AI can monitor biodiversity through animal sounds

Researchers have demonstrated that artificial intelligence (AI) models can detect biodiversity from animal sounds in tropical rainforests, specifically in reforested areas. 

The goal of the research, led by the University of Würzburg, is to monitor the return of biodiversity in previously deforested tropical zones.

Monitoring biodiversity 

Tropical forests play an essential role in keeping our planet healthy. They provide habitats for unique wildlife,  play a role in the carbon cycle, and influence the climate on a global scale. Yet, these rainforests are under increasing pressure from relentless overexploitation.

“Tropical forests play a key role in the global carbon cycle and are central to Nature-based Climate Solutions, both in terms of climate adaptation and mitigation,” wrote the study authors. 

“They are also fundamental to global biodiversity conservation, harboring 62% of terrestrial vertebrate species. As such, restoring tropical forests is key to counteract two of the major crises of our times, biodiversity loss and climate change.”

Sound analysis

In tropical regions that are getting a much-needed boost from reforestation, it becomes crucial to monitor biodiversity. The researchers have recently shown that animal sounds provide a helpful monitoring tool 

“To be effective, all conservation measures require cost-efficient and robust biodiversity monitoring, which is lagging behind carbon monitoring due in part to the lack of scalable, reproducible and cost-effective sampling methodologies,” explained the study authors. 

Focus of the study

The researchers chose northern Ecuador as their study site, specifically focusing on abandoned pastures and former cocoa plantations undergoing forest reclamation. 

The goal of the study was to determine if autonomous sound recorders, combined with AI, can recognize species compositions of birds, amphibians, and mammals.

What the researchers learned 

“The research results show that the sound data reflect excellently the return of biodiversity in abandoned agricultural areas,” said Professor Jörg Müller, head of the Ecological Station Fabrikschleichach at Julius-Maximilians-Universität (JMU) Würzburg 

A notable finding was the alignment of vocalizing species communities with recovery gradients. A preliminary set of 70 AI bird models described entire species communities, even capturing shifts in nocturnal insects.

Future research 

The research team plans to refine these AI models. The goal is to record an even broader spectrum of species and establish these models in other protected zones, including the Sailershausen JMU Forest and Germany’s venerable national park in the Bavarian Forest.

“Our AI models can be the basis for a very universal tool for monitoring biodiversity in reforested areas,” said Müller. He suggests their applicability in biodiversity credits, which are comparable to carbon dioxide emissions trading. 

These credits, which offset adverse environmental impacts of activities, can be procured by companies or organizations.

Study implications 

“Our results demonstrate that automated bioacoustics monitoring can be used to track tropical forest recovery of animal communities from agricultural abandonment beyond vocalizing vertebrates, suggesting its broad use to assess restoration outcomes,” wrote the researchers.

The study was designed within the framework of the research group Reassembly, which is funded by the German Research Foundation (DFG).

The research is published in the journal Nature Communications.

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.

—-

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.