopes of probit-generated dosage-mortality curves determine the degree of homogeneity or heterogeneity of a test populati on. Larger slopes (> 2.0) indicate a more homogeneous population, low slopes (< 2.0) indicate a heterogeneous population in which some individuals of the test population are more tolera nt of an insecticide than other individuals in the same test. Slopes obtained in our study t ypically ranged between 1.7 and 2.5 (Table 5-1), but slope s as steep as 4.9 were seen in some tests, usually with permethrin. A test populations LC50 and its 95% CIs were compared against 2 other tests, all with the same population and pesticide exposure, for overlap. Overlap of these intervals ensures continuity of the tests; only those tests with overlapping 95% CIs were used in determining the mean LC50 for a test population of mosquitoes from a given location for each insecticide used. The LC50 was used to determine the presence of resistance, as opposed to the LC95. The logical reason for comparisons at this level is that an insects response to toxicants are less vari able at the dose producing an LC50 than at any other LC dose. Stated another way, the fiducial limits of the probit-generated dose-mortality curve are narrowest at the LC50 level, giving the most accurate interpretation of dose-response data, whereas LC95 fiducial limits are extremely large in comparison and often overlap
123 despite potential tremendous differences between LC95 values (Simon Yu, personal communication). Another method used to determine suscepti bility in adult mosquitoes is the CDC bottle bioassay test that measures the timemortality response in test populations (Brogdon and McAllister 1998b). This test detects resistance in field mosquitoes quickly. Advantages of the time-mortalit y test include resistance detec tion in less than 2 h and use of registered pesticides making preparation by mosquito control practitioners practical (with use of off-the-shelf insecticides availabl e at the work location). A single diagnostic dose (2 x LC95 or 2 x LC99 of a susceptible species) is used as a benchmark in determining resistance. Far fewer mosquitoes are required to obtain results with the timemortality method compared with the coated-v ial technique. Disadvantages of the timemortality technique are that dose-mortality LC s, CIs, and RRs are not generated, all of which are useful in determining the degree of resistance severity and for quantitative comparisons between different populations of the same mosquito species. The CDC bottle bioassay has been used to monitor re sistance in field mosquitoes by PHEREC throughout Florida since 1999 (Pet ersen et al. 2004). Our st udy used the coated-vial technique to assess insec ticide susceptibility in Fl orida field populations of Ae. albopictus by comparing LC50s, LC95s, and CIs of field populati ons to susceptible colony populations. Results shown in Tables 5-1 through 57 indicate that neither tolerance or resistance was present among the 6 field populations of Ae. albopictus to any of the insecticides examined over the c ourse of our study, at the LC50 level. On 2 occasions, field populations had LC95 values exceeding 10 x, however, corresponding LC50 values
124 for these test populations had RRs < 4 x, indicating that these populations were heterogeneous, but still susceptible, to the insecticide in question. Populations exposed to high selection pressure with malathion (Citrus and Escambia Counties) had elevated RRs of 3.33 and 3.03, respectively, well below the RR threshold of 10 x, but approaching a level of tolerance (4 x) as defined by Brown and Pal (1971). Escambia County Ae. albopictus had a malathion LC95 of 35.34 g/vial, 55 x higher than the LC95 of the susceptible colony strain. These data tend to support the idea that part of that population was resistant to malathion, though this was not evident at the LC50 level. Other field populations of Ae. albopictus had malathion RRs ranging between 1.07 (Gadsden county) and 1.95 (Indian River County). The Gadsden County site was purportedly treated with malathion once per week in the summer month of 2003, but the field LC50 was only slightly higher than the colony LC50 (0.163 g/vial and 0.153 g/vial, respectively). Adult Ae. albopictus from Houston and New Orleans have shown low levels of resistance to malathion (Robe rt and Olson 1989, Khoo et al. 1988). Sames et al. (1996) tested the susceptibility of Ae. albopictus to commonly used adulticides in field populations collected from the Lower Rio Gr ande Valley in Texas and Mexico. The TAMU colony strain malathion LC50 was 0.130 g/vial, very similar to the LC50 of the USDA ARS colony (0.153 g/vial). Field populations of Florida Ae. albopictus were more susceptible to malathion than Texas populations, which had an LC50 of 0.700 g/vial (Sames et al. 1996), still higher than the Citrus County population LC50 of 0.509 g/vial (highest of all Flor ida populations sampled). Robe rt and Olson (1989) obtained
125 LC50s of 0.150and 0.130 g/vial for 2 field populations of Liberty County and Houston Ae. albopictus, respectively, much lower than LC50s found in the Rio Grande Valley. At the time of those findings (1989), Ae. albopictus had been present for only 2 or 3 years in upper Texas Gulf Coast. The higher LC50s found in Rio Grande Valley populations of Ae. albopictus may have been due to several extra y ears of exposure to malathion in this population at the time of that study (1996), as Ae. albopictus is known to have occurred in the Rio Grande Valley since at least 1990 (F rancy et al. 1990). Malathion was also the most commonly used agricultu ral insecticide in Texas in 1995, and farms in the Rio Grande Valley undoubtedly received their fair share of it to control pests of citrus and vegetables, the 2 biggest cas h crops in the valley ( http://www.texasep.org/html/pes/pes _2tex.html ). Intense use of agricultural pesticides in the vicinity of mosquito breeding grounds has been shown to induce resistan ce in anopheline and cu licine mosquitoes (Mouchet 1988), and it is certainly possibl e that Rio Grande Valley populations of Ae. albopictus had been exposed as a direct result of agricultural pest control operations. Larval susceptibility studies revealed that the Florida Medical Entomology Laboratory (FMEL) Ae. albopictus colony was susceptible to all of 5 OP insecticides except malathion, which produced an LC90 of 1.043 ppm. As such, it was considered to be tolerant (Ali et al. 1995). Examination of larvae from one of the first infestations discovered in Kentucky demonstrated that the sample was as susceptible to malathion as a susceptible Hawaiian strain (Cilek et al. 1989). Aedes albopictus larvae from Mobile, Alabama, were recently found to be tolerant of malathion (RR of 3 x) (Liu et al. 2004, in press), although larval tolerance as defi ne by the W.H.O. begins at RRs of 10 x in mosquito larvae (Brown and Pal 1971). T hus, limited testing of larvae and adult Ae.
126 albopictus from scattered locations within the United States seems to indicate that some tolerance or resistance is deve loping to malathion. This wa s probably true of the founder populations as they were introduced into the United States from infested Japanese and Korean tires (W.H.O. 1986, Reiter 1998). A second OP insecticide, naled, was test ed against colony and field populations of Ae. albopictus. The USDA ARS susceptible colony had a naled LC50 of 0.082 g/vial and an LC95 of 0.397 g/vial. The slope generate d from Probit PC was 2.395, indicating a homogenous response of the susceptible colony to this insecticid e. Resistance ratios ranged from a low of 0.64 (Citrus County) to a high of 2.71 (Indian River County). Thus, all samples tested with naled were susceptible. In the only other report on Ae. albopictus response to naled, Robert and Olson (1989) obtained LC50 and LC95 values of 0.07and 0.35 g/vial in a Houston strain, respectively. These results are very similar to the ones obtained in our study for naled (Table 5-1). Naled is sometimes used by larger MCDs that employ aircraft to treat sizeable tracts of land for floodwater mosquitoes, such as Indian River MCD. Naled is extremely corr osive to aluminum and seldom applied by truck-mounted ULV machines. Advantages of using naled include quick knockdown of mosquitoes and rapid degradation into harmless byproducts. Minimizing the exposure time of a pesticide to a pest is one way of reducing selection pressure for resistance. Organophosphate resistance in Aedes (Stegomyia) mosquitoes has been slow to develop. The extensively studied yellow fever mosquito, Ae. aegypti (closely related to Ae. albopictus), was not found to be resistant to OP insecticides until the early 1970s when tolerance to several OP compounds was seen in Vietnamese populations (Mouchet 1972). As late as 1970, 7 Indian strains of Ae. aegypti were still susceptible to 5 OP
127 insecticides (including malath ion) despite routine use of these compounds for better than 10 years in areas from which test larvae we re collected (Madhukar and Pillai 1970). In spite of placing intense selection pressure on these mosquitoes for 20 generations, their strains of Ae. aegypti only developed a 2 x to 6 x increased tolerance to OPs. In contrast, Culex mosquitoes demonstrated resistance to malathion as early as 1957 (Gjullin and Isaak 1957). The first report of low levels of malathion resistance in Ae. albopictus came from Vietnam in 1969, when larvae collect ed from the demilitarized zone (DMZ) exhibited an increased toleran ce to this insectic ide (Stasiak et al. 1970). The author attributed this development to the application of over 45,000 gallons of 57% and 95% malathion in the affected area during the pr evious 12 months as part of an on-going malaria control program. Despite its use in vector c ontrol operations for almost 50 years, serious malathion resistance ( 100 x) in Aedes (Stegomia) mosquitoes has yet to appear. In only 2 cases did American Ae. albopictus strains exceed a RR of 10 x (Wesson 1990, Cilek et al. 1989). Wesson observed a 22 x RR in 1 Chicago strain of Ae. albopictus; all others were susceptible to malathion. It should be noted that the Chicago strains RR was determined at the LC95 level. The population was highly he terogeneous, with a log dose-probit mortality slope of 1.32. Using the topical application method, a RR of approximately 20 was observed in a Houston strain at both the LD50 and LD95 levels (Khoo et al. 1988). In no other case did LC50 RRs exceed 10 x (Cilek et al. 1989, Robert and Olson 1989, Sames et al. 1996, Liu et al. in press). Samples tested in th ese last 4 studies represented strains from a broad geographic area (T exas, Kentucky, Alabama, and Florida).
128 The 3 pyrethroids used in our study, pe rmethrin, d-phenothrin, and resmethrin, were commercial formulations commonly used by Florida vector control agencies. Pyrethroids have been used for mosquito contro l in the United States since the late 1970s and offer some important advantages over OP insecticides. Advantages include different target sites, effectiveness at extremely low doses (often in oz/acre as compared with lb/acre for many OP insecticides), rapid knoc kdown, and lower mammalian toxicity than most OPs (Ware 1983). Pyrethroid insec ticides offer both long-term residuals (permethrin, persistence to 3 weeks) and short-term resi duals (resmethrin, d-phenothrin, persistence < 24 h). These adul ticides are formulated with piperonyl butoxide (PBO), a mixed function oxidase (MFO) enzyme inhib itor added to counter MFOs present in mosquitoes. Mixed function oxidases serve to make xenobiotics more polar, and thus more rapidly excretable, in mammals a nd arthropods (Wilkinson 1983). Due to the presence of PBO in the pyrethroid tests conduc ted in our study, there was a chance that resistance otherwise exposed by testing with the a.i. alone might be masked. However, the results given here reflect the actual effec tiveness of registered adulticides currently used in the field to control Ae. albopictus in Florida, which was the goal of our study. Permethrin proved to be the most t oxic of all adulticides tested; the LC50 of the susceptible colony was 0.037 g/vial and field population LC50s were usually less than 0.1 g/vial. Resistance ratios were between 1 and 2, only the C itrus County population had a RR exceeding 3 x. In every case, LC95 RRs were less than LC50 RRs, indicating a fairly homogenous response of Ae. albopictus to this insecticide. The Gainesville field LC95 was only half of the colony LC95, an unusual finding and suggests that this
129 population had no previous exposure to permethrin Permethrin was used at test sites in Jacksonville, Citrus County, Indian River County, and Escambia County. D-phenothrin is a nonresidual pyrethroid th at is rapidly degraded by ultraviolet radiation (sunlight). It wa s the least toxic compound (LC50 = 0.358 g/vial) to the susceptible colony (Table 5-1). The slope wa s less than 2.0, indicating heterogeneity in the populations response to this pesticide. Dphenothrin is the newe st of all adulticides tested in our study, only Citrus County fi eld populations were exposed to it. No published literature was found assessing Ae. albopictus response to this insecticide. Recently, 2 susceptible colonies of Culex quinquefasciatus and Ochlerotatus taeniorhynchus were tested with d-phenothrin (P etersen et al. 2004). Log dose probitmortality generated slopes were 1. 53 and 1.65, respectively, and LC50 and LC95 values were considerably higher in these species than in Ae. albopictus. Aedes albopictus colony response to resmethrin was the anomaly in our study. Previous work with susceptible Ae. albopictus colonies indicated that this species in highly susceptible to resmethrin, with LC50 values similar to permethrin (Sames et al. 1996). Considering the slope of the susceptibl e colony (1.348), it is likely that this colony had experienced previous resmethrin expose prior to its establishment as a USDA ARS colony in 1992. The USDA ARS colony LC50 was 5.9 x higher than the TAMU susceptible colony (0.235 g/vial and 0.04 g /vial, respectively). A similarly high LC50 response was observed in a FMEL susceptible co lony from Vero Beach, Florida, thus, the USDA susceptible colony was used for all resmethrin comparisons. Due to these results, a second sample of Scourge insecticide was obtained from the Bayer Environmental Science Division as a check against a bad samp le, but similar results were obtained with
130 that sample. Regardless, only on 1 occasion did a RR exceed 10 x (LC95, Citrus County). The LC50 RR for that site was less than 1 x. Apparently, a proportion of the Citrus County sample population is resistant to resm ethrin (heterogeneous), but the population as a whole is still susceptible to this adulticide (LC50 results). Overlap of field and susceptible colony LC50 CIs indicate similar susceptibilities between those populations to a given insecticid e. A lack of overlap in CIs indicates some degree of insecticide tolerance or re sistance, depending on the magnitude of the difference. Resistance ratios have historically served as a quick check for tolerance or resistance between test populat ions, but lack statistical si gnificance. Resistance ratios exceeding 10 x (resistance) and 4 x (tolerance) at the LC50 level for all test samples of Ae. albopictus indicated that it is still susceptible to these 5 adulticides, despite almost 20 years of exposure (Jacksonville, Gainesville, and Pensacola) (OMear a 1993) to at least 2 of these compounds (malathion, resmethrin). This supports earli er statements that Aedes (Stegomyia) mosquitoes are slow to devel op resistance (Madhukar and Pillai 1970). Although the precise biochemical mech anism for elevated tolerance in field populations (RR to 3.3 x in 1 case) is unknown, it is likel y that the mosquitos response to these insecticides involves metabolic enzymatic detoxification with hydrolyases, MFOs and/or glutathion S-transferases (GSTs) (Brogdon and McAllister 1998a). Metabolic resistance often pertains to increased production of MFOs and GSTs. Both enzymes are ubiquitous in aerobic or ganisms (Wilkinson 1980, Enayati et al. 2005) and have evolved within insect s (and other animals) to detoxify environmental, animal (host), and plant toxicants. They make foreign lipophilic compounds more water soluble (polar) before accumulation in fat bodies and tissues can poison the insect. By-products
131 from interaction with these enzymes are dire ctly excreted or c onjugated with endogenous sugars, amino acids, sulfates, or phosphate s and excreted (Wilkinson 1983). These enzymes are inducible (increased enzyme s ynthesis) in both sus ceptible and resistant insects experiencing chemical stress (Plapp and Wang 1983). Resistance ratios at the LC50 and LC95 levels not exceeding 4 x, as seen in the majority of field samples tested in our study (Tables 5-2 throu gh 5-7), are likely the results of background environmental stresses as sociated with living in field environments as opposed to protected laboratory environs. Diet (allelochemicals and host immune response) and breeding site s (waste tires) may suppl y xenobiotic and chemical compounds that induce increased metabolic production of detoxification enzymes in Ae. albopictus. However, as stated earlier, sample sites used in our study were specifically chosen because they were under high selection pr essure with adulticides. In light of this, it is not surprising to find LC95 RRs exceeding 10 x in 2 field populations of Ae. albopictus to resmethrin (Table 5-2) and malathion (Table 5-5). The author believes this was due to routine treatment of these 2 sites with Scourge (resmethrin) and Fyfanon ULV (malathion) adulticides, 2 commonly used insecticides for adult mosquito control since 1986, when Ae. albopictus first arrived in Florida. Mosquito abatement control efforts th at rely heavily on 1 control method sometimes experience control failure. This is true of any pest control effort and was the impetus for developing integrated pest manage ment programs. Routine reliance on just 1 or 2 pesticides can quickly l ead to selection for resistan ce to those compounds as has been demonstrated in Florida mosquitoes with DDT and several OP insecticides (Breaud 1993). Rotation between different classes of insecticides can stem or reverse the
132 development of resistance (Plapp and Wang 1983 ), but ultimately, different methods of control are necessary to alleviate selection pressure s brought on by continuous reliance on pesticides. Aedes albopictus prefer to breed in artificial containers, especially tires. Tire piles are sometimes treated with larvicides to control Ae. albopictus. Several mosquito larvacides that have different modes of acti on from OP and pyrethroid insecticides (i.e., microbial insecticides, insect growth regulators, surface oils, and chitin synthesis inhibitors) would make excellent rotation insecticides for OP or pyrethroid tolerant (or resistant) populations of Ae. albopictus. Source reduction is probably the best method of quickly reducing nuisance populations of this mosquito (Watson 1967, Hawley 1988, Estrada-Franco and Craig, 1995). Fewer Ae. albopictus are produced from natural breeding sites than from artificial containers. Except for Evergreen cemetery in Gainesville, all study sites had tires producing excessive numbers of this pest. Tire cleanup is necessary where waste piles exist, and tire shredding is the preferred corrective measure because sh redded tires cannot hold water. Shredded tires can be mixed with macadam to improve road qua lity or disposed of in landfills. The Solid Waste Act of 1988 adds a small ta x to all tires sold in Florida, and a portion of this money goes to fund Mosquito Control/Waste Tire Abatement Grants provided to MCDs. While most large waste tire sites have been eliminated from the state since the passage of this act, many smaller (often illegal) dumps still exist around the state and property owners of such dumps ar e reluctant to remove tires due to costs associated with cleanup and turn-in at recyc ling centers (greater th an $2/tire at 1 study
133 site). The Florida Department of Environmental Protection allows storage of up to 1,500 tires on dealer sites ( http://www.dep.state.fl.us/wast e/quick_topics /publications /shw/tires/tires.pdf ) and requires that tires to be main tained in mosquito free condition, but does not state how this should occur. Tire piles of less than 1,500 resulted in landing/biting rates a pproaching 50/minute in Gadsden C ounty, thus, small retail sites can produce considerable problem s for their respective MCDs. Mosquito control districts must ultimat ely control mosquitoes and prevent or mitigate insecticide resistance a nd it will be their responsibilit y to educate the proprietors of retail tire businesses in preventive techni ques. Public educati on is a key preventive strategy in reducing conditions conducive to Ae. albopictus breeding. Elimination of these breeding sites in Flor ida by an educated public will reduce our reliance upon insecticides for control, lessen selection pr essure, and help to delay or reverse the development of insecticide resistance in this pest.
134 Table 5-1. Baseline insecticide susceptibili ty bioassay results for adult females of a colonized USDA ARS strain of Aedes albopictus. n = 150. Insecticide Slope SE LC50 (g/vial) 95% CI (g/vial) LC95 (g/vial) 95% CI (g/vial) Malathion 2.485 0.12 0.153 0.118-0.197 0.666 0.469-1.123 Permethrin 2.183 0.13 0.037 0.027-0.049 0.261 0.149-0.804 Resmethrin 1.348 0.09 0.235 0.156-0.413 3.894 1.526-23.883 D-phenothrin1.943 0.11 0.358 0.266-0.500 2.513 1.448-6.345 Naled 2.395 0.14 0.082 0.062-0.106 0.397 0.267-0.773
135Table 5-2. Insecticide susceptibility resu lts for Inverness, Citrus County, Florida and USDA ARS colony pop ulations of adult f emale Aedes albopictus. Insecticide Field LC50 (95% CI)1 Field LC95 (95% CI) Colony LC50 Colony LC95 RR50 4 RR95 ( g/vial) ( g/vial) ( g/vial) ( g/vial) Malathion 0.394 (0.285-0.586) 3.654 (1.890-11.768) 0.153 (0.118-0.197) 0.666 (0.469-1.123) 3.33 7.02 0.692 (0.502-1.008) 6.553 (3.510-18.494) 0.440 (0.324-0.612) 3.826 (2.248-8.763) Mean2 0.509 4.678 Permethrin 0.107 (0.082-0.142) 0.580 (0.369-1.202) 0.037 (0.027-0.049) 0.261 (0.149-0.804) 3.02 1.74 0.116 (0.089-0.152) 0.564 (0.368-1.128) 0.048 (0.037-0.062)3 0.219 (0.149-0.417) Mean 0.112 0.454 Resmethrin 0.237 (0.166-0.337) 3.184 (1.712-8.661)3 0.235 (0.156-0.413) 3.894 (1.526-23.883)0.92 27.01 0.163 (0.077-0.406) 75.585 (8.780-28,062) 0.250 (0.120-0.794) 134.915 (12.579-125,834) Mean 0.217 D-phenothrin 0.341 (0.239-0.498) 5.121 (2.604-15.537) 0.358 (0.266-0.500) 2.513 (1.448-6.345) 0.99 1.60 0.288 (0.202-0.414) 3.990 (2.086-11.465) 0.439 (0.330-0.592) 2.942 (1.822-6.251) Mean 0.356 4.018 Naled 0.063 (0.047-0.093) 0.468 (0.217-4.809) 0.082 (0.062-0.106) 0.397 (0.267-0.773) 0.64 1.93 0.042 (0.014-0.067) 1.004 (0.265-19,762) 0.173 (0.123-0.506)3 0.825 (0.315-78.772) Mean 0.053 0.766 1LC = Lethal Concentration, CI = Confidence Interval. 2Mean of 2 or 3 field trial LC50s. 3Data excluded from mean, no overlap in confidence intervals with other field results. 4RR = resistance ratio between field and colony LCs. LC50 = field LC50/colony LC50, LC95 = field LC95/colony LC95.
136Table 5-3. Insecticide susceptibility re sults for Quincy, Gadsden County, Florida an d USDA ARS colony populations of adult fem ale Aedes albopictus. Insecticide Field LC50 (95% CI)1Field LC95 (95% CI) Colony LC50 Colony LC95 RR50 4 RR95 ( g/vial) ( g/vial) ( g/vial) ( g/vial) Malathion 0.129 (0.098-0.169) 0.711 (0.471-1.347) 0.153 (0.118-0.197) 0.666 (0.469-1.123) 1.07 1.14 0.174 (0.132-0.226) 0.801 (0.552-1.294) 0.187 (0.147-0.239) 0.757 (0.531-1.293) Mean2 0.163 0.756 Permethrin 0.096 (0.075-0.125) 0.580 (0.369-1.202) 0.037 (0.027-0.049) 0.261 (0.149-0.804) 2.30 1.98 0.079 (0.059-0.106) 0.564 (0.368-1.128) 0.079 (0.059-0.106) 0.411 (0.268-0.826) Mean 0.085 0.518 Resmethrin 0.223 (0.133-0.500) 10.342 (2.578-249.222) 0.235 (0.156-0.413) 3.894 (1.526-23.883)0.87 1.61 0.251 (0.197-0.317) 0.888 (0.642-1.468)3 0.139 (0.091-0.202) 2.207 (1.087-8.042) Mean 0.204 6.275 D-phenothrin 0.554 (0.398-0.795) 5.504 (2.935-16.217) 0.358 (0.266-0.500) 2.513 (1.448-6.345) 1.16 1.48 0.343 (0.231-0.507) 2.718 (1.473-8.683) 0.352 (0.253-0.489) 2.942 (1.822-6.251) Mean 0.416 3.721 Naled 0.188 (0.142-0.249) 1.116 (0.658-2.824) 0.082 (0.062-0.106) 0.397 (0.267-0.773) 2.32 3.52 0.209 (0.161-0.273) 1.057 (0.615-2.961) 0.173 (0.121-0.245) 2.024 (1.079-5.988) Mean 0.190 1.399 1LC = Lethal Concentration, CI = Confidence Interval. 2Mean of 2 or 3 field trial LC50s. 3Data excluded from mean, no overlap in confidence intervals with other field results. 4RR = resistance ratio between field and colony LCs. LC50 = field LC50/colony LC50, LC95 = field LC95/colony LC95.
137Table 5-4. Insecticide susceptibility re sults for Vero Beach, Indian River County, Florida and USDA ARS colony populations of adult female Aedes albopictus. Insecticide LC50 (95% CI)1 LC95 (95% CI) Colony LC50 Colony LC95 RR50 4 RR95 ( g/vial) ( g/vial) ( g/vial) ( g/vial) Malathion 0.282 (0.212-0.382) 1.811 (1.122-3.898) 0.153 (0.118-0.197) 0.666 (0.469-1.123) 1.95 2.07 0.299 (0.229-0.395) 1.575 (0.937-3.960) 0.312 (0.232-0.432) 0.757 (0.531-1.293) Mean2 0.298 1.381 Permethrin 0.074 (0.060-0.092) 0.225 (0.162-0.408) 0.037 (0.027-0.049) 0.261 (0.149-0.804) 2.05 0.99 0.076 (0.060-0.097) 0.293 (0.183-0.738) 0.077 (0.056-0.106) 2.253 (1.322-5.456)3 Mean 0.076 0.259 Resmethrin 0.636 (0.363-1.632) 32.002 (6.647-1,644) 0.235 (0.156-0.413) 3.894 (1.526-23.883)2.21 7.21 0.507 (0.295-0.959) 30.642 (7.983-700.220) 0.416 (0.242-0.735) 21.613 (6.267-356.809) Mean 0.520 28.086 D-phenothrin 0.478 (0.336-0.696) 5.732 (2.912-18.703) 0.358 (0.266-0.500) 2.513 (1.448-6.345) 1.26 1.54 0.323 (0.225-0.459) 3.714 (2.012-10.696) 0.553 (0.433-0.708) 2.223 (1.514-4.260) Mean 0.451 3.890 Naled 0.257 (0.142-0.532) 4.252 (1.430-75.932) 0.082 (0.062-0.106) 0.397 (0.267-0.773) 2.71 5.81 0.198 (0.158-0.247) 0.643 (0.397-1.669) 0.210 (0.165-0.267) 2.024 (1.079-5.988) Mean 0.222 2.306 1LC = Lethal Concentration, CI = Confidence Interval. 2Mean of 2 or 3 field trial LC50s. 3Data excluded from mean, no overlap in confidence intervals with other field results. 4RR = resistance ratio between field and colony LCs. LC50 = field LC50/colony LC50, LC95 = field LC95/colony LC95.
138Table 5-5. Insecticide susceptibility results for Pensacola, Escambia County, Flor ida and USDA ARS colony populations of adult female Aedes albopictus. Insecticide LC50 (95% CI)1 LC95 (95% CI) Colony LC50 Colony LC95 RR50 4 RR95 ( g/vial) ( g/vial) ( g/vial) ( g/vial) Malathion 0.602 (0.425-0.986) 5.566 (2.596-24.255) 0.153 (0.118-0.197) 0.666 (0.469-1.123) 3.03 53.07 0.501 (0.313-0.896) 22.696 (7.269-203.090) 0.287 (0.140-0.607) 77.768 (12.468-7,875) Mean2 0.463 35.34 Permethrin 0.033 (0.022-0.046) 0.281 (0.169-0.678) 0.037 (0.027-0.049) 0.261 (0.149-0.804) 1.05 0.84 0.042 (0.033-0.053) 0.190 (0.129-0.365) 0.042 (0.032-0.054) 0.189 (0.099-0.237) Mean 0.039 0.220 Resmethrin 0.548 (0.406-0.754) 3.986 (2.324-9.801) 0.235 (0.156-0.413) 3.894 (1.526-23.883) 2.14 1.00 0.396 (0.316-0.483) 1.081 (0.819-1.733)3 0.567 (0.422-0.779) 3.766 (2.250-8.979) Mean 0.504 3.876 D-phenothrin 0.372 (0.250-0.575) 8.349 (3.689-34.001) 0.358 (0.266-0.500) 2.513 (1.448-6.345) 1.39 2.29 0.447 (0.334-0.602) 2.955 (1.810-6.632) 0.673 (0.489-0.967) 5.924 (3.243-16.664) Mean 0.497 5.743 Naled 0.121 (0.092-0.163) 0.713 (0.445-1.530) 0.082 (0.062-0.106) 0.397 (0.267-0.773) 2.18 3.72 0.201 (0.143-0.289) 2.396 (1.065-11.755) 0.215 (0.163-0.284) 1.332 (0.811-2.946) Mean 0.179 1.480 1LC = Lethal Concentration, CI = Confidence Interval. 2Mean of 2 or 3 field trial LC50s. 3Data excluded from mean, no overlap in confidence intervals with other field results. 4RR = resistance ratio between field and colony LCs. LC50 = field LC50/colony LC50, LC95 = field LC95/colony LC95.
139 Table 5-6. Insecticide susceptibility re sults for Jacksonville, Duval County, Florid a and USDA ARS colony populations of adult female Aedes albopictus. Insecticide LC50 (95% CI)1 LC95 (95% CI) Colony LC50 Colony LC95 RR50 3 RR95 ( g/vial) ( g/vial) ( g/vial) ( g/vial) Malathion 0.147 (0.107-0.199) 1.161 (0.706-2.580) 0.153 (0.118-0.197) 0.666 (0.469-1.123) 1.19 1.83 0.181 (0.133-0.246) 1.427 (0.860-3.231) 0.219 (0.169-0.285) 1.065 (0.718-1.943) Mean2 0.182 1.218 Permethrin 0.063 (0.048-0.082) 0.316 (0.206-0.645) 0.037 (0.027-0.049) 0.261 (0.149-0.804) 1.78 1.08 0.064 (0.048-0.084) 0.338 (0.220-0.693) 0.070 (0.058-0.086) 0.193 (0.139-0.358) Mean 0.066 0.282 D-phenothrin 0.601 (0.445-0.834) 4.445 (2.580-11.024) 0.358 (0.266-0.500) 2.513 (1.448-6.345) 1.92 2.06 0.790 (0.578-1.142) 6.416 (3.488-18.309) 0.672 (0.500-0.933) 4.670 (2.718-11.802) Mean 0.688 5.177 1LC = Lethal Concentration, CI = Confidence Interval. 2Mean of 2 or 3 field trial LC50s. 3RR = resistance ratio between field and colony LCs. LC50 = field LC50/colony LC50, LC95 = field LC95/colony LC95.
140Table 5-7. Insecticide susceptibility re sults for Gainesville, Alachua County, Flor ida and USDA ARS colony populations of adul t female Aedes albopictus. Insecticide LC50 (95% CI)1 LC95 (95% CI) Colony LC50 Colony LC95 RR50 3 RR95 ( g/vial) ( g/vial) ( g/vial) ( g/vial) Malathion 0.228 (0.167-0.318) 2.009 (1.150-5.056) 0.153 (0.118-0.197) 0.666 (0.469-1.123) 1.54 2.60 0.250 (0.186-0.342) 1.809 (1.088-4.127) 0.227 (0.172-0.303) 1.371 (0.879-2.756) Mean2 0.235 1.730 Permethrin 0.035 (0.026-0.045) 0.217 (0.123-0.625) 0.037 (0.027-0.049) 0.261 (0.149-0.804) 0.84 0.50 0.034 (0.025-0.045) 0.174 (0.116-0.346) 0.025 (0.016-0.035) 0.193 (0.139-0.358) Mean 0.031 0.131 D-phenothrin 0.514 (0.378-0.714) 4.125 (2.380-10.308) 0.358 (0.266-0.500) 2.513 (1.448-6.345) 1.55 1.47 0.680 (0.508-0.942) 4.603 (2.700-11.291) 0.470 (0.360-0.614) 2.350 (1.370-6.943) Mean 0.555 3.693 1LC = Lethal Concentration, CI = Confidence Interval. 2Mean of 2 or 3 field trial LC50s. 3RR = resistance ratio between field and colony LCs. LC50 = field LC50/colony LC50, LC95 = field LC95/colony LC95.
141 Figure 5-1. Aedes albopictus egg collection sites, north and central Florida. 1) Pensacola, Escambia County. 2) Quincy, Gadsden County. 3) Jacksonville, Duval County. 4) Gainesville, Alachua County. 5) Inverness, Citrus County. 6) Vero Beach, Indian River County.
142 Figure 5-2. Aedes albopictus egg collection apparatus. Si xteen-oz black plastic cups lined with seed germination paper were secured to 30 cm nails. Cups were filled with approximately 200 mL of tap water and left in place 7 to10 days before papers were collected.
143 Figure 5-3. Partitioned box hol ding insecticide-coated 20 mL scintillation vials. Adult female mosquitoes were narcotized with 500 mL/min CO2 injected directly into a mechanical aspirator collection tube, removed, separated by sex, and added to insecticide-coated vials with forceps.
144 CHAPTER 6 LABORATORY RESPONSE OF Aedes albopictus TO LIGHT EMITTING DIODES OF EIGHT DIFFERENT COLORS AND ULTRAVIOLET LIGHT OF EIGHT DIFFERENT FLICKER FREQUENCIES Introduction Visual cues play a key role in successf ul host location by biting insects (Laarman 1955, Allan et al. 1987, Lehane 1991). Refl ected and transmitted light, movement, size, contrast, and color are components of these cues (Brown 1953, 1954). Over the last 70 years, researchers have tested a large array of traps incorporating artificial light of different color, intensity, and/or frequency in attempts to improve trap capture (Service 1993, Bidlingmayer 1994). Trap color (reflected light) and lamp color (transmitted light) have been among the most intensel y studied of these visual cues. Diurnally active biting insects such as Ae. albopictus and Ae. aegypti are thought to be more sensitive to color (spectral reflectance) than crepuscular or nocturnally active biters (Brett 1938, Gjullin 1947, Brown 1954). Conversely, nocturnally active biting insects are suspected of being more sensitive to movement than day biters and may have a heightened ability to detect intensity contrast (Allan et al. 1987). Attraction studies have therefore focused on the color preferences for both reflected and transmitted light among diurnally, crepuscular and nocturnally f eeding mosquitoes in attempts to enhance trap capture (Gjullin 194 7, Sippell and Brown 1953, Thurman and Thurman 1955, Gilbert and Gouck 1957, Haufe 1960, Barr et al. 1963, Breyev 1963, Fay 1968, Miller et al. 1969, Fay and Prince 1970, Bidlingmayer 1980, Wilton and Kloter 1985, Ali et al. 1989, Burkett et al. 1998).
145 Crepuscular and nocturnally active mosquitoes have generally shown preference for several colors of transmitted light regardless of species. Ultraviolet (UV), blue, green, and incandescent light have fared bette r than most other colors in attracting Aedes, Ochlerotatus, Coquillettidia, Culiseta, Culex, Mansonia, and Psorophora mosquitoes (Headlee 1937, Breyev 1963, Ali et al. 1989, Burk ett et al. 1998). Mosquitoes have also shown color preference (reflected light) to traps and targets. Brown (1954) demonstrated color preference among Canadian woodland Aedes mosquitoes for black, red, blue, and green colored cloth over yello w, orange, and tan cloth. Browne and Bennett (1981) found Canadian Aedes and Mansonia mosquitoes more attracted to black, red, and blue targets over white or yellow ta rgets. It is well known th at many anopheline mosquitoes are attracted to black and red color; these colors are used in resting boxes to obtain surveillance data (Goodwin 1942, Laarman 1955, Service 1993). Medically important an d diurnally active mosquitoes such as Ae. aegypti and Ae. albopictus are known to be attracted to dark surfaces and to a lesser degree certain shades of red (Brett 1938, Brown 1966, Kusakabe a nd Ikeshoji 1990, Estrada-Franco and Craig 1995). Based on woodland Aedes response to colored cloth, Brown (1954, 1956, 1966) concluded that Aedes mosquitoes can discern color be tween wavelengths of 475 nm and 625 nm (blue-green to orange) and are not attracted to those colors per se, but to their spectral reflectance. He surmised that re flected light above a nd below this range appeared black to these mosquitoes and that some of these frequencies were attractive to them. Apparently, these visual qu alities are also attributes of Ae. aegypti and Ae. albopictus as they are attracted to black and red colors (Brett 1938). This finding led to the development of traps incorporati ng black color specifically targeting Ae. aegypti (Fay
146 1968, Fay and Prince 1970, Fr eier and Francy 1991). Aedes aegyptis attraction to black is increased by adding alternating or checker ed white patterns (Sippell and Brown 1953, Brown 1956, Fay and Prince 1970). It is not cl ear why this occurs, but increased contour may cause a flicker effect that the mosquito finds attractive (Brown 1966). Despite extensive research and published data focusing on the attractiveness of transmitted light to mosquitoes, little is known of colored lig ht preference of Ae. albopictus. Negative phototaxis is a characteristic of Ae. albopictus and Ae. aegypti although they readily bite in open daylight (Christophers 1960). Adults are seldom collected in adult light tr aps such as the NJLT and CDC light trap (Thurman and Thurman 1955). Indeed, it has been demons trated that illumination of black surfaces with incandescent light as low as 100 lux decr eased the attractiveness of black surfaces to Ae. aegypti (Wood and Wright 1968). Lack of host-seeking response in Ae. albopictus to most mosquito traps led to the use of ovi position traps for surv eillance (Service 1993). Light other than incandescent light might be useful in attracting Ae. albopictus. Burkett (1998) investigated the response of Ae. albopictus to an artificial host illuminated with light of 9 different colors. Us ing a multiport olfactometer, he exposed Ae. albopictus to filtered light separated by 50 nm wavelength intervals, from 350 nm through 700 nm. He found light of 600 nm (yellow-orange) most attractive followed by 500 nm (blue-green), white, 450 nm (blue), 400 nm (violet), and no light attracting significantly more adults than the other fr equencies tested (350 nm, 550 nm, and 650 nm) (UV, yellow-green, and red light, respectivel y). Apart from this study, nothing was found about Ae. albopictuss light preference.
147 The multiport olfactometer was developed by Dr. J. F. Butler of the University of Florida. This olfactometer is capable of evaluating the relationship between biting activity (measured in sec) and color preference for most hematophagous insects of interest. Eight different wavelengths of transmitted light were presented to adult female Ae. albopictus to determine color preference. The most favorable of these colors was then selected for testing with different rates of flicker. Data collected from our study might be exploited in traps to aid in surveilla nce, control, or resear ch endeavors targeting Ae. albopictus. Materials and Methods Visualometer A pie-shaped olfactometer (Butler and Ka tz 1987, Martin et al. 1991, Butler and Okine 1994) designed by Dr. Butler electronicall y monitors and quantif ies insect feeding activity simultaneously on 10 arti ficial hosts. The olfactomet er, electrical amplification boxes, and CO2 input and exhaust systems are shown in Fig. 6-1. The apparatus is termed visualometer when measuring biting response to light as opposed to chemical attractants. Ten identical artificial hosts were embedded in a transparent Plexiglas ceiling and each was illuminated with light of a diffe rent color and/or frequency. Feeding activity on artificial host was measured over an 8 h period and quantified as biting-sec. Feeding mosquitoes rested on an insulated wire screen and completed an electrical circuit after inserting their proboscis into artifici al host. A computer recorded, logged, and analyzed biting response. Li ght was provided inside the vi sualometer from a Plexiglas false floor in which 10 holes were drilled to accommodate Light Emitting Diodes (LEDs). Holes were centered immediately below artificial media to allow direct illumination by LEDs. An additional white lig ht was provided from the aluminum floor
148 of the visualometer as a ninth light source by fiber optic light guide from an external tungsten halogen source. The final artific ial host was not illuminated (control). Temperature and humidity inside the visualometer was maintained at 32oC and 70% relative humidity. The visualometer was enclosed in a Fa raday cage room (Lindgren Enclosures, Model no. 18-3/5-1) to protect against outsi de electrical interference and extraneous sources of light. The Faraday cage was maintained between 28oC and 32oC. All visualometer surfaces exposed to mosquitoes were disposable or cleaned with soapy water between trials. Light Emitting Diodes Light emitting diodes of 8 different colors (wavelengths) were evaluated for their attractiveness to Ae. albopictus in a visualometer loaded w ith artificial host. The LEDs were obtained from Digi-Key Corporation (Thief River Falls, MN). Color, part number, wavelength and millicandela (mcd) chosen fo r testing were ultraviolet (67-1831-ND, 380 nm, 70 mcd), violet (67-1830-ND, 410 nm 120 mcd), blue (P466-ND, 470 nm, 650 mcd), green (67-1755-ND, 502 nm, 1,500 mcd), orange (67-1113-ND, 610 nm, 2,500 mcd), red (67-1611-ND, 660 nm, 1,800 mcd) infrared (LN77L-ND, 860 nm), and infrared (67-1001-ND, 940 nm). Infrared diodes are not mcd-rated; infrared radiation is invisible. Light emitting diode s were 8.6 mm long by 5.0 mm in diameter with round lens. Viewing angles were 30o except for IR 860 (20o). White light produced from a wide spectrum tungsten-halogen bulb (Sylva nia, no. DNF, Danvers, MA) was provided by fiber optic cables embedded into the alumi num floor of the visualometer. All were blocked with copper tape except 1 to provide a white light source (pos itive control). The
149 final artificial host remained unlit as a no-light control. Light Emitting Diodes were randomly assigned positions around the vi sualometer for every replication. Artificial Host Refrigerated food attractant, or artificial host, was prep ared at the beginning of every trial. It consisted of agar gel, bovine blood, and attr actants. Thirty-three mL of fresh (less than 3 weeks old) bovine blood was added to 100 mL of BSS Plus Ocular solution (Alcon Laboratories, Inc., Fort Wort h, TX) and 1.66 g of powdered agar (U.S. Biochemical Corp., Cleveland, OH). Twelve g of calcium citrate was added per gallon of bovine blood to prevent coagulation. Ocul ar solution mimics natural composition of human tears and is attractive to mosquitoes. Ocular solution (500 mL ) contained 7.14 mg sodium chloride, 0.38 mg potassium chloride 0.154 mg calcium chloride dehydrate, 0.2 mg magnesium chloride hexahydrate, 0.42 mg dibasic sodium phosphate, 2.1 mg sodium bicarbonate, 0.92 mg dextrose, 0.184 mg glutathion disulfid e, and hydrochloric acid or sodium hydroxide to bring pH to 7.4. Ocular solution was brought to a boil, allowed to cool slightly, and agar was added. The so lution was heated to a boil twice again and allowed to cool until warm to the touch. Citrated bovine blood was added and the media was poured into 100 mL plastic syringes with the tips removed. Plastic 35 mm film canister lids were filled with media and covere d with a 10 m to 15 m silicon membrane (Butler et al. 1984). A 1/8 in slice cut off th e top of the 35 mm film canister was used to hold the membrane tight against the lid. This assembly was inserted membrane side down into the Plexiglas port in the ceiling of the visualometer. New artificial host was used at the beginning of each trial. Carbon dioxide was provided to the visualometer as an additional attractant at a rate of 20 mL to 30 mL/min th rough 10 air inlet ports evenly distributed around the visualometers perimeter.
150 Mosquitoes Aedes albopictus from the USDA ARS Gainesville mo squito rearing facility. The colony was established in 1992 (Ruide Xue, Director, Anastasia Mosquito Control District, personal communication) and kept under a 14:10 (L:D) phot operiod. Rearing room temperature was maintained between 27oC and 32oC and approximately 50% to 60% RH. One hundred sixty nulliparious adult females from 7 to 20 days old were used in all trials. Adults were destroyed after each test. Flicker Response Trials Highly attractive Ultraviole t (380 nm) LEDs were chosen for frequency testing on Ae. albopictus. Eight frequencies were tested with 2 controls (1 LED constantly on (incan) and 1 LED always off (blank), posit ive and negative contro ls, respectively). Frequency is measured in Hertz (Hz), or the number of on/off cycles per second. Frequencies tested were 10-, 30-, 40-, 60-, 120-, 150-, 200and 500 Hz. Frequency selection was based on research indicating that slow-flying Dipt era (mosquitoes) are capable of discerning flicker to approxi mately 40 Hz (Mazokhin-Porshnyakov 1969) and the supposition that mosquitoes might possibly use reflected light at wing beat frequency (up to 600 Hz in Ae. albopictus (Brogdon 1994)) to locate mate s (Jerry Butler, personal communication). The LED flashing circuit was constructed from a 555 precision timer integrated circuit (IC) connected for stable operation so th at it operated as a multivibrator (Fig. 6-2). Detailed description of the circuit func tion and capacitor components is given in Appendices D and E, respectively.
151 Statistical Analysis Statistical analysis was conducted on 10 replications for both color and flicker preference trials. All bite contact sec were r ecorded for 8 h and analyzed over this period of time. Results from 10 trials were c onsolidated and data normalized by SQRT (n + 1) conversion. A randomized complete block de sign was used for both color and frequency preference trials and anal yzed for significance ( = 0.05) with respec t to treatment and replication using a 2-way ANOVA (SAS In stitute, 2001). The Ryan-Einot-GabrielWelsh Multiple Range Test ( = 0.05) was used to delineate significant differences between colored light and flicker frequency means. Results Aedes albopictus Response to Light of Different Color Color preference of Ae. albopictus in 10 randomized trials was determined in the visualometer. Selection of 10 trials from a pool of 26 trials was dependent upon 1) no treatment failures or malfunctions within the visualometer for a given trial and 2) all trial means within 50% of the group mean (b ite-time averages above 50% of the group mean implied sensor malfunction, bite-tim e averages below 50% of the group mean implied poor mosquito quality). Bi te-second data was normalized (SQRT (n + 1)) over an 8 h period for all trials to obtain treatme nt means and the standard error of means. Results of the 10 trials in our expe riment are presented in Table 6-1. Significant differences were seen for treatm ent (F = 2.95, df = 9, p= 0.0044) but not between trials (F = 1.09, df = 9, p = 0.38). The normalized group mean ( standard error of the mean) was 24.02 bite-sec ( 5.70) for all trials with a 50% group mean range from 12.01 sec to 36.03 sec (Table 6-1). Bite-second response ranged from a high of 43.64 sec (380 nm) to a low of 7.38 sec (860 nm). These results indi cate that ultraviolet
152 light (380 nm) was the mo st attractive color to Ae. albopictus (Fig. 6-3). Ultraviolet light outperformed the second most attr active light (blue, 470 nm) by 25%. Aedes albopictus fed significantly longer on UV-lit artificial host than on incandescent(p = 0.0147) or infrared-lit (860 nm) (p = 0.0016) artificial hos t. Blue (470 nm), violet (410 nm), and green (502 nm) light were about equally a ttractive (32.74-, 28.98-, and 27.28 bite-sec, respectively). Orange (610 nm), red (660 nm), and infrared (940 nm) (24.09-, 21.65-, and 18.88 bite-sec, respectively) were least attractive in comparison with UV light, but not significantly different in results. Incandescentand in frared (860 nm) light was least attractive to Ae. albopictus, and at least 31% less attract ive than low-average attractive infrared 940 nm. Aedes albopictuss attraction to colored light generally decreased as wavelength increased (from 380 nm through 940 nm). Aedes albopictus Response to Flickering Li ght of Different Frequencies Aedes albopictuss response to artificial host illuminated with flickering light of different frequencies was determined from ni ne randomized trials. Ultraviolet light (380 nm) was used at all flicker frequencies and all replicates in this experiment. Ultra violet light was chosen due to superi or results achieved in the color preference trials. Criteria for selection of 9 trials from a pool of 26 trials are mentioned above. Bite-second data was normalized (SQRT (n + 1)) over the 8 h period for all trials to obtain treatment means and the standard error of means. Results of the 9 trials used in this experiment are given in Table 6-2. No significant differe nce was observed in Ae. albopictuss response to flickering light treatments (F = 1.22, df = 9, p = 0.30) or between trials (F = 0.86, df = 8, p = 0.55). The normalized (SQRT (n + 1)) group mean ( standard error of the mean) was 13.67 bite-sec ( 3.51) for all trials with a 50 % group mean range from 6.69 sec to 20.50 sec
153 (Table 6-2). Bite-second tr eatment ranged from a high of 20.94 (500 Hz) to a low of 6.50 (10 Hz). Biting activity tended to increase as flicker frequency increased (from 10 Hz to 500 Hz). Non-flickering treatment (stea dy light) response was higher than the group mean (16.09 Hz and 13.67 Hz, respectivel y) while the negative control (no light) treatment was slightly lower than the group mean (11. 51 Hz and 13.67 Hz, respectively). Discussion Colored Light Preference Transmitted light attraction studies with di urnally active mosquitoes are scarce, as most emphasis with these insects has been placed into traps or targets of different colors (reflected light) (Howlett 1910, Brett 1938, Brown 1954, Gilbert and Gouck 1957, Fay 1968, Fay and Prince 1970, Wilton and Kloter 1985). Transmitted light studies are usually directed towards crepuscular a nd nocturnally active species (Headlee 1937, Thurman and Thurman 1955, Barr et al. 1963, Brey ev 1963, Ali et al. 1989, Burkett et al. 1998). It is generally accepted among mosquito control practitioners and researchers that incandescent light provided by CDC li ght traps is a poor attractant for Ae. aegypti and Ae. albopictus; published data draw the same c onclusions (Thurman and Thurman 1955, Freier and Francy 1991, Service 1993, Jensen et al. 1994). Due to this fact, traps designed to capture Ae. aegypti and Ae. albopictus have incorporated colors (reflected light) attractive to these species with highl y attractive black most commonly used (Fay 1968, Fay and Prince 1970, Wilton and Klot er 1985, Freier and Francy 1991). Unfortunately, the lackluster response of these species to incandescent light has resulted in a belief that transmitted light in general is probably an ineffective attractant for both species without regard to color.
154 In addition to laboratory and field studies that quantify mosquito response to light, spectral sensitivities in mosquitoes can be determined thought elec troretinograph (ERG) techniques. Electroretinographs may provide clues to a species choice of colored light, should that insect be attracted to light. Ironically, the onl y 2 published mosquito spectral sensitivity studies were both conducted on diurnally active Ae. aegypti (Snow 1971, Muir et al. 1992). Snow (1971) investig ated the spectral sensitivity of Ae. aegypti during oviposition and Muir et al. ( 1992) used ERGs to determine the spectral sensitivity range of Ae. aegypti. Bimodal sensitivities were noted in both studies with peaks occurring near 325 nm (ultraviolet) and 535 nm (yello w-green). Bimodal visual sensitivity is characteristic of other hematophagous insect s including black flies, tabanids, horn flies and tsetse flies (Lehane 1991, Smith and Butler 1991). Results of visualometer color pr eference trials demonstrate that Ae. albopictus responds most frequently to UV (380 nm), blue (470 nm), violet (410 nm), and green (502 nm) light. Although no significant differe nces exist between th ese 4 treatments, UV light outperformed the next most attractive light, blue, by 25%. Bl ue, violet, and green light were almost equally attractive to Ae. albopictus. Results of these trials corresponded well with the results of Snow (1971) and Muir et al. (1992) in that Ae. albopictus responded most often to colors in the same frequency ranges as those to which Ae. aegypti responded (UV and green light). A lthough the 2 mosquitoes are different species, they are very closely rela ted taxonomically (both of subgenus Stegomyia) and behaviorally (both ar e diurnal biters). Incandescent light used in the visualometer was significantly less attractive to Ae. albopictus than UV light. Only infrared (860 nm) performed more poorly than
155 incandescent light. Unlit artificial media was almost twice as attractive (1.76 x) to Ae. albopictus than incandescent-lit media. Similar trends were observed in a Vietnam field study comparing Ae. albopictus capture in litand unlit-CDC traps (Herbert et al. 1972). These results also agree well with the genera l lack of success of in candescent light traps compared to other traps not using light (F reier and Francy 1991, Jensen et al. 1994, Dennett et al. 2004). Unexpectedly, infrared (960 nm) light elicited a moderate amount of biting activity, a lthough its bite-time mean was lo west among all midrange results (Fig. 6-3). Apart from this aberration, th ese data support several aspects of what is known about mosquito attraction to transm itted colored light: 1) light of shorter wavelengths are usually more attractive to mo squitoes; UV, blue, and violet light being the most attractive in our test, 2) green light is attractive to Aedes mosquitoes and is in the range of frequencies associated with the bi modal sensitivity spectrum associated with Ae. aegypti, 3) as wavelength lengthens across the visual spectrum attractancy decreases, with reds and infrareds being at the visual limit of Aedes mosquitoes as determined by ERGs (approximately 650 nm) (Muir et al. 1992), and 4) with regard to diurnally active mosquitoes, incandescent light tends to be repellent compared with no light. The results of our study warrant field resear ch as to the effectiveness of UV, blue, and/or green LED-modified light traps in collecting Ae. albopictus. Additionally, traps targeting Ae. aegypti (and Ae. albopictus) (Fay-Prince, Wilton trap) could be supplemented with LEDs in an attempt to enha nce capture rates. The miniscule amount of energy required to power LEDs (approximately 20 mA/h) w ould not adversely impact battery life.
156 Flickering Light Preference There are 2 good arguments for developing flicke ring light in insect traps. First, a substantial amount of battery life could be conserved using circuitry that provided a 50% on/off duty cycle. Nearly half the energy requ ired to operate a standard CDC light trap (model 512, J.W. Hock Company, Gainesville, FL) is consumed by the lamp. Using this circuitry, battery life could be extende d by 25% and might result in money and manpower savings. Second, 1 or several flic ker frequencies might prove more attractive to target mosquito species than steady light and possibly improve su rveillance results or enhance control efforts incorporating mech anical trapping (Mboera and Takken 1997). Third, even if flickering light was no more attractive than steady light, additional battery life conservation would result from a 50% on/off duty cycle. Three published studies reported on the at tractiveness of flickering light to mosquitoes. Varva et al. (1974) modified CDC li ght traps to flicker at a rate of 2 Hz with the goal of reducing operating costs. Flic kering light traps caught significantly less mosquitoes than steady light traps. Ross a nd Service (1979) used a flickering fluorescent light (20 x/min to 30 x/min, ~ 0.5 Hz) to capture mosquitoes and sand flies. The flickering light trap captured as many or more biting flies than steady light traps. Lang (1984) modified NJLTs to flicker at rates of 1-, 5-, 10-, 20-, 30-, 45and 59 Hz. Comparable numbers of mosquitoes were captu red in traps flickering at 10-, 20-, 30and 45 Hz as compared to steady light traps. These experiments show that at certain flic ker rates, trap capture in modified traps can match capture rates of unmodified traps while conserving battery life. Results of Langs study demonstrated that flickering li ght at 45 Hz produced optimum results, and of the 3 studies, this was the only frequency at which more mosquitoes were caught
157 compared to steady light. Mazokhin-Pors hnyakov (1969) has shown that slow-flying insects (mosquitoes) have the ab ility to distinguish flickering light in the range of 10 Hz to 40 Hz. Above 40 Hz, the ability to see flicke r is lost and light appe ars to be constant. The point at which the eye can no long di scern flicker is known as flicker fusion frequency (FFF) and is a function of the rec overy time of photoreceptors (Lehane 1991). Sensitivity to flicker relates to the ability of an organism to detect movement, thus, mosquitoes are probably able to detect move ment to the same degree as man (FFF of 20 Hz to 30 Hz) (Lehane 1991). No significant difference between treatmen ts (n = 10, p = 0.29) was observed in visualometer trials with 8 flicker rates (10-, 30-, 40-, 60, 120-, 150-, 200and 500 Hz) and 2 controls (negative, no light, positive, steady light) with Ae. albopictus. At flicker frequencies believed discernable to the vision of this mosquito, biting-time means were below the group mean of 13.67 bite-sec (10 Hz mean of 6.50 bite-sec, 30 Hz mean of 9.38 bite-sec). At 40 Hz flic ker, the bite-sec mean was a bout equal with the group mean (13.70 Hz and 13.67 Hz, respectively). The lit control (incan, Fig. 6-4) achieved slightly higher bite-sec means than did the unlit negati ve control. Although these differences are not significant, these findings do support th e proposition that flic kering UV light (380 nm) might be as good a visual attractant in modified trap s compared with unmodified traps. Results of the light color preference trials tend to support this argument (43.64 bite-sec mean for UV treatment vs. 22.64 bite-sec mean for unlit treatment; nearly twice the biting activity on UV lit media compared to un lit media). Indeed, Kloter et al. (1983) caught significantly more Ae. aegypti in UV lit CDC traps than in lit and unlit CDC traps.
158 Findings of no significant di fference between treatments in our flickering light trials support the use of 555 circuitry-modified light traps. These circuits are rugged, small and lightweight (Fig. 6 5). They could easily be attached to the underside of the rain shield and integrated into trap circui try with quick disconn ect attachments. If Ae. albopictus can not discern between flickering light and if UV, blue, or green light is in fact more attractive than standard incandes cent light, trap performance could be improved while extending the operational life of batteries used to power them (i.e., smaller power requirements of LEDs vs. CM-47 bulbs a nd 50% reduction in power necessary to energize light sources from 555 circuits). Thes e results indicate that future trials are needed to ascertain the impact of UV, blue, and/or green LEDs in enhancing trap capture of Ae. albopictus in the field. Equivalent or supe rior results with modified traps over current light traps could lead to savings in time and manpower required for routine surveillance activities and might someday be in corporated into an integrated mosquito management program.
159 Table 6-1. Average number of bite-sec for 8 h exposure of Ae. albopictus to artificial host illuminated by light of different colors. Trial Treatment (wavelength nm) Trial mean Color Incan1Blank2UVViolBlueGreenOrangeRedIR860IR940 Wavelength 380410470502610660860940 1 4.23.735.69.222.214.171.1249.020.83.818.61 2 126.96.36.1996.734.927.011.914.711.52.527.22 3 4.523.423.073.92.330.116.12.911.58.419.61 4 37.04.041.623.12.129.015.919.91.646.922.10 5 19.26.857.111.551.969.730.345.93.53.029.91 6 3.824.726.7188.8.131.52184.108.40.2068.819.22 7 12.150.678.87.560.439.114.665.216.425.537.033 8 15.026.646.826.050.61.08.4220.127.116.110.42 9 2.96.041.730.92.618.058.22.94.218.918.62 10 21.024.527.29.451.718.104.22.168.057.427.43 Trt average 12.8822.6443.6428.9832.7427.2824.0921.657.3818.88 Group mean 24.02 50% Above average 36.03 50% Below average 12.01 1Incandescent light of all waveleng ths of the visible spectrum. 2Blank treatment; no light (control). 3Trial mean is slightly above upper 50 % average, however, review of trial biting data indicates no malfuncti on of visualometer equipment. Table 6-2. Average number of bite-sec for 8 h exposure of Ae. albopictus to artificial host illuminated by flickering light of di fferent frequencies. All light from ultraviolet light emitting diodes (380 nm). Trial Treatment (flicker Hz) Trial mean 1030 4060120150200500Blank Incan 1 10.72.2 22.214.171.1245.6126.96.36.199 37.713.38 2 1.62.0 188.8.131.52184.108.40.206.5 14.710.68 3 6.91.2 19.3220.127.116.113.82.421.3 13.112.55 4 1.534.7 18.104.22.168.79.331.57.1 19.712.16 5 1.615.7 14.723.610.21.314.58.817.3 10.311.79 6 19.44.4 23.46.038.64.322.214.813.9 15.116.20 7 10.94.0 22.214.171.1240.140.149.218.6 17.118.69 8 1.04.9 1.149.3126.96.36.1991.010.2 12.018.12 9 4.915.4 26.53.03.116.188.8.131.52 5.09.45 Trt average 6.509.38 13.7014.9215.1611.7216.7820.9411.51 16.09 Group mean 13.67 50% Above average 20.50 50% Below average 6.69
160 A B Figure 6-1. Visualometer used in color prefer ence tests. A) Set up. B) In operation.
161 Figure 6-2. Diagram of a 555 integrated circ uit frequency generator. Numbers 1 through 8 represent capacitors of varying cap acitance, R1 through R4 are ohm ratings of circuit resistors.
162 b b ab ab ab ab ab ab a ab 0 10 20 30 40 50 blank380410470502610660860940incan Wavelength (nm)Mean SQRT feeding time (sec) over 8 h Figure 6-3. Duration of feeding (sec) over an 8 h period (mean SQRT (n + 1) SEM) for Aedes albopictus on artificial host illu minated with light of different colors. Means within each treatment with the same letter are not significantly different ( = 0.05, Ryan-Einot-Gabriel-Welsh Multiple Range Test).
163 a a a a a a a a a a0 5 10 15 20 25 30 10304060120150200500blankincan Frequency (Hz)Mean SQRT feeding time (sec) over 8 Figure 6-4. Duration of feeding (sec) over an 8 h period (mean SQRT (n + 1) SEM) for Aedes albopictus on artificial host i lluminated with ultraviolet (380 nm) light emitting diodes of different frequencies. Means within each treatment with the same letter are not significantly different ( = 0.05, Ryan-Einot-GabrielWelsh Multiple Range Test).
164 CHAPTER 7 EVALUTION OF LIGHTAND MOTOR-MO DIFIED CENTERS FOR DISEASE CONTROL TRAPS FOR WOODLAND MO SQUITOES IN NORTH CENTRAL FLORIDA Introduction Adult mosquito surveillance is important to state and national vector control agencies for several reasons, chief among thes e is determination of species composition, abundance through time, and identification of potential disease vectors. Centers for Disease Control and Prevention light traps are routinely used by vector control agencies in Florida for mosquito surveillance (Flori da Coordinating Council on Mosquito Control 1998). These battery-powered tr aps are especially useful in remote locations in which electricity-powered surveillance traps such as the New Jersey Light Tr ap cannot be used. CDC light traps use incandes cent light and sometimes CO2 as attractants. They are powered by 6 V rechargeable batteri es or 4 1.5 V D cell batteries. An operational limiting factor associated w ith use of CDC light traps is length of battery life. A typical 10or 12 ampere-h (A-h) rechargeable battery will effectively power a CDC trap for approximately 36 h before battery needs to be replaced. This is enough power for 1 effective day of trapping. A CDC light trap left operating on the same 6 V battery for 48 continuous h will not be operating at all, or only at a minimal level. After 48 h of use, suction velocity is insufficient to prevent mosquito escape (capture nets are situated below the suction cylinder; once airspeed is unable to contain mosquitoes, adults instinctively fly upward a nd escape out of the trap opening). Thus, to effectively monitor adult populations, the CD C light trap must be attended to daily.
165 Light traps are effective, in part, due to lamp color and intensity characteristics. The attractiveness of transmitted and reflected light to mosquitoes has been investigated by many authors (Headlee 1937, Brett 1938, Gjullin 1947, Brown 1951, Sippell and Brown 1953, Brown 1954, Gilbert and Gouc k 1957, Barr et al. 1963, Breyev 1963, Fay 1968, Wood and Wright 1968, Vavra et al. 1974, Browne and Bennett 1981, Lang 1984, Ali et al. 1989, Burkett 1998). In general, blue, green, red, and incandescent transmitted light has proven to be attractive to the major ity of mosquitoes in th ese studies. In fact, most hematophagous flies are attr acted to light of short wave length, especially ultraviolet light (Breyev 1963, Lehane 1991). Attraction to reflect ed light has been shown, in most cases, to be inversely proportional with the reflectivity or brightness of the surface from which light was reflected (Gjullin 1947). Darker, less reflective colors are usually favored over brighter colors (Brett 1938). Cana dian field species were less attracted to reflected light as wavelength increased from 475 nm through 625 nm (Brown 1954). Black, blue, and red surfaces are more attrac tive to many species than lighter colors (Gilbert and Gouck 1957, Fay 1968, Browne and Bennett 1981). Light intensity plays a role in the attrac tion of nocturnally active mosquitoes. Breyev (1963) and Ali et al. (1989) found little difference in mosquito preference between lamps of different intensities as long as intensity was low (i .e., at or below 200 W). Barr et al. (1963) found that several speci es of California Riceland mosquitoes were increasingly attracted to br ighter light, from 25 W through 100 W. Above a certain intensity, light becomes repellent to mosquito es (Service 1993) and some species are not attracted to transmitted light at all (Thurman and Thurman 1955, Fay and Prince 1970).
166 The goal of our study was to compare woodland mosquito capture rates between a standard CDC light trap and traps modified w ith a combination of energy efficient motors and/or highly attractive blue LEDs. Light em itting diodes were oriented in 2 directions, one direction (perimeter orientation) provi ded direct transmitted light and the other (cluster orientation) provided reflected light as visual attr actants. It is hoped that comparable capture rates and similar species diversity can be obt ained with modified CDC light traps that effectively run 3 to 4 x longer than standard tr aps thus reducing time and manpower requirements necessary fo r routine mosquito surveillance. Materials and Methods Trap Rotation and Collection Three identical field trials were cond ucted using 6 CDC model 512 light traps (John W. Hock Company, Gainesville, FL). One unmodified trap (control) and 5 modified CDC light traps were used in all tr ials. Traps were activated between 0800 and 1000 and allowed to run for 24 h before coll ection and rotation (s equentially in a 6 x 6 Latin square design). Trap intake was se t 150 cm (5 ft) above ground per manufacturer recommendation and traps were spaced at least 200 m apart. Traps were placed such that they were not visible from other traps. Ca rbon dioxide was provided to all traps from a 9 kg (20 lb) compressed gas cylinder. A flow rate of 500 mL/min was achieved with a 15psi single-stage regulator equi pped with an inline microregul ator and filter (Flowset 1, Clarke Mosquito Control, Roselle, IL). Ca rbon dioxide was delivered to the trap through 2-m long, 6.4 mm diamet er clear plastic Tygon tubing (S-50-HL, Saint-Gobain Performance Plastic, Akron, OH) taped to the un derside of the rain sh ield 5 cm from the trap inlet. Mosquitoes drawn into the trap inlet were blown thr ough a screen funnel and into a 1-quart glass Mason ja r containing a paper-wrapped 1 x 1 in dichlorvos vinyl strip
167 used as a killing agent (Hot Shot no-pest stri p, United Industries, St. Louis, MO). A 6 V, 12 A-h, rechargeable gel cell battery (Battery Wholesale Distributors, Georgetown, TX) provided power. Trapped mosquitoes were anesthetized with CO2, transferred to labeled paper cups (Solo Cup Company, Urbana, IL) and stored in a freezer until identified. Mosquitoes were identified to species using the ke ys of Darsie and Morris (2000). All Anopheles quadrimaculatus Say and An. crucians Wiedemann were pooled as these taxa could not be distinguished with confidence. Trap Modification One unmodified CDC light trap served as a control, 5 other traps were modified by replacing standard CM-47 incandescent bulbs with blue LEDs and/or replacing standard 6 V direct current (DC) mo tors with smaller 3 V DC motors (Model H202, 3 V DC, International Components Corpor ation, Westchester, IL). Trap combination and energy consumption is given in Table 7-1. Three volt motors operate in a range from 3 V to 6 V and perform well under a 6 V load. Light emitting diodes were arranged in both perimeter and cluster formations (Burkett 1998). Trap comparisons between transmitted light (perimeter arrangement) and reflected light (cluster arrangement, shining directly onto aluminum rain shields) (Fig. 7-1) we re made with this these arrangements. Standardand small motor-equipped traps were provided for each light arrangement for a total of 6 combinations. Four blue LEDs (P-466-ND, Digi-Key Co rporation, Thief River Falls, MN) were wired in parallel to motors (Fig. 7-2). Bl ue LEDs were rated at 650 millicandela, 470 nm wavelength, and a 30o viewing angle. Each diode was provided with a 180-ohm resistor to prevent over driving the LED. The clus ter arrangement of 4 adjacent LEDs was glued
168 to the lamppost with Welder All-Purpose adhesive (Homax Products, Inc., Chicago, IL). Perimeter arrangement consisted of 4 LEDs pointing outward, 90o apart from each other, and glued to the top of the trap body, just below the intake (Fig. 7-1). Standard model 512 CDC light traps cons ume 320 milliamperes of power per h (mA/h); the CM-47 lamp draws 150 mA/h and the standard motor draws 170 mA/h. The LEDs used in modified traps drew 20 mA. Small 3 V motor and LED combinations draw 120 mA/h. Standard motor and LED combin ations draw 150 mA/h. The 120 mA/h and 150 mA/h ratings were obtained with a di gital voltmeter (Extech model MM560, Omni Controls Inc., Tampa, FL), with an RS -232 computer interface, using data logger software. A mean average was taken from a 16 h burn-in period on each unit. Values for both unit specifications were rounded off to the nearest tenth unit (e.g., 118.56 mA/h rated as 120 mA/h). Airflow at trap intake was measured with a Turbo Meter anemometer (Davis Instruments Corp., Hayward, CA) 1.5 mm directly above the trap inle t with rain shield removed. The 6 V trap fan produced an air current of 2.9 m/s (6.5 mph), the small 3 V motor produced an air curre nt of 0.7 m/s (1.7 mph). Table 7-1 summarizes trap combination components, energy consumption, and days of effective operational use. Trial Location Two trials were conducted at the Univer sity of Floridas Horse Teaching Unit (HTU) and 1 trial at the universitys Austin Cary Memorial Forest (ACMF). The 65-acre HTU is a cleared pasture surr ounded by pine flatlands and hardwood forest. In the center of the unit is a shallow 8-acre pond supporting extensive growth of emergent vegetation predominated by cattail (Typha spp.), pickerelweed (Pontederia cordata), and duckweed (Lemna spp.), ideal for breeding Culex, Mansonia, and Coquillettidia mosquitoes. Austin
169 Cary Memorial Forest is a 2,040-acre pine flatwoods forest located 6 miles north of Gainesville. Pine flatwoods consist of dense to open c over pine forests with saw palmetto (Serenoa repens) and/or gallberry (Illex spp.) understory interspersed with swamps, grassy areas, wet prairies, or sa vannas (Minno et al. 2005) A cypress swamp borders the southern edge of ACMF. Criteri a used for site selection included large mosquito populations and gatedsecurity, present at both locat ions. Trials 1 and 2 were conducted at the HTU from 7-13 July and 12-18 September 2003. Trial 3 was conducted in ACMF from 18-24 September 2003. Statistical Analysis Three identical trials were run at 2 tests sites. Six traps were randomly set and rotated sequentially th rough 6 sites in a 6 x 6 Latin square design. Trap (treatment), position, and day effects were evaluated us ing a 3-way ANOVA (SAS Institute 2001) for the total number and most common mosquito species caught by the traps. Multiple comparisons were made using the Ryan-Ei not-Gabriel-Welsh (REGW) multiple range test ( = 0.05). All data were log10 (n + 1)-transformed prior to analysis. The 3 data sets were analyzed separately (not pooled) as mosquito species composition and population changed from mid to late summer. Results Trial 1. Trial 1 yielded 9,987 mosquito es over 6 trap days, including 96 males (0.96% of the catch). Means, standard errors, p-values, a nd significant differences for 14 species collected in sufficient numbers for anal ysis is included in Table 7-2. Mosquitoes excluded from analysis due to low numbers included Ae. vexans (2), Oc. atlanticus (2), and Ps. howardii (1). Collection totals cons isted primarily of 4 species, Culex erraticus, Mansonia titillans (Walker), Anopheles crucians s.l., and Coquillettidia perturbans,
170 which accounted for 93.9% of the total. Sign ificant trap, position, and day effects were seen for several species (Table 7-2). No significant difference was seen between trap capture means for each treatment (F = 1.80, df = 5, p = 0.15). Significant diffe rences between treatments were observed for 3 species: An. crucians s.l. (F = 2.74, df = 5, p = 0.048), Cx. salinarius (F = 4.89, df = 5, p = 0.004), and Cq. perturbans (F = 3.61, df = 5, p = 0.017). Significantly more An. crucians s.l. were caught in standard perimeter trap s than in small perimeter traps (p = 0.049). Culex salinarius was caught in significantly higher numbers in standard perimeter traps than in control, CDC small, and small cluster traps. Standard perimeter and standard cluster traps with blue LEDs caught more Cq. perturbans than did control or CDC small traps with incandescen t lights. More mosquito species were caught in control and standard perimeter traps ( 13 each) than in other trap combinations. The small cluster arrangement caught the fewest number of speci es (9). Standard cluster and standard perimeter traps caught the largest number of mosquitoes (means of 454.4 and 403.3, respectively), outperforming the cont rol (mean of 203.5) by 55.2% and 49.5% respectively, effectively doub ling the number of mosqu itoes caught (Table 7-2). Trial 2. A total of 83,886 mosquitoes we re collected over 6 days, of which 297 were males (0.35%). Means, standard errors and p-values are in cluded for 14 mosquito species listed in Table 7-3. Species trapped but not included in Table 7-3 (due to small counts) included Ae. vexans (1), Oc. atlanticus (1), Orthopodomyia signifera (Coquillett) (1), and Ps. howardii (3). Again, Cx. erraticus, Mn. titillans, An. crucians s.l., and Cq. perturbans were the 4 most abundant species, comp rising 98.5% of all mosquitoes taken. Significant differences in position and day effects were seen for several species (Table 7-
171 3). Most noticeable was the 8 x increase in September mosquitoes counts compared to July totals at same site (trial 1). No significant differences were observed between trap capture means (F = 2.46, df = 5, p = 0.06). Treatment was significant for 6 of the 14 species listed in Table 7-3. Anopheles crucians s.l. and An. quadrimaculatus s.l. were caught most often in standard perimeter traps as compared with CDC small traps. Significantly less Cq. perturbans were caught in CDC small traps than in othe r treatments (Table 7-3) Highly significant differences (F = 15.75, df = 5, p = 0.0001) were seen between Cx. erraticus capture means with standard cluster and control trap s outperforming small cluster and CDC small traps. Mansonia titillans was taken in significantly highe r numbers (F = 8.78, df = 5, p = 0.0219) in control, standard perimeter, small perimeter and standard cluster traps. No significant differences in Uranotaenia lowii counts was seen, however, counts were small. Only 1 trap (standard cluster) collected fewer Ur. sapphirina than other traps (F = 3.46, df = 5, p = 0.02). As seen in trial 1, the highest capture means were obtained in the standard cluster trap (3,187.6) followed by the standard perime ter trap (3,000.0). In this trial, the control did nearly as well as st andard cluster and standard perimeter traps (2,945.2). Total species captured per treatmen t included 13 (CDC small), 12 (control and small perimeter), and 11 (standard perimete r, standard cluster, and small cluster). Trial 3. A total of 33,996 mosquitoes were collected from ACMF over 6 nights, of which 299 were male (0.88%). Means, standard errors, p-values, and significant differences for 18 species collected are incl uded in Table 7-4. Other species caught but excluded from analysis due to low counts included Ae. albopictus (1), An. punctipennis (1), Oc. c. canadensis (1), and Ps. cyanescens (1). Species occurring most abundantly in
172 all traps includes An. crucians s.l., Cx. erraticus, Cs. melanura, Cx. nigripalpus, and Oc. atlanticus. Anopheles crucians s.l. and Cx. erraticus accounted for 37.5% and 33.6% of the entire catch, respectively. No significant difference was seen be tween traps means (F = 0.65, df = 5, p = 0.67). Significant treatment differences were seen in 5 of the 18 species listed in Table 74. Anopheles crucians s.l. responded less favorably to the CDC small trap than to all other (p = 0.05, REGW). Standard clus ter traps collected significantly more Cq. perturbans than did CDC small traps (p = 0.03) and Cs. melanura was collected more often in control traps than in CDC smallor small perimeter traps. Standard perimeter traps caught more Cx. erraticus than did other traps, and significantly more than did CDC small traps (p = 0.03). Control tr aps caught the la rgest number of Ps. columbiae. Eighteen mosquito species were collected in control, standard perimeter and standard cluster traps. The small perimeter trap collected 17 species followed by the CDC small and small cluster traps with 16 apiece. Tr ap means across all species were control (1,214.5) > standard cluster (1,155.3) > standard perimete r (1,068.8) > small perimeter (937.0) > small cluster (749.7) > CDC small (540.7). Discussion Medically important Florida mosquito spec ies collected during these trials include Cq. perturbans, Cq. titillans, and Cs. melanura, all involved with th e enzootic cycle of eastern equine encephalitis virus (Florida Coordinating Council on Mosquito Control 1998), Cx. nigripalus and Cx. salinarius, vectors of St. Louis en cephalitis and West Nile virus (Florida Coordinating Council on Mos quito Control 1998, Ru tledge et al. 2003), and An. crucians s.l. and An. quadrimaculatus s.l., malaria vectors in previous times when malaria was endemic to the Southeast United States (Harwood and James 1979).
173 Disease vectors and nuisance mosquito sp ecies are monitored throughout the state by vector abatement agencies as they gather fi eld data from mechanical trap collections. Decisions to apply pestic ide is often based trap counts, ther efore, it is necessary that any surveillance trap, regard less of model or brand used, be as good at attracting mosquitoes as CDC lightor New Jersey light traps. Mosquito surveillance is a la bor-intensive endeavor. Larv al surveillance is used to locate and treat problems breeding sites before biting adults can emerge. As this is the most time consuming survey and control activit y, mechanical traps are frequently used to monitor adult species composition and abundance in specific areas. Advantages of using battery-powered mechanical traps include sm all size, quick setup, and use in remote locations away from AC electricity sources Unfortunately, the energy required by CDC light traps from standard 6 V batteries precl udes trapping for more than 24 h. Daily attendance to these traps is costly in term s of manpower, labor, and time. CDC light traps modified to run for several days before collecting would serve to reduce surveillance costs, especi ally in remote areas. The response of Florida woodland mosquito es to different light color has been investigated in 2 studies. A li et al. (1989) used incandescen t lamps of 6 different colors and 4 intensities to trap mosquitoes in NJ LTs. They found that color was much more important than intensity to trap results. Of 6 colors, blue light was the most attractive among 17 species of mosquitoes (primarily Culex and Psorophora mosquitoes) caught, outperforming white, yellow, orange, green, and red light. In several comparison studies, Burkett et al. (1998) substituted different colored LEDs for incandescent lamps in CDC light traps. They found signi ficant color preferences among so me mosquito species given
174 a choice between CO2-baited CDC light traps equipped with incandescent, red, orange, yellow, green, infrared, and blue light. In candescent, blue, and green light-bait traps were found to be the most attractive among 17 mosquito species collected. These results led to a second series of attractancy test s between incandescent lamps, green LED, and blue LED equipped light trap s (Burkett 1998). Light emitting diodes were oriented in 2 directions: outward to assess transmitted light attraction and upward facing the rain shield to assess reflected light attr action. No significant difference was seen in collection totals between treatments in that study. From 23 mosquito species colle cted, only 4 species showed significant preferences among treatments. Positive results obtained in that study led to our studies incorporat ing small motors in combination with LEDs and incandescent lamps to extend the operational life of CDC light traps. No significant differences were seen between treatment means in trial 1. Small cluster and small perimeter traps caught a pproximately equal numbers of mosquitoes (means = 222.8 and 191.5, respectively) as the co ntrol (mean = 209.0). Of the 14 species collected in significant numbers, 13 species we re represented in the control, 10 species from small perimeter traps and 9 species from small cluster traps. The 4 major species present at that site (Cq. perturbans, Mn. titillans, Cx. erraticus, and An. crucians s.l.) were represented equally well among these 3 traps (Table 7-2). Significantly more Cx. salinarius and Cq. perturbans were caught in standard pe rimeter and standard cluster traps than in incandescent traps (control a nd CDC small traps). These findings suggest and Cx. salinarius and Cq. perturbans are more strongly attracte d to blue light (470 nm) than to incandescent light.
175 Standard perimeter and standard cluster tr aps are capable of 3 days of continuous operation (Table 7-1). They caught approxima tely twice as many mosquitoes than the control. Species composition included 13 for standard perimeter traps and 11 for standard cluster traps, comparable to cont rol species composition (13). The CDC small trap caught the least number of mosquitoes (mean = 167.8, 11 species). Results of trial 1 indicate that all LED equipped traps (standard and small motor traps) perform as well or better than the CDC light tr aps while operating 3 or 4 x as long (Table 7-1). No significant differences were seen between trap means in trial 2. Standard cluster (mean = 3,187.2), standard perimeter (3,000), and control traps (2,945.2) caught approximately twice as many mosquitoes as did small cluster (mean = 1,411.2) and CDC small traps (1,398.5). The small perimeter trap collected a mean of 2,039 mosquitoes, intermediate between standard and small mo tor trap results. The same 14 mosquito species were represented in trial 1 and tria l 2. Species composition agreed well with the results of Campbell (2003) trapping at th e same site a year earlier. Again, Cq. perturbans, Mn. titillans, Cx. erraticus, and An. crucians s.l. were the 4 predominate species trapped. As in trial 1, An quadrimaculatus s.l., Ps. columbiae, Ps. ferox, Ps. ciliata, and Ur. lowii were trapped in the smallest numbers and on occasions, absent from trap collections, usually from traps fitted with small motors. Results of trials 1 and 2 indicate that traps equipped with a standard CDC light trap motor in combination with blue LEDs are as effective at monitoring mos quito populations as are standard CDC light traps in open pasture/field se ttings in North Florida. No significant difference was seen between tr eatment totals in trial 3. Austin Cary Memorial Forest was characterized by pine flatwoods and cypress swamps, different
176 from the HTUs open pasture environment. Species composition was slightly larger, but predominate species were caught in smaller numbers compared with trials 1 and 2. Species composition and counts at this site were similar that that found by Burkett (1998). Anopheles crucians s.l., Cx. erraticus, Cs. melanura, Cx. nigripalpus, and Oc. atlanticus were the most abundant mosquitoes collect ed in this trial, the former 3 species showed significant differences between tr eatments. No preference was seen for incandescent light, reflected blue light, and transmitted blue light in An. crucians s.l., Cx. nigripalpus, and Oc. atlanticus. Culex erraticus preferred transmitted blue light to either reflected blue or incandescent light and Cs. melanura was most strongly attracted to incandescent light. Control, standard cluster, and standa rd perimeter traps obtained similar means of 1,214.5, 1,155.3 and 1,068.8, respectively. Results of our study indicate that modified CDC light traps are as efficient as standard traps in collecting Florida woodl and mosquitoes. Transmitted, reflected, and incandescent light produced no si gnificant difference in trap means in both forest and pasture habitats. These findings are simila r to those of Burkett (1998) in which no significant difference was found between blue LED light traps, green LED light traps, and unmodified CDC light traps. In our st udy, some mosquito species were caught in significantly smaller numbers in small motor traps as compared with standard motor traps, however, there was no significant diffe rence between trap totals (means) for all species. The obvious benefit in using LED-modified traps is that surv eillance-associated costs are greatly reduced as time, labor, and material (battery replacement, gasoline) is conserved collecting traps every third day as opposed to daily collecting. Further savings
177 could be obtained using small motor + LED tr ap combinations (serviced every fourth day), however, a small reduction in species co mposition and catch is likely to occur. Modified traps would best serve those agenci es surveying at locations distant from the office or military surveillance and vect or control teams with limited resources.
178 Table 7-1. Power consumption of standard a nd modified CDC light traps with effective operating days produced from 6 V, 12 A-h rechargeable gel cell batteries. Model 512 trap Lamp Motor Milliamps/h1 Operating days2 Control3 Incandescent Standard 320 1 CDC small Incandescent Small 230 2 Std. perimeter LED Standard 150 3 Small perimeter LED Small 120 4 Std. cluster LED Standard 150 3 Small cluster LED Small 120 4 1Average hourly energy consumption. 2Effective operating days. Excludes subsequent days in which a battery failed to maintain e ffective motor speed or completely discharged over the course of 24 h. 3Control trap is an unmodified J.W. Hock model 512 CDC light trap.
179Table 7-2. Trial 1 results of m odified light and motor CDC light trap counts with 500 mL/min CO2 (means SEM) at the Horse Teaching Unit. Means within each row having the same letter are not significan tly different (Ryan-Einot-Gabriel-Welsh Multiple Range Test). n = 6 days. Species Control CDC small Std. perim. Small perim. Std. cluster Small cluster p-value An. crucians s.l.1,2 30.37.6ab 27.37.7ab 38.010.3a 21. 87.6b 28.23.5ab 29.313.5ab 0.048 An. quadrimaculatus s.l. 1.00.4a 0.80.7a 1.70.7a 0. 50.3a 0.80.3a 0.80.4a 0.75 Cs. melanura 0.20.2a 0.00.0a 0.20.2a 0. 00.0a 0.00.0a 0.00.0a 0.61 Cx. erraticus2 85.313.0a 73.823.9a 173.598.1a 110. 225.7a 253.562.0a 63.528.1a 0.35 Cx. nigripalpus2 2.71.5a 1.51.1a 3.20.8a 1. 00.4a 2.81.0a 1.50.8a 0.39 Cx. salinarius1,2 2.50.8b 2.30.8b 12.33.3a 4.7 1.9ab 9.54.6ab 3.72.1b 0.004 Cq. perturbans1 13.34.4b 13.54.2b 43.524.3a 13. 83.5ab 38.79.1a 17.06.9ab 0.017 Mn. titillans1 63.533.2a 41.011.6a 124.087.3a 63. 215.3a 115.037.3a 68.541.7a 0.11 Oc. infirmatus1,2 2.81.7a 3.21.3a 4.21.0a 4. 31.8a 4.52.0a 2.51.8a 0.23 Ps. ferox 0.30.3a 0.20.2a 0.50.3a 0. 50.2a 0.70.7a 0.00.0a 0.55 Ps. ciliata 0.20.2a 0.00.0a 0.20.2a 0. 00.0a 0.00.0a 0.00.0a 0.61 Ps. columbiae 0.00.0a 0.00.0a 0.00.0a 0. 00.0a 0.20.2a 0.00.0a 0.44 Ur. lowii 0.30.3a 0.30.3a 0.80.5a 0. 00.0a 0.00.0a 0.00.0a 0.31 Ur. sapphirina1,2 1.71.0a 1.31.0a 1.20.5a 0. 50.3a 0.50.3a 0.70.2a 0.43 Trap mean3 SEM 209.054.4a 167.840.5a 412.2215.8a 222.839.4a 461.296.0a 191.593.7a 0.15 1Significant position eff ect (p < 0.05). 2Significant day effect (p < 0.05). 3Trap mean = trap sum of all mosquito species divided by 6 collection days.
180Table 7-3. Trial 2 results of m odified light and motor CDC light trap counts with 500 mL/min CO2 (means SEM) at the Horse Teaching Unit. Means within each row having the same letter are not significan tly different (Ryan-Einot-Gabriel-Welsh Multiple Range Test). n = 6 days. Species Control CDC small Std. perim. Small perim. Std. cluster Small cluster p-value An. crucians s.l.1,2 83.331.4ab 66.526.1b 106.230.4a 85. 720.3ab 70.88.8ab 65.710.8ab 0.05 An. quadrimaculatus s.l. 1.70.8ab 0.20.2b 2.30.8a 1.7 0.7ab 2.00.7ab 0.50.3ab 0.03 Cs. melanura 0.20.2a 0.20.2a 0.00.0a 0. 00.0a 0.00.0a 0.00.0a 0.61 Cx. erraticus1,2 868.5220.3a 269.769.7c 855.0153.6a 587.8125.7ab 1,020.0234.4a 377.369.2bc 0.0001 Cx. nigripalpus1,2 9.84.2a 8.85.5a 9.53.6a 9. 33.5a 9.23.3a 7.02.3a 0.92 Cx. salinarius1 2.81.5a 2.81.0a 4.81.4a 3. 21.4a 3.81.4a 2.20.7a 0.48 Cq. perturbans1,2 836.2197.6a 456.0152.3b 1,100.0328.2a 667.773.3a 1,054.2202.7a 623.2151.2ab 0.002 Mn. titillans1,2 1,112.0348.3a578.2182.2b 898.0283.0a 662.0187.7a 1,005.5342.4a 326.0115.0b 0.0002 Oc. infirmatus1 2.81.7a 3.32.5a 2.31.2a 5. 23.3a 5.23.5a 3.71.3a 0.47 Ps. ferox 0.00.0a 0.20.2a 0.00.0a 0. 00.0a 0.00.0a 0.00.0a 0.44 Ps. ciliata 0.00.0a 0.00.0a 0.00.0a 0. 20.2a 0.30.2a 0.00.0a 0.08 Ps. columbiae1,2 9.35.9a 4.32.4a 5.52.8a 4. 21.6a 3.81.4a 1.20.3a 0.37 Ur. lowii1 0.80.5a 1.00.6a 1.81.3a 1. 00.8a 0.30.2a 0.30.2a 0.17 Ur. sapphirina1 3.52.2a 2.71.5a 2.81.4a 1.51.1a 0.50.5b 0.70.4a 0.02 Trap mean3 SEM 2,945.2712.5a1,398.5330.9a 3,000.0550.8a 2,039.0266.6a 3,187.2719.8a 1,411.2306.1a 0.06 1Significant position effect (p < 0.05). 2Significant day effect (p < 0.05). 3Trap mean = trap sum of all mosquito species divided by 6 collection days.
181 Table 7-4. Trial 3 results of m odified light and motor CDC light trap counts with 500 mL/min CO2 (means SEM) at Austin Cary Memorial Forest. Means within each row having the same letter are not significan tly different (Ryan-Einot-Gabriel-Welsh Multiple Range Test). n = 6 days. Species Control CDC small Std. perim. Small perim. Std. cluster Small cluster p-value An. crucians1,2 367.8209.7a 196.282.0b 369.7221.8a 394.2189.7a 537.0328.0a 262.7100.4ab 0.003 An. quadrimaculatus s.l.1,2 2.30.9a 0.50.2a 1.80.6a 0. 70.3a 0.70.2a 1.50.8a 0.052 Cq. perturbans1,2 9.24.0ab 1.50.4b 4.73.0ab 3. 01.0ab 17.39.4a 7.75.1ab 0.03 Cs. melanura1,2 104.732.7a 37.011.7b 76.734.2ab 47. 016.9b 79.225.9ab 74.033.1ab 0.002 Cx. erraticus1,2 362.3106.9ab 167.867.1b 421.7144.0a 346.7112.3ab 330.383.5ab 275.2117.5ab 0.03 Cx. nigripalpus2 101.829.4a 53.010.6a 67.812.3a 43. 711.3a 70.718.7a 39.59.7a 0.11 Cx. salinarius1,2 16.09.5a 5.82.6a 5.71.9a 7. 33.0a 13.29.1a 9.07.8a 0.63 Mn. titillans1 11.54.8a 10.57.9a 6.03.1a 8. 25.7a 13.57.5a 8.75.3a 0.52 Oc. atlanticus1,2,4 37.518.1a 30.57.0a 62.225.2a 47. 817.0a 64.529.8a 41.813.2a 0.25 Oc. dupreei1 3.52.4a 2.21.1a 3.51.7a 1. 20.7a 2.71.4a 1.81.5a 0.57 Oc. infirmatus1 8.32.5a 6.82.5a 6.33.5a 2. 51.5a 3.22.2a 8.23.5a 0.11 Oc. triseriatus1 0.30.2a 0.00.0a 0.70.4a 0. 20.2a 0.20.2a 0.00.0a 0.16 Ps. ciliata 1.81.0a 0.50.3a 0.70.2a 0. 30.2a 0.30.2a 0.80.5a 0.72 Ps. columbiae1,2 11.24.7a 2.81.0ab 2.31.3b 6. 03.0ab 6.04.4ab 3.81.7ab 0.02 Ps. ferox1 9.35.9a 4.22.6a 9.76.9a 7. 75.9a 2.81.9a 2.82.1a 0.16 Ps. howardii 0.20.2a 0.00.0a 0.20.2a 0. 00.0a 0.30.2a 0.00.0a 0.35 Ur. lowii1 2.31.4a 1.80.9a 2.01.0a 2. 20.9a 0.30.2a 0.20.2a 0.04 Ur. sapphirina 20.310.7a 12.55.7a 15.06.7a 8. 33.1a 6.02.4a 5.02.2a 0.07 Trap mean3 SEM 1,214.5335.9a 540.7108.6a 1,068.8398.7a 937.0276.0a 1,155.3486.0a 749.7188.7a 0.67 1Significant position eff ect (p < 0.05). 2Significant day effect (p < 0.05). 3Trap mean = trap sum of all mosquito species divided by 6 collection days. 4Could not be distinguished from Oc. tormentor.
182 A B Figure 7-1. Light emitting diode-modified CDC li ght traps. A) cluster arrangement. B) perimeter arrangement. A B Figure 7-2. Wiring schematic of light emitting diode-m odified CDC light traps. A) Cluster arrangement of 4 LEDs mount ed on lamppost with light directed upward onto rain shield. B) Perimeter arrangement with LEDs spaced 90o apart with light directed outward. All measurements made in mm. Not drawn to scale.
183 CHAPTER 8 SURVEILLANCE AND CONTROL OF Aedes albopictus: THE IMPORTANCE OF TRAPS, ATTRACTANTS AND ADULTICIDES Introduction Aedes albopictus, the Asian Tiger mosquito, is a newly introduced pest to North America. As such, only recently has intense st udy of this mosquito in the United States begun. Overseas, it is a known vector of several important pathogens including the viruses of dengue and yellow fever. Circum stantial evidence incrimin ated this mosquito as the vector in a small outbreak of dengue fever in Hawaii during 2001 and 2002. Stateside, it has been found na turally infected with eastern equine encephalitis (Mitchell et al. 1992), West Nile viru s (Holick et al. 2002), Cache Valley virus (Mitchell et al. 1998) and La Crosse encephalitis virus (Gerha rdt et al. 2001). In addition, it is an efficient vector of dog heartworm in the Un ited States (Estrada-Franco and Craig 1995). Aedes albopictus spread rapidly across much of the Southeast and Midwest United States, primarily through interstate transport of egg-laden tires. In less than 20 years, it colonized more than 1,000 counties in 32 stat es following its initial discovery in Harris County, Texas. First discovered at a Jacksonville tire repository in 1986, Ae. albopictus had colonized all 67 Florida counties by 1995. Its expansion has been slowed only in areas with an average winter isotherm of -5o C or less and/or by dry, arid climates typical of the western United States. Thus, this spec ies represents a relatively new surveillance and control challenge to Ameri can vector control agencies.
184 Traps, Trapping, and Attractants One of the operational pillars of organized mosquito control agencies is the use of mechanical surveillance gear to monitor adult mosquito populati ons through time and space. These devices enable operators to determine the seasonal abundance and distribution of mosquito species in areas of concern. Traps are also important for vector species identification during mosquito-borne disease outbr eaks. In both routine and disease surveillance, data gained are used to determine where and when control efforts are needed. Two adult mosquito traps have been the mainstay of mosquito surveillance programs in the United States, the New Jersey Light trap (NJLT) and the Centers for Disease Control and Prevention (CDC) light trap Both traps use light as an attractant; the NJLT is powered by AC electricity and the CDC trap with batteries. Unfortunately, Ae. albopictus does not respond well to light traps (Thurman and Thurman 1955, Service 1993). Neither does its closely related sibling species, Ae. aegypti (Christophers 1960). Aedes aegypti has been present in the United States since colonial times and traps speci fically designed to capture it have recently been developed (Fay and Prince 1970, Wilton and Kloter 1985). The Fay-Prince trap and Wilton trap attracts and catches Ae. albopictus as well. These traps are Aedes surveillance devices only, they have yet to prove themselves useful as part of an integrated control program. Two investigative aspects of this disserta tion were to determine if newly developed residential traps were more effective than surveillance traps in collecting Ae. albopictus and if commercially available attractants could positively influence trap capture. During the past 10 years, a number of reside ntial mosquito control traps have been developed and marketed for homeowner use. Th ey occur in a variety of sizes, shapes and attraction factors. Attraction factors include visual features (black color, contrasting
185 black and white interfaces or patterns), s ound, water vapor, heat, light, and generation of CO2 to improve catch rates. Several models can be supplemented with octenol and lactic acid, the only two commercia lly marketed attractants. Several residential traps were chosen for field trials based on results obtained in large-cage trials (Chapter 2). Residentia l traps were field tested against those surveillance traps best suited for capturing Ae. albopictus in suburban environments where this pest was prevalent and both trap types were likely to be used by homeowners and vector control agencies. A CDC light tr ap, typically used in mosquito surveillance efforts, was added as a control. Significantly more Ae. albopictus were captured in residential traps than in surveillance traps (p < 0.0001). Residential traps have additional attraction factors lacking in surveillance traps: octenol and lactic acid baits, water vapor, and heat, which when added to CO2 simulates breath. Results of our study agree with the findings of Dennett et al. (2004) in which a re sidential trap achieved significantly better Ae. albopictus capture than other surveillance traps. This is not surprising given that Aedes mosquitoes are strongly attracted to heat, water vapor, and contrasting color (Peterson and Brown 1951, Sippell and Brown 1953, Brown 1953, Christophers 1960, Wood and Wright 1968). Except for the Fay-Prin ce trap, as the number of trap attractant features increased, so did the traps effectiveness. Color contrast, black, heat, and water vapor were attributes of effective traps. Light, sound, and adhesive traps were less effective. The effects of commercial at tractants on trap capture of Aedes albopictus field populations were assessed in Chapter 3. Mos quito Magnet Pro traps baited with octenol, lactic acid and octenol + lactic acid were tested against an u nbaited (control) MM Pro
186 trap. Attractants used in combination were more effective than when used separately, octenol + lactic acid-baited tr aps caught the largest number of Ae. albopictus. Octenol, blended with lactic acid, attracted significantly more Ae. albopictus than did octenol alone and considerably more than lactic acid and control treatments. This was an interesting finding in light of the fact that octenol-baited traps captured the least number of Ae. albopictus. Lactic acid has been shown to be more effective when blended with acetone, dichloromethane or dimethyl disulfide than by itself as an attractant for Ae. aegypti (Bernier et al. 2003). These two attractants, in the presence of CO2, apparently better mimic hosts than either attractant with CO2 or CO2 alone. Lactic acid-baited traps caught the second larg est number of Ae. albopictus. While lactic acid is repellent to Ae. albopictus in concentrations high er than occurs on human skin (Shirai et al. 2001 ), it is a proven attractant for this and other Aedes (Ochlerotatus) spp. in concentrations found on human skin (Acree et al. 1968, Klin e et al. 1990, Ikeshoji 1993). Although lactic acid recep tors have not been found on Ae. albopictus, they are known to occur on the antennae of Ae. aegypti (Acree et al. 1968, Davis and Sokolove 1976). Lactic acid-baited trap s performed better than the control traps. It should be remembered that the MM Pro generates its own CO2 and that CO2 is essential for positive trap results (Gillies 1980, Mboera and Takke n 1997). Acree et al. (1968) caught a much higher percentage of adult Ae. aegypti with CO2 + lactic acid baited tubes as opposed to CO2 alone, the results of our field trials agree with these laboratory results. Lactic acid depressed captur e rates of other mosquito species in our study. The four most abundant species, Culex nigripalpus, Cx. erraticus, Psorophora ferox and Ochlerotatus infirmatus were strongly attracted to c ontrol or octenol-baited traps.
187 Similar depressive effects with lactic acid attr actants were seen previous studies in other Culex species (USAEHA 1970, Stryker and Young 1970) and Oc. taeniorhynchus (Kline et al. 1990). Octenol-baited traps collect ed the fewest number of Ae. albopictus in the field, but not significantly less than the cont rol. Octenol lightly depressed Ae. albopictus capture rates, similar to previous Hawaiian field results in MM Pro trap s (Sean Bedard, ABC, personal communication). These results also agree with a study in which approximately equal numbers of Ae. albopictus were captured in Fay-Prince traps baited with octenol + CO2 and traps baited with just CO2 (Shone et al. 2003). Octenol production in mammals is most commonly associated with ruminant s. Two blood meal analysis studies of Ae. albopictus collected within the continental United States indicated that ruminants were important hosts. Savage et al. (1993) found deer to be the second most common host and Niebylski et al. (1994) found th at cattle were the fourth most common blood source in field collected Ae. albopictus. Blood hosts also included ruminants, humans, rodents, turtles, and birds. This is not surprising c onsidering the opportunistic feeding behavior of Ae. albopictus (Watson 1967). Thus, if the goal of tr apping is to collect more than just Ae. albopictus, octenol should be used as it is attr active to many mosquito species. If Ae. albopictus is the target of trapping, octenol + la ctic acid baits are the best for optimal capture of this species. Trap collections of Ae. albopictus were significantly reduced between the first trial and the later two trials (p = 0.0002, trials 1 and 2; p = 0.0005, trials 1 and 3, Tukeys multiple comparison test) from 4 test sites. Aedes albopictus populations throughout Gainesville increased slightly during this time (Appendix B). Based on this result,
188 octenol + lactic acid-baited MM Pro traps would be a candidate for use with an integrated mosquito control program. Removal trappi ng has long been a goal of vector control agencies and recent successes with attractan t-baited tsetse fly traps have facilitated replacement of routine aerial insecticide a pplications for insecticide-treated traps in Zimbabwe (Vale 1993). Mansonia mosquito populations was reduced using insecticidebaited sound traps (Ka nda et al. 1990) and Ae. albopictus was greatly reduced over a month using insecticide-treate d sound traps in Malaysia (I keshoji and Yap 1990). Thus, some highly effective residential mosquito traps could be used to control, not just survey, targeted species. Use of traps for mosquito co ntrol could fit nicely into an integrated pest management program that included water management, sanitation, pesticide rotation, biological control agents, and other as pects of mosquito control programs. Surveillance traps designed specifically to catch Ae. aegypti are also effective in capturing Ae. albopictus. The Fay-Prince trap caught significantly more Ae. albopictus than the Wilton trap and more than the CDC li ght trap. Color contrast was shown to be more important than black or incandescent light in surveillance traps. Residential traps, which performed significantly better than surveillance traps, had between 6 and 8 attractant factors (Table 4-1). The addition of heat, water vapor, octenol + lactic acid, and CFG technology found in residential traps played a decisive role in superior Ae. albopictus capture rates. Pesticide Response Adulticides are the largest selling and mo st commonly used insecticides of most Florida mosquito control agencies (Flori da Coordinating Council on Mosquito Control 1998). Source reduction is the most efficient method in reducing Ae. albopictus populations. Larvaciding to control this mos quito is often impractical because of the
189 small number of eggs laid in tree holes, ra in gutters, birdbaths, and other natural and artificial containers. Larvacidi ng is best practiced on those mo squito species that breed in large numbers in temporary ground water situat ions such as roadside ditches and rain water pools in pastures, sites in which Ae. albopictus does not breed (Watson 1967). Therefore, adulticides are often used to provide quick c ontrol of biting adult females in suburban settings and at tire repositories or waste piles (a lthough larvacides are often effectively used on tire piles). Should Ae. albopictus become resistant to the limited number of adulticides available to mosqu ito control agencies, control personnel would find it difficult to manage this mosquito in suburban neighborhoods. Aedes albopictus is still susceptible to 5 di fferent chemical compounds most frequently used for adult mosquito control in Florida today. Da ta from 6 separate locations demonstrated that at the lethal concentration 50 (LC50) level, resistance ratios (RRs) were less than 10 x that of susceptible laboratory mosquitoes (RRs are obtained by dividing field LC50s by susceptible colony LC50s). Resistance ratios exceeding 10 x indicate resistance (Sames et al. 1996). At the higher LC95 level, 2 field populations demonstrated a greater than 10 x RR, one to resmethrin and one to malathion. A proportion of those populations are showing sign s of resistance, but those population as a whole are still susceptible. Although th e biochemical mechanisms responsible for elevated levels of tolerance in these populations were not expl ored, it is very likely that detoxification enzymes are responsible. Mixed function oxidases, hydrolases, and glutathion transferases are most often res ponsible for metabolic pesticide resistance (Wilkinson 1983).
190 Increased production of esterases and oxidases are commonly responsible for increased tolerance to insecticides in Ae. aegypti and Ae. albopictus. Excessive production of esterases and oxidases were responsible for elevated tolerance in Venezuelan populations of Ae. aegypti to OP and carbamate insecticides, respectively (Mazzarri and Georghiou 1995). Elevated este rase activity was responsible for resistance in Virgin Island samples of Ae. aegypti to an OP larvacide, temephos (Wirth and Georghiou 1999). Malathion resistance in Ae. albopictus was reported from Vietnam in 1970 (Herbert and Perkins 1973), but only in areas intensely treated with malathion for malaria control. It was likely due to in creased esterase activ ity. This study found Ae. albopictus resistant to DDT, which had been used as a local larvacide for many years. Chinese strains of Ae. albopictus have been reported to be resistant to DDT because of increased production of DDT-dehydrochlorinase, another detoxification enzyme (Neng et al. 1992). Fortunately, the more serious knockdown resistance mechanism, in which nerve cell target site s (sodium channels) have changed to prevent pesticide binding, has not been observed in this species. These data suggest that the developmen t of resistance to insecticides in Ae. albopictus is slow and that increased enzyme production is the main component of resistance. Organophosphate-resistant Ae. aegypti was largely unknown until the early 1970s despite 20 years of prior malathion use. By that time, many other mosquito species developed resistance to insecticid es (W.H.O. 1986). Sporadic treatment of Ae. albopictus trouble sites (tire piles), rotation between OP and pyr ethroid insecticides, use of larvacides with different modes of action, and the diurnal feeding habit of Ae. albopictus, possibly removing it from harms way in the course of late evening and
191 nighttime ULV spray operations, have all played roles in lessening selection pressure on this mosquito and consequently retarding resistance. Research, development, and registration of new classes of insecticides with novel modes of action for mosquito control is cu rrently underway at the USDA ARS Center for Medical, Agricultural and Veterinary Entomology (CMAVE) in Gainesville, Florida. Compounds now registered for ur ban pest control may soon prove effective as mosquito larvacides or adulticides. These include ch loronicotinyls (imidacloprid, a nicotinergic acetylcholine receptor antagonist), phenyl pyrazoles (fipronil, a gamma amino butyric acid receptor antagonist), avermectins (glutamate receptor agonists) and pyrroles (chlorfenapyr, an inhibitor of oxidative phosphorylation). In a ddition, novel compounds demonstrating insecticidal ac tivity will be tested. Integr ation of compounds that use different modes of action to kill insects from traditional adulticides could prove valuable in controlling mosquitoes resistant to OP and pyrethroid insecticides. Intensive research into new residential mosquito control traps and attractants is also ongoing at CMAVE. As seen in this and othe r research at CMAVE, some of these traps and attractants show promise as control devices that could be used in integrated mosquito control programs. Light emitting diodes laboratory studies (Chapter 6) indicate that UV light (380 nm) and blue light ( 470 nm) is more attractive to Ae. albopictus than incandescent light. Orientation of LEDs to make use of transmitted and reflected light resulted in no significant preferences among most mosquito species in field tests (Chapter 7). The way is now set to test UV and bl ue light LED-modified traps in multi-trap comparison studies targeting Ae. albopictus. Future control practices targeting Ae. albopictus, a new exotic nuisance and disease vector recently introduced into the United
192 States, may soon rely on integrated contro l programs using traps, attractants, and adulticides recommended from our study.
193 APPENDIX A LARGE-CAGE Aedes albopictus CAPTURE RESULTS WITH RESIDENTIAL AND SURVEILLANCE MOSQUITO TRAPS The following data are capture results for various commercial and surveillance traps used in large screened cages at the United States Department of Agricultures Center for Medical, Agricultural and Veterinary Entomology Unit, Gainesville, FL. All data represent raw numbers of Ae. albopictus caught. Traps were set and adult females were released between 0800 and 1200 and co llected 24 h later. Single-border boxes indicate evening release of adult females (approximately 1 h before sunset). Doublebordered boxes represent half th e original trap capture to adjust for later reductions in release rates (from 1,000 to 500). CO2 = 500 mL/min, MM Liberty CO2 = 420 mL/min, MM Pro CO2 = 520 mL/min. Table A-1. Trial counts, means, and treatments (trap type) of Ae. albopictus in large-cage trials at USDA ARS Gainesville, Florida. Trap Trial Mean 1234567 MM Pro 48475010230112 64.83 Bites/station 1 987215310 18.00 2 620916213 12.50 3 6314842192 24.67 total 215964799315 55.17 Cage 112221 MM Pro + Oct 1681515387105126 115.00 Bites/station 1 791801213 9.83 2 225621183 18.50 3 232344014 15.83 total 52888312120 44.17 Cage 121122 MM Liberty 149651014359297574.43 Bites/station 1 18031571418.29 2 5248101715013.86 3 8062251242021.29 total 312113503671143.43 Cage 2211211 MM Liberty + Oct 591301494410628209103.57 Bites/station 1 75342027213.57 2 171833225210.00 3 1214321766.43 total 36851073591030.00 Cage 2211121
194 Table A-1. Continued Trap Trial Mean 123456 7 Wilton 3' + CO2 112719154715 4525.57 Bites/station 1 10901285 209.14 2 11191621 66.57 3 38211156 010.29 total 59492191512 2626.00 Cage 111222 2 Wilt 15-20" + CO2 60.526140686923 4862.07 Bites/station 1 44221621 09.57 2 105276013 119.14 3 253512708 012.43 total 174392129222 141.14 Cage 211221 CDC 5' + CO2 1536632553115 51.17 Bites/station 1 832040946 27.00 2 2717320316 15.83 3 146190313 9.17 total 124435501575 52.00 Cage 221112 CDC 15-20" + CO2 236645765358 53.50 Bites/station 1 19061314 7.17 2 40241424 11.00 3 2142103 5.17 Total 6131448311 23.33 Cage 212211 Fay-Prince + CO2 127161854324241 113.50 Bites/station 1 526007 3.33 2 1453120 4.17 3 5410013 9.83 Total 73891310 17.33 Cage 112122 MM-X + CO2 93.5713939146258 107.75 Bites/station 1 21329612 3.33 2 34040914 4.17 3 3510331302 9.83 total 40631022828 17.33 Cage 1 12122 MM-X, oct + CO2 2912.59910854260 93.75 1 181611924 60 2 11176543 46 3 1124193113 71 total 405736171710 167 Cage 121212
195 Table A-1. Continued Trap Trial Mean 1234567 Bugjammer 1.534 2.83 1 1021 4.33 2 2511 9.00 3 2183 10.67 total 56115 24.00 Cage 121 Mosq Deleto 2200 10.51 0.83 1 2288 12.67 2 63232 23.33 3 22424 16.67 total 306464 52.67 Cage 212
196 APPENDIX B Aedes albopictus CAPTURE TOTALS IN CDC LI GHT TRAPS AT SIX SITES IN GAINESVILLE, FLORIDA Table B-1. Gainesville Ae. albopictus counts from 6 light traps in Gainesville, Florida. Trap CDC 1 CDC 2 CDC 3 CDC 4 CDC 5 CDC 6 Monthly Site (street) NW 13 PL NW 7 LN NW 22 ST NW 14 AVW CLUB FINLEYS Total August 4 0 3 2 0 4 2 5 0 2 1 0 0 1 6 0 2 3 0 2 0 11 0 3 4 0 2 1 12 0 6 4 1 2 6 13 0 5 4 T/F 2 1 18 0 5 6 1 1 2 19 2 2 4 0 T/F 2 20 3 6 1 0 4 1 25 1 1 T/F 1 3 4 26 2 3 1 1 3 4 27 2 1 0 1 0 1 31 1 2 3 1 5 6 Total 11 41 33 6 28 31 150 Sept.4 0 0 5 0 3 1 9 0 0 8 4 5 1 10 2 2 8 2 8 2 15 6 4 6 6 3 0 16 4 1 3 5 6 3 17 2 3 4 6 2 1 22 3 6 0 0 1 6 23 1 5 1 2 4 4 24 3 1 3 1 0 1 29 2 1 3 2 0 4 30 4 6 3 3 0 1 Total 27 29 44 31 32 24 187
197 0 5 10 15 20 25 30 35 40 45 50 CDC1CDC2CDC3CDC4CDC5CDC6 CDC trapMonthly Ae. albopictus capture August Sept. Figure B-1. August and September 2003 Aedes albopictus trap totals for each of 6 CDC light traps set in residential nei ghborhoods in Gainesville, Florida.
198 APPENDIX C PESTICIDE DILUTIONS FOR SUSCEPTIBILITY STUDY Pesticide Dilutions from labeled con centration to end use concentration. 1. Fyfanon ULV (Malathion) = 9.9 lb/gal 9.9 lbs./gal. x 453.6 g/lb. = 4490.64 g a. i./gal x 1 gal./3.785 l = 1186.431 g/L 1186.431 mg malathion/mL Fyfanon ULV 1186.431 mg = 1000 mg 1 mL X mL X mL = 1000 mg = .843 mL = 843 l malathion 1186.431 mg 843 l Fyfanon ULV = 1000 mg malathion .843 mL Fyfanon ULV = 1,000,000 g + 9.157 mL acetone 10 mL solution = 1,000,000 g malathion (100,000 g/mL) 1 mL solution = 100,000 g malathion + 9 mL acetone 10 mL solution = 100,000 g malathion (= 10,000 g/mL) 1 mL solution = 10,000 g malathion + 9 mL acetone 10 mL solution = 10,000 g malathion (= 1,000 g/mL) 1 mL solution = 1,000 g malathion + 9 mL acetone 10 mL solution = 1,000 g malathion (= 100 g/mL) 1 mL solution = 100 g malathion + 9 mL acetone 10 mL solution = 100 g malathion (= 10 g/mL) 1 mL solution = 10 g malathion + 9 mL acetone 10 mL solution = 10 g malathion (= 1 g/mL)
199 1 mL solution = 1 g malathion + 9 mL acetone 10 mL solution = 1 g malathion (= 0.1 g/mL)
200 2. Dibrom (Naled) = 14.1 lb/gal 14.1 lbs./gal x 453.6 g/lb. = 6395.76 g/ gal. x 1 gal./3.785 l = 1689.765 g/L 1689.765 mg/mL 1689.765 mg = 1000 mg 1 mL X mL X mL = 1000 mg = .592 mL = 592 l Dibrom 1689.765 mg 592 l Dibrom = 1000 mg naled 0.592 mL naled + 9.408 mL acetone 10 mL solution = 1000mg naled (= 100,000 g/mL naled) 1 mL solution = 100,000 g naled + 9 mL acetone 10 mL solution = 100,000 g naled (= 10,000 g/mL naled) 1 mL solution = 10,000 g naled + 9 mL acetone 10 mL solution = 10,000 g naled (= 1,000 g/mL naled) 1 mL solution = 1,000 g naled + 9 mL acetone 10 mL solution = 1,000 g naled (= 100 g/mL naled) 1 mL solution = 100 g naled + 9 mL acetone 10 mL solution = 100 g naled (= 10 g/mL naled) 1 mL solution = 10 g naled + 9 mL acetone 10 mL solution = 10 g naled (= 1 g/mL naled) 1 mL solution = 1 g naled + 9 mL acetone 10 mL solution = 1 g naled (= 0.1 g/mL naled)
201 3. Anvil 10 + 10 ULV (D-phenothrin) = 0.74 lb/gal 0.74 lbs./gal x 453.6 g/lb. = 335.664 g/gal. x 1 gal/3.785 l = 88.683 g/L 88.683 mg/mL 88.683 mg = 100 mg 1 mL X mL X mL = 100 mg = 1.13 mL (1130 l) Anvil ULV = 100 mg d-phenothrin 88.683 mg 1.13 mL Anvil ULV + 8.87 mL acetone 10 mL solution = 100 mg d-phenothr in (10,000 g d-phenothrin/mL solution) 1 mL = 10,000 g d-phenothrin + 9 mL acetone 10 mL solution = 10,000 g d-phenothrin (1,000 g/mL) 1 mL solution = 1,000 g d-phenothrin + 9 mL acetone 10 mL solution = 1,000 g d-phenothrin (100 g/mL) 1 mL solution = 100 g d-phenothrin + 9 mL acetone 10 mL solution = 100 g d-phenothrin (10 g/mL) 1 mL solution = 10 g d-phenothrin + 9 mL acetone 10 mL solution = 10 g d-phenothrin (1 g/mL) 1 mL solution = 1 g d-phenothrin + 9 mL acetone 10 mL solution = 1 g d-phenothrin (0.1 g/mL)
202 4. Biomist 4 + 4 ULV (Permethrin) = 0.3 lb/gal 0.3 lbs./gal. x 453.6 g/lb. = 136.08 g/gal x 1 gal/3.785 l = 35.952 g/L 35.952 mg/mL 35.952 mg = 100 mg 1 mL X mL X mL = 100 mg = 2.78 mL Biomist ULV = 100 mg permethrin 35.952 mg 2.78 mL Biomist ULV = 2780 l Biomist ULV = 100 mg permethrin 2.78 mL + 7.22 mL acetone 10 mL solution = 100 mg permethr in (10,000 g permethrin/mL solution) 1 mL = 10,000 g permethrin + 9 mL acetone 10 mL solution = 10,000 g permethrin (1,000 g/mL) 1 mL = 1,000 g permethrin + 9 mL acetone 10 mL solution = 1,000 g permethrin (100 g/mL) 1 mL = 100 g permethrin + 9 mL acetone 10 mL solution = 100 g permethrin (10 g/mL) 1 mL = 10 g permethrin + 9 mL acetone 10 mL solution = 10 g permethrin (1 g permethrin/mL) 1 mL = 1 g permethrin + 9 mL acetone 10 mL solution = 1 g permethrin (0.1 g permethrin/mL)
203 5. Scourge 4 + 12 (Resmethrin) = 0.3 lbs/gal 0.3 lbs./gal. x 453.6 g/lb. = 136.08 g/gal x 1 gal/3.785 l = 35.952 g/L 35.952 mg/mL 35.952 mg = 100 mg 1 mL X mL X mL = 100 mg = 2.78 mL Scourge = 100 mg resmethrin 35.952 mg 2.78 mL Scourge = 2780 l Scourge = 100 mg resmethrin 2.78 mL + 7.22 mL acetone 10 mL solution = 100 mg resmethr in (10,000 g resmethrin/mL solution) 1 mL = 10,000 g resmethrin + 9 mL acetone 10 mL solution = 10,000 g resmethrin (1,000 g/mL) 1 mL = 1,000 g resmethrin + 9 mL acetone 10 mL solution = 1,000 g resmethrin (100 g/mL) 1 mL = 100 g resmethrin + 9 mL acetone 10 mL solution = 100 g resmethrin (10 g /mL) 1 mL = 10 g resmethrin + 9 mL acetone 10 mL solution = 10 g resmethrin (1 g /mL) 1 mL = 1 g resmethrin + 9 mL acetone 10 mL solution = 1 g resmethrin (0.1 g/mL)
204 APPENDIX D CIRCUIT DESCRIPTION OF 555 FREQUENCY GENERATORS The LED flashing circuit was constructed from a 555 precision timer integrated circuit (IC) connected for stable operation so that it operates as a multivibrator. The frequency of the multivibrator is determined by the time it takes the capacitor C1 to charge and discharge between the threshold-vo ltage level and the tri gger-voltage level of the IC. Capacitor C1 (various capacitance, Appendix C) is charged through resistors R1 (22,000 ohms), R2 (180,000 ohms) and R3 (50,000 ohms) and discharges through R2 and R3. In this circuit, R2 and R3 are connected in series and form a single resistance for the charge and discharge path. The combined resistance of R2 and R3 is approximately 10 x the resistance of R1 that produces a duty cycl e near 50%. Resistor 3 is variable to allow the frequency to be adjusted through a limited range. The LED is connected to the output of the 555 IC in series with the load resist or R4 (270 ohms). During half of each cycle, the output of the IC is pulle d to ground potential and curre nt flows through R4 and the LED causing the LED to emit light. During the ot her half of the cycle, the output of the 555 IC is held high and no current flows thr ough the LED. The value of R4 sets the maximum current through the LED to approx imately 22 mA during the on period. The maximum forward steady current is 25 mA.
205 APPENDIX E CAPACITANCE IN MICRO FARADS OF TEN DIFFERENT FREQUENCY GENERATING 555 INTEGRATED CIRCUITS Table E-1. Capacitance of 10 different frequency-generating capacitors. Frequency C1 Capacitor rating ( F) 10 0.33 30 0.11 40 0.082 60 0.054 120 0.03 150 0.02 200 0.015 500 0.0068
206 LIST OF REFERENCES Abbott, W. S. 1925. A method of computing th e effectiveness of an insecticide. J. Econ. Entomol. 18:265-267. Acree, F. Jr., R. B. Turner, H. K. Gouck a nd M. Beroza. 1968. L-lactic acid: a mosquito attractant isolated from humans. Science 161:1346-1347. Ali, A. and J. K. Nayar. 1997. Inva sion, spread, and vector potential of Aedes albopictus in the USA and its control possibiliti es. Med. Entomol. Zool. 48(1):1-9. Ali, A., J. K. Nayar, J. W. Knight and B. H. Stanley. 1989. Attraction of Florida mosquitoes (Diptera: Culicidae) to artificial light in th e field. Proceedings Paper 57th Annual Conference of th Califor nia Mosquito and Vector Control Association:82-88. Ali, A., J. K. Nayar and R. Xue. 1995. Comp arative toxicity of se lected larvacides and insect growth regulators to a Florida laboratory population of Aedes albopictus. J. Am. Mosq. Control Assoc. 11(1):72-76. Allan, S. A., J. F. Day and J. D. Edman. 1987. Visual ecology of biting flies. Ann. Rev. Entomol. 32:297-316. Barr, A. R., T. A. Smith, M. M. Boreham and K. E. White. 1963. Evaluation of some factors affecting the efficiency of light traps for collecting mosquitoes. J. Econ. Entomol. 56:123-127. Bernier, U. R., D. L. Kline, K. H. Posey, M. M. Booth, R. A. Yost and D. R. Barnard. 2003. Synergistic attraction of Aedes aegypti to binary blends of L-lactic acid and acetone, dichloromethane, or dimethyl di sulfide. J. Med. Entomol. 40(5):653-656. Bidlingmayer, W. L. 1994. How mosquitoes se e traps: role of visual responses. J. Am. Mosq. Control Assoc. 10(2):272-279. Bidlingmayer, W. L. and D. G. Hem. 1980. Th e range of visual attr action and the effect of competitive visual attractants upon mosquito (Diptera: Culicidae) flight. Bull. Entomol. Res. 70(2):321-342. Black, R. J. 1993. Florida climate data. Circ ular EES-5. University of Florida, Florida Cooperative Extension Service. 4 pp Last retrieved Feb. 28, 2005 from http://edis.ifas.ufl.edu/pdffiles/EH/EH10500.pdf
207 Boike, A. H. Jr., C. B. Rathburn, Jr., T. G. Floore, H. M. Rodriquez and J. S. Coughlin. 1989. Insecticide tolerance of Culex nigripalpus in Florida. J. Am. Mosq. Control Assoc. 5(4):522-528. Bonnet, D. D. and D. J. Worcester. 1946. The dispersal of Aedes albopictus in the territory of Hawaii. Am. J. Trop. Med. 26:465-476. Breaud, T. P. 1993. Insecticide resistance in Florida mosquitoes: a review of published literature. J. Fla. Anti-Mosq. Assoc. 64(1):14-21. Breeland, S. G. 1972. Studies on the diurnal resting habits of Anopheles albimanus and A. pseudopunctipennis in El Salvador. Mosq. News. 32:99-106. Brett, G. A. 1938. On the relative attractiveness to Aedes aegypti of certain coloured cloths. Trans. R. Soc. Trop. Med. Hyg. 32:113-124. Breyev, K. A. 1963. The effect of various light sources on the numbers and species of blood-sucking mosquitoes (Dip tera: Culicidae) collected in light traps. Entomol. Rev. 42:155-168. Brogdon, W. G. 1994. Measurement of fli ght tone differences between female Aedes aegypti and A. albopictus (Diptera: Culicidae). J. Med. Entomol. 31:700-703. Brogdon, W. G. and J. C. McAllister. 1998a. In secticide resistance and vector control. Emer. Infect. Dis. 4(4):605-613. Brogdon, W. G. and J. C. McAllister. 1998b. Si mplification of adult mosquito bioassays through use of time-mortality determination in glass bottles. J. Am. Mosq. Control Assoc. 14:159-165. Brown, A. W. A. 1951. Studies of the responses of the female Aedes mosquito. Part IV. Field experiments on Canadian specie s. Bull. Entomol. Res. 42:575-582. Brown, A. W. A. 1953. Studies of the responses of the female Aedes mosquito. Part V. The role of visual factors. Bull. Entomol. Res. 44:567-574. Brown, A. W. A. 1954. Studies on the responses of the female Aedes mosquito. Part VI. The attractiveness of coloured cloths and Canadian species. Bull. Entomol. Res. 45:67-78. Brown, A. W. A. 1956. Factors which attract Aedes mosquitoes to humans. Proc. 10th Int. Congr. Entomol. 3:757-763. Brown, A. W. A. 1966. The attraction of mosquitoes to hosts. J.A.M.A. 196(3):249252. Brown, A. W. A. and A. G. Carmichael. 1961. Lysine and alanine as mosquito attractants. J. Econ. Entomol. 54:317-324.
208 Brown, A. W. A. and R. Pal. 1971. Ins ecticide resistance in arthropods. 2nd ed. Geneva, World Health Organization. Monogr. Ser. 38. 491 pp. Brown, A. W. A., D. S. Sarkar ia and R. P. Thompson. 1951. Studies on the responses of the female Aedes mosquito. Part I. The search for attractant vapours. Bull. Entomol. Res. 42:105-114. Browne, M. S. and G. F. Bennett. 1981. Re sponses of mosquitoes (Diptera: Culicidae) to visual stimuli. J. Med. Entomol. 18:505-521. Burkett, D. A. 1998. Light color attracti on and dietary sugar composition for several mosquito (Diptera: Culicidae) species found in north central Florida. Ph.D. dissertation. University of Florida, Gainesville. 141 pp. Burkett, D. A., J. F. Butler and D. L. Klin e. 1998. Field evaluation of colored lightemitting diodes as attractants for woodland mosquitoes and other Diptera in north central Florida. J. Am. Mos q. Control Assoc. 14(2):186-195. Burkett, D. A., R. Kelly, C. H. Porter and R.A. Wirtz. 2004. Comm ercial mosquito trap and gravid trap oviposition media evalua tion, Atlanta, Georgia. J. Am. Mosq. Control Assoc. 20(3):233-238. Burkett, D. A., W. Lee, K. Lee, H. Kim, H. I. Lee, J. Lee, E. Shin, R. A. Wirtz, H. Cho, D. M. Claborn, R. E. Coleman, W. Kim and T. A. Klein. 2002. Late season commercial mosquito trap and host seeking activity evaluation against mosquitoes in a malarious area of the Republic of Kor ea. Korean J. Parasitol. 40(1):45-54. Burkett, D. A., W. J. Lee, K. W. Lee, H. C. Ki m, H. I. Lee, J. S. Lee, E. H. Shin, R. A. Wirtz, H. W. Cho, D. M. Claborn, R. E. Coleman and T. A. Klein. 2001. Light, carbon dioxide, and octenol-baited mos quito trap and host-seeking activity evaluations for mosquitoes in a malarious area of the Republic of Korea. J. Am. Mosq. Control Assoc. 17(3):196-205. Butler, J.F., W.R. Hess, R.G. Endris and K. H. Holscher. 1984. In vitro feeding of Ornithodoros ticks for rearing and assessmen t of disease transmission. pp. 10751081. In: Acarology VI, Vol. 2. New York, Wiley & Sons. Butler, J. F. and I. Katz. 1987. Process for determination of repell ency and attractancy. International Flavors & Fragrances Inc., NY a nd University of Florida, Gainesville. USPN: 4,759,228. Appl. No. 114,424. Butler, J. F. and J. S. Okine. 1995. Developm ent of attractants and repellents for future management of pest fly populations. pp. 117-126. In: Proceedings of Nuisance Concerns in Animal Manure Management: Odors and Flies. Gainesville, Florida. Calisher, C. H., M. Nuti, J. S. Lazuick, J. D. M. Ferrari and K. D. Kappus. 1981. Dengue in the Seychelles. Bull. World Health Organ. 59:619-622.
209 Campbell, C. B. 2003. Evaluation of five mosquito traps and a horse for West Nile vectors on a north Florida equi ne facility. M.S. Thesis. University of Florida, Gainesville. 37 pp. Carpenter, S. J. and W. J. LaCasse. 1955. Mosquitoes of North America (north of Mexico). Los Angeles, Universi ty of California Press. 360 pp. Carson, R. 1962. Silent Spring. Boston, Houghton Mifflin. 368 pp. Christophers, S. R. 1960. Aedes aegypti (L.), The Yellow Fever Mosquito: Its Life History, Bionomics and Structure. L ondon, Cambridge University Press. 739 pp. Cilek, J. E., G. D. Moorer, L. A. Del ph and F. W. Knapp. 1989. The Asian tiger mosquito, Aedes albopictus, in Kentucky. J. Am Mosq. Control Assoc. 5(2):267268. Clements, A. N. 1992. The Biology of Mos quitoes. Vol. 1. Development, Nutrition and Reproduction. London, Chapman & Hall. 509 pp. Clements, A. N. 1999. The Biology of Mosquitoes. Vol. 2. Sensory Reception and Behaviour. New York, Co mmonwealth Agricultural Bureau International Publishing. 740 pp. Darsie, R. F. Jr. and C. D. Morris. 2000. Keys to the Adult Females and Fourth Instar Larvae of the Mosquitoes of Florida (Diptera, Culicidae). Florida Mosquito Control Assoc., Inc., Ft. Meyers, FL. 159 pp. Davis, E. E. and P. G. Sokolove. 1976. Lac tic acid-sensitive receptors on the antennae of the mosquito, Aedes aegypti. J. Comp. Physiol. 105:43-56. Dennett, J. A., N. Y. Vessey and R. E. Pars ons. 2004. A comparison of seven traps used for collection of Aedes albopictus and Aedes aegypti originating from a large tire repository in Harris County (Houston), Te xas. J. Am. Mosq. Control Assoc. 20(4):342-349. Deonier, C. C. and I. H. Gilbert. 1950. Re sistance of salt-marsh mosquitoes to DDT and other insecticides. Mosq. News 10:138-143. Dodd, C. S. and N. R. H. Burgess. 1995. Why do insects bite? A review of blood sucking behaviour. J. R. Army Med. Corps 141:151-156. Eads, R. B. 1972. Recovery of Aedes albopictus from used tires shipped to United States ports. Mosq. News 32:113-114. Enayati, A. A., H. Ranson and J. Hemingwa y. 2005. Insect glutathion transferases and insecticide resistance. In sect. Mol. Biol. 14(1):3-8.
210 Estrada-Franco, J. G. and G. B. Craig, Jr. 1995. Biology, disease relationships, and control of Aedes albopictus. WHO. Panam. Health Org. Tech. Paper No. 42. Washington DC. 49 pp. Farkas, S. R. and H. H. Shorey. 1972. Ch emical trail-following by flying insects: a mechanism for orientation to a di stant odour source. Science 178:67-68. Fay, R. W. 1968. A trap based on visual re sponses of adult mosquitoes. Mosq. News 28(1):1-7. Fay, R. W. and W. H. Prince. 1970. A modified visual trap for Aedes aegypti. Mosq. News 30:20-23. Fehrenbach, P. 1990. Lawsuits: tips from an expert. Pest Control 58(6):36,38. Finney, D. J. 1971. Probit Analysis, 3rd ed. London, Cambridge University Press. 333 pp. Florida Coordinating Council on Mosquito Contro l. 1998. Florida mosquito control: the state of the mission as defined by mos quito controllers, regulators, and environmental managers. University of Florida. 207 pp. Francy, D. B., C. G. Moore and D. A. E liason. 1990. Past, present and future of Aedes albopictus in the United States. J. Am. Mosq. Control Assoc. 6:127-132. Freier, J. E. and D. B. Francy. 1991. A dupl ex cone trap for the collection of adult Aedes albopictus. J. Am. Mosq. Control Assoc. 7(1):73-79. Gerhardt, R. R., K. L. Gottfried, C. S. Appe rson, B. C. Davis, P. C. Erwin, A. B. Smith, N. A. Panella, E. E. Powell and R. S. Nasc i. 2001. First isolation of La Crosse Virus from naturally infected Aedes albopictus. Emerg. Infect. Dis. 7:807-811. Gilbert, I. H. and H. K. Gouck. 1957. Influence of surface color on mosquito landing rates. J. Econ. Entomol. 50:678-680. Gilbertson, W. E. 1945. Sanitary aspects of the control of th e 1943-1944 epidemic of dengue fever in Honolulu. Am. J. Pub. Health 35:261-270. Gillett, J. D. 1972. The Mosquito: Its Life, Activities and Impact on Human Affairs. New York, Doubleday. 358 pp. Gillies, M. T. 1980. The role of carbon dioxi de in host-finding by mosquitoes (Diptera: Culicidae): a review. Bull. Entomol. Res. 70:525-532. Gillies, M. T. and T. J. Wilkes. 1974. The ra nge of attraction of birds as bait for some West African mosquitoes. Bull. Entomol. Res. 63(4):573-581.
211 Gjullin, C. M. 1947. Effect of clothi ng color on the rate of attack of Aedes mosquitoes. J. Econ. Entomol. 40:326-327. Gjullin, C. M. and L. W. Isaak. 1957. Pr esent status of mosquito resistance to insecticides in the San Joaquin Valley in California. Mosq. News 17:67-70. Goodwin, M. H. 1942. Studies on artificial resting places of Anopheles quadrimaculatus Say. J. Nat. Malar. Soc. 1:93-99. Gubler, D. J. 1970. Comparisons of reproductive potentials of Aedes (Stegomyia) albopictus Skuse and Aedes (Stegomyia) polynesiensis Marks. Mosq. News 30:201-209. Gubler, D. J. and G. Kuno. 1997. Dengue and dengue hemorrhagic fever: its history and resurgence as a global public health problem. pp. 1-22. In: Dengue and Dengue Hemorrhagic Fever. New York, Commonwealth Agricultural Bureau International. Hall, D. R., P. S. Beevor, A. Cork, B. F. Nesbitt and G. A. Vale. 1984. 1-octen-3-ol: a potent olfactory stimulant and attractant for tsetse isolated from cattle odours. Insect Sci. Appl. 5:335-339. Harrington, L., J. D. Edman and T. W. Scott. 2001. Why do female Aedes aegypti (Diptera: Culicidae) feed pr eferentially and frequently on human blood? J. Med. Entomol. 38(3):411-422. Harrison, B. A., C. J. Mitchell, C. S. Appers on, G. C. Smith, N. Karabatsos, B. R. Engber and N. H. Newton. 1995. Isolation of Potosi virus from Aedes albopictus in North Carolina. J. Am. Mosq. Control Assoc. 11(2):225-229. Harwood, R. F. and M. T. James. 1979. Entomology in Human and Animal Health. New York, Macmillan. 548 pp. Haufe, W. O. 1964. Visual attraction as a pr inciple in design of mosquito traps. Can. Entomol. 96:118. Haufe, W. O. and L. Burgess. 1960. Design and efficiency of mosquito traps based on visual response to patterns. Can. Entomol. 92:124-140. Hawley, W. A. 1988. The biology of Aedes albopictus. J. Am. Mosq. Control Assoc. 4(Suppl. 1):1-40. Headlee, T. J. 1932. The development of mechanical equipment for sampling the mosquito fauna and some results of its use. Proceedings of the Annual Meeting of the N.J. Mosquito Extermin ation Association 19:106-126. Headlee, T. J. 1937. Some facts underlying th e attraction of mosquitoes to sources of radiant energy. J. Econ. Entomol. 30(2):309-312.
212 Herbert, E. W., R. P. Meyer and P. G. Turbes. 1972. A comparison of mosquito catches with CDC light traps and CO2 baited traps in the Republic of Vietnam. Mosq. News 32:212-214. Ho, B. C., Y. C. Chan and K. L. Chan. 1973. Field and laboratory observations on landing and biting periodicities of Aedes albopictus (Skuse). Southeast Asian J. Trop. Med. Public Health 4(2):238-244. Hodgson, E. and P. E. Levi. 2001. Metabolism of pesticides. pp. 531-562. In: Handbook of Pesticide Toxicology Princi ples. London, Academic Press. Holick, J., A. Kyle, W. Ferraro, R. R. Dela ney and M. Iwaseczko. 2002. Discovery of Aedes albopictus infected with West Nile virus in southeastern Pennsylvania. J. Am. Mosq. Control Assoc. 18(2):131. Hornby, J. A. and T. W. Miller, Jr. 1994. Aedes albopictus distribution, abundance and colonization in Lee County, Florida and its effect on Aedes aegypti-two additional seasons. J. Fla. Anti-Mosq. Assoc. 65:21-27. Howlett, F. M. 1910. The influence of temperature upon the biting of mosquitoes. Parasitology 3:479-484. Ikeshoji, T. 1986. Distribution of the mosquitoes, Culex tritaeniorhynchus, in relation to disposition of sound traps in a paddy fiel d. Jap. J. Sanit. Zool. 37(2):153-159. Ikeshoji, T. 1993. Attractive physical and ch emical stimuli for mosquito control. Sumitomo Pyrethroid World 20:3-8. Ikeshoji, T. and K. Ogawa. 1988. Field catc hing of mosquitoes with various types of sound traps. Jpn. J. Sanit. Zool. 39(2):119-123. Ikeshoji, T., M. Sakakihara and W. K. Reisen. 1985. Removal sampling of male mosquitoes from field populations by sound-trapping. Jpn. J. Sanit. Zool. 36(3):197-203. Ikeshoji, T. and H. H. Yap. 1990. Impact of the insecticide-trea ted sound traps on an Aedes albopictus population. Jpn. J. San it. Zool. 41(3):213-217. Jensen, T. O., R. Willis, T. Fukuda and D. R. Barnard. 1994. Comparison of Bidirectional Fay, omni-directional, CDC, and duplex cone traps for sampling adult Aedes albopictus in north Florida. J. Am. Mosq. Control Assoc. 10(1):74-78. Kahn, M. C., W. Celestin and W. Offenhaus er. 1945. Recording of sounds produced by certain disease-carrying mos quitoes. Science 101:335-336. Kambhampati, S., W. C. Black and K. S. Rhi. 1991. Geographic origin of US and Brazilian Aedes albopictus inferred from allozyme analysis. Heredity 67:85-94.
213 Kanda, T., W. H. Cheong, K. P. Loong, T. W. Lim, K. Ogawa, G. L. Chiang and S. Sucharit. 1987. Collection of male mo squitoes from field populations by sound trapping. Tropical Biomedicine 4:161-166. Kanda, T., V. Kerdpibule, T. Deesin, S. Thongrungkiat, S. Leemingsawat and G. L. Chiang. 1990. Strategies for mosquito cont rol by using a sound trap system and an insect growth regulator (pyriproxyfen)a review. Tropical Biomedicine 7:159-174. Khan, A. A., H. I. Maibach, W. G. Strauss and W. R. Fenley. 1966. Quantification of effect of several stim uli on the approach of Aedes aegypti. J. Econ. Entomol. 59(4):690-694. Khoo, B. K., D. J. Sutherland, D. Sprenger, D. Dickerson and H. Nguyen. 1988. Susceptibility status of Aedes albopictus to three topically app lied adulticides. J. Am. Mosq. Control Assoc. 4(3):310-313. Kline, D. L. 1994a. Introduction to sympos ium on attractants for mosquito surveillance and control. J. Am. Mos q. Control Assoc. 10(2):253-257. Kline, D. L. 1994b. Olfactory attractants fo r mosquito surveillance and control: 1-octen3-ol. J. Am Mosq. Cont rol Assoc. 10(2):280-287. Kline, D. L. 1999. Comparison of two American Biophysics mosquito traps: the professional and a new counterflow geometry trap. J. Am. Mosq. Control Assoc. 15(3):276-282. Kline, D. L. 2002. Evaluation of various mode ls of propane-powered mosquito traps. J. Vector Ecol. 27(1):1-7. Kline, D. L., D. A. Dame and M. V. Meis ch. 1991a. Evaluation of 1-octenol-3-ol and carbon dioxide as attractants for mosquitoes associated with irri gated rice fields in Arkansas. J. Am. Mosq. Control Assoc. 7(2):165-169. Kline, D. L. and G. F. Lemire. 1995. Field evaluation of heat as an added attractant to traps baited with carbon dioxide and octenol for Aedes taeniorhynchus. J. Am. Mosq. Control Assoc. 11(4):454-456. Kline, D. L. and M. O. Mann. 1998. Evaluation of butanone, carbon dioxide, and 1octen-3-ol as attractants for mosquitoes associated with north central Florida bay and cypress swamps. J. Am. Mosq. Control Assoc. 14(3):289-297. Kline, D. L., W. Takken, J. R. Wood and D. A. Carlson. 1990. Field studies on the potential of butanone, carbon dioxide, honey extract, 1-octe n-3-ol, L-lactic acid and phenols as attractants for mosquitoes. Med. Vet. Entomol. 4:383-391. Kline, D. L., W. Takken, J. R. Wood and J. A. Cornell. 1991b. Interactive effects of 1octen-3-ol and carbon dioxide on mosquito (Diptera: Culicidae) surveillance and control. J. Med. Entomol. 28:254-258.
214 Kloter, K. O., J. R. Kaltenbach, G. T. Carmichael and D. D. Bowman. 1983. An experimental evaluation of six different suction traps for attracting and capturing Aedes aegypti. Mosq. News 43(3):297-301. Kusakabe, Y. and T. Ikeshoji. 1990. Compar ative attractancy of physical and chemical stimuli to aedine mosquitoes. Jpn. J. Sanit. Zool. 41(3):219-225. Laarman, J. J. 1955. The host-seeking behaviour of the malaria mosquito Anopheles maculipennis atroparvus. Acta Leiden 25:1-144. Lamche, G. D. and P. I. Whelan. 2003. Variab ility of larval identi fication characters of exotic Aedes albopictus (Skuse) intercepted in Da rwin, Northern Territory. Commun. Dis. Intel. 27:105-109. Lang, J. T. 1984. Intermittent light as a mosquito attractant in New Jersey light traps. Mosq. News 44(2):217-220. Lehane, M. J. 1991. Biology of Blood-Suck ing Insects. London, Chapman & Hall. 288 pp. Linley, J. R. 1989. Comparative fine structure of the eggs of Aedes albopictus, Ae. aegypti, and Ae. bahamensis (Diptera: Culicidae). J. Med. Entomol. 26(6):510521. Linthicum, K. J., V. L. Kramer, M. B. Madon and K. Fujioka. 2003. Introduction and potential establishment of Aedes albopictus in California in 2001. J. Am. Mosq. Control Assoc. 19(4):301-308. Liu, H., E. W. Cupp, A. Guo and N. Liu. 2004. Insecticide resistance in Alabama and Florida mosquito strains of Aedes albopictus. J. Med. Entomol. 41(5):946-952. Madhukar, B. V. R. and M. K. K. Pilla i. 1970. Development of organophosphate resistance in Indian strains of Aedes aegypti (L). Bull. World Health. Organ. 43:735-742. Mahler, H. R. and E. H. Cordes 1971. Biological Chemistry. 2nd ed. New York, Harper & Row. 1009 pp. Martin, A. B., C. B. Warren and J. F. Butler. 1991. Method for repelling Aedes aegypti using 3,7 dimethyl-6-octenenitrile a nd/or 2(3,3-dimethyl-2-norbornylidene) ethanol-1. International Flavors and Fragranc es Inc. NY, and University of Florida, Gainesville. USPN: 5,134,892. Appl. No. 157,403. Mazokhin-Porshnyakov, G. A. 1969. Insect vision. New York, Ple num Press. 306 pp. Mazzarri, M. B. and G. P. Georghiou. 1995. Characterization of resistance to organophosphate, carbamate, and pyrethroid insecticides in fi eld populations of Aedes aegypti from Venezuela. J. Am. Mosq. Control Assoc. 11(3):315-322.
215 Mboera, L. E. G. and W. Takken. 1997. Ca rbon dioxide chemotropism in mosquitoes (Diptera: Culicidae) and its potential in vector surveillance and management programmes. Rev. Med. Vet. Entomol. 85:355-368. Mboera, L. E. G., W. Takken and E. Z. Sambu. 2000. The response of Culex quinquefasciatus (Diptera: Culicidae) to traps baited with carbon dioxide, 1-octen3-ol, acetone, butyric acid and human foot odour in Tanzania. Bull. Entomol. Res. 90:155-159. McDonald, P. T., S. M. Asman, M. M. Mil by, J. Bruen and R. Ainsley. 1978. Outdoor cage tests of genetic strains of Culex tarsalis for future field releases. Proc. 46th Annu. Conf. Calif. Mosq. C ontrol. Assoc.:105-109. McNelly, J. R. 1995. An introduction to the ABC trap. Proc. N.J. Mosq. Control Assoc.:47-52. Miller, T. A., R. G. Stryker, R. N. Wilkerson and S. Esah. 1969. Notes on the use of CO2 baited CDC miniature light traps for mosquito surveillance in Thailand. Mosq. News 29(4):688-689. Minno, M. C., J. F. Butler and D. W. Hall. 2005. Florida Butterfly Caterpillars. Gainesville, University of Florida Press. 341 pp. Mitchell, C. J. 1995. The role of Aedes albopictus as an arbovirus vector. Parassitologia 37:109-113. Mitchell, C. J. and D. J. Gubler. 1987. Vector competence of geographic strains of Aedes albopictus and Aedes polynesiensis and certain other Aedes (Stegomyia) mosquitoes for Ross River virus. J. Am. Mosq. Control Assoc. 3(2):142-147. Mitchell, C. J., L. D. Karabatsos, N. Smit h, G. C. Smith and V. J. Starwalt. 1998. Isolation of La Crosse, Cache Valley, and Potosi virus from Aedes mosquitoes (Diptera: Culicidae) collected at used-t ire sites in Illinois during 1994-1995. J. Med. Entomol. 35(4): 573-577. Mitchell, C. J., M. L. Niebylski, G. C. Smit h, N. Karabatsos, D. Martin, J. P. Mutebi, G. B. Craig, Jr. and M. J. Mahler. 1992. Isolation of Eastern equine encephalitis virus from Aedes albopictus in Florida. Science (257):526-527. Moore, A., J. R. Miller, B. E. Tabashnik a nd S. H. Gage. 1986. Automated identification of flying insects by analysis of wingbeat frequencies. J. Econ. Entomol. 79:17031706. Moore, C. G. 1999. Aedes albopictus in the United States: curre nt status and prospects for further spread. J. Am. Mosq. Control Assoc. 15(2):221-227. Moore, C. G. and C. J. Mitchell. 1997. Aedes albopictus in the United States: ten-year presence and public health implications Emerg. Infect. Dis. 3(3):329-334.
216 Mouchet, J. 1972. La resi stance aux insecticides des Aedes dans les regions dAsie du sud-est et du Pacifique. Cah. Orstom., Ser. Entomol. Med. Parasitol. 10(4):301308. Mouchet, J. 1988. Agriculture and vector re sistance. Insect Sci. Appl. 9(3):297-302. Muir, L. E., M. J. Throne and B. H. Kay. 1992. Aedes aegypti (Diptera: Culicidae) vision: spectral sensitivity a nd other perceptual parameters of the female eye. J. Med. Entomol. 29:278-281. Murphy, M. W., R. F. Dunton, M. J. Perich and W. A. Rowley. 2001. Attraction of Anopheles (Diptera: Culicidae) to volatile chemicals in western Kenya. J. Med. Entomol. 38(2):242-244. Nayar, J. K., N. Karabatsos, J. W. Knight, M. Godsey, J. Chang and C. J. Mitchell. 2001. Mosquito hosts of arboviruses from Indi an River County, Florida, during 1998. The Fla. Entomol. 84(3):376-379. Neng, W., X. Yan, H. Fuming and C. Dazong. 1992. Susceptibility of Aedes albopictus from China to insecticides, and mechan ism of DDT resistance. J. Am. Mosq. Control Assoc. 8(4):394-397. Niebylski, M. L., H. M. Savage, R. S. Na sci and G. B. Craig, Jr 1994. Blood hosts of Aedes albopictus in the United States. J. Am Mosq. Control Assoc. 10(3):447450. OMeara, G. F., L. F. Evans, Jr., A. D. Gettman and J. P. Cuda. 1995a. Spread of Aedes albopictus and decline of Ae. aegypti (Diptera: Culicidae) in Florida. J. Med. Entomol. 32:554-562. OMeara, G. F., L. F. Evans, Jr., A. D. Gettman and A. W. Patterson. 1995b. Exotic tank bromeliads harboring immature Aedes albopictus and Aedes bahamensis (Diptera: Culicidae) in Florida. J. Vector Ecol. 20(2):216-224. OMeara, G. F., A. D. Gettman, L. F. Evans and G. A. Curtis. 1993. The spread of Aedes albopictus in Florida. Am. Entomol. 39(3):163-172. Peacock, B. E., J. P. Smith, P. G. Gregory, T. M. Loyless, J. A. Mulrennan, Jr., P. R. Simmonds, L. Padgett, Jr., E. K. Cook and T. R. Eddins. 1988. Aedes albopictus in Florida. J. Am. Mosq. Control Assoc. 4:362-365. Perkins, R. C. L. 1913. Fauna Hawaiiensis Cambridge University Press. Vol. 1, clxxxi. Petersen, J. L., T. G. Floore and W. G. Brogdon. 2004. Diagnostic dose of synergized dphenothrin for insecticide susceptibility testing by bottle bioassay. J. Am. Mosq. Control Assoc. 20(2):183-188.
217 Peterson, D. G. and A. W. A. Brown. 1951. Studies of the responses of the female Aedes mosquito. Part III. The response of Aedes aegypti (L.) to a warm body and its radiation. Bull. Entomol. Res. 42:535-541. Plapp, F. W. 1971. Insecticide resistance in Heliothis: tolerance in larvae of H. virescens as compared with H. zea to organophosphate insecticides. J. Econ. Entomol. 64:999-1002. Plapp, F. W. and T. C. Wang. 1983. Genetic origins of insecticide resistance. pp. 47-70. In: Pest Resistance to Pesticides. New York, Plenum Press. Pratt, J. J., R. H. Hetrick, J. B. Harrison and L. Haber. 1946. Tires as a factor in the transportation of mosquitoes by ships. Milit. Surg. 99:785-788. Qiu, F., H. Zhang, L. Shao, X. Li, H. Luo and Y. Yu. 1981. Studies on the rapid detection of dengue virus antigen by immunofluorescen ce and radioimmuno-assay. Chinese Med. J. 94:653-658. Rawlings, J., C. Burgess, L. Tabony, R. Chapman, K. Hendricks and G. Stevenson. 1996. Dengue fever at the U.S.-M exico border, 1995-1996. MMWR Morb. Mortal. Wkly. Rep. 45(39):841-844. Reeves, W. C. and W. M. Hammon. 1942. Mo squitoes and encephalitis in the Yakima Valley, Washington. IV. A trap for collecti ng live mosquitoes. J. Infect. Dis. 70:275-277. Reiter, P. 1998. Aedes albopictus and the world trade in us ed tires, 1988-1995: the shape of things to come? J. Am. Mosq. Control Assoc. 14(1):83-94. Reiter, P. and R. F. Darsie. 1984. Aedes albopictus in Memphis, Tennessee (USA): an achievement of modern transportation? Mosq. News 44:396-399. Robert, L. L. and J. K. Olson. 1989. Susceptibility of female Aedes albopictus from Texas to commonly used adulticides. J. Am. Mosq. Control Assoc. 5(2):251-253. Rodhain, F. and L. Rosen. 1997. Mosquito vectors and dengue virus-vector relationships. pp. 45-60. In: Dengue and Dengue Hemorrhagic Fever. New York, Commonwealth Agricultural Bureau International. Roessler, P. and A. W. A. Brown. 1964. Studies on the responses of the female Aedes mosquito. X. Comparison of oestrogens and amino-acids as attractants. Bull. Entomol. Res. 55:395-403. Rose, R. I. 2001. Pesticides and public health: integrated methods of mosquito management. Emerg. Infect. Dis. 7(1):17-23.
218 Rosen, L., L. E. Roseboom, D. J. Gubler, J. C. Lien and B. N. Chaniotis. 1985. Comparative susceptibility of mosquito sp ecies and strains to oral and parenteral infection with dengue and Japanese ence phalitis viruses. Am. J. Trop. Med. Hyg. 34(3):603-615. Rosen, L., D. A. Shroyer, R. B. Tesh, J. E. Freier and J. C. Lien. 1983. Transovarial transmission of dengue viruses by mosquitoes: Aedes albopictus and Aedes aegypti. Am. J. Trop. Med. Hyg. 32(5):1108-1119. Rudolfs, W. 1922. Chemotropism of mosquito es. Bull. N.J. Agric. Exp. Station, No. 367. 23 pp. Russo, R. 1978. Substrate texture as an oviposition stimulus for Aedes vexans (Diptera: Culicidae). J. Med. Entomol. 15(1):17-20. Rutledge, C. R., J. F. Day, C. C. Lord, L. M. Stark and W. J. Tabachnick. 2003. West Nile virus infection rates in Culex nigripalpus do not reflect transmission rates in Florida. J. Med. Entomol. 40(3):253-258. Sabin, A. B. 1952. Research on dengue dur ing World War II. Am. J. Trop. Med. Hyg. 1:30-50. Sames, W. J., IV, R. Bueno, Jr., J. Hayes and J. K. Olson. 1996. Insecticide susceptibility of Aedes aegypti and Aedes albopictus in the lower Rio Grande Valley of Texas and Mexico. J. Am. Mosq. Control Assoc. 12(3):487-490. SAS Institute. 2001. SAS/STAT users manual, ver. 8.2. SAS Institute, Cary, N.C. Savage, H. M., M. L. Niebylski, G. C. Smith, C. J. Mitchell and G. B. Craig, Jr. 1993. Host-feeding patterns of Aedes albopictus (Diptera: Culicidae) at a temperate North American site. J. Med. Entomol. 30(1):27-33. Schoeler, G. B., S. S. Schleich, S. A. Manw eiler and V. L. Sifuentes. 2004. Evaluation of surveillance devices for monitoring Aedes aegypti in an urban area of northeastern Peru. J. Am. Mos q. Control. Assoc. 20(1):6-11. Service, M. W. 1993. Mosquito Ecology: Field Sampling Methods. London, Elsevier Applied Science. 988 pp. Shirai, Y., H. Funada, K. Kamirua, T. Se ki and M. Morohashi. 2002. Landing sites on the human body preferred by Aedes albopictus. J. Am. Mosq. Control Assoc. 18(2):97-99. Shirai, Y., K. Kamimura, T. Seki and M. Mo rohashi. 2001. L-lactic acid as a mosquito (Diptera: Culicidae) repellent on human and mouse skin. J. Med. Entomol. 38(1):51-54.
219 Shone, S. M., P. N. Ferrao, C. R. Lesser, G. E. Glass and D. E. Norris. 2003. Evaluation of carbon dioxideand 1-octen-3-ol-baited centers for disease control Fay-Prince traps to collect Aedes albopictus. J. Am. Mosq. Control Assoc. 19(4):445-447. Shroyer, D. A. 1990. Vertical maintenance of dengue-1 virus in sequential generations of Aedes albopictus. J. Am. Mosq. Control Assoc. 6(2):312-314. Sippell, W. L. and A. W. A. Brown. 1953. Studies on the responses of the female Aedes aegypti. Part V. The role of visual fact ors. Bull. Entomol. Res. 43:567-574. Skuse, F. A. A. 1896. The banded mosquito of Bengal. Indian Museum Notes Vol. 3, (5):20. Smith, J. and E. Cope. 2004. Efficacy of the ABC Mosquito Magnet Liberty trap for control in a residential environment. Flor ida A&M University. Last retrieved Dec. 20, 2004 from http://pherec.org/DECS/Trapping%20Systems/ Smith, J., J. Walsh and R. Huss. 2002. Mosquito-capturing prowess of several commercial traps. Florida A&M Univers ity. Last retrieved Dec. 20, 2004 from http://pherec.org/DECS/Trapping%20Systems/ Smith, J. P. and J. D. Walsh. 2003. Evalua tion of seven commercial mosquito control systems. Florida A&M University. Last retrieved Dec. 20, 2004 from http://pherec.org/DECS/Trapping%20Systems/ Smith, W. C. and J. F. Butler. 1991. Ultr astructure of the Ta banidae compound eye: unusual features for Diptera. J. Insect Physiol. 37:287-296. Snow, F. W. 1971. The spectral sensitivity of Aedes aegypti (L.) at oviposition. Bull. Entomol. Res. 60:863-696. Sprenger, D. and T. Wuith iranyagool. 1986. The disc overy and distribution of Aedes albopictus in Harris County, Texas. J. Am Mosq. Control Assoc. 2:217-219. Stasiak, R. S., R. H. Grothaus and W. F. Miner. 1970. Susceptibility of Aedes albopictus (Skuse) larvae from South Vietnam to fi ve insecticides in 1969. Mosq. News 30(2):246-249. Stryker, R. G. and W. W. Young. 1970. Effe ctiveness of carbon diox ide and L(+) lactic acid in mosquito light traps with and w ithout light. Mosq. News 30(3):388-393. Sudia, W. D. and R. W. Chamberlain. 1962. Battery-operated light trap, an improved model. Mosq. News 22(2):126-129. Sutcliffe, J. F. 1987. Distance orientation of bi ting flies to their hosts Insect Sci. Appl. 8:611-616.
220 Sutton, O. G. 1953. Micrometeorology; a st udy of physical processes in the lowest layers of the earth's atmosphere. New York, McGraw-Hill, 1953. 333 pp. Takken, W. and D. L. Kline. 1989. Carbon dioxide and 1-octen-3-ol as mosquito attractants. J. Am. Mosq. Control Assoc. 5(3):311-316. Thirapatsakun, L., P. Tauthong and B. Phan thumachinda. 1981. Surface preferences for oviposition of Aedes aegypti in Thailand. Southeast Asian J. Trop. Med. Public Health 12(2):209-212. Thurman, D. C., Jr. and E. B. Thurman. 1955. Report of the initial operation of a light trap in northern Thailand Mosq. News 15:218-224. Toma, T., I. Miyagi, T. Chinen and H. Hat azoe. 1992. Insecticidal susceptibilities of Aedes albopictus larvae in different islands of Ok inawa Prefecture, Japan. Jpn. J. Sanit. Zool. 43(4):331-336. Turell, M. J., C. L. Bailey and C. L. Beam an, Jr. 1988. Vector competence of a Houston, Texas Strain of Aedes albopictus for Rift Valley fever virus. J. Am. Mosq. Control Assoc. 4(1):94-96. Turell, M. J. and J. R. Beaman. 1992. Expe rimental transmission of Venezuelan equine encephalomyelitis virus by a strain of Aedes albopictus (Diptera: Culicidae) from New Orleans, Louisiana. J. Med. Entomol. 29(5):802-805. United States Army Environmental Hygiene Agency. 1970. Field evaluation of trap, insect survey, CO2 April-September 1970. USAEHA Edgewood Arsenal, MD Study No. 31-005-70/71. 12 pp. Usinger, R. 1944. Entomological phases of the recent dengue epidemic in Honolulu. Public Health Rep. 59:423-430. Vale, G. A. 1993. Development of baits fo r tsetse flies (Dipte ra: Glossinidae) in Zimbabwe. J. Med. Entomol. 30:831-842. Vale, G. A. and D. R. Hall. 1985. The role of 1-octen-3-ol, acet one and carbon dioxide in the attraction of tsetse flies, Glossina spp. (Diptera: Glossinidae), to ox odour. Bull. Entomol. Res. 75:209-217. Valles, S. M., P. G. Koehler and R. J. Bre nner. 1997. Antagonism of fipronil toxicity by piperonyl butoxide and S,S,S-tribut yl phosphorotrithioat e in the German Cockroach (Dictyoptera: Blattellidae ). J. Econ. Entomol. 90:1254-1258. Van Essen, P. H. A., J. A. Kemme, S. A. Ritchie and B. H. Kay. 1994. Differential responses of Aedes and Culex mosquitoes to octenol or light in combination with carbon dioxide in Queensland, Australia Med. Vet. Entomol. (8):63-67.
221 Vavra, R. W., Jr., R. R. Carestia, R. L. Frommer and E. J. Gerberg. 1974. Field evaluation of alternative light sources as mosquito attractants in the Panama Canal zone. Mosq. News 34(4):382-384. Vythilingam, I., C. G. Lain and C. S. Thim. 1992. Evaluation of carbon dioxide and 1octen-3-ol as mosquito attractants. S outheast Asian J. Trop. Med. Public Heath 23(2):328-331. Ware, G. W. 1983. Pesticides Theory and Application. New York, W. H. Freeman and Company. 308 pp. Watson, M. S. 1967. Aedes (Stegomyia) albopictus (Skuse): A literature review. Dep. Army, Ft. Detrick, MD. Misc. Publ. 22:1-38. Wesson, D. M. 1990. Susceptibility to organophosphate insec ticides in larval Aedes albopictus. J. Am. Mosq. Control Assoc. 6(2):258-264. Wilkinson, C. F. 1980. The metabolism of xenobiotics: a study in biochemical evolution. pp 251-268. In: The Scientific Basis of Toxicity Assessment. New York, Elsevier/North-Holland Biomedical Press. Wilkinson, C. F. 1983. Role of mixed-func tion oxidases in insect icide resistance. pp. 175-205. In: Pest Resistance to Pesticides New York, Plenum Press. Wilton, D. P. 1968. Oviposition site selection by the tree-hole mosquito, Aedes triseriatus (Say). J. Med. Ento mol. 5(2):189-194. Wilton, D. P. and K. O. Kloter. 1985. Pre liminary evaluation of a black cylinder suction trap for Aedes aegypti and Culex quinquefasciatus (Diptera: Culicidae). J. Med. Entomol. 22(1):113-114. Wirth, M. C. and G. P. Georghiou. 1999. Se lection and characterization of temephos resistance in a population of Aedes aegypti from Tortola, British Virgin Islands. J. Am. Mosq. Control Assoc. 15(3):315-320. Wood, P. W. and R. H. Wright. 1968. Some responses of flying Aedes aegypti to visual stimuli. Can. Entomol. 100(5):504-513. World Health Organization. 1957. Expert committee on insecticides: resistance of insects to insecticides. Se venth Report of the Expert Committee on Vector Biology and Control. World Health Organ. Tech. Rep. Ser. 125. 31 pp. World Health Organization. 1986. Resistance of vectors and reservoirs of disease to pesticides. Tenth Report of the Expert Committee on Vector Biology and Control. World Health Organ. Tech. Rep. Ser. 737. 87 pp. Wright, R. H. 1958. The olfactory guidance of flying insects. Can. Entomol. 90:81-89.
222 Yamanishi, H. 1999. Susceptibi lity of Chinese strains of Aedes albopictus and Aedes aegypti to Chikungunya virus. Med. Entomol. Zool. 50(1):61-64.
223 BIOGRAPHICAL SKETCH David Franklin Hoel was born on September 28, 1959 in Columbia, South Carolina. He is the first child of Frank and Patricia Hoel. He moved to Chelsea, Oklahoma and finished high school there in 1978. Shortly afte rwards, he enlisted in the U.S. Army for 3 years and served as an ammunition specialist. Having always been interested in biting insects, he attended Te xas A&M University and graduated with a B.S. in Entomology in 1986. After graduation, he wo rked for a pest control company in Ft. Lauderdale, Florida and joined the Army Rese rves serving as an Environmental Health Specialist. He went back to Texa s A&M and studied mosquito biology (Culex salinarius) under Dr. J. K. Olson and finished with his M.S. degree in May 1993. He worked as a salesman for B&G Chemicals & Equipment Company in Houston, Texas for 2 years before receiving a commission in th e U.S. Navys Medical Service Corps as a medical entomologist. His duty stations include the Navy Disease Vector Ecology Control Center, Alameda, CA; Second Medi cal Battalion, Camp Lejeune, NC; and Navy Environmental and Preventive Medicine Unit Six, Pearl Harbor, HI. His duties have included pest and vector control consulta tion; entomology training for military and DOD civilian personnel; vector c ontrol; USDA-mandated inspect ions of Navy ships and material to prevent the introduction of exotic pests; force health protection; on-site preventive medicine support dur ing contingencies, exercises and operations; and serving as medical detachment commander. He served in disaster relief mi ssions in Puerto Rico and Guatemala; provided malaria and dengue pr evention and control co nsultation in East
224 Timor; served as medical-detachment commander on an exercise in Norway; and provided preventive medicine services in Au stralia, Spain, Estonia, and Namibia. Routine inspections and training missions ha ve taken him to Japan, Okinawa, South Korea, Guam, and Diego Garcia. Lieutena nt Commander Hoel was selected for Duty Under Instruction by the Medi cal Service Corps and began school in August 2002. He anticipates assignment to Naval Medical Research Unit No. 3 (NAMRU-3), Cairo Egypt after graduation. He and his wife Joyce have two children, Michael and Caroline.