| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To understand the ecology of these viruses in Brisbane, the objectives of this study were to document the feeding patterns of the mosquito species considered to be vectors or potential vectors on urban vertebrates and show that these vertebrates were naturally infected.
Blood meals were analyzed from wild-caught mosquitoes from the inner and outer suburbs of Brisbane, and animal sera from the Brisbane region were collected and tested for the presence of RRV and BFV antibodies. An animal species that develops high levels of viremia capable of infecting recipient mosquitoes under experimental conditions is inconsequential if not fed on by appropriate vectors under natural conditions. Similarly, the status of animals that develop low levels of viremia or that are infected in relatively small numbers when exposed to virus may be elevated in terms of importance if, under natural circumstances, they are the preferred food source of efficient mosquito vectors.
It is difficult to quantify mosquito feeding patterns, especially because blood-fed mosquitoes usually account for only a small proportion of adult mosquitoes collected. It is also difficult to link mosquito feeding patterns with host availability, mostly because of the difficulties associated with gaining accurate census data and because there are a range of ecological and behavioral factors (pertaining to both mosquitoes and hosts) that may influence mosquito host selection, regardless of host availability. We used the Feeding Index (FI)6 as our method of comparative analysis to examine host-specific feeding success for those mosquito species collected in sufficient numbers. This method does not rely on an exhaustive census, which may be difficult to do for nocturnal and cryptic host species.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mosquitoes were collected twice weekly between September 2000 and April 2001 inclusively. Mosquitoes were collected using modified Centers for Disease Control (CDC) light traps baited with CO2 (as dry ice) and supplemented with 1-octen-3-ol.7 At each collection site, two traps were placed out of the line of sight of each other in the front and back yards. At the Herston site, traps were placed on each side of a building surrounded by gardens. Traps were set at 2–3 hours before sunset and retrieved early the following morning. In addition, resting adult mosquitoes were collected at the Herston site using a hand-held vacuum aspirator8 powered by a 12-V battery, with these collections concentrating in the areas around walls and near vegetation to collect resting mosquitoes.
Identification of available hosts and quantification of blood feeding. Residents at each collection site and all adjoining properties were provided with a questionnaire aimed at determining the relative number of available hosts (humans, dogs, cats, brushtail possums, bats, birds, and others) in the immediate vicinity of the trapping sites. The number of vertebrates estimated to be available as blood sources for each study site included those at the trapping site itself and at adjoining properties, an area averaging 3,600–4,000 m2. At each site, total animal numbers were estimated from the averaged response from each group of residents answering the questionnaire. Animal numbers were similar at all sites except for the following: 1) at Carseldine in the outer suburbs, there were 30 horses recorded, whereas they were absent at the other sites; 2) the numbers of humans present at the sites was underestimated because of some properties being vacant at the time of census, and at the Herston site (inner suburban site), the numbers of visitors and patients at this hospital locality were not estimated.
The FI6 is expressed as: FI = (Ne/Ne')/(Ef/Ef'), where Ne = the number of feeds on host 1, Ne' = the number of feeds on host 2, Ef = the expected proportion of feeds on host 1, and Ef' = the expected proportion of feeds on host 2. An FI of 1 indicates equal feeding on the hosts being compared, and figures greater or less than 1 indicate greater or lesser feeding on the first host relative to the second host, respectively. For quantitative analysis, feeding indices were estimated for combined data because of similarities in animal abundance and because of the low numbers of blood-fed mosquitoes collected.
Identification of mosquito blood meals. Providing that reliable antisera are available, blood meals could be identified up to 48 hours after ingestion, although the reliability of the tests decreased after 24 hours.9 Commercially produced forensic antisera were used to test the bulk of the blood meals as follows: anti-bird and anti-kangaroo (Bethyl Laboratories, Montgomery) and anti-cat, anti-dog, anti-horse, and anti-human (ICN Pharmaceuticals, Costa Mesa). Brushtail possum and flying fox antisera (reacting to four Australian species, reference number 144), respectively, were kindly provided by the Marsupial Cooperative Research Center, Newcastle, Australia, and the Queensland Department of Primary Industries and Fisheries, Brisbane, and adapted for use after determining optimum concentrations.
Human, dog, cat, horse, brushtail possum, and bird blood meals were identified using agar gel diffusion,10,11 but because the flying fox antiserum was of low titer, a modified dot immunobinding assay on nitrocellulose (DOT-PAP) method12 was developed for use. As with agarose diffusion, the test gives a simple positive or negative result.
To test the specificity and sensitivity of the antisera, positive controls were run for each of the antisera. Blood was sourced from humans, dogs, cats, chickens, cockatoos, brush-tail possums, kangaroos, and flying fox species P. alecto and P. poliocephalus. Each blood sample was diluted with phosphate-buffered saline (PBS) at dilutions of 1:50, 1:80, and 1:200 and tested against each of the antisera. In addition, colony mosquitoes that had fed on dogs, cats, humans, flying foxes, and mice were ground in PBS, and the supernatant was diluted as above. Each dilution of these positive controls was tested against each antisera to check for specificity. No cross-reactions between antisera and other species were observed. As negative controls, supernatant from unfed colony mosquitoes were ground in PBS at a dilution of 1:80 and used with each run with both methods. The dilution of 1:80 was found to give a clear result while allowing sufficient volume for an adequate number of tests.
Mosquito blood meals were prepared by grinding each mosquito in 80 µL of PBS. Each mosquito sample was centrifuged at 5,000 rpm for 10 minutes. Supernatant was removed, and debris was discarded. Each sample was divided into 20-µL lots and held at –20°C until needed. Holes were punched into each gel coat using an Ouchterlony 9-hole gel punch and rack (Gelman Instrument, Pall Life Sciences, MI). The resulting pattern was a central well surrounded by eight wells. Two such patterns were punched on each slide. Seven microliters of each blood meal was added to the outer wells, whereas the central wells were used to hold 7 µL of the relevant antisera. By following this format, it was possible to test 16 individual mosquito blood meals against one antiserum (potential host) per slide. Slides were stored face up in humidified boxes at room temperature and checked daily. After 24–48 hours, antisera reacting to blood meals formed immunoprecipitates, characterized by strong opaque bands at a point of equivalence.
The DOT-PAP method was used as follows: Nitrocellulose membrane film (Amersham Hybond C extra) was marked into 10 by 10-cm grids (allowing 100 samples to be processed per test) and positioned in plastic weighing trays. Each mosquito blood meal was diluted 1:80 with PBS, and 5 µL was dotted in each square of the marked film. The films were allowed to dry at room temperature. Once dry, the films were soaked in Tris EDTA and NaCl (TENTC) blocking buffer for 1–2 hours to block the non-specific protein binding sites. Membranes were incubated in Mab (anti-flying fox mixed species) diluted 1:20 in TENTC blocking buffer for 1 hour. After 1 hour, the membranes were washed in PBS/TWEEN three times, soaking the membranes in fresh PBS/TWEEN for 5 minutes between rinses. After rinsing, the membranes were transferred to a suitable anti-mouse conjugate diluted 1:4,000 in TENTC blocking buffer and held for 1 hour. The membranes were washed again as above and drained before 40 mL of substrate solution (diaminobenzoate [DAB]; Sigma Fast DAB with Metal Enhancer; Sigma Aldrich, Castle Hill, Australia) was added. The reactions occurred within several minutes, after which membranes were rinsed in milliQ H2O and air-dried. Reactions that were significantly darker than the negative controls (no Mab) were considered positive.
Vertebrate serology. Serum samples were obtained from dogs, cats, and horses residing in southeast Queensland, within an ~75-km radius of Brisbane City. Sera were obtained from Veterinary Pathology Services (VPS) in Brisbane and sorted according to the address of the veterinary practice that obtained the samples. The locality of the veterinary practice taking the original blood sample was assumed to be the residential locality of the animal. Dog (N = 481), cat (N = 579), and horse (N = 379) samples were collected from March to November 1999. The breed and ages of the donor animals were also recorded.
Flying fox sera were obtained from Queensland Department of Primary Industries and Fisheries serum banks sourced from wild colonies in the urban Brisbane area. Between April 1997 and September 1998, flying fox sera (N = 165) were obtained from wild P. alecto Temminck and P. poliocephalus Temminck at East Brisbane and Indooroopilly (both inner urban areas).
Brushtail possums (N = 100) were captured in wire box traps in urban Brisbane between April 1998 and October 1999 under the Nature Conservation Act 1992 (permit no. E5/ 000015/97/SAA). Blood samples were collected on site from anesthetized possums (using Domitor). Animals were ear-tagged to ensure they were only included once and released. This protocol was approved by the University of Queensland and QIMR Animal Ethics Committees (reference no. A97011).
Detection of antibodies and evaluation of testing methods. To select the most efficient and accurate method for antibody testing, a subsample of sera was tested for RRV antibodies using hemagglutination inhibition (HI), neutralization (Nt), and competitive ELISA.13,14
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Vertebrate host survey. Numbers of dogs, cats, birds, and flying foxes were similar, regardless of locality. In the inner city compared with outer city areas, respectively, an average of 2.1 to 0.6 brushtail possums per residence was reported, but because such sightings were done by night, this may not constitute a real difference. A total of 84 humans were recorded from 29 properties, but the census at Herston was an underestimate because it did not include hospital patients and their visitors. At Carseldine in the outer suburbs, two residences had 30 horses, but the other 27 properties in the other seven localities had none.
Blood-fed mosquitoes and host feeding patterns. Of the mosquitoes collected, 1,119 (3.7%) were blood fed, and host blood source was identified from 865 (77.3%). At the Herston site where both collection methods were used, 27% and 4.0% of those collected by aspirators and light traps were blood fed, respectively. From the aspirator, Ae. notoscriptus blood meals were identified as 8 brushtail possums, 1 human, and 3 unidentified, whereas from the light trap, the results were 16 brushtail possum, 1 human, 2 dog, 1 bird, and 12 unidentified.
The most commonly identified blood meal from all sites was dog (average of 37.4% of all identified blood meals), followed by bird (18.4%), horse (16.8%), brushtail possum (13.3%), human (11.6%), cat (1.7%), flying fox (0.7%), and macropod (0.2%) (Table 1
).
|
) in sufficient numbers from the important vector species of RRV, Cx. annulirostris (N = 450), Ae. vigilax (N = 129), and Ae. notoscriptus (N = 170), for comparative analyses using the FI. For other species from which RRV has been isolated,15 blood meals were identified in relatively small numbers; Culex australicus Dobrotworsky and Drummond (N = 22), Culex quinquefasciatus Say (N = 25), Cx. sitiens (N = 14), Aedes procax (Skuse) (N = 27), Aedes vittiger (Skuse) (N = 5), Coquillettidia linealis (Skuse) (N = 5), and Coquillettidia xanthogaster Edwards (N = 18).
Culex annulirostris was associated with dogs in both inner (32.6%) and outer (48.2%) urban areas and averaged 42.8% of all Cx. annulirostris blood meals identified (Table 1). This species also fed on a variety of hosts in the inner and outer urban localities, including brushtail possums (30.1% and 3.7%, respectively), horses (0% and 33.3%, respectively), humans (19.0% and 5.7%, respectively), and birds (16.3% and 6.7%, respectively), with a lesser degree of contact with cats, flying foxes, and macropods.
Based on relative abundance, the estimated FIs (Table 2) for Cx. annulirostris indicated that dogs were fed on disproportionately compared with brushtail possums (FI = 5.6), cats (FI = 9.2), humans (FI = 10.7), birds (FI = 16.9), and flying foxes (FI = 187.2). For the outer urban Carseldine site where horses occurred (data not shown), the FIs for horses compared with dogs, humans, and birds, were 0.8, 6.6, and 5.1, respectively.
|
). FIs (Table 3) showed strong feeding of Ae. notoscriptus on dogs compared with brushtail possums (FI = 2.5), humans (FI = 4.9), birds (FI = 9.1), cats (FI = 6.9), and flying foxes (FI = 15.0) relative to host abundance.
|
), the most common blood sources were dogs (31.8%), birds (30.2%), humans (14.0%), and brushtail possum (9.3%), with meals also on cats and a flying fox. When blood meals from host species were compared against each other (Table 4), dogs were most often fed on over other species, e.g., compared with brushtail possums (FI = 5.7), humans (FI = 5.8), birds (FI = 4.1), cats (FI = 6.8), and flying fox (FI = 39.8). (Table 4).
|
A total of 1,706 sera from five animal species were tested (Table 5). Sera from dogs and horses were most commonly positive for antibodies to RRV, with antibodies present in 22.5% and 25.5% of samples, respectively. Brushtail possum and flying fox sera had a RRV antibody prevalence of 17.6% and 13.3%, respectively. Cat sera accounted for the lowest proportion of antibody to RRV at 10.5%.
|
). In all cases, antibody to BFV was less common than antibody to RRV, but antibodies to both RRV and BFV were found in 11% of horses, 0.2% of dogs, and 0.3% of cats. |
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The ecology of BFV presently is poorly defined, but potential or actual vectors include Ae. vigilax, Ae. notoscriptus, Cq. linealis, and possibly, Cx. annulirostris. For BFV, it is known that dogs and cats remain aviremic and that brushtail possums may develop a low-level viremia after infection, but little else. HI antibodies have been detected in horses, cattle, and goats, but mainly in macropods.16,17
Vertebrate blood-meal sources. Most of the blood meals identified were from the CO2- and octenol-baited light traps, and so it is possible that some of the blood-fed mosquitoes collected (they were also attracted by light) were still host seeking, as a result of attempting to feed on intolerant rather than sedentary hosts. This could bias the results. Our only indication that this was not the case is that for Ae. notoscriptus at Herston, where both aspirator and light trap methods indicated that the primary blood meal source was brushtail possum.
In Brisbane, dogs were a major blood source for Cx. annulirostris, Cx. quinquefasciatus, Ae. notoscriptus, Ae. vigilax, and Ae. procax. For those species collected in adequate numbers detailed for analysis, Culex annulirostris, Ae. notoscriptus, and Ae. vigilax, this predominance was consistent in all eight sampling localities in Brisbane. Of all the blood meals tested, only those of Cq. linealis did not contain dog blood. These findings indicate that dogs are highly attractive to mosquitoes, as has been found previously for Cx. annulirostris at Kowanyama,18 Charleville,19 and in several northern Queensland and Papua New Guinean localities,20 but this depended on the abundance and availability of these hosts.
Although dogs were the predominant blood source (42.9%) for Cx. annulirostris, in cases where other mammals were available in relatively large numbers, this taxon was found to feed on other commonly available hosts. For example, at outer urban Carseldine, Cx. annulirostris commonly fed on horses, whereas at the inner city locations where horses were absent, dogs, brushtail possums, humans, and birds were the most common blood sources. Overall, brushtail possums accounted for 12.7% of Cx. annulirostris blood meals, resulting in an FI of 1.9 compared with humans, an FI of 3.0 compared with birds, an FI of 1.6 compared with cats, and FI of 33.2 compared with flying fox.
There was evidence of contact between humans and Cx. annulirostris, Ae. vigilax, Cx. sitiens, Ae. notoscriptus, and Ae. procax but in numbers less than expected based on census data. Although human numbers at the Herston site were underestimates, lower feeding rates would also be influenced by inaccessibility to mosquitoes because of low spatial and temporal concurrence, screened houses, and the use of personal protection. Because mosquitoes were collected outdoors, this would also create a negative bias against humans. This is in direct contrast to the accessibility of animals housed outdoors (e.g. dogs and horses) or those that were active outside nocturnally (e.g. brushtail possums, flying foxes).
Similarly, Ae. notoscriptus is widely accepted as an opportunistic species, feeding naturally on a wide range of species including humans, dogs, marsupials, cattle, pigs, rabbits, birds, and sheep.18–21 Because of its peri-domestic ecology, Ae. notoscriptus has a closer spatial association with humans, compared with most other mosquito species, because of its exploitation of domestic larval habitats.22 Because Ae. notoscriptus do not disperse over long distances,23 this suggests that Ae. notoscriptus could be of greater significance as a vector in more densely populated inner city and older areas. Ae. notoscriptus also feeds throughout the night, with peaks at dusk and dawn, and will also bite during the day21,22 when human blood meals may be more accessible. Importantly, the FI showed abundant feeding of Ae. notoscriptus on brushtail possums (24.7%) and some interaction in the inner suburbs with flying foxes (2.6%).
Dogs were the most common blood source (31.8%) for Ae. vigilax, followed by birds (30.2%) and horses when they were present (24.6% with a dog: horse FI of 1.2 at outer urban Carseldine), but blood sources also included human (14.0%), brushtail possum (9.3%), cat (1.5%), and flying fox (0.8%).
For the blood-fed species collected in small numbers, the results showed few clear trends. In the United States, Cx. quinquefasciatus has been shown to feed on a wide range of hosts, including mammals, birds, reptiles, and amphibians,24 and a similar pattern occurs in Queensland.18–20 Cx. australicus is largely ornithophagic, with 63.6% of their identified blood meals being from birds, but also from dogs, brushtail possums, and horses. Other data also suggest they do not generally feed on humans.21 The Ae. vittiger blood meals identified suggest that, as with Ae. procax, Ae. vigilax, and Cx. annulirostris, this species is opportunistic, with blood meals identified mainly from horses but also from dogs and possums.
About 22.7% of blood meals remained unidentified, and this was particularly evident with Ae. procax. This was presumably because of the restricted range of available antisera and/or advanced digestion of the blood meals. Other marsupials in the region such as ringtail possums (Pseudocheirus peregrinus Boddaert, a different genus, with no cross-reaction with brushtail possum antisera) and bandicoots (Isoodon macrourus Gould), along with a number of bat species, rats, native mice, and even reptiles (lizards, geckos, snakes) and amphibians (frogs and toads), may account for these. For bird species, the mixed species antiserum was reactive to several groups, including pigeons and doves (Columbiformes), ducks, geese and swans (Anseriformes), herons and ibis (Ciconiiformes), chickens, pheasants, and turkeys (Galliformes), and sparrows (Passeriformes). It is likely that further cross-reaction occurs between this antiserum and other bird species and that the majority of blood meals of bird origin would have been detected.
Vertebrate serology. Our survey showed that antibody to BFV is less prevalent than antibody to RRV for all animal species tested. For humans, this is also the case. In a generalized survey,25 31.6% and 6.5% prevalence for RRV and BFV antibodies was found in human sera from Queensland, with overall annual seroconversion rates of 0.59% and 0.23%, respectively. Seroconversion rates of up to 0.72% for RRV and 0.28% for BFV for Queensland notification data have been estimated.26 Between the years 1995 and 2001, 32,216 cases of RRV disease were reported compared with 5,268 for BFV during the same period (Australian National Notifiable Diseases Surveillance System Combined Data). Thus, these findings for these animal sera are consistent with human data, which could indicate that the level of transmission of RRV is three to six times greater than that of BFV.
Neutralizing antibody to RRV was detected in 22.5% of Brisbane dogs. When relative age of the dogs and cats tested in our survey is accounted for, the greater seroprevalence of dogs to cats represents a 2.8-fold difference. From experimental infection studies,5 both hosts are refractory to both viruses and developed a similar neutralizing antibody response. This difference, therefore, may be caused by the reduced exposure of cats to mosquito attack (blood meals on cats were approximately 1/20th of that for dogs) and/or because of lack of attractiveness, avoidance behavior, active disruption of mosquito feeding or preening, or environmental factors such as sleeping indoors.
Horses had the highest proportion of antibodies to RRV (25.5%) compared with the other species assessed. Because RRV is known to cause clinical disease in horses,27,28 this may be of some concern to the racing and performance horse industry, where loss of performance equates to economic loss. In Brisbane, some densely populated urban areas contain racing stables, racetracks, and thoroughbred training facilities that bring large numbers of horses in close proximity to the city. Horses infected with RRV may develop high-level viremias of up to 106.3 suckling mouse intracerebral LD50/mL and can infect recipient mosquitoes.29
When expected seroprevalence of horses to RRV was adjusted for average relative age compared with cats and dogs, rates could be expected to be 3.8 and 1.3 times higher, respectively. Horses had much higher rates of dual seropositivity (11%) compared with the other animal species tested. The exact meaning of this is unclear, as is the case with many factors relating to the ecology of BFV.
Brushtail possums, with RRV and BFV antibodies detected in 17.6% and 10.7% of samples, respectively, are clearly exposed to the viruses under natural conditions. Given that they can develop high viremia after exposure to RRV3 and that they are commonly fed on by vector mosquitoes in urban areas along with humans, brushtail possums are likely to be a major vertebrate host of RRV in urban Brisbane. In view of the largely negative experimental transmission results of BFV to recipient Ae. vigilax, it is suspected that brushtail possums may generally play a minor role, especially because the presence of one BFV topotype throughout Australia may suggest more mobile hosts.30
Flying foxes had a slightly higher RRV antibody prevalence (13%) than did cats (10.5%), but it was lower than that for dogs and horses. P. poliocephalus are believed to be low-grade hosts of RRV,2 and this virus has been isolated from mosquitoes collected near a flying fox camp in Cairns, north Queensland.31 In the context of our study in urban Brisbane, most of the limited interaction with mosquitoes concerned Ae. notoscriptus, which is known to be less competent as a vector of RRV than either Ae. vigilax or Cx. annulirostris.32 In contrast, no BFV antibody was detected in flying foxes.
Epidemiologic implications. Transportation of virus between urban, peri-urban, and rural areas may occur as vertebrate hosts, including humans, move between localities. In relation to the urban conduit hypothesis2 based on ecological study at Indooroopilly Island in inner Brisbane, mosquitoes were shown to feed on flying foxes, which were suggested as one means of transportation of RRV from their peri-urban camps into urban backyards. Outer urban areas, where horses, humans, and marsupials (brushtail possums and macropods) are present, may act as zoonotic foci for RRV, because there is ample interaction with vectors.
In the inner urban areas of Brisbane, our study showed an association between flying foxes and the accepted RRV and BFV vectors Ae. notoscriptus and Ae. vigilax, as well as the important RRV vector Cx. annulirostris. It also showed considerable interaction between seven mosquito species, but notably Ae. notoscriptus, Ae. vigilax, and Cx. annulirostris, with brushtail possums and humans.
Our serological survey showed widespread natural infection of RRV in horses, brushtail possums, flying foxes, dogs, and cats, which suggests that those known to become viremic (the first three hosts and humans) may be involved in transmission cycles in Brisbane.
In terms of the control of RRV, two options are worthy of consideration: 1) reduction or separation of host from vector and 2) reduction of numbers of vectors. Regarding the former, because blood feeding is dependent on host accessibility and availability, pigs as major vertebrate hosts of Japanese encephalitis virus have been relocated in the Torres Strait to reduce vector and host contact.20 This approach is clearly not practical in Brisbane, given that, in the case of RRV, the important host species besides humans are predominantly native and protected species (marsupials and flying foxes) or livestock (horses). It is also possible to consider reduction of suitable (habitat and food) trees for both possums and flying foxes, but this would detract from the aesthetics of the city and would be contrary to strategies for reduction of greenhouse emissions. However, because dogs remain aviremic after infection with either RRV or BFV and because they represented 37.3% of total mosquito blood meals, we would suggest that, in Brisbane, as at Kowanyama,18 they act as a zooprophylactic agent to mitigate transmission to some degree.
Therefore, the only practical solution would seem to lie with the reduction of the number of susceptible vectors to mitigate host–vector contact. Currently in the Brisbane region, there are highly developed and proactive programs to reduce populations of the salt marsh breeding vector, Ae. vigilax, but those for freshwater groundpool mosquitoes (e.g., Cx annulirostris and Ae. procax) are rudimentary and often ad hoc.
For mosquitoes such as Ae. notoscriptus that breed in urban backyards, education to promote community awareness of personal protection and participation in reducing larval habitats is essential. Production of educational materials and some house-to-house visitation to identify breeding sites may also be useful. In the western suburbs of Brisbane, natural habitats, garden accouterments, and discards were the most productive habitats, accounting for 86% of Ae. notoscriptus larvae.22 While government programs can reduce the number of available breeding sites for other species, public motivation is necessary to reduce or treat natural (tree holes, bromeliad axils, palm fronds) and artificial (sumps, drains, roof gutters, tires, general trash) containers for Ae. notoscriptus. Given that this species is probably a major vector of RRV and BFV in Brisbane, upgrading these strategies should reduce the number of locally acquired cases.
Received September 4, 2006. Accepted for publication November 18, 2006.
Acknowledgments: This study forms part of A. M. Boyd’s PhD dissertion at the University of Queensland. She thanks Kim Pham, Andrew van den Hurk (UQ), and John Aaskov (Queensland University of Technology) for advice on blood-meal analyses and Leith Poulsen for technical assistance. The authors thank Veterinary Pathology Services, Brisbane. and Scott Smith and Hume Field, Animal Research Institute, Brisbane. for supplying sera for testing and Kim Halpin, Queensland Department of Primary Industries and Fisheries, for providing the flying fox antisera used for blood-meal analyses.
Financial support: This study was financed by the National Health and Medical Research Council, Canberra.
* Address correspondence to Brian H. Kay, Queensland Institute of Medical Research, Post Office Royal Brisbane Hospital, Queensland 4029, Australia. E-mail: brian.kay{at}qimr.edu.au 
Authors’ addresses: Brian H. Kay and Peter A. Ryan, Queensland Institute of Medical Research, Post Office Royal Brisbane Hospital, Queensland 4029, Australia, Telephone: 61-7-3362-0350, Fax: 61-7-3362-0104, E-mails: brian.kay{at}qimr.edu.au and peter.ryan{at}qimr.edu.au. Ann Marie Boyd, formerly Queensland Institute of Medical Research, currently Pine Rivers Shire Council, PO Box 5070, Strathpine Queensland 4500, Australia, E-mail: Ann-Marie.Boyd{at}pinerivers.qld.gov.au. Roy A. Hall, School of Molecular and Microbial Sciences, University of Queensland, St Lucia, Queensland 4072, Australia, E-mail: roy.hall{at}uq.edu.au.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
