Dryland Salinity and the Ecology of Ross River Virus: The Ecological Underpinnings of the Potential for Transmission

Abstract

Alterations in transmission of vector-borne zoonoses are often linked to environmental change. However, ecological processes that determine variability in potential for transmission are generally not well understood. Ross River virus (RRV, Togoviridae: Alphavirus) is a mosquito-borne zoonosis in Australia with a significant human disease burden. The inland southwest (Wheatbelt) of Western Australia (WA) is substantially affected by an anthropogenic salinization of agricultural land (dryland salinity). Aedes camptorhynchus Thomson (Diptera: Culicidae) is the dominant vector of RRV in southwest WA and is halophilic. As such, dryland salinity may influence potential for RRV transmission by influencing interactions between Ae. camptorhynchus and mammalian hosts. We surveyed areas of the Wheatbelt with varying salinity impacts and found Ae. camptorhynchus was more abundant in saline areas, whereas sheep Ovis aries (Linnaeus 1758, Bovidae) declined with increasing salinity. We used a deterministic model to examine interactions between Ae. camptorhynchus and mammals, and we assessed potential for RRV transmission. We found variation in potential for RRV transmission was positively related to increasing salinity and abundance of Ae. camptorhynchus and negatively associated with increasing abundance of Macropus fuliginosus (Desmarest 1817, Macropodidae). Abundance of Ae. camptorhynchus determined more variation in potential for RRV transmission than other variables. Accordingly, dryland salinity increases the zoonotic potential for RRV transmission primarily by facilitating abundance of Ae. camptorhynchus. Human RRV notifications do not currently reflect the salinity-RRV transmission potential in the Wheatbelt but appear to be associated with RRV activity in the enzootic coastal zone. We speculate dryland salinity is a determinant of potential for RRV transmission but not activity. Dryland salinity is predicted to expand two- to four-fold by 2050. Preservation and restoration of freshwater ecosystems may ameliorate the potential for transmission of RRV and possibly incidence of human disease.

Introduction

I dentifying ecological processes that govern pathogen transmission is essential to understand how environmental change will affect variability in dynamics of human and wildlife diseases (Ezenwa et al. 2007). Although change in land cover is often tied to spatial variability in disease risk and occurrence (Daszak et al. 2001, Derraik and Slaney 2007, Klinkenberg et al. 2004, Molyneux 2003, Patz et al. 2000, Ryan et al. 2001, Walsh et al. 1993), ecologic underpinnings of these correlations are often unknown (Ezenwa et al. 2007). In Australia, Ross River Virus (RRV, Togoviridae: Alphavirus) is a mosquito-borne zoonosis, spilling over into the human population and causing a significant disease burden (Harley et al. 2001, Russell 2002). Numerous anthropogenic influences on the environment are linked to human incidence and risk of RRV infection (Foley et al. 2004, Harley et al. 2005, Jardine et al. 2004, 2007, Kelly-Hope et al. 2002, Lindsay et al. 2007, Russell 1998, 1999, Ryan et al. 2001, 2006), and disease burden appears to be governed by zoonotic transmission (Harley et al. 2001, Lindsay et al. 1996, Russell 2002). Yet investigations that specifically consider zoonotic transmission between vectors and hosts are rare (Carver et al. 2008b, Glass 2005, Kay et al. 1987b). Here, we use a mechanistic model to investigate links between anthropogenic salinization of the environment and the host-vector ecology of RRV transmission.

Environmental factors influence hosts and vectors and ultimately transmission of vector-borne zoonoses. Empirical and modeling studies have implicated a number of ecological processes that determine variability in transmission risk. Vector abundance is traditionally considered a determinant of transmission risk (Jardine et al. 2004, Kilpatrick et al. 2006b, Overgaard et al. 2003, Russell 1986), but this is not supported by all studies (Kanojia et al. 2003, Lindsay et al. 2005, Ryan et al. 1999). Vectors are not obligate generalists, consequently heterogeneities in vector feeding behavior may influence transmission (Tirados et al. 2006, van den Hurk et al. 2003), such as for West Nile virus (WNV, Flaviviridae: Flavivirus; Kilpatrick et al. 2006a, 2006b). More recently, diversity of hosts has also been identified as a determinant of transmission risk (Dobson 2004), such as for WNV, Lyme disease bacterium (Borrelia burgdorferi) and louping-ill virus (Flaviviridae: Flavivirus; Ezenwa et al. 2006, 2007, Hudson et al. 1995, LoGiudice et al. 2003, Ostfeld and Keesing 2000a, 2000b). Studies that examine links between environmental variables and vector-borne disease risk may benefit from considering ecological interactions between hosts and vectors (Carver et al. 2008a).

In southwest Western Australia (WA), RRV is principally transmitted by the halophilic mosquito Aedes camptorhynchus Thomson (Diptera: Culicidae), which feeds on a variety of, mostly mammalian, vertebrate species (Jardine 2007, Lindsay et al. 1993a, 1996, 1998, 2005, 2007). The putative reservoir host of RRV is thought to be the western grey kangaroo, Macropus fuliginosus (Desmarest 1817, Macropodidae), but other mammals are likely to influence transmission, due to the feeding behavior of Ae. camptorhynchus (Carver et al. 2008a). Anthropogenic salinization of land (dryland salinity) in inland southwest WA (the Wheatbelt) is postulated to influence potential for RRV transmission, by providing large areas of suitable habitat for Ae. camptorhynchus breeding (Biggs and Mottram 2008, Jardine et al. 2007, Lindsay et al. 2007). This is supported empirically by studies of mosquito larvae (Carver et al. submitted), but not supported by human clinical notifications (Jardine et al. 2008). The host-vector ecology of RRV transmission, in relation to salinity, has not been assessed. Currently more than one million hectares are saline in the Wheatbelt, and dryland salinity is expected to expand to between three and four and a half million hectares by 2050 (George et al. 2006, Jardine et al. 2007).

In this study, we examine the relationship between dryland salinity and ecology of potential RRV transmission in the WA Wheatbelt. Initially, we examine empirical associations between salinity, abundance of Ae. camptorhynchus, and assemblages of mammals. Using a simple susceptible-exposed-infectious-recovered (SEIR) model of vector-host interactions, we simulate zoonotic interactions that influence temporal and spatial variability in potential for RRV transmission. We then retrospectively examine which factors (salinity, mosquito, or mammal) explained the greatest amount of variation in potential for RRV transmission. We hypothesize that salinity will promote the spatial potential for RRV transmission. Humans are not specifically considered in our analysis, but by extension our results are relevant to humans. In general, the model used here is widely applicable to the ecology of RRV across Australia and other mosquito-borne zoonoses with multiple host species.

Methods

Study area

This study was undertaken in the Great Southern meteorologic district of the WA Wheatbelt (Fig. 1). The study area has a Mediterranean climate with hot dry summers and mild wet winters. Annual rainfall declines from approximately 600 mm at the western boundary of our study area to 350 mm in the east (Australian Bureau of Meteorology). Approximately 80–90% of the region has been cleared for agriculture (Halse et al. 2004). Sites were separated by 5 km or more, located in remnant native vegetation, and chosen to correspond with locations previously used by the WA Department of Environment and Conservation to assess impacts of secondary salinization (Keighery et al. 2004).

FIG. 1.

Sites surveyed for Ae. camptorhynchus and mammal fauna in the Wheatbelt of southwest Western Australia.

Modeling

To investigate links between salinity and zoonotic RRV transmission in the Wheatbelt, we simulated the ecology of RRV transmission between Ae. camptorhynchus and multiple mammalian host species, using a simple deterministic SEIR model (Eq. 1, Table 1). \documentclass{aastex} \usepackage{amsbsy} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{bm} \usepackage{mathrsfs} \usepackage{pifont} \usepackage{stmaryrd} \usepackage{textcomp} \usepackage{portland, xspace} \usepackage{amsmath, amsxtra} \pagestyle{empty} \DeclareMathSizes {10} {9} {7} {6} \begin{document} \begin{align*}&\frac{dS_{M}}{dt} = - r \sum\limits_H T_{H,M}I_{{H_i}})S_M - \mu S_M \\&\frac{dE_M}{dt} = r ( \sum\limits_H T_{H,M}I_{{H_i}})S_M - \sigma_M E_M - \mu E_M \\&\frac{dI_M}{dt} = \sigma_M E_M - \mu I_M \\&\frac{dS_{H_i}}{dt} = - rf_i T_{MH_i} I_M S_{H_i}\\&\frac{dE_{H_i}}{dt} = rf_i T_{MH_i} I_M S_{h_i} - \sigma_{H_i} E_{H_i}\\&\frac{dI_{H_i}}{dt} = \sigma_{H_i} E_{H_i} - \gamma_{H_i} I_{H_i} \\&\frac{dR_{H_i}}{dt} = \gamma_{H_i} I_{H_i} \\&{\rm Where}: \ r = \frac{b}{N_H} \\& R_0 = \sum\limits_H \frac{(bf_{H_i})^2 T_{{H_i}M} T_{MH_i}N_M}{\mu \gamma_{H_i}N_{H_i}} \tag{{\rm Eq. 1}} \end{align*} \end{document}

Table 1.

Definitions of Parameter Used in the Model

b	bite rate of mosquitoes
r	rate at which hosts are bitten by mosquitoes
f _{H_i}	fraction of vectors that feed on host species i
T _{MH_i}	transmission probability (following a bite) from mosquito to host species i
T _{H_i} _M	transmission probability from host species i to mosquito
σ_i	rate at which individuals of species i move from exposed to infectious classes
γ _{H_i}	recovery rate for host species i
μ	per capita death rate for mosquitoes
S_i	number of individuals of species i that are susceptible
E_i	number of individuals of species i that are exposed
I_i	number of individuals of species i that are infectious
R_i	number of individuals of species i that are immune
N_i	population size of species i

Equation 1 is the model of RRV transmission between Ae. camptorhynchus and multiple host species, and calculation of R ₀. Mosquitoes (M) are divided into susceptible, exposed and infectious (S, E, and I, respectively) classes. Hosts (H) are divided into susceptible, exposed, infectious, and recovered (R) classes. For definition of parameters see Table 1.

Our model of epizootics assumes RRV is introduced into sites, where local transmission occurs, briefly, and is followed by local extinction of the virus. This assumption is consistent with observations of human cases by Jardine et al. (2008). Arboviruses can be transported by infected vectors (Kay and Farrow 2000) and hosts (Aaskov et al. 1981, Kelly-Hope et al. 2002, Lindsay et al. 1993c, McBride 2008, Ryan et al. 1997). We assumed RRV would most likely be introduced to a site by either a single infectious Ae. camptorhynchus or host. Initial trials of our model found RRV became extinct for all sites and dates, unless an artificially large vector population (2–10 million) was simulated. Consequently, further simulations were based on the introduction of a single infectious host, the hypothesized RRV reservoir in southwest WA, M. fuliginosus (Lindsay et al. 1996). Due to the brief temporal nature of RRV in the Wheatbelt, births and deaths of hosts were considered to be negligible during epizootics (Eq. 1). We made conservative assessments of Ae. camptorhynchus production for each simulation, assuming mosquitoes were introduced in a single pulse, death rate was constant, and infection with RRV chronic (Eq. 1, Ballard and Marshall 1986). The salinity of water bodies was assumed to have no significant effect on survival and developmental rate of larval Ae. camptorhynchus (Barton and Aberton 2005).

Our model assumes RRV transmission follows mass action (de Jong et al. 1995). Inclusion of the parameter r (Eq. 1, Table 1) means mosquito-to-host transmission is frequency-dependent, whereas host-to-mosquito transmission is density-dependent (Keeling and Rohani 2008). Simulations were conducted using MATLAB^® R2007a (The MathWorks Inc., Natick, MA). Two assessments of potential for RRV transmission were calculated: epidemic size (ES, the number of infectious mosquitoes produced over the duration of the simulation) and reproductive ratio (R ₀, the number of secondary infections in mosquitoes generated, assuming all hosts are susceptible, Eq. 1; Keeling and Rohani 2008).

Ae. camptorhynchus and mammal abundance parameters

The model was parameterized by our empirically derived estimates of mosquito and mammal abundance (Table 2). Surveys of larval Ae. camptorhynchus were conducted from September to December 2005 (spring) and February to May 2006 (late summer and autumn), and mammals from September to December 2006 and February to May 2007. These seasons were chosen to reflect the times of year when rainfall occurs, Ae. camptorhynchus were most abundant, and temperatures permissive for viral replication.

Table 2.

Salinity Category for Each Site and the Associated: Abundance of Ae. Camptorhynchus (Larval Abundance Adjusted to Number Surviving to Emergence) in a 1-Hectare Area (Empty Cells Indicate Dates When Sites Were Dry), for Each Sampling Period, and Average Abundance and Diversity of Mammals Occurring in a 1-km Radius Around the Mosquito Sampling Site

Site	DA10	DA13	DU02	DU09	DA03	DA12	DA05	DA02	DA07	DU06	DU08	DU13
Salinity	1	1	1	1	2	2	3	4	4	4	4	4
Ae. camptorhynchus
12-14 Sep	0	0	5	0	1531	0	0	898	611	26490	0	580
24-26 Sep	0	0	0	0	227	0	0	3468		6101	4	3619
8-10 Oct	0	0	66	0	0	0	0	23882		3523	29	0
25-26 Oct	825		0		0	0		7750		7751	0	0
7-8 Nov	0		0		0	0		117		8342		0
27 Feb			0
12 Mar			0
27 Mar			0
11 Apr			0
24 Apr			0					0		9
8 May								30		27
Mammals
Abundance	1502	5993	2061	3299	3550	1955	1272	1722	4605	5029	3753	3367
Diversity	5	6	4	5	5	5	4	5	6	4	5	6
O. aries ^a	742	669	515	432	668	454	391	301	299	53	192	197
M. fuliginosus	47	39	143	54	81	95	171	12	75	74	51	13
M. domesticus ^a	701	5257		2804	2103	1402	701	1402	1402	4205	3504
O. cuniculus ^a	2	2	2	7		3	10	3	10		5	2
V. vulpes ^a		2		3	3	1					1	1
T. vulpecula ^a					696				2799	696
P. calura ^a			1402									701
N. mitchelli ^a												2453
M. irma ^a								5
C. hircus ^a										20
E. caballus ^a		24
S. scrofa ^a	10

Common name, genus and species, and authority: sheep Ovis aries (Linnaeus 1758, Bovidae), house mouse Mus domesticus (Schwarz & Schwarz 1943, Muridae: syn. Mus musculus Linneaus 1758), European rabbit Oryctolagos cuniculus (Linnaeus 1758, Leporidae), fox Vulpes vulpes (Linnaeus 1758, Canidae), brush-tailed possum Trichosurus vulpecula (Kerr 1792, Phalangeridae), red-tailed phascogale Phascogale calura (Gould 1844, Dasyuridae), Mitchell’s hopping mouse Notomys mitchelli (Ogilby 1838, Muridae), brush wallaby Macropus irma (Jourdan 1837, Macropodidae), goat Capra hircus (Linnaeus 1758, Bovidae), horse Equus cabalus (Linnaeus 1758, Equidae), and pig Sus scrofa (Linnaeus 1758, Suidae).

Surveys of Ae. camptorhynchus larvae were undertaken in (0.2 × 10⁻⁵–0.97 ha) rain-fed ephemeral water bodies fortnightly across 12 sites (Fig. 1, described in detail by Carver et al., submitted). At each site, a 1-hectare quadrat was established in a low-lying area, and on each fortnightly sampling occasion, the surface area of the quadrat occupied by ephemeral water bodies calculated. In total, Ae. camptorhynchus larvae were sampled from 180 water bodies during the course of the surveys. Surveys of Ae. camptorhynchus were undertaken using a standard D-frame 500-mm diameter FBA pond net (250-μm mesh size: Australian Entomological Supplies Pty. Ltd., Bangalow, New South Wales), which was swept through all water bodies and microhabitats within each water body at each site, and the area sampled for Ae. camptorhynchus was recorded. Samples were preserved in 70% ethanol and returned to the laboratory, where 6266 Ae. camptorhynchus were identified under a microscope (Liehne 1991). Ae. camptorhynchus abundance at each site for each sampling interval was calculated by multiplying the area sampled by the area occupied by standing water within each 1-hectare quadrat.

Surveys of mammals were undertaken using a combination of transect counts and trapping across 12 sites (Fig. 1, described in detail by Carver 2008c). Transect surveys were undertaken during the hour immediately after dawn and the half hour before dusk. Surveys consisted of walking an established line transect through remnant vegetation or slowly driving a transect along a farm track/dirt road through or adjacent to the remnant. Each site was surveyed repeatedly (6–8 times), and surveys were averaged to account for temporal and spatial variability in observations (n = 87 surveys). Densities of mammals at each site were calculated as the number observed per hectare, adjusted for observational area within each habitat type. Trapping grids were laid in areas of remnant vegetation at each site using box traps (90 × 90 × 350 mm, Elliott Scientific Equipment, Upwey, Victoria) and wire cage traps (220 × 220 × 550 mm, Sheffield Wire Products, Welshpool WA), baited with peanut butter (Sanitarium^®, New South Wales) and rolled oats (Asia/Pacific Wholesalers Pty. Ltd., Parramatta, New South Wales). Traps were set just before dusk, then cleared and closed shortly after dawn. Each site was trapped twice (once in spring and once in autumn) for 3 consecutive nights, giving a total of 6480 trap nights. Trapping was used to estimate the minimum number alive of any mammal within the trapping grid and approximate densities per hectare for each site.

Abundance of adult Ae. camptorhynchus was calculated by adjusting larval abundance by survival rate (0.844 at 20°C, Barton and Aberton 2005; Table 2). Dispersal studies have found the majority (∼90%) of adult Ae. camptorhynchus dispersal occurs within a 1-km radius of release site (Jardine 2007, Robertson 2006). As such, our measurement of mammal density (number per hectare) was scaled to abundance occurring in a 1-km radius around the mosquito sampling site (3.14 km², Table 2). This radius is also within the transect survey boundaries from mammal surveys and similar to the size of vegetation remnants in which mammal trapping and surveys were conducted.

Salinity parameters

Salinity of sites was designated into four categories by the WA Department of Environment and Conservation (Keighery et al. 2004, McKenzie et al. 2003, van Gool and Moore 1999), and these categories were modified and supplemented by our own observations of dead trees (stags) and samphire in saline areas, which are known indicators of soil salinity and shallow water tables (Table 2): (1) nonsaline, due to absence of stags and samphire, high position in the landscape, low water table, high soil permeability, and/or the low salt store in the regolith; (2) low salinity impact or risk, due to stags and samphire being uncommon, small variation in local relief and geology where rising water tables may not affect all the land area, or where rising water tables are not currently saline, and the salt store in the regolith is low; (3) intermediate salinity impacted, due to common occurrence of stags and samphire, salinity occurring in limited areas, or from shallow saline groundwater with a rising trend; and (4) saline land, due to widespread occurrence of stags and samphire and obvious salt effects (such as large patches of bare earth) in the area. These salinity categories provided a good indication of the continuum of salinity changes in the field and are equivalent to those used by Carver (2008c), Keighery et al. (2004), and McKenzie et al. (2003).

RRV transmission parameters

Ae. camptorhynchus were predicted to seek a blood meal at a rate (b) of once per gonadotrophic cycle (7 days, Barton and Aberton 2005). Infected Ae. camptorhynchus become infectious at a rate (σ_M ) of 1/3 days (Ballard and Marshall 1986). A specific mortality rate for adult Ae. camptorhynchus is unknown, so a general figure (μ = 0.1) for Aedes sp. was used (Canyon et al. 1999, Carron et al. 2008, Harrington et al. 2001, Leisnham et al. 2008, Muir and Kay 1998, Niebylski and Craig 1994, Watson et al. 2000). Blood meal analysis of Ae. camptorhynchus in the Wheatbelt were conducted by Jardine (2007), and this information was used to determine f for each mammal species (Tables 1 and 3). For cases where mammal species blood was not examined by Jardine (2007), such as M. domesticus, f was approximated from Lindsay et al. (1998). The remaining parameters for hosts were derived from reports of RRV infections of mammals and approximated, where data were unavailable (Table 3).

Table 3.

Host Parameters and Assumptions for the Model

Species	f	T_MH	T_HM	σ	γ	Reference
O. aries	0.478	0.7^a	0.5^b	0.5–1^c	0.25–1^c	Jardine 2007, Spradbrow 1973
M. fuliginosus	0.290	1^a	0.5^b	1^b	0.167	Jardine 2007, Kay and Aaskov 1989
M. domesticus	0.014	0.25–1^c	0.75	1	0.125–1^c	Ballard and Marshall 1986, Lindsay et al. 1998, Seay et al. 1981
O. cuniculus	0.027	0.5–0.778^a,c	0.5^b	1	0.333–1^c	Jardine 2007, Kay and Aaskov 1989, Whitehead 1969
V. vulpes	0	0.1	0	0	0	Boyd and Kay 2002, Jardine 2007
T. vulpecula	M. fuliginosus ^d	0.3	0.53	1	1	Boyd et al. 2001, Jardine 2007
P. calurae	M. domesticus ^f	1^a	0.5^b	1	0.167–1^c	Lindsay et al. 1998, Whitehead 1969
N. mitchelli	M. domesticus ^f	1^a	0.5^b	1	0.167–1^c	Lindsay et al. 1998, Marshall and Miles 1984, Whitehead 1969
M. irma	M. fuliginosus ^d	1^a	0.5^b	1^b	0.333–0.5^c	Jardine 2007, Marshall and Miles 1984
C. hircus	O. aries ^d	0.7^a	0.5^b	0.5	0.25^c	Jardine 2007, Kay et al. 1987a
E. caballus
S. scrofa	0	0.909–1^a,c	0.5^b	1	0.2–1^c	Jardine 2007, Kay and Aaskov 1989, Spradbrow 1973

See Table 1 for definitions of parameters f, T_MH, T_HM, σ, and γ.

Transmission to host by subcutaneous inoculation.

In the absence of empirical data, T_HM and σ assumed to equal 0.5 and unity, respectively.

Where a range of values were available, a middle value was used.

When blood meal data could cross-react with another host species at a site, f was partitioned between the cross-reacting species, based on the density of both hosts.

Parameter values based on Bandicoot (Isoodon macrourus) from Whitehead (1969), Marsupial mouse (Antechinus sp.) from Whitehead (1969), Tammar wallaby (Macropus eugenii) from Marshall and Miles (1984), and O. aries from Spradbrow (1973).

The contact rate (f) of P. calura and N. mitchelli with Ae. camptorhynchus was assumed to be consistent with the similar-sized mammal, M. domesticus.

Analyses

Linear regression analysis was used to examine the relationship between salinity category and the abundance of Ae. camptorhynchus, abundance and diversity of mammals, and abundance of mammal species (occurring at three or more sites). Linear regression was preferred over analysis of variance (in this and subsequent analyses) because salinity categories were indicative of a linear continuum of environmental change in the field, and there were few sites occurring at low and intermediate salinity categories (Cottingham et al. 2005). Aedes camptorhynchus exhibit temporal fluctuations (Table 2) associated with rain and temperature (Barton et al. 2004), and this is not easily controlled for in our analysis. Rather than using an average abundance of Ae. camptorhynchus per site, which would mask the observed natural variation, all samples were used in analyses, which contributed variation but provided a more holistic view of the relationship between abundance of Ae. camptorhynchus and salinity category. Abundance data of mosquitoes were log transformed to achieve normality. For all statistical tests, the significance level was set at 0.05. But to reduce the probability of committing type I statistical errors, α was adjusted for multiple comparisons using the sequential Bonferroni method (Rice 1989), where k is the number of analyses testing the same hypothesis. Analyses were undertaken using SPSS 15.0 (SPSS Inc., Chicago, IL).

After calculating R ₀ and ES from our simulations, we retrospectively examined which factors (salinity, mosquito, or mammal) explained the greatest amount of variation in the potential for RRV transmission. We used linear regressions to examine the relationship of R ₀ and ES to salinity category, abundance of Ae. camptorhynchus, abundance and diversity of mammals, and abundance of mammal species (occurring at three or more sites). In cases where Ae. camptorhynchus or mammals were significantly related to salinity, we examined residual variation to control for salinity interrelationships in regressions. R ₀, ES, and abundance data of mosquitoes and mammals were log-transformed to achieve normality, and á was adjusted for multiple comparisons using the sequential Bonferroni method (Rice 1989). Analyses were undertaken using SPSS 15.0 (SPSS Inc.).

Results

Vectors, mammals, and salinity

Aedes camptorhynchus were predominantly detected in the highest salinity category (Table 2 and Fig. 2a). However, this mosquito was also occasionally detected at low saline and nonsaline sites (DA10, DU02, and DA03, Table 2 and Fig. 2a). There was temporal variability in the existence of standing water between sites and the number of Ae. camptorhynchus within and between sites (Table 2). The overall abundance of Ae. camptorhynchus increased positively with salinity category (Fig. 2a). For mammals, O. aries decreased in association to increasing salinity category (Fig. 2b). The remaining measures of host species were not significantly related to salinity (Table 2, linear regressions p > 0.05 in all instances, Carver, 2008c).

FIG. 2.

Relationship between salinity category (1 = nonsaline, 2 = low saline, 3 = intermediate saline, and 4 = saline; Table 1) and (a) the abundance of Aedes camptorhynchus in 1-hectare area, at each sampling session, and (b) average abundance of Ovis aries in 1-km radius surrounding mosquito sampling site (linear regressions plus predictor and constant coefficients (± standard error): r ² = 0.305, F _1,6 = 24.597, p < 0.001, 0.624 (0.126), and − 0.491 (0.359); and r ² = 0.720, F _1,10 = 25.665, p < 0.001, − 131.34 (25.93), and 748.77 (75.21) respectively; critical α-value 0.007, Table 2). Figure 2b modified and reproduced with permission from Carver, 2008c.

Modeling results

R ₀, as a measure of potential for RRV transmission, did not exceed unity at any site or sampling time in the first three salinity categories (Table 4). Ae. camptorhynchus were encountered on 5/34 sampling occasions in the first three salinity categories (Table 1), meaning that both R ₀ and ES values for our model were 0 for 25/34 sampling occasions (Table 4). Comparatively, Ae. camptorhynchus was detected on 18/24 sampling occasions in the fourth salinity category (Table 4). R ₀ and ES values varied between sites and over time within sites (Table 4). Both R ₀ and ES values were higher during spring and at DA02, DU06, and DU13 than in autumn and at other sites (Table 4). Epidemic size values from introducing a single infectious host (M. fuliginosus) were small, only exceeding unity for 2/58 simulations (Table 4).

Analysis of R ₀ and ES found these measures of potential for RRV transmission were significant and positively related to salinity category and abundance of Ae. camptorhynchus (Table 5, Fig. 3a and 3b). Additionally, there was a significant negative relationship between R ₀ and the abundance of M. fuliginosus (Fig. 3c). Regressions between R ₀ and the abundance of V. vulpes and ES and the abundance of M. fuliginosus and V. vulpes were all negative, but not significant (Table 5), implying that sampling of a greater number of sites may have picked up an effect. Abundance of Ae. camptorhynchus explained more variation in R ₀ and ES potential for RRV transmission than any other variable (Table 5, Fig. 3b).

FIG. 3.

Relationships of R ₀ with (a) salinity category, (b) the abundance of Aedes camptorhynchus and (c) Macropus fuliginosus (linear regressions: r ² = 0.225, F _1,56 = 16.292, p < 0.001; r ² = 0.438, F _1,56 = 43.614, p < 0.001; and r ² = 0.196, F _1,56 = 13.686, p < 0.001, respectively, Table 5).

Table 4.

R ₀ and Epidemic Size Values (R ₀/ES) for Sites and Dates in Which Ephemeral Water Bodies Were Sampled for Ae. Camptorhynchus (Salinity Category for Each Site Also Listed)

Site	Salinity	12–14 Sept	24–26 Sept	8–10 Oct	25–26 Oct	7–8 Nov	24 Apr	8 May
DA10	1	0/0	0/0	0/0	0.96/0.12	0/0
DA13	1	0/0	0/0	0/0
DU02^a	1	0.002/0.001	0/0	0.03/0.01	0/0	0/0	0/0
DU09	1	0/0	0/0	0/0
DA03	2	0.08/0.09	0.01/0.01	0/0	0/0	0/0
DA12	2	0/0	0/0	0/0	0/0	0/0
DA05	3	0/0	0/0	0/0
DA02	4	2.39/0.11	9.22/0.43	63.5/2.98	20.61/0.95	0.31/0.01	0/0	0.08/0.04
DA07	4	0.06/0.03
DU06	4	13.22/1.11	3.05/0.26	1.76/0.15	3.87/0.32	4.16/0.35	0.004/0.004	0.01/0.001
DU08	4	0/0	0.004/0.0002	0.03/0.002	0/0
DU13	4	2.43/0.04	15.14/0.23	0/0	0/0	0/0

Ephemeral water bodies also occurred and were sampled on February 27, March 12, March 27, and April 11, 2006, but no Ae. camptorhynchus were detected on these dates (Table 1) giving an R ₀/ES value of 0/0 for each.

Table 5.

Regression Analyses of R ₀ and ES Against Salinity, Ae. Camptorhynchus, Mammal Abundance, Diversity, and the Abundance of the Most Common Mammals

	R₀			Regression coefficients (± SE)		ES			Regression coefficients (± SE)
	r	F_1,56	p ^a	Predictor	Constant	r²	F_1,56	p ^a	Predictor	Constant
				0.140	−0.171				0.027	−0.032
Salinity	0.225	16.292	<0.001	(0.035)	(0.099)	0.133	8.600	0.005	(0.009)	(0.026)
				0.208	0.182				0.045	0.035
Ae. camptorhynchus	0.438	43.614	<0.001	(0.031)	(0.040)	0.336	28.354	<0.001	(0.008)	(0.011)
				0.092	−0.314				0.002	0.028
Abundance	0.002	0.120	0.731	(0.267)	(0.916)	0.001	0.001	0.974	(0.066)	(0.227)
				0.005	0.116				−0.009	0.077
Diversity	0.001	0.005	0.944	(0.078)	(0.379)	0.004	0.206	0.652	(0.019)	(0.093)
				−0.000	0.182				−0.000	0.035
O. aries	0.001	0.024	0.878	(0.001)	(0.053)	0.001	0.009	0.925	(0.001)	(0.013)
				−0.482	1.026				−0.087	0.187
M. fuliginosus	0.196	13.686	<0.001	(0.130)	(0.233)	0.104	6.495	0.014	(0.034)	(0.061)
				0.029	0.111				0.012	0.007
M. domesticus	0.012	0.656	0.421	(0.036)	(0.103)	0.031	1.773	0.188	(0.009)	(0.025)
				−0.237	0.293				−0.048	0.058
O. cuniculus	0.031	1.775	0.188	(0.178)	(0.098)	0.020	1.156	0.287	(0.044)	(0.024)
				−0.461	0.258				−0.106	0.053
V. vulpes	0.062	3.674	0.060	(0.240)	(0.065)	0.053	3.165	0.081	(0.060)	(0.016)
				0.043	0.155				0.011	0.028
T. vulpecula	0.017	0.954	0.333	(0.044)	(0.060)	0.018	1.047	0.311	(0.011)	(0.015)

R ₀, ES, and Ae. camptorhynchus were log-transformed for the analyses. Regressions were against residuals for Ae. camptorhynchus and O. aries to control for salinity interrelationships. Significant results adjusted for multiple comparisons using the sequential Bonferroni method and shown in bold. SE, standard error.

Critical α-values for R ₀ = 0.007 and ES = 0.006.

Discussion

RRV transmission is governed by ecological interactions between zoonotic hosts and vectors (Harley et al. 2001, Russell 2002). Accordingly, anthropogenic impacts on the environment that affect the dynamics of these interactions may influence potential for transmission. This study supports the hypothesis that salinity promotes the spatial potential for RRV transmission. We also observed temporal variability in transmission potential within our study sites. Our modeling results are best explained by salinity and variation in abundance of Ae. camptorhynchus and M. fuliginosus.

This study found abundance of Ae. camptorhynchus explained more variation in potential for RRV transmission than any other variable. Abundance of Ae. camptorhynchus also determined temporal variability in potential for RRV transmission, because there were fewer standing water bodies during late summer and autumn, reflecting low regional rainfall. Ae. camptorhynchus larvae have a broad salinity tolerance, and their abundance is positively associated with increasing salinity, irrespective of the temporal variability in their abundance, due to standing water and rainfall. It is possible the positive relationship between Ae. camptorhynchus reflects release from interspecific competition (Carver et al. submitted). It should, however, be noted that our estimates of mosquito abundance, and by extension potential for RRV transmission, are conservative. In each simulation, we assumed all mosquitoes in space were derived from our 1-hectare sampling site. We also assumed no additional mosquitoes were produced during the temporal period of each simulation. Both of these assumptions are likely to underestimate Ae. camptorhynchus abundance. As such, our estimates of potential for transmission are spatially and temporally conservative.

A possible caveat involving Ae. camptorhynchus, in this study, is when the percentage of blood meals obtained from each mammal species were added together, 16.9% of blood meals were not accounted for. We assumed these blood meals were derived from vertebrate species, which are incompetent RRV hosts. However, if some of these blood meals are derived from competent hosts, our estimates of potential for RRV transmission may increase. Our model also did not consider vertical transmission of RRV, though this is known to occur for Ae. camptorhynchus (Dhileepan et al. 1996). Vertical transmission may influence persistence of RRV (Glass 2005, Lindsay et al. 1993b), but we considered it unlikely to influence the difference between saline and nonsaline areas (the focus of this study). Finally, other mosquito species occur in the WA Wheatbelt (Lindsay et al. 2007), and it is possible these species may contribute to RRV transmission. Aedes camptorhynchus is the dominant vector in southwest WA (Lindsay et al. 1996, 2007). Other mosquito species identified in the Wheatbelt have not been shown to transmit RRV or are infrequent in occurrence, suggesting Ae. camptorhynchus is likely to be the only relevant vector mosquito (Lindsay et al. 2007).

The negative relationship of R ₀ to abundance of M. fuliginosus is in contrast to the putative reservoir status of this host. However, there is a negative correlation between the abundance of M. fuliginosus and Ae. camptorhynchus (Spearman correlation, ρ = − 0.325, p = 0.013). This correlation suggests the relationship between potential for RRV transmission and abundance of M. fuliginosus may be due to a correlation with Ae. camptorhynchus, rather than a causal relationship. It is possible M. fuliginosus may avoid areas where they are likely to receive many nuisance bites from Ae. camptorhynchus or this correlation could reflect the effect of a combination of salinity and vegetation variables on abundance of M. fuliginosus not captured in this study. Abundance of Ae. camptorhynchus, according to our model, appears to be a better predictor of potential for RRV transmission in the Wheatbelt than abundance of M. fuliginosus. In addition, O. aries comprise the majority (47.8%) of mosquito blood meals in the Wheatbelt (Jardine 2007). This mammal declined in association with increasing salinity, reflecting degradation of agricultural land and a reduction in pasture for grazing (Barrett-Lennard 2003, Dorrough et al. 2004, McFarlane and Williamson 2002, Spradbrow 1973). Ovis aries has a moderate capacity as a host for RRV but did not explain variation in the potential for RRV transmission when regressions were controlled for salinity.

Our estimates of abundance of some trapped mammals at some sites, such as T. vulpecula, M. domesticus, P. calura, and N. mitchelli, appear relatively high, reflecting high trap success. Consequently, our estimates of abundance over the 3.14 km² area, used in the modeling, were also high, and it is possible mammal abundance results may influence our estimates of potential for RRV transmission. The abundance of trapped mammals presented in this study are either within the range of other published accounts or have not previously been documented (Brockie et al. 1997, Brown et al. 1997, De Long 1967, Efford et al. 2005, Fox et al. 2003, How and Hillcox 2000, Isaac 2005, Jacob et al. 2007, Sinclair et al. 1990, Southgate and Masters 1996, Wayne et al. 2005). Notwithstanding a more detailed study of mammal abundance at each of our sties, the results here do reflect spatial differences in abundance of mammals between sites. In addition, none of these mammals were either common enough between our sites and/or had a significant effect on our measure of potential for RRV transmission.

Spatial and temporal variability in Ae. camptorhynchus blood meals may influence variability in potential for RRV transmission. In this study, our values for f were based on the total percentage of blood-engorged Ae. camptorhynchus collected in the Wheatbelt by Jardine (2007). However, Jardine (2007) did demonstrate some spatial and temporal variability within Ae. camptorhynchus blood meals. We did not incorporate this variability into our model due to difficulties in matching sites spatially and confidently attributing salinity categories. However, given dominance of Ae. camptorhynchus in explaining potential for RRV transmission, spatial and temporal variability in f may only have a minor influence on our results. It is also possible T_MH may be an overestimated parameter where subcutaneous inoculations were used, and R ₀ and ES may be influenced by the use of a central value for T_MH , σ, and γ, rather than a range of values. But again, these possible caveats do not influence our spatial comparison between sites and salinity categories.

It is important acknowledge that population cycles differ for Ae. camptorhynchus and mammal species in the Wheatbelt. In addition to salinity, results from this study reflect temporal variability in abundance of Ae. camptorhynchus, due to standing water and rainfall. However, the abundance of individual mammal species may vary over longer temporal periods (i.e., seasonal or interannual), which could influence the outcome of our model, but is beyond the timescale of this study. Further research is needed to assess how long-term cycles in the abundance of host species and communities, such as those caused by long-term drought, influence the potential for transmission of vector-borne diseases.

In contrast to studies of WNV and Lyme disease (Ezenwa et al. 2006, 2007, LoGiudice et al. 2003), we did not find an association between potential for RRV transmission and host diversity. The variability in mammal species diversity, between sites, was low (4–6) in this study. Consequently, we cannot confidently assess whether host diversity influenced potential for RRV transmission. The contribution of host diversity to RRV transmission requires more detailed investigation (Carver et al. 2008a). Additionally, the relative density and community of mammals necessary for RRV transmission (zoonotic, epizootic, or epidemic) is not well understood. Greater focus on hosts of RRV and their role in transmission is warranted (Carver et al. 2008a).

This study suggests that potential for transmission of RRV is greater in saline areas of the Wheatbelt, particularly during the months of September to November, when rainfall and Ae. camptorhynchus are common and temperatures are permissive for viral replication. Heavy rain, which facilitates abundance of Ae. camptorhynchus, may also promote potential for RRV activity at other times of the year (i.e., summer). However, human notification data does not currently reflect the salinity-RRV potential (Jardine et al. 2008). It is possible the examination by Jardine et al. (2008) of a sparse human population over a large geographical area may be an inappropriate scale of spatial variability in RRV transmission. In support, other studies have found evidence of variability in vector-borne transmission and transmission risk over smaller spatial scales (Ezenwa et al. 2006, 2007, Hu et al. 2004, Ostfeld et al. 2002, 2006, Ryan et al. 2006). In fact, our site with greatest potential for RRV transmission, DA02, occurred within the lowest saline area of Jardine et al. (2008). It is also possible salinity may influence potential for RRV transmission, but not activity. RRV appears to be episodic and nonpersistent in the Wheatbelt (Jardine et al. 2008). For example, a limited outbreak of RRV occurred in Wickepin in early 2006 but not elsewhere in the region (Jardine 2007). Comparison of RRV clinical notifications between the Wheatbelt and the endemic coastal zone indicates RRV activity in the Wheatbelt is linked to coastal epidemics and movement of infected vectors or hosts inland (Spearman correlation of published RRV notifications from Wheatbelt with southwest coastal and metropolitan Perth 1988–89, 1991–92, 1995–96, and 2001–04, ρ = 0.750, p = 0.05; Jardine et al. 2008, Lindsay et al. 1997, 2005). The Peel region, 70–130 km south of Perth, appears to be significant source from which coastal RRV activity may originate (Lindsay et al. 1996, 1997, 2005). We speculate that although salinity influences potential for RRV transmission, coastal RRV activity is a more parsimonious determinant of RRV activity in the Wheatbelt.

This study suggests dryland salinity is the proximate cause of enhanced potential for RRV transmission, but not activity, in the WA Wheatbelt. Our results indicate the ecologic determinant of this potential is predominantly enhanced abundance of Ae. camptorhynchus, which is halophilic. Other studies have also found links between vector-borne disease activity and salinity. Biggs and Mottram (2008) linked dryland salinity, abundance of Aedes vigilax Skuse (Diptera: Culicidae), and RRV notifications in Queensland; Temel (2004) identified salinity to be positively associated with malaria (Plasmodium falciparum) incidence in Azerbaijan; and Klinkenberg et al. (2004) found malaria incidence in southern Punjab was negatively related to increasing salinity and changes in the composition of Anopheles stephensi Liston (Diptera: Culicidae) and Anopheles culicifacies sensu lato (Diptera: Culicidae). Currently, more than one million hectares are saline in the Wheatbelt, and dryland salinity is expected to expand to between three and four and a half million hectares by 2050 (George et al. 2006, Jardine et al. 2007). This expansion of dryland salinity is likely to enhance the potential for RRV transmission in inland southwest WA and may influence RRV activity. Preservation and restoration of the natural environment and, by extension, intact freshwater communities may represent a valuable ecosystem approach to ameliorating potential for transmission of RRV.

Footnotes

Acknowledgments

We would like to acknowledge the contribution of several volunteers that helped with all aspects of the field work. Peter Speldewinde assisted with preparation of . Richard Ostfeld, Jesse Brunner, Andrew Jardine, and two anonymous reviewers provided valuable comments and insights that greatly improved this paper. This work followed all relevant environmental and institutional regulations in the collection of data presented here.

Disclosure Statement

This study was supported by the Ray Hart Memorial Award for Integrative research, the Australian and Jean Rogerson Postgraduate awards, and the CRC for Plant Based Management of Dryland Salinity. The Department of Environment and Conservation granted authority to enter land (CE001070) and provided licenses to survey fauna (SF005093). Approval was also granted by the University of Western Australia Animal Ethics Committee (06/100/566).

References

Aaskov

, Mataika

, Lawrence

, Rabukawaqa

et al. An epidemic of Ross River virus infection in Fiji, 1979. Am J Trop Med Hyg, 1981; 30:1053–1059.

Ballard

JWO

, Marshall

. An investigation of the potential of Aedes camptorhynchus as a vector of Ross River virus. Aust J Exp Biol Med Sci, 1986; 64:197–200.

Barrett-Lennard

. The interaction between waterlogging and salinity in higher plants: causes, consequences and implications. Plant Soil, 2003; 253:35–54.

Barton

, Aberton

. Larval development and autogeny in Ochlerotatus camptorhynchus (Thomson) (Diptera: Culicidae) from Southern Victoria. Proc Linnean Soc NSW, 2005; 126:261–267.

Barton

, Aberton

, Kay

. Spatial and temporal definition of Ochlerotatus camptorhynchus (Thomson) (Diptera: Culicidae) in the Gippsland Lakes system of eastern Victoria. Aust J Entomol, 2004; 43:16–22.

Biggs

AJW

, Mottram

. Links between dryland salinity, mosquito vectors, and Ross River virus disease in southern inland Queensland-an example and potential implications. Aust J Soil Res, 2008; 46:62–66.

Boyd

, Hall

, Gemmell

, Kay

. Experimental infection of Australian brushtail possums, Trichosurus vulpecula (Phalangeridae: Marsupialia), with Ross River and Barmah Forest viruses by use of a natural mosquito vector system. Am J Trop Med Hyg, 2001; 65:777–782.

Boyd

, Kay

. Assessment of the potential of dogs and cats as urban reservoirs of Ross River and Barmah Forest viruses. Aust Vet J, 2002; 80:83–86.

Brockie

, Ward

, Cowan

. Possums (Trichosurus vulpecula) on Hawke’s bay farmland: Spatial distribution and population structure before and after a control operation. J Roy Soc NZ, 1997; 27:181–91.

10.

Brown

, Singleton

, Kearns

, Griffiths

. Evaluation and cost-effectiveness of strychnine for control of populations of wild house mice (Mus domesticus) in Victoria. Wildlife Res, 1997; 24:159–172.

11.

Canyon

, Hii

JLK

, Muller

. Effect of diet on biting, oviposition, and survival of Aedes aegypti (Diptera : Culicidae) J Med Entomol, 1999; 36:301–308.

12.

Carron

, Bichaud

, Platz

, Bicout

. Survivorship characteristics of the mosquito Aedes caspius adults from southern France under laboratory conditions. Med Vet Entomol, 2008; 22:70–73.

13.

Carver

, Bestall

, Jardine

, Ostfeld

. The influence of hosts on the ecology of arboviral transmission: potential mechanisms influencing dengue, Murray Valley encephalitis and Ross River virus in Australia. Vector-Borne Zoonotic Dis, 2008aAhead of print10.1089/vbz.2007.0040.

14.

Carver

, Sakalidis

, Weinstein

. House mouse abundance and Ross River virus notifications in Victoria, Australia. Int J Infect Dis, 2008b. 12:528–533.

15.

Carver

. Dryland salinity, mosquitoes, mammals and the ecology of Ross River virus [Ph.D.] Perth, Western Australia: University of Western Australia, 2008.

16.

Cottingham

, Lennon

, Brown

. Knowing when to draw the line: designing more informative ecological experiments. Frontiers Ecol Environment, 2005; 3:145–152.

17.

Daszak

, Cunningham

, Hyatt

. Anthropogenic environmental change and the emergence of infectious diseases in wildlife. Acta Trop, 2001; 78:103–116.

18.

de Jong

MCM

, Diekmann

, Heesterbeek

. How does transmission of infection depend on population size? Mollison

. Epidemic Models: Their Structure and Relation to Data. Cambridge: Cambridge University Press, 1995; 84–94.

19.

de Long

. Population ecology of feral house mice. Ecol, 1967; 48:611–634.

20.

Derraik

JGB

, Slaney

. Anthropogenic environmental change, mosquito-borne diseases and human health in New Zealand. Ecohealth, 2007; 4:72–81.

21.

Dhileepan

, Azuolas

, Gibson

. Evidence of vertical transmission of Ross River and Sindbis viruses (Togaviridae, Alphavirus) by mosquitoes (Diptera, Culicidae) in southeastern Australia. J Med Entomol, 1996; 33:180–182.

22.

Dobson

. Population dynamics of pathogens with multiple host species. Am Nat, 2004; 164:S64–S78.

23.

Dorrough

, Yen

, Turner

, Clark

et al. Livestock grazing management and biodiversity conservation in Australian temperate grassy landscapes. Aust J Agric Res, 2004; 55:279–295.

24.

Efford

, Warburton

, Coleman

, Barker

. A field test of two methods for density estimation. Wildlife Soc Bul, 2005; 33:731–738.

25.

Ezenwa

, Godsey

, King

, Guptill

. Avian diversity and West Nile virus: testing associations between biodiversity and infectious disease risk. Proc R Soc Lond B Biol Sci, 2006; 273:109–117.

26.

Ezenwa

, Milheim

, Coffey

, Godsey

et al. Land cover variation and West Nile virus prevalence: patterns, processes, and implications for disease control. Vector-Borne Zoonotic Dis, 2007; 7:173–180.

27.

Foley

, Russell

, Bryan

. Population structure of the peridomestic mosquito Ochlerotatus notoscriptus in Australia. Med Vet Entomol, 2004; 18:180–190.

28.

Fox

, Taylor

, Thompson

. Experimental manipulation of habitat structure: a retrogression of the small mammal succession. J Anim Ecol, 2003; 72:927–940.

29.

George

, Clarke

, English

. Modern and palaeogeographic trends in the salinisation of the Western Australian Wheatbelt: Proceedings of the Australian Earth Sciences Convention 2006. Melbourne. http://www.earth2006.org.au/papers/extendedpdf/George%20Richard%20-%20Modern%20and%20palaeogeographic-extended.pdf. 2006 September 22.

30.

Glass

. Ecological mechanisms that promote arbovirus survival: a mathematical model of Ross River virus transmission. Trans R Soc Trop Med Hyg, 2005; 99:252–260.

31.

Halse

, Lyons

, Pinder

, Shiel

. Biodiversity patterns and their conservation in the wetlands of the Western Australia Wheatbelt. Records of the Western Australian Museum Supplement No. 67 2004; 337–364.

32.

Harley

, Ritchie

, Bain

, Sleigh

. Risks for Ross River virus disease in tropical Australia. Int J Epidemiol, 2005; 34:548–555.

33.

Harley

, Sleigh

, Ritchie

. Ross River virus transmission, infection, and disease: a cross-disciplinary review. Clin Microbiol Rev, 2001; 14:909–932.

34.

Harrington

, Buonaccorsi

, Edman

, Costero

et al. Analysis of survival of young and old Aedes aegypti (Diptera: Culicidae) from Puerto Rico and Thailand. J Med Entomol, 2001; 38:537–547.

35.

How

, Hillcox

. Brushtail possum, Trichosurus vulpecula, populations in south-western Australia: demography, diet and conservation status. Wildlife Res, 2000; 27:81–89.

36.

, Nicholls

, Lindsay

, Dale

et al. Development of a predictive model for Ross River virus disease in Brisbane, Australia. Am J Trop Med Hyg, 2004; 71:129–137.

37.

Hudson

, Norman

, Laurenson

, Newborn

et al. Persistence and transmission of tick-borne viruses: Ixodes ricinus and louping-ill virus in red grouse populations. Parasitology, 1995; 111:S49–S58.

38.

Isaac

. Life history and demographics of an island possum. Aust J Zool, 2005; 53:195–203.

39.

Jacob

, Hinds

, Singleton

, Sutherland

et al.

Is the reproductive potential of wild house mice regulated by extrinsic or intrinsic factors?

Austral Ecol, 2007; 32:202–209.

40.

Jardine

. The impact of dryland salinity on Ross River virus in south-western Australia: an ecosystem health perspective [PhD dissertation] Perth, Western Australia, Australia: University of Western Australia, 2007.

41.

Jardine

, Lindsay

, Heyworth

, Weinstein

. Dry-season mosquito breeding associated with irrigation in the northeast Kimberley region of Western Australia: potential impact on mosquito-borne disease transmission. Ecohealth, 2004; 1:387–398.

42.

Jardine

, Speldewinde

, Carver

, Weinstein

. Dryland salinity and Ecosystem Distress Syndrome: human health implications. Ecohealth, 2007; 4:10–17.

43.

Jardine

, Speldewinde

, Lindsay

, Cook

et al.

Is there an association between dryland salinity and Ross River virus disease in southwestern Australia?

Ecohealth, 2008; 5:58–68.

44.

Kanojia

, Shetty

, Geevarghese

. A long-term study on vector abundance & seasonal prevalence in relation to the occurrence of Japanese encephalitis in Gorakhpur district, Uttar Pradesh. Indian J Med Res, 2003; 117:104–110.

45.

Kay

, Aaskov

. Ross River virus (epidemic polyarthritis) Monath

. The Arboviruses: Epidemiology and Ecology. Boca Raton, F: CRC Press, Inc., 1989; 93–112.

46.

Kay

, Farrow

. Mosquito (Diptera: Culicidae) dispersal: implications for the epidemiology of Japanese and Murray Valley Encephalitis viruses in Australia. J Med Entomol, 2000; 37:797–801.

47.

Kay

, Pollitt

, Fanning

, Hall

. The experimental infection of horses with Murray Valley encephalitis and Ross River viruses. Aust Vet J, 1987a. 64:52–55.

48.

Kay

, Saul

, McCullaugh

. A mathematical model for the rural amplification of Murray Valley encephalitis virus in southern Australia. Am J Epidemiol, 1987b. 125:690–705.

49.

Keeling

, Rohani

. Modeling infectious diseases in humans and animals. Princeton, NJ: Princeton University Press, 2008.

50.

Keighery

, Halse

, McKenzie

, Harvey

. A biodiversity survey of the Western Australian agricultural zone. Perth, Western Australia: Records of the Western Australian MuseumSupplement ; no. 67, Western Australian Museum2004.

51.

Kelly-Hope

, Kay

, Purdie

, Williams

. The risk of Ross River and Barmah Forest virus disease in Queensland: implications for New Zealand. Aust N Z J Public Health, 2002; 26:69–77.

52.

Kilpatrick

, Daszak

, Jones

, Marra

et al. Host heterogeneity dominates West Nile virus transmission. Proc R Soc Lond B Biol Sci, 2006a. 273:2327–2333.

53.

Kilpatrick

, Kramer

, Jones

, Marra

et al. West Nile virus epidemics in North America are driven by shifts in mosquito feeding behavior. PLos Biol, 2006b. 4:e82.

54.

Klinkenberg

, Konradsen

, Herrel

, Mukhtar

et al. Malaria vectors in the changing environment of the southern Punjab, Pakistan. Trans R Soc Trop Med Hyg, 2004; 98:442–449.

55.

Leisnham

, Sala

, Juliano

. Geographic variation in adult survival and reproductive tactics of the mosquito Aedes albopictus. J Med Entomol, 2008; 45:210–221.

56.

Liehne

PFS

. An atlas of the mosquitoes of Western Australia. Perth, Australia: Health Department of Western Australia, 1991.

57.

Lindsay

, Oliveira

, Jasinska

, Johansen

et al. An outbreak of Ross River virus disease in southwestern Australia. Emerg Infect Dis, 1996; 2:117–120.

58.

Lindsay

, Broom

, Oliveira

, Jasinska

et al. Western Australian arbovirus surveillance and research program, Annual Report: 1997–1998. Western Australia, Australia: Arbovirus Surveillance and Research Laboratory and the Western Australian Centre for Pathology and Medical Research, 1998.

59.

Lindsay

, Jardine

, Johansen

, Wright

et al. Mosquito (Diptera: Culicidae) fauna in inland areas of south west Western Australia. Aust J Entomol, 2007; 46:60–64.

60.

Lindsay

, Mackenzie

, Condon

. Ewan

, Calvert

, Bryant

, Garrick

. Ross River virus outbreaks in Western Australia: epidemiological aspects and the role of environmental factors. Health in the Greenhouse: The Medical and Environmental Health Effects of Global Climate Change. University of Wollongong, Canberra, Australia: AGPS Press, 1993a. 85–100.

61.

Lindsay

, Oliveira

, Jasinska

, Johansen

et al. Western Australia’s largest recorded outbreak of Ross River virus disease. Arbovirus Res Aust, 1997; 7:147–152.

62.

Lindsay

MDA

, Breeze

, Harrington

, Johansen

et al. Ross River and Barmah Forest viruses in Western Australia, 2000/01–2003/04: contrasting patterns of disease activity. Arbovirus Res Aust, 2005; 9:194–201.

63.

Lindsay

MDA

, Broom

, Wright

AET

, Johansen

et al. Ross River virus isolations from mosquitoes in arid regions of Western Australia—implication of vertical transmission as a means of persistence of the virus. Am J Trop Med Hyg, 1993b. 49:686–696.

64.

Lindsay

MDA

, Coelen

, Mackenzie

. Genetic heterogeneity among isolates of Ross River virus from different geographical regions. J Virol, 1993c. 67:3576–3585.

65.

LoGiudice

, Ostfeld

, Schmidt

, Keesing

. The ecology of infectious disease: effects of host diversity and community composition on Lyme disease risk. Proc Natl Acad Sci U S A, 2003; 100:567–571.

66.

Marshall

, Miles

JAR

. Ross River virus and epidemic polyarthritis. Current Topics in Vector Research. New York: Praeger, 1984; 31–56.

67.

McBride

WJH

. Infections in travellers arriving from Australia. Trans R Soc Trop Med Hyg, 2008; 102:312–313.

68.

McFarlane

, Williamson

. An overview of water logging and salinity in southwestern Australia as related to the “Ucarro” experimental catchment. Agricul Water Management, 2002; 53:5–29.

69.

McKenzie

, Burbidge

, Rolfe

. Effect of salinity on small, ground-dwelling animals in the Western Australian Wheatbelt. Aust J Botany, 2003; 51:725–740.

70.

Molyneux

. Common themes in changing vector-borne disease scenarios. Trans R Soc Trop Med Hyg, 2003; 97:129–132.

71.

Muir

, Kay

. Aedes aegypti survival and dispersal estimated by mark-release-recapture in northern Australia. Am J Trop Med Hyg, 1998; 58:277–282.

72.

Niebylski

, Craig

. Dispersal and survival of Aedes albopictus at a scrap tire yard in Missouri. J Am Mosq Control Assoc, 1994; 10:339–343.

73.

Ostfeld

, Keesing

. The function of biodiversity in the ecology of vector-borne zoonotic diseases. Canadian J Zool, 2000a. 78:2061–2078.

74.

Ostfeld

, Canham

, Oggenfuss

, Winchcombe

et al. Climate, deer, rodents, and acorns as determinants of variation in Lyme-disease risk. PLoS Biol, 2006; 4:1058–1068.

75.

Ostfeld

, Keesing

. Biodiversity and disease risk: the case of Lyme disease. Conserv Biol, 2000b. 14:722–728.

76.

Ostfeld

, Keesing

, Schauber

, Schmidt

. Aguirre

, Ostfeld

, Tabor

, House

, Pearl

. Ecological context of Lyme disease—Biodiversity, habitat fragmentation, and risk of infection. Conservation Medicine: Ecological Health in Practice. New York: Oxford University Press, 2002; 207–219.

77.

Overgaard

, Ekbom

, Suwonkerd

, Takagi

. Effect of landscape structure on anopheline mosquito density and diversity in northern Thailand: implications for malaria transmission and control. Landscape Ecol, 2003; 18:605–619.

78.

Patz

, Graczyk

, Geller

, Vittor

. Effects of environmental change on emerging parasitic diseases. Int J Parasitol, 2000; 30:1395–1405.

79.

Rice

. Analyzing tables of statistical tests. Evolution, 1989; 43:223–225.

80.

Robertson

. Biology of the Ross River virus vector mosquito Ochlerotatus camptorhynchus in an urban environment [Honours thesis] Perth, Australia: University of Western Australia, 2006.

81.

Russell

. Seasonal activity and abundance of the arbovirus vector Culex annulirostris Skuse near Echuca, Victoria, in the Murray Valley of southeastern Australia 1979–1985. Aust J Exp Biol Med Sci, 1986; 64:97–103.

82.

Russell

. Mosquito-borne arboviruses in Australia: the current scene and implications of climate change for human health. Int J Parasitol, 1998; 28:955–969.

83.

Russell

. Constructed wetlands and mosquitoes: Health hazards and management options—an Australian perspective. Ecol Engineering, 1999; 12:107–124.

84.

Russell

. Ross River virus: ecology and distribution. Annu Rev Entomol, 2002; 47:1–31.

85.

Ryan

, Alsemgeest

, Gatton

, Kay

. Ross River virus disease clusters and spatial relationship with mosquito biting exposure in Redland Shire, southern Queensland, Australia. J Med Entomol, 2006; 43:1042–1059.

86.

Ryan

, Alsemgeest

, Thomas

, Kay

. Heterogeneity of risk of Ross River virus disease: implications for residential development and vector control programs. Arbovirus Res Aust, 2001; 8:331–336.

87.

Ryan

, Do

, Kay

. Spatial and temporal analysis of Ross River virus disease patterns at Maroochy Shire, Australia: association between human morbidity and mosquito (Diptera : Culicidae) abundance. J Med Entomol, 1999; 36:515–521.

88.

Ryan

, Martin

, Mackenzie

, Kay

. Investigation of gray-headed flying foxes (Pteropus poliocephalus) (Megachiroptera: Pteropodidae) and mosquitoes in the ecology of Ross River virus in Australia. Am J Trop Med Hyg, 1997; 57:476–482.

89.

Seay

, Diane

, Griffin

, Johnson

. Experimental viral polymyositis: age dependency and immune responses to Ross River virus infection in mice. Neurology, 1981; 31:656–661.

90.

Sinclair

ARE

, Olsen

, Redhead

. Can predators regulate small mammal populations—evidence from house mouse outbreaks in Australia. Oikos, 1990; 59:382–392.

91.

Southgate

, Masters

. Fluctuations of rodent populations in response to rainfall and fire in a central Australian hummock grassland dominated by Plectrachne schinzii. Wildlife Res, 1996; 23:289–303.

92.

Spradbrow

. Experimental infection of sheep and pigs with Ross River virus. Aust Vet J, 1973; 49:403–404.

93.

Temel

. Malaria from the gap: need for cross-sector. co-operation in Azerbaijan. Acta Trop, 2004; 89:249–259.

94.

Tirados

, Costantini

, Gibson

, Torr

. Blood-feeding behaviour of the malarial mosquito Anopheles arabiensis: implications for vector control. Med Vet Entomol, 2006; 20:425–437.

95.

van den Hurk

, Johansen

, Zborowski

, Paru

et al. Mosquito host-feeding patterns and implications for Japanese encephalitis virus transmission in northern Australia and Papua New Guinea. Med Vet Entomol, 2003; 17:403–411.

96.

van Gool

, Moore

. Land Evaluation Standards for Land Resource Mapping. Resource Management Technical Report 181, 2nd. Perth: Agriculture Western Australia, 1999.

97.

Walsh

, Molyneux

, Birley

. Deforestation: effects on vector-borne disease. Parasitology, 1993; 106:S55–S75.

98.

Watson

, Saul

, Kay

. Aedes notoscriptus (Diptera : Culicidae) survival and dispersal estimated by mark-release-recapture in Brisbane, Queensland, Australia. J Med Entomol, 2000; 37:380–384.

99.

Wayne

, Cowling

, Ward

, Rooney

et al. A comparison of survey methods for arboreal possums in Jarrah Forest, Western Australia. Wildlife Res, 2005; 32:701–714.

100.

Whitehead

. Experimental infection of vertebrates with Ross River and Sindbis viruses, two group A arboviruses isolated in Australia. Aust J Exp Biol Med Sci, 1969; 47:11–15.