Australian Mosquitoes Research Articles

Assessment of Local Mosquito Species Incriminates Aedes aegypti as the Potential Vector of Zika Virus in Australia.

BackgroundWithin the last 10 years Zika virus (ZIKV) has caused unprecedented epidemics of human disease in the nations and territories of the western Pacific and South America, and continues to escalate in both endemic and non-endemic regions. We evaluated the vector competence of Australian mosquitoes for ZIKV to assess their potential role in virus transmission.Methodology/Principal FindingsMosquitoes were exposed to infectious blood meals containing the prototype African ZIKV strain. After 14 days incubation at 28°C and high relative humidity, infection, dissemination and transmission rates were assessed. Infection in Culex annulirostris and Cx. sitiens could not be detected. 8% of Cx. quinquefasciatus were infected, but the virus did not disseminate in this species. Despite having infection rates > 50%, Aedes notoscriptus and Ae. vigilax did not transmit ZIKV. In contrast, Ae. aegypti had infection and transmission rates of 57% and 27%, respectively. In susceptibility trials, the virus dose required to infect 50% (ID50) of Ae. aegypti was106.4 tissue culture infectious dose50 (TCID50)/mL. Additionally, a threshold viral load within the mosquito of at least 105.1 TCID50 equivalents/mL had to be reached before virus transmission occurred.Conclusions/SignificanceWe confirmed Ae. aegypti to be the most likely mosquito vector of ZIKV in Australia, although the restricted distribution of this species will limit the receptive zone to northern Queensland where this species occurs. Importantly, the role in ZIKV transmission of Culex and other Aedes spp. tested will be negligible. Despite being the implicated vector, the relatively high ID50 and need for a high titer disseminated infection in Ae. aegypti suggest that high mosquito population densities will be required to facilitate epidemic ZIKV transmission among the currently immunologically naïve human population in Australia.

The insect-specific Palm Creek virus modulates West Nile virus infection in and transmission by Australian mosquitoes.

BackgroundInsect-specific viruses do not replicate in vertebrate cells, but persist in mosquito populations and are highly prevalent in nature. These viruses may naturally regulate the transmission of pathogenic vertebrate-infecting arboviruses in co-infected mosquitoes. Following the isolation of the first Australian insect-specific flavivirus (ISF), Palm Creek virus (PCV), we investigated routes of infection and transmission of this virus in key Australian arbovirus vectors and its impact on replication and transmission of West Nile virus (WNV).MethodsCulex annulirostris, Aedes aegypti and Aedes vigilax were exposed to PCV, and infection, replication and transmission rates in individual mosquitoes determined. To test whether the virus could be transmitted vertically, progeny reared from eggs oviposited by PCV-inoculated Cx. annulirostris were analysed for the presence of PCV. To assess whether prior infection of mosquitoes with PCV could also suppress the transmission of pathogenic flaviviruses, PCV positive or negative Cx. annulirostris were subsequently exposed to WNV.ResultsNo PCV-infected Cx. annulirostris were detected 16 days after feeding on an infectious blood meal. However, when intrathoracically inoculated with PCV, Cx. annulirostris infection rates were 100 %. Similar rates of infection were observed in Ae. aegypti (100 %) and Ae. vigilax (95 %). Notably, PCV was not detected in any saliva expectorates collected from any of these species. PCV was not detected in 1038 progeny reared from 59 PCV-infected Cx. annulirostris. After feeding on a blood meal containing 107 infectious units of WNV, significantly fewer PCV-infected Cx. annulirostris were infected or transmitted WNV compared to PCV negative mosquitoes. Immunohistochemistry revealed that PCV localized in the midgut epithelial cells, which are the first site of infection with WNV.ConclusionsOur results indicate that PCV cannot infect Cx. annulirostris via the oral route, nor be transmitted in saliva or vertically to progeny. We also provide further evidence that prior infection with insect-specific viruses can regulate the infection and transmission of pathogenic arboviruses.

Desiccation survival time for eggs of a widespread and invasive Australian mosquito species, Aedes (Finlaya) notoscriptus (Skuse).

The Australian native mosquito Aedes (Finlaya) notoscriptus (Skuse) is closely associated with natural and artificial water holding receptacles. Eggs are laid in habitats where they are exposed to drying conditions as water levels fluctuate. Withstanding desiccation enables survival in challenging environments and increases the potential for establishment in non-native habitats. Until now, the desiccation resistance of Ae. notoscriptus eggs has been unknown despite the historical invasive success of this important dog heartworm and arbovirus vector. Viability and mean survival times of eggs from two Ae. notoscriptus populations (metropolitan areas of Sydney, NSW and Adelaide, SA) were evaluated, with eggs stored under three dryness conditions for up to 367 days. Our results revealed that Ae. notoscriptus eggs can withstand desiccation for extended periods, under a variety of conditions, with approximately 9-13% egg viability recorded after one year. This prolonged egg survival reflects the widespread distribution of this mosquito in Australia and its history of incursions and subsequent establishment in non-native habitats. Differences in mean egg volume were recorded in addition to significantly different egg length to width ratios for the two populations, which may reflect adaptation to biotope of origin and an associated likelihood of drought and drying conditions. The results of this study suggest that the desiccation resistant eggs of Ae. notoscriptus make this species highly adaptable, increasing the risk of movement to non-endemic regions of the world.

A novel insect-specific flavivirus replicates only in Aedes-derived cells and persists at high prevalence in wild Aedes vigilax populations in Sydney, Australia

To date, insect-specific flaviviruses (ISFs) have only been isolated from mosquitoes and increasing evidence suggests that ISFs may affect the transmission of pathogenic flaviviruses. To investigate the diversity and prevalence of ISFs in Australian mosquitoes, samples from various regions were screened for flaviviruses by ELISA and RT-PCR. Thirty-eight pools of Aedes vigilax from Sydney in 2007 yielded isolates of a novel flavivirus, named Parramatta River virus (PaRV). Sequencing of the viral RNA genome revealed it was closely related to Hanko virus with 62.3% nucleotide identity over the open reading frame. PaRV failed to grow in vertebrate cells, with only Aedes-derived mosquito cell lines permissive to replication, suggesting a narrow host range. 2014 collections revealed that PaRV had persisted in A. vigilax populations in Sydney, with 88% of pools positive. Further investigations into its mode of transmission and potential to influence vector competence of A. vigilax for pathogenic viruses are warranted.

Enhanced arbovirus surveillance with deep sequencing: Identification of novel rhabdoviruses and bunyaviruses in Australian mosquitoes

Viral metagenomics characterizes known and identifies unknown viruses based on sequence similarities to any previously sequenced viral genomes. A metagenomics approach was used to identify virus sequences in Australian mosquitoes causing cytopathic effects in inoculated mammalian cell cultures. Sequence comparisons revealed strains of Liao Ning virus (Reovirus, Seadornavirus), previously detected only in China, livestock-infecting Stretch Lagoon virus (Reovirus, Orbivirus), two novel dimarhabdoviruses, named Beaumont and North Creek viruses, and two novel orthobunyaviruses, named Murrumbidgee and Salt Ash viruses. The novel virus proteomes diverged by ≥50% relative to their closest previously genetically characterized viral relatives. Deep sequencing also generated genomes of Warrego and Wallal viruses, orbiviruses linked to kangaroo blindness, whose genomes had not been fully characterized. This study highlights viral metagenomics in concert with traditional arbovirus surveillance to characterize known and new arboviruses in field-collected mosquitoes. Follow-up epidemiological studies are required to determine whether the novel viruses infect humans.

The Role of Australian Mosquito Species in the Transmission of Endemic and Exotic West Nile Virus Strains

Recent epidemic activity and its introduction into the Western Hemisphere have drawn attention to West Nile virus (WNV) as an international public health problem. Of particular concern has been the ability for the virus to cause outbreaks of disease in highly populated urban centers. Incrimination of Australian mosquito species is an essential component in determining the receptivity of Australia to the introduction and/or establishment of an exotic strain of WNV and can guide potential management strategies. Based on vector competence experiments and ecological studies, we suggest candidate Australian mosquito species that would most likely be involved in urban transmission of WNV, along with consideration of the endemic WNV subtype, Kunjin. We then examine the interaction of entomological factors with virological and vertebrate host factors, as well as likely mode of introduction, which may influence the potential for exotic WNV to become established and be maintained in urban transmission cycles in Australia.

Adult mosquito trap sensitivity for detecting exotic mosquito incursions and eradication: a study using EVS traps and the Australian southern saltmarsh mosquito, Aedes camptorhynchus

Adult mosquito traps are commonly used in biosecurity surveillance for the detection of exotic mosquito incursions or for the demonstration of elimination. However, traps are typically deployed without knowledge of how many are required for detecting differing numbers of the target species. The aim of this study was to determine the sensitivity (i.e., detection probability) provided by carbon dioxide-baited EVS traps for adult female Australian southern saltmarsh mosquitoes, Aedes camptorhynchus, a recent biosecurity problem for New Zealand. A mark-release-recapture study of three concurrently released cohorts (sized 56, 296, and 960), recaptured over four days with a matrix of 20 traps, was conducted in Australia. The detection probability for different numbers of traps and cohorts of different sizes was determined by random sampling of recapture data. Detection probability ranged from approximately 0.3 for a single trap detecting a cohort of 56 mosquitoes to 1.0 (certainty of detection) when seven or more traps were used. For detection of adult Ae. camptorhynchus around a known source, a matrix of traps provides a strong probability of detection. Conversely, the use of single traps deployed over very large areas to detect mosquitoes of unknown entry pathway is unlikely to be successful. These findings have implications for the design of mosquito surveillance for biosecurity.

Pyriproxyfen for the Control of Australian Salt-Marsh Mosquito, Aedes vigilax

The efficacy of pyriproxyfen against the Australian salt-marsh mosquito, Aedes vigilax, was examined in 2 laboratory and 1 semi-field study using both technical grade and formulated products. In a dose-response study, the median emergence inhibition (EI50) and EI95 values were determined to be 0.019 and 0.076 ppb, respectively, for pyriproxyfen technical grade, 0.021 and 0.092 ppb for a microencapsulated formulation (Sumilarv 90CS), and 0.054 and 0.236 ppb for the formulated s-methoprene product, Altosid Liquid Larvicide. A further laboratory comparison of the microencapsulated formulation of pyriproxyfen and Altosid, at the nominal field rate for Altosid, showed that both products provided 100% emergence inhibition and this was confirmed in a semi-field study, which also included a granular formulation of pyriproxyfen (Sumilarv 0.5G).

Vector Competence of Australian Mosquitoes for Yellow Fever Virus

The vector competence of Australian mosquitoes for yellow fever virus (YFV) was evaluated. Infection and transmission rates in Cairns and Townsville populations of Aedes aegypti and a Brisbane strain of Ae. notoscriptus were not significantly different from a well-characterized YFV-susceptible strain of Ae. aegypti. After exposure to 10⁷·² tissue culture infectious dose (TCID₅₀)/mL of an African strain of YFV, > 70% of Ae. aegypti and Ae. notoscriptus became infected, and > 50% transmitted the virus. When exposed to 10⁶·⁷) TCID₅₀/mL of a South American strain of YFV, the highest infection (64%) and transmission (56%) rates were observed in Ae. notoscriptus. The infection and transmission rates in the Cairns Ae. aegypti were both 24%, and they were 36% and 28%, respectively, for the Townsville population. Because competent vectors are present, the limited number of travelers from endemic areas and strict vaccination requirements will influence whether YFV transmission occurs in Australia.

The acute toxicity of fipronil to two non-target invertebrates associated with mosquito breeding sites in Australia

Mosquito-borne arboviruses are a significant health issue in the irrigation areas of south-eastern Australia. Fipronil, a pyrazole insecticide with strong activity against larval Culex species, was tested for its acute effects on Simocephalus elizabethae (Daphniidae) and Polypedilum nubiferum (Chironomidae), two non-target invertebrates associated with Australian rice field mosquito habitats. Technical and formulated fipronil were assessed in the presence or absence of particulate artificial diets in 48h static bioassays. LC50 values for neonate S. elizabethae ranged from 11.13 to 19.12μgl−1 whilst those for final instar P. nubiferum ranged from 0.89 to 2.18μgl−1. Feeding during exposure significantly reduced the susceptibility of P. nubiferum to both technical and formulated fipronil. The effect of feeding was less consistent in Simocephalus bioassays, where much less food was present. We investigated whether adsorption to unconsumed food particles may have limited fipronil bioavailability by using solid-phase microextraction and gas chromatography to measure the available fipronil from fed and unfed systems 24h after establishment. Differences between the systems were not significant (P>0.05). The significantly higher LC50 values in the fed Polypedilum bioassays do not appear to be a consequence of reduced fipronil bioavailability. Observed differences in toxicity probably reflect increased stresses associated with food deprivation in the unfed bioassays. Our results support published data on the toxicity of fipronil to aquatic invertebrates which suggest that the use of this material as a mosquito larvicide may cause disruption to aquatic ecosystems.

Eggs of the Australian saltmarsh mosquito, Aedes camptorhynchus, survive for long periods and hatch in instalments: implications for biosecurity in New Zealand

The Australian saltmarsh mosquito, Aedes camptorhynchus (Diptera: Culicidae), is a significant biting pest and disease vector and is the subject of an eradication programme in New Zealand (NZ), where it has been resident for more than 10 years. To better understand the ecology of this common and widespread pest, we studied egg longevity and hatching patterns in the laboratory. By regularly testing for the presence of viable embryos, we found that eggs may last more than 15 months when stored dry (13% viable at this time). Eggs display instalment hatching, with no more than 56% of a batch hatching upon first inundation. Further hatching may occur for at least six inundations and some unhatched eggs may remain viable even after this. Variation in hatching rates can be observed using different water types, with weaker hatching media stimulating lower hatching rates spread over more inundations. By applying average hatching rates to a non-linear model of natural egg attrition, we showed that egg batches exposed to three inundations should be exhausted (zero live eggs present) in approximately 11 months at the conditions tested here. These findings have implications for the current eradication programme for Ae. camptorhynchus in NZ and for our understanding of the ecology of a widespread and common disease vector in Australia.

Vector Competence of Australian Mosquitoes for Chikungunya Virus

Chikungunya virus (CHIKV) is a globally emerging arbovirus responsible for unprecedented outbreaks in the western Indian Ocean, the Indian subcontinent and Italy. To assess the receptivity of Australia to CHIKV, we exposed 10 Australian mosquito species to a 2006 strain of CHIKV isolated from a viremic traveler from Mauritius. In susceptibility trials, the infectious dose required to infect 50% of the mosquitoes was 10(0.6) cell culture infectious dose (CCID)(50)/mosquito for Aedes procax, 10(1.7) CCID(50)/mosquito for Aedes albopictus, 10(2.1) CCID(50)/mosquito for Aedes vigilax, and 10(2.6) CCID(50)/mosquito for Aedes aegypti and Aedes notoscriptus. When exposed to blood meals containing between 10(3.5) and 10(4.1) CCID(50)/mosquito of CHIKV, infection rates in these five species, plus Coquillettidia linealis, were >or=81%. Subsequent transmission rates ranged between 20% for Ae. notoscriptus and 76% for Ae. vigilax. In contrast, Culex spp. were poor laboratory vectors, with infection and dissemination rates <or=20% and <or=12%, respectively. Although Australia has efficient laboratory vectors, the role a mosquito species plays in potential CHIKV transmission cycles will also depend on its geographical and temporal abundance, longevity, and association with humans.

Vector Competence of Australian Mosquito Species for a North American Strain of West Nile Virus

Since the establishment of West Nile virus (WNV) into the United States, concern has arisen that this virus may also pose a serious threat to Australian biosecurity. The vector competence of 19 Australian mosquito species for a North American strain of WNV was evaluated. Mosquitoes collected from Cairns, Brisbane, and Sydney were exposed to blood containing 10(4.0+/-0.3) cell culture infectious dose(50)/mosquito WNV that was isolated from a crow during the 1999 New York outbreak. Mosquitoes were tested 12-15 days later to determine their infection, dissemination, and transmission rates. A number of Culex spp. demonstrated a high vector competence for this virus, with some populations of Culex annulirostris, the primary Australian Kunjin virus vector, displaying transmission rates up to 84%. Similarly, Cx. quinquefasciatus and Cx. gelidus were highly competent, with infection and transmission rates of >80% and >50%, respectively. Common Aedes spp., including Aedes notoscriptus, Ae. vigilax, and Ae. procax, were moderately susceptible, and some Verrallina spp. and Coquillettidia spp. were relatively refractory to infection. Thus, Australia possesses a number of competent mosquito species that could facilitate local transmission of WNV, should it be introduced.

Biologically based insecticides for the control of immature Australian mosquitoes: a review

Abstract Our review outlines changes in insecticide usage for mosquito control against Australian mosquitoes, with a focus on biologically based insecticides. The most widely used microbial insecticide is Bacillus thuringiensis var. israelensis de Barjac (Bti) and it is regularly used in estuaries and freshwater. In comparison, the use of Bacillus sphaericus (Neide) is limited to freshwater. The only insect growth regulator registered for mosquito control in Australia is s‐methoprene, and it is used in both estuaries and freshwater. Although biologically based insecticides have been used for more than 10 years, there remains knowledge gaps surrounding the use of specific formulations in specific habitats: for example, VectoBac WG (Bti) and VectoBac G (Bti) in containers containing freshwater, or Altosid pellets (s‐methoprene) in ground‐pool freshwater habitats. Where broad‐scale mosquito control programs have been implemented in Australia, a reduction in the incidence of mosquito‐borne disease, mainly Ross River virus, has been recorded. Therefore, the future application of these products is supported. The use of insecticides should be integrated with public education, biological control, physical habitat modification and coupled with spatial risk assessment.

Effects of invasive cane toads on Australian mosquitoes: Does the dark cloud have a silver lining?

Research on the ecological impacts of invasive organisms typically looks only for negative impacts, ignoring the possibility that the wider community might see benefits in some of these effects. To truly understand the impact of invasive species, we need to look as broadly as possible, and incorporate studies on a diversity of variables. The spread of the South American cane toad (Bufo marinus) through tropical Australia is widely viewed as an ecological catastrophe, but anecdotal reports suggest that the invasion of toads may reduce the numbers of mosquitoes (and thus, potentially, the risk they pose to human health). We conducted experiments to determine whether the presence of toad tadpoles affects survival rates, adult body sizes and/or rates of oviposition of four species of disease-carrying mosquitoes. In the laboratory, the presence of toad tadpoles significantly reduced the sizes of adult mosquitoes at emergence, and also reduced survival rates of the larvae of one mosquito species. In field trials, mosquitoes were less likely to oviposit in waterbodies containing toad tadpoles. Accordingly, these data suggest (but do not prove) that toad invasion may reduce mosquito abundance. More generally, any overall evaluation of the impact of an invasive species needs to consider possible benefits (e.g. to human health) as well as negative effects (e.g. to native species). Both types of information are essential to inform community decisions about the management of feral taxa such as the cane toad in Australia.

EXPERIMENTAL INFECTION OF CULEX ANNULIROSTRIS, CULEX GELIDUS, AND AEDES VIGILAX WITH A YELLOW FEVER/JAPANESE ENCEPHALITIS VIRUS VACCINE CHIMERA (CHIMERIVAX™-JE)

Australian mosquitoes from which Japanese encephalitis virus (JEV) has been recovered (Culex annulirostris, Culex gelidus, and Aedes vigilax) were assessed for their ability to be infected with the ChimeriVax-JE vaccine, with yellow fever vaccine virus 17D (YF 17D) from which the backbone of ChimeriVax-JE vaccine is derived and with JEV-Nakayama. None of the mosquitoes became infected after being fed orally with 6.1 log(10) plaque-forming units (PFU)/mL of ChimeriVax-JE vaccine, which is greater than the peak viremia in vaccinees (mean peak viremia = 4.8 PFU/mL, range = 0-30 PFU/mL of 0.9 days mean duration, range = 0-11 days). Some members of all three species of mosquito became infected when fed on JEV-Nakayama, but only Ae. vigilax was infected when fed on YF 17D. The results suggest that none of these three species of mosquito are likely to set up secondary cycles of transmission of ChimeriVax-JE in Australia after feeding on a viremic vaccinee.

Toxicity of VectoLex (&lt;I&gt;Bacillus sphaericus&lt;/I&gt;) Products to Selected Australian Mosquito and Nontarget Species

Toxicity of VectoLex (&lt;I&gt;Bacillus sphaericus&lt;/I&gt;) Products to Selected Australian Mosquito and Nontarget Species

Toxicity of VectoLex (Bacillus sphaericus) products to selected Australian mosquito and nontarget species.

Laboratory and field bioassay studies were conducted in southeast Queensland, Australia, on the efficacy of VectoLex Control Granule (CG; active ingredient [AI]:50 Bacillus sphaericus [B.s.] International Toxic Units [ITU]/mg) and VectoLex Water Dispersible Granule (WDG) (AI: 650 B.s. ITU/mg) formulations against third-instar larvae of Culex annulirostris Skuse, Culex quinquefasciatus Say, Culex sitiens Wiedemann, Ochlerotatus rigilax (Skuse), Ochlerotatus. notoscriptus (Skuse), and Aedes aegypti (L.). The Gompertz model for log-linear mortality data were used to determine laboratory 48-h LC95 values. As with foreign evaluations, the B.s. formulations were most effective against Culex spp., with the WDG 10-100 times more effective than the CG on an ITU/mosquito basis. Consequently, Cx. annulirostris and Cx. quinquefasciatus were selected as target species for small-plot field evaluation of WDG efficacy over time. Weekly cohorts of caged third-instar Cx. annulirostris were exposed to replicated low (250 g/ha), medium (500 g/ha), and high (1,000 g/ha) dosages of WDG. Concurrent assessment of Cx. quinquefasciatus mortality outside the cages was also conducted. In water with high organic content, the low rate produced > 99% Cx. annulirostris mortality at 48 h, decreasing to 79% at week 3 and no control at week 4. The medium and high rates resulted in 100% Cx. annulirostris mortality for 2 wk posttreatment, decreasing to 95% at week 3, and no control at week 4. The WDG was equally effective against Cx. quinquefasciatus. Treatment did not affect water quality or nontarget shrimp and fish species survival.

Vector competence of Australian mosquitoes (Diptera: Culicidae) for Japanese encephalitis virus.

Australian mosquitoes were evaluated for their ability to become infected with and transmit a Torres Strait strain of Japanese encephalitis virus. Mosquitoes, which were obtained from either laboratory colonies and collected using Centers for Disease Control and Prevention light traps baited with CO2 and octenol or reared from larvae, were infected by feeding on a blood/sucrose solution containing 10(4.5 +/- 0.1) porcine stable-equine kidney (PS-EK) tissue culture infectious dose50/mosquito of the TS3306 virus strain. After 14 d, infection and transmission rates of 100% and 81%, respectively, were obtained for a southeast Queensland strain of Culex annulirostris Skuse, and 93% and 61%, respectively, for a far north Queensland strain. After 13 or more days, infection and transmission rates of > 90% and > or = 50%, respectively, were obtained for southeast Queensland strains of Culex sitiens Wiedemann and Culex quinquefasciatus Say, and a far north Queensland strain of Culex gelidus Theobald. Although infection rates were > 55%, only 17% of Ochlerotatus vigilax (Skuse) and no Cx. quinquefasciatus, collected from far north Queensland, transmitted virus. North Queensland strains of Aedes aegypti L., Ochlerotatus kochi (Dönitz), and Verrallina funerea (Theobald) were relatively refractory to infection. Vertical transmission was not detected among 673 F1 progeny of Oc. vigilax. Results of the current vector competence study, coupled with high field isolation rates, host feeding patterns and widespread distribution, confirm the status of Cx. annulirostris as the major vector of Japanese encephalitis virus in northern Australia. The relative roles of other species in potential Japanese encephalitis virus transmission cycles in northern Australia are discussed.

Toxicity of insecticides for control of freshwater Culex annulirostris (Diptera: Culicidae) to the nontarget shrimp, Caradina indistincta (Decapoda: Atyidae).

Laboratory evaluations were conducted in southeastern Queensland, Australia, to determine the toxicities of two organophosphate compounds (temephos and pirimiphos-methyl), an insect growth regulator (s-methoprene), and an entomopathogenic bacterium (Bacillus thuringiensis variety israelensis de Barjac [B.t.i.]) to Culex annulirostris (Skuse), an Australian freshwater mosquito vector of arboviruses, and to Caradina indistincta Calman, a co-habiting nontarget shrimp species. S-methoprene and B.t.i. were safest for Cx. annulirostris control with lethal dose ratios (LC95 nontarget/LC95 target) of 3,300 and 846,000, respectively. In contrast, lethal dose ratios for temephos and pirimiphos-methyl were 0.05 and 0.00005, respectively, suggesting that they are environmentally unsuitable. Based on their high lethal dose ratios, s-methoprene and B.t.i. are recommended for control of larval Cx. annulirostris in Australian freshwater habitats.