Scientists are increasingly using bioacoustic imaging to automatically analyze sounds that identify animal species in forests. Microphones and AI are used for this.

“Monitoring biodiversity has always been an expensive and difficult task,” says entomologist and ecologist David Donoso of the National Polytechnic School of Ecuador. “The problem only gets worse when you consider that good monitoring programs require many years of data to fully understand system dynamics.”

The researchers selected more than 40 sites with different types of landscape, including active farmland, plantations that had been abandoned for decades, and pristine old forests.

In the photo of the instrument on the left, you can see the microphone that recorded the sound. On the right is a light trap for catching insects.

After the team obtained these recordings, they turned to experts to identify birds and amphibians by their vocalizations, and used DNA from light traps to identify nocturnal insects. They also used artificial intelligence to identify the species of birds by sound.

“We can say that the scientific part is basically solved, so the AI models work,” says ecologist Jörg Müller of the University of Würzburg in Germany.

Several years of such records will track how the forest ecosystem evolves over time, when species populations increase or decrease, when new arrivals colonize the area, or when climate change affects birds.

Recall that astronomers witnessed how two planets the size of Neptune collided with each other, forming a cloud of gas and dust that resembles a donut.

Commentary