DECEMBER 1991 MoseurT o VpctoR SYMPoSIUM

MOSQUITO VECTOR CONTROL AND BIOLOGY IN LATINAMERICA-A SYMPOSIUM'

GARY G. CLARK'� enn MARCO F. SUAREZ (ORGANIZERS)

San Juan Laboratories, Centers for Disease Control, San Juan, PR 00936

ABSTRACT. The first Spanish language symposium ever presented by the American MosquitoControl Association was held as part of thJSTth Annual Meeting in New Orleans, LA, in March 1991-A major objective was to increase and stimulate participation in the Annual Meeting by vector.controlspecialists and public health workers from Latin Americi. This publication summarizes the 25 individualpresentations tirat encompassed the following areas: mosquito biology and ecology, evaluation ofchemicaland biological control miasures for mosquiloes, epidemiology of dengue, and current and developingdengue prevention programs.

INTRODUCTION

The American Mosquito Control Association(AMCA) is recognized as the premier organiza-tion of its type in the world. To better serve theinternational vector control and public healthcommunity, a symposium with presentations inSpanish was planned, organized and held at the57th Annual Meeting of the AMCA in NewOrleans, LA. The objectives were to encouragecolleagues from Latin America to attend theAMCA meeting, discuss their experiences invector control, present results of recent studies,promote greater interaction with AMCA mem-bers and stimulate future collaboration in theresolution of vector control and vector-bornedisease problems.

In planning for this symposium, all AMCAmembers and about 60 other governmental oruniversity vector control personnel in LatinAmerica were invited to submit abstracts. As aresult of the enthusiastic response, 25 presen-tations covering a wide variety of topics wereincluded in this session. The quality of the pres-entations, enthusiasm of the speakers and at-tendance at the session contributed to its suc-cess. The symposium's objectives were clearlymet and provided a forum that has led to greaterinteraction between AMCA members and theirneighbors in Latin America where vector-bornediseases, Iike malaria and dengue, are problemsthat must be dealt with on a daily basis.

Special recognition for support ofthe sympo-sium goes to the Vector Biology and Control(VBC) Project (Andrew A. Arata), managed forthe Agency for International Development by

I Presented at the 57th Annual Meeting of theAmerican Mosquito Control Association, New Orle-ans, Louisiana, on March 19, 1991.

2 Reprints may be requested from Gary G. Clark.

Medical Sewice Corporation International ofArlington, VA; the New Orleans Mosquito Con-trol Board (Edgar S. Bordes, Jr.), AmericanCyanamid (William Jany), Vectec (Isaac S.Dyals) and Sumitomo Chemical America (S.Ohutsuki) for providing financial support formany of the international visitors. The VBCProject also provided copies ofthe abstracts (inEnglish) for non-Spanish speakers attending thesession and provided funds for publication ofsymposium abstracts in the Journal of theAMCA. As a result of the response from partic-ipants and the AMCA leadership, it is expectedthat this unique forum will be included in futuremeetings of the organization.

SYMPOSIUM ABSTRACTS

Resting sites for Aed.es aegypti inPanam6 (Sitios de reposo de Aed.es

aegypti en Panama)G. Ddvila, M. J. Nelson, A. Turner and A. Garcia

Comision del Canal de Panama, OrganizacionPanarnericana de Ia Salud y Uniuersidad de

Pananta, Panama

The ground application of ultra-low volume(ULV) insecticidal fog by truck is the mostcommon emergency measure to control Aedesaegypti during a dengue epidemic. Studies indi-cate that these mosquito populations recoverwithin 3-7 days and that it is not a very effectivecontrol measure. These questions, and the pos-sibility of imported cases of dengue enteringPanama from nearby countries and because lar-val infestations (house indices) have reachedl0% and above in certain metropolitan and out-lying districts during 1988-89, stimulated us toinvestigate the resting behavior of. Ae. aegyptiinrespect to the effectiveness of ULV spraying.

This project received support from the Pan-ama Health Ministry and SNEM mosquito con-trol, who assigned 4 technicians to work on the

oJ4 JounNer, oF THE AMERTcIN Mosqurro CoNrRor, AssocrarroN VoL. 7, No. 4

project. We measured each mosquito's restingplace in relation to height, location light inten-sity and relative humidity.

Data were gathered over 30 days during Oc-tober and November 1990. Collections weremade during a 6-h period between 0800 and 1600hours. Seventy-six houses, 20 mechanics' shopsand 30 manholes were inspected. We collectedan average of5 mosquitoes per house. The high-est number of Ae. aegyptl collected in any singledwelling was 42 females and 23 males.

Our study compared distribution of the mos-quitoes resting inside the house vs. those col-Iected outside. We recorded the nature of theresting site, the number collected in each roomand the number collected on other surfaces.Females were dissected to determine the paritystate, and if possible, their physiological age(number of ovarian dilatations) and degree offeeding vs. ovarian development (Christophers'stages).

Seventy-seven percent of adult Ae. aegyptimosquitoes were collected inside the dwellineand 23% were found resting outside. Fifteeipercent were males and 85% were females. TheIesser number of males may be related to thescarcity of larval habitats near the houses orbecause females may have dispersed some dis-tance since emergence in search of a bloodmeal.Males are known to disperse over shorter dis-tances during their much shorter life span.

The largest number of female mosquitoes werefound in bedrooms, but more males occurred inbathrooms. Seventy-one percent of the Ae. ae-gypti females collected were resting in protectedareas under sinks or within rustic china cabi-nets. Fifty-seven percent of the Ae. aegypti werccollected less than 1 m above the floor and94%less than 2 m above the floor.

Dissections showed that 55% of the parousfemales had one dilatation, 25% had 2 dilata-tions and 20% had 3 dilatations, indicating thatat least 45% had, fed and laid eggs 2-3 times andthat they were more likely to have come intocontact with an infectious agent. Only one mos-quito was found with 6 dilatations.

Gravid and semigravid females were foundresting on unexposed surfaces such as clothing,under and behind furniture and other articles inthe house. They were also taken from restingsites such as hanging garments, pictures, por-celain vases and ornaments on the walls wherethey would be more protected from ULV insec-ticide droplets. The remaining (427o) ones werefound on exposed surfaces. Most nulliparousfemales appear to rest on sites that would bemore exposed to ULV sprays, and we found that66% of males rest in similar areas that wouldalso be exposed to the spray.

Several insights were gained from our study,but we consider the following the most impor-tant:

1. A higher percentage of parous females,which represent a greater epidemiological riskregarding dengue transmission, rest in more pro-tected areas and on surfaces where ULV spravdroplets are less likely to drift.

2. Higher percentages of males and nullipa-rous females were found resting in the moreexposed sites and would therefore be killed bvroutine ULV spraying.

Relationship between larval indices andadult densities of Aed.es aegypti inEl

Progreso, Ifonduras, 1989-90 (Relaci6nentre indices larvarios y densidad de

adultos de Aedes aegypti en El Progreso,Ilonduras, 1989-90)

E. Gil-BellorinProyecto Rockefelbr, El Progreso, Honduras

To determine if there was a direct relationshipbetween larval indices and adult densities, weperformed 4 entomological surveys in 100% ofthe households in 8 "barrios" in El Progreso.They were conducted in June 1989 when therainy season begins; in October 1989 when thereis more rainfall; in February 1990 when the rainyseason ended; and in July 1990 when the nextrainy season began. We also captured adultAedes aegypti mosquitoes with the CDC back-pack aspirator in 30 houses in 6 "barrios" locatedin 8 that were surveyed for larvae. Adults werecaptured once per week in each "barrio."

We observed that the larval indices increasedat the beginning of the rainy season, with thehighest peak (100%) for the month of greatestrainfall and decreased at the end of the rainyperiod. In contrast, the adult density increasedat the beginning of the rainy season and laterdecreased, remaining constant during themonths of more rainfall with a small increase atthe end of the season. This variation in adultdensities was directly related to the incidence ofdengue cases. We observed that there was anincrease in adult densities followed by the ap-pearance of dengue cases 1-2 months later.When the adult density decreased, the denguecases also decreased almost immediately.

We conclude that there was no direct relation-ship between the larval indices in householdsand the adult population, most Iikely becausethe breeding containers that become positiveduring the rainy season are not good productionsites for adult mosquitoes. The increase in adultdensities at the beginning and end of the rainyseason occurred because of an increase in waterlevels, but in a certain way, it had to do with thestabilization of the breedinq containers. There-

Drcnlrnrn 1991 MoseuITo Vrcron Svvposrurra 635

fore, we must seek other risk indicators asidefrom those we have been dealing with untilrecently (e.g., presence of larvae), because thereal success of a good risk indicator is that itindicates when we should act to prevent a prob-lem (e.g., a dengue epidemic). Therefore, denguecontrol programs should include investigation ofthe number of pupae, type of breeding con-tainers, seasonal variation, as well as abundanceof adult mosquitoes, using the CDC aspiratorwhich offers an operational advantage.

Biology and vectorial capacity ofAnopheles albimanus on the southern

coast of Guatemala (Estudio de biologia ycapacidad vectorial de Anopheleo

albimanus en la costa sur de Guatemala)F. V. Zamora and M. L. Calder6n

Ministerio de Salud Piblica y AsistenciaSocial, Guatemaln

From October 1987 to March 1990, we studiedAnophcles albimanus in 2 areas on the southerncoast of Guatemala. They were the village of LaBlanca in Ocos San Marcos and the village ofEl Aceituno in Escuintla.

We collected mosquitoes between 0500 and1200 h daily and dissected 9L7o of the Anophelesfemales. Indoor collections yielded 1,783 Anoph-eles, of which 340 were nulliparous and 1,279were parous. Outdoor collections yielded 2,917of which 515 were nulliparous and 2,114 wereparous.

We marked 220 Anopheles using fluorescentdusts and recaptured 40. The marked mosqui-toes were observed resting 369 times on spray-able surfaces. The average time at rest was 14min and 40 seconds. Of the females recaptured,12% died, within the 24-h observation period.

At La Blanca, we dissected 5,681 Anopheles.Of these only 3 (0.05%) were positive for Plos-modium uiuax sporozoites. None was positive forP. f alciparum sporozoites.

We found that 3Vo of the Anopheles sampledhad taken blood from humans and domesticanimals with the highest percentage from hu-mans. Comparing the source of the animalblood, we demonstrated by the Forage Indexthat the highest rate was from bovines in LaBlanca and from equines in El Aceituno.

According to results from both localities, theblood meal analysis from human bait and animalshelter collections demonstrated that the femaleAn. albim.anus are very long-lived. The bitingactivity is more peridomiciliary; the restingplace inside the house is less than 1 m above thefloor; and with predominant peridomiciliary ref-uges in stalls and woodpiles.

Ecology and biology of AnoPhelesalbimanus in a locality of the Pacific

Coast of Nicaragua (Ecologia y biologiade Anopheles albimanus en una localidad

del Pacifico de Nicaragua)P. Rivera Mendoza and M. M. LoPez

Ministerio d.e Salud, Managua, Nicaragua

To obtain information on the biology andecology of Anopheles albimanus on the PacificCoast of Nicaragua, we selected a locality witha high mosquito density and continuous malariatransmission for study.

Mosquitoes were collected indoors and out-doors between 0600 and 1200 h with human baitin 2 different houses, every month. A total of41,013 specimens were captured (6,199 indoorsand 34.814 outdoors). Fluorescent dust (lumo-gen) was used to mark267 femaleAn. albimanusafter feeding on human bait in order to observetheir postfeeding behavior. The number offlights, as well as height and resting time, wererecorded. The physiological age was determinedby ovarian dissection.

In addition, mortality and survival by gono-trophic cycle were estimated. The collectionsrevealed a bimodal activity pattern for An. al-birnanus, with a peak between 2000 and 2200 hand another (less sharp) between 0500 and 0600hours.

The maximum density was observed in Sep-tember and October, and the lowest density wasseen between the months of November andApril. These differences in densities corre-sponded to the months of highest and lowestrainfall, respectively. The mean physiologicalage of mosquitoes was calculated at 1.44 days(36.2% of female mosquitoes were parous and63.8% were nulliparous). A small percentage offemale mosquitoes had 4 dilations, thus indicat-ing that transmission may be due to a highfrequency of mosquito-man contact and not tothe abundance of old female mosquitoes. Mostmosquitoes preferred resting on walls at 0-1 mheight and on nonsprayable surfaces.

The majority of mosquitoes left the houseswithin 30 min after feeding, and very few re-mained indoors after 60 minutes. Lastly, themortality by cycles among old female mosqui-toes was 83.9%.

Current strategy for the successfulcontrol of malaria in El Salvador

(Estrategia actual de control de malaria ylogros aleanzados en el Salvador)

M. SauerbreyM edical Se ruices Corpo ration I nternational,

San Saluador, EI Saluador

Since 1979, the malaria control program ofthe Ministry of Public Health has epidemiolog-

o,ib JouRraL oF THE ArrapRtclN Mosqurro CoNrRor, AssocrerroN V o L . 7 , N o . 4

ically stratified the problem. At the same time.it decided to change its policy toward a morerealistic control strategl that was based uponan epidemiological approach. This replaced theold method for vector control using pesticideswith an integrated control strategy which in-cluded the following measures:

1. Massive drug administration in high-riskareas.

2. Intradomiciliary spraying with residual in-secticide in 30,000 houses, three cycles per year(bendiocarb).

3. Larvicide application, in 10-day cycles,ttsing 44% Abate@ (temephos).

4. Ultra-low volume spraying as a response tohigh vector density.

5. Distribution of combined drugs (chloro-quine-primaquine) through the voluntary col-Iaborators network and the general health serv-ices.

6. Source reduction by construction of smalland large drainage works with community par-ticipation.

In addition to the above measures, some ad-aptations were introduced. These included de-centralization of diagnosis and timely treatmentof cases, by incorporating the 2,600 membervoluntary collaborator network in the provisionof a 5-day curative treatment and decentraliza-tion of planning in decision-making at opera-tional level. Finally, vector source reduction ac-tivities were emphasized and supported by sig-nificant community participation. AII controlmeasures and new adaptations were prioritizedamong the different locations, according to theepidemiological importance of the problem.

The significant improvement reached in ma-laria control in El Salvador is evidenced by acontinuous decline in morbidity, reaching theIevel of 9,000 reported cases per year for the last3 years. There was a g0% reduction in the num-ber of reported malaria cases between 1980 and1990, with very few cases of P. falciparunl. Theseachievements are the result of the application ofthe new control strategy.

Interagency vector control during floodsin Honduras (Manejo interagencial en elcontrol de vectores en Ifonduras en casos

de inundaciones)J. R. G6mez

Ministerio de Salud. Tegucigalpa, Honduras

The Ministry of Health of Honduras lacks theresources to face the health problems resultingfrom natural disasters alone. As a consequence,there was a need to seek assistance from othergovernmental agencies during October and No-vember 1990 when extensive flooding occurredin northern Honduras.

The main purpose for developing a coopera-tive plan was to prevent the spread of malariaepidemics, before the flooding occurred. Thiswas done through the execution of control ac-tivities against larval and adult anopheline pop-ulations:

1. With the cooperation of the road district ofthe Ministry of Communications and PublicWorks, draining activities were carried out toeliminate 19 breeding sites for Anopheles albi-fil,o,nus and An. crucians.

2. The army provided human resources andmade ULV applications of insecticide in intra-domiciliary sites and temporary shelters. Weused fenitrotion L-100 and made 5 applicationsat 17 day intervals.

3: With the financial assistance of the U.S.Agency for International Development (USAID)project, Bacillus thuringensis var. israelensis wasapplied every 7 days for one month to 32 an-opheline breeding sites (flooded fields) in theurban and rural sectors in the most malariousarea of Honduras.

Although we did not quantity the number ofmalaria cases, we felt that this control strategyhelped prevent a malaria outbreak.

Indoor low-volume insecticide spray as anew method for the control of Anophelesalbimanus in southern Mexico. Village-scale trials of bendiocarb, deltamethrin

and cyfluthrin (Aplicaci6n de insecticidaa bajo volirmen dentro de las casas como

nuevo metodo para el control deAnopheles albimanue en el sur de M6xico.

Pruebas de bendiocarb, deltametrin ycyflutrin a nivel de pueblos)

J. I. Arredondo-Jim6nez, M. H. Rodriguez, D. N.Bown and E. G. Loyola

Secretar[a de Salud, Tapachula, Chiapas, Mexico yOrganizacion Panam,ericana de Ia Salud,

Guatemala, Guatemala

The increasing costs of DDT-alternative in-secticides and their application in malaria vectorcontrol programs in developing countries haveresulted in the need to optimize the use of in-secticides in terms of cost-effectiveness.

With this aim, a comparative study was car-ried out during 1988-89 to determine the effi-cacy of low-volume (LV) indoor spray of bendi-ocarb, deltamethrin and cyfluthrin for the con-trol of Anopheles albimantrc in a coastal plain ofsouthern Mexico. Three villages were used forLV treatments (one per insecticide), a fourthvillage was used to spray wettable powder (WP)bendiocarb, and a fifth village remained un-treated and served as the control.

Low volume spraying was conducted withknapsack mistblowers Fontan R-12@, with No.

DECEMBER 1991 Moseurro VEcroR SYtr,tPostuu 637

1.0 nozzles, giving an average discharge rate of215 ml/min, which deposited droplets of 50-100(MVD). With this technique, it was possible foreach sprayman to treat 25 houses per day (ascompared with a maximum of 8 houses/day/sprayman using the conventional Hudson X-Pert@ compression pumps). Insecticide efficacywas measured by mortality indicators, feedingsuccess, and indoor activity patterns of mosqui-toes.

Low volume spray mortality from wall-bioas-says, indoor resting and human bait collectionsand house curtains showed a residual effect for10 wk, compared with the 13 wk with the WP.Repellency was lower on surfaces treated withthe LV, but no change in feeding success wasobserved.

With this technique the spraying costs de-creased by 37% with bendiocarb and 76% withdeltamethrin. The spraying time was reduced bytwo-thirds and an effectiveness similar to WPwas achieved.

Evaluation of the effectiveness of Reldan40 WP chlorpyrifos-methyl insecticide inthe laboratory for the control of malaria

vectors in Colombia (Evaluaci6n de laefectividad del insecticida Reldan 40 WPclorpyrifos-metil, a nivel de laboratorio

en el control de los vectores de malaria enColombia)

M. L. Quifrones and C. A. LoraMini*terio d,e Salud., Bogotd, Colnmbia

Beyond our principal objective of reducingmalaria incidence. an additional aim of our ma-laria control program is to evaluate insecticidesthat might be used as an alternative to DDT.Susceptibility tests for adults and bioassaystests, following WHO methods, were used byestablishing the LTee.e (lethal time) for the prin-cipal vector populations. We estimated the re-sidual activity on 3 different materials sprayedwith 1 or 2 g/^'. Additionally, tests of irritationcaused by contact with the insecticide and hu-man bait catches, inside and outside of thehouses, were conducted.

Populations of Anopheles darlingi (Villavicen-cio, Choco), An. nuneztouori (Circuta, Uraba)and An. albimanus (Uraba) were highly suscep-tible, having a LTgg.g of 3-21 min, at a concen-tration of 5%. Residual activity on cement withwater paint and on baked brick was 6-10 wk andon wood was 10-14 weeks. There was significantresidual difference between the doses used.Anophclcs d.arlingi from Villavicencio was irri-tated when it contacted the insecticide and in-creased its exophagic behavior in sprayedhouses.

The recommended Reldan dose is 1 g/m'�, withan average residual life between 2 and 3 months.Understanding that exophagic and irritable be-havior reduce the effectiveness of insecticide,Reldan is a good alternative to organophos-phates of low toxicity especially in areas wherethe main construction material is wood.

Entomological aspects of the Arenal-Tempisque irrigation project, Guanacaste,

Costa Rica. Preliminary observations(Aspectos entomol6g:icos del Proyecto de

Irrigaci6n de Arenal-Tempisque'Guanacaste, Costa Rica. Observaciones

preliminares)M. Vargas V. and I. Coto A.

Uniuersid.ad de Costa Rica, San Jos6, Costa Rica

The northern part of Costa Rica, consideredas the dry Pacific zone, has a great agriculturalpotential, but due to its annual drought, produc-tivity is limited to the wet season.

Servicio Nacional de Aguas Subterraneas,Riego y Avenamiento is a national institutionin charge of a vast irrigation project of about66,675 hectares. This funding agency requires acomplete study of the environmental impact ofthe project.

Centro de Investigaci6n y Diagn6stico en Par-asitologia has initiated a study of the entomo-Iogical aspects under a vector monitoring planto detect possible changes, as a result of contin-uous irrigation of population dynamics and di-versity of mosquitoes of medical and veterinaryimportance. The field strategy will include useof ovitraps, Iarval collections, CDC, Fay, SolidState Army Miniature light trap and Weinbergadult traps; and human and animal bait collec-tions. In the laboratory, Belkin's methodolory(1967. Contrib. Am. Entomol. Inst.) will be fol-lowed for specimen preparation.

Preliminary observations from January-De-cember 1990 consisted of 8 field collections ofculicid and simuliid larvae in the area irrigatedin the region of Caias, in which the dam "Ar-

enal-Corobici" is located.The species so far identified include Culicidae:

Culex (Cux.) coronator, Cx. (Cux.) pipiens, Cx.(Cux.) nigripalpus, Cx. (Cux.) declarator,C x.(C ux.) interrogator, C x. (Cux.) ottsqua f coro -nator, Cx. (Cux.) corniger, Cx. MeL) erraticus;Psorophora (Gra.) confinnis, Ps. (Gra.) cingulata,Ps.(Gro.) howardii; Aedes (Och.) spp.; Haema-gogrrs spp.; Arnphclcs (Nys.) albimanus, An.(Nys.) argyritarsis, An. (Ano.) pseu.d.opunctipen-nrs; and Simuliidae: Simulium (Ps.) haemato-,r:Y,S. (H.) puluerulentum, S. (S.) metalli-

638 Jounual oF THE AMERTcAN Mosqurro CoxrRor, AssocrarroN VoL. 7, No. 4

Peritrophic membrane and pupal-adultecdysis in mosquitoes (Membrana peri_

trofica y 6cdisis de pupa-adulto bnmosquitos)

A. Moncayo. K. Lerdthusnee and W. S. RomoserTropical and Geographical Diseose Institute. Ohio

Uniuersity, Athens, Ohio, USA

The peritrophic membrane (pM) in Dipterais of special interest in the field of vector biologynot only because of its physiologically importantrole in insects, but also because of its potentialinfluence on disease transmission. The pM is achito-proteinous, non-cellular sheet enclosinethe food in the midgut of most insects that hav6been studied.

Several functions have been attributed to thePM. Among these are: 1) protecting the midgutcells from abrasion by food particles; 2) servingas a barrier to parasites; and 3) acting as aselectively permeable membrane, allowing somemolecules to pass while excluding others.

Although PMs in mosquitoes have beenknown for a long time to occur in larvae and inadult females following a blood meal, only re-cently has the PM been described in associationwith pupal stage. Romoser (1928) described 2PMs (PMl and PM2) which appeared to formaround the meconium, i.e., the sloughed larvalmidgut epithelium which persists throughoutthe pupal period and into the adult as a mass inthe developing adult midgut. In our study, pre-liminary observations suggested that pM1 foimsearly in the pupal stage while PM2 forms inrecently emerged adults.

On the basis of dissections of Aedes aegyptipupae ofknown ages post-pupation and recentlyemerged adults of known ages post-emergence,we have determined: 1) PM1 forms between 15and 24 h following pupation and its formationoccurs afber apolysis; 2) Pl'{z does not formduring the pupal stage, but rather very soonafber pupal adult ecdysis. It apparently does notform in every individual; 3) since PM2 forms atsome point after the distension of the midgut bygas during ecdysis, we suggest that PM2 is anal-ogous to the PM that forms around a bloodmeal, its formation being induced by the disten-tion of the midgut epithelium; 4) the meconiumand PM1 are voided from the midgut as early as3-4 h following pupal-adult ecdysis, but there ismuch variation in this; and 5) PM2 appears toform only in female mosquitoes.

Functional response of some aquatic pred-ators on Culex pipiens (Diptera: Culici-

dae) larvae (Respuesta funcional de algu-nos depredadores acu6ticos sobre larvas

de Culex pipiensl

H. Quiroz-Mtz., M. H. Badji, M. L. Segoviano-Rdz.and M. L. Deleado

IJniuersidad Autonoma de Nuiuo kon, Monterrey,N.L., Mexico

The functional response (as a statistical toolin predation) was used to evaluate the potentialof aquatic insects to regulate larval mosquitopopulations. The tests were carried out in glasscontainers with 750 ml of dechlorinated waterand fourth instar Culex pipiens larvae exposedto 9 insect genera commonly found as predatorsin the,mosquito breeding habitat. Aftei 24 h, werecorded the number of prey consumed. Therygult9 we,re analyzed according to Holling's(1959) and Rogers' (1972) functional responsemodels.

The 9 predators were: dragonfly naiads (pon.-tala sp.; Odonata: Libellulidae), creeping waterbug (Ambrysus sp.; Hemiptera: Naucoridae), thebackswimmers (Buenoa sp. and Notonecta sp.;Hemiptera: Notonectidae), waterscorpion (Rin-atra sp; Hemiptera: Nepidae), predaceous div-ing beetles (Laccophilus sp. and Therrnonectessp.; Coleoptera: Dytiscidae), larvae of the waterscavenger beetles (Tropisternus sp. and Hydro-philus sp.; Coleoptera: Hydrophilidae).

We estimated the searching capacity and"handling time" as probable indicators to choosethe most effective biological control agent. Thesearching capacity is the ability to move in anarea or volume of water to find prey. The ,.hand-ling time" is the time that the entomophagouspredator spends quelling, killing and eating theprey. With Holling's model, the backswimmer(Notonecta sp.) had the highest value of search-ing capacity (a' : 0.043) with the handlins timeTt :0.172; in second place came the predalceousdiving beetle (Thermonectes sp.) with a, :0.042and Tr, : 0.0357.

With Rogers' model, the backswimmer (No_torlecta sp.) had the highest values of searchinscapacity (a' : 0.414) and the handling time walTr, : 0.900. In second place came the predaceousdiving beetle (Thermonectes sp.) with a, : 0.8g1and Tr, : 0.503. We considered only the valuesfor 24 h exposure tests.

Evaluation and persistence of two Booil-lus sphaericus 2362 formulations on Cu-

lex quinquefasciatus under laboratorvconditions (Evaluaci6n y persistencia iedos formulaciones de Bacillus sphaerieus

2362 en Culex qujnquefasciatus bajocondiciones de laboratorio)

E. Fonseca, M. L. Rodriguez"I.,L.H. Morales R.and L. J. Gal6n

Uniuersi.dad Aut6rwmn de Ntrcuo l*on, Monteney,N.L., Mexico

DECEMBER 1991 MoseulTo VEcroR SYMPoSIUM 639

We evaluated the toxicity and residual activ-ity of 2 commercial formulations of Bacillussphaericus (strain 2362) against field and labo-ratory strains of Culex quinqtrcfosciafus larvae.The larvicidal activity of these formulations wasdetermined by the methods of Lacey (1984) andDe Barjac (1988). Six doses were prepared froma l% stock of each formulation; one granular(ABG-6185 from Abbott Laboratories, Chicago,IL) and the other a concentrated solution (F.C.,BCP- 2 Solvay, Brussels, Belgium). At the sametime, we conducted a bioassay with the Inter-national Standard RST 88 of B. sphaericusstrain 2362. Mortality was recorded at 48 and72 h to calculate the LCso and LCgo for eachformulation. The residual effect was determinedby larvicidal activity and counting of spores,following the WHO (1986) protocol. The basicdose with the LCso and LCgo for each formula-tion was used in 4 repetitions with 20 larvaeeach. Mortality was recorded at 48, 72,96 and168 hours. Later, we eliminated the survivinglarvae and repeated this for every 7 days through28 days. Spores were counted, using the methodof Karch (1990) before treatment and on days7, 14, 2I, and 28 posttreatment. The residualeffect was only made with field-collected larvae.

The larvicidal activity at72h for laboratory-reared larvae for the granular formulation was:LCbo 0.04 mg/liter and LCgo 0.22 mg/liter andfor the concentrated solution the values wereLCso 0.30 mg/liter and LCgo 0.24 mg/liter. Forfield-collected larvae, the granular formulationproduced an LCso of 0.19 mg/liter and an LCe6of 0.66 mg/liter while the concentrated solutionyielded an LCso of 0.24 mg/liter and an LCgo of0.83 mglliter.

The residual effect was reflected in larvalmortality for a period of 28 days following asingle application of the LCso or LCno dose ofeither formulation. The number of coloniesformed increased exponentially over time. Theinitial spore count for the granular formulationwas from 2I9 to 549 spores/ml for the LCso andLCgo, respectively, and from 6.44 x 1010 and 2.18X 1011 spores/ml in the final count for the LCsoand LCso, respectively. For the concentrated so-Iution, the initial count was from 148 to 295spores/ml and the final count was from 2.18 x1011 to 5.5 x 1010 spores/ml for the LCso ani.LCes, respectively.

Effect of sublethal doses of Abate@ onsome biological parameters of Aed.es ae-

gypti (Efecto de dosis subletales de Abate@sobre algunos par6metros biol6gicos de

Aed.es aegyptilF. Reyes-Villanueva, H. de la Garza-Garza and A.

Flores-Leal

(Jniuersid.ad Autonomd de Nueuo Leon, San Nicolasde los Garza, N.L., Mexico

The effect of Abate(D applied at concentrationsof 0.016, 0.020 and 0.025 mg/liter to fourth stageAedes aegypti larvae and the effect on theweights of pupae and adults, quantity of bloodingested, and adult Iongevity were studied' Weused the insecticide from the standard solutionsprovided with the WHO test kit and diluted itin distilled water to obtain the above concentra-tions. The larvae were in the Fz generation andhad been collected in an urban area of Monter-rey, Mexico, and reared on an artificial diet.Females were fed with a l0% solution of honeyand rabbit blood during each gonotrophic cyclewhile males received honey alone. For each con-centration, we randomly selected 20 larvaewhich were then individually separated. Later,the pupae and recently emerged adults wereweighed to 0.01 mg. Later, each female wasconfined with 3 males in a small cage (10 x 10x 10 cm) and a 25 cm2 piece of moist paper wasprovided for oviposition. The females were alsoweighed again after their first bloodmeal.

We observed a decrease in weight in pupaeand in adults of both sexes from those individ-uals that survived the 3 concentrations of Abate.The lowest weights found in the pupae andadults (1.09 + 0.07 and 0.54 + 0.07 mg) were 31and 33% lower than their respective controls.The lowest pupal (2.55 + 0.27 mg) and adult(1.20 + 0.09 mg) weights were only 10 and 6%,respectively, less than controls that were notexposed to sublethal concentrations of Abate.The average weight of blood ingested by femalesexposed to the maximum concentration of Abatewas 0.036 + 0.06 mg. This reduction of 10% wasnot statistically different from the untreatedcontrols.

Longevity of males exposed to the insecticidewas 30.96 + 2.19 days and20Vo longer than thecontrols. No difference in female longevity wasfound.

Evaluation of oviposition rates in prelimi-nary releases of Toxorhynchites theo-

baldi in suburban areas in Mexico (Evalu-acion de la tasa de oviposici6n en libera-ciones preliminares de Toxorhynchitestheoboldi a nivel suburbano en Mexico).J. A. Alvarado-Castro, Filiberto Reyes-Villanueva

and J. A. Flores-Leal(Jniuersidad Autorutma de Nueuo I'eon, San Nicokn

de los Garza, N.L., M'itico

Toxorhynchites theobaldi is a mosquito pred-ator that has been considered as a control agentfor Aedes aegypti, the mosquito vector of dengueviruses. Previous experience with other Toxor-

640 JounNnr, oF THE Arurnrcar Moseurro CoNrRor, Assoctarrorl Vol. 7, No. 4

hynchites species has demonstrated that beforebeginning a control program using these mos-quitoes, it is necessary to know as much aspossible about their biological characteristics,especially those pertaining to oviposition.

From the spring through the 6ll of 1990, wereleased 20,20,40 and 80 gravid females (9-11days old) in a walnut grove at the edge of anurban area of Monterrey, Mexico. The adultsthat we used in these experiments were obtainedfrom a laboratory colony. For oviposition sites,we provided 10 used automobile tires, placedvertically at the base of tree trunks. in a rectan-gular pattern of 20 x 40 m. Oviposition wasrecorded daily for 17 days.

The average daily oviposition rate varied be-tween releases. The release of 80 females hadthe highest average oviposition rate (242.g)while that of 40 females was higher (97.5) thanthose (35.9 and 48.2) of 20 females. The ovipo-sition pattern was adjusted to the inverse .egre.-sion equation (Y : a + b/X), where y equalsthe number of eggs deposited per day and X isthe number of days post-release. We observedthis tendency in the fall release of 80 females.

In terms of positive tires, the release of 80females (69.4%) was equal to that of the releaseof 40 females ga.l%) and larger than that ofthe 2 releases of 20 females (48.2 and 4b.g%).We found a negative linear correlation betweerrpercent of positive tires and days post-release.This was quite notable for the releases of 20females, less so for the release of 40 and poorfor the release of 80.

The lower fall temperatures affected flightactivity of the released females and the ovipo-sition pattern of 80 females was very irreguiar,but they followed the increases and decreaies oftemperature within the range of 13-28"C (r, :0.81). The increased number of released femaleshad a tendency to find a greater number ofoviposition sites. Additional studies of the dis-persion and oviposition in larger urban areas arenecessary to determine the potential of this spe-cies as a control agent for Ae. aegypti.

Functional response of Toxorhynchiteetheoboldi on Aedes aegypti larvae (Res-

puesta funcional de Toxorhynchites theo-bald.i sobre larvae de Aedes aegypti)

A. Flores-Leal and F. Reyes-VillanuevaUniuersidad Autonorna de Nu.euo Leon, San Nicotas

de los Garza, N.L., Mexico

We studied the functional response of Toxor-hynchites theobaldi larvae on Aedes aegyptilar-vae. The variation in number of larval Ae. ae-gypti consumed by each larval predator was re-corded for the 4 larval predator stages and at 8larval prey densities of 10, 20, 30, 40, 50, 60, T0

and 80. Tests were conducted in an individualPetri dish and replicated 8 times. After 4g h ofexposure, we recorded the number of larvae con-sumed. The insectary had an ambient tempera-ture of 28"C, 48% relative humidity and 14:10light:dark cycle.

We plotted the number of larvae consumedby the prey species at different prey densitiesand tested the observed points with the chi-square test, to the linear model (Na: a + bNo)and the Holling II (Na = "Ita,ll * Tha,) whereNa equals prey consumed at the density No, anda and b are the intercept and the angle of thelinear equation. In Holling, Tt is the time ofpredator exposure (in hours), Th is handlinetime for the prey, and a' is the rate of instanltaneous attack.

The functional response was described by thelinear model for the second and fourth predatorinstars. The equations were Na : 5.609b +0.6149No and Na : 2.0632 + 4ZNo 0.8479No.respectively. Holling II describes the best func-tional response with the first and third instarswith equations Na + (0.02448) (48) No/1 +(0.02448) (0.47385) No and 56 : (0.024bb) (48)No/l + (0.02455) (0.4643) No, respectively.

Aedes aegypti larval control using cy-clops in Puerto Rico (Control larvario de

Aedes aegypti con ciclops en Puerto Rico)M. F. Suarez, M. M. Rivera, C. Aranda, G. G. Clark

and G. MartenProyecto Rochefeller y CDC, San Juan Laboratories.

San Juan, Puerto Rico

Cyclops are one of the commonest forms offreshwater zooplankton in the world. These tinycrustaceans have recently been recognized asexceptionally promising biological controlagents in some Aedes aegypti production siteswhere they attack and kill first-instar larvae.

We collected 18 cyclops species from theirnatural habitats in Puerto Rico and identifiedand cultured 3 good larvivorous candidates.Through laboratory bioassays, we determinedthat Mesocyclops aspericornis was the best can-didate for field trial applications, followed byMacrocyclops ahidus and Mesocyclops n.sp.

In April 1990, we added Me ospericornis and.Ma. albidus to 25 tires located outdoors undernatural conditions. About 50 individuals wereapplied per tire. They have survived, multipliedand maintain a constant population, while pre-venting the appearance of Ae. aegypti larvae. Inadjacent tires and during this same period, Ae.aegypti production continued unaffected.

In previous entomological surveys in a com-munity in Caguas, Puerto Rico, we identified b5gallon drums, discarded appliances and cisternsas the most common containers that would lend

DECEMBER 1991 Mosqurro VpcroR Svrraposruu 641

themselves to the application of cyclops. In ad-dition, large tire accumulations in public dumpswere identified as good candidates for cyclopsapplications.

Recently, we initiated field applications of Me.aspericornis in the above containers. After hav-ing gained the community's confidence duringprior project activities, residents accepted, with-out resistance, this larval control procedure andsome have asked for multiple applications.Along with the cyclops application, residentsreceived a pamphlet with basic informationabout these predators. The principal reactionwas one of general disbelief, since the residentsthought that the cyclops were too small to killAe. aegypti larvae. Preliminary results indicatethat the cyclops survive and continuously preyon larvae, unless the container becomes com-pletely dry.

Effectiveness, risks and costs of spatialULV applications with Malathion CE-96and Sumithion L-IOO against Aedes ae-

gppti in Panama (Eficacia, riesgo y costode las aplicaciones espaciales ULV con

Malati6n CE-96 y Sumition L-IOO contraAed.es aegypti en Panam6)

E. Adames and J. R. RoviraUniuersi.dad de Panami y Ministerio de Satu.d de

Panami, Panarnd

This study was conducted on the universitycampus and 3 districts of the metropolitan areaof Panam6 City, Panam6 to evaluate ULV ap-plications of malathion and Sumithion againstAedes aegypti. For evaluation we used bioassaycages with 10 bloodfed females, plastic jars withfifteen 3rd and 4th stage larvae in 50 ml of waterand silicone-coated slides. Each of these B eval-uation methods was located inside and adjacentto each, systematically selected house and leftfor a 2-h exposure period. Before and after theULV treatment, we determined the level ofserum cholinesterase (CHE) in operators andresidents. During the test, we recorded temper-ature, humidity and wind.

Malathion CE-96, applied at the rate of !27ml/min (103 ml/ha), produced an average mor-tality (recordedat24 h) of 56 and,82Vo in adultsand 6.8 and28.6% in larvae, respectively. Simi-larly, Sumithion L-100 applied at 122.b ml/min(94.2 ml/ha) produced an average mortality of56 and 83% in adults and 14.6 and 84.7% inIarvae, inside and outside the houses, respec-tively. When the average bioassay mortality wascompared at 2 and 24 h post-treatment, theresults obtained were 34 and. G8.GVo for mala-thion and 31 and 69.6% for Sumithion.

The levels of CHE before and after treatmentwere not significantly different in operators or

residents. The concentration of insecticide drop-lets was from24-33/cm2 for Malathion and from26 to 50 for Sumithion. The average diameter(DMV) fot 85% of the droplets was 11.5 micronsfor rnalathion and 12.0 microns for Sumithion.During the 8 field tests, weather conditions var-ied between 26.5-28.5"C, 60-76% relative hu-midity and windspeed from 2 to 10 km/hour.

The project included 8 ULV applications inan area of 100 ha, where we selected 115 houses.Personnel used in this trial included 3 techni-cians, 2 operators and 2 investigators to com-plete 160 h of work. The total cost for thevaluation of the test was $968, equivalent to$9.68/ha for insecticide, personnel, CHE analy-sis and fuel.

Measurement of aerial and ULV sprayingefficacy in Antioquia, Colombia (Medici6nde la eficacia de fumigaciones UBV a6rea

y con LECO en Antioquia,Colombia)

J. L Zapata, J. P. Escobar and M. A. RodriguezSeruici.o Seccionnl de Salud de Antinquia, Metro

Salud. and Seruicin de Erradi.caciin de tn Malnrin,Medellin, Colornbia

To prevent a dengue epidemic in 1989, theColombian government alerted the communitythrough different campaigns and took safetyprecautions. In the state of Antioquia, therewere at least 100 suspected and 3 confirmedcases of dengue hemorrhagic fever.

To reduce the adult mosquito population, ap-plications of malathion were made with vehicle-mounted LECO units andby aircraft. By LECO,the application rate was 500 cc/ha in 4 localitiesin Medellin. For each outbreak case, the oper-ating range covered was 200 meters. The samedosage was sprayed by aircraft in the munici-palities of Carepa and Chigorodo.

Prior to each application, SVo of the dwellingswere surveyed by collecting Aedes aegypti mos-quitoes for 20 min in each house. After sprayingwas completed, the procedure was repeated tocalculate the post-spray index, which was thencompared with the pre-spray index. The resultsfavored the LECO applications where the adultmosquito population was reduced by more than50% in the 4 study areas. The indices did notshow any important reduction with the aerialapplication.

The selective application showed good resultsfor its significant reduction ofadult population.The aerial application was highly ineffective,mainly because of the difficult weather condi-tions, the workers' inexperience and the type ofhouses. With its high costs, this operation canonly be recommended for special cases.

642 JounNnr, oF THE ArraauceN Moseurro CoNrRor, Assocre'rroN VoL. 7, No. 4

Utility of ultra-low volume application ofinsecticides for the control of Aed.es ae-

Erypfi during dengue and dengue hemor-rhagic fever epidemics (La utilidad de in-secticidas aplicados por medio de volumen

ultra-reducido (VUB) para el control deAed.ee aegypti durante epidemias de den-

gue y fiebre hemorrdgica de dengue).M. J. Nelson

Organizacion Panamericana de la Salu.d, Panarna,Panamd

In recent years there has been a controversyconcerning the usefulness of ultra-low volume(ULV) appllsation of insecticides for the controlof diseases transmitted by Aedes oegyptl. Nu-merous trials from the 1960s until the presenthave shown widely different results, varyingfrom no detectable control to virtually completeelimination of the adult vector population. Ofthe multitude of factors contributing to thesedifferences, the most important are apparentlydosage, insecticide susceptibility, house con-struction, block configuration and route of thespray vehicle.

One of the most important reasons for differ-ing results has been differences in dosages ofinsecticide applied. For example, in Puerto Ricowhere ULV application has resulted in little orno control, Iess than 73 ml/hectare of malathionwas applied. However, in most of Latin Americawhere better reductions have been observed, theusual target dosage is 305 ml/hectare. In bothcases, the same discharge rate (127 ml/min) wasemployed, but in Puerto Rico application ismade only once along each street and only onthose streets with the same compass orientation(i.e., only North to South and South to North,but not East to West), whereas in Latin Americathe vehicle circles each city block, resulting inpasses in both directions along every street.

What degree of reduction of the natural mos-quito population, and for how long is it necessaryto interrupt, or at least reduce transmission ofdengue or yellow fever? This is very difficult todetermine in the field because we are unable topredict what the transmission would have beenin the absence of the treatment, and transmis-sion in untreated comparison areas cannot beassumed to be comparable to the treated areas.However, we can design mathematical modelsbased on what we know or assume about thedynamics of disease transmission. Preliminarymodels show that ULV application alone willhave very little effect on the total number ofcases of dengue because the adult mosquito pop-ulations recover very rapidly to previous densi-ties, but it will decrease the rate of transmission,spreading it over a longer period of time. Ifelimination of breeding sites (source reduction)

were done instead of ULV application, therewould be a reduction in total number of casesbecause of the longer lasting reduction of mos-quito densities, but the implementation wouldbe so slow that the epidemic would be overbefore a great reduction could be achieved. Ifthe 2 methods were used together, they wouldhave a synergistic effect, resulting in greaterreduction in transmission than the sum of thereductions when they were each used alone,ULV application delaying the peak of the epi-demic thereby giving more time for the slower

;:::* of source reduction to reduce number of

Architecture, ventilation, Aedes aegy ptiand ULV (Arquitectura, ventilaci6n,

Aed.es aegypti V VUB)P. Reiter

San Juan Laboratories, Centers for Disease ControL,San Juan, Pwrto Rico

Aedes aegyptj is an indoor species that favorsheavily sheltered sites when it is not feeding orovipositing. Vehicle-dispensed and aerial ULVapplications for Ae. aegypti control are thereforeessentially drift spraying operations from thestreet or the sky into the closet. To understandthis drift, we must consider airflow into andaround buildings. The mechanics of such airfloware complex, but predictable. Discussion in thispaper is limited to the more important principlesrelevant to single story houses.

Air is deflected over and around buildings,creating calm zones ofpressure differential, pos-itive on the windward sides and negative toleeward. Air flows from inlets in positive areasto outlets in areas of lower pressure. The largerthe negative calm area, the greater the potentialfor indoor flow. The proportions of a structuredetermine the form of the air movement. Forexample, behind a house of equal height, widthand depth, the depth of the calm area is twiceits height. The orientation of the house affectsair movement by the same basic parameters:height, overall width to windward and depth.Because of the frictional drag of the roof, a thinstructure, such as a wall, provides a larger calmarea than a house ofthe same height and width,and the deeper the house, the less deep is thecalm area. A wide house has a wider, deepercalm area than a narrow one, but roof-heightstill determines its height. Roof slope and pitchgreatly increase the height and depth ofthe calmarea. Buildings on stilts channel airflow underthe floor as well as over the roof. Buildings inlines may protect downwind buildings from po-tential airflow, and tend to channel airflow alongthe streets.

DECEMBER 1991 Moseurro VEcroR Sylr,tposlurvr 643

Air has momentum-it will circulate in theinterior of a house if the inflow is oblique to thealignment of the inlet/outlet, but will passthrough in a straight, narrow stream if exteriorand interior flow are in line. Velocitv of air isdetermined by the inlet/outlet ratio. il4aximuminterior airflow occurs when the outlet is largerthan inlet, because negative pressure is estab-lished inside the building. If inlets are Iargerthan outlets, positive pressure is established in-doors, so more airflow is deflected around thehouse. Vegetation, window forms, insect screens,overhangs, awnings and many other structuresare also important.

In view of this complexity we can expect greatvariation in the potential efficacy of outdoorULV applications. We certainly cannot expectcomparable results against Ae. aegypti in theCaribbean as we do with an exophilic speciessuch as Ae. taeniorhynchus in a typical mosquitoabatement district in the USA. Moreover, theenormous differences in architecture and streetplans, between countries and even within cities,means that the impact on Ae. aegypti in a villageof lattice-walled wooden houses on stilts inSouth East Asia may be very different from thatin a cinder-block and concrete urbanization inSouth America. For these reasons, the efficacvof ULV in any area can only be known witlrcertainty if careful evaluations are made in situ.

Dengue activity in Puerto Rico in 1990(Actividad de dengue en Puerto Rico en

1990)G. G. Clark, A. V. Vorndam, I. Gomez and E. Versne

San Juan Laboratories, Centers for Disease Control,San Juan, Puerto Rico

During the last 3 decades, Puerto Rico hasexperienced 11 epidemics ofdengue and, begin-ning in 1986, from 6,000 to 11,000 cases ofdengue-like illness have been reported annuallv.During the last 5 years, three slrotypes (DEN-1, DEN-2 and DEN-4) have been isolated. Theidentification of these serotypes and serologicdocumentation of dengue virus transmission hisbeen the result of a cooperative, laboratory-based surveillance program between the puerioRico Department of Health and the Centers forDisease Control. During 1990, 2,660 cases werereported to the Dengue Branch, of which 1,225were confirmed by laboratory testing.

Although there is transmission throughoutthe year, the annual seasonal increase in denzueactivity in Puerto Rico normally begins in Au-gust, usually peaking in October or November,and diminishing sharply by January or Febru-ary. Increased dengue activity in 1990 began inAugust and was first recognized in the city ofLares in the western mountains. The outbreak

was first thought to be caused by measles. Fromthere, significant dengue activity was observedin the Ponce region on the southern coast withcases eventually confirmed from all municipali-ties. Highest incidence rates were seen on thesouthern coast, in several mountain municipal-ities and along the eastern end ofthe island.

Dengue-2 was the most frequently (72%) iso-lated serotype, followed by DEN-4 (21%), and,was found throughout much of the island. It wasisolated during every month of the year andfrom all municipalities where virus isolates wereobtained. All 3 serotypes were isolated from 4municipalities.

The highest percentage (36%) of laboratoryconfirmed cases and highest incidence (0.65/1,000 population) were in the age-group lessthan 15 years of age. Severe cases of denguecontinue to be detected in Puerto Rico. This wasreflected in 1990 by the number of cases tharwere hospitalized (150/L,000 confirmed cases)and the substantial number of cases with hem-orrhagic manifestations (195/1,000 confirmedcases).

In 1990, we serologically identified a presump-tive case of dengue with a fatal outcome. Thecase involved a 18 year-old female who residecin a locality west of the San Juan metropolitanarea. This was the first such case since 1986when 3 fatal cases were attributed to dengue/dengue hemorrhagic fever.

Socioeconomic impact of an outbreak ofdengue in Lares, Puerto Rico (Impacto so-

cioecon6mico de un brote de dengue enLares, Puerto Rico)

M. I. Torres and G. G. ClarkUniuersity of Connecticut, Storrs and Centers for Dis-

eose Control, San Juan, Puerto Rico

We present preliminary findings obtainedfrom a household survey conducted after anoutbreak of dengue in Lares, Puerto Rico. Datawere collected to document both the geographi-cal distribution and the socioeconomic imoactof the outbreak.

A random sample of households was selectedbased on the total number of occupied housingunits. The number of household membersranged between 2 and 11 (i : 4). A total of 417individuals were included and only female headsof households were interviewed. Of those sur-veyed, 237o reported at least one denzue casewithin the family. Among these 22 households,there were 56 suspected cases reported (lB% otthe 41,7 persons surveyed).

Symptomatology was consistent with the clin-ical picture of classical dengue. Children (n :19) represented 55% of the cases and 68% ofthose who visited a doctor during the outbreak.

644 JouRNu, op rnn AuonrcAN Mosourro CoNTRoL Assocrnrrou VoL. 7, No. 4

The urban area produced 67% of the cases.Communities characterized by a low socioeco-nomic status reported the highest number ofcases.

Documentation on direct cost of medical careindicated that while the outbreak did not pro-duce serious cases requiring secondary or ter-tiary level of care, primary care was sought atleast once by the majority of households whoreported cases (86%). The local health centerwas the main source (66%) of this care. Medicaidwas the health insurance most frequently men-tioned as a source of payment for medical treat-ment. Other direct costs for transportation, pre-scription and over-the-counter medicines, andlaboratories were not significantly high.

Indirect costs were also documented. Sur-veyed households reported 300 sick-days amongthe cases with clinical dengue, an average of 5sick-days per person. Students missed 9 school-days, an average of 3 days per student. Althoughmost of the adult cases were in housewives andunemployed males, employed persons lost 38work-days, an average of 4 days per employee.Lost income per day ranged from $29 to $84.Housework was negatively impacted by the out-break. The vast majority (9I%) of the femalehomemakers interviewed reported not beingable to complete all their housework normallyperformed during the days when they or theirfamily members were sick with dengue.

Psychological costs were also documented.The majority (86%) of interviewed female care-takers reported a high level of stress related tolack of information about the illness and logis-tics (e.g., transportation to the health provider)as well as emotional stress (e.g., feelings of anx-iety and nervousness) while they or their familymembers were sick with dengue.

Experience of community control of den-gue in llonduras and future perspectives(Experiencias de control comunitario de

dengue en llonduras y perspectivasfuturas)

E. Fernindex and R. SotoProyecto Rochefellcr, Ministerio d.e Salud de Hondu-

ras, Tegucrgalpa, Honduras

Dengue fever is endemic in Honduras wherethe vector, Aedes aegypti was reintroduced inthe 1960s. In 1977 , the dengue virus was presentin Honduras, and during 1978, the first epidemicoccurred in the northern coast when about100,000 cases were reported nationwide. Severaloutbreaks have occurred in subsequent years.Vector control methods such as the ultra-lowvolume application of insecticides against theadult mosquitoes or biological control are not

adequate to deal with the problem when theyare used alone.

Since Ae. aegypti is a domestic mosquito andits breeding sites are in and around humandwellings, human attitude and behavior regard-ing real and potential breeding sites can be partof the problem. Since they can be part of thesolution, community participation in vector con-trol programs is being emphasized.

During 1989 and 1990, a baseline study wasconducted in EI Progreso (population 70,000),Honduras. Epidemiological, entomological andsocial science data were gathered and analyzed.A small-scale educational program was devel-oped which showed that knowledge and attitudetoward Ae. aegypti breeding sites can be posi-tively modified to improve vector control.

During 1991, community participation activi-ties for Ae. aegypti control will be expanded to10 communities in El Progreso (previous phaseincluded only four) and with a wider scope ofwork: 1) community organization/mobilization,2) school involvement, and 3) development ofeducational materials and programs.

After evaluating the success of the first phase,Ministry of Health officers have been requestedto extend the benefits of this experience to otherurban centers in the country as well as sharingexperiences with health personnel in the othersanitary regions of Honduras, seeking to adaptmethodology to their local vector control pro-gTams.

Simultaneously, a biological control researchcomponent will be initiated in EI Progreso tobegin application of MesocycLops (crustaceans)as a complementary Aedes larval control methodat the community level. Later, the applied bio-logical control technology will be shared withcommunity residents to support their efforts, aswell as to maximize vector reduction.

The project is a result of an agreement forcooperation between the Honduran Ministry ofHealth and the Rockefeller Foundation in Aedesaegypti control.

The dengue control program in Mexicoand perspectives for the future (El pro-grama de control de dengue en Mexico y

perspectivas para el futuro)H. G6mez

Proyecto Rochefeller, Secretarin de Salud, Mdxico, D.F., Mexico

The recent history of dengue transmission inMexico started in 1978 along the southern bor-der with Guatemala. Since then, dengrre hasspread in epidemic fashion and is now endemicin 29 of 32 states in the country. The totalnumber of cases reported through 1990 was227,229, and 867o of the cases were found in

DECEMBER 1991 MoseulTo VEcroR Svl'lPosIuu 645

only 13 states. After the eradication of Aedesaegypti in 1963, efforts to keep the country freeof the vector were fruitless. The first dissemi-nation of the infection and the recent transmis-sion of dengue in high attitude areas demon-strated how difficult the control has been.

During the 1980s, the control program wasdependent on the use of chemical tools for in-secticide spraying and larval control. As in manyother Latin American countries, the programfaced problems and Iimitations in budget,trained human resources and equipment. Thesewere not new but became more severe as trans-mission in different regions and the risk of den-gue hemorrhagic fever increased. Since the so-Iution to vector-borne diseases was traditionallywith the government, the community perceivedthat the solution for dengue or malaria wasoutside its domain. The present situation hall-marks a point where the feasibility of investingin a technically based solution where specializedpersonnel, spraying machines and insecticide orlarvicide use is out-of-hand. This is also becausethe vector is a domestic mosquito that breeds inman-made containers in the domestic setting,generated by specific human behavior and pat-terns of consumption. There is a need to createsocial awareness and recognize our participationand responsibility as a community, in generatinga solution to the problem.

The program in Mexico is beginning to de-velop educational tools where the community isactively involved in their development and con-tent. Different approaches to control are.alsobeing tested where breeding sites are controlledby routine activities like trash recycling schemesand biological control strategies, which will bedeveloped and implemented.

Prevention strategies for dengue epidem-ics in Panama (Estrategias en la preven-cion de epidemias de dengue en Panamri)

B. E. Dutary, P. H. Peralta, G. de Jaen, M. A.V6squez and E. Quiroz

Gorgas Mernorinl Laboratory y Ministerin de Salu'd'Panama, Panamd

Many of the dengue epidemics in several LatinAmerican countries since the 1970s could havebeen controlled by integrated health activities.The specific activities to be employed in suchstrategies include: 1) education of health person-nel, 2) community education, 3) monitoring offebrile illnesses. 4) use of sentinel centers foractive surveillance, and 5) preparation for aneventual outbreak. The case of Panama will beexamined to illustrate how these activities arecoordinated to produce reduced larval indices,monitor dengue virus activity and prepare foran outbreak.

The absence of dengue virus in Panama issurprising, given the fact that Panama was rein-fested with Aedes aegyptl in August 1985 (theoverall house index has exceeded 5% every rainyseason since 1986) and Panama's population ishighly susceptible to dengue, since the last out-break ended in 1942. In an effort to combat thepossibility of a dengue virus epidemic, GorgasMemorial Laboratory, the Ministry of Health,social security system, PAHO and communityleaders have joined efforts in 1990 to implementa program to prevent epidemic dengue. Thisprogram included all 5 strategies.

Larval indices have shown a downward trendwithout the use of chemical control. Laboratorytests on 1,780 suspected cases ofdengue studiedsince September 1988, employing virus isolation,IFAT, IgM-ELISA, HI and NT, indicated nodengue virus activity in Panama, except for 2imported cases detected through the establishedsurveillance system.

This is the first example in which a dengue-free country with Ae. aegypti reinfestation hasmade a concerted effort to reduce larval indices,while simultaneously monitoring dengue virusactivity and preparing for an eventual outbreakof dengue.