Pest animal research

Anti-predator behaviour in livestock

Research officer: Lee Allen

Experienced adult cattle are generally unaffected from wild dog attacks but death of, or bites to, calves and weaners can be a significant source of economic loss, although notably, impacts vary greatly between years and properties. In northern Australia, calf loss between pregnancy diagnosis and weaning is twice that of other regions; most calf loss occurs within 10-14 days of birth and 40-50% of these are attributed to “unknown causes”. A joint DAF/UQ ‘CashCow’ study indicated a mean 5% greater calf loss on properties whose owners ‘believed’ wild dogs negatively impacted their businesses. There is ~10% greater calf loss in 1st calf cows compared to mature cows, substantially greater calf loss in north Queensland compared to southern and south-western Queensland yet no difference in calf loss between owners that bait wild dogs several times per year, bait annually, infrequently or did nothing at all. The CashCow study prompted further UQ/QAAFI (Queensland Alliance for Agriculture and Food Innovation) research into calf loss, husbandry practices and how management interventions might reduce losses. This project, initially proposed within a larger “Calf Wastage” QAAFI proposal, supports those studies by examining how cow anti-predator behaviours develop in calves, weaners and 1st calf mothers, whether defence behaviours can be learned from observing experienced associates (i.e. social learning), and whether experienced ‘coacher cows’, pastured with younger cattle, train and/or protect these younger animals.

Cluster fencing monitoring and evaluation

Research officers: Lee Allen and Peter Elsworth

In 2013 South-West Natural Resource Management contracted graziers to erect several exclusion fences around ‘clusters’ of properties to facilitate the removal of wild dogs and control of kangaroos and other pests inside the fenced area by denying immigration. With public funding support, the construction of additional exclusion/cluster fences in Queensland has exceeded all expectations. To evaluate the impacts and longer-term benefits of cluster fences on livestock production, land condition, regional economies and biodiversity, a project was developed to monitor wildlife and vegetation trends within cluster fences. Initially intended as a 5-year study, dry seasonal conditions, ongoing Commonwealth and State investment and funding from the Centre for Invasive Species Solutions (CISS), this project will be extended further to capture the expected longer-term impacts.

This project monitors the abundance of kangaroos, wild dogs and other wildlife and pasture biomass and condition before and after the erection of cluster fences to quantitatively evaluate the cluster fence strategy. Our monitoring contrasts pest abundance and pasture condition on individual properties within the cluster with that of properties immediately outside the cluster fence. Ultimately, the success of cluster fencing will be determined by the extent to which livestock production improves (in addition to other indirect economic and social benefits) relative to livestock production in comparable areas outside the cluster, less the cost of establishing and maintaining the cluster fence and reduced pest populations. To that end and with support from CISS, this project will (1) add to the work at two intensive sites by undertaking comprehensive economic and productivity assessments of cluster fencing,(2) use regional modelling and remote-sensing to more broadly evaluate the effects of cluster fences and (3) determines the initial and likely future impact of fencing on some key environmental indicators..

Non-target impacts of 1080 pig baits

Research officers: Matthew Gentle and Peter Elsworth

This project examines two feral pig 1080 baiting practices - application of meat baits in the absence of pre-feeding or bait-stations, and the use of baits prepared from fruit and vegetable materials - that have a long history of use in Queensland to protect agriculture and the environment. The Australian Pesticides and Veterinary Medicines Authority (APVMA) has initially rejected the inclusion of these methods in the future registration of the Queensland 1080 Concentrate, given the limited assessments available on non-target species impacts. However, they have agreed to permit continued legacy use while studies are undertaken to collect and collate relevant data. We have reviewed the literature and other available data to define the issue and the appropriate research questions. We have then designed and completed six field studies in areas of Queensland where baiting for feral pigs with various bait substrates containing 1080 is common: Ingham (banana and mango substrates, in horticulture and canefields), Gore (meat substrate in sheep production enterprises), Moonie (meat substrate on conservation lands), Greenvale (meat substrate in cattle enterprises), Hebel and Hungerford (meat substrate in conservation and cattle production areas). Data collected includes efficacy data on feral pig populations, the identity of non-target species visiting and consuming bait material, population counts of theoretically-susceptible birds, and activity of varanids pre and post baiting. Control sites were used to ensure treatment (baiting) effects (if any) can be identified. A considerable amount of field data has been collected, but data collation and analysis is time consuming. Project staff have prioritised approaches and will complete study in 19-20 financial year.

Feral cat ecology and management

Research officers: Matthew Gentle, James Speed, Bronwyn Fancourt and Peter Elsworth

Feral cat populations are notoriously difficult to control. Intensive control to benefit threatened and other native species is often labour-intensive and costly. There is also often limited monitoring of the efficacy and/or benefits of control programs. Broadscale control options are limited. A field trial conducted during this project suggests the registered bait for feral cats in Qld (6mg 1080, 125g fresh meat bait) deployed under approved usage patterns (e.g ground baiting, trails/tracks) is problematic given low encounter rates and bait availability and palatability issues (Fancourt et al. 2016; and unpublished data). In recent times, a chipolata-style sausage containing either 1080 or PAPP toxin, and known as Eradicat or Curiosity bait (respectively) has been developed. These baits have shown some success for the broadscale control (ie via aerial baiting) of feral cats in other jurisdictions. Alternatively, other means of presenting poison to cats are being developed. An example is a device that sprays a toxin onto the coat of the cat, which then ingests the toxin during normal grooming behaviour (Read et al. 2014). While this product remains in development, it could be tested once commercially available. The use of detector dogs to detect/capture feral cats, and the cost-effectiveness of trapping or shooting could also be assessed. An assessment of the efficacy, specificity and feasibility (e.g. cost) of a range of possible techniques to control cats is needed and specifically in Queensland environs. By necessity, it may also incorporate refinement of methods to monitor cat populations, to ensure that control techniques can be adequately assessed. Although reducing feral cat abundance is important, the focus of any pest control campaign should be on reducing damage, not numbers killed. Thus, there also needs to be a better understanding of the predator-prey dynamics, requiring research beyond control efficacy. This project aims to determine the efficacy of feral cat broadscale control options, ultimately determining the benefits to native species from cat control, and recommend long-term management strategies for feral cats.

Monitoring the efficacy of new rabbit biocontrol

Research officer: Peter Elsworth

The national RHDV1-K5 program rolled out the release of this additional bio-control virus for rabbit management in March 2017. Monitoring following the release showed generally good declines of rabbit populations at the release site, but not beyond those sites from RHDV1-K5. The intensive monitoring undertaken did show that the RHDV2 virus that arrived in Australia in early 2015 had spread extensively and was reducing rabbit numbers across Australia. Additionally, the release of RHDV1-K5 encouraged some landholders and NRM groups to undertake warren ripping and harbour control to further reduce their rabbit impacts. To further our understanding of how the bio-control viruses interact with other and rabbits, and the importance of harbour destruction in long-term control of rabbits, on-going monitoring at the Wallangarra site, and additional monitoring of ripping programs around Highfields and other sites as available, will be undertaken. Queensland currently has confirmed outbreaks of five lethal viruses (RHDV-CAPM351, RHDV1-K5, RHDVa-chinese, RHDV2, and Myxo) as well as the non-pathogenic RCV-A1. Understanding how these viruses impact rabbits will improve rabbit management.

Ecology and management of chital deer in north Queensland

Research officers: Tony Pople, Matt Amos and Michael Brennan

There are two components seeking to improve the management of wild deer in Queensland.

  1. Determine the ecology, impacts and capacity for increase and spread of chital deer (Axis axis) in north Queensland to develop long-term management strategies. Limiting factors are likely to be a combination of wild dog predation and food supply, particularly availability of water and high-quality food during the dry season. Basic ecological data (diet, reproduction, population dynamics, movements) have been collected. Future work assesses habitat use, regional distribution, disease risk, fawn survival and causes of mortality.
  2. Develop two cost-effective tools for controlling wild deer in Australia: aerial shooting and ground shooting. The tools will be tested on three deer species in four states: fallow deer on agricultural properties in New South Wales and in native habitats in Tasmania, chital deer on pastoral properties in the Charters Towers area of Queensland, and sambar deer in the montane forests of Victoria. The effectiveness of the tools will be evaluated on cost, reduction in deer abundance and humaneness.

Abundance estimation of wild deer using drones and machine learning

Research officers: Tony Pople, Matt Amos and Michael Brennan

Deer present a current and growing problem in peri-urban and rural areas. Land managers undertake control, but struggle to monitor its effectiveness. Current monitoring methods are typically limited, particularly in the peri-urban environment, with problems of high costs and high effort to collect data that are often low in spatial and temporal resolution. Deer can be detected with airborne thermal imagery, but detection probability is often unknown though essential for accurate monitoring. Detection probability can be improved with machine learning rather than relying on human assessment of imagery.

There is now an opportunity to combine advanced sensor technology with state of the art analytical methods in machine learning and data fusion to create an accurate, practical and cost-efficient monitoring method for wild deer in complex landscapes. Queensland University of Technology (QUT) have achieved excellent results for cryptic canopy species such as koalas and so much of the necessary basic research has already been conducted and proven. Research is needed however to establish the efficacy of drones and thermal imaging to determine deer abundance as part of a management program. Abundance will be estimated at selected sites with a drone survey and compare that with conventional ground surveys.

Detecting, delimiting and eradicating small, isolated deer populations

Research officers: Tony Pople, Matt Amos and Michael Brennan

There are isolated populations throughout Australia of the five problematic species in peri-urban and rural settings. Management authorities and landholders tend to adopt reactive strategies such as shooting on crops and trapping in residential areas without considering the overall strategy (control, containment, eradication). Current Centre for Invasive Species Solutions (CISS) projects assess control (impact reduction) techniques and strategies, but do not address eradication that requires a completely different set (removal only, close monitoring, time limited) with relatively high benefit:cost in the medium-long term. Eradication may well be a feasible and cost-effective option for some populations. However, the tools and strategies need assessment.

One difficulty is the detection and delimitation of small populations which are necessary for any eradication attempt. There has been some success in using remote cameras to monitor deer at low densities. A range of other techniques and strategies can be employed, such as aerial surveys, ground surveys for pellets and other sign and water point monitoring during the dry season. While local eradication is attractive, it requires high rates of removal, all individuals to be vulnerable to removal and no immigration. The probability of eradication given available resources needs to be determined.

Management of peri urban wild dogs and deer

Research officers: Matt Gentle, James Speed, Tony Pople, Matt Amos, Michael Brennan and Lana Harriott

Small landholdings, varied land use, high human density and media exposure create distinct difficulties for peri-urban pest managers. Peri-urban local governments have identified the need for tools and strategies for control of wild dogs and deer. Managers have had some initial successes in controlling wild dogs, but these need to be extended and tailored to other areas. Broad-scale baiting in peri-urban areas is rarely an option for wild dogs because of the number of properties and regulations for using poisons. Additionally, public sentiment and reaction to the traditional toxin (1080) can stifle or at worst, even halt the usage of 1080 within local government areas. This problem may be somewhat alleviated with recent availability of a relatively humane toxin PAPP with an antidote, but it will need field assessment before widespread acceptance and best application methods are known. New tools such as canid ejectors and lethal trap devices will need similar assessment as they are introduced into peri-urban pest management. Recent IA CRC work has also shown that there are self-sustaining peri-urban populations of wild dogs that are not simply spill-over from surrounding bushland. This suggests management at ‘point-of-impact’ (e.g. using ejectors/traps) may be more effective than pursuing broad-scale population suppression.

In contrast, for deer, a more fragmented distribution suggests management of source populations may be more fruitful. Deer have only modest rates of increase, so suppression of small populations is theoretically feasible. The difficulty has been removing sufficient animals. Control tools include various traps, baiting devices and shooting but efficacy is not well known and some methods are untenable in built-up areas. Monitoring methods are well developed overseas, but need broader application and refinement in Australia to guide and assess management action.

Particularly for deer but increasingly for wild dogs, public opposition to culling programs has thwarted control effort so community engagement is needed to determine and guide appropriate management strategies. Behavioural science and engagement approaches are required to successfully identity, implement and monitor the success of strategies to manage pest animals in the landscape. Initially, workshops will be conducted prior to implementing management interventions to gauge community attitudes to wild dogs and deer issues, and guide the type, level and involvement in interventions required to reach ‘acceptable levels of pest impact’. Findings from these workshops will guide the interventions conducted. Similar workshops will be completed during and post-intervention to determine the success of the management intervention as judged by the community. This would mesh with more traditional assessments of control strategies that focus on the monitoring of impacts (e.g. incident records) and activity (e.g. sightings, activity) as metrics of success.