Department of the Environment

About us | Contact us | Publications

Header imagesHeader imagesHeader images

Publications archive - Biodiversity

Disclaimer

Key departmental publications, e.g. annual reports, budget papers and program guidelines are available in our online archive.

Much of the material listed on these archived web pages has been superseded, or served a particular purpose at a particular time. It may contain references to activities or policies that have no current application. Many archived documents may link to web pages that have moved or no longer exist, or may refer to other documents that are no longer available.

Review of existing Red Fox, Wild Dog, Feral Cat, Feral Rabbit, Feral Pig, and Feral Goat control in Australia. I. Audit

Ben Reddiex, David M. Forsyth, Eve McDonald-Madden, Luke D. Einoder, Peter A. Griffioen, Ryan R. Chick, and Alan J. Robley.
Department of the Environment and Heritage, 2004

1. Introduction

The Department of the Environment and Heritage (DEH) commissioned the Arthur Rylah Institute for Environmental Research (Department of Sustainability and Environment, Victoria) to undertake a project to increase understanding of the threats to native species and ecological communities from foxes, wild dogs, feral cats, feral rabbits, feral pigs, and feral goats. The key aims of the project were to investigate; 1) control activities currently undertaken across Australia for from foxes, wild dogs, feral cats, feral rabbits, feral pigs, and feral goats; and 2) pest control that is necessary to secure the recovery of affected native species and ecological communities, especially those listed as threatened (under the Environmental Protection and Biodiversity Conservation Act 1999 (EPBC Act)). The project is being completed in three stages. This report is the first stage, and describes an audit of existing pest animal control activities in Australia. The next stages include identification of gaps in information on control activities and recommendations for filling these gaps (see Reddiex and Forsyth 2004), and designing a process to determine priority ranking for control of pest animals in order to minimise threats to native species and ecological communities.

2. Background

Red foxes (Vulpes vulpes), wild dogs (Canis lupus familiaris, Canis lupus dingo, and hybrids), feral cats (Felix catus), feral rabbits (Oryctolagus cuniculus), feral pigs (Sus scrofa), and feral goats (Capra hircus) separately and in various combinations are believed to be responsible for the extinction or decline of a wide range of native species and for adverse changes in ecological communities in Australia. Predation by foxes and feral cats are key threatening process under the EPBC Act, whilst competition with native species and land degradation by feral rabbits, feral pigs and feral goats are also listed as key threatening processes under that Act. Some of these species also have important impacts on agricultural values (through competition for resources and depredation of livestock) and may impact on historic cultural heritage and act as vectors of animal and human diseases (Braysher 1993).

The belief that pest animals have caused these declines in native species (and damaged production values) is reflected in legislation and has led to many attempts to control these pests. Many agencies and organisations including Federal, State and Local governments commit significant resources managing these species. However, there is limited hard evidence that this management has led to a reduction in threats and to a reversal in the decline (e.g., Hone 1994; Dickman 1996). Benefits of pest control are likely to depend on a wide range of factors, including the intensity and frequency of pest control, pest abundance following control, the size of an area controlled, and the ability of impacted species or resources to recover (Hone 1994; Choquenot and Parkes 2000; Coomes et al. 2003).

To our knowledge, this is the first national audit of pest animal control operations in Australia. The distributions and abundances of some pest animals have been reviewed for some states of Australia (West and Saunders 2003), but there have been no detailed reviews of the characteristics of existing pest animal control operations. This review reports on pest animal control information collected in interviews conducted across all states and territories of Australia. Since the key focus of this review was to increase understanding of the threats by pest animals on native species and ecological communities, an emphasis was placed on collecting data from 'conservation' focused rather than 'agricultural' focused control activities. Monitoring of the impact of pest animal control is not absent but less likely to be undertaken by private landholders in the agricultural sector.

3. Objectives

The objectives of this study were:

  1. To conduct a comprehensive audit and evaluation, of all available information on existing control activities/programs, for foxes, wild dogs, feral cats, feral rabbits, feral pigs, and feral goats, across Australia.
  2. To review all available information on control activities/programs that include measures of the outcomes of differing levels/types of control activities, in terms of reductions in pest species and recovery of native species/ecological communities, especially those listed under the EPBC Act, and those known or perceived to be threatened by foxes, wild dogs, feral cats, feral rabbits, feral pigs, and feral goats.

4. Study species

Foxes (Vulpes vulpes)

The red fox was deliberately introduced into Australia in the mid to late 1800's. Foxes are now common throughout most of Australia, except the tropical north and some offshore islands (Figure 4.1). Foxes occupy many habitats, including urban, alpine and arid areas, but are most common in woodland and semi-open habitats (Saunders et al. 1995). Foxes have been shown to eat a wide range of native species (reviewed in Robley et al. 2004) and are thought to have played a major role in the decline of many ground-nesting birds, small to medium sized mammals, and reptiles (see Table 7.4; Pg 48, for a list of native species for which foxes have been identified as a known or potential threat).

Wild dogs (Canis lupus familiaris, Canis lupus dingo, and hybrids)

Dingoes were introduced to Australia from Asia about 3,500 years ago. Domestic dogs arrived in Australia with European settlers in the late 1700's. The range of dingoes has reduced by approximately 60%, but wild dogs and hybrids are now widely distributed throughout the dingo's former range (Fleming et al. 2001). Dogs are present across most of Australia except for the sheep and wheat growing areas of southeastern Australia (Figure 4.1). Wild dogs are primarily carnivores, and have been shown to eat a wide range of native species (reviewed in Robley et al. 2004). The impacts associated with wild dogs are complex as they are likely to impact endangered wildlife (see Table 7.4; Pg 48, for a list of native species for which wild dogs are considered as a threat), and predation by wild dogs is a major concern for agriculture. However, dingoes may play an important functional role in natural ecosystems, and are protected as a native species under legislation in some States and Territories (Fleming et al. 2001).

Feral cats (Felis catus)

Cats probably became established in Australia soon after the arrival of the first Europeans. Feral populations now occupy most parts of the mainland, Tasmania and many offshore islands (Figure 4.1). Cats eat a wide range of native wildlife (reviewed in Robley et al. 2004), and for this reason are thought to have a major impact on many native species, especially on islands (see Table 7.4; Pg 48, for a list of native species for which feral cats have been identified as a known or potential threat).

Feral rabbits (Oryctolagus cuniculus)

The feral rabbit is one of the most widely distributed and abundant mammals in Australia (Williams et al. 1995). Rabbits were first released in 1859 in Geelong, Victoria, and spread rapidly to cover most of Australia, except the far north, by 1910 (Figure 4.1). Feral rabbits occur in many habitats, but are sparsely distributed in the arid north and are most abundant in areas with deep and sandy soils. They are predominantly grazers and are thought to compete with native wildlife for resources. They may also alter the distribution and abundance of native plant species and physically alter habitats (Williams et al. 1995). Feral rabbits have been implicated in the extinction of a number of small mammals in Australia's arid regions, and may have contributed to the decline in numbers of many native plant and animal species (see Table 7.4; Pg 48, for a list of native species for which feral rabbits have been identified as a known or potential threat) (Williams et al. 1995).

Feral pigs (Sus scrofa)

Domestic pigs were introduced to Australia by European settlers, and populations of feral pigs were widespread by the 1880s. Feral pigs are now common in the Northern Territory, Queensland, Australian Capital Territory and New South Wales, and less common in western Victoria, Western Australia, and on a few offshore islands (Figure 4.1). Feral pigs are omnivorous habitat generalists, occupying subalpine grasslands, woodlands, tropical forests and, semi-arid and monsoonal floodplains. The primary environmental impacts of feral pigs are habitat degradation and predation of native species. By wallowing and rooting feral pigs modify streamsides, increase erosion, and decrease food resources and habitat for native wildlife (Choquenot et al. 1996). Feral pigs are also thought to compete with native animals for food, eat the eggs of ground-nesting species, and spread environmental weeds. Feral pigs have destroyed breeding sites and degraded key habitats for a number of species (see Table 7.4; Pg 48; Choquenot et al. 1996).

Feral goats (Capra hircus)

Feral populations of goats established in Australia from the escape, abandonment, or deliberate release of domestic goats (Parkes et al. 1996). Feral goats live in all States and Territories and on many offshore islands, but are most common in areas of western New South Wales, South Australia, Western Australia, and Queensland (Figure 4.1). The diet of feral goats includes grasses, leaves, bark, flowers, fruit, and the roots of many plant species (Parkes et al. 1996). Feral goats are thought to have major effects on native vegetation, and may also compete with native wildlife and stock for food, water and shelter (see Table 7.4; Pg 48, for a list of native species for which feral goats have been identified as a known or potential threat).

Figure 4.1. Distribution of foxes, wild dogs, feral cats, rabbits, feral pigs, and feral goats in 2001, in Australia (electronic distributions are from subregional or bioregional scale data from the Natural Land and Water Resources Audit, Landscape Health in Australia database, 2001).

Figure 4.1. Distribution of foxes, wild dogs, feral cats, rabbits, feral pigs, and feral goats in 2001, in Australia (electronic distributions are from subregional or bioregional scale data from the Natural Land and Water Resources Audit, Landscape Health in Australia database, 2001).

5. Gaining reliable knowledge about pest animal control

The key objective of this report is to assess the relationships between (i) levels of pest animal control and the change in pest abundance, and (ii) the relationship between the level of reduction in pest animal abundance and the subsequent abundance or change in condition of conservation resources. Hence, we seek to establish predictable responses (relationships) between pest animal control actions and their outcomes. This report does not investigate whether pest species are the agents of decline to native species and ecological communities.

The strongest inferences (or most reliable knowledge) in science come from experiments (Platt 1964). Experiments enable us to make the most reliable inferences about the effects of manipulating a system. Experimentation has four basic tenets: (i) treatments, (ii) non-treatments, (iii) replication (of treatments and non-treatments), and (iv) random assignment of treatments and non-treatments. Given that a control program is a manipulation of a system, the concept of reliable knowledge is relevant to the aims of this report.

A pest animal control program can be considered as a 'treatment' (Macnab 1983); the abundance of pest animals within a defined area is manipulated (hopefully reduced) with the aim of achieving a stated objective (see below). If the objective is protection of native species and/or habitats (termed 'resource'; the focus of this report), then managers can only know if the objective has been achieved if they monitor that resource. Hence, the weakest possible inference in pest animal control comes from the situation where there is one treatment area and only the resource is monitored (Table 5.1). If the monitoring showed an increase in the abundance of the resource then it would be tempting to conclude that it was a result of the pest control action. However, such a conclusion is unreliable because it is not known if the control actually reduced the abundance of the pest. Hence, a more reliable inference can be made if changes in the abundance of the pest were also monitored. But, even if monitoring revealed that the abundance of the pest declined and the abundance of the resource increased following control, one or both changes may be a consequence of something other than the control (e.g., changes in food availability). The only way to establish whether these changes were the result of the control is to compare these changes to a non-treatment area.

The incorporation of non-treatment areas, one of the four tenets of experimental design, thus leads to a substantial increase in the reliability of inferences (Table 5.1). Even if a treatment and a non-treatment area are monitored, site differences (e.g., in the availability of potentially suitable habitat for a threatened species) can still mean that inferences about the effects of pest animal control are false. This problem can be overcome by the random assignment of the treatment and non-treatment sites, and by replicating the number of treatment and treatment sites. Although both randomisation and replication are desirable, there is a consensus that replication is more important than randomisation (e.g., Walters and Green 1997; Johnson 2002). Randomisation involves randomly assigning (e.g., by a fair coin) a treatment or non-treatment to each site. We acknowledge that random allocation of treatment areas is not always possible. Replication involves increasing the number of independent treatment and non-treatment areas and can be achieved by partitioning a single contiguous study area into smaller study areas, or by using multiple study areas separated in space. A key component of replication is that the control must be applied to each treatment area independently (see Hurlbert 1984). When there are 2 treatment and non-treatment areas, the effects of the treatment are averaged relative to the non-treatment areas, with the confidence in the estimate of the effect of pest control increasing with the number of treatment and non-treatment areas. Hence, the strongest possible inferences in pest animal control come from a program which has multiple treatment and non-treatment areas, random assignment of those treatment and non-treatment areas, and where both the resource and pest are monitored (Table 5.1).

Table 5.1. The relationship between reliability of inferences and the design of pest animal control programs.
Reliability inferences of
Design
Monitoring
  Treatment areas (number) Non-treatment areas (number) Resource Pest
Least reliable 1 0 Yes No
arrow 1 0 Yes Yes
1 1 Yes Yes
1, random 1, random Yes Yes
≥2 ≥2 Yes Yes
Most reliable ≥, random ≥, random Yes Yes

Monitoring can also influence the reliability of inferences about pest animal control. Pre-control monitoring is an essential component of estimating the change in the abundance of a pest in a control operation (e.g., Veltman and Pinder 2001). Pre-control monitoring may also be useful for understanding differences between sites in the abundance of pests and resources (or other factors that may be important, such as food availability); such information can be used to improve the design of the control program (e.g., by adopting a matched-pair design). Post-control monitoring is required to evaluate the response of the pest and native species to pest animal control.

The above discussion is important because the strength of the inferences about the benefits of pest animal control for resources varies depending upon the design of the program and it's monitoring. We therefore address the issues of experimental design and monitoring when attempting to evaluate the benefits of pest animal control for native species and habitats.