Abstract

Koala (Phascolarctos cinereus) populations in the wild are in sharp decline in Australia due to deforestation, road accidents, dog attacks, and disease from infection with sexually transmitted Chlamydia spp. Severely diseased koalas that are captured are euthanized for humane reasons because antibiotics are not effective. Paradoxically, we propose that euthanizing more koalas could help to increase koala population numbers. We investigated the potential impact of systematically euthanizing diseased koalas. Using data from a well-studied koala population, and an individual-based computer simulation model, we predict that such a program would result in a larger population of koalas after 7 yr than would exist without the program. If terminally diseased and sterile koalas are euthanized and other infected captured koalas are given antibiotics, chlamydial infection could be eliminated and population growth observed after 4 yr. The practical implementation of such a program would be facilitated with further development of tools to diagnose infection and internal disease in the field.

INTRODUCTION

The koala (Phascolarctos cinereus) is an iconic Australian animal, attracting many tourists to wildlife sanctuaries. However, the number of koalas in the wild is rapidly declining across the continent. The Australian state of Queensland had the largest koala population in 1990, with an estimated 295,000; in just 20 yr that number has decreased by over 40% (Threatened Species Scientific Committee 2012). The Queensland Koala Coast population decreased in size by 68% between 1996 and 2010 (Department of Environment and Resource Management 2010). While habitat degradation, dog attacks, and road accidents contribute to koala population decline, reducing disease burden is probably the most essential requirement for population stability (Rhodes et al. 2011). An important disease affecting koalas is caused by sexually transmitted Chlamydia spp. Koalas can be infected by both Chlamydia pecorum and Chlamydia pneumoniae; the former is thought to have crossed to koalas from introduced cattle and sheep during the last 300 yr, whereas the latter has probably existed in Australia for much longer (Polkinghorne et al. 2013). Chlamydia pecorum is the more pathogenic of the two species and the one to which we refer in this study.

Polkinghorne et al. (2013) found that the median prevalence of chlamydial infection was almost 50%. Chlamydia spp. are efficiently transmitted sexually (Katz 1992; Althaus et al. 2012), and in koalas the infection causes infertility and blindness, and can lead to death (Brown and Grice 1986; Brown et al. 1987). Visible signs of infection include “dirty tail,” which is caused by urinary tract infections and incontinence.

Sick or injured koalas are sometimes brought by the public or field workers into animal care centers for treatment. Animals from populations facing impact from human development are also captured and monitored. If these captured animals have disease they may be treated with antibiotics, requiring daily injections of antibiotics for 14–28 d, along with close monitoring and feeding while they remain in the care facility. When they are terminally diseased, they may be euthanized to prevent further suffering. The euthanasia of a koala with chlamydial infection will have secondary benefits for the population, as the koala will not be able to infect other koalas with the disease-causing bacteria. As a corollary, culling of koalas that have disease beyond recovery and who still have chlamydial infection (i.e., beyond those who would normally be euthanized) may have a positive effect on the long-term population size. To investigate this hypothesis we used a computer simulation model of a typical koala population with endemic chlamydial infection, based on the Queensland Koala Coast population.

MATERIALS AND METHODS

The computer simulation model we employed has been used elsewhere to describe koala population dynamics and help plan for the implementation of future vaccination programs (Craig et al. 2014). The model is stochastic and individual-based, run with monthly time steps. A schematic diagram of the infection and disease processes is presented in Figure 1. All koalas in the simulation model are categorized as 1) uninfected and potentially susceptible to infection (or reinfection), 2) infected, or 3) recovered from a past infection and currently temporarily immune. For each of these infection states, koalas may be in one of four disease states: 1) not diseased, 2) sterile, 3) terminal, or 4) sterile and terminal. Thus, there are 12 combinations of infection and disease states that define the health of koalas in the simulation model (Fig. 1). We note that this disease terminology differs from that in the study in which this model was first introduced (Craig et al. 2014) but the underlying model is the same. We also note that it is possible for a koala to be susceptible to infection (i.e., not currently infected) and still have disease, namely, residual disease (e.g., from induced tissue scarring that can occur from the immune response to an infection), which remains after resolution of infection(s).

Figure 1. 

Schematic of the structure of a computer simulation model of the infection and disease process for Australian koalas (Phascolarctos cinereus) infected with Chlamydia pecorum showing the movement of numbers of koalas in the population in various infection and disease states. See Table 1 for parameter descriptions. Death is not explicitly shown; it may occur from any state, based on natural lifespan for a nonterminal koala and on a separate rate for a terminally diseased koala.

Figure 1. 

Schematic of the structure of a computer simulation model of the infection and disease process for Australian koalas (Phascolarctos cinereus) infected with Chlamydia pecorum showing the movement of numbers of koalas in the population in various infection and disease states. See Table 1 for parameter descriptions. Death is not explicitly shown; it may occur from any state, based on natural lifespan for a nonterminal koala and on a separate rate for a terminally diseased koala.

Koalas enter the simulated population as susceptible or infected (from vertical transmission). Once in the population, they may acquire chlamydial infection through sexual transmission (described below). Infection lasts for 10–18 mo, after which the koala has recovered and has temporary partial immunity from reinfection; their probability of reinfection is reduced by 20–60% for 6–12 mo (Table 1). Infected females may become temporarily infertile, with a probability of 1–80%, for the duration of their infection. Additionally, upon acquiring chlamydial infection, some females may develop permanent sterility, and males and females may develop terminal disease with a probability of 1–42%. Terminal disease is fatal after 1–6 mo.

Table 1. 

Model input infection parameters of a computer simulation model of the infection and disease process for Australian koalas (Phascolarctos cinereus) infected with Chlamydia pecorum. Parameters appear in Figure 1. Ranges before calibration indicate the ranges from which values were sampled for the calibration process; the range of parameter values used was identified after passing through a calibration process of reconciliation between model projections and observational data.

Model input infection parameters of a computer simulation model of the infection and disease process for Australian koalas (Phascolarctos cinereus) infected with Chlamydia pecorum. Parameters appear in Figure 1. Ranges before calibration indicate the ranges from which values were sampled for the calibration process; the range of parameter values used was identified after passing through a calibration process of reconciliation between model projections and observational data.
Model input infection parameters of a computer simulation model of the infection and disease process for Australian koalas (Phascolarctos cinereus) infected with Chlamydia pecorum. Parameters appear in Figure 1. Ranges before calibration indicate the ranges from which values were sampled for the calibration process; the range of parameter values used was identified after passing through a calibration process of reconciliation between model projections and observational data.

For intervention in practice, koalas may be diagnosed with a chlamydial infection using a PCR test or an enzyme immunoassay, with the latter better suited to use in the wild, albeit with some challenges (Hanger et al. 2013). Terminal disease is readily apparent but techniques such as ultrasound may be necessary to diagnose sterility (Markey et al. 2007). Infection can be treated with antibiotics and it is generally accepted that there is complete resolution of infection in the vast majority of animals (assumed to be 100% in this analysis). However, sterility and terminal disease do not revert upon resolution of infection either naturally or through antibiotic treatment. It is assumed that disease develops with infection; the only experimental transmission study using koalas found that almost all developed disease within 1 mo (the time step of the model) (Brown and Grice 1986).

In each simulated month in the model, sexual partnerships are determined randomly and if one of the sexual partners is already infected they may transmit the infection to the partner. The chance of breeding for male koalas is dependent on the animal’s weight (Ellis and Bercovitch 2011); we established weight growth curves as a function of age for each koala based on data from 95 age and weight measurements for 38 koalas captured in the wild near Brisbane, Queensland, using methods described by Jones et al. (2012). In the model, a female does not mate while pregnant or with a dependent joey. If a mother with a dependent joey dies, then the joey is simulated to die also. The fertility rates of female koalas were taken by scaling the distribution of births per month recorded for Southeast Queensland (Ellis et al. 2010). The general life stages of growth, fertility, and mortality for male and female koalas are shown in Supplementary Material Figure S1.

Model parameter values were informed by the published literature and field data (see Table 1 for infection parameters, and Supplementary Material Table S1 for demographic parameters); the distributions of all nonconstant model parameters were assumed to be uniform. For each simulation-month, the various processes are resolved in the following order: captures, breeding and determination of new infections, deaths, births, and then the recording of the population status (number of koalas, prevalence etc.). The model was implemented in Matlab, using a modified third-party function (Edwards 2009). The model code and input data are publicly available (Craig and Wilson 2015). Model calibration involved choosing parameter sets from the parameter space, and using a two-stage Monte Carlo filtering process: 1) demographic calibration: simulating chlamydia-free populations, retaining only demographic parameter sets that resulted in the populations remaining stable or growing with a doubling time no greater than 2.7 yr (Martin and Handasyde 1990; the fastest doubling time we found in the literature); 2) disease calibration: simulating an infected population over the 10 yr preceding 2014, retaining disease and demographic parameter sets that produced a population of around 1,350 koalas in 2014 (based on previous population estimates), along with a population halving time of 5–10 yr, with chlamydial infection and disease (as defined by Craig et al. [2014]) prevalence between 35% and 70%, as is estimated for the field trial population. Parameter sets were chosen from the parameter space using Latin hypercube sampling (McKay et al. 1979), because of its efficiency compared with completely random sampling and sampling on a grid. In addition to this variation in the parameter values, there is also stochasticity within each model simulation (e.g., in determining whether transmission actually occurs for a given mating event between infected and uninfected koalas).

Of the 100,000 parameter sets, 66,927 passed the demographic calibration and 127 of these also passed disease calibration. Of the parameter sets that passed both stages of calibration, 100 were randomly selected for simulation of different scenarios; the ranges of the parameter values in these 100 sets are shown in Table 1 and Supplementary Material Table S1. For all of these parameter sets, in the absence of an intervention, population extinction did not occur but was close in some simulations with just four koalas after 20 yr (in 2034). We evaluated the sensitivity of the forecasted population size in year 2034 to the model parameters by using SaSAT software to calculate partial rank correlation coefficients (Hoare et al. 2008).

In addition to projecting continuation of current practices (“no intervention”), we assessed the expected potential of the following three intervention programs: 1) “cull only”: whereby sterile, terminal, and infected koalas are euthanized; 2) “treat only”: whereby sterile, terminal, and infected koalas are treated with antibiotics; and 3) “cull or treat”: whereby sterile and terminal koalas are euthanized and infected koalas who are not sterile or terminal are treated with antibiotics. Such programs might be considered in koala populations that are relatively small and therefore amenable to capturing animals to reach sufficient coverage for impact. We consider that it is feasible to cull/treat up to 10% of the current population each year (i.e., around 130–140 koalas), or the number of koalas that are eligible for culling/treatment, whichever is fewer. That is, we assumed that approximately the same number of koalas (130–140) would be culled/treated each year. Thus, in a declining population a larger proportion of the population would be captured each year. We also investigated lower culling/treatment rates of 5% (around 65–70 koalas) and 2.5% (around 32–35 koalas) of the population each year. Captures were simulated to occur over two intensive efforts at different times of the year. The model was used to simulate what is expected to occur if these programs were implemented. Each scenario was simulated once for each of the 100 parameter sets.

RESULTS

We compare model projections of the expected koala population sizes with and without an intervention in Figure 2 (point-wise medians only; medians with 5th and 95th quantiles are shown in Supplementary Material Fig. S2). Under no intervention, we estimate that by the year 2030 there would be approximately 185 koalas remaining in this population. Under the “cull only” intervention, there would be a sudden decrease in the number of koalas that would be observed over approximately the first 4 yr of the intervention; however, after 7 yr there would be more koalas in the population than there would be under no intervention. The “cull or treat” intervention would be even more successful, with the population size overtaking the no-intervention population size after 4 yr. In all of the 100 simulations of the “cull or treat” intervention, the koala population size overtakes the population size for the “treat only” intervention between 2 and 7 yr after the start of the intervention. We argue that, paradoxically, an effective way to save koala populations from extinction is to cull koalas.

Figure 2. 

The projected numbers of koalas (Phascolarctos cinereus) in a population infected with Chlamydia pecorum, without intervention and according to several intervention programs. The model values from 2004 are the point-wise medians from 100 model simulations, with the model values prior to 2004 being a cubic spline interpolation. The population data are from the Koala Coast in the Australian state of Queensland.

Figure 2. 

The projected numbers of koalas (Phascolarctos cinereus) in a population infected with Chlamydia pecorum, without intervention and according to several intervention programs. The model values from 2004 are the point-wise medians from 100 model simulations, with the model values prior to 2004 being a cubic spline interpolation. The population data are from the Koala Coast in the Australian state of Queensland.

The reason for the population increase under the “cull or treat” intervention can be seen in Figure 3; specifically, the prevalence of chlamydial infection would decrease substantially. Indeed, simulations reveal that there is a strong chance of eliminating chlamydia from the population within 4 yr under the “cull or treat” intervention. Of the 100 simulations for the “cull or treat” intervention, the earliest that the prevalence drops to zero is during year 2, and the latest is during year 5 (see Supplementary Material Fig. S3 for uncertainty results). Lower prevalence of chlamydial infection results in reduced incidence of new infections and therefore reduced mortality and fertility, and this is the cause of the change in population dynamics. Under no intervention, the prevalence remains roughly constant even as the population size decreases, as expected given that the model structure assumes frequency-dependent transmission.

Figure 3. 

Projected chlamydia prevalence in Australian koalas (Phascolarctos cinereus) in a population infected with Chlamydia pecorum, without intervention and according to several intervention programs. The values shown are the point-wise medians from 100 model simulations.

Figure 3. 

Projected chlamydia prevalence in Australian koalas (Phascolarctos cinereus) in a population infected with Chlamydia pecorum, without intervention and according to several intervention programs. The values shown are the point-wise medians from 100 model simulations.

Currently, the only real alternative to culling for control of chlamydial infection and reversing population decline in koalas is through the widespread use of antibiotics. This strategy is highly resource-intensive and in practice requires in-facility monitoring and daily treatment for up to a month. We found that a “treat only” intervention can eliminate chlamydial infection in around 7 yr compared with 4 yr for a “cull or treat” intervention (Fig. 3); interventions involving only antibiotic treatment are projected to result in a smaller koala population size after 20 yr compared with either the “cull only” or “cull or treat” interventions (Fig. 2).

Sensitivity analysis reveals that the koala population size after 20 yr (in 2034) under no intervention is most sensitive to the annual survivorship of adult koalas, the probability of an infected koala developing terminal disease, the probability of an infected koala developing sterility, the probability of breeding producing young, and the probability of transmission from an infected male to a female (Table 2).

Table 2. 

Results of a sensitivity analysis of a computer simulation model of the infection and disease process for Australian koalas (Phascolarctos cinereus) infected with Chlamydia pecorum. The parameters shown are those to which the predicted koala population size in year 2034 under no intervention is most sensitive, as determined by partial rank correlation coefficients. Only the 10 parameters with the strongest correlation (>0.3) are shown.

Results of a sensitivity analysis of a computer simulation model of the infection and disease process for Australian koalas (Phascolarctos cinereus) infected with Chlamydia pecorum. The parameters shown are those to which the predicted koala population size in year 2034 under no intervention is most sensitive, as determined by partial rank correlation coefficients. Only the 10 parameters with the strongest correlation (>0.3) are shown.
Results of a sensitivity analysis of a computer simulation model of the infection and disease process for Australian koalas (Phascolarctos cinereus) infected with Chlamydia pecorum. The parameters shown are those to which the predicted koala population size in year 2034 under no intervention is most sensitive, as determined by partial rank correlation coefficients. Only the 10 parameters with the strongest correlation (>0.3) are shown.

In practice, budget constraints and other practical issues will make annual capture of 10% of the population difficult, but our main findings hold even under capture rates of 5% and 2.5% (Fig. 4). A capture rate of 5% produces a reversal in population decline after around 6 yr under the “cull or treat” intervention (compared with a reversal of population decline after around 4 yr at a 10% capture rate). Under a 2.5% capture rate, reversal of population decline would be delayed until around 10 yr.

Figure 4. 

The projected numbers of koalas (Phascolarctos cinereus) in a population infected with Chlamydia pecorum, without intervention and according to several intervention programs, with (A) 5% of koalas captured annually (as a proportion of the initial population), and (B) 2.5% of koalas capture annually (as a proportion of the initial population). The model values from 2004 are the point-wise medians from 100 model simulations, with the model values prior to 2004 being a cubic spline interpolation. The population data are from the Koala Coast in the Australian state of Queensland.

Figure 4. 

The projected numbers of koalas (Phascolarctos cinereus) in a population infected with Chlamydia pecorum, without intervention and according to several intervention programs, with (A) 5% of koalas captured annually (as a proportion of the initial population), and (B) 2.5% of koalas capture annually (as a proportion of the initial population). The model values from 2004 are the point-wise medians from 100 model simulations, with the model values prior to 2004 being a cubic spline interpolation. The population data are from the Koala Coast in the Australian state of Queensland.

DISCUSSION

Culling of koalas is a controversial subject. Its use was considered on Kangaroo Island in South Australia, where a chlamydia-free population of introduced koalas increased until it overbrowsed its food trees. However, there was vocal public opposition to culling and ultimately a sterilization and translocation program was implemented (Duka and Masters 2005). Although euthanasia of very sick animals for ethical reasons is generally accepted, a serious proposal of a program to capture and euthanize koalas would likely be met with opposition. However, in 2015 it was reported that, in the preceding 2 yr, almost 700 koalas were secretly culled in the Australian state of Victoria (ABC 2015).

It has been estimated that decline in koala populations can be completely reversed only if there was 100% forest restoration, no vehicle collisions with koalas, and no dog attacks on koalas, or if chlamydial disease mortality is decreased by about 60% (Rhodes et al. 2011). Alternatives to a culling program as we have proposed for the reduction of chlamydial infection prevalence include vaccines (Craig et al. 2014) and mass antibiotic treatment across the vast majority of the population. Mass treatment with antibiotics would be logistically difficult due to the resources required to treat the majority of the population over a few weeks. The development of antibiotic resistance following widespread use is also a possibility, although stable resistance in Chlamydia species in vivo is very rare (Sandoz and Rockey 2010). The ultimate goal of a vaccine may convey lasting protection, removing the need to treat all koalas in a limited time, and could be administered in the field. Significant advances in the development of a chlamydial vaccine for koalas have been made (Carey et al. 2010; Kollipara et al. 2012, 2013a, b, c), which could be utilized on its own or in conjunction with other management strategies. We argue that, in the absence of a vaccine, intensive capturing programs that lead to culling of terminally diseased and sterile koalas and antibiotic treatment of infected koalas could be the most effective approach to reversing population declines. The optimal approach in the future may be a combination of vaccination, antibiotic treatment, and culling, along with conservation of habitats.

Other euthanasia programs of infected or diseased animals have been effective. Four million cattle were destroyed in the US between 1917 and 1940 in a successful campaign to prevent bovine tuberculosis from spreading to humans (Olmstead and Rhode 2004). In 2001, a culling campaign was used to successfully control mad cow disease (Tildesley et al. 2009). Culling of the Tasmanian devil was used unsuccessfully to combat facial tumour disease (Lachish et al. 2010); modelling suggests that the rate of culling necessary to eliminate the disease would be impractically high (Beeton and McCallum 2011).

There would be practical challenges in implementing culling programs. Identifying the koalas that should be culled or treated would not be trivial. Although terminally diseased koalas are clearly unwell, it is not always possible to identify without hospital equipment the internal disease that causes sterility (Markey et al. 2007). It would clearly be impractical to regularly hospitalize a sizeable portion of the population to test them for disease.

Additionally, identifying infected koalas, which we predict is necessary for effective culling programs, would require diagnostic tests to be performed until the required number of these koalas was found. This is possible with enzyme immunoassays but is not simple to implement in the wild (Hanger et al. 2013). Thus, the practical implementation of these interventions would be facilitated with further development of the tools to assist in diagnosis of infection and internal disease in the field. This is a further argument for vaccination if and when it becomes possible. Previous modelling found that vaccines would be most effective if given to young females (Craig et al. 2014), which could be identified visually without the need for diagnostic tests.

Some limitations of the individual-based simulation model we used, its parameter values, and its data, have been discussed (Craig et al. 2014). In addition, we assumed that infected koalas will be equally infectious irrespective of whether they are diseased, but more shedding may occur in animals with chronic, active disease (Wan et al. 2011). Another limitation is the lack of data to inform the infection and disease parameters. Although cross-sectional prevalence studies have been conducted, longitudinal transmission experiments with koalas are impossible in practice. To reflect the lack of certainty in infection and disease parameters we used wide ranges for their estimates (Table 1 and Supplementary Material Table S1). The sensitivity analysis (Table 2) showed that the three parameters that most strongly influenced the longer-term size of the koala population under no intervention were the annual probability of surviving for a healthy adult (greater values obviously result in a larger population), and the probabilities of developing terminal disease or sterility (greater values obviously result in a smaller population). Better estimates of these parameters will improve the accuracy of predictions of both the population size under no intervention or various intervention strategies. Any or all of these values may vary between populations (e.g., the survivorship parameter implicitly includes death by dog attack and car impact, which we would expect to differ by region). A population in rapid decline with low disease and infection prevalence might decrease even in the absence of chlamydia. Even at that the extremes of the parameter ranges we considered, our general conclusions are robust: for the “cull or treat” intervention, all of the simulations between the 5th and 95th percentiles result in population growth compared with none of the simulations under no intervention (Supplementary Material Fig. S2). Therefore, our results can be considered broadly robust to the parameter uncertainty.

We recommend that if decision-makers choose to implement a culling program in practice, special effort should be made to estimate the parameters to which the results are most sensitive for the population in question. For example, the population could be monitored prior to the intervention, and the model recalibrated to the specific population numbers and chlamydial infection prevalence of that population to yield more precise estimates of the parameters. Such surveillance and monitoring should continue during the implementation of the intervention to confirm that the initial population decline caused by culling is accompanied by a reduction in chlamydial infection prevalence.

In summary, we predict that aggressively culling koalas affected by chlamydia would initially cause a sharp decrease in the population size, but then lead to population increases after around 4 yr. The most effective approach would be to cull terminally diseased and sterile koalas while treating infected koalas with antibiotics. We hope to see further innovation in programs designed to halt the decline of koala populations and safeguard their stability into the future.

ACKNOWLEDGMENTS

This study was funded by a Queensland University of Technology/Department of Employment, Economic Development and Innovation—National and International Research Alliances Program Shared Grant/Subcontract Australia—Canada–India Chlamydia Research Alliance Improved detection treatment and control of chlamydial infections and the Australian National Health and Medical Research Council. The Kirby Institute is funded by the Australian Government, Department of Health. The views expressed in this publication do not necessarily represent the position of the Australian Government. The Kirby Institute is affiliated with the University of New South Wales. We thank the two anonymous reviewers and Assistant Editor Michael D. Samuel for their comments, which greatly improved the manuscript.

SUPPLEMENTARY MATERIAL

Supplementary material for this article is online at http://dx.doi.org/10.7589/2014-12-278.

LITERATURE CITED

ABC
.
2015
.
Starving koalas secretly culled at Cape Otway, ‘overpopulation issues’ blamed for ill health
. .
Althaus
CL
,
Heijne
JCM
,
Low
N.
2012
.
Towards more robust estimates of the transmissibility of Chlamydia trachomatis
.
Sex Transm Dis
39
:
402
404
.
Beeton
N
,
McCallum
H.
2011
.
Models predict that culling is not a feasible strategy to prevent extinction of Tasmanian devils from facial tumour disease
.
J Appl Ecol
48
:
1315
1323
.
Brown
AS
,
Girjes
AA
,
Lavin
MF
,
Timms
P
,
Woolcock
JB.
1987
.
Chlamydial disease in koalas
.
Aust Vet J
64
:
346
350
.
Brown
AS
,
Grice
RG.
1986
.
Experimental transmission of Chlamydia psittaci in the koala
.
In:
Chlamydial infections: Proceedings of the sixth international symposium on human chlamydial infections,
.
Sanderstead, Surrey, 15–21 June;
Cambridge University Press
,
Cambridge, UK
, pp.
349
352
.
Carey
AJ
,
Timms
P
,
Rawlinson
G
,
Brumm
J
,
Nilsson
K
,
Harris
JM
,
Beagley
KW.
2010
.
A multi-subunit chlamydial vaccine induces antibody and cell-mediated immunity in immunized koalas (Phascolarctos cinereus): Comparison of three different adjuvants
.
Am J Reprod Immunol
63
:
161
172
.
Craig
AP
,
Hanger
J
,
Loader
J
,
Ellis
WAH
,
Callaghan
J
,
Dexter
C
,
Jones
D
,
Beagley
KW
,
Timms
P
,
Wilson
DP.
2014
.
A 5-year Chlamydia vaccination programme could reverse disease-related koala population decline: Predictions from a mathematical model using field data
.
Vaccine
32
:
4163
4170
.
Craig
AP
,
Wilson
DP.
2015
.
GitHub KoalaMatLabCode
. .
Department of Environment and Resource Management, The State of Queensland
.
2010
.
Koala Coast: Koala population report 2010
. .
Duka
T
,
Masters
P.
2005
.
Confronting a tough issue: Fertility control and translocation for over-abundant koalas on Kangaroo Island, South Australia
.
Ecol Manag Restor
6
:
172
181
.
Ellis
W
,
Bercovitch
F.
2011
.
Body size and sexual selection in the koala
.
Behav Ecol Sociobiol
65
:
1229
1235
.
Ellis
W
,
Bercovitch
F
,
FitzGibbon
S
,
Melzer
A
,
de Villiers
D
,
Dique
D.
2010
.
Koala birth seasonality and sex ratios across multiple sites in Queensland, Australia
.
J Mammal
91
:
177
182
.
Hanger
J
,
Loader
J
,
Wan
C
,
Beagley
KW
,
Timms
P
,
Polkinghorne
A.
2013
.
Comparison of antigen detection and quantitative PCR in the detection of chlamydial infection in koalas (Phascolarctos cinereus)
.
Vet J
195
:
391
393
.
Hoare
A
,
Regan
D
,
Wilson
D.
2008
.
Sampling and sensitivity analyses tools (SaSAT) for computational modelling
.
Theor Biol Med Model
5
:
4
.
Jones
D
,
Dexter
C
,
Bernede
L
,
Scott
J
,
Sullivan
K
,
Pickvance
J
,
Cousins
S.
2012
.
Koala retrofit works program—Evaluation and monitoring report
.
Report to Department of Transport and Main Roads.
Prepared by Applied Road Ecology Group, Environmental Futures Centre, Griffith University
,
Brisbane, Queensland, Australia
, pp.
11
18
.
Katz
BP.
1992
.
Estimating transmission probabilities for chlamydial infection
.
Stat Med
11
:
565
577
.
Kollipara
A
,
George
C
,
Hanger
J
,
Loader
J
,
Polkinghorne
A
,
Beagley
K
,
Timms
P.
2012
.
Vaccination of healthy and diseased koalas (Phascolarctos cinereus) with a Chlamydia pecorum multi-subunit vaccine: Evaluation of immunity and pathology
.
Vaccine
30
:
1875
1885
.
Kollipara
A
,
Polkinghorne
A
,
Beagley
KW
,
Timms
P.
2013a
.
Vaccination of koalas with a recombinant Chlamydia pecorum major outer membrane protein induces antibodies of different specificity compared to those following a natural live infection
.
PLoS One
8
(
):
e74808
.
Kollipara
A
,
Polkinghorne
A
,
Wan
C
,
Kanyoka
P
,
Hanger
J
,
Loader
J
,
Callaghan
J
,
Bell
A
,
Ellis
W
,
Fitzgibbon
S
,
et al
.
2013b
.
Genetic diversity of Chlamydia pecorum strains in wild koala locations across Australia and the implications for a recombinant C. pecorum major outer membrane protein based vaccine
.
Vet Microbiol
167
:
513
522
.
Kollipara
A
,
Wan
C
,
Rawlinson
G
,
Brumm
J
,
Nilsson
K
,
Polkinghorne
A
,
Beagley
K
,
Timms
P.
2013c
.
Antigenic specificity of a monovalent versus polyvalent MOMP based Chlamydia pecorum vaccine in koalas (Phascolarctos cinereus)
.
Vaccine
31
:
1217
1223
.
Lachish
S
,
McCallum
H
,
Mann
D
,
Pukk
CE
,
Jones
ME.
2010
.
Evaluation of selective culling of infected individuals to control Tasmanian devil facial tumor disease
.
Conserv Biol
24
:
841
851
.
Markey
B
,
Wan
C
,
Hanger
J
,
Phillips
C
,
Timms
P.
2007
.
Use of quantitative real-time PCR to monitor the shedding and treatment of chlamydiae in the koala (Phascolarctos cinereus)
.
Vet Microbiol
120
:
334
342
.
Martin
R
,
Handasyde
K.
1990
.
Population dynamics of the koala (Phascolarctos cinereus) in southeastern Australia
.
In:
Biology of the koala
.
Lee
AK
,
Handasyde
KA
,
Sanson
GD
,
editors
.
Surrey Beatty and Sons
,
Sydney, Australia
, pp.
75
84
.
McKay
MD
,
Beckman
RJ
,
Conover
WJ.
1979
.
Comparison of three methods for selecting values of input variables in the analysis of output from a computer code
.
Technometrics
21
:
239
245
.
Morrison
RP
,
Caldwell
HD.
2002
.
Immunity to murine chlamydial genital infection
.
Infect Immun
70
:
2741
2751
.
Olmstead
AL
,
Rhode
PW.
2004
.
An impossible undertaking: The eradication of bovine tuberculosis in the United States
.
J Econ Hist
64
:
734
772
.
Polkinghorne
A
,
Hanger
J
,
Timms
P.
2013
.
Recent advances in understanding the biology, epidemiology and control of chlamydial infections in koalas
.
Vet Microbiol
165
:
214
223
.
Rhodes
JR
,
Ng
CF
,
de Villiers
DL
,
Preece
HJ
,
McAlpine
CA
,
Possingham
HP.
2011
.
Using integrated population modelling to quantify the implications of multiple threatening processes for a rapidly declining population
.
Biol Conserv
144
:
1081
1088
.
Sandoz
KM
,
Rockey
DD.
2010
.
Antibiotic resistance in Chlamydiae
.
Future Microbiol
5
:
1427
1442
.
Threatened Species Scientific Committee
.
2012
.
Listing advice for Phascolarctos cinereus (koala)
.
Tildesley
MJ
,
Bessell
PR
,
Keeling
MJ
,
Woolhouse
MEJ.
2009
.
The role of pre-emptive culling in the control of foot-and-mouth disease
.
Proc R Soc Lond B Biol Sci
276
:
323
3248
.
Wan
C
,
Loader
J
,
Hanger
J
,
Beagley
K
,
Timms
P
,
Polkinghorne
A.
2011
.
Using quantitative polymerase chain reaction to correlate Chlamydia pecorum infectious load with ocular, urinary and reproductive tract disease in the koala (Phascolarctos cinereus)
.
Aust Vet J
89
:
409
412
.

Supplementary data