Doceunnn 1990 Mtcnocvctops rN Trnes 689

ELIMINATION OF AEDESALBOPIC?US FROM TIRE PILES BY INTRODUCINGMACROCYCLOPS ALBIDUS (COPEPODA, )

GERALD G. MARTEN1

Diuision of Vector-BorneInfectious Diseases,Centers for DiseaseControl, Fort CoIIins,CO 80522and New Orl.eansMosquito Control Board, Neu Orleans, LA 20126

ABSTRACT. The-copepo4_Mocrocyclopsalbidus is an unusually promising new form of biological corttrol for larvae. When introduced to two isolated tire pile;, M. atb-iduseliminated all iedes albopictuslarvae from both piles within 2 months. Adult Ae. albopiitus around the tire piles disappeared within another month. Completesuppression of Ae. albopicttn laivae was still in effect in all treated tires a year later.

INTRODUCTION The tire piles were 500 m from one another in a triangular anangement. They were The Centersfor DiseaseControl and the New more than 500 m from other Ae. albopictus Orleans Mosquito Control Board recently col- breeding arcas.Aedes albopictus populations at each pile laboratedin field trials to evaluatethe abiiitv of were isolated by a paucity of discarded 6 speciesof cyclopoidcopepods to control Aedes con- tainers and tree holes in the surrounding woods larvae in tires (Marten 1989). al- to serve as breeding sites.Aedes bidus (Jurine) was the most effective predator. albopirctusdoes not normally dispersemore than a Once establishedin tftes, Macrocyclopsalways few hundred meters(Hawley 1988). The very low populations killed all first instar larvae of Aedes albopicttu of Ae. albopictusin the surrounding (Skuse),Ae. aegypti (Linn.) and.Ae. triseriatus woods dur- ing (Say) (Marten 1990a). May-November 1989were verified by means of ovipositiontraps. A small percentageof discardedtires in New Each pile had approximately 15,000immature Orleans contain natural populations of Macro- Ae. albopirtus at the end of July, when cyclopsthat appear to have been in the tires for 10 adult female Macrocyclops albidus years (Marten 1989). Using Macrocyclops for were introduced into every tire in 2 of the piles (designated biological control of mosquito larvae is a matter "treated" piles). The Macrocyclaps of introducing Macrocyclop.sinto tires that do came from a laboratory colony initiated not already have them naturally. Macrocyclops 3 months earlier from a natural population in were introduced to experimental tire piles to a tire. No Macro- cyclopswere introduced to the third pile, assesstheir impact on both mosquito larvae in which servedas a control. the tires and the population of adult mosquitoes around the tires. Immature and adult mosquitoes were moni- tored at each tire pile from June through No- vember. A random sample of 10 tires MATERIALS AND METHODS at each pile was examined every 2 weeks for larvae and In June 1989,three tire piles infested with Ae. pupae.AII water and detritus were removedfrom albopictuswere establishedin a wooded area of each tire with a plastic cup and examined in a New Orleans(Big Oak Island), which was vir- large shallow glass dish with intense illumina- tually free of this mosquito previously. Each pile tion from the top, bottom and sides.Larvae and consistedof 100 naturally infested tires brought pupaewere removedfrom the sample,identified from another area. Only tires without a natuial and counted live with a stereomicroscope,and population of Macrocyclopsor other larvivorous returned with water and detritus to the tire, cyclopoids were used. The tires were stacked Standard errors of estimates of the number of against each other, all touching the ground and immature Ae. albopictu.sin each tire pile varied arranged so that the water inside could be re- from 17 to 43% of the estimate.Everv tire in moved for examination without moving the the Macrocyclops-treatedpiles was examinedon tires. No maintenance was provided. The only everysampling occasion during September-No- source of water for the tires was rainfall. Some vember,when there werefew or noA e. albopictus of the tires dried out from time to time. larvaein thosepiles; the standarderrors ofthose sampleswere zero. Macrocyclopsin the treated piles t were moni- Current address:New Orleans Mosquito Control tored by counting copepodidsand adults in the Board, 6601South Shore Harbor Drive, New Orleans, same samplesthat were taken for mosquito lar- LA70726. vae and pupae. Standard errors of estimates of JouRNnr,oF THE Atunmca.uMoseutro Coxrnor- AssochtroN VoL.6,No.4 the number of Macrocyclopsin each tire pile 15000 varied from 16to 30% ofthe estimate. a o The adult mosquito population was assessed o J at weekly intervals, using 2 different measures. o 10000 The first was landing rate on human bait. Two C) persons stood simultaneously at different parts o () of the sametire pile, using dry ice as an attrac- 5000 Macrocyclops-treatedPiles tant and collecting mosquitoeswith hand-held = aspirators.Each person collectedfor a total of 0 10 min in two 5-min sessionsat different parts ofthe pile. Landing rates were alwaysconducted in immediate successionat all piles between 0900and 1100h the sameday. The samepeople 30000 measuredlanding rates for the entire study.

Numerous speciesof adult mosquitoes were IJJ collected:Aedes atlantic&s Dyar and Knab, Ae. cL 2oooo f sollicitans (Walker), Ae. triseriatus, Ae. uexans a! (Meigen),Anopheles crucian s Wied., Coquillet- tidia perturba;asDyar, Culex nigripalpus Theo- UJ bald, Cr. salinaritn Coq. and Psorophora ferox E (von Humboldt). Only Ae. albopictus were J counted. Standard errors of Ae. albopictu,sIand- ing ratesranged from 11 to 55% ofthe estimate. The relatively high standard enors were asso- ciated with very low landing rates. The secondmeasure of the adult population was oviposition. The ovitraps were black plastic cups containing water in which rabbit food pel- z lets had previously been incubated. Eggs were I a collectedon red velour paper strips (10 x 2.5 o cm) clipped to the inside of each cup and im- =6 o mersedhalfway in the water. Ten ovitraps were z set amongthe tires at eachpile. Ovistrips were J collectedweekly. F Aedes triseriatus eggs appeared on the ovi- o strips occasionally,but only Ae. albopictus eggs were counted.Standard errors ofAe. albopictus eggcounts rangedfrom t4 to 29% of the average. Macrocyclopswere killed in most of the tires by a severefreeze in December1989, when the temperaturestayed below -6'C for 3 consecu- tive days. Water in the tires froze solid. v roo IJJ Macrocyclopswere reintroduced to all tires in lr.t the 2 treatedpiles in March 1990,at which time = o- immature and adult Ae. albopictuswere observed E piles. In June 1990,all tires at the treated F at the a50 a piles were examined for Macrocyclopsand mos- ,J quito Iarvae and pupae; a sample of 10 tires at o uJ the control pile was examined for the same. Landing rates, basedon 4 personscollecting for a total of 60 man-rninutesat each pile, were Sept measured at the same time. Ten oviposition populations at tire piles traps were operatedfor a week at each pile. Fig. l. Aedes albopictus after introducing Macrocyclops albidus: A, number of Macrocyclnps in the 2 tire piles to which Macrocyclops were introducedt B, number of Ae. albopictus larvae RESULTS and pupae in each tire pile; C, Ianding rate of adult Ae. albopictus on hurnan bait at each tire pile; D, It took several weeks after introduction for oviposition of Ae. albopictrzs in ovitraps at each tire the Macrocyclopsto build up their numbers in pile. The time axis of the graphs starts on the date the treated tires (Fig. 1A). By one month, Mo- that Macrocyclops were introduced into the tires. DECEMBER1990 MAcBocYCLoPsrN TrREs 691 crocyclnpsnumbers averagedmore than 100/tire October, but the decline started after Ae. albo- in each of the treated piles, but 5-10 tires in pictus populations at the treated piles were al- eachpile contained fewer than 20 individuals. A ready closeto zero. month later each treated pile contained at least No adult Ae. albopictuswere collected at the 50 Macrocycbps,and most tires had more than treated piles during November (Fig. 1C), nor 100. Macrocyclopswere present in all treated wereanyAe. albopictuseggs collectedby ovitraps tires until December. at the treated piles during that month (Fig. 1D). Small numbers of Ae. triseriatus larvae (av- Adult landings of Ae. albopictus at the control eraging less than 1 /tire) were observedin pile averaged2.3 per min during November, a the control pile throughout the study. Similar total of 138 + 12 landings during that period. numbers were found in the treated piles during Oviposition at the control pile averaged13 eggs/ June-September, but no Ae. triseriatus lawae ovistrip/week during November, a total of 650 were observedin the treated piles during Octo- t 78 eggs. ber-November. Large numbers of Cx. salinarius Macrocyclopswere present in all tires in the larvae were present in all piles, particularly dur- treated piles when they were examined in June ing October-November, with no difference be- 1990. No Ae. albopictu.slarvae or pupae were tween treated and control piles. Larvae of Tox- found in either of the treated piles (Table 1). orhynchitesrutihts (Coq.) were sometimesfound However, very small numbers of adult Ae. albo- in tires at both treated and control piles. Most pictu.s werc collected at both treated piles in tires with Tx. rutilus in the control pile con- June, and small numbers of Ae. albopictus eggs tained substantial numbers of Ae. albopictttslar- were collected in ovitraps at the treated piles. vae and pupae. Aedes albopictr.rclarvae, pupae, adult landings The averagenumber of Ae. albopicfuslarvae and oviposition at the control pile were normal and pupae in the control pile was more than for that time of year. 200/tfte until the middle of September(Fig. 1B). pupae The number of Ae. albopicfirslarvae and DISCUSSION in the treated piles started to decline in the middle of August, 2 weeks aft,er Macrocyclops In summary, Macrocyclapseliminated all Ae. introduction. Larvae and pupae were nearly ahopictns larvae from the treated tire piles gone from the treated piles by the middle of within 2 months. It took an additional month September, 8 weeks after Mocrocyclopsintro- for the adult mosquitoes to disappear. Aedes duction. No Ae. albopictuslarvae or pupae were albopictus larvae did not reappear in the tires found in the treated tire piles during October the following summer, but adults reappearedin and November. The control pile averaged43 t small numbers. 9.8 (SE) immature Ae. albopictusper tire during It is possible that a natural decline due to Octoberand November. autumn temperatures and diapausecontributed The landing ratesof most speciesof mosqui- to the disappearanceof adult Ae. albopictusfrom toes were substantial at all 3 tire piles through- the treated piles during October-November. out the study. In contrast, Ae. albopictusIanding However, diapause was incomplete that year; rates and oviposition at the 2 treated piles Ae. albopictuspopulations remained substantial started to decline about 1 week after the decline at the control pile and elsewherein New Orleans of Iarvae and pupae, i.e., 3 weeks after Macro- until December. cycLopsintroduction (Fig. 1, C and D). Landing Adult populations of Ae. albopictusat treated rates and oviposition at the treated piles were and control piles were drastically different by closeto zeroby the llth week after Macrocyclops November; a null hlpothesis that the treated introduction. Aedesalbopictrzs landing rates and piles and control pile had the same number of oviposition at the control pile declined during mosquitoescan definitely be rejected.Assuming

Table 1. Aedesahopictus at the 3 experimental tire piles in June 1990." Larvae * Adult Tire piles pupaeb landings" Ovipositiono M ac r ocStc lops -tr e atnd" 0 0.05+ 0.03 9+3 Macrocyclops-treatet 0 0.03a 0.02 '7 -t- 3 Control pile' 142+ 3l 2.20+ 0.2r 118-f 13 " Three months after reintroducing Macrocyclnpsto the 2 treated tire piles. b Averagenumber of immature A edes albopictusper tire t SE. " Averagelandings per minute + SE. o Averagenumber of eggsper ovistrip in I week + SE. " No larvivorous copepodsin control tire pile. 692 Jounr.rar,oF THEAupnrclN Mosqurro Coxrnol Assocrauou VoL.6,No.4 that adult landings followed a Poisson distribu- A point of concern is the time that it took tion, if the expected number of landings at a Macrocyclapsto eliminate all larvae from the treated pile was actually the 138 landings ob- treated tires during July-August 1989.The delay servedat the control pile during November, the was due in part to the time it took for Macro- probability of no observedlandings at a treated cyclopsto build up their numbers in the tires. value) - pile would be P[0] = g-(exPeted e-138 - This part of the delay could be eliminated by 10-60.For oviposition, assumingthat a mosquito introducing more Macrocyclnpsat the beginning lays an averageof 10 eggseach time it deposits (e.g.,100 individuals/tire). eggsat a container, the 650 eggsat the control A second and equally important source of pile represent approximately 65 independent delay arisesbecause cyclopoid copepodsare not sets of eggson the ovistrips. The probability of large enoughto kill third and fourth instar mos- no observedeggs at a treated pile would be P[0] quito larvae. Any large larvae that are in a tire - : s-65 10-'e.The hypothesesthat eachtreated when Macrocyclopsis introduced will escapepre- pile was the same as the control pile can there- dation and may eventually become adult mos- fore be rejected at an extremely high level of quitoes.If a tire is crowdedwith larvae, the food significance:p: 1g-eofor adult landings;P : supply is reduced, larval development is slow, 10-2sfor oviposition. and it can take more than a month for some of We do not know whether there was still a the larvae to emergeas adults. small population of adult Ae. albopictus at the The seconddelay could be eliminated by ap- treated piles during November,but if there were plying a larvicide at the sametime Macrocyclops any adults, their numbers were extremely small. is introduced. The larvicide kills all larvae, mos- Assuming a Poisson distribution, the fact that quito production is terminated, and the Macro- no landings or eggs were observed at either cyclopscontinue the treatment after the larvi- treated pile during November gives a 99Volevel cide loses its potency. Several commonly used of confidencethat the total expectednumber of larvicides (e.g.,diesel o1l,B.t.i., and methoprene) Iandings or egg sets at each treated pile during are compatible with Macrocyclops(W. Y. Che Novemberwas lessthan 1.15-much less than and G. G. Marten, unpublisheddata). the 138 adult landings and 65 eggssets observed at the control pile during the sameperiod. ACKNOWLEDGMENTS Becausethere were no Ae. albopicrtuslarvae in the treated tires in June 1990,it is unlikely that I wish to thank Raymond Bailey, Donald Elia- the small numbersof adultAe. albopictusaround son, Bruce Francy, Jerome Freier, Thomas the treated piles at that time camefrom the piles Monath and ChesterMoore of the Division of themselves. The experimental introduction of Vector-Borne Infectious Diseases,Centers for Ae. albopictus appears to have established a DiseaseControl, for their suggestionsand sup- small population in the surrounding woods.This port. I am highly grateful to Edgar Bordes, Mi- interpretation is supported by the observation chael Carroll, Edward Freytag, Brooks Hart- of low levels of oviposition and adult landings man, LawrenceKabel, Jack Leonard, Mieu Ngu- in the woodsbetween the tire piles in June 1990. yen, Stephen Sackett, Charles Spizale and The results of this study suggestthat Macro- others on the staff of the New Orleans Mosquito cyclopsalbidus should be consideredseriously as Control Board for their numerouscontributions a biological control agent for operational use in to the study. Earl Anthony, Jr., Mary Klinger, tires. The prospectsfor an inexpensiveand con- Willie McKinney, Giai Ngo and Peter Omonde venient supply are promising (Rividre et al. provided Iaboratory and field assistance.The 1987,Marten 1990b).Techniques for masspro- author was supported by a National Research duction, storageand transport are currently un- Council seniorresearch associateship. der developmentat the New Orleans Mosquito Control Board. REFERENCES CITED Methods for applying Macrocyclopsto large tire piles remain to be worked out in detail. Hawley,W. A. 1988.The biologyof Aedesalbopictus. Macrocyclopscan be squirted into tires with a J. Am.Mosq. Control Assoc. 4(Suppl. 1):1-40. hand-held, pressure-pumpedlarvicide sprayer. Marten,G. G. 1989.A surveyof cyclopoidcopepods Broadcast spraying is also a possibility; Macro- for control of Aedesalbopictus larvae. Bull. Soc. cyclopssurvive when sprayed from a backpack VectorEcoJ. 14:232-236. Marten, air-blast sprayer. BecauseM acrocyclopssurvive G.G. 1990a.Evaluation of cyclopoidcopepods for Aedesahopictus control in tires. J. Am. Mosq. without water as long as they are damp, it may ControlAssoc. 6:681-688. be possibleto apply them with a granular carrier Marten, G. G. 1990b. Issues in the development of similar to ones that have proved effective for cyclopsfor mosquito control, pp. 159-164.In: M. F. delivering larvicides to tire piles (Novak et al. Uren, J. Blok, L. H. Manderson(eds.), Proc. Fifth 1990). Arbovirus Research in Australia. CSIRO and DECEMBER1990 MAcnocYcLoPs rN TrREs 693

QueenslandInst. Med. Res.,Australia. Rividre, F., Y. Sechan and B. H. Kay. 1987. The Novak, R. J., B. A. Steinly,L. Haramis,J. Clarke,B. evaluation of predators for mosquito control in Farmer and R. Cieslik. 1990. Penetration rate of FrenchPolynesia, pp. 150-154. In: T. D. St. George, two pesticide carriers at a large used-tire storage B. H. Kay and J. Blok (eds.),Proc. Fourth Symp. facility in Chicago, Illinois. J. Am. Mosq. Control Arbovirus Research in Australia. CSIRO and Assoc.6:188-196. QueenslandInst. Med. Res.,Australia.