How machine learning can help us protect chimpanzees 

The future of chimpanzee conservation relies on smart strategies, which in turn require detailed data that is not always easy to obtain.

Ecologist Adrienne Chitayat has led important research in Tanzania’s Mahale Mountains National Park, becoming the first to systematically survey chimpanzee population density across the entire park. 

Chitayat’s doctoral dissertation, conducted at the University of Amsterdam, will likely serve as a crucial baseline for future conservation in this region – which is home to Tanzania’s largest chimpanzee population. 

This area is significant not only for its biodiversity but because it represents the most southern range of the eastern chimpanzee, an endangered subspecies.

Estimating chimpanzee population density 

Chimpanzees, along with bonobos, are humans’ closest living relatives. While the Greater Mahale Ecosystem is vast, covering nearly 20,000 square kilometers, there has been limited research on chimpanzee populations across the whole park.

“There have been estimates based on small-scale or localized surveys, but no comprehensive baseline data. And without that data, it is difficult to understand chimpanzee population patterns and develop good conservation strategies. My goal was to fill that gap,” Chitayat said.

Chimpanzees build new nests each night, rarely reusing them, which makes nest counting a reliable method for estimating population density. 

Chitayat conducted these counts park-wide, determining density figures between 1.1 and 3.7 chimpanzees per square kilometer, depending on the specific habitat characteristics in each area.

Ecosystems threatened by climate change

The Mahale Mountains feature diverse landscapes, from dense forests to expansive savannahs. Chitayat observed that chimpanzees occupy all these habitats, even the open vegetation, rather than sticking solely to forested areas. This discovery is crucial as climate change threatens these ecosystems. 

“By looking at how the chimpanzees use the landscape and move within it, we can conclude that not only the densely forested areas are important for the protection of the animals,” said Chitayat.

Protecting this wide range of habitats, including establishing corridors, could support long-term population health and resilience.

Monitoring chimpanzees with acoustic detection

Traditional nest counting, while effective, is labor-intensive and costly. Recognizing the limitations, Chitayat developed a new monitoring approach: a deep learning-based acoustic detector. 

This technology uses passive acoustic monitoring to capture ambient sounds, including the calls of unhabituated chimpanzees that are difficult to observe up close. 

“Passive acoustic monitoring is a revolutionary technique that automatically records all the sounds in the vicinity of the acoustic device, including the sounds of chimpanzees. It can be used to find out where the chimpanzees are, how often, at what times, and how many individuals,” said Chitayat.

“The problem is that you have hours, days, even weeks of sound recordings that are far too laborious to listen to manually.”

Fortunately, her deep learning algorithm is able to automate the identification of chimpanzee sounds, saving time and resources.

Understanding chimpanzee communication

Creating the acoustic detector presented unique challenges. Chimpanzees communicate through a complex range of sounds, from loud, long-distance “pant-hoots” to softer grunts. 

Chitayat focused her detector on pant-hoots, which are vital for group cohesion and individual identification but difficult to interpret due to their variability. 

Developing the detector to recognize these calls required extensive training data and sophisticated algorithm design, marking a substantial achievement in bioacoustic technology.

Future of chimpanzee conservation

Chitayat’s work opens important doors for future innovations. She envisions refining the acoustic detector to distinguish individual chimpanzees, offering insights into group demographics, social dynamics, and age and sex distributions. 

Such data would be invaluable for shaping targeted conservation strategies that address specific population needs. “The more we learn, the better we can protect chimpanzees,” said Chitayat.

This technology could enhance our understanding of not just population size but also the social structure and health of groups over time.

Building resilience through data-driven strategies

With her research, Chitayat aims to integrate ecological data into conservation planning for chimpanzees in the Greater Mahale Ecosystem. 

By tracking population density, movement patterns, and habitat usage, conservationists can design interventions that align with the needs of this vulnerable species. 

In an era of climate change, such comprehensive data enables proactive measures, from protecting essential habitats to creating corridors that allow chimpanzees to adapt to shifting environments.

Fuller understanding of chimpanzee populations 

Chitayat’s contributions highlight the power of combining field ecology with advanced technology. By using nest counts, acoustic monitoring, and deep learning, her work brings a fuller understanding of chimpanzee populations and offers a model for innovative conservation strategies. 

As climate change reshapes ecosystems, tools like Chitayat’s detector will be essential for adapting and responding to the challenges facing chimpanzee populations.

—–

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.

—–