Blueprint for today’s climate change?
HAMBURG . UNIVERSITY
The end of the Permian is characterized by the largest mass extinction event in Earth’s history. 252 million years ago, a series of volcanic eruptions in Siberia resulted in the release of large amounts of greenhouse gases. Over the next few millennia, the climate eventually warmed by 10 degrees. As a result, on land, about 75% of organisms have gone extinct; In the oceans, this number is about 90%.
By analyzing how extinct sea creatures once lived, Dr Foster and his team were able to directly link their extinction to the following climate changes: reduced oxygen levels in the water, heat rising water levels and, most likely, ocean acidification.
These changes are similar to the current trend. “Needless to say, our findings on the Permian cannot be applied to modern climate change one-on-one. Foster, a geologist, says the two climate systems are too different. “However, they show which traits are important for an organism’s survival or extinction – under similar conditions. This can provide us with valuable indicators of who or what will be at greatest risk in the future. ”
Specifically, the team analyzed more than 25,000 records of 1283 fossil marine genera such as bivalves, snails, sponges, algae and crustaceans from the South China region – all with bones. or mineral shells. Their fossil remains can be dated using a special method, providing insights into marine ecosystems dating back millions of years. The team also drew on a huge database that provides additional information on various ecological aspects of how these creatures live.
For each genus, twelve of these criteria were analyzed. Did certain characteristics make a given organism more likely to survive the conditions prevalent in the late Permian – or not? With the help of machine learning, a method from the field of Artificial Intelligence, all these factors were analyzed simultaneously and simultaneously. In the process, the machine will essentially make certain rational decisions on its own. When this was done, the team compared the results: which organisms were present before, during, and after the mass extinction?
Their findings reveal the four most important factors for whether organisms survived the Permian: where they lived in the water, the mineralization of the shell, the diversity of species within the genus, and the sensitivity of the species. their sensitivity to acidification.
“But with previous machine learning applications, we couldn’t say how the machine made decisions.” Using a new approach developed from game theory, Dr. Foster has now succeeded in shedding light on this aspect: “Some animals live in deeper water. Here, the machine shows that the worsening hypoxia poses a risk. In contrast, animals that live near the surface face rising water temperatures. Plus, when you only have a limited habitat, you have nowhere to go when that particular habitat becomes uninhabitable.” As such, the results show which characteristics of the organism are identified as potentially lethal. The team were finally able to confirm that the mass extinction could be directly caused by deoxidation, rising water temperatures and acidification – which suggests that, in a future climate crisis, , these could also be the three main causes of long-term extinction.
Foster WJ, Ayzel G, Münchmeyer J, Rettelbach T, Kitzmann N, Isson TT, Mutti M, Aberhan M (2022): Machine learning identifies patterns of ecological selection through the late Permian mass extinction; Palaeontology;
DOI: 10.1017 / pab.2022.1
JOURNEYS
Palaeontology
DOI
RESEARCH METHODS
Data analysis / statistics
RESEARCH SUBJECTS
Animals
ARTICLE TITLE
Machine learning identifies patterns of ecological selection during the late Permian mass extinction
ARTICLE PUBLICATION DATE
March 1, 2022
REPORT REPORT
No conflict of interest
Related
