New and old ideas about Earth's oxygen history
Textbooks assert that oxygen first appeared in Earth's atmosphere 2.3 billion years ago during the Great Oxidation Event, when photosynthesizing bacteria started to release oxygen into the atmosphere. The free oxygen reacted with iron dissolved in the oceans, and settled to the ocean floor as iron oxide.
Although rusty red bands in the sedimentary record support that idea, geochemical data prompt questions on the timing of free oxygen in the atmosphere and the related evolution of photosynthesis itself.
The second major rise of atmospheric oxygen preceded the Cambrian explosion of 550 million years ago, when many lifeforms diversified. Recently, more detailed geochemical studies have started to complete the framework by which oxygenation and its cause-and-effect relationship with biological evolution may be understood.
If you want to understand the origin of life, you have to understand the origin of oxygen. “At 2.45 billion years ago, you get the red iron bands. The question is, between the oldest rocks and 2.45 billion years, how much oxygen was there and where did it come from?” says Mark Thiemens, an atmospheric chemist at the University of California, San Diego.
These red, banded-iron formations in South Africa are 3 billion years old. CREDIT: Nic Beukes
The composition of rocks does not reveal oxygen levels. On our water-filled planet, original signatures that might have been left by oxygen have become unreliable. Other elements and isotopes are therefore required as proxies for determining oxygen levels over Earth's history. Thiemens and geologist James Farquhar, now at the University of Maryland, made a significant advance in that direction when looking at Martian rock samples to find evidence of water as revealed by sulfur isotopes.
“If you have a reaction that occurs in solids or liquids, it's highly unlikely that you will fractionate isotopes in a way that does not preserve a certain linear relationship between all the different isotopes,” says Farquhar. The magnitude of an effect that results from the fractionation of 34S from 32S will be about twice the magnitude that results from the fractionation of 33S from 32S.
But lab work shows that in gas-phase chemistry, other factors produce isotope variation. It turns out that sulfate, which forms when sulfur is oxidized, is a very complicated molecule in gas-phase reactions. In particular, reaction products that include the 33S isotope are anomalously large compared with those that contain 34S or 32S. Such anomalies are called mass-independent fractionation.
On Mars, UV light photolyzes sulfate and leads to fractionation. Laboratory experiments revealed that sulfur isotopes in Martian rocks show a mass-independent distribution. If there were no oxygen in Earth's atmosphere, UV light would get through as well, and similar anomalies would appear in the sulfur isotope record. “We know that sulfur in today's atmosphere gets oxidized by peroxide in ozone,” explains Thiemens.
Farquhar did indeed see this isotope variation in 2.45 billion-year-old Earth rocks. “We realized that this variation existed and the only mechanism we knew of that could produce it was related to gas-phase reactions,” he says. All that remained was to figure out the atmospheric conditions.
“The sulfur allows you to see at most how much oxygen there was at the time,” says Thiemens. During the period prior to 2.45 billion years ago, the photolytic effect can be seen. Then, when there was sufficient oxygen in the atmosphere, normal mass-dependent chemistry appears in the isotope record. Models suggested that this oxygen threshold is about 10 ppm.
“We have been able to show fine-scale differences between the [sulfur isotopes] that show up. They fall into bins in different places in Earth's early record, and might give insight to what's happening in the atmosphere and oceans at any given times,” summarizes Farquhar.
The sulfur isotope study provided one of the first continuous records that could be used to demonstrate a long-advocated change in oxygen levels. Other studies have used metals to study reactions that might be related to air oxidation states and redox reactions that occur in oceans and during weathering.
Earlier or later?
“There are polarized views in the geochemical and early Earth community about when oxygenic photosynthesis evolved,” says Sean Crowe of the University of British Columbia. One camp insists that oxygen appeared in the atmosphere 2.3-billion years ago, arguing that any signs of earlier oxygen are related to post-depositional processes that occurred after the formation of the sediments and after atmospheric oxygenation. The other camp votes for earlier evolution.
Using chromium isotopes as a sensitive indicator for oxidative weathering, Crowe found evidence of oxygen traces in 3-billion-year-old Nsuze paleosol, an ancient soil sandwiched between volcanic and sedimentary rocks in South Africa. Lighter isotopes of chromium react more rapidly in chemical and biological reactions, so the ratio of heavy to light chromium is used to trace these reactions. It turns out that only reactions caused by separation of heavy and light chromium involve oxygen.
Crowe and colleagues also looked at ancient soils and marine sediments in the Pongola Supergroup in South Africa. The sequence provides relatively pristine samples that have not been subjected to high levels of metamorphism or modern day weathering. The sediments show enrichment of heavy relative to light chromium, while the soils show depletion. “That's what we would expect when weathering reactions that cause soil formation take place in the presence of oxygen,” says Crowe.
Since the soil samples were recovered from a drill core several hundred meters below the surface, any modern-day weathering would be unlikely. Rather, signs of oxygen would be in the mineralogy and geochemistry. And while some marine sediments come from exposed outcrops, primary textures are preserved. What's more, the chromium isotope signal would be reversed by any post-depositional processes.
Earlier signs of oxygen mean that “permanent accumulation of higher levels of oxygen was delayed from the time that photosynthesis evolved,” says Crowe. Ultimately, a combination of biological and geological processes had to combine, and volcanic reduced gases would have rapidly consumed most oxygen produced.
The debate continues. Jena Johnson and colleagues at Caltech suggest that a form of photosynthesis using manganese as an electron donor preceded the evolution of cyanobacteria and the rise of oxygen around 2.4 billion years ago.
Crowe plans to conduct chromium tests on other rocks, and study modern environments to “learn more about how the proxy works and how it would be altered by post-depositional processes.” Additionally, independent genomic analyses help to constrain the evolution of different genes involved directly in oxygenic photosynthesis, as well as of other cyanobacteria elements; these analyses will move us closer to an overall timing constraint.
The next big thing
A billion years of anoxic oceans followed the Great Oxidation Event. Geochemical evidence points to the general oxygenation of the oceans around 635 million years ago, likely requiring a minimum of 10% present atmospheric levels. Whether through cause or effect, the ecosystem began to diversify at around the same time.
“Right now, we lack solid constraints on the baseline of atmospheric oxygen leading up to the second rise,” says Noah Planavsky of Yale University. “That makes it difficult to understand the potential consequence, but also what geochemical change would have induced fundamental changes in oxygen levels.”
But the “boring billion” years between the two major oxygen rises brought with them dropping oxygen levels and constrained evolutionary activity. University of Tasmania geologist Ross Large analyzed ancient seafloor muds to track trace elements absorbed from the seawater over time in pyrite iron sulfate, a unique mineral that forms in the muds.
“Some trace elements are more soluble in the ocean when oxygen increases,” explains Large. “Thus we can use the trends in trace elements in pyrite to predict periods when atmosphere oxygen rose and fell.” A laser analysis that detects lower levels than was previously possible showed that many nutrient trace elements decrease in the ocean from 1800 to 800 million years ago.
Trace elements and oxygen likely declined because tectonics were stable and little erosion occurred on the continents. The lack of nutrients meant a drop in bacteria, leading to less oxygen being liberated to the atmosphere. Ultimately, however, a better global database of samples and laser geochemistry is needed to produce more robust curves for trace elements and oxygen.
“We speculate that the increased trace elements [at 750 million years ago] led to a rapid increase in bacteria and plankton in the oceans, which in turn released oxygen to the atmosphere and sped up oxidative erosion of the continents,” says Large. The increased ocean productivity led to increased carbon burial on the sea floor, which drew trace elements and sulfur to the muds as sedimentary pyrites. That burial process increased atmospheric oxygen even more, and the Cambrian explosion of life was primed.
There were definitely no animals before the 2.3 billion-years-ago event, but the big question is whether the late neoproterozoic rise 600 million years ago was necessary before the appearance of animals,” says geobiologist Lewis Ward of Caltech. Fossils of sponges—the simplest animals, and indeed ones whose physiology has remained the same throughout history—exist from 635 million years ago, suggesting that the first animals did start to appear before the second major oxygen rise.
Daniel Mills, also of Caltech, Ward, and colleagues created a small aquarium in which they could vary, measure, and control oxygen levels, and observe the survival of sponges and their respiration rates at different concentrations. They determined that modern sponges can survive for reasonably long timescales at 0.5 to 4% of current oxygen levels. While Precambrian oxygen levels are not tightly constrained, general consensus is that most of the proterozoic had 1–10% of modern oxygen levels. That's enough to support sponge-grade animals.
One of the Halichondria panicea sea sponges that Daniel Mills and his collaborators used to investigate the likely oxygen demands of early animals. CREDIT: D. Mills
“We now think that oxygen was extremely low in the period before the appearance of sponges,” agrees Tim Lyons of the University of California, Riverside. But metabolically active animals, which demand energy and oxygen to diversify and grow in size, appeared later. “Oxygen levels could have been high enough to allow metozoans to thrive, but evolution could have triggered the rise in oxygen,” adds Planovsky.
The chicken-or-egg question thus prevails. One curious anomaly is the Lomagundi carbon isotope excursion at around 2 billion years ago. Oxygen from photosynthesis can combine with organic matter, and carbon isotopes get heavy when organic matter is buried. Thus a record of heavy carbon isotopes indicates a net gain of oxygen in the atmosphere. “That [excursion] followed a very dramatic fall [in oxygen levels]. It's a fascinating time,” adds Lyons.
“What's interesting, as we construct more records from the time period, is that oxygen may have actually risen to significant levels during this time interval. This supports intervals of significant rise, then fall, in oxygen,” adds Planovsky.
He and Lyons, like Crowe, argue that oxygen production happened early on, and that periods of increasing and decreasing oxygen happened in short-term cycles. Abundance rose and fell depending on how much was produced by bacteria, and how much was consumed by sinks or reaction with reduced ions.
Earth's oxygenation history appears more complex than a unidirectional rise. Piecing it together from biological, chemical, geological, and other lines of evidence remains a work in progress.