374 J. Au. Mosg. CoNrnol Assoc. Vor-.I, No.3

the larvae and pupae were being collected. Two Restifo,R. A. and G. C. Lanzaro. 1980.The occur- of these adults, both female, were collected, rence of Aedesatropalpus (Coquillett) breeding in along with 35 larvae and pupae. The larvae and tires in Ohio and Indiana. Mosq. News 40:292-294. pupae were reared for identification as adults. White, D. J. and C. P. White. 1980.Aedts atropalpus The, mosquitoes were maintained through one breeding in artificial containers in Suffolk County, New York. Mosq.News 40:106-107. additional generation in a large masonjar, indi- Zavortink, T. 1972. Mosquito studies (Diptera: cating stenogamy. The colony proved to be au- J. Culicidae).XXVIII. The New World speciesfor- togenous; viable eggs were produced without merly placed in Aedcs(Finlaya). Contrib. Am. En- blood,. restuans Theobald larvae were also tomol. Inst. (Ann Arbor) 8(3):l-206. collected from the rock holes. The rock holes were located on the surface of a granite outcrop about 8 m from the shoreline and elevated about 5 m above the water level. The largest of the three pools was 3 m long by 2 m across at the time of collection. The maximum depth of this pool was about 20 cm. The other pools were much smaller. The most CULEX THEILERI AND SINDBIS VIRUS; prominent types of vegetation were grassesand SALIVARY GLANDS INFECTION IN moss growing along the edge of the largest pool RELATION TO TRANSMISSION and in cracks in the granite. All of the pools P. c. JUPP contained leaf litter and other organic material. Arbovirus Research Unit, National Institute for Recently, numerous reports of Ae. atropalpus Virology, Private Bag X4, Sandringham 2131, utilizing discarded tires have appeared in the Johannesburg, , and Dept. of Virology, literature (Covell and Brownell lg7g, Restifo University of the Witwatersrand, Johannesburg, and Lanzaro 1980, White and White 1980). The South Africa. tire breeding habit brings this species into close proximity of human habitations and allows for The competence of a mosquito species as a large numbers to develop in a small ared. Fur- vector of an arbovirus depends upon both ex- thermore, tire breeding strains are avid blood trinsic and intrinsic factors (Reeves 1967. World feeders even though autogeny is maintained. Health Organization 1972, Hardy et al. 1983). Aedzsatropalpzs has been shown to utilize a wide Extrinsic factors include, among others, the rel- range of host species,including humans, under ative abundance and biology of the mosquito natural conditions in a tireyard (Berry and vector and its vertebrate host as well as envi- Craig 1984). The demonsrration of the abil- ronmental conditions. The latter affect not only ity of Ae. atropalpus to transmit La Crosse en- the bionomics of both invertebrate and verteb- cephalitis virus orally and transovarially (Freier rate hosts of the virus but also the duration of and Beier 1984) indicates rhar this species the extrinsic incubation period of the virus in merits further study, especially with regard to the mosquito. The main criteria pertinent to a the tire breeding strains. quantitative laboratory assessment of vector Voucher specimens of adults discussed in this competence are the susceptibility of the insect report have been deposited in the Iowa State to infection, the length of the extrinsic incuba- Insect Collection at Iowa State Universitv. tion period and the transmission rate. The transmission rate is defined in our laboratory as the proportion of infected mosquitoes which RderencesCited successfully transmit the virus. In transmission tests to determine this rate we expose individual Beadle,L. D. 1963.The mosquitoesof Isle Royale, susceptible animals to individual infected mos- Michigan. Proc. New JerseyMosq. Exterm. Assoc. 50:133-139. quitoes, the choice of animal depending on the virus in question. Berry, W. J. and G. B. Craig,Jr. 1984.Bionomics of However, tests with large Aedesatropalpus breeding in scrap tires in northern numbers of animals have the disadvantage of Indiana. Mosq. News 44:476-484. being expensive, cumbersome and time con- Cassani,J.R. and H. D. Newson.1980. An annotated suming. Furthermore, certain species of mos- list of mosquitoesreported from Michigan. Mosq. quitoes may be reluctant to take a second News 40:356-368. blood-meal, notably aedine species in the sub- Covell, C. and A. Brownell. J., Jr. J. 1979.Aedes at- genera Ochlerotatus, Neomelaniconion and ropalpusin abandoned tires in County, Jefferson Aedimorphus Kentucky. Mosq. News 39:142. Qupp, unpublished work). Alter- native ways for determining the transmission Freier, J. E. and J. C. Beier. 1984. Oral and transovarial transmissionof La Crosse virus by rate are the in aitro capillary method of Aitken Aedcsatropalpus. Am.J. Trop. Med. Hyg. 33:70&- (1977) or testing the salivary glands of each 714. mosquito for infectivity (Chamberlain 1968). SnprrN4srn.1985 J. Av. Mosq. CorvrRolAssoc. 375

According to Chamberlain (1968), the presence then triturated with small needles in a second of virus in the salivary glands of mosquitoes drop of saline. The suspension was then trans- indicates the probable ability of the species to ferred to a tube containing 0.3 ml of bovine transmit by bite. The experiment reported phosphate albumin. This salivary gland suspen- below was undertaken to investigate whether sion and another prepared from the decapi- individual mosquitoes with infected salivary tated body of each mosquito were tested sepa- glands always transmitted virus by bite. rately for virus by inoculating them intracereb- The mosquitoes used were aged 2-18 days, rally into 2-3-day-old mice. the 3rd generation of a laboratory colony of The decapitated bodies of all 23 mosquitoes Culex theileri Theobald which originated from tested were found to be infected but only t had gravid females collected at Lake Chrissie in the infected salivary glands. Only 2 of these 9 mos- eastern Transvaal highveld. The experiment quitoes transmitted virus to the chicks. Thus it was undertaken in an insectary where the mos- appears in the case of Cx. theileri with Sindbis quitoes were maintained at 75-85Vo RH and virus that, although salivary glands are infected, 24-26'C. They were infected by feeding on a transmission will not necessarily follow when blood-Sindbis virus mixture through a chicken the mosquito bites the susceptible host. Hence skin membrane. The virus used was the AR86 the salivary gland infection rate cannot be used strain of Sindbis virus (Weinbren et al. 1956) at as a measure of the transmitting ability. This its 4th intracerebral mouse passage level. result is shown in Table I together with the Methods for preparation and titration of the results obtained previously by other workers infective blood-meal, feeding the mosquitoes, using other virus-mosquito vector combina- and determining the infection rate were the tions. It can be seen that a good agreement same as described previously (fupp and McIn- between salivary gland infectivity and transmis- tosh 1970) except that the infecting titer of 7.2 sion by bite (transmission rate) has usually been logs is expressed as the log mouse L\o per ml. demonstrated in experiments, with the notable Fourteen days after their infective feed the exception of California encephalitis virus and mosquitoes were allowed to feed singly on indi- (Linn.). vidual 2-week-old chicks. Three weeks after this It can be concluded from the present ex- transmission feed the chicks were bled and their periment and the results of others, that al- sera tested for antibody against Sindbis virus by though salivary glands infectivity would often the haemagglutination inhibition test (Clarke be a reliable expression of the transmission rate, and Casals 1958). The presence of antibody this is not always the case. Why are some mos- indicated successful transmission. On the day quitoes unable to transmit virus while feeding following the transmission feed those mos- even though infectious virus or viral antigens quitoes which had fed were frozen at -70'C can be detected in the salivary glands? Hardy et and 12 days later were removed from storage al. (1983) suggested 3 possible explanations: (l) and allowed to thaw at room temperature. The Viral infection is at an early stage so that little or salivary glands were removed from the mos- no infectious virus has been released into the quito, rinsed in a drop of normal saline, and salivary glands, (2) the mosquito could have re-

Table l. Experiments comparing salivary gland infection rates and transmission rates.

Salivary gland Transmission Virus Mosquito vector infection rate rate Reference Alplwvinu Sindbis Culex theileri 91231 2123' Present report Whataroa Aedzs australis 6/10 6/10 Miles et al. 1973 Western equine encephalitis Culex tarsalis 4/r0 3/103 Thomas 1963 Western equine encephalitis Culex tarsalis t7n7 17l17 Kramer l98l Flaaivinu Dengue2 Aefusaegypti ll/lt I l/l I Mcleanetal.1974(TableS) Bunyaairus Californiaencephalitis Aed.esaegypti 16t23 4122 Mclean et al. 1974 (Table l) r Numerator:number of mosquitoes with salivary glands infected; denominator:number tested. 2 Numerator = number of mosquitoes transmitting by bite, denominator = number of infected mosquitoes feeding. 3 This rate could have been 2/ l0 but actual result unclear as groups of more than one mosquito fed on chicks in the transmission attempts (Thomas 1963). 376 J. Aer. Mosq. CoNrnol Assoc. Vol. l, No. 3

cently ejected the virus while feeding on a car- lated from culicine mosquitoes in the Union of bohydrate source and the virus has yet to be South Africa. S. Afr. Med. J. 30:631-636. replenished into the salivary duct, and (3) W.H.O. 1972. Vector Ecology. Tech. Rep. Ser. "salivary there might be a gland escape bar- 50 I : l-38. rier" which prevents virus leaving these glands. This would be the situation when the produc- tion of infectious virus has been modulated to low levels. Such an escape barrier could explain TOXORHYNCHITES MOCTEZUMA, A the lowered transmission rates which have beerr POTENTIAL BIOLOGICAL CONTROL obtained after prolonged incubation periods AGENT IN TRINIDAD AND (Mclntosh and 1970, Mangiafico l97l). Jupp TOBAGO, W. I.1 I wish to thank the Director-General of National Health and Welfare, Pretoria. for DAVED. CHADEE' permission to publish. Ministry of Health and Environment, Insect Vector Control Division, f3 Queen Street, St. Joseph, Trinidad, West Indies

ReferencesCited Toxorhynchitesmoctezuma (Dyar and Knab) fas Aitken, T. H. G. 1977. An in vitro feeding technique theobaldi (Dyar and Knab) in Knight and Stone for artificially demonstrating virus transmission by (1977)l has a wide distribution from in mosquitoes. Mosq. News 37:l3G- 133. Central America to in South Chamberlain, R. W. 1968. Arboviruses, the America (Rubio et al. 1980; S. J. Heinemann, arthropod-borne animal viruses. Current Topics personal communication). In the Caribbean Microbiol. Immunol. 42:3U58. Basin, it has a limited distribution and is found Clarke, D. H. and Casals. 1958. Techniques J. for only in (Heinemann et al. haemagglutination and haemagglutination- 1980). It should be noted that Tx. theobaldi is inhibition with arthropod-borne viruses. Amer. J. now considered Trop. Med. Hyg. 7:561-573. a synonym of Tx. moctezuma (S. Heinemann, personal Hardy, J. L., E. J. Houk, L. D. Kramer and W. C. J. communication). Reeves, 1983. Intrinsic factors affecting vector In Trinidad and Tobago, threeToxorhynchites competence of mosquitoes for arboviruses. Annu. species have been recorded. Toxorbnchites moc- Rev. Entomol. 28:229-62. tezumn, Tx. iris (Dyar) [as mariae (Bowrroul) in J,rpp, P. G. and B. M. Mclntosh. 1970. Quantitative Knight and Stone (1977)l andTx. superhn (Dyar experiments on the vector capability of Culex '\Niedemann and Knab), of which Tx. iris andTx. superbus are (Culex) pipiens fatigans with West Nile found primarily in terresrrial and epiphytic and Sindbis viruses. Med. Entomol. 7:353-356. J. bromefiads (Heinemann et al. 1980). Toxorhyn- Kramer, L. D., L. Hardy, S. B. Presser and E. J. J. chites moctezuma, Houk. 1981. Dissemination barriers for Western on the other hand, has been equine encephalomyelitis virus in Culex tarsalis in- collected from bromeliads, treeholes. in tires and cans (Heinemann fected after ingestion of low viral doses. Am. J. et al. 1980, Chadee et al. Trop. Med. Hyg. 30:190-197. 1984). Mangiafico, J. A. 1971. Chikungunya virus infecrion In 1976, rhe Insect Vecror Control Division and transmission in five speciesof mosquito. Am. J. (IVCD) of the Ministry of Health and Envi- Trop. Med. Hyg. 20:642-45. ronment embarked on an Aedcs aegpti (Lin. Mclntosh, B. M. and P. G. Jupp. 1970. Attempts to naeus) Eradication program in Trinidad, W.I. transmit chikungunya virus with 6 species of mos- Tobago, however, is relatively free of Ae. aegypti quito. J. Med. Entomol. 7:615-618. but routine vigilance Mclean, D. M., A. M. Clarke, et al. 1974. Vector surveys are still conducted capability of Aed.esaegyph for California encephalitis twice yearly (Chadee et al. 1984). During sur- veys, larvae and dengue viruses at various temperatures. Can. J. and pupae of Tx. moctezumawere Microbiol. 20:255-262. collected in artificial containers in both Miles, J. A. R., J. S. Pillai and T. Maguire. 1973. Trinidad and Tobago. All immature stages Multiplication of Whataroa virus in mosquitoes. J. were collected and transported to the IVCD Med. Entomol. l0: 176- 185. laboratory, St. Joseph, Trinidad, where they Reeves, W. C. 1967. Factors that influence the proba- were subsequently identified by the author. bility of epidemics of Western equine, St. Louis and Table I summarizes the total collection of Tx. California encephalitis in California. Vector Views l4: I 3- 18. moctezunutfrom artificial containers in Trinidad Thomas, L. A. 1963. Distribution of the virus of t Western equine encephalomyelitis in the mosquito Published with approval from the Ministry of vector Culetctarsalis. Am. J. Hyg. 78:150-165. Health and Environment, Trinidad and Tobago, W.I. Weinbren, M. P., R. H. Kokenot and K. C. 2 Present Address: Department of Biological Sci- Smithburn. 1956. Strains of Sindbis-like virus iso- ences,The University. Dundee. DDI 4HN. UK.