Artificial intelligence finds surprising patterns in Earth’s biological mass extinctions

Artificial intelligence finds surprising patterns in Earth's biological mass extinctions

Extinction is a normal part of Earth’s history. Most of the organisms that have ever lived are extinct. Sometimes a large number of species go extinct in a short period of time. It is the so-called mass extinction .

After a mass extinction, many habitats are no longer inhabited by organisms because they have become extinct. With new habitats available, some species will adapt to new environments. The process by which many new species evolve in a short period of time to fill available niches is called adaptive radiation .

There is no causal relationship

But a new study led by scientists affiliated with the Earth-Life Science Institute (ELSI) at the Tokyo Institute of Technology has used machine learning to examine the coexistence of fossil species and has found that radiation and extinctions are rarely present. connected and therefore mass extinctions probably rarely cause radiation of a comparable scale .

H Chimantensis2 Heliamphora chimantensis, a carnivorous plant that is only found in the Gran Sabana, Venezuela.

This study has compared the impacts of both extinction and radiation over the period for which fossils are available, the so-called Phanerozoic Eon . The Phanerozoic (from the Greek meaning ‘apparent life’), which represents the most recent ~ 550 million year period of Earth’s total history ~ 4.5 billion years, and is very important to paleontologists: before this period Most of the organisms that existed were microbes that did not easily form fossils, so the previous evolutionary record is difficult to observe .

The new study suggests that creative destruction is not a good description of how species originated or became extinct during the Phanerozoic, and suggests that many of the most notable periods of evolutionary radiation occurred when life entered new evolutionary and ecological scenarios, as during the explosion of animal diversity and the carboniferous expansion of forest biomes in the Cambrian. It is not known if this is true during the previous ~ 3 billion years dominated by microbes, as the paucity of recorded information on such ancient diversity did not allow for a similar analysis .

The team used a new machine learning application to examine the temporal coexistence of species in the Phanerozoic fossil record, examining more than a million entries in a huge curated public database that includes nearly 200,000 species.

Using their objective methods, they found that the ‘big five’ mass extinction events previously identified by paleontologists were picked up by machine learning methods as one of the top 5% of significant outages in which extinction outpaced radiation or vice versa. , as well as seven additional mass extinctions, two combined mass extinction-radiation events, and 15 mass radiation events. Surprisingly, in contrast to earlier narratives that emphasized the importance of post-extinction radiation, this work found that massive radiation and the most comparable extinctions were rarely coupled in time, refuting the idea of ​​a causal relationship between them .