Issue: August 2026

At the Arthur Rylah Institute (ARI), our terrestrial ecology teams deliver science to inform evidence-based decisions by governments and communities.

With over 50 skilled scientists, we bring together extensive expertise in fauna and flora research, ecological modelling and data interpretation. We work closely with national, state and local agencies, universities and community partners to support the management of Victoria’s biodiversity.

Semi-arid woodland in eastern Murray-Sunset. Photo by ARI.

Semi-arid woodlands were once widespread across the Victorian Mallee but they are now a threatened ecosystem due to land clearing and grazing pressure. Since 2012, ARI has partnered with Parks Victoria to monitor how well grazing management actions help these slow-growing woodlands recover in north-west Victoria.  

Semi-arid woodlands are characterised by tree species such as Buloke (Allocasuarina luehmannii), Belah (Casuarina pauper), Slender Cypress Pine (Callitris gracilis) and Sugarwood (Myoporum platycarpum). They also have a sparse shrub layer and a groundcover of herbs, saltbushes and grasses.   

Our research established 300 monitoring sites across key conservation areas, including Murray-Sunset, Hattah-Kulkyne and Wyperfeld National Parks, Lake Albacutya Park and several Flora and Fauna Reserves (Yarrara, Mallanbool, Dunstans, Dering and Turriff). The results of this long-term monitoring show that controlling browsing herbivore numbers is important and effective in supporting the recovery of semi-arid woodlands.  

In some areas with grazing controls, there has been an increase in young plants, large shrubs and overall native plant diversity. Continued grazing control, fire prevention and targeted replanting are needed to ensure these threatened woodlands survive into the future.   

This work is supported by Parks Victoria and DEECA. The initial stages of the monitoring program were also supported by funding from the Mallee Catchment Management Authority.  

Spotted-tailed Quoll (Dasyurus maculatus) captured on a camera trap. Photo by ARI.

Camera trapping is an efficient and cost-effective way to monitor wildlife distributions, abundance and behaviours. It involves setting up a digital camera in the field that automatically takes images when it detects a moving heat source, such as an animal passing by. One of the advantages of camera trapping is that these cameras can be left out on site for long periods to collect data.  

ARI has extensive experience using camera traps to study a wide range of animals, including deer, dingoes, foxes, Leadbeater’s possums, potoroos, bandicoots, spotted-tailed quolls, smoky mouse and dunnarts. Since 2012, ARI has deployed more than 12,500 cameras across different projects. These have involved infrared-flash cameras that capture black-and-white images at night for medium to large animals, and white-flash cameras that capture colour images at night for small mammals.   

Data collected from identifying animals in camera trap images can be combined with advanced modelling techniques to estimate wildlife abundance across large areas. These methods have been used to predict statewide population densities of species such as deer and dingoes.  

This information is important to support wildlife management and conservation by enhancing our understanding of native species populations, and the impacts of pest animals and wildlife responses to disturbances such as fires and floods. ARI’s camera trapping projects have also contributed species observations to the Victorian Biodiversity Atlas, which helps inform biodiversity management across Victoria.  

ARI provides support for camera trapping projects at all stages, from study design, deployment methods, to developing sophisticated analyses of camera data to answer wildlife management questions. Our fleet of high-quality camera traps, in-house image database and use of artificial intelligence (AI) tools to identify species are also helping to make the process more efficient and cost-effective. 

Territorial dispute between Red-necked Stints (Calidris ruficollis). Photo by Danny Rogers.

On the western shore of Victoria’s Port Phillip Bay, the Western Treatment Plant (WTP) is a wetland of international importance recognised under the Ramsar Convention. While the WTP treats around half of Melbourne’s wastewater, it also provides valuable habitat for tens of thousands of waterbirds. 

Since 2000, ARI has partnered with Melbourne Water to monitor waterbird populations at the WTP, creating one of Australia’s most significant long‑term waterbird datasets. Spanning more than 25 years, the program tracks the number of waterfowl (including ducks, swans, coots and grebes), shorebirds, large wading birds (such as, ibis, egrets, herons), and freshwater terns.  

The monitoring program has shown that the WTP hosts large numbers of freshwater terns and provides important feeding grounds for shorebirds and ibis. Over time, species have shown different population trends.

For example, numbers of Red-necked Stint and Sharp-tailed Sandpiper have remained stable at the WTP, although both species have declined in other parts of Australia. On the other hand, the Eastern Curlew is now thought to be locally extinct at the WTP and Straw-necked Ibis numbers have continued to decline.   

Beyond counting birds, ARI studies waterbird movements across habitats and investigates how future changes in climate and habitat conditions may affect their populations. This research has helped Melbourne Water manage and conserve the WTP’s wetland and help guide management decisions about sea‑level rise, nutrient management and pond operations to maintain waterbird populations at the WTP. 

Audiomoth deployed at McMillians Bushland Reserve. Photo by Lachlan Francis.

ARI has developed a world-leading artificial intelligence (AI) software called ARISA that can identify the calls of birds, frogs, bats and other mammals from sound recordings collected in the field. As more acoustic monitoring data is gathered, using AI makes it much faster and more efficient to process these recordings. 

The accuracy of ARISA depends on how the recordings are collected. Factors such as the sampling rate, device sensitivity and where the recorder is positioned in the field can affect how well animal calls are detected.  

For the best results, recordings should be made at a higher sampling rate, which gives the AI more information to work with. Recorders should also be placed carefully and directed away from noise sources. 

Feel free to forward this to others you feel may be interested.

Tracey Regan
Co-Science Manager
Community Ecology
Arthur Rylah Institute for Environmental Research

Jenny Nelson
Section Leader
Wildlife Ecology
Arthur Rylah Institute for Environmental Research

For general enquiries, please contact research.ari@deeca.vic.gov.au

Stay in touch with ARI's latest research with our updates including the ARI eNews, Quarterly Updates and ARI Seminars.