Machine Learning Helps Rewrite Earth’s History
Darwin’s theory regarding speciation, the process by which new species emerge from existing ones, concerned the process by which flora and fauna adapt to their environments.
But there is more to planet Earth than individual plants and animals. They are part of a vast ecosystem — think all the members of a coral reef or rainforest — destroyed and reborn many times over in Earth’s history.
Ecosystem destruction, and subsequent rebirth, have long been thought to be a case of clear-cut cause and effect. As an extinction happens, new species rise from the ashes and fill in the gap.
Now, for the first time, a team of earth and life science researchers from across the globe have used machine learning to examine massive amounts of information about how life develops on our planet. What they discovered challenges what we thought we knew about extinctions, ecosystem rebirth, and the emergence of species.
The group’s findings also have implications for medicine.
What are mass extinctions?
In a mass extinction, a vast number of species die off within a relatively short period of time. As National Geographic puts it, “In a handful of times in the last 500 million years, 75 [percent] to more than 90 percent of all species on Earth have disappeared in a geological blink of an eye.”
The “ Big Five ” were the largest extinctions.
Ordovician-Silurian, 440 million years ago.
Devonian, 365 million years ago.
Permian-Triassic, 250 million years ago.
Triassic-Jurassic, 210 million years ago.
Cretaceous-Tertiary, 65 million years ago.
How does machine learning help?
Industry leader IBM defines machine learning as “a branch of artificial intelligence (AI) focused on building applications that learn from data and improve their accuracy over time without being programmed to do so.” These applications, called algorithms, are “’trained’ to find patterns and features in massive amounts of data in order to make decisions and predictions based on new data.” For this study, the researchers’ algorithm analyzed over one million fossils from a public database, covering over 200,000 species, to track how extinctions, and subsequent life, occurred.
“Some of the most challenging aspects of understanding the history of life are the enormous timescales and numbers of species involved,” said lead author Jennifer Hoyal Cuthill, PhD, in a press release. “New applications of machine learning can help by allowing us to visualise this information… we can, so to speak, hold half a billion years of evolution in the palms of our hands, and gain new insights from what we see.”
In the conventional model, mass extinctions are part of a process called “creative destruction.” Many new species then emerge in a burst called a “radiation.” One of the most well-known was the rise of mammals after dinosaurs disappeared during the Cretaceous-Tertiary extinction.
What the researchers found, however, was that “comparable mass radiations and extinctions were only rarely coupled in time, refuting the idea of a causal relationship between them.” In fact, the researchers found what they call “destructive creation” — evidence that radiations may themselves cause major changes in the ecosystem. The rate of species turnover– how long it takes before the majority of species have been replaced by new ones — sped up with both mass extinctions and mass radiations.
The scientists noted that the Quaternary period, which began 6 million years ago and continues today, has already experienced massive changes in climate as ice ages begin and end. Now, a so-called Sixth Extinction, “mainly caused by human activity including hunting and land-use changes caused by the expansion of agriculture,” is “eroding biodiversity that was already disrupted.”
Why does this matter?
Biodiversity plays an important but underappreciated role in medicine. The WHO says that “Loss in biodiversity may limit discovery of potential treatments for many diseases and health problems.” Just this year, researchers found tantalizing clues on how to fight Covid-19 by analyzing the immune systems of animals like llamas and sharks.
Dr. Cuthill told Phys.org that, “Each extinction that happens on our watch erases a species that may have existed for millions of years up to now, making it harder for the normal process of new species origination to replace what is being lost.”
You’ve probably heard of dodo birds and passenger pigeons as common examples of man-made extinctions. According to a 2008 study published in the Journal of the National Cancer Institute , “current extinction rates are at least 100 to 1,000 times higher than natural background rates” and 15,000 medicinal herbs are at risk. Even before industrialization, human activity is believed to have destroyed valuable plants like the legendary Roman wonderdrug silphium.
Climate change has been called “the defining crisis of our time” by the UN Refugee Agency due to its impact on where and how people — and by extension, other life– can continue to live. Extensive research shows that climate change can accelerate extinctions many times over. One study in Proceedings of the National Academy of Sciences asserts that “climate change may be a major threat to biodiversity in the next 100 years,” with as many as 44% of plants surveyed in one area already experiencing local extinctions. Another study identified hundreds of species at risk.
The ecosystem has so far been capable of self-correcting after mass extinctions and radiations, but the time scale this occurs in is staggering. According to the Nature study’s press release, after the Sixth Extinction, “the authors suggest it will take at least 8 million years for it to revert to the long term average of 19 million years.”
The Take Home
Environmental protection movements have gained steam in the last couple decades, but few people appreciate the toll extinctions can take on medical advancement. The new findings are a sobering reminder that the effects of climate change and man-made extinctions could be even longer-lasting and more difficult to reverse than we think.
Sean Marsala is a health writer based in Philadelphia, Pa. Passionate about technology, he can usually be found reading, browsing the internet and exploring virtual worlds.