Nature
PUBLISHED
The process of evolution can be too complex, too slow, and too subtle to appreciate through the limited perceptions of us fleshy humans, but new research is showing how artificial intelligence could help pick out the patterns we can't easily see.
Theory suggests organisms don't change at a steady pace, but in pulses and bursts followed by long stretches of quiet. This pattern – often hinted at in the fossil recrod but not so easy to prove – explains how an array of new species with novel features emerge over a relatively short timescale before settling back into a groove.
"This is really important for evolutionary theory because there's a long history, going back 100 years, that predicts the emergence of new groups, called evolutionary radiations, is often associated with an explosive burst of diversification," said Jake Berv at the University of Michigan, in a statement.
"This could be because of a new ecological opportunity, or it could be because a group dispersed to a new continent, resulting in dramatic accelerations in their rate of evolution. The idea is that, over time, there's less opportunity as evolution proceeds, and so it slows down," he explained.
Berv's group has now published data in the form of more than 170,000 measurements taken from the skeletons of over 2,000 species of passerines, which is the largest order of birds, comprising the majority of all bird species.
This information was then plugged into a new AI model developed by the researchers that could precisely measure a dozen bones across specimens, allowing them to track changes in bird body shape over the group's 45-million-year history.
The analysis found that birds experienced a rapid burst of body-shape evolution around 35 million years ago. The timing is no coincidence. Known as the Eocene-Oligocene Transition, it saw a pivotal climate shift when Earth rapidly moved from a warm, ice-free "greenhouse" world into a cooler "icehouse" state.
Amid the global upheaval, new ecological niches opened up and natural selection went into overdrive. But eventually, it ran out of steam.
Around 15 million years ago, the analysis showed a cluster of evolutionary slowdowns, coinciding with a further period of global cooling, by which point most ecological niches had already been filled, tried, and tested.
“Our findings have definitely shifted my thinking about how the world works,” said Brian Weeks, also at the University of Michigan.
“This pattern we found with rare, big increases in the rates of evolution and lots of small decreases in the rate of evolution is really consistent with a pattern where lineages are exploring new ecological space and changing rapidly to take advantage of that opportunity,” Weeks added.
Sifting through huge amounts of data at speed to find patterns that aren't immediately obvious to the human eye is exactly the kind of discovery AI is well-suited to (and arguably a much better use of it than churning out artwork or trying to chat with fellow humans).
The new study is published in the journal Nature Ecology & Evolution.
1
