Research
Science for Wildlife creates and shares scientific knowledge to empower wildlife conservation. Our aim is to bridge the implementation gap between research and conservation, so that the best possible wildlife science and technology is applied on the ground, where it counts. To do this we work together with land managers and communities. We're working across a diversity of conservation challenges, including bushfire recovery for critical weight range mammals in NSW Australia, with support from Greater Sydney Local Land Services and Wettenhall Environment Trust.
The environmental issue
The Greater Blue Mountains World Heritage Area (GBMWHA) was significantly impacted by the 2019/2020 megafires, with the full extent of biodiversity loss unknown. In the two years following the fires, Science for Wildlife (S4W) found that across three separate sites within the GBMWHA, unburnt areas were significantly more important for critical weight range (CWR) mammals compared to burnt areas of any severity. Conversely, it was found that invasive mammal diversity and occurrence, including invasive predators, was highest in areas of ‘extreme burn’, and lowest in ‘unburnt’, highlighting the importance of unburnt vegetation as a refuge for CWR mammals during and post fire (initial project results).
Now 5+ years post fire, it is essential for fire management to understand how CWR mammals are utilising the recovering vegetation and when burnt habitats become available to them again. This is particularly important in light of the increasingly frequent and extreme fires expected under climate change, and the accompanying need to manage fire refugia. Unburnt patches can provide essential pyric refugia while species and the surrounding landscape recover from fire impacts. If areas are protected as unburnt refuges, then the question of how long to protect them for is a vital one to answer.
The project
Fifty camera traps were deployed across urban/peri-urban and protected areas within the Hawkesbury LGA in October 2025 and will remain in situ until February 2026 to allow for a minimum of 12 months of data collection over two summer seasons. Survey locations were stratified by fire severity history and included four categories: unburnt, low, moderate, and high/extreme. Prior to uploading these images onto Zooniverse for classification, we utilized Artificial Intelligence to prescreen the images to remove blank images as well as images with vehicles and people.
Bait stations are a common tool used in small mammal camera trap surveys to attract animals into the camera's field of view, increasing the likelihood of capturing clear, classifiable images. In this project, bait stations consist of a simple mixture of peanut butter and oats — a combination that is highly attractive to a wide range of small mammals due to its strong scent and palatability. This bait is not accessible for animals to be able to ingest, however the scent can still escape the lure.
Each bait station is constructed from a PVC pipe lure mounted on a wooden/metal stake. The camera itself is mounted on a separate stake or tree and set 2.5 metres away from the lure, ensuring the entire bait station and the area around it is within the camera's field of view.
When classifying images, you may notice animals investigating or feeding directly at the PVC pipe station — this is entirely expected and is a deliberate part of the survey design. The use of bait does not harm the animals and is a standard, ethically approved method in wildlife monitoring across Australia.
Project outcomes
Outcomes of this project include an increased resilience of CWR mammals to fire events through the uptake of results into fire management plans, for improved fire management under climate change. Results of this project will strengthen data on the rates of wildlife and habitat recovery 5+ years after the 2019/2020 megafires, as well as changes in invasive mammal distribution and occurrence, and their drivers. Gaining a better understanding of fire impacts, immediately after and in the mid to long-term, supports more effective management.
Additionally we aim to increase community knowledge and capacity for wildlife conservation. The project involves different community audiences. Camera trap deployment and data collection will incorporate citizen science at two levels: I) Community-based local volunteers are involved in camera trap deployment and maintenance, via training and supervised participation; II) Analysis of camera trap images involves a much larger online audience, combining citizen science and artificial intelligence (AI) to process image data via the online platforms Wildlife Insights and Zooniverse. This will upskill online volunteers in identification of native and invasive species, grow their knowledge of fire impacts, and raise awareness of key conservation issues and solutions.
