Scientists try to listen to learn about wildlife

A team of scientists is studying the sound of an Ecuadorian forest to learn how artificial intelligence (AI) can track animal life in a restoration environment.

When scientists want to measure new forest growth, they can survey large areas of soil with tools such as satellites and lidar.

But understanding how quickly and in what numbers wildlife is returning to an area is more difficult. Sometimes this requires an expert to listen to audio recordings and pick out animal calls.

Jörg Müller is an expert on birds at the Biocenter of the University of Würzburg in Germany. He wondered if there was any other way.

Mueller told the French news agency AFP. “I saw gap that we need, especially in the tropics, better methods quantify huge variety… improve conservation operations.

So he turned to bioacoustics, which uses sound to learn more about the lives of animals and the environments in which they live.

The tool has been used by scientists for some time. But recently, researchers are using it with computer learning to study large amounts of data more quickly.

Muller and his team recorded wildlife sounds at sites in the Choco region of Ecuador. The habitats they recorded included areas that were once used for agriculture and cattle ranching as well as old-growth forests.

They first had experts listen to the recordings and index the sounds of different animals. They then examine the sound quality to measure the environment.

In the end, they spent two weeks making recordings using an artificial intelligence computer program trained to understand 75 different bird calls.

More recordings are needed

The program was able to select the calls it was trained on. However, the scientists were interested in whether the program could correctly determine the number of different types of plants and animals in each habitat.

To see if the program could do this, the team used two different controls. One was from experts who listened to the recordings, and the second was based on examples of each medium that could be used for understanding; biodiversity.

Because the number of available sounds used for training is limited, the artificial intelligence software can identify only one-fourth of the bird calls that experts could. But it was still able to correctly measure the level of biodiversity in each habitat, the study said.

The study was recently published Communications of nature. In the study, the scientists’ findings show that the AI ​​program is a powerful tool to measure the recovery of animal communities. tropical forests.

The study notes that biodiversity detected from recordings can be quantified in a cost-effective and holistic manner. And it is said to be able to measure environments ranging from active agriculture to regenerating and old-growth forests.

There are still areas for improvement, including the lack of animal sounds to train the AI ​​models on. And the method can only attract animals that use sound to communicate.

I’m Gregory Stachel.

This is reported by Agence France-Presse. Gregory Stachel adapted the story for VOA while learning English.

_________________________________________________

Words in this story

lidar n:. a remote sensing method that uses light in the form of a pulsed laser to measure ranges (variable distances) on Earth

gap n:. missing part

quantify v:. find or calculate the quantity or amount of (something).

variety n:. the quality or state of having many different forms, types, or ideas

conservation n:. protection of animals, plants and natural resources

tropical regions n:. the part of the world near the equator where the weather is very hot