[UPDATE: I wrote an updated version of this piece for a talk I delivered at EAG 2017. It has more sources and better info about the extinctions.]
This post has also been published on the Global Risk Research Network, a group blog for discussing risks to humanity. Take a look if you’d like more excellent articles on global catastrophic risk.]
Several times in evolutionary history, the arrival of an innovative new evolutionary strategy has lead to a mass extinction followed by a restructuring of biota and new dominant life forms. This may pose an unlikely but possible global catastrophic risk in the future, in which spontaneous evolutionary strategies (like new biochemical pathways or feeding strategies) become wildly successful, and lead to extreme climate change and die-offs. This is also known as a ‘biotic replacement’ hypothesis of extinction events.
- Biotic replacement in past extinctions
- Is this still a possible risk?
- Risk factors from climate change and synthetic biology
- The shape of the risk
- What next?
Identifying specific causes of mass extinction events may be difficult, especially since mass extinctions tend to be quickly followed by expansion of previously less successful species into new niches. A specific evolutionary advantage might be considered as the cause when either no other major physical disruptions (asteroids, volcanoes, etc) were occurring, or when our record of such events doesn’t totally explain the extinctions.
1. Biotic replacement in past extinctions
There are five canonical major extinction events that have occurred since the evolution of multicellular life. Biotic replacement has been hypothesized as either the major mechanism for two of them: the late Devonian extinction and the Permian-Triassic extinction. I outline these, as well as four other extinction events.
Great oxygenation event
Cyanobacteria became the first microbes to produce oxygen (O2) as a waste product, and began forming colonies 200 million years before the extinction event. O2 was absorbed into dissolved iron or organic matter, and the die-off began when these naturally occurring oxygen sinks became saturated, and toxic oxygen began to fill the atmosphere.
The event was followed by die-offs, massive climate change, and permanent alteration of the earth’s atmosphere, and eventually the rise of the aerobic organisms.
The Ediacaran period was filled with a variety of large, autotrophic, sessile organisms of somewhat unknown heritage, known today mostly by fossil evidence. Recent evidence suggests that one explanation for this is the evolution of animals, able to move quickly and and re-shape ecosystems. This resulted in the extinction of Ediacaran biota, and was followed by the Cambrian explosion in which animal life spread and diversified rapidly.
Late Devonian extinction
19% of families and 50% of genera became extinct.
Both modern plant seeds and modern plant vascular system developed in this period. Land plants grew significantly as a result, now able to more efficiently transport water and nutrients higher – with maximum heights changing from 30 cm to 30 m. Two things would have happened as a result:
- The increase in soil content produced more weathering in rocks, which released ionic nutrients into rivers. The nutrient levels would have increased plant growth and then death in oceans, resulting mass anoxia.
- Less atmospheric carbon dioxide would have cooled the planet.
96% of marine species, and 70% of land vertebrate species went extinct. 57% of families and 83% of general became extinct.
One hypothesis explaining the Permian-Triassic extinction events posits that an anaerobic methanogenic archaea, Methanosarcina, developed a new metabolic pathway allowing them to metabolize acetate into methane, leading to exponential reproduction and consuming vast amounts of oceanic carbon. Volcanic activity around the same time would have released large amounts of nickel, a crucial but rare cofactor needed for Methanosarcina’s enzymatic pathway.
Dead members of especially efficient fern genus built up in the ocean over 800,000 years and created a massive carbon sink, leading to a snowball earth scenario and mass global cooling.
Quaternary and Holocene extinction events
The evolution of human intelligence and human civilization has lead to mass climate alteration by humans. Another set of adaptations among human society (IE agriculture, use of fossil fuels) could be considered here, but in terms of this hypothesis, the evolution of human intelligence and civilization could be considered to be the driving evolutionary innovation.
Minor extinction events
Any single species that goes extinct due to a new disease can be said to have become extinct due to another organism’s innovative adaptation. These are less well described as “biotic replacement”, because the new pathogen won’t be able to replace its extinct hosts, but it was still an evolutionary event that caused the disease. A new disease may also attack the sole or primary food source of an organism, leading to its extinction indirectly.
2. Is this still a possible risk?
It seems unlikely that all possible disruptive evolutionary strategies have already happened: Disruptive new strategies are rare – while billions of new mutations arise every day, any new gene must meet stringent criteria in order to spread: Is actually expressed, is passed on to progeny, immediately conveys a strong fitness benefit to its bearer, serves any vital function of the old version of the gene, is supported by the organism’s other genes and environment, and the organism isn’t killed by random chance before having the chance to reproduce. For instance, an unusually efficient new metabolic pathway isn’t going to succeed if it’s in a non-reproducing cell, if its byproducts are toxic to the host organism, if its host can’t access the food required for the process, or if its host happens to be born during a drought and starves to death anyways.
Environmental conditions that make a pathway more or less likely to be ridiculously successful, meanwhile, are constantly changing. Given the rareness of ridiculously successful genes, it seems foolhardy to believe that evolution up til now has already picked all low-hanging fruit.
How worried should we be? Probably, not very. The major extinction events listed above seem to be spaced by 100-200 million years, suggesting a 1-in-100,000,000 chance of occurring in any given year. For comparison, NASA estimates that asteroids causing major extinction events strike the earth every 50-100 million years. These threats are possibly on the same orders of magnitude.
(This number requires a few caveats: This is a high estimate, assuming that evolutionary advantages were a major factor in all cases. Also, an advantage that “starts” in one year may take millions of years to alter the biosphere or climate catastrophically. Once in 100 million years is also an average – there’s no reason to believe that disruptive evolutionary events, or asteroid strikes for that matter, occur on regular intervals.)
On a smaller scale, entire species are occasionally wiped out by a single disease. This is more likely to happen when species are already stressed or in decline. Data on how often this happens, or what fraction of extinctions are caused by a novel disease, is hard to find.
3. Risk factors from climate change and synthetic biology
Two risk factors are worth noting which may increase the odds of a biotic replacement event – climate change and synthetic biology.
Historically, a catastrophic evolutionary innovation seems to follow other massive climate disruption, as in the Permian-Triassic extinction explanation that followed volcanic eruptions. A change in conditions may select for innovative new strategies that quickly take over and produce much more disruption than the instigating geological event.
While the specific nature of the next disruptive evolutionary innovation may be nigh-impossible to predict, this suggests that we should give more credence to environment alteration as a threat – via climate change, volcanic eruptions, or asteroids – as changing environments will select for disruptive new alleles (or resurface preserved strategies.) This means that a minor catastrophic event could snowball into a globally catastrophic or existential threat.
The other emerging source of alleles as-of-yet unseen in the environment comes from synthetic biology, as scientists are increasingly capable of combining genes from distinct organisms and designing new molecular pathways. While genes crossing between wildly different organisms is not unheard of in nature, the increased rate at which this is being done in the laboratory, and the fact that an intentional hand is selecting for viability and novelty (rather than natural selection and random chance), both imply some cause for alarm.
A synthetic organism designed for a specific purpose, may disperse from its intended environment and spread widely. This is probably especially a risk for organisms using completely synthetic and novel pathways unlikely to have evolved in nature, rather than previously evolved genes – otherwise, the naturally occurring genes would have probably already seized low-hanging evolutionary fruit and expanded into possible niches.
4. The shape of the risk
How does this risk compare to other existential risks? It is not especially likely to occur, as described in Part 2. The precise shape or cause of the risk is harder to determine than, say, an asteroid strike. Also, as opposed to asteroid strikes or nuclear wars, which have immediate catastrophic effects, evolutionary innovations involve significant time delays.
Historically, two time delays appear to be relevant:
- Time for the evolution to become widespread
Presumably, this is quicker in organisms that disperse/reproduce more quickly. EG, this could be fairly quickly for an oceanic bacteria with a quick generation cycle, but slowly for the 180,000 years it took between the first appearance of modern humans, and their eventual spread to the Americas.
- Time between the organism’s dispersal and the induction of a catastrophe
EG, during the global oxygen crisis, it took 200 million years from the evolution of the species, to when the possible oxygen sinks filled up, for a crisis to occur. (At least some of this time included the period required for cyanobacteria to diversify and become commonplace.)
During the azolla event, azolla ferns accumulated for 800,000 years causing steady climate change. The modern threat from anthropogenic global warming is much steeper than that.
What are the actual threats to life?
- Climate change
- The great oxygenation event and the Permian-Triassic extinction hypothesis involve the dispersal of a microbe that induces rapid, extreme climate change.
- Other events such as volcanoes erupting may change the environment such that a new strategy becomes especially successful, as in the Permian-Triassic extinction event.
- Faster, stronger, cleverer predation
- The Ediacaran extinction event and the Holocene extinction event involved the dispersal of an unprecedentedly capable predator – animals and humans, respectively.
- This seems unlikely to be a current risk. The risk from runaway artificial intelligence somewhat resembles this concern.
- Death from disease
- Any event in which a novel disease causes a species to go extinct has a direct impact. Additionally, a disease might cause one or more major food sources to go extinct (for humans or animals.)
- Globalization and global trade has increased the risk of a novel disease spreading worldwide. This also mirrors current concerns over engineered bioweapons.
5. What next?
Disruptive evolutionary innovation is problematic in that there don’t appear to be clear ways of preventing it – evolution has been indiscriminately optimizing away for billions of years, and we don’t appear to be especially able to stop it. Building civilization-sustaining infrastructure that is more robust to a variety of climate change scenarios may increase our odds of surviving such a catastrophe. Additionally, any such disruptive event is likely to happen over a long period of time, meaning that we could likely mitigate or prepare for the worst effects. However, evolutionary innovation hasn’t been explored or studied as an existential risk, and more research is needed to clarify the magnitude of the threat, or which – if any – interventions are possible or reasonable to study now.
Questions for further study:
- How common are extinction events due to disruptive evolutionary innovation?
- What factors make these evolution events more likely?
- How often do species go extinct due to single disease outbreaks?
- Can small-scale models help us improve our understanding of the likelihood of global warming inducing “runaway” scenarios involving microbial evolution?
- What man-made environmental changes could potentially lead to runaway microbial evolution?