ArticlePDF Available

Mammal responses to fox control - Glenelg Ark

Authors:
  • Arthur Rylah Institute for Environmental Research
Glenelg Ark 2005–2013: Evidence of
the Benefits for Native Mammals of
Sustained Fox Control
March 2014
Arthur Rylah Institute for Environmental Research
Unpublished Client Report
A. Robley, A. Gormley, B. Triggs, R. Albert, M. Bowd,
C. Hatfield, R. McDonald, C. Rowe, K. Scott, and A. Smith
iii
Glenelg Ark 2005–2013: Evidence of the
Benefits for Native Mammals of
Sustained Fox Control
Alan Robley, Andrew Gormley, Barbara Triggs, Ray Albert, Michael Bowd, Chris
Hatfield, Robert McDonald, Chris Rowe, Kenneth Scott, and Alieen Smith
Arthur Rylah Institute for Environmental Research
123 Brown Street, Heidelberg, Victoria 3084
March 2014
Arthur Rylah Institute for Environmental Research
Department of Environment and Primary Industries
Heidelberg, Victoria
iv
Report produced by: Arthur Rylah Institute for Environmental Research
Department of Environment and Primary Industries
PO Box 137
Heidelberg, Victoria 3084
Phone (03) 9450 8600
Website: www.depi.vic.gov.au/ari
© State of Victoria, Department of Environment and Primary Industries 2014
This publication is copyright. Apart from fair dealing for the purposes of private study, research, criticism or review as
permitted under the Copyright Act 1968, no part may be reproduced, copied, transmitted in any form or by any means
(electronic, mechanical or graphic) without the prior written permission of the State of Victoria, Department of
Environment and Primary Industries. All requests and enquiries should be directed to the Customer Service Centre,
136 186 or email customer.service@depi.vic.gov.au
Citation: Robley, A., Gormley A., Triggs, B., Albert, R., Bowd, M., Hatfield. C., McDonald, R., Rowe, C., Scott, K.,
and Smith, A. (2014) Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control.
Arthur Rylah Institute for Environmental Research Technical Report Series. Department of Environment and Primary
Industries, Heidelberg, Victoria.
Disclaimer: This publication may be of assistance to you but the State of Victoria and its employees do not guarantee
that the publication is without flaw of any kind or is wholly appropriate for your particular purposes and therefore
disclaims all liability for any error, loss or other consequence which may arise from you relying on any information in
this publication.
Accessibility:
If you would like to receive this publication in an accessible format, such as large print or audio, please telephone
136 186, or through the National Relay Service (NRS) using a modem or textphone/teletypewriter (TTY) by dialling
1800 555 677, or email customer.service@depi.vic.gov.au
This document is also available in PDF format on the internet at www.depi.vic.gov.au
Front cover photo: Long-nosed potoroo (Alan Robley)
Authorised by: Victorian Government, Melbourne
Printed by: PRINTROOM 77 St Georges Rd, Preston 3072
v
Contents
Acknowledgements ......................................................................................................................... vii
Summary ............................................................................................................................................ 1
1 Introduction ............................................................................................................................. 2
2 Methods .................................................................................................................................... 4
2.1 Study area ................................................................................................................................. 4
2.2 Monitoring and evaluation design ............................................................................................ 4
2.3 Measuring changes in foxes ...................................................................................................... 5
2.3.1 Index of abundance.................................................................................................... 5
2.4 Measuring populations of key native species ........................................................................... 5
2.4.1 Site-occupancy modelling ......................................................................................... 5
2.5 Fox and feral cat diet................................................................................................................. 7
2.6 Rainfall ...................................................................................................................................... 8
3 Results ...................................................................................................................................... 9
3.1 Fox control program ................................................................................................................. 9
3.1.1 Overall reductions in fox abundance ......................................................................... 9
3.2 Native species response .......................................................................................................... 11
3.2.1 Common brushtail possum ...................................................................................... 12
3.2.2 Long-nosed potoroo................................................................................................. 14
3.2.3 Southern brown bandicoot ....................................................................................... 17
3.2.4 Fox and feral cat diet ............................................................................................... 19
3.2.5 Fox and feral cat stomach and gut contents ............................................................. 22
4 Discussion .............................................................................................................................. 24
References ........................................................................................................................................ 28
Appendix 1. Ecological Vegetation Classes within each treatment and non-treatment area ... 33
Appendix 2.1 Location and year of detection for long-nosed potoroos at LGNP-south and
north. Location where species recorded shown by coloured dots, records over multiple years
shown as separate colours. Black dots – hairtube stations where species not recorded. .......... 34
Appendix 2.2 Location and year of detection for long-nosed potoroos at Cobboboonee and
Mt Clay. Location where species recorded shown by coloured dots, records over multiple
years shown as separate colours. Black dots – hairtube stations where species not recorded. 35
Appendix 2.3 Location and year of detection for long-nosed potoroos at Hotspur and Annya.
Location where species recorded shown by coloured dots, records over multiple years shown
as separate colours. Black dots – hairtube stations where species not recorded. ...................... 36
Appendix 3.1 Location and year of detection for southern brown bandicoots at LGNP-south
and north. Location where species recorded shown by coloured dots, records over multiple
years shown as separate colours. Black dots – hairtube stations where species not recorded. 37
vi
Appendix 3.2 Location and year of detection for southern brown bandicoots at Cobboboonee
and Mt Clay. Location where species recorded shown by coloured dots, records over
multiple years shown as separate colours. Black dots – hairtube stations where species not
recorded. .......................................................................................................................................... 38
Appendix 3.3 Location and year of detection for southern brown bandicoots at Hotspur and
Annya. Location where species recorded shown by coloured dots, records over multiple
years shown as separate colours. Black dots – hairtube stations where species not recorded. 39
vii
Acknowledgements
This project was funded by the Pests and Weeds on Public Land Initiative of the Department of
Environment and Primary Industries through the Glenelg Ark project and Parks Victoria. W.
Burns, Georgina Neave and S. Balharrie provided valuable support to the project and the
development of the monitoring and evaluation program, as did J. Cook, S. Nicol and D.
MacKenzie. Members of the Glenelg Ark Working Group (including R. Hill and B. Hoare) also
provided guidance and input throughout the years. Dave Forsyth and Matt Bruce provided
comments that improved this report. This work was conducted under the Department of
Environment and Primary Industries Animal Ethics Committee Permit Numbers 10/23, 9/15 and
8/28.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 1
Summary
The Glenelg Ark project was established in 2005 under the Department of Sustainability and
Environment Weeds and Pests on Public Land Initiative. The project aims to facilitate the recovery
of native mammal populations at risk from predation by foxes in far south-west Victoria by
undertaking broad-scale, continuous fox baiting using buried Foxoff® baits across 100 000 ha of
state forest and national park.
Management agencies routinely invest considerable amounts of scarce public funding into reducing
the impacts of invasive predators. However, limited published information exists from large-scale
management experiments demonstrating that the reduction in the abundance of invasive predators
benefits native prey species.
We conducted a nine year landscape-scale management experiment to evaluate the benefits of
reducing red foxes (Vulpes vulpes) to three native ground-dwelling mammal prey species (southern
brown bandicoot Isoodon obesulus, long-nosed potoroo Potorous tridactylus, and common
brushtail possum Trichosurus vulpecula). We hypothesised that sustained fox control would result
in a difference in fox abundance between treatment and non-treatment sites, and that occupancy,
colonisation and persistence rates of the three native species would be higher on sites subject to
sustained fox control.
Differences in indices of fox abundance between three treatment and three non-treatment areas
were used to measure the effects of control. We used dynamic Bayesian modelling to assess
changes in colonisation, persistence and occupancy of the three native prey species.
There was a 76.7% difference in bait take following the commencement of baiting, with a
significant decline in bait take from 2005 to 2013 on the treatment sites and a significant difference
between treatment and non-treatment sites between 2005 and 2013. Overall, each of the three native
prey species showed a significant positive response in site occupancy to fox control. However,
colonisation and persistence rates were not uniformly positive for all species or all treatment areas.
The most common item in fox diet was the common ringtail possum, wallabies and kangaroos.
Southern brown bandicoot (3%) and long-nosed potoroos (1.7%) were present but recorded
infrequently. Foxes had a broader diet on treated areas compared to non-treated areas and the diets
of foxes and feral cats overlapped.
The results of our nine-year management experiment demonstrate that foxes can be reduced and
maintained at relatively lower levels in open, fragmented landscapes, and that fox control has had a
generally positive effect on the occurrence of the three native mammal species. Alternative
hypotheses that could explain the observed variable response on treatment areas are that bottom-up
processes may limit the population's capacity to persist in newly colonised sites, or that foxes have
been replaced by a mesopredator. These alternative explanations require further investigation.
While fox control can be beneficial to native species, land managers may need to consider a wider
range of actions than simply reducing foxes in order to maintain positive benefits to native species.
The findings of our research will help policy makers, and the community, to assess the success and
future direction of broad-scale mainland fox control programs.
2 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
1 Introduction
Predators introduced to new environments can have a dramatic impact on the ecology of native
species (Groombridge 1992; Vitousek et al. 1997). Often naive native prey lack the co-evolved
adaptations needed to reduce encounter rates or increase the probability of escape once detected by
introduced predators (Lima and Dill 1990). Introduced predators can maintain themselves at high
population densities as they are no longer limited by their natural enemies, competitors, or parasites
(Elton 1958). Also, most introduced predators are generalists, whose numbers are not greatly
limited by the decline in any one prey species. Predation can act to limit population growth directly
by reducing recruitment and survival rates, either when the prey species is the main food item and
is consumed as prey density increases, or when prey are a secondary item, avoided at low density
but actively sort as prey density increases (Holling 1959; Pech et al. 1995; Sinclair et al. 1997).
Predation can also have sub-lethal effects on prey behaviour and physiology through stress
(Boonstra et al. 1998). While the reduction or removal of introduced predators can have significant
positive effects on native prey (Pech et al. 1995; Kinnear et al. 2002, 2010; Salo et al. 2010).
The red fox (Vulpes vulpes) is the most widespread species of its genus, naturally occurring
throughout most of the Palaearctic ecozone (Lloyd 1980). It was introduced into the north Americas
near Maryland sometime in the mid-eighteenth century and into Australia in 1871 (Saunders et al.
1995) where it is now both widespread and common. The red fox is listed by the International
Union for Conservation of Nature (IUCN) as among the world’s worst 100 invaders
(www.issg.org/database), and predation by red foxes has been recognised as a key threatening
process to a range of native animals around the world (Environment Protection and Biodiversity
Conservation Act 1999, Australia; Endangered Species Act 1973, USA).
In Australia, the red fox has been implicated as a primary cause in the complete or regional
extinction of a range of native mammal species, most of which are small to medium-sized (35g to
5500g) and ground dwelling (Burbidge and McKenzie 1989; Short and Smith 1994). Salo et al.
(2010) concluded that the impact of red fox predation on the Australian native fauna was greater
than in any other area where foxes had invaded. To mitigate this threat, fox control is conducted for
threatened species protection on an estimated 10.5 million ha of land in Australia per year (Reddiex
et al. 2004), with the economic cost to the environment estimated at $190 million dollars annually
(McLeod et al. 2008).
Increasingly, land managers are required to justify the investment of scarce public funds in ongoing
fox control by demonstrating positive population growth in native species which can only result
once predator densities are reduced below their level of impact. Effective fox control in open areas
can be successful when undertaken continuously and over relatively large areas (Saunders and
McLeod 2007; Salo et al. 2010). Measuring the success of landscape scale management actions is
problematic. Adaptive landscape scale management (including randomisation and replication of
sites) and modelling incorporating on-ground management actions (Walters 1986; Walters and
Holling 1990) has been suggested as a way of understanding the effectiveness of managing
introduced species (Innes et al. 1999; Park 2004). However, there are few examples of where an
experimental approach has been applied to assess the management of introduced predators and the
benefits to prey species (Reddiex and Forsyth; 2006; Salo et al. 2007; McLeod et al. 2008).
The impacts of introduced predators on native species is well documented (Salo et al. 2007, 2010),
despite this there remains a large amount of uncertainty on how to effectively measure the response
of threatened species to management intervention (Park 2004). It has been suggested that indirect
measures of prey response could be used to assess whether fox predation is regulating a prey
population (Pech et al. 1995). Site occupancy (an estimate of the proportion of sites in an area that
are occupied) has been used as a metric that reflects the current state of a population (MacKenzie et
al. 2003). Occupancy of locations is driven by the processes of species persisting at sites from one
year to the next (akin to the survival), and the colonisation of new, previously unoccupied sites
(akin to recruitment). The use of multi-season occupancy modelling has been used to assess
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 3
changes in occupancy associated with habitat disturbance and fragmentation and predator–prey and
predator–predator interactions (Frey et al. 2011; Cove et al. 2012; Lazenby and Dickman 2013).
The Glenelg Ark project was established in July 2005 to facilitate the recovery of native animal
populations considered at risk from fox predation by undertaking landscape scale, continuous fox
baiting across 100 000 ha of state forest and national park in south-western Victoria, Australia. To
ensure ongoing government commitment and community support for Glenelg Ark, its benefits to
Victoria’s biodiversity must be demonstrated. The monitoring and evaluation component of
Glenelg Ark is measuring: a) the response of foxes to control activities; and b) the response of a
range of native species that are at risk from fox predation. Without such a program, management
will have no capacity to: a) justify reinvestment of scarce public conservation funds; b) improve
management actions based on scientific information about the effectiveness of previous
management actions; and c) maintain community support. Thus, the monitoring and evaluation
form part of management and are not an imposition or adjunct to it.
Three native mammal species that are present in the Glenelg Ark project area that are in low
numbers (Robley et al. 2011), and with patchy distributions (Menkhorst 1995), were selected for
monitoring. These were the southern brown bandicoot (Isoodon obesulus), long-nosed potoroo
(Potorous tridactylus), and the common brushtail possum (Trichosurus vulpecula). The bandicoot
and potoroo are medium-sized ground-dwelling mammals (c. 1.0 kg and c. 1.2 kg respectively)
with high and moderate fecundity respectively (Lobert and Lee 1990). Both species are known to
be preyed on by foxes (Seebeck 1978) and have been reported to positively respond to the reduction
in foxes (Kinnear et al. 2002; Arthur et al. 2012). Brushtail possums are semi-arboreal species
weighing c. 3.0 kg, have low rate of fecundity (Kerle and How 2008), are known to occur in the
diet of foxes (Triggs et al. 1984), and respond to fox control (Kinnear et al. 2002).
Given the role that foxes have played in the decline and extinction of Australian mammals and
examples of recovery following the sustained reduction in foxes, and our knowledge of the current
status of the targeted prey species, we reasoned that once foxes had been reduced the prey species
would be able to escape limitation. It is predicted that colonisation and persistence would be higher
at sites with fox control, relative to those without, leading to an increased number of sites occupied.
The outcome of our research is that land managers, policy makers, and the community can make
informed, evidence-based decisions on the success and future direction of broad-scale mainland fox
control operations.
4 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
2 Methods
2.1 Study area
The general study area is located in far south-west Victoria, near the township of Heywood
(38°07'50''S, 147°37'45''E), situated on six forested blocks of state forest and national park. The
main vegetation communities across all six blocks are heathy woodland, lowland forest, herb-rich
woodland and wet heathland (Appendix 1). The area receives an average annual rainfall of 70 mm,
and an average minimum and maximum temperature of 8.1 °C and 17.6 °C respectively.
2.2 Monitoring and evaluation design
We use a conditional case-control study design to identify whether changes in site occupancy by
native species (the ‘case’) is conditional on the presence of foxes by comparing sites that have
foxes present but are otherwise similar (the controls) to those that have a substantially reduced
number of foxes.
Three monitoring areas known as Treatment Monitoring Areas (TMAs, i.e. areas that receive fox
control) and three Non-Treatment Monitoring Areas (NTMA, i.e. areas that do not receive fox
control) (Figure 1) are used to assess the benefits of fox control. In the TMAs and NTMAs there
had been little previous fox control prior to 2005. In order to achieve a broad-scale reduction in
foxes across the public land areas, treatment for fox control was consolidated in the southern half of
the overall project area (Figure 1). This meant that random allocation of treatment and non-
treatment sites was not feasible. The six monitoring areas are:
1. Lower Glenelg National Park-south (LGNP-south; TMA; 8954 ha)
2. Lower Glenelg National Park-north (LGNP-north; NTMA; 4659 ha); separated by the Glenelg
River
3. Cobboboonee National Park (TMA; 9750 ha)
4. Annya State Forest (NTMA; 8520 ha)
5. Mount Clay State Forest (TMA; 4703 ha)
6. Hotspur State Forest (NTMA; 6940 ha).
This design will enable the identification of any patterns of association between a reduction in foxes
and an increase in targeted native species but does not allow any statistical interpretation of
causality (Lande et al. 1994).
Figure 1. Glenelg Ark monitoring and evaluation areas
0 10 205 Kilometers
¯
Hotspur
Heywood
Annya
Mt Clay
Cobboboonee
LGNP-north
LGNP-south
Glenelg R.
Legend
Bait Stations
No fox control
Fox control
Parks and Reserves
State Forest
Softwood Plantation
Glenelg River
Roads
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 5
2.3 Measuring changes in foxes
In each TMA bait station were constructed at 1 km intervals, and a single 1080 Foxoff® bait was
buried to a depth of 10 cm at each bait station. Bait stations were checked and all baits replaced on
a fortnightly basis throughout each year between October 2005 and October 2013. NTMA bait
stations were constructed in the same manner and non-toxic Foxoff® baits used. The number of
bait stations ranged from 46 to 78 per TMA. An assessment of which species was likely to have
taken bait was recorded at the time of bait checking. Monitoring the results of the baiting program
has two components:
1. demonstrating the impact of the initial knockdown of the fox population, and
2. demonstrating the sustained reduction of foxes through time, relative to the non-treatment areas.
2.3.1 Index of abundance
At each site, visitations by foxes to bait stations were recorded. Because baits become unavailable
to other foxes once bait has been removed the relationship between fox density and bait take is not
linear. In addition, more than one animal may visit a station, but this was recorded as one visit. This
can be accounted for with the use of a frequency–density transformation (Caughley1997):
ν = –log
e
(1 – ƒ)
where ƒ is the frequency of visitation to bait stations by foxes, and v is the mean density of the
occurrence of fox sign per bait station (Fleming 1997).
To establish an index of pre-baiting fox activity, free-feed (i.e. non-toxic) baits were used for
several weeks prior to the commencement of poisoning on TMA sites. Progressively higher
frequencies of bait take occur as time passes during the free-feed phase, as foxes become familiar
with the location of bait stations and the presence of the baits. The assessment period for pre-toxic
bait take was commenced once variation in daily bait take had stabilised (<10–15% variation).
Contagion causes the daily frequencies of bait take to form a curve that flattens out at high values.
The mean of three or more days after the plateau is reached is used to determine the index of
abundance.
The difference in the percentage of free-feed bait taken and the percentage of poison bait taken was
determined to quantify the immediate effect of 1080 poisoning on the fox population. The
percentage change in the index of abundance was calculated as follows:
% Change = (pre-baiting index – post-baiting index) / pre-baiting index x 100
Once the initial knockdown of foxes had been established, continued low levels of bait take
provided evidence of a sustained reduction in foxes.
NTMAs also had free-feed bait laid initially to determine the presence of foxes. Free-feeding was
undertaken once a year between 2005 and 2013 on each NTMA. Free-feed bait take between the
yearly estimates were compared to test if foxes remain unchanged on the NTMAs.
2.4 Measuring populations of key native species
2.4.1 Site-occupancy modelling
Site occupancy of the three target species (long-nosed potoroo, southern brown bandicoot and
common brushtail possum) were monitored annually at 40 monitoring stations established within
each TMA and NTMA (Figure 1). The selection of locations of monitoring sites was based on
descriptions of the habitat of the target species (Menkhorst 1995). Site location was then stratified
according to the proportion of habitat, based on Ecological Vegetation Division (EVD), within each
TMA and NTMA (Appendix 1). At each monitoring site, nine handiglaze hair tubes (Murray 2005),
baited with peanut butter, rolled oats and honey, were set and checked daily for four consecutive
6 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
days, with tapes replaced each day. These daily surveys represented four repeat surveys of the
monitoring site per sampling period.
Monitoring was undertaken from 2005 to 2013, typically in spring (2005, 2008–2013). Initial
sampling, prior to the commencement of poison baiting was conducted in winter 2005. In 2006
sampling was undertaken in late winter due to staff resource issues, and the spring 2007 sampling at
Mt Clay and Hotspur was delayed due to staff being allocated to fire-fighting duties; as a result
monitoring was undertaken in summer 2007/2008.
The data for each species was summarised by monitoring station (i.e. 40 stations in each of the six
monitoring areas), such that the observed presence–absence of the species on each day j at site i in
year t was indicated by Y
i,j,t
= 1 and 0 respectively for j = 1 – 4 (corresponding to the four days of
monitoring) and years t = 2005,…,2013, (i.e. regardless of how many of the nine hair tubes at a site
contained a hair of a target species).
It was considered likely that the probability of detection using hair tubing could differ among years
because of seasonal effects on foraging behaviour. For each species, the true, but unknown,
occurrence at a site in a given year (z
i,t
) was modelled as a random variate from a Bernoulli
distribution with the parameter ψ
area[i],t
, the probability of occurrence (i.e. occupancy) in a
monitoring area in year t:
z
i,t
~ Bern(ψ
area[i],t
).
The repeat surveys allowed us to construct a detection history for each site, and thus estimate a
separate detection probability for each monitoring area in each year, conditional on the site being
occupied. The observed presence–absence data were modelled as:
Y
i,j,t
~ Bern (z
i,t
× p
t
),
where p
t
is the detection probability in year t and z
i,t
is the true occurrence estimated above.
Modelling the data in a Bayesian context allows us to use the posterior estimates of true occurrence
at each site in each year to derive other quantities of interest. We pooled estimates of true
occurrence to derive estimates of the total number of occupied sites in each year pooled across
TMAs and NTMAs, enabling a direct assessment of an effect of fox control on species occupancy.
We also used the estimates of the true occurrence to derive estimates of average site colonisation
and extinction for each TMA and NTMA. Colonisation (γ
t
) is analogous to the reproduction rate of
a population, and is the probability that a site that was not occupied at time t1 becomes occupied at
t2 (MacKenzie et al. 2003). Persistence (ø
t
) is analogous to survival and is the probability that sites
that were occupied at t1 remain occupied at t2.
Colonisation and persistence are important as they are the processes that drive occupancy, i.e.
occupied sites will remain occupied with probability ø, and unoccupied sites will be colonised with
probability γ. They can either be estimated directly (MacKenzie et al. 2003), or derived from the
estimates of true occupancy as we have done here.
Depending on the distribution of the species, some parameters may be poorly estimated. For
example, if all sites within an area were occupied, then an estimate of colonisation (γ
t
) for the
following interval will not be possible as there are no unoccupied sites to colonise. Similarly, if all
sites are unoccupied, then estimates of persistence (ø
t
) will not be possible.
Treatment effects were determined by calculating the posterior distribution of the difference in
occupancy between treatment and non-treatment areas. Strong, moderate and weak evidence of a
positive treatment effect is indicated by 0.99 or more of the posterior distribution of the difference
being above 0, between 0.95 – 0.99, and between 0.9 – 0.95 respectively. Correspondingly, strong,
moderate or weak evidence of a negative treatment effect are indicated by small probabilities (e.g.
0.01, 0.05 and 0.1 respectively).
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 7
The probabilities correspond to the 95% credible intervals (CI). In general, strong evidence of a
positive treatment effect is indicated by the entire 95% CI being above zero. Similarly, strong
evidence of a negative treatment effect is indicated when the 95% CI is below zero. Where the 95%
CI includes 0, there will be either weak or no evidence for a treatment effect.
The effect of brushtail possums on detection and occupancy of other species
Previous site-occupancy modelling (Robley et al. 2011) suggested that southern brown bandicoots
and long-nosed potoroos were less likely to be detected at sites where common brushtail possums
(CBTP) were present relative to sites where they were absent. We therefore modelled all three
species together in a single model, allowing species-specific parameters as described above. We
assumed the presence of common brushtail possums has an effect on the detection of the other two
species rather than on their probability of occupancy. Specifically, we modelled the detection
probability of long-nosed potoroos and southern brown bandicoots as:
logit(p
i,t
) = α
t
+ (β × z
CBTP
,t
)
where z
(CBTP)
i,t
= 1 if common brushtail possums are present and 0 otherwise at location i in time t,
and α
t
is the probability detection without possums (on the logit scale) and β is the effect of
brushtail possum presence on detection of long-nosed potoroos and/or southern brown bandicoots,
estimated separately for each species.
Independence of mammal detections
Preliminary modelling indicated that conditioning on locations where species were detected at least
once during the four-day sampling period, the probability of detection was consistently higher on
the day following a positive detection compared to the probability of detection following no
detection, indicating a lack of independence in detection between the four-day sampling period.
In order to reduce the lack of independence we pooled days one and two and days three and four
into periods A and B respectively within each year, whilst maintaining the required repeated
sampling. Six models were run with varying specifications of detection probability. All models
were specified to allow detection to vary by species.
All parameters were estimated using OpenBUGS 3.2.1 (Lunn et al. 2009).
2.5 Fox and feral cat diet
Red fox scats were collected opportunistically from roads and tracks and from bait stations
throughout 2012 from the three non-treatment monitoring areas: Annya State Forest, Hotspur State
Forest and Lower Glenelg National Park (LGNP)-north. Scats were also collected opportunistically
during January 2013 from 200 ha plots in the three above sites, plus from a 200 ha area within the
Cobboboonee National Park and Mt Clay State Forest as part of a research project investigating the
interaction between the use of planned burning and predation and the impact on native wildlife.
As part of a supplementary red fox control program using soft-jaw leg-hold traps at LGNP-south
and north, stomachs of feral cats and foxes were removed and frozen within six hours of capture
until content identification could be undertaken by Barbara Triggs.
Determining species consumed and an approximate proportion of the volume of each scat or
stomach sampled is relatively straightforward (Scott 1941). Mammalian hair has distinguishing
features that often allow species-specific differences to be identified under a microscope (Hausman,
1920). Such features have been long used in wildlife studies, largely to identify dietary components
from the stomachs or scats of predators (e.g. Coman 1973, Brunner et al. 1975, Catling et al. 2001).
However, results can provide only an approximate indication of the overall diet, due to differences
in digestibility and passage time (Putman 1984, Litvaitis 2000).
8 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
The hair, bones and teeth of prey items consumed by foxes are indigestible and are passed with
other faecal matter following digestion (Brunner and Coman 1974, Triggs 2004). The
morphometric features of the recovered guard hairs, such as size, medulla pattern, cross section
shape and scale pattern, allow the identity of the prey item to be determined with a high degree of
accuracy for most species (Lobert et al. 2001).
The importance in the diet of each prey type was quantified by the frequency of occurrence
determined as number of faeces containing each prey type x 100/total number of faeces. We
grouped prey items into four categories: insect and plant material; small mammals (e.g. antechinus
species); medium-sized mammals (e.g. possums, potoroos and bandicoots); and large mammals
(e.g. macropods). We used chi-squared tests to investigate differences in dietary items between
sites, seasons and sites with and without fox control.
2.6 Rainfall
Rainfall records were collected from Dartmoor (Station no. 69055; latitude –37.26°S, longitude
150.05°E; approximately 20 km north of the centre of the study site). We used yearly rainfall
deviations from the long-term mean based on winter–winter rain (1 July to 30 June). Since 1907,
rainfall has averaged 70 mm year-1 (SD 46 mm) with a minimum of 7 mm and a maximum of 302
mm. From the beginning of the study period in 2005 until 2009 rainfall was generally below the
long-term average, with a significant rainfall event in 2011 (Figure 2).
Figure 2. The average yearly rainfall from 2000-2012 (1 July to 30 June) (line) and the yearly
rainfall deviance during this period from the long-term average (bars).
-20
-15
-10
-5
0
5
10
15
20
25
30
00 01 02 03 04 05 06 07 08 09 10 11 12
Yearly rainfall deviation (mm)
0
20
40
60
80
100
120
Yearly average rainfall (mm)
Yearly Rainfall Deviation (mm)
Yearly Mean Rainfall (mm)
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 9
3 Results
3.1 Fox control program
3.1.1 Overall reductions in fox abundance
Free-feeding in the TMAs and the NTMAs commenced in July 2005 and continued on the TMAs
until late September 2005. Across all three TMAs 191 poison baits (LGNP-south = 75,
Cobboboonee = 71, Mt Clay 45) were first laid in October 2005. Baits were checked and replaced
every fortnight for the next 10 weeks. Free-feeding continued on 172 bait stations in the NTMAs
(LGNP-north = 46, Annya = 78, Hotspur = 48) until the end of November 2005, with baits being
checked and replaced fortnightly.
There was an 85% reduction in the index of fox abundance on the TMAs following the
implementation of the baiting program (Figure 3).
Figure 3. Percentage change in bait take between July–September 2005 and October–December
2005 pooled across treatment monitoring sites. Bars are 95% confidence limits.
Across each site the level of percentage change in the index of fox abundance was similar with a
79% change at LGNP-south, a 81% change at Cobboboonee and a 76% change at Mt Clay.
Fox numbers remained low on the treatment sites for the duration of the monitoring (Figure 4).
There was a significantly lower index in the TMAs compared to the NTMAs as shown by the non-
overlapping 95% confidence limits in most years (Figure 4).
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
Free Feed Poison
Average index of fox abundance
10 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Figure 4. Index of fox abundance during the poisoned and free-feed period pooled across all
monitoring sites.
Free-feeding on the NTMAs occurred in autumn each year (2005–2013). Bars are
95% confidence intervals.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 11
3.2 Native species response
In each of the sampling years the main species of interest (long-nosed potoroo, southern brown
bandicoot and common brushtail possum) were detected. Figures in Appendices 2 and 3 show the
monitoring stations and years of detection for long-nosed potoroos and southern brown bandicoots
respectively.
Independence of mammal detections
Six models were run with varying specifications of detection probability. All models were specified
to allow detection to vary by species (Table 1). The best model (in terms of minimum deviance)
assumed detection varied only by species and was consistent between areas and by year, but that
there was a relationship between the presence of common brushtail possums on the detection of the
other two species (Table 1). Models that allowed detection to differ by area were the worst relative
to similar nested models. This is primarily due to the low rates of occupancy by species at the site
level, resulting in poor estimates of detection (i.e. very wide credible intervals).
Table 1. Deviances of the six models investigating the influence of common brushtail possum
presence on other species
. The model with the lowest minimum deviance is shown in bold.
Model Deviance
p(spp, Year) 2596
p(spp, Year, Area) 2701
p(spp, Year, common brushtail
possum)
2675
p(spp) 2550
p(spp, Area) 2637
p(spp, common brushtail possum) 2545
The probability of detection was lower at sites that contained common brushtail possum for both
long-nosed potoroos and southern brown bandicoots (Figure 5). We posit that this is due to the hair
tubes becoming ‘swamped’ with common brushtail possum hairs, making identification of the other
species problematic. The detection probabilities are interpreted as the probability of detection of a
species at least once over a two-night period, given it is present in the sampling location. Detection
probabilities of common brushtail possum in any two-night period were high (0.79, 95% CI = 0.76
– 0.81). Detection probabilities of long-nosed potoroo were high in the absence of common
brushtail possum (0.74, 95% CI = 0.67 – 0.80), but low in their presence (0.28, 95% CI = 0.15 –
0.46). Detection probabilities of southern brown bandicoot were the lowest of all three species
(0.52, 95% CI = 0.40 – 0.61) in the absence of common brushtail possum, and only 0.2, 95% CI =
0.12 – 0.31 in the presence of common brushtail possum.
12 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Figure 5. Detection probabilities of common brushtail possum (BTP), and for long-nosed
potoroos (LNP) and southern brown bandicoot (SBB) with and without common brushtail
possums.
3.2.1 Common brushtail possum
The number of sites at which common brushtail possums were detected varied across locations
(Table 2). Common brushtail possums were most common at LGNP-south and north, and least
common at Mt Clay.
Table 2. Actual number of sites at which common brushtail possums were detected for each
location from 40 possible sites.
Site 2005
2006
2007
2008
2009
2010
2011
2012
2013
Cobboboonee (TMA) 6 9 9 16 14 14 17 24 21
LGNP-south (TMA) 22 23 32 34 36 37 36 37 35
Mt Clay (TMA) 2 2 3 1 1 2 2 3 2
Annya (NTMA) 3 4 8 7 12 12 10 12 8
Hotspur (NTMA) 6 8 12 17 18 13 9 8 11
LGNP-north (NTMA) 17 15 21 25 23 20 17 21 21
Total 56 61 85 100 104 98 91 105 98
The estimated number of locations occupied by common brushtail possum in treated areas showed
an apparent increase over time and was higher when compared to sites in non-treatment areas
(Figure 6). This gives rise to an increasing treatment effect of fox control on the distribution of
common brushtail possum (Figure 7).
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 13
Figure 6. Number of sites occupied by common brushtail possums in non-treatment and
treatment areas over time. Bars are 95% credible limits.
Figure 7. Difference in the number of sites occupied by common brushtail possum (treatment –
non-treatment). Bars are 95% credible limits.
The increase in common brushtail possum occupancy appears to be driven by a relatively higher
rate of site persistence in treated areas (Figure 8). Colonisation of unoccupied sites was similar
between treated and untreated areas.
Figure 8. Average probability of colonisation (left) and persistence (right) for common brushtail
possums over time by non-treatment and treatment areas. Bars are 95% credible limits.
0
20
40
60
80
2005
2006
2007
2008
2009
2010
2011
2012
2013
Overall
2005
2006
2007
2008
2009
2010
2011
2012
2013
Overall
Year (Non-treatment) Year (Treatment)
Occuppied sites (n=120)
-10.0
0.0
10.0
20.0
30.0
40.0
50.0
2005 2006 2007 2008 2009 2010 2011 2012 2013
Difference in occupied sites (TAs -
NTAs)
Year
14 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
There was wide variation in the occupancy probabilities between the six study areas (Figure 9).
Average occupancy of common brushtail possum in the three non-treatment areas remained
relatively constant, despite fluctuations over time. Occupancy in two of the treatment areas
(Cobboboonee and LGNP-south) increased since fox control began, with no apparent change in the
Mt Clay area where occupancy remained consistently low.
Figure 9. Average annual probability of occupancy for each of the six study areas for common
brushtail possums.
3.2.2 Long-nosed potoroo
Long-nosed potoroos were detected in low numbers at all sites with the exception of Mt Clay
(Table 3).
Table 3. Actual number of hairtube sites at which long-nosed potoroos were detected for each
location from 40 possible sites.
Location 2005
2006
2007
2008
2009
2010
2011
2012
2013
Cobboboonee (TMA) 0 6 4 0 3 4 4 4 3
LGNP-south (TMA) 5 11 3 2 2 1 1 0 3
Mt Clay (TMA) 8 11 5 9 11 14 11 9 4
Annya (NTMA) 2 3 1 1 1 1 4 1 2
Hotspur (NTMA) 7 0 0 0 0 1 1 1 0
LGNP-north (NTMA) 4 8 4 2 3 0 1 1 1
Total 26 39 17 14 20 21 22 16 13
The estimated number of locations occupied by long-nosed potoroo in treated areas remained
relatively stable over time compared to a decrease in the non-treatment areas (Figure 10). The
difference in occupied sites increased initially following fox control and has remained at similar
levels since (Figure 11).
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 15
Figure 10. Number of sites occupied by long-nosed potoroo in non-treatment and treatment
areas over time. Bars are 95% credible limits.
Figure 11. Difference in the number of sites occupied by long-nosed potoroo (treatment – non-
treatment). Bars are 95% credible limits.
Similar to common brushtail possums, the higher rates of long-nosed potoroo occupancy in
treatment areas appears to be driven by a relatively higher rate of site persistence in treated areas
compared to non-treatment areas (Figure 12). The decrease in site occupancy in non-treated areas
appears to be driven by a very low rate of colonisation of unoccupied sites, dropping to nearly zero
since 2007.
16 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Figure 12. Average probability of colonisation (left) and persistence (right) of Long-nosed
potoroos over time by non-treatment and treatment areas. Bars are 95% credible limits.
There was little variation in the occupancy probabilities of long-nosed potoroo between the six
study areas (Figure 13). Average occupancy of long-nosed potoroo in the three treatment areas
initially increased following the beginning of fox control, but has since remained relatively constant
in two of the study areas, and decreased in the third (LGNP-south). Occupancy in all three of the
non-treatment areas has remained low throughout the entire study period, with an average
occupancy of c. 0.08.
Figure 13. Average annual probability of occupancy for each of the six study areas for long-nosed
potoroos.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 17
3.2.3 Southern brown bandicoot
Southern brown bandicoots were also detected in low numbers across all locations (Table 4), being
slightly more common at Mt Clay until 2012.
Table 4. Actual number of hairtube sites at which southern brown bandicoots were detected for
each location from 40 possible sites.
Location 2005
2006
2007
2008
2009
2010
2011
2012
2013
Cobboboonee (TMA) 0 1 3 3 5 3 2 2 0
LGNP-south (TMA) 5 8 5 8 4 2 2 0 1
Mt Clay (TMA) 5 12 9 9 6 8 4 1 0
Annya (NTMA) 0 1 0 2 0 1 3 7 4
Hotspur (NTMA) 4 0 0 0 0 0 0 0 2
LGNP-north (NTMA) 4 9 5 7 1 4 2 6 1
Total 18 31 22 29 16 18 13 16 8
The estimated number of locations occupied by southern brown bandicoot in treated areas initially
rose following the initiation of fox control, from 15 location in 2005 to 36 in 2008 (Figure 14).
Since that time, however, the number of occupied locations decreased substantially to an estimated
six out of 120 sites in treated areas in 2013. In the non-treatment areas, the number of occupied
locations remained relatively stable over time. The recent decline in occupied sites in the treatment
areas resulted in a higher proportion of occupied locations in the non-treatment area compared to
the treatment area in 2013 (Figure 15).
Figure 14. Number of sites occupied by southern brown bandicoot in non-treatment and
treatment areas over time. Bars are 95% credible limits.
0.0
20.0
40.0
60.0
80.0
2005
2006
2007
2008
2009
2010
2011
2012
2013
Overall
2005
2006
2007
2008
2009
2010
2011
2012
2013
Overall
Year (Non-treatment) Year (Treatment)
Occupied sites (n=120)
18 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Figure 15. Difference in the number of sites occupied by southern brown bandicoot (treatment –
non-treatment). Bars are 95% credible limits.
The decline in southern brown bandicoot occupancy in treatment areas appears to be driven by both
a decrease in colonisation of new sites and a decrease in site persistence (Figure 16). In non-
treatment areas, colonisation has remained low, but stable. Persistence has varied widely, however,
this is somewhat of an artefact of the low numbers of occupied sites each year (c. 10 out of 120).
Figure 16. Average probability of colonisation (left) and persistence (right) for southern brown
bandicoots over time by non-treatment and treatment areas. Bars are 95% credible limits.
The proportion of sites occupied on the NTAs was low, and generally lower than on the TAs
(Figure 17). No southern brown bandicoot have been detected in LGNP-north since 2005. The
notable exception being at Annya from 2010 onwards, which has seen steady increase in occupied
sites and occupancy at Hotspur has tended to increase from 2011 onwards. The proportion of sites
occupied at these sites is now greater than on the three TAs. Average occupancy at two of the three
treatment areas has decreased from 2008 to 2013 to levels below those at the beginning of the
program.
-30.0
-20.0
-10.0
0.0
10.0
20.0
30.0
40.0
50.0
2005 2006 2007 2008 2009 2010 2011 2012 2013
Difference in occupied sites (Tas -
NTAs)
Year
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 19
Figure 17. Average annual probability of occupancy for each of the six study areas for southern
brown bandicoots.
3.2.4 Fox and feral cat diet
3.2.4.1 Fox scats
Fox scats were collected from Annya and Hotspur State Forests and LGNP-north (all no fox control
sites) in autumn, winter and spring 2012 and summer 2013. Additional scats were collected at these
sites plus at Cobboboonee and Mt Clay in summer 2013 (Table 5).
Table 5. Location, season and number of fox scats collected during this study.
Site Treatment Aut-12 Win-12 Spr-12 Sum-13 Total
Annya No Fox Control
26 9 26 12 73
Hotspur No Fox Control
24 16 34 11 85
LGNP-north No Fox Control
28 3 2 11 44
Cobboboonee Fox Control
14 14
Mt Clay Fox Control
4 16 20
Total
78 32 62 64 236
Nineteen prey items were identified from fox scats including 13 native species, 5 introduced species
and insect/plant/reptile material. Overall common ringtail possum (Pseudocheirus peregrinus;
29%), insect, plant and reptile remains (27%), and western grey kangaroos (Macropus fuliginosus;
15%) were the three main food items in fox scats, common brushtail possum and swamp wallaby
(Wallabia bicolor) made a minor contributions. Southern brown bandicoots (3%) and long-nosed
potoroo (1.7%) were also identified in scats (Figure 18).
0.0
0.2
0.4
0.6
0.8
1.0
2005
2006
2007
2008
2009
2010
2011
2012
2013
2005
2006
2007
2008
2009
2010
2011
2012
2013
Year (Non-Treatment) Year (Treatment)
Proportion of area occupied (POA)
Annya Cobboboonee Hotspur
Mt Clay LGNP North LGNP South
20 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Figure 18. Overall proportion of identified food items from fox scats in this study.
Chi-squared tests revealed a significant difference in prey items and sites (X
2
=16.9, DF=3, P<0.05),
with more insect/plant/reptile material and less small mammal prey consumed than expected at
Annya, and at LGNP more insect/plant/reptile material and less small mammal prey were
consumed than expected. There were only sufficient numbers of prey items for comparison between
Annya, Hotspur and LGNP.
The most common items identified in fox scats at Annya were common ringtail possums (42%) and
plant/ insect/reptile material (36%). Western grey kangaroo formed 18% of the diet, most likely as
carrion (Figure 19). At Hotspur plant, insect and reptile remains made up 24% and common ringtail
possums 22%. Both long-nosed potoroo (3%) and southern brown bandicoot (5%) were found in
scats at Hotspur. At LGNP-north common ringtail possum and common brushtail possums made up
20% each with western grey kangaroos and swamp wallaby making up 12% each and swamp rat
(Rattus lutreolus) comprising 10%. Long-nosed potoroo were also a small component (8%) of the
diet at LGNP-north. The two most common prey items at Cobboboonee were insect/plant/reptile
(23%), and dusky antechinus (Antechinus swainsonii; 23%), followed by ringtail possums (15%)
and western grey kangaroos (15%). At Mt Clay insect/plant/reptile material (40%), western grey
kangaroo (30%) and common ringtail possum (20%) were the three most abundant prey items.
These two sites has less species recorded in the diet samples, however, this in part is likely to be
due, at least in part, to the small sample size at these sites.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 21
Figure 19. Proportion of items identified in red fox scats across all sites with season combined.
There were no statistically significant differences in groups of prey between autumn, winter, spring
2012 and summer 2013 for all sites combined (X
2
= 13.7, DF = 9, P>0.15). In autumn common
ringtail possums and plant/insect/reptile remains made up 32% and 27% respectively, with swamp
wallaby comprising 13% (Figure 8). In winter common ringtail possums were still a significant
item (22%) with plant, insects and reptile remains (16%) and western grey kangaroos (16%) both
forming important components, and common brushtail possums making up 11%. In spring food
items were made up of common ringtail possums and plant material (32% each). In summer
western grey kangaroos (22%) and ringtail possums (21%) made up the bulk of the diet, with
insect, plant and reptile remains contributing 16% and swamp wallabies 11%.
22 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
To investigate differences between treated and non-treated sites we combined data from
Cobboboonee and Mt Clay in summer 2013 (treatment sites) and from Annya and Hotspur for
summer 2013 (non-treatment sites) and grouped prey items into four main categories: insect, plant
material; small mammals; medium-sized mammals; and large mammals. There was no statistical
difference in grouped prey items (X
2
= 0.9, DF = 1, P>0.25).
3.2.5 Fox and feral cat stomach and gut contents
Eighteen foxes were collected from LGNP, 10 from the north and 8 from the south. Overall the
three main food items identified were common ringtail possum, common brushtail possum and red-
necked wallaby (Macropus rufogriseus), which collectively made up greater than 60% of food
items (Figure 20).
Seven native species and four introduced species were identified from 60 feral cat stomach and gut
samples, 48 from LGNP-south and 12 from LGNP-north. Common ringtail possum and swamp rat
were the two most common items in feral cat diet overall, together comprising more than 50% of
food items.
Figure 20 shows that both species overlapped in their diet, sharing common ringtail possums as a
significant food item. A wider range of food items were identified in feral cat stomach and gut
samples; however, this could be a reflection of the larger sample sizes.
Figure 20. Proportion of items identified in fox and feral cat stomach and guts from LGNP.
The larger sample size of feral cats allowed us to look at the difference between LGNP-south (fox
control) and north (no fox control). Swamp rat and common ringtail possum were most common in
the south, while these and antechinus species were the three most common items in feral cat diet in
the north (Figure 21).
The diet of feral cats differed slightly between the fox control and no fox control area. Ten prey
items were identified in the south while only eight were found in the smaller sample size from the
north. Sugar gliders and swamp wallaby were found in samples from the south but not the north,
and bush rat were included in the north but not found in LGNP-south.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 23
Figure 21. Frequency of occurrence of prey items identified from feral cat gut and stomach
analysis from LGNP-south (fox control area) and north (no fox control area).
24 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
4 Discussion
Our results show that the nine-year landscape scale management experiment was able to reduce
foxes immediately following poison baiting and that a new lower level of foxes could be
maintained continuously on treatment sites. The decline in the proportion of baits taken in our study
is similar to those reported elsewhere in Australia (66–73%: Thompson and Fleming 1994; 97%:
Dexter and Meek 1998; 95%: Thomson et al. 2000) and in Europe and North America (40–78%:
reviewed by Trewhella et al. 1991). We showed a general positive response in three native prey
species, although the level of response varied between species. This result is supported by similar
finding in Western Australia (southern brown bandicoot, brushtail possum: Kinnear et al. 2002) and
eastern Victoria (same three species as in our study: Dexter and Murray 2009) that report positive
responses to the reduction in foxes.
Variable prey responses to a reduction in predation were found by Salo et al. (2010), with one in six
experiments (16.4%, 19 of 116) in their worldwide review not showing a predator manipulation
effect (i.e., effect size was 0). McLeod et al. (2008) found similar outcomes in their review of
Australian fox control programs, with 35% (5 of 14) of control operations resulting in no or
negative responses in native species to the reduction in foxes. The lack of effect may be related to
the implementation of the management operation. For example, a reduction in predator abundance
may produce no impact if predators are not reduced below the threshold at which prey species' rate
of increase can be positive (Sinclair and Krebs 2002). Predator species may compensate for
mortality, especially if predator removal allows the release of mesopredators that in turn suppress
the prey species (Soule et al. 1989; Courchamp et al. 1999; Glen and Dickman 2005), or if
resources are limiting at some sites (e.g. food; Sinclair 1975).
The threshold at which prey species escape limitation by predation has not been established for any
of the prey species in this study. Sinclair and Krebs (2002) showed that it may not be necessary to
remove predators totally to increase prey populations. The level of reduction in predators depends
on the underlying rate of increase in the prey population, and how close this is to the point where it
shifts from negative to positive. If the rate of increase is close to positive, only a modest level of
reduction in predation could result in an increase in the prey species (e.g. possibly brushtail
possums in this study); if the rate is strongly negative then near total removal may be required.
While it is clear that it is possible to maintain reduced numbers of foxes, what is less clear is the
level to which foxes need to be reduced to allow native mammal species to escape any predator
induced limitation. While we have quantified the degree of change in relative abundance of foxes
pre- and post-baiting and between treated and non-treated sites, fully understanding the
density/damage relationship is considerable more complicated.
The mesopredator release hypothesis (Soule et al. 1989) predicts that the decline of the most
common large predator would result in the ecological release of native or exotic mesopredators, and
that increased predation by these effective predators would result in higher mortality and local
extinctions of prey species. The domestic cat was most likely introduced into Australia at multiple
points along the coastline during the period 1824–1886 (Abbott 2002) with feral populations now
widespread and common in the landscape. Supplementary fox control that commenced in 2010 at
LGNP-south and north using leg-hold traps captured feral cats as a non-target species; this and
camera survey data indicate that the feral cat population at LGNP-south was three times higher than
on the non-treatment site (Robley et al. 2010). While no pre-fox control data are available on cats at
these sites, it suggests that cats have the potential to assume the role as the apex mammalian
predator. Southern brown bandicoots declined following an initial increase after fox control
commenced, a possible explanation is that after escaping limitation by foxes bandicoots increased;
feral cats may have also responded to the release of foxes and subsequently reduced bandicoot
populations. The hypothesis is supported by Arthur et al. (2012) who found that feral cats either
responded to bandicoots in a bottom-up relationship, i.e., cats increased in response to the increase
in bandicoots, or the top-down effects of cat predation contributed to the decline in bandicoots.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 25
The overall derived estimate of occupancy indicates that bandicoots have declined markedly at the
treatment sites since 2009. Although we have not measured population size directly, a decreasing
population would likely lead to this pattern of decreasing persistence (individuals unable to
maintain a site), and decreasing colonisation (a shrinking population would not be able to spread to
inhabit new sites).
Occupancy of locations is driven by the processes of species persisting at sites from one year to the
next, and the colonisation of new, previously unoccupied sites. These processes are analogous to
survival and reproduction in populations. To persist at sites three things are needed: 1) suitable
resources are present and remain so at new sites, 2) predation (direct or indirect) is absent (or below
a critical threshold) or habitat allows for avoidance of predation, and 3) competition (real or
apparent) is absent or below a critical threshold. To colonise new sites two things must happen: 1)
immigration from other ‘sites’, which in this case implies successful breeding and recruitment into
the population, and 2) suitable habitat is available (and that we are sampling in that habitat).
Habitat, defined as an area with a combination of resources (e.g. food, water, cover) and
environmental conditions (e.g. temperature, precipitation, presence or absence of predators and
competitors), that promotes occupancy by species allowing for survival and reproduction (Morrison
et al. 2006) may be limited at some sites within the treated areas. Hence despite the adequate
reduction in predators, occupancy of new sites is limited and a general response to the reduction of
foxes is variable. This may be the case for long-nosed potoroos. This species increased shortly
following fox control, but decreased in subsequent years. The suggestion of habitat limitation is
supported by findings of Claridge et al. (2010) that investigated activity levels of forest-dwelling
vertebrates in areas with and without fox control with occupancy lower in 2012 than at the start.
These authors found that while overall prey species activity was greater in areas with fox control,
this response was not universal across all sites, and that habitat quality may have influenced species'
response. At least at one of our sites, Cobboboonee, southern brown bandicoots may be limited by
habitat. In a study investigating the distribution of feeding sites, Rees and Paul (2000)found this
species at only one site from 30 randomly allocated sites within the southern section of this large
public land block. There was no fox control in place at the time of their surveys, so interpretation of
the relative role of predation and habitat is confounded.
Other possible causes for prey not increasing after predator removal could be if they represent
alternative prey for the predator (Norrdahl and Korpimäki 2000), or if the predator preys on a major
competitor of the studied prey (Abrams 1992). These alternatives and those discussed above
suggest that while predator limitation may be general, more focus is needed on cases where
limitation does not seem to occur (Claridge et al. 2010; Salo et al. 2010).
Independence of the treatment and non-treatment sites is an important issue as we compare rates of
change in foxes between sites. A trend towards a reduction in the index of abundance in foxes was
observed on the LGNP-north site; however, relative to the treatment site the index remained higher.
At the remaining two non-treatment sites the index of abundance also remained higher, with a
marked increase at two of the three sites in 2010 to 2013. In general we did not observe any effect
of a regional wide decline in foxes from baiting a few, large areas as purported in Dexter and
Murray (2009). The reduction in the fox abundance index and relative difference between treatment
and non-treatment sites in our study enabled us to investigate the benefits of long-term fox control
to native mammals.
Dietary studies of foxes in Australia have shown that foxes consume a wide array of foods. Where
introduced European rabbits (Oryctolagus cuniculus) are widely present in the landscape, this
species is dominant in the diet (McIntosh 1963, Coman 1973, Croft and Hone 1978, Catling 1988,
Molsher et al. 2000). In areas where rabbits are not present or uncommon, native mammals are the
dominant dietary component. This and other studies support the observation that foxes have a
highly diverse diet lacking significant specialisation (Diment 2010, Saunders et al. 1995, 2004).
26 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Feral cats are similarly opportunistic predators and also eat a wide variety of foods. Their diet at
any time will usually consist of those species of prey most available to them. Dietary studies have
shown that the rabbit is the major food item of feral cats in Victoria, but mice, smaller native
mammals, reptiles, birds and invertebrates are also eaten. However, when live prey is scarce, feral
cats will also scavenge food scraps.
Interpretation of the results from this study needs to be undertaken with caution. There are two
widely used approaches to analysis of diet in predators: 1) percentage occurrence of identified
items, as used in this study, and 2) percentage volume. Both approaches have limitations. The first
tends to overestimate the importance of minor components and the second suffers from differential
rates of digestion, inclusion of non-digestible items, and subjectivity of estimates (Hyslop 1980).
Common ringtail possum was a commonly shared prey item between red foxes and feral cats at
LGNP, and both species often made use of common brushtail possums. Red fox and feral cats
shared seven species of prey, albeit in different proportions. Feral cats appear to have slightly
broader diets than foxes.
Roberts et al. (2006) studied changes in fox diets in relation to the control of foxes. They noted
great increases in the frequency of ringtail possums and long-nosed bandicoots (Perameles nasuta)
in fox scats several years after baiting. They surmised that dietary changes were related to an
increase in abundance in the prey animals as a result of reduced predation pressure following fox
control.
If reducing foxes allows for feral cats to increase in abundance there may be increased levels of
predation pressure on those shared prey items. There is some evidence for this from LGNP where
feral cats have a higher rate of occupancy at the LGNP-south site (reduced fox abundance)
compared to the north side (relatively higher of abundance) (Robley et al. 2010).
Where foxes have been reduced (LGNP-south), they had a greater diversity of native mammals in
the diet. This may reflect a generalised response of native mammals to the reduction in foxes, i.e.,
there is more prey species available or encountered thus appearing more (but in low proportions) in
their diet.
Ringtail possums were a common item identified in the stomach and gut of foxes. This combined
with their prevalence in the scats from both Annya and Hotspur indicates this species is the rabbit
of the bush in terms of fox diet, and is the species that is most likely supporting fox populations in
this environment. An important implication of this finding is the potential impact of planned burns
and wildfire on this species. Russell et al. (2003) found that fire significantly reduced a population
of ringtail possums following wildfire and that native and introduced predators suppressed the
population and prevented recovery. It is generally accepted that foxes have a catholic diet and
consume prey in relation to the prey abundance (Saunders et al. 1995). The potential for prey-
switching would appear to be significant in areas where ringtail possum dominate the diet and are
significantly reduced in abundance following fire.
The results of this work clearly demonstrate that foxes have been reduced and remain at relatively
low levels due to the fox control, and that the use of occupancy modelling as applied in this study is
capable of measuring changes in native species. However, the variable results in colonisation and
persistence across species suggest that blanket, broad-scale fox control may not be the only factor
limiting recovery or protection of species threatened by fox predation, and that a greater
understanding of both bottom-up and top-down processes may lead to more effective and targeted
management of species that are rare in the landscape.
Developing multi-species dynamic occupancy models would explicitly link the probabilities of
foxes and native species occupying areas, allowing for more direct estimates of the link between a
reduction in fox abundance (and potentially the presence of other predators) and any increase in
native species. These models could include site-specific covariates to help understand the non-
uniform response of target species. More complex models could be used to explore additional
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 27
factors of occupancy and detection processes, including: inter-seasonal variation in detection
probability due to weather; habitat variables such as vegetation structure, rainfall, temperature,
exposure to solar radiation; spatial effects such as adjacency to known occupied or unoccupied
sites; and the impact of fuel reduction burning and wildfires.
Ideally we would have assessed the rates of population increase of the three prey species prior to
baiting, used direct assessments of abundance of prey and predators (both foxes and feral cats),
randomised the treatment and non-treatment sites, and had a larger number of replicates. No
published studies on the response of native species to reductions in predators have included all
these facets, and many have had only a few components of experimental design (Platt 1964).
Reddiex and Forsyth (2006) reviewed 633 fox control operations from 1990–2003, with 82%
having the stated objective of species conservation. Of these 15 (2.4%) had a single treatment and
non-treatment area and 6 operations (0.9%) had >2 non-randomised treatment and non-treatment
areas. No operations had completely randomised treatment and non-treatment areas. McLeod et al.
(2008) identified a similar number (n=467 fox control operations) of which only 14 (3%) had
treatment and non-treatment sites (median = 4), with monitoring of both foxes and native species
(median < 2 years), they noted that few results of fox control operations are reported.
While there are limitations to our study, the three treatment and non-treatment areas monitoring
fox and native species numbers is more rigorous than other studies and can be used to guide future
studies. Non-manipulative studies have demonstrated that predation plays a role in limiting primary
prey populations, but they cannot be used to unequivocally assess whether predation is a limiting
factor because of the potential confounding effect of other factors as discussed above. The
landscape scale management experiment of foxes in our study has demonstrated that continuous
suppression of the fox population is possible and that in general prey species respond positively,
however, response is complicated and potentially limited by other factors. Conservation managers
will need to increasingly consider the role of these other factors and develop management strategies
if broad-scale recovery of native species is to be achieved.
28 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
References
Abbott, I. (2002). Origin and spread of the cat, Felis catus, on mainland Australia, with a discussion
of the magnitude of its early impacts on native fauna. Wildlife Research 29, 51-74.
Abrams, P.A. (1992). Predators that benefit prey and prey that harm predators: unusual effects of
interacting foraging adaptations. The American Naturalist 140, 573-600.
Agresti, A. (2002). Categorical Data Analysis. Wiley Interscience, New York.
Allen, L., Engerman, R. and Krupa, H. (1996). Evaluation of three relative abundance indices for
assessing dingo populations. Wildlife Research 23, 197–206.
Anson, J.R., Dickman, C.R., Boonstra, R., Jessop, T. S. (2013). Stress triangle: do introduced
predators exert indirect costs on native predators and prey? PLoS One 8: e60916. doi:
10.1371/journal.pone.0060916. Print 2013.
Arthur, A.D., Catling, P.C. and Reid, A. (2012). Relative influence of habitat structure, species
interactions and rainfall on the post-fire population dynamics of ground-dwelling
vertebrates. Austral Ecology 37, 958–970. doi: 10.1111/j.1442-9993.2011.02355.x.
Boonstra, R., Hik, D. and Singleton, G.R. (1998). The impact of predator-induced stress on the
snowshoe hare cycle. Ecological Monographs 68, 371–394.
Brunner, H. and Coman, B.J. (1974). The Identification of Mammalian Hair. Inkata Press,
Melbourne.
Brunner, H., Lloyd, J. and Coman, B. (1975). Fox scat analysis in a forest park in south-eastern
Australia. Wildlife Research 2, 147–154.
Burbidge, A.A. and McKenzie N.L. (1989). Patterns in the modern decline of Western Australia's
vertebrate fauna: causes and conservation implications. Biological Conservation 50, 143–
198.
Burnham, K.P. and Anderson D.R. (2010). Model selection and multi-model inference: A practical
information-theoretic approach. Springer-Verlag, New York.
Catling, P.C., Coops N.C. and Burt R.J. (2001). The distribution and abundance of ground-dwelling
mammals in relation to time since wildfire and vegetation structure in south-eastern
Australia. Wildlife Research 28, 555565.
Caughley, G. (1977). Analysis of vertebrate populations. John Wiley, London.
Claridge, A.W., Cunningham, R.B., Catling, P.C., and Ried, A.M. (2010). Trends in activity levels
of forest-dwelling vertebrate fauna against a background of intensive baiting for foxes.
Forest Ecology and Management 260, 822-832.
Coman, B. (1973). The diet of red foxes, Vulpes vulpes, in Victoria. Australian Journal of Zoology
21, 391–401.
Courchamp, F., Langlais, M., and Sugihara, G.(1999). Cats protecting birds: modelling the
mesopredator release effect. Journal of Animal Ecology 68, 282–292. DOI:10.1046/j.1365-
2656.1999.00285.x.
Cove, M.V., Jones, M., Bossert, A. J., Clever, D. R., Dunwoody, R. K., White, B. C. and Jackson,
V. (2012). Use of camera traps to examine the mesopredator release hypothesis in a
fragmented midwestern landscape. The American Midland Naturalist 168, 456–465.
Croft, J. and Hone, L. (1978). The stomach contents of foxes, Vulpes vulpes, collected in New
South Wales. Wildlife Research 5, 85–92.
Dexter, N., and Meek, P. (1998). An analysis of bait-take and non-target impacts during a fox
control exercise. Wildlife Research 25, 147-155.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 29
Dexter, N. and Murray, A. (2009). The impact of fox control on the relative abundance of forest
mammals in East Gippsland, Victoria. Wildlife Research 36, 252–261.
Diment, A.N. (2010). Monitoring the ecological impacts of invasive predator control. PhD thesis,
University of Sydney.
Elton, C.S. (1958). The ecology of invasions by animals and plants. Methuen and Company,
London, UK.
Fleming, P. (1997). Uptake of baits by red foxes (Vulpes vulpes): Implications for rabies
contingency planning in Australia. Wildlife Research 24, 335–346.
Frey, S. J. K., Strong, A. M. and McFarland, K. P. (2011). The relative contribution of local habitat
and landscape context to metapopulation processes: a dynamic occupancy modelling
approach. Ecography 35, 581–589.
Groombridge, B. (1992). Global biodiversity. London: Chapman & Hall.
Hausman, L.A. (1920). Structural characteristics of the hair of mammals. The American Naturalist
54, 496–523.
Holden, C. and Mutze, G. (2002). Impact of rabbit haemorrhagic disease on introduced predators in
the Flinders Ranges, South Australia. Wildlife Research 29, 615–626.
Holling, C.S. (1959). The components of predation as revealed by a study of small mammal
predation of the European pine sawfly. Canadian Entomology 91, 293–320.
Hosmer, D.W. and Lemeshow, S. (2000). Applied logistic regression. Wiley Interscience, New
York.
Hyslop, E.J. (1980) Stomach content analysis: a review of methods and their applications. Journal
of Fish Biology 17, 411–429.
Innes, J., Hay, R., Flux, I., Bradfield, P., Speed, H. and Jansen, P. (1999). Successful recovery of
North Island kokako Callaeas cinerea wilsoni populations, by adaptive management.
Biological Conservation 87, 201–214.
Kerle, J.A. and How, R.A. (2008). Common Brushtail Possum. Pp. 274–275 In S. Van Dyck and R.
Strahan (eds). The Mammals of Australia 3
rd
Edition. Reed New Holland, Sydney,
Australia.
Kinnear, J., Summer, N.R. and Onus, M.L. (2002). The red fox in Australia – an exotic predator
turned biocontrol agent. Biological Conservation 108, 335–359.
Kinnear, J., Krebs, C., Pentland, C., Orell, P., Holme, C.C. and Karvinen, R.T. (2010). Predator-
baiting experiments for the conservation of rock-wallabies in Western Australia: 25-year
review with recent advances. Wildlife Research 37, 57–67.
Krebs, C.J., Boutin, S. and Boonstra, R. (2001). Ecosystem dynamics of the Boreal Forest: The
Kluane Project. Oxford University Press, New York.
Lande, N., Ryan, L., Billard, L. and Farber, D. (1994). Case studies in biometry. John Wiley and
Sons, New York.
Lazenby, B.T. and Dickman, C.R. (2013). Patterns of detection and capture are associated with
cohabiting predators and prey. PLoS ONE 8(4): DOI:10.1371/journal.pone.0059846.
Lima, S.L. and Dill, L.M. (1990). Behavioural decisions made under the risk of predation: a review
and prospectus. Canadian Journal of Zoology 68, 619–640.
Litvaitis, J. (2000). Investigating food habits of terrestrial vertebrates. Pp. 165–190 In Research
techniques in animal ecology: controversies and consequences. Columbia University Press,
New York.
30 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Lloyd, H. G. (1980). The red fox. B. T. Batsford, London.
Lobert, B. and Lee, A.K. (1990). Reproduction and life history of Isoodon obesulus in Victorian
heathlands. Pp. 311–318 In J.H. Seebeck, P.R. Brown, R.L. Wallis and C.M. Kemper (eds).
Bandicoots and Bilbies. Surrey Beatty and Sons, Chipping North, New South Wales.
Lobert, B., Lumsden, L., Brunner, H. and Triggs, B. (2001). An assessment of the accuracy and
reliability of hair identification of south-east Australian mammals. Wildlife Research 28,
637–641.
Lunn, D., Spiegelhalter, D., Thomas, A. and Best, N. (2009). The BUGS project: Evolution,
critique, and future directions, Statistics in Medicine 28, 3049–3067.
Lunney, D. (2001). Causes of the extinction of native mammals of the western division of New
South Wales: an ecological interpretation of the nineteenth century historical records.
Rangelands Journal 23(1), 44–70.
MacKenzie, D.I., Nichols, J D., Hines, J.E., Knutson, M.G. and Franklin, A.B. (2003). Estimating
site occupancy, colonization, and local extinction when a species is detected imperfectly.
Ecology 84, 2200–2207.
MacKenzie, D.I., Nichols, J.D., Royle, J.A., Pollock, K.H., Bailey, L.L. and Hines, J.E. (2006).
Occupancy estimation and modelling: inferring patterns and dynamics of species
occurrence. Elsevier, San Diego.
McIntosh, D.L. (1963). Food of the fox in the Canberra district. CSIRO Wildlife Research 8, 1–20.
McLeod, L.J., Saunders, G.R. and Kabat, T.J. (2008). Do control interventions effectively reduce
the impact of European Red Foxes on conservation values and agricultural production in
Australia? Systematic Review No. 24. Collaboration for Environmental Evidence,
www.environmentalevidence.org.
Menkhorst, P. (1995). (Ed.) Mammals of Victoria, distribution, ecology and conservation. Oxford
University Press, South Melbourne.
Molsher, R., Gilford, E. and McIlroy, J. (2000). Temporal, spatial and individual variation in the
diet of red foxes (Vulpes vulpes) in central New South Wales. Wildlife Research 27, 593–
601.
Morrison, M., Marcot, B., and Mannan, W. (2006). Wildlife-habitat relationships: concepts and
applications. Island Press, Washington, D.C.
Murray, A. (2005). A new low-cost hairtube design for the detection of the spotted-tailed quoll
Dasyurus maculatus in south-eastern Australia. Australian Mammalogy 27, 81–84.
Norrdahl, K., and Korpimäki, E. (2000). The impact of predation risk from small mustelids on prey
populations. Mammal Review 30, 147–156.
Park, K. (2004). Assessment and management of invasive alien predators. Ecology and Society 9,
12–28.
Pech, R.P., Sinclair, A.R.E. and Newsome, A.E. (1995). Predation models for primary and
secondary prey species. Wildlife Research 22, 55–64.
Platt, J. R. (1964). Strong inference. Science 146, 347-354.
Putman, R.J. (1984). Facts from faeces. Mammal Review 14, 79–97.
Quinn, G.P. and Keough, M.J. (2002). Experimental Design and Data Analysis for Biologists.
Cambridge University Press, Cambridge.
Reddiex, B. and Forsyth, D.M. (2006). Control of pest mammals for biodiversity protection in
Australia. II. Reliability of knowledge. Wildlife Research 33, 711.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 31
Rees, M., and Paull, D. (2000). Distribution of the southern brown bandicoot (Isoodon obesulus) in
the Portland region of south-western Victoria. Wildlife Research 27, 539-545.
Risbey, D.A., Calver, M.C., Short, J., Bradley, J.S. and Wright, I.W. (2000). The impact of cats and
foxes on the small vertebrate fauna of Heirisson Prong, Western Australia. II. A field
experiment. Wildlife Research 27, 223–235.
Roberts, M.W., Dexter, N., Meek, P.D., Hudson, M. and Buttemer, W.A. (2006). Does baiting
influence the relative composition of the diet of foxes? Wildlife Research 33, 481–488.
Robley, A. Gormley, A., Woodford, L., Lindeman, M., Whitehead, B., Albert, R., Bowd, M. and
Smith, A. (2010). Evaluation of sampling designs used to determine change in occupancy
rate and abundance of feral cats. Arthur Rylah Institute for Environmental Research
Technical Report No. 201. Department of Sustainability and Environment, Heidelberg,
Victoria.
Robley, A., Gormley A., Albert, R., Bowd, M., Hatfield. C., McDonald, A. Thorpe, C., Scroggie,
M., Smith, A. and Warton, F. (2011). Glenelg Ark 2005–2010: evidence of sustained
control of foxes and benefits for native mammals. Arthur Rylah Institute for Environmental
Research Technical Report Series No. 226. Department of Sustainability and Environment,
Heidelberg, Victoria.
Russell, B.G., Smith, B. and Augee, M.L. (2003). Changes to a population of common ringtail
possum (Pseudocheirus peregrinus) after bushfire. Wildlife Research 30, 389–396.
Salo, P., Korpimaki, E., Banks, P.B., Nordstrom, M. and Dickman, C. (2007). Alien predators are
more dangerous than native predators to prey populations. Proceedings of the Royal Society
B. 274, 1237–1243.
Salo, P., Banks, P.B., Dickman, C.R. and Korpimäki, E. (2010). Predator manipulation
experiments: impacts on populations of terrestrial vertebrate prey. Ecological Monographs
80, 531–546. http://dx.doi.org/10.1890/09-1260.1.
Saunders, G., Coman, B., Kinnear, J. and Braysher, M. (1995). Managing Vertebrate Pests: Foxes.
Australian Government Publishing Service, Canberra.
Saunders, G., Berghout, M., Kay, B., Triggs, B., van de Ven, R. and Winstanley, R. (2004). The
diet of foxes (Vulpes vulpes) in south-eastern Australia and the potential effects of rabbit
haemorrhagic disease. Wildlife Research 31, 13–18.
Saunders, G. and McLeod, L. (2007) Improving fox managements strategies in Australia.
Australian Government Publishing Service, Canberra.
Saunders, G.R., Gentle, M.N. and Dickman, C.R. (2010). The impacts and management of foxes
Vulpes vulpes in Australia. Mammal Review 40, 181–211.
Seebeck, J. H. (1978). Diet of the fox Vulpes vulpes in a western Victorian forest. Australian
Journal of Ecology 3, 105–108.
Scott, T.G. (1941). A method for estimating the red fox population. Iowa State College Journal of
Science, 15, 155–159.
Short, J. and Smith, A. (1994). Mammal decline and recovery in Australia Journal of Mammalogy
75, 288–297.
Sinclair, A.R.E. (1975). The resource limitation of trophic levels in tropical grassland ecosystems.
Journal of Animal Ecology 44, 497-520.
Sinclair, A.R.E., Pech, R., Dickman, C.R., Hik, D., Mahon, P. and Newsome, A.E. (1997).
Predicting the effects of predation on conservation of endangered prey. Conservation
Biology 12, 564–575.
32 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Sinclair, A. and Krebs, C. (2002). Complex numerical responses to top-down and bottom-up
processes in vertebrate populations. Pp. 127–147 In R.M. Siby, J. Hone and T.H. Clutton-
Brock (eds). Wildlife Population Growth Rates. Cambridge University Press, Cambridge,
UK.
Soule, M.E., Bolger, D.T., Alberts, A.C., Wright, J., Sorice, M., and Hill, S. (1989). Reconstructed
dynamics of rapid extinctions of chaparral-requiring birds in urban habitat islands.
Conservation Biology 2, 75-92.
Thomson, P.C., Marlow, N.J., Rose, K., and Kok, N.E. (2000). The effectiveness of a large-scale
baiting campaign and an evaluation of a buffer zone strategy for fox control. Wildlife
Research 27, 465-472.
Thompson, J.A. and Fleming, P. (1994). Evaluation of the efficacy of 1080 poisoning of red foxes
using visitation to non-toxic baits as an index of fox abundance. Wildlife Research 21, 27–
40.
Trewhella, W.J., Harris, S., Smith, G.S., and Nadian, A. K. (1991). A field trial evaluating bait
uptake by an urban fox (Vulpes vulpes) population. Journal of Applied Ecology 28, 454-
466.
Triggs, B. (2004). Tracks, Scats and Other Traces: A Field Guide to Australian Mammals. Oxford
University Press, South Melbourne, Victoria.
Triggs, B., Brunner, H., Cullen, J.M. (1984). The food of fox, dog and cat in Croajingalong
National Park, southeastern Victoria. Australian Wildlife Research 11, 491–499.
Vitousek, P.M., Mooney,H. A., Lubchenco, J. and Melillo, J.M. (1997). Human domination of
Earth’s ecosystems. Science 277, 494–499. DOI:10.1126/science.277.5325.494.
Walters, C.J. (1986). Adaptive Management of Renewable Resources. McMillan: New York.
Walters, C.J. and Holling, C.S. (1990). Large-scale management experiments and learning by
doing. Ecology 71, 2060–2068.
Zuur, A. F., Ieno E. N. and Smith G.M. (2007). Analysing Ecological Data. Springer, New York.
34 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Appendix 2.1 Location and year of detection for long-nosed potoroos at LGNP-south and north. Location where species recorded shown by
coloured dots, records over multiple years shown as separate colours. Black dots – hairtube stations where species not recorded.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 35
Appendix 2.2 Location and year of detection for long-nosed potoroos at Cobboboonee and Mt Clay. Location where species recorded shown by
coloured dots, records over multiple years shown as separate colours. Black dots – hairtube stations where species not recorded.
36 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Appendix 2.3 Location and year of detection for long-nosed potoroos at Hotspur and Annya. Location where species recorded shown by coloured
dots, records over multiple years shown as separate colours. Black dots – hairtube stations where species not recorded.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 37
Appendix 3.1 Location and year of detection for southern brown bandicoots at LGNP-south and north. Location where species recorded shown by
coloured dots, records over multiple years shown as separate colours. Black dots – hairtube stations where species not recorded.
38 Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control
Appendix 3.2 Location and year of detection for southern brown bandicoots at Cobboboonee and Mt Clay. Location where species recorded shown
by coloured dots, records over multiple years shown as separate colours. Black dots – hairtube stations where species not recorded.
Glenelg Ark 2005–2013: Evidence of the Benefits for Native Mammals of Sustained Fox Control 39
Appendix 3.3 Location and year of detection for southern brown bandicoots at Hotspur and Annya. Location where species recorded shown by
coloured dots, records over multiple years shown as separate colours. Black dots – hairtube stations where species not recorded.
lenelg Ark 2005–2012
rthur Rylah Institute for Environmental Research Technical Report Series No. xxx 1
ResearchGate has not been able to resolve any citations for this publication.
Book
Occupancy Estimation and Modeling: Inferring Patterns and Dynamics of Species Occurrence, Second Edition, provides a synthesis of model-based approaches for analyzing presence-absence data, allowing for imperfect detection. Beginning from the relatively simple case of estimating the proportion of area or sampling units occupied at the time of surveying, the authors describe a wide variety of extensions that have been developed since the early 2000s. This provides an improved insight about species and community ecology, including, detection heterogeneity; correlated detections; spatial autocorrelation; multiple states or classes of occupancy; changes in occupancy over time; species co-occurrence; community-level modeling, and more. Occupancy Estimation and Modeling: Inferring Patterns and Dynamics of Species Occurrence, Second Edition has been greatly expanded and detail is provided regarding the estimation methods and examples of their application are given. Important study design recommendations are also covered to give a well rounded view of modeling.
Article
Most populations of southern brown bandicoot breed seasonally, with births occurring between July-December in Victoria, and July-February in Tasmania. Onset and cessation of breeding are highly predictable in S Victoria, but less so in Tasmania. Litter size varies from one to six. Gestation period is <15 days and lactation c60 days. Females produce up to three litters per breeding season in Victoria, and 4 litters per breeding season in Tasmania. Litter size and the numbers of litters produced per year appear to be related to food abundance. Dispersal and mortality in Victoria heathlands are highest among juveniles and young adults of both sexes. Maximum longevity of wild animals is 3.5-4 yr. Females at Cranbourne first breed in the breeding season following birth, whereas in Tasmania females may breed in the season of birth. Tasmanian animals grow faster and weigh more than Victoria animals as adults. Tactical responses of I. obesulus include facultative variations in length of the breeding season, number of litters produced by females within the breeding season and in litter size. -from Authors