By Julio Cesar Rivas |
Toronto (Canada) (EFE).- The combination of Artificial Intelligence (AI) and the use of hydrophones -underwater microphones- makes it possible to observe and monitor the growing impact of human activities in the oceans, as well as the habits of marine life , with a precision that is astounding scientists.
For decades, Jesse Ausubel, director of the Program for the Human Environment at Rockefeller University in New York (USA), has studied the oceans and the life they contain.
Ausubel was one of the promoters of the Census of Marine Life (CVM), a large-scale project that for more than a decade cataloged all known marine species and in which thousands of scientists from all over the world worked between 2000 and 2010.
The American researcher is also one of the promoters of the innovative technique of environmental DNA analysis that makes it possible to identify the species that inhabit a certain area by extracting genetic material from seawater samples.
Acoustic surveillance to study marine life
And for years, Ausubel and his colleagues have been amassing acoustic material, picked up by hundreds of hydrophones deployed around the world to study marine life.
Since 2021, scientists have been using those acoustic recordings to create the so-called Global Library of Underwater Biological Sounds (Glubs), where each sound corresponds to an animal.
The director of Glubs, Miles Parsons, a scientist at the Australian Institute of Marine Science, explained in a statement that passive acoustic monitoring is especially effective “in remote, deep, dark, turbid or rapidly changing places” and that it will allow the discovery of as yet unknown marine species. .
Ausubel clarified to EFE that until recently it was “a huge task”: “There was a large amount of information but very few means to digest it. If I put a hydrophone on a coral reef and leave it there for a year, it would take me three years to listen to the recorded material. And then it is not certain that he could understand what he was hearing, ”he said.
“But now with machine learning or Artificial Intelligence, a computer can listen to a whole year of recordings in a few seconds and compare it with the content of Glubs. Now we can begin to understand the sounds of the ocean”, added the American scientist.
ocean habits
One of the first applications of this novel pairing between machine learning and ocean sounds has been made by groups of scientists from the International Quiet Ocean Experiment (IQOE).
In a study published this Wednesday in the Journal of the Acoustical Society of America (Jasa), a team of researchers, including Bishwajit Chakraborty, from IQOE, explained how the combination of hydrophones and artificial intelligence have enabled the identification of different species of fish in the Zuri estuary, India, and reveal their habits.
The scientists pointed out that thanks to AI they now know that species concentrate their activities in “shift work”: some are active between 3am and early afternoon, while others opt for a “night shift”. , from 2 in the afternoon until 2.45 in the morning.
In addition, the scientists were able to verify the different levels of abundance of marine life before and after the monsoon season, the storms that affect the region.
Thanks to the sounds captured, they identified different species of shrimp, including the most precious ones for commercial exploitation.
“A milestone”
All in all, the scientists concluded that hydrophones combined with AI are a powerful tool for monitoring the abundance of fish in the ecosystem.
Ausubel called the study “a milestone” for ocean science and expressed surprise at the finding of clearly differentiated “shift work” between different species.
“I never thought they would be so marked,” he said.
Ausubel insisted on the enormous potential offered by the analysis of marine sounds with AI, especially in combination with the analysis of the genetic material extracted from the waters.
“Given the growing number of concerns about the industrialization of the oceans and the effects of climate change, the combination of sound analysis and environmental DNA allows us to have timely information on changes in marine life associated with modifications caused by man or other causes”, explained the scientist.
“This combination will really produce the kind of knowledge we want to have about how marine life is recovering or degrading,” he concluded.