Table : Comparison of alternative control measures

Based on available assumptions it is not possible to estimate the impact of these options on the level of dog attack. There is insufficient data and evidence of cause and effect to conclude a relationship between the level of dog take under current and future control measures and levels of dog attack. Five key variables are unknown:

the current size and growth of the wild dog population in Victoria;

bait take by target and non-target species;

the relationship between the intensity of controls within the buffer zone and private land and the incidence of new attacks by wild dogs;

the effectiveness of trapping and baiting to remove existing ‘problem dogs’, i.e. wild dogs known to have been responsible for previous attacks; or dogs that have come through the baiting containment lines and are likely to attack; and

the rate of re-colonization of areas where dogs have been removed.

Likely impact of scenarios

In Table the likely impact of the change in wild dog management arrangements is estimated. These impacts show that:

increased ground baiting will increase dog take, albeit at a higher cost than base case operations under 72, 48 and 24 hour trap inspections (Options A1, B1 and C1);

doubling the current frequency of aerial baiting from once to twice a year will increase dog take but is very expensive compared to alternatives (Options A2, B2 and C2). Aerial baiting is a controlled action under the EPBC Act. Careful site selection is used to address risks associated with potential off target take;

replacing ground baits with CPEs produces similar wild dog take and costs (e.g. base case 72 hours compared to Option A3). However, it does reduce non-target take (e.g. native animals), noting that several key threatened species may be at increased risk;);

replacing trapping with CPEs increases dog take at a higher cost than base case operations under 72, 48 and 24 hour trap inspections. (Options A4, B4 and C5). However, it does reduce the spectrum of non-target species (e.g. native animals) that could be impacted – noting that while dingoes are attracted to the baits several other key threatened species are not and do not have the force to risk setting off the trigger mechanism; and

non-target take is the highest with options in which ground or aerial baiting is increased.

We also note that there are likely to be a range of changes to animal welfare outcomes as a result of any changes to the mix of control measures given that;

some control measures such as baiting and CPEs are considered to be more humane than trapping; and

shorter trapping inspection times are considered to be more humane.

Table : Estimate impacts of alternative control scenarios

7.3Optimal design mix issues

Caution is required interpreting our analysis in Table to draw conclusion on the most preferred control investment scenario.

Based on available data we are unable to estimate an optimal mix of approaches because:

there are significant uncertainties over the relative efficacy of instruments;

optimisation is dependent on the control strategy to be pursued;

key constraints on the application of control measures; and

some options are simply not practical and would not keep the optimise the deployment of the workforce.

These caveats relate to the public control measures in of themselves. More broadly the extent of private control activities such as fencing and stock management can ultimately influence overall efficacy.

7.3.1 Significant uncertainties over the efficacy of instruments

A significant limitation to estimating optimal control strategies is the significant uncertainty over the key performance variables that impact on the efficacy of control measures.

This includes uncertainties over the:

level of target and non-target species take by baiting and aerial baiting; and

the relative effectiveness of baiting and trapping to reduce dog attacks as opposed to dog take.

Level of target and non-target species take

Our analyses and conversations indicate that there is a degree of uncertainty over the level of dog take and non-target species take by baiting.

The risk of non-target species take of baits is reduced by the requirements on the program to bury baits and constraining the use of baiting in areas where there is a higher risk of non-target species take, in particular Spot tailed Quoll.

Aerial baiting

Previous research undertaken by the Arthur Rylah Institute (ARI) has found that there is no demonstrated relationship between the level of aerial baiting and the level of dog attack. However, it is noted that community and wild dog controller feedback provides the opposite view.

In our empirical analysis we have made the following assumptions about aerial baiting:

achieves a similar level of bait take per bait laid;

has a similar level non-target species take, noting that in practice, this is highly unlikely given the fact that the bait is not buried and non-target species have a greater risk of taking baits when compared to the use of buried baits; and

has a substantially higher cost per bait laid than ground baiting.

Remote ground baiting is therefore less costly, more target specific than aerial baiting and can achieve an equivalent level of wild dog take.

Relative effectiveness of trapping and baiting to reduce dog attacks

The program has generally assumed that increasing the level of dog take is the mechanism to achieve reductions in dog attacks.

There are two dimensions to reducing the level of dog attacks:

reducing the number of dogs that have attacked; and

reducing the number of dogs that could attack in the future.

Trapping and baiting can conceivably achieve both since both control measures achieve take within close proximity to where attacks have or could occur. However, they have different likelihoods of take for a given level of effort and have different strategic applications since:

some dogs are either less likely to take baits or are less likely to be trapped — there is no single blanket control measure;

trapping is primarily focussed on taking dogs that have attacked in order to stop ongoing attacks as soon as possible or remove dogs in close proximity to livestock that are likely to attack in the short term; and

baiting is primarily focussed on reducing the capacity of populations of dogs to produce new attacks by intercepting dogs before they reach private property.

Given neither control measure can be fully successful in achieving its primary focus, neither can be solely employed.

7.3.2Control strategy to be pursued

By definition the control measures are complimentary to one another. The question of issue for this assessment is whether there is an optimal mix of both control measures.

The answer to this, in part, depends on the control strategy to be employed namely either:

containment — undertake control to reduce the number of dogs reaching private property; or

sustained control — undertake control to a new defined acceptable level of attack.

Our conversations with various stakeholders indicate that in the short to medium term (1-3 years) reactive trapping underpins a containment strategy (preventing an increase in the level of dog attack) but in the longer term (beyond 3 years) needs to be supported by baiting in order to address the risk of growth in dog attacks from growth in the wild dog population. In the long run, substantially higher levels of baiting are required to achieve sustained control.

Important questions are whether:

How can sustained control best be achieved?

What does sustained control look like, how is it measured, how do we know we are achieving it, how do we know if we are being successful in protecting livestock?

What is the optimum balance of control on public and private land to achieve sustained dog control at a landscape scale over the longer term?

There is sufficient data to answer any of these questions at this point in time? It is likely that levels of control will vary across Wild Dog Management Zones.

It is not clear to us if this type of intelligence is overlaid within the Dogbytes database, thereby enabling detailed nuanced empirical assessment of control activities and outcomes.

There is some indication that the level of attacks has fallen in some regions with the increases in baiting that have occurred within the 3km buffer since 2013. It is not clear, however, given the short time frame of analysis whether further falls will occur or if the level of attack will stabilise.

It would be valuable to continue current baiting levels for a further 12 months to assess if the emerging trends in some areas can be sustained and to understand if trends emerge in areas where they are presently unclear.

It would also be valuable to estimate and monitor baiting intensities and attack levels and their relationships at localised areas of control.

7.3.3Factors related to the application of control measures

As noted in section 1 there are range of factors related to the application of some control measures and this presently influences the flexibility of program managers to alter the mix of control measures.

DELWP has a service delivery commitment. The level of response may range from the provision of advice to trapping i.e. landholders who experience wild dog incidents where stock is maimed, killed or harassed are provided advice on immediate actions to mitigate further losses, and if it is verified that stock losses are due to wild dog attacks an assessment of current controls undertaken in the area may determine whether any modifications are needed. The wild dog controller may implement additional control using reactive tools and techniques such as trapping, opportunistic shooting and baiting.

The most pressing issue is the requirement in the WDMZ workplans to respond individually to each attack incident and continue control activities in the area for extended periods.

These operating arrangements have the practical effect of concentrating most resources into reactive control and, in-turn with trap inspection requirements, results in most of the available labour effort being directed into reactive trapping.

This Term of Reference is addressed in Report 1.

In Report 1 we concluded there are opportunities to improve the governance arrangements of the Wild Dog Control Program.

We recommended that a new more broadly based advisory body be established. We also recommended that a new Terms of Reference and operating guidelines for the body be established.

In Report 1 we reviewed whole of government arrangements and concluded that on balance the current allocation of program responsibilities across the departments be maintained.

We also found that given current institutional structures, accountabilities and capacities, it would not be appropriate at this point to devolve all operations and governance of the program to regional or local community institutions.

Suggestions for improving the efficacy of the Wild Dog Control Program is constrained because of uncertain relationships between inputs, outputs and outcomes.

In our view there is an urgent need to invest in research and development to inform future program strategy design and operations.

There is an urgent need to undertake comprehensive research on the levels of target and non-target species take by traps, ground (and aerial) baiting and CPEs. We are concerned that there are wide variations in the estimated level of take of baiting in particular.

Another aspect of affecting the assessment of efficacy is the absence of empirical data on the efficacy of targeted trapping. Trapping of wild dogs is primarily undertaken to take problem dogs. Anecdotal evidence is that targeted trapping can be highly effective but the empirical evidence is that trapping is more resource intensive and requires substantial skills. Research should be undertaken to test this evidence.

There is limited understanding of the size of the wild dog population and the empirical relationship between population size, buffer zone size and characteristics, and the level of attack. Improving understanding of these relationships within the Dogbytes database would provide insight into the levels of investment required to reduce dog attacks.

At present there is no systematic method of assessing the incremental contribution of individual controls on the level of attack. In effect, de-facto experimentation is being undertaken on a broad scale without appropriate controlling for other influences. Piloting of alternative mixes and levels of control measures should be considered in targeted locations (to minimise the impact of neighbouring area control influences).

Further consultation with community is required to determine what might be a desirable level of dog take given likely outcomes on attacks, non-target species and program costs.

Fox, C. 2014a, The Status of the Virginia Cooperative Coyote Damage Control Program - Fiscal Year 2013, United States Department of Agriculture, Animal and Plant Health Inspection Service, Wildlife Services, 105 B Ponderosa Drive, Christiansburg, Virginia.

Fox, C. 2014b, Predators in Peril, The Federal Governments War on Wildlife, Project Coyote. http://www.projectcoyote.org/newsreleases/news_predatorinperil.html

Kowlton, F, Gese, E and Jeager M 1999, ‘Coyote depredation control : an interface between biology and management’, Journal of Range Management, September 52:398-412.

Louney, M., Houben, J., and Eggborn, P 1997, Development of the Virginia Cooperative Coyote Control Program to Protect Livestock, Digital Commons, University of Nebraska.

Robley, A., Woodford, L., Lee, P., Kingston, V., Peters, W., Klippell D., and Gormley, A. 2009, Assessing the effectiveness of ground-based baiting for the control of wild dogs, Arthur Rylah Institute for Environmental Research Technical Report Series No. 193, Department of Sustainability and Environment, Heidelberg, Victoria.