See discussions, stats, and author profiles for this publication at:
Can we mitigate cane toad impacts on northern quolls?
Technical Report · June 2015
DOI: 10.13140/RG.2.2.20438.24643
5 authors, including:
Some of the authors of this publication are also working on these related projects:
The response of the Kangaroo Island dunnart to a feral cat control program View project
Ecology of an arboreal marsupial in the tropical savannas of northern Australia View project
Sarah Legge
Australian Wildlife Conservancy
Katherine Tuft
Arid Recovery
Teigan Cremona
Charles Darwin University
All content following this page was uploaded by Teigan Cremona on 15 September 2016.
The user has requested enhancement of the downloaded file.
Report Title
Author’s name
Can we mitigate cane toad impacts
on northern quolls? | Final report
By Jonathan Webb, Sarah Legge, Katherine Tuft, Teigan Cremona
and Caitlin Austin

We thank Damien Stanioch, Lynda Veyret, Dion Wedd, and the Territory Wildlife Park staff, in
particular Sarah Hirst and Noel Riessen for raising quolls, training, and assistance in the feld. We
particularly thank the Traditional Owners for supporting the project and allowing us to work in
the East Alligator region. We are indebted to the Kakadu Rangers, trainees, and all the volunteers,
especially Sandra Riemer, Libby Dwyer, Belinda McCarthy, and Tim Dempster, who provided assistance
in the feld. We thank Steve Winderlich, Anne O’Dea, Greg Sattler, Matthew Dunn, and Patrick
Shaughnessy for advice and logistical support. We thank Dr Peter Spencer, Murdoch University, for
carrying out the parentage analyses, and Mia Hillyer and the WA Department of Parks and Wildlife for
technical assistance and processing quoll DNA samples. Quoll monitoring in Kakadu was supported
fnancially by grants from the National Geographic Conservation Trust, the Mazda Foundation, the
Australian Research Council, and NERP. We thank the Australian Wildlife Conservancy staff and
volunteers who conducted the central Kimberley quoll surveys and collaborated on feld trials of
taste-aversion baits at Mornington Wildlife Sanctuary. Part of the research carried out at Mornington
Wildlife Sanctuary was funded by the generous supporters of the Australian Wildlife Conservancy.
This research on northern quolls was carried out in accordance with protocols approved by the
University of Technology Sydney Animal Care and Ethics Committee (UTS ACEC 2012-432A) and
permits obtained from the Western Australia Department of Parks and Wildlife (Licences SF009187
and SF009829).
Photographs by Jonathan Webb
© Charles Darwin University, 2015
Can we mitigate cane toad impacts on northern quolls? is published by Charles Darwin
University and is for use under a Creative Commons attribution 4.0 Australia licence. For licence
conditions see:
The citation for this publication is as follows:
Webb, J., Legge, S., Tuft, K., Cremona, T., Austin, C. (2015). Can we mitigate cane toad impacts on
northern quolls?
Darwin: Charles Darwin University.
ISBN 978-1-925167-16-0
Printed by Uniprint
Charles Darwin University, Darwin Northern Territory, 0909, Australia.

Summary������������������������������������������������������������������������������������������������������������������������������������������� 1
1 Focus and signifcance of the project����������������������������������������������������������������������������������� 1
2 Distinctiveness of issue to this landscape����������������������������������������������������������������������������� 2
3 Knowledge status and constraints ��������������������������������������������������������������������������������������� 3
4 Methodological approaches������������������������������������������������������������������������������������������������� 3
4�1 Parentage analyses and monitoring of ‘toad smart’ quolls in Kakadu NP ������������� 3
4�2 Camera trap methods to estimate quoll population size ���������������������������������������� 4
4�3 Quoll surveys in the central Kimberley���������������������������������������������������������������������� 5
4�4 Field trials of toad-aversion baits at Mornington Willdlife Sanctuary �������������������� 5

5 Lessons learnt for this landscape������������������������������������������������������������������������������������������ 6
5�1 Parentage analyses and monitoring of ‘toad smart’ quolls in Kakadu NP – each
generation of quolls learns to avoid toads ���������������������������������������������������������������� 6
Camera trap methods to estimate quoll population size ���������������������������������������� 7
Distribution of quolls in the central Kimberley �������������������������������������������������������� 7
Field trials of toad-aversion baits at Mornington Willdlife Sanctuary �������������������� 9

6 National implications of lessons learnt������������������������������������������������������������������������������ 10
7 Problems addressing the focus and how to overcome these ������������������������������������������� 10
7�1 Cane toad invasion slower than anticipated ���������������������������������������������������������� 10
7�2 Toad-aversion bait shelf life������������������������������������������������������������������������������������� 10
8 Towards implementation���������������������������������������������������������������������������������������������������� 11
9 Looking ahead – future needs�������������������������������������������������������������������������������������������� 11
References �������������������������������������������������������������������������������������������������������������������������������������� 13

The spread of the toxic cane toad Rhinella marina threatens populations of the endangered northern
Dasyurus hallucatus. We identifed quoll populations at risk from toad invasion in the central
Kimberley and explored whether free ranging quolls would consume ‘toad-aversion’ baits that induce
aversions to live toads. A long-term study in Kakadu National Park showed that each generation
of quolls learns to avoid toads, so one deployment of toad-aversion baits could protect quolls from
toads. Encouragingly, 50% of wild quolls at Sir John Gorge, Mornington Wildlife Sanctuary (central
Kimberley) consumed toad-aversion sausages. More research on captive quolls is necessary to develop
long-lasting toad-aversion baits suitable for aerial deployment.
1 Focus and signifcance of the project
Northern Australia is currently experiencing a rapid, widespread collapse of its small mammal fauna
et al. 2010; Woinarski et al. 2011). One species at risk of extinction is the northern quoll,
Dasyurus hallucatus. Populations of northern quoll have rapidly gone extinct across northern Australia
as cane toads invaded their range (Woinarski
et al. 2014). Although quolls were declining before toads
arrived, this decline was mild relative to the catastrophic impact of toads. Quolls readily attack toads
but have little resistance to the toads’ toxin, and die after mouthing large toads (Covacevich & Archer
1975). If we do nothing, it is likely that the toads will eventually cause the widespread extinction of
quolls in Australia’s Kimberley and Pilbara regions. Unfortunately, we cannot do anything to prevent
the spread of toads through the Kimberley (Tingley
et al. 2013). What we can do, however, is teach
wild quolls to avoid eating cane toads (O’Donnell, Webb & Shine 2010). Populations of ‘toad smart’
quolls that avoid cane toads as prey will have a much lower risk of extinction than populations of toad
naïve quolls.
Potentially, we could train wild quolls to avoid eating cane toads by deploying ‘toad-aversion’ baits
(toad sausages containing a nausea inducing chemical) ahead of the toad invasion front. Quolls that
consumed such baits would become ill, and would subsequently associate the smell and taste of cane
toads with illness, and some individuals would ignore live cane toads (O’Donnell, Webb & Shine 2010).
This process is called conditioned taste aversion (CTA), and it occurs when predators ingest novel
toxic prey, become ill, and subsequently associate the smell and taste of prey with illness, and avoid
consuming the prey (Garcia, Hankins & Rusiniak 1974). Provided some ‘toad smart’ female quolls
survive in a toad-infested landscape, then their offspring will learn to avoid toads. This aversion is likely
to be transmitted to subsequent generations via social learning (i.e. when juveniles forage with their
mothers), or via CTA. For example, juvenile quolls that attack or ingest small, non-lethal metamorph
toads are likely to become ill and subsequently reject toads as prey, as do smaller dasyurid predators
(Webb, Pearson & Shine 2011).
The key aims of this project were to:
Determine whether each generation of quolls learns to avoid toads as food
Develop camera trap methods for estimating quoll population size
Identify northern quoll populations in the central Kimberley at risk of toad invasion
Evaluate whether wild Kimberley quolls would consume toad-aversion baits.
Can we mitigate cane toad impacts on northern quolls? | Final report
The research is signifcant because it tackles an important conservation issue; northern quolls are
critically endangered in the NT, and populations in WA are vulnerable to extinction from cane toads.
Our project aims to develop a method for preventing declines in northern quolls driven by the invasion
of the cane toad. In halting the ongoing decline of quolls across northern Australia, we would not only
be saving numerous local populations from extinction, but we will also be saving their unique genetic
heritage, giving the species a greater chance of persisting through future threats (e.g. climate change).
The quoll is also of cultural importance to many of the Indigenous people of northern Australia, and so
conserving local quoll populations, or bringing them back to country, also allows the conservation of
relevant Indigenous culture.
2 Distinctiveness of issue to this landscape
Northern quoll populations have declined across northern Australia over the last 40 years through the
combined impacts of grazing, feral predators, and altered fre regimes (Braithwaite & Griffths 1994).
Cane toads are now dispersing very rapidly across northern Australia, at a rate of around 40-60 km
per year (Phillips
et al. 2007), and have already invaded large parts of the eastern Kimberley. At their
present rate of spread, toads will have completely colonised the rest of the Kimberley within a decade
(B. Phillips, personal communication 2014). Thus, we have little time left to act if we are to prevent
widespread extinctions of quolls in the Kimberley.
Figure 1: A quoll stalking a cane toad�
3 Knowledge status and constraints
The broad-scale deployment of toad-aversion baits to quoll populations prior to toad invasion has the
potential to reduce toad impacts on this endangered species. However, research into the application
of toad-aversion baits to train wild quolls to avoid eating cane toads was not considered as an
action in the national recovery plan for northern quolls because the authors suggested that ‘the
likelihood of such a treatment having an effect past the initial generation of quolls is small’ (Hill &
Ward 2010). Thus, long-term studies are necessary to determine whether ‘toad-smart’ quolls transmit
toad-avoidance to their offspring. More importantly, we do not know if wild quolls from the central
Kimberley will consume toad-aversion baits. Nor do we know if wild quolls that eat toad-aversion baits
have higher survival following toad invasion compared to toad-naïve quolls.
There is also limited knowledge on the distribution of quolls in the central Kimberley or the size
of extant populations. There is uncertainty surrounding the best methods for detecting quolls and
estimating population sizes. The traditional method for estimating density is to live trap quolls with
wire cages; however, this method is labour intensive, and is impractical for surveying large areas.
Cameras provide an alternative non-invasive method for estimating density, and do not involve capture
stress for target and non-target species. A recent study showed that individual quolls can be identifed
by unique spot patterns on their pelt (Hohnen
et al. 2013). Hence, it should be possible to estimate
population size in this species using remote cameras and capture-recapture analysis. At present, there
is no consensus on how far apart cameras should be spaced to detect quolls and provide reliable
estimates of population size.
4 Methodological approaches
4.1 Parentage analyses and monitoring of ‘toad smart’ quolls
in Kakadu NP
To determine whether each generation of quolls learns to avoid eating cane toads we monitored
a population of ‘toad smart’ quolls in Kakadu National Park from 2010-2014. Fifty captive reared
juvenile quolls (28 males, 22 females) at the Territory Wildlife Park were trained not to eat cane
toads by feeding them a small dead toad infused with the nausea inducing chemical thiabendazole
(O’Donnell, Webb & Shine 2010). These ‘toad smart’ quolls were introduced to East Alligator ranger
station in December 2009 and February 2010, and their long-term survival was monitored via trapping
over four years. Tissue samples were taken from all individuals captured, and parentage analyses
were used to determine the identity of parents of all juveniles in the population (Marshall
et al. 1998;
Kalinowski, Taper & Marshall 2007).
Can we mitigate cane toad impacts on northern quolls? | Final report
4.2 Camera trap methods to estimate quoll population size
Live trapping and camera traps were used to estimate quoll population size at Sir John Gorge (17°
31.78’S, 126° 13.08’E) at Mornington Wildlife Sanctuary in the Kimberley, Western Australia. Live
trapping was done over three consecutive nights in September 2013. Small cage traps (Tomahawk
Live Traps, Hazelhurst USA) baited with one tablespoon of tuna in oil were spaced approximately
30 metres apart on two transects (20 on the north transect, 10 on the south). We dusted Coopex
(Bayer Environmental Science) around each cage trap to ward off ants. We checked cages at dawn
and photographed each quoll, injected a microchip for identifcation, and recorded sex, mass and
reproductive status prior to release.
We set up camera traps (Bushnell Trophy Cam, Bushnell Outdoor Products, 2012) immediately after live
trapping along the same transects. We deployed 4 cameras along the south transect and 11 cameras
along the north transect, with each camera spaced approximately 80 metres apart. We secured cameras
to trees or rocky ledges approximately one metre from the ground facing directly downwards using EzAim 2 Game Camera Mount (Slate River, LLC, Milwaukee) or a webbed strap (Fig. 2). We set cameras
to high sensitivity and programmed them to take three consecutive photographs for each trigger with
a delay of 10 seconds between triggers. We placed one layer of cream masking tape over the LED light
of each camera to reduce the harshness of the flash, and avoid overexposure of photographs (De Bondi
et al. 2010). A perforated plastic vial flled with tuna was affxed to a rock below each camera (Fig. 2). A
pilot study showed that tuna was the most effective bait for detecting quolls (Austin 2014).
Figure 2: Photographs of two camera stations set up on a rock ledge (left) and in a tree (right)�
Using photographs collected from the cameras, we identifed individuals from spot patterns (Hohnen
et al. 2013) and compiled capture histories for each individual. We estimated population size using
Huggins P and C population models for closed populations in Program MARK (version 7.2). To
determine how many cameras should be deployed to provide reliable estimates of population size,
we ran a block bootstrapping analysis in R, using cameras along the northern transect (n = 11) with
a moving block of three cameras. To determine the ideal spacing of cameras we calculated the mean
number of individuals recorded from all possible combinations of cameras, alternating cameras (160
metres), every third camera (320 metres), every fourth camera (400 metres) and every sixth camera
(480 metres).
4.3 Quoll surveys in the central Kimberley
Northern Quolls were surveyed across Australian Wildlife Conservancy Sanctuaries to produce baseline
distribution data ahead of the arrival of cane toads. Twenty-nine sites were surveyed in 2014. At
each site, six white-flash camera traps (Reconyx, PC850 Hyperfre White Flash) were set each 100 m
apart in a transect following likely quoll habitat (creekline, cliffline, gorge). Cameras were positioned
to face down, as described previously. We baited traps with oats, peanut butter and fsh and
retrieved cameras two weeks later. We identifed individual quolls from images by inspecting unique
arrangements of spots to generate the minimum number known to be alive for each site.
4.4 Field trials of toad-aversion baits at Mornington Willdlife
We made toad-aversion baits by removing the legs from freshly thawed adult cane toads that were
collected by community groups during ‘toad busts’. We placed both skinned and unskinned toad legs
(90 and 10% by volume respectively) into a blender to create a toad mince. We added approximately
60 milligrams of the nausea inducing chemical thiabendazole to each 15 g portion of mince. Toad
meatballs consisted of 15 g portions of mince, whereas toad sausages consisted of 15 g of mince
encased within a synthetic sausage skin.
Baiting trials were carried out on the southern side of Sir John Gorge in October and December
2014. Cane toads had not yet colonised this site, but could invade during early 2015. For each trial,
we deployed 30 cage traps along a transect running parallel to the gorge with each trap spaced
approximately 20 m apart. We placed the toad-aversion bait (either meatball or sausage) inside each
trap at dusk, and checked the traps the following morning. For each trap, we recorded whether
the quoll had eaten or partially eaten the bait, and recorded the microchip number, sex, mass, and
reproductive status of the quoll. New quolls were injected with a microchip in the loose fur on the
scruff of the neck. New baits were placed in each trap each evening, and traps were open for four
consecutive nights in October, and three consecutive nights in December 2014.
Once trapping was completed, we set up nine camera traps (Reconyx, PC850 Hyperfre White Flash)
spaced approximately 80 m apart along the trapping transect. Each camera was set up facing down,
and was programmed to high sensitivity motion trigger as well as a time-lapse setting to take photos
every 30 minutes. A toad sausage or toad meatball was placed underneath each camera, and was
surrounded by a ring of Coopex to discourage ants from eating the bait. Cameras were brought
in four days after deployment, and any quolls taking or investigating the toad-aversion baits were
identifed from their spot patterns
Can we mitigate cane toad impacts on northern quolls? | Final report
5 Lessons learnt for this landscape
5.1 Parentage analyses and monitoring of ‘toad smart’ quolls in
Kakadu NP – each generation of quolls learns to avoid toads
Parentage analyses of juvenile quolls captured at East Alligator revealed that each generation of quolls
learns to avoid eating cane toads (Cremona
et al. 2015). For example, in May 2011, there were fve
breeding females on the study site: two trained reintroduced females, one wild female, the offspring
of a wild female, and the offspring of a trained female. These females reproduced, and in 2012 we
captured 14 juveniles: 2 could not be assigned a mother, 4 were offspring of a trained reintroduced
female breeding in her second year and 4 were second generation offspring of a reintroduced female
et al., 2015). Thus, some quolls learn to avoid eating cane toads. Juvenile quolls spend
extended periods with their mothers (Fig. 3) and may have learnt not to eat toads by watching their
mothers sniff and reject toads, or they may have ingested small non-lethal sized toads that induce
nausea and long-lasting aversion to live toads in other dasyurids (Webb
et al. 2008; Webb, Pearson &
Shine 2011).
Figure 3: A quoll and her young�
Radio-telemetry revealed that predation by camp dogs and dingoes were the major sources of
mortality for quolls on the study site. Our study site was burnt annually during the late dry season,
and there was little cover available to quolls in December when young quolls begin foraging. Previous
studies have shown that quolls are most at risk from dingo predation in recently burnt areas (Oakwood
2000). Population viability modeling suggests that high mortality from predation, coupled with
frequent late dry season burning, is preventing the quoll populations in Kakadu National Park from
recovering after toads invaded (Cremona, Crowther & Webb 2015).
5.2 Camera trap methods to estimate quoll population size
Camera traps provided good estimates of quoll population size. Live trapping yielded an estimate of 14
animals, while cameras yielded an estimate of 15 animals. Ten of the 11 cameras along the northern
bank of Sir John Gorge recorded 971 photographs of quolls, of which 938 (96.6%) could be used
to identify individuals. In total, we identifed 14 individuals (6 females, 7 males and one unknown).
Comparing photographs taken during live trapping to photos from cameras, we determined that all
live captured individuals were captured on remote cameras. An additional individual was captured on
camera but not during live trapping. Bootstrapping revealed that the number of individuals recorded
on camera decreased as cameras were more widely spaced (Fig. 4). At Sir John Gorge, spacing of
cameras 80 m apart provided reliable estimates of quoll population size (Fig. 4).
Figure 4: Mean number of individual quolls detected at different camera spacing (± standard errors)�
5.3 Distribution of quolls in the central Kimberley
Twenty-nine sites were surveyed in 2014 (Fig. 5). Additional data collected between 2011 and 2014 is
included in the map for sites in the north-western parts of Artesian Range where quolls are common.
Quolls were present at 9 of the 25 sites surveyed on Mornington, Marion Downs and Tableland
sanctuaries, all of which were within a discrete area of the King Leopold Ranges in the south-east
of Mornington (Fig. 6). At these sites, the minimum number of quolls detected on cameras ranged
between 1 and 12 (Table 1). These small populations are vulnerable to extinction from toad invasion
and environmental stochasticity.
8 6 4 2
0 100 200
Spacing of cameras (m)
300 400 500 600
Can we mitigate cane toad impacts on northern quolls? | Final report
Figure 5: Distribution of the northern quoll on AWC sanctuaries in the Kimberley� The current location (September
2014) and direction of spread of the cane toad invasion front is shown in green� The next nearest quoll
population to the northwest is shown in the orange ellipse; there are no quoll populations to the northeast of the
Mornington population�
Table 1� Northern quoll populations on Mornington and the neighbouring crown land with minimum number
known to be alive from most recent survey data, data source and watercourse�

Site Minimum number known
to be alive
Data source Watercourse
Sir John Gorge 12 Trapping 3 x annually Fitzroy River
Tin Can Gully 9 Camera traps 2014 Fitzroy River
Rose’s Pool 8 Camera traps 2014 Spider Creek
Cowendyne South 9 Camera traps 2014 Cowendyne Creek
Sir John Gorge Upper 3 Camera traps 2014 Fitzroy River
Cliftoniana Gully 10 Camera traps 2014 Fitzroy River
Collis Creek 1 Camera traps 2014 Fitzroy tributary
Cowendyne North Trapping 2011 Cowendyne Creek
Narrie West 4 Camera traps 2014 Fitzroy tributary
Narrie Range Sir John 3 Camera traps 2014 Fitzroy River
Slingshot Gap Trapping 2011 Fitzroy tributary

Figure 6: Northern Quoll sites on Mornington and neighbouring crown land with density in the most recent round
of surveys indicated: high (more than 4 individuals detected in 84 trap-nights using 6 camera traps) and low (4 or
fewer individuals)� All sites were surveyed in 2014 except for Slingshot Gap and Cowendyne North which were last
surveyed in 2011 and thus indicated only as present�
5.4 Field trials of toad-aversion baits at Mornington Wildlife
In October, we trapped fve quolls in traps baited with toad-aversion baits, and fve quolls were
identifed on camera taking baits. In December, we trapped two quolls; one new female (not trapped
in October) consumed the bait, while one male trapped in October did not consume the bait. Quolls
investigated baits at fve of the cameras, sometimes rubbing their bellies over the baits. No additional
quolls to those that inspected baits in October were recorded. One individual investigated baits nine
times at four different cameras but didn’t eat them. This individual previously took two toad-aversion
baits in the October camera trap trial. A second quoll investigated baits seven times at fve cameras
but did not eat them. Again, this female had previously eaten a bait in the October trial. Two quolls
from the October trial that hadn’t taken baits investigated baits in December.
Can we mitigate cane toad impacts on northern quolls? | Final report
6 National implications of lessons learnt
Our research in Kakadu National Park demonstrated that each generation of quoll learns to avoid
toads as prey. Thus, toad aversion baiting would only need to be done once to protect quoll
populations from cane toads. Thereafter, each generation of quolls can learn to avoid eating cane
toads, perhaps via social learning, or via ingestion of small non-lethal toads that induce aversions to
live toads.
Camera traps, in conjunction with identifcation of individuals via spot patterns, can provide reliable
estimates of quoll population size in the central Kimberley. Using this methodology, we identifed
another eight quoll populations near Mornington Wildlife Sanctuary, all of which are small, and thus,
vulnerable to extinction (Table 1).
Six of 12 quolls (50%) at Sir John Gorge ingested toad-aversion baits. The major problem that we
observed was the toad aversion baits went rancid overnight, which presumably made them less
palatable to quolls. Hence, we either need to develop long-lasting toad-aversion baits that are
more palatable to quolls, or we need to bring quolls into captivity prior to toad invasion to prevent
population extinctions. Our research in Kakadu National Park shows that trained ‘toad-smart’ quolls
can be reintroduced following toad invasion, so the latter option should be feasible.
7 Problems addressing the focus and how to
overcome these
7.1 Cane toad invasion slower than anticipated
The invasion of cane toads was slower than anticipated, and cane toads did not invade the study sites
in 2014. Consequently, we were unable to experimentally test whether wild quolls that consumed
toad-aversion sausages had higher survival than toad naïve individuals following toad invasion.
This problem will be overcome by deploying toad-aversion baits to experimental sites in 2015, and
monitoring the subsequent survival of radio-collared quolls at control and experimental sites before
and after cane toads invade the study sites.
7.2 Toad-aversion bait shelf life
We did not impregnate toad-aversion baits with salts, preservatives or fats, and consequently the
toad-aversion baits rapidly desiccated and went rancid overnight during the December baiting trials
when day time temperatures exceeded 45 °C. Future studies are necessary to develop long-lasting
toad-aversion baits suitable for aerial deployment. A captive colony of quolls is crucial for testing the
effcacy of the toad-aversion baits for inducing aversions to live toads. Currently, no captive quolls are
available to do this research.
8 Towards implementation
Toads are expected to arrive at Mornington Wildlife Sanctuary either in February-March 2015, or early
2016. Our plan is to feld test whether toad-aversion baiting can reduce the impacts of cane toads on
quoll populations in the central Kimberley. We will deploy toad-aversion baits in the feld, and monitor
the fate of experimental (toad-aversion baits deployed) and control (no baits deployed) populations
via camera trapping. A more intensive study will be done at Sir John Gorge. At this site, toad-aversion
baits will be placed in cage traps, and the fate of quolls that consume baits (trained quolls) or do not
eat baits (untrained quolls) will be monitored via radio-telemetry before and after toad invasion.
9 Looking ahead – future needs
Cane toads are rapidly spreading across the Kimberley, and will likely cause quolls in that region to go
locally extinct. There are two strategies that could be used to prevent quoll populations from going
1. Develop long-life toad-aversion baits suitable for aerial deployment, and deploy these baits ahead of
the toad invasion front.
We need to develop toad aversion baits with a long shelf life suitable for aerial deployment. The
baits we tested in this study rapidly desiccate and go rancid overnight, and thus quickly become
unpalatable to quolls. The incorporation of toad fats and tasteless, odourless preservatives into
the next generation of baits could help to solve this problem. A captive colony of quolls will be
necessary to test whether this next generation of toad-aversion baits can induce an aversion to
live toads in quolls. The Territory Wildlife Park has purpose built quoll enclosures available to house
quolls and TWP personnel have considerable expertise in maintaining and breeding quolls. Wild
quolls could be obtained from Pobassoo and Astell Islands off Gove. These island quolls are derived
from wild stock obtained from the Darwin and Kakadu regions, and because they are toad-naïve,
they are ideal subjects for testing the effcacy of next generation toad-aversion baits.
After development and testing on captive quolls, next generation toad-aversion baits could
be deployed via helicopter ahead of the toad invasion front. We have identifed several quoll
populations that could serve as control populations (no baiting) and experimental populations
(baiting) near Mornington Wildlife Sanctuary (Table 1). These populations could be monitored via
camera trapping before, during, and after toad invasion to test whether toad-aversion baiting
prevents quoll populations from going locally extinct.
2. Capture wild quolls before toads invade and maintain the quolls inside toad-proof enclosures at
Mornington Wildlife Sanctuary
Bringing wild quolls into captivity before the toads invade would prevent local extinctions from
occurring. Once cane toads have invaded, the offspring of wild quolls could be trained not to eat
cane toads, by feeding them a small dead toad laced with thiabendazole (O’Donnell, Webb &
Shine 2010). These ‘toad smart’ quolls could then be reintroduced to the wild after the toads have
invaded (Fig. 7). The research at Kakadu National Park suggests that some of these quolls are likely
to survive and reproduce, and their offspring are also likely to learn not to eat cane toads.
Can we mitigate cane toad impacts on northern quolls? | Final report
Figure 7: A quoll is released after being microchipped and having its sex, mass and reproductive status recorded�
Figure 8: Cane toad
Rhinella marina.
Austin, C. (2014) Can remote cameras accurately estimate populations of the endangered northern
quoll? Honours thesis, University of Technology Sydney.
Braithwaite, R.W. & Griffths, A.D. (1994) Demographic variation and range contraction in the
northern quoll,
Dasyurus hallucatus (Marsupialia: Dasyuridae). Wildlife Research, 21, 203-217.
Covacevich, J. & Archer, M. (1975) The distribution of the cane toad,
Bufo marinus, in Australia and its
effects on indigenous vertebrates.
Memoirs of the Queensland Museum, 17, 305-310.
Cremona, T., Crowther, M.S. & Webb, J.K. (2015) Predation by a subsidised apex predator prevents the
recovery of an endangered mesopredator.
Animal Conservation, in review.
Cremona, T., O’Donnell, S., Spencer, P., Shine, R. & Webb, J.K. (2015) Avoiding the last supper: taste
aversion provides long-term protection to endangered predators from invasive prey.
Journal of Applied
in review.
De Bondi, N., White, J.G., Stevens, M. & Cooke, R. (2010) A comparison of the effectiveness of
camera trapping and live trapping for sampling terrestrial small-mammal communities.
, 37, 456-465.
Garcia, J., Hankins, W.G. & Rusiniak, K.W. (1974) Behavioural regulation of the Milieu Interne in man
and rat.
Science, 185, 824-831.
Hill, B.M. & Ward, S.J. (2010) National Recovery Plan for the northern quoll
Dasyurus hallucatus.
Department of Natural Resources, Environment, The Arts and Sport, Darwin., Darwin.
Hohnen, R., Ashby, J., Tuft, K. & McGregor, H. (2013) Individual identifcation of northern quolls
Dasyurus hallucatus) using remote cameras. Australian Mammalogy, 35, 131-135.
Kalinowski, S.T., Taper, M.L. & Marshall, T.C. (2007) Revising how the computer program CERVUS
accommodates genotyping error increases success in paternity assignment.
Molecular Ecology, 16,
Marshall, T.C., Slate, J., Kruuk, L.E.B. & Pemberton, J.M. (1998) Statistical confdence for likelihoodbased paternity inference in natural populations.
Molecular Ecology, 7, 639-655.
O’Donnell, S., Webb, J.K. & Shine, R. (2010) Conditioned taste aversion enhances the survival of an
endangered predator imperilled by a toxic invader.
Journal of Applied Ecology, 47, 558-565.
Oakwood, M. (2000) Reproduction and demography of the northern quoll,
Dasyurus hallucatus, in the
lowland savanna of northern Australia.
Australian Journal of Zoology, 48, 519-539.
Phillips, B.L., Brown, G.P., Greenlees, M., Webb, J.K. & Shine, R. (2007) Rapid expansion of the cane
toad (
Bufo marinus) invasion front in tropical Australia. Austral Ecology, 32, 169-176.
Tingley, R., Phillips, B.L., Letnic, M., Brown, G.P., Shine, R. & Baird, S.J.E. (2013) Identifying optimal
barriers to halt the invasion of cane toads
Rhinella marina in arid Australia. Journal of Applied Ecology,
50, 129-137.
Can we mitigate cane toad impacts on northern quolls? | Final report
Webb, J.K., Brown, G.P., Child, T., Greenlees, M.J., Phillips, B.L. & Shine, R. (2008) A native dasyurid
predator (common planigale, Planigale maculata) rapidly learns to avoid a toxic invader.
, 33, 821-829.
Webb, J.K., Pearson, D. & Shine, R. (2011) A small dasyurid predator (
Sminthopsis virginiae) rapidly
learns to avoid a toxic invader.
Wildlife Research, 38, 726-731.
Woinarski, J.C.Z., Armstrong, M., Brennan, K., Fisher, A., Griffths, A.D., Hill, B., Milne, D.J., Palmer, C.,
Ward, S., Watson, M., Winderlich, S. & Young, S. (2010) Monitoring indicates rapid and severe decline
of native small mammals in Kakadu National Park, northern Australia.
Wildlife Research, 37, 116-126.
Woinarski, J.C.Z., Legge, S., Fitzsimons, J.A., Traill, B.J., Burbidge, A.A., Fisher, A., Firth, R.S.C.,
Gordon, I.J., Griffths, A.D., Johnson, C.N., McKenzie, N.L., Palmer, C., Radford, I., Rankmore, B.,
Ritchie, E.G., Ward, S. & Ziembicki, M. (2011) The disappearing mammal fauna of northern Australia:
context, cause, and response.
Conservation Letters, 4, 192-201.

Northern Biodiversity
Northern Biodiversity
Email: [email protected]
Phone: 08 8946 6761
This research was supported by
funding from the Australian
Government’s National Environmental
Research Program
4.1 – 06/15
View publication stats