Impact of Minor Reductions in Adult and Larval Survival, Fecundity and Hatch on the Population Dynamics of Psorophoracolumbiae: a Simulation Study1,2

476 JouRNel oF THEAnnnrceN Mosqurro CoNrnol AssocurtoN Vol. 7, No. 3

IMPACT OF MINOR REDUCTIONS IN ADULT AND LARVAL SURVIVAL, FECUNDITY AND HATCH ON THE POPULATION DYNAMICS OF PSOROPHORACOLUMBIAE: A SIMULATION STUDY1,2

D. A. FOCKS

Medical and VeteriruaryEntomology Research Laboratory, ARS, USDA, GainesuiLl.e,FL 32604

ABSTRACT. Laboratory work has shown that mosquitoesobtaining blood mealsfrom animals treated with ivermectin exhibit lowered adult survival, fecundity, egg hatch and larval survival. Computer simulation evaluatedthe consequencesof this phenomenonin field populations of Psorophoracolurnbiae feeding on cattle in the rice agroecosystem.Results suggestthat rather minor reductions, on the order of 10% below normal, in these life history parameterswould significantly affect the population dynamics of this speciesin this particular system. Significant reductions in the amount of insecticide used for mosquito abatement are also projected.

INTRODUCTION the control of human and animal vector-borne diseasesif administration of the drug at suitable The avermectins are a class of macrocyclic dosageto vertebrate hosts could causethe death Iactones produced by fermentation of the soil of hematophagousarthropod vectors after a sin- microorganism Streptomyces auermitilis. The gle blood meal on the treated animals. Where low toxicity of ivermectin, a mixture of dihy- domestic animals are the primary host for the droavermectins,in most vertebrates and its ac- arthropod, Iower, non-lethal dosessufficient to tivity against a broad range parasites of at ther- reduce one or more reproductive parameters apeutic dosesof <1 mg AI/kg permits this agent could contribute to population suppression to be used as a systemic to control insects, (Langleyand Roe 1984). acarines and helminths in domestic animals My interest in ivermectin stems from field (Cambell et al. 1983). Administered orally or (Mclaughlin and Focks 1990) and computer parenterally, ivermectin is effective against ar- simulation studies (Focks et al. 1988a, 1988b) thropod endoparasitessuch as the common cat- on the population dynamics of Psorophora co- tle grub, Hypoderma lineatum (Villers) and the lumbiae (Dyar and Knab). These studies indi- common horse bots, Gasterophilus nasalis cated that the abundance primary (Linn.) andG. intestinolis(DeGeer) (Drummond of cattle, the blood source for this mosquito, is a key factor 1985). Ectoparasites affected by ivermectin in- determining Ps. columbiaedensities in some clude many species of ticks, mites and lice rice-cattle agroecosystems. (Miller et al. 1986). The biological de- pendencyof Ps.columbloe upon cattle led Ivermectin reducesadult survival and subse- Kuntz et al. (1982) to propose a form of host quent fecundity and egg hatch of intermittent management, the treatment of cattle with insecticide, ectoparasitessuch as tsetse flies (Langley and as a potential control method. Computer Roe 1984), stable and horn flies (Miller et al. simu- Iations showed that host management 1986),mosquitoes (Focks et al. 1991a)and cer- tech- niques,by exerting continual pressure atopogonids(Standfast et al. 1984). Given a on female survival and fecundity, sufficiently high dose of ivermectin, a single could result in significant, Iong-term population (Focks blood meal taken on mice (Pampiglione et al. suppression 1985)or rabbits (Koopman et al. 1989)will kill and Mclaughlin 1988).In addition to attacking mosquitoes.Pampiglione et al. (1985)proposed the key regulatory factor in the life history of that these compoundsmight be appropriate for Ps. columbiae in this environment. host management has the additional advantageof being, unlike current practice,3spatially confined to 'This the host animals which are visited bv virtuallv researchwas conductedby the U.S. Depart- ment of Agriculture, Agriculture ResearchService in cooperationwith the USDA, CSRS Southern Regional 'Olson, Project S-230involving state Agricultural Experiment J. K. 1983. Final report of the Riceland station personnel located in Arkansas, California, Mosquito Management Program for the development Louisiana, Mississippi and Texas. of strategies optimizing non-chemical approaches to 2 This paper reports the results of research only. managing mosquito populations in freshwater irri- Mention of a commercial or proprietary product does gated cropping systems using the riceland agroecosys- not constitute a recommendationor an endorsement tem as a model. Texas A & M Research Foundation. of this product by the U.S, Department of Agriculture. College Station, TX. SEpTEMBER1991 Sruur-erIoN SruoY or PS. Cot uuartE POPULATION DyNeulCS

Table 1. Estimates of various life history parameters as a function of dose of ivermectin and time since drug administration. These reductions were used in the simulation studies and are based on data from Focks et al- (1991a).

Response in treated mosquitoes as a percentage of response in untreated

Days after Adult Egg Larval Dose treatment survival Fecundity hatch survival Averaget 2 mg Allkg body 3 100.0 87.6 76.0 95.0 89.7 weight 10 100.0 80.8 66.3 85.0 83.0 1n 100.0 95.8 93.9 86.7 94.1 o/ 100.0 97.2 100.0 86.7 96.0 10 mg AI/kg body 3 74.0 L3.2 0.0 23.2 weight 10 76.0 15.9 0.6 2t.8 28.6 JJ.:) 13.0 44.7 47.8 t7 100.0 't7.0 24 100.0 73.9 68.0 65.9 JA 100.0 86.0 84.9 80.4 87.8 38 96.0 86.4 86.5 88.4 89.3 98.0 86.4 87.2 92.2 91.0 52 89.1 86.4 87.8 93.5 89.2 I Arithmetic average of reductions in adult survival, fecundity, egg hatch and larval survival- all adult females.A preliminary field test of this Iaboratory model, Aedes aegypti (Linn.) and concept reduced Ps. columbiae populations by ivermectin-treatedrabbits, were used (Focks , et more than 80% (Fockset al. 1991b,Nasci et al. al. 1991a).In that study,Ae. aegpti wereblood 1990),suggesting that host managementmay be fed on rabbits which had been injected 3-52 a viable method for suppressingpopulations of daysearlier with either 10 or 50 times the labeled Ps.columbiae. dose (2 or 10 mg AI/kg body weight) of ivermec- Ivermectin-caused reductions in adult sur- tin.a Depressions in adult survival, fecundity, vival, fecundity, hatch and/or subsequentlarval egg hatch, and larval survival were monitored survival might impact Ps.columbiae populations during the first gonotrophic cycle for females in a fashion similar to that produced by killing fed up to 52 days after injection. Table 1 pre- cattle-visiting females. In this study, computer sents the depressionsin these parameters due simulation was usedto project the consequences to ivermectin treatment as a percentageof nor- of suppressingthese life history parameters in mal values. field populations of Ps. columbiae.This study is Briefly, the results were as follows: As more conceptually similar to a previous simulation time betweendrug administration and the blood study of another noncatastrophic method. meal transpired, drug titers in the rabbit's blood Throne (1989)suggested that resistant varieties, declinedand the effect of ivermectin diminished. in causing minor changes to some life history At the lower dose,adult female survival was not parameters(increased development time leading influenced even among females feeding on rab- to increasedimmature mortality and decreased bits treated 3 days previously. The other param- fecundity) of the flat grain beetle, Cryptolestes eters were only depressedslightly ranging be- pusillus (Schonherr), could have significant im- tween ca. 75 and L00% of normal. Females fed pact on population growth when applied over an on rabbits within a few weeks of drug adminis- entire growing season.He stated that the com- tration at the higher dose exhibited significant bination of noncatastrophic techniques with reductions in all parameters-adult survival was may improve the cost other methods of control ca. 75% of normal, and fecundity and larval of pest management techniques effectiveness survival werc <2070 of normal; of the few eggs even if the noncatastrophic control methods do Iaid, virtually none hatched. As drug titers de- not by themselvesprovide adequatesuppression. clined with time, the reductions were less severe but never did completely return to normal val- DATA SOURCES AND SIMULATION ues-blood mealstaken up to 52 days after drug METHODS Iuermectin: There are no reports in the literature on the consequencesof Ps. columbineob- aThe material used was Ivomec, MK-933, a com- taining blood mealsfrom ivermectin-treatedcat- mercially available (Merck and Company, Rahway, tle. In lieu of using completely hypothetical val- NJ 07065)1% formulation of ivermectin for veterinary ues in the simulation studies. data from a use in cattle and reindeer. JouRNlr, oF THEAuBnlcln Mosqurro CoNrRor,AssocretroN VoL.7,No.3 treatmentwhen the experimentwas terminated, istration and blood meal, were used. At the were still reducingall parametersto about 90% higher dose,one or more parametervalues seen of normal. for days 3, 10, 17 and 24 were so low that zero Sirnulation methods:Computer simulation es- populations were projected. For parameter val- timated the consequencesin field populations of ues observed with lapses of 31 or more days, reductions in adult and larval survival. fecunditv population densities were projected to average and hatch due to ivermectin. The value of the 55% of normal (range:3l-73%). For all simula- simulation studies is that realistic projections tions at the high dose(days 3-52), adult densi- can be made of the dynamics of field popula- ties averaged28% of normal. These reductions tions, reflecting both the influence of ivermectin reflect the combined effect of pressureon adult on reproductiveparameters and the variousden- and larval survival, fecundity and egg hatch. sity dependent feed back mechanisms which Figures1A and 18 also presentthe projected counter control measures,The model used, number of adulticide applications required as a PcSim, is a dynamic life table model of the function of dose and elapsed time after drug population dynamics of Ps. columbioe in the administration. Adulticide applications were rice/cattle agroecosystem(Focks et al. 1988a). triggered when projected populations of Ps. co- PcSim predicts populations (light trap captures lumbiae adults exceeded2.81m2 (an operational and absolute estimates of adult, larval and egg threshold for mosquito control operations); densities)by integrating information on weather spray mortality was assumedtobe g0% and the (date and amount of rainfall, maximum and proportion of land treated assumedto be \5% minimum temperature and pan evaporation (seeFocks and Mclaughlin 1988 for additional rates), agricultural practices (host density, rice details). Becauseit is peak populationswhich acreage,planting and harvesting dates), mos- require insecticide treatments (Focks and quito control activities (datesand areasof space Mclaughlin 1988),the relationship betweenthe spray and larvicidal operations),the influence number of spray applications and averagepop- of the residual animal treatments, and a host of ulation densitiesis not linear but rather logistic. abiotic and biotic factors. The model has been Therefore, although population densities aver- validated with field data (Focks et al. 1988b) aged ca. 55% of normal at the low dose, the and used to evaluate and optimize management number of spray applications during the season strategiesfor this mosquito (Focks and Mc- averaged<20% of normal, ranging between 8 Laughlin 1988).PcSim also has been used to and25% (Fig.1A). At the higherdose (Fig. 1B), evaluate a field trial in Louisiana involving in- no insecticide was required through day 24; secticide-treatedcattle for the suppressionof thereafter spray frequenciesaveraged <20% of Ps.columbiae; here, PcSim- based predictions ol' normal. Over the entire 52-day period, spray adult abundance were within 9% of observed frequency was reduced to a single spray during populations(Focks et al. 1991b). the season,an averageca. 9% of the normal 12 The projected effect of ivermectin presented applications. belowassumes an on-going,wide area,and asyn- chronous treatrnent of cattle at rates and fre- CONCLUSIONS quenciessuch that the aueragetiter among host animals is sufficient to produce in Ps. columbine These simulations are not offered as a precise the depressionsin reproductive function seenin prediction of the effect of ivermectin on Ps. Table 1. For detailson modelequilibration, spe- colurnbiaepopulations in the rice agroecosystem cific values used for land use, agricultural prac- since the estimates of the effect of ivermectin- tices, etc., see Focks and Mclaughlin (1988); treated host animals were based on data ob- Iargehost animal densitieswere assumedto be tained from rabbits by usingAe. aegypti.Rather, 0.33/ha. the simulations were conductedto call attention to the fact that even relatively minor reductions, RESULTS AND DISCUSSION on the order of 10% below normal, in certain life history parameters, however caused,could The projections in Fig. 14. indicate that even significantly suppress field populations of Ps. minor suppressionsin the life history parame- colurnbiae.The simulations suggest that even ters associatedwith the lower treatment rate more dramatic reductions in insecticide use would have a significant impact on Ps. columbiae could be expectedin this system.In light of the populations.For this dose,season-long, average role of cattle as the primary blood meal source treated population densities averaged 55Vo of for Ps. coluntbiae,the current labeled use of normal and ranged between 34 and 70% of nor- ivermectin in cattle, and the discovery of novel mal dependingon which setofvalues, associated avermectins with unprecedented activity (ca. with various Iapsesof time betweendrug admin- 1.500-foldmore active than ivermectin lMrozik SEPTEMBER1991 Sruur-erroN Sruov or Ps. cot'uuaas Popur.etloN DvNeurcs

NA A14

N a fr r 12 c.) rJ .i 1n oL i' a a4. C!B

U.J F.l P i- o i 02 * tr Q4 a0

Sr 0 /2 0.1 z

0.0 310 17 24 Conlrol Days between drug administration and blood meal

B14 0.6

N a - t12 05 !

n, F -U) (!8 q) d AQ -l p - rl o.2 tr qr4 b0

d tr vr > z

0.0 3 10 17 24 31 38 45 52 Control Days between drug administration and blood meal

Fig. fA and 18. Required adulticide applications and average season-long adult Psorophora columbiae densitiesbased on PcSim projections when adult and larval survival, fecundity and hatch in Ps.columbiae arc reduced to the degree seen in Table 1. Adulticide applications triggered when populations of Ps. columbiae adults exceeded2.8/m2; spray mortality was assumedto be 907o,proportion of land treated was assumedto be 15% (SeeFocks and Mclaughlin (1988) for additional details).

et al. 19891),more effort should be devoted to suggeststhat the method could have application developing management for Ps. columbiap via for other riceland mosquito speciesas well. the use of ivermectin-basedsystemics. The re- Further researchquestions include the follow- cent report of significant correlations between ing: Could existing methods for administration most mosquito speciesin this environment and of insect growth regulatorsto cattle, e.g.,ground cattle density (Mclaughlin and Focks 1990) feed (Bay and Boyde 1987), sustained-release 480 JouRNer,oF THE AvnnrcnN Mosgurro CoNrnor, Assocrenot VoL. 7, No.3 boluses (Miller et al. 1979), drinking water of methoprene mineral blocks to cattle for control. (Miller et al. 1977)or mineral blocks (Harris et J. Econ. Entomol. 67:384-386. al. 1974),be adapted for delivery of ivermectin? Koopman, J. P., P. M. Scholtem,T. Van Zutphen and What are the dose/responserelationships in the J. B. A. Hooghof. 1989. Effect of ivermectin on Ps. columbiaecattle system? What about other psoroptic ear mange in rabbits. Tijdschr. Dierge- speciesof mosquitoes?Ivermectin is expen- neeskd.114:825-828. Kuntz. K. J.. J. K. sive-are there situations where costswould not Olson and B. J. Rade.1982. Role be prohibitive? of domestic animals as hosts for blood-seekingfe- males of Psorophoracolumbiae and other mosquito speciesin Texasricelands. Mosq. News 42:202-210. ACKNOWLEDGMENTS Langley,P. A. and J. M. Roe. 1984.Ivermectin as a possible control agent for the tsetse fly, Glnssina I thank J. K. Olson,Texas A & M University, morsitans.Entomol. Exp. Appl. 36:137-143. CollegeStation, TX, R. J. Nasci,McNeese Stale Mclaughlin, R. E. and D. A. Focks. 1990.Effects of University, Lake Charles,LA, and J. A. Hog- cattle density on New Jersey light trap mosquito sette, Medical and Veterinary Entomology Re- captures in the rice/cattle agroecosystemof south- search Laboratory, ARS, USDA, Gainesville, western Louisiana. J. Am. Mosq. Control Assoc. FL, for critical reviews of the manuscript. 6:283-286. Miller, J. A., J. L. Eschle,D. E. Hopkins,F. C. Wright and J. J. Matter. 1977. REFERENCES CITED Methoprene for control of horn flies: a suppressionprogram on the Island of Bay, D. E. and L. S. Boyde. 1987.Effect of horn fly Molokai, Hawaii. J. Econ. Entomol. 70:417-423. larval density on methoprene toxicity. Southwest. Miller, J. A., F. W. Knapp, R. W. Miller and C. W. Entomol. 12:259-267. Pitts. 1979. Sustained-releaseboluses containing Cambell,W. C.,M. H. Fisher,E. O. Stapley,G. Albers- methoprenefor control of the horn fly and face fly. Schdnbergand T. A. Jacob. 1983. Ivermectin: a Southwest.Entomol. 4:195-200. potent new antiparasitic agent. Science 221:823- Miller, J. A., D. D. Oehler,A. J. Siebenalerand S. 828. Kunz. 1986. Effect of ivermectin on survival and Drummond, R. O. 1985.Effectiveness of ivermectin fecundity of horn flies and stable flies (Diptera: for control of arthropod pests of livestock. South- Muscidae).J. Econ.Entomol. 79:1564-1569. west.Entomol. (Suppl.7):34-42. Mrozik, H., P. Eskola,B. O. Linn, A. Lusi, T. L. Shih, Focks,D. A. and R. E. Mclaughlin. 1988.Computer M. Tischler, F. S. Waksmunsik, M. J. Wyvratt, N. simulation of managementstrategies for Psorophora J. Hilton, T. E. Anderson, J. R. Babu, R. A. Dybas, coLumbiaein the rice agroecosystem.J. Am. Mosq. F. A. Preiserand M. H. Fisher. 1989.Discovery of Control Assoc.4:399-413. novel avermectins with unprecedentedinsecticidal Focks, D. A., R. E. Mclaughlin and B. M. Smith. activity.Experientia 45:315-316. 1988a. A dynamic life table model of Psorophora Nasci, R. S., R. E. Mclaughlin, D. A. Focks and J. coLum,biaein the southern Louisiana rice agroeco- Billodeaux. 1990.Effect of topically treating cattle system with supporting hydrologic submodel, Part with permethrin on Psorophoracolumbiae (Diptera: 1. Analysisof literature and model development.J. Culicidae) blood feeding in a southwest Louisiana Am. Mosq. Control Assoc.4:266-281. rice-pastureecosystem. J. Med. Entomol. 27:1031- Focks, D. A., R. E. Mclaughlin and B. M. Smith. 1034. 1988b. A dynamic life table model of Psorophora columbiaein the southern Louisiana rice agroeco- Pampiglione,S., G. Majori, G. Petrangeliand R. Romi. system with supporting hydrologic submodel, Part 1985.Avermectins, MK-933 and MK-936, for mosquito 2. Model validation and population dynamics.J. control. Trans. R. Soc. Trop. Med. Hyg. Am. Mosq. Control Assoc.4:282-299. 79:797-799. Focks,D. A., R. E. Mclaughlin and S. B. Linda. 1991a. Standfast,H. A., M. J. Muller and D. D. Wilson. 1984. The effects of ivermectin (MK-933) on the repro- Mortality of Culicoides breuitarsis fed on cattle ductive rate of.Aedes aegypti (Diptnra: Culicidae). J. treatedwith ivermectin,pp. 611-616.In:T.L.Bar- Med. Entomol. 28:501-505. ber and M. M. Jochim (eds.).Progress in clinical Focks,D. A., R. E. Mclaughlin and R. S. Nasci.1991b. and biologicalresearch, Vol. 178. Bluetongueand Elfect of permethrin treatment of cattle on Psoro- related orbiviruses; Meeting, Monterey, CA, USA, phora columbiaepopulations: preliminary field test Jan. 16-20,Alan R. Liss, Inc., New York. of a host managementconcept. J. Am. Mosq. Con- Throne, J. E. 1989.Effects ofnoncatastrophic control trol Assoc.7:1-6. technologiesthat alter life history parameters on Harris, R. L., W. F. Chamberlainand E. D. Frazar. insect population growth: a simulation study. En- 1974.Horn llies and stableflies: free-choice feedins viron. Entomol. 18:1050-1055.