Researchers at James Cook University used advanced artificial intelligence applications to analyze more than 300,000 hours (about 36 years) of mammal calls across eastern Australia.
This data is expected to significantly improve wildlife monitoring and conservation efforts in the region.
The university’s research brief said the data covered species from far north Queensland to southern New South Wales.
The research team compared different methods of tracking the health of mammal populations, including passive acoustic monitoring, camera traps, and direct observation.
They noted that while the initial technique worked well for vocalizations in birds, it had not been extensively tested in mammals, which led to their study, published in the journal Methods in Ecology and Evolution.
Lead researcher and JCU Postdoctoral Researcher Dr Sebastian Hofer explained that Australia has one of the worst mammal extinction rates in the world, most of which are not found anywhere else. Although they play an important role in ecological health and biodiversity, their range has been severely restricted by human activities, making them vulnerable to environmental change.
TCD Recommendations » Upway Spotlight
💡At Upway, you can easily find up to 60% off on premium electric bike brands
Therefore, monitoring their movement and population health is important for conservation, but large-scale observation is not easy. Thankfully, the open source BirdNET deep learning model eliminates many of the challenges, allowing teams to access vast amounts of data with far less effort.
By inputting mammalian sounds into the system, such as the sounds of a male koala, the AI platform can detect similar sounds across all datasets.
“It worked very well, especially in long-term monitoring,” Dr. Hoefer said.
“We can now rely on acoustic monitoring and AI for vocal mammals, freeing up more time and funding to monitor non-singing mammals.”
Using the BirdNET system in conjunction with remote sensing technology will give the team new insights into when certain species are most active, helping them “plan targeted management and monitoring at that time of year.”
Some environmental advocates have warned of the potential negative effects of AI, including the use of large amounts of energy and water, the depletion of resources from hardware manufacturing, and higher energy bills for consumers.
However, AI is proving essential to wildlife conservation because it can analyze large datasets to monitor species, detect poachers, predict habitat loss, and inform land use planning.
For example, conservation teams in China are using AI to monitor the endangered Yangtze finless porpoise. Thanks to these and other restoration efforts, the species’ population has increased by more than 23 percent in recent years, the first increase in decades.
In the United States, an Upper Midwest-based drone company and an AI company have teamed up to create POLLi, a conservation software platform to identify milkweed for monarch butterfly conservation.
Monarch butterflies are important pollinators and support ecosystem health in many regions. Despite the impressive population recovery in both the West and East, the U.S. Fish and Wildlife Service has proposed listing it on the federal Endangered Species List under the Endangered Species Act at the end of 2024, emphasizing the need for continued conservation efforts.
The latest use of AI to track mammalian calls could have important implications for global research and species recovery. In the future, the researchers aim to create a “species-specific recognition device” to track the mammals they record during their studies.
“This will make it easier to detect and monitor mammals across Australia in the future,” Dr Hofer said.
Get TCD’s free newsletter for easy tips to save more, waste less, and make smarter choices. Plus, earn up to $5,000 towards clean upgrades with TCD’s exclusive Rewards Club.
