Assessment of the biodiversity, economic and productivity gains from exclusion fencing (QLD)


Substantial investments have been made into constructing pest-proof netting fences (‘cluster fences’) around multiple grazing properties in western Queensland. Effective control of many vertebrate pests (e.g. wild dogs, kangaroos, feral pigs and feral goats) is now possible across large areas, by denying immigration, offering widespread and substantial benefits to agriculture and the environment. Similar fences are proposed for more arid areas in southern rangelands of WA, but the optimal cluster size and likely benefits for particular land types and production systems are unknown. This project brings together and expands the scope of existing studies to inform future cluster fencing activities. 




  1. Determine the cost-effectiveness of cluster fencing in the short and long term through the reduction in predation by wild dogs and reduced competition from kangaroos. This requires an assessment of the effectiveness of pest control as done by landholders, improvements in pasture production and, ultimately, improvements to livestock production, all relative to unfenced areas. This will draw on results of the subprojects below. 
  2. Determine the gains in livestock production from reduced predation and kangaroo competition resulting from cluster fencing. This involves collation and analysis of data on, for example, livestock weight gain, wool production, reproductive output, survival and injuries. It will also require evaluation of the relative contributions of kangaroos and livestock to grazing pressure. 
  3. Determine the economic benefits of cluster fencing for individual landholders and the broader community. The costs of constructing and maintaining the fences need to be determined as do the costs of managing pests in unfenced areas. 
  4. Identify practices and arrangements that strengthen collaborative management of cluster fences. 
  5. Estimate broad-scale changes in land condition, including vegetation biomass and composition, from a combination of regional modelling of pasture productivity and remotely sensed vegetation cover (satellite imagery). This will enable comparisons between areas inside cluster fences and areas that are unfenced. Estimates will be calibrated with empirical data from fixed field sites that are regularly recorded. If successful on intensively monitored clusters, the technique can be used for other clusters and under a range and mix of domestic, feral and native herbivore numbers. 
  6. Estimate broad-scale changes in the abundance and diversity of mammalian and avian taxa both directly and through linkages between biodiversity and land condition. 

Project Leader

Dr Malcolm Kennedy
Project Team
  • Dr Lee Allen, QDAF 
  • Dr Peter Elsworth, QDAF 
  • Bob Karfs, QDAF 
  • John Carter, QDES 
  • Dr Malcolm Kennedy, WA DPIRD 
  • Dr Peter Fleming, NSW DPI 
  • Megan Star, Central Queensland University
Project Partners
  • Queensland Government Department of Agriculture and Fisheries, (QDAF)  
  • Queensland Government Department of Department of Environment and Science (QDES) 
  • New South Wales Department of Primary Industries (NSW DPI)
  • Western Australia Department of Primary Industries and Regional Development (WA DPIRD) 
  • Meat and Livestock Australia (MLA)
  • Central Queensland University (CQU)

The project receives funding from the Australian Government Department of Agriculture, Water and Environment


February 2019: 

In the Morven region, weather conditions have been very dry, resulting in poor pasture conditions. Pasture dry matter is being monitored inside and outside the Morven cluster. Prior to closure of the cluster fence, pasture biomass was higher outside, but that difference is less now. 

Since the Morven cluster was closed in 2015, the number of wild dogs captured inside the fenced area has declined rapidly. Contract trappers removed 16 wild dogs from within the Morven cluster during 2018. A total of 523 wild dogs have been trapped and destroyed since fence construction commenced. 

Wild dog activity across multiple properties in the Tambo cluster shows the expected annual breeding peaks in activity in autumn but there is an encouraging underlying trend of a gradual reduction over time inside the cluster. This suggests producers or sub-clusters are making incremental progress at removing this problem species. 

The estimated abundance of kangaroos inside and outside of the Morven cluster has declined substantially since August 2017. The pattern for kangaroo density at Tambo is similar to that found in Morven with consistently higher densities inside the clusters than outside. This is attributable to the combined impacts of drought, predation and less productive land being located outside clusters. 

August 2019 update:

Pasture, wild dog and wildlife monitoring, assessment of livestock productivity and economic analysis are currently underway for the Morven and Tambo cluster sites.

Wild dog activity inside and outside the Morven cluster is very low. Removal of wild dogs has occurred since 2014, with decreasing numbers trapped each year. The Tambo cluster has similar population trends, apart from April-August 2018, when activity outside the fence was high. Dry conditions have resulted in low wildlife activity at both sites.

Kangaroo density at the Morven cluster is much lower than at Tambo, higher activity inside the clusters at both sites is most likely due to more fertile land within the clusters. Cattle producers at the Morven cluster have utilised lower kangaroo numbers by stocking sheep and goats, while maintaining cattle numbers. Sheep production within the Morven cluster has declined due to below average rainfall. The Tambo cluster received higher rainfall, though cold weather resulted in losses of drought-affected sheep and a large reduction in lambing rates.

Remotely-sensed vegetation cover at a paddock scale is not well correlated with visual observations at the small plot scale. The reasons are being investigated. Bird species abundance and richness is being sampled. Potential positive and negative effects of the cluster fences may influence different species.

February 2020 update:

A post-doctorate position has been filled, allowing the economic analysis to commence and ensuring that all project components are now underway. Other components underway include pasture monitoring via satellite, monitoring of wild dogs, kangaroos and other wildlife, and assessment of livestock productivity.


Wild dog activity inside and outside the Morven is very low. Dry conditions have resulted in low wildlife activity at both sites. Kangaroo abundance in the Morven and Tambo clusters is higher than outside the clusters. Existing differences in habitat suitability for kangaroos inside and outside the two clusters are likely to account for the disparities.

The impacts of cluster fencing on diversity of granivorous birds has proved challenging to assess. Emus are being investigated as a potential species to monitor. Long-term, broad-scale aerial survey data for emus and kangaroos available will be analysed for trends in abundance in areas with and without cluster fences.


Within the Morven cluster, where the fence was completed in 2013, sheep numbers and lambing rates reached a peak in 2015, but then declined to pre-fence levels in line with successive years of below average rainfall. The Tambo properties have received more rain over the same period which enabled them to maintain sheep numbers and lambing rates.

The absence of sheep properties outside of the clusters has prevented comparison of sheep and wool production on properties within clusters with that of sheep properties which do not have exclusion fencing. Given this, monitoring of livestock productivity is being expanded to include beef cattle.


Field surveys carried out in the Morven and Tambo clusters have recorded the lowest biomass during the five years that has been collected. Remote sensing of vegetation cover is ongoing. Remotely sensed vegetation cover at a paddock scale is not well correlated with visual observations at the small plot scale. The differences can be attributed to remote sensing recording litter and woody debris, which does not break down in low rainfall conditions.

Subscribe to our newsletter