During the summer 1994 outbreak of epidemic polyarthritis in suburban Brisbane, 29,931 adult female mosquitoes were collected by octenol-CO2 light traps and tested for virus by species in pools of approximately 20 using an in situ enzyme-linked immunoassay. Overall, 63 isolations of Ross River (RR) virus were made from 7 different mosquito species, including 23 from freshwater-breeding Culex annulirostris Skuse, 13 from peridomestic Aedes notoscriptus (Skuse), 4 from Aedes procax (Skuse), 12 from the brackish water-breeding Aedes funereus (Theobald), 9 from saltmarsh Aedes vigilax (Skuse), and 1 each from Culex sitiens Wiedemann and Aedes alternans (Westwood). The RR virus minimum infection rate in mosquitoes ranged from 1.6 to 2.5/1,000 from March to June 1994. This study implicates freshwater and brackish water mosquitoes as important suburban vectors of RR virus and indicates the need for refocusing mosquito control priorities.

Previous studies have shown that antibodies produced against strategic flavivirus epitopes play an important role in recovery and immunity. Definition of the conformation and location of these epitopes and the degree of their conservation among flaviviruses is important to understanding the humoral response to flavivirus infection. In this study we have examined epitopes recognized by 14 monoclonal antibodies (MAbs) produced to the envelope (E) and nonstructural (NS1) proteins of Murray Valley encephalitis virus (MVE). These antibodies were analysed for specificity, neutralization, haemagglutination inhibition (HI) and competitive binding. We have identified six distinct epitopes on the E protein which are located in four non-overlapping domains. MAbs to epitopes in one domain neutralized virus, were specific for MVE and Japanese encephalitis virus, and reacted with epitopes resistant to reduction. Two other E domains, one specific to MVE and the other shared by all flaviviruses, also contained neutralization sites and were stabilized by disulphide bonds. The fourth domain on E was conserved among the flaviviruses, sensitive to SDS denaturation and did not induce neutralizing antibody. Studies with MVE NS 1 MAbs revealed that they were mostly type-specific, unreactive with conserved epitopes, and unreactive in HI and neutralization tests. The six epitopes identified on NS 1 did not overlap and represent antigenic domains either resistant or sensitive to reduction. Immunoblotting of viral proteins in MVE-infected C6/36 cells revealed two distinct forms of NS1 and high Mr proteins of 97K and 108K that represented disulphide-linked heterodimers of E and NS1.

Eleven weanling horses were inoculated with Murray Valley encephalitis and Ross River viruses either by intravenous injection or by the bite of Culex annulirostris or Aedes vigilax mosquitoes infected orally. Five of the 11 horses circulated trace amounts of MVE virus for 1 to 5d and they infected 7/408 Cx annulirostris which subsequently fed on them. Haemagglutination-inhibiting antibody persisted at detectable levels for the 24-week observation period. With Ross River virus, only one of 11 horses inoculated developed a viraemia detectable by inoculation of suckling mice but 5 horses contained virus sufficient to infect 41/383 Cx annulirostris that fed on them 3 to 4 days after inoculation. On primary inoculation with Ross River virus, only 2 horses developed HI antibody but late responses occurred in 3 horses following probable naturally acquired re-infections. With both viruses, most horses remained normal, some developed mild pyrexia and transient clinical signs. This paper, therefore, indicates that horses are unlikely to be efficient amplifiers of either MVE or RR viruses and does little to incriminate them as important pathogens.

Blood meals of 7552 mosquitoes of 13 species were identified from sylvan and urban areas of Charleville, southwestern Queensland, Australia, between 1976 and 1979. Over 80% of the meals identified were from Culex annulirostris, the major vector of Murray Valley encephalitis in Australia. The feeding patterns of Cx. annulirostris snowed remarkable consistency between years, but the relative abundance of available hosts within and outside Charleville played a major role in influencing host selection. In sylvan sites mammalian feeding ranged from 85 to 93%, whereas in town sites where chickens were common this was reduced to between 57 and 67%. Culex quinquefasciatus and Cx. australicus fed predominantly on birds (domestic poultry and passerines, respectively). Anopheles annulipes, An. amictus, and 6 Aedes spp. fed mainly on mammals, whereas Cx. (Lophoceraomyia) cylindricus fed predominantly on poikilotherms. These host-feeding data are in keeping with other studies in Australia suggesting that Cx. annulirostris prefers to feed on mammalian rather than avian blood if it is available.

Summary Australian populations of Culex annulirostris Skuse from 10 localities, Brisbane, Townsville, Cairns, Charleville, Kowanyama (Queensland), Darwin (Northern Territory), Mildura (Victoria) and Port Hedland, Karratha and Mt. Tom Price (Western Australia) showed considerable heterogeneity in response to oral infection with low passage level Murray Valley encephalitis (MVE) and Kunjin (KUN) viruses. However, there was no consistent pattern of either high or low vector efficiency for populations collected from different regions. Overall, Cx annulirostris was a more competent vector of MVE virus than KUN virus, both in oral susceptibility (suckling mouse intracerebral ID50 of 101.7‐103.9 compared to 102.7‐104.8/mosquito, respectively) and in ability to transmit virus to chicks after 10 d extrinsic incubation (50–89% compared to 0–55%). High or low susceptibility with one virus did not correlate with susceptibility for the other. Although these data were probably subject to some seasonal variation, they generally support the view that Cx annulirostris is the principal vector of MVE in Australia. Poor to moderate susceptibility with KUN suggests that (1) it is only a minor vector in some localities, (2) vertebrate hosts have high viraemias or (3) the strain used may have been atypical.