For a long time, the narrative was straightforward: for the first 2 billion years of Earth’s history, there was virtually no free oxygen in the atmosphere or oceans. Then, around 2.45 billion years ago, the ancestors of photosynthetic cyanobacteria began saturating the planet with oxygen, creating the conditions that would allow complex life to evolve and thrive. This turning point, known as the Great Oxidation Event (GOE), is described by geochemist Mojtaba Fakhraie of the University of Connecticut as “the most astonishing and dramatic change in the functioning of Earth’s surface in its entire history.”
However, recent findings are adding new layers to this story. Mineral analyses indicate that oxygen‑producing microorganisms appeared hundreds of millions of years before the GOE, leaving a vast, unexplained gap during which oxygen levels remained low. Moreover, evidence now suggests the GOE was not a single event but a series of phases over roughly 200 million years, with oxygen levels rising and falling repeatedly — a pattern that included what researchers call “Great Deoxygenation” episodes following periods of oxidation.
Emerging hypotheses are beginning to offer possible explanations. At a conference honoring Goldschmidt, one presentation proposed that oxygen could not accumulate at the surface until extensive shallow‑water areas had formed in the world’s oceans. Another talk argued that spikes in oxygen levels reflect the periodic transformation of early Earth into a “Snowball Earth,” with ice extending to low latitudes.
For decades, scientists have used a sharp shift in the ratio of three sulfur isotopes in rocks as evidence of global glaciation. This isotopic signal reflects changes in the breakdown of sulfur dioxide in the atmosphere under ultraviolet (UV) radiation. Those reactions slowed once rising oxygen levels led to the formation of an ozone layer that blocked UV radiation. This isotope shift has long been used to pinpoint the GOE at 2.45 billion years — a date still widely accepted.
Yet research previously published in Nature identified carbonate deposits formed by oxygen‑producing cyanobacteria as early as 2.85 billion years ago. Estimates based on mutation rates in modern microbes push the origin of photosynthesis even further back, to around 3.5 billion years ago. For two decades, researchers have tracked so‑called oxygen “spikes” preceding global glaciations, as recorded in minerals that can only form in the presence of oxygen.
Taken together, these studies paint a compelling picture: global oxygenation did not begin immediately after the emergence of photosynthetic microbes but only hundreds of millions of years later, Fakhraie explains. The central question remains: why?
Not all scientists agree that oxygenic photosynthesis evolved so early. Jena Johnson, a geochemist at the University of Michigan, has shown that microbes may have first developed a form of photosynthesis that did not produce oxygen. Furthermore, some metal oxides once thought to signal biological oxygen have been proven to form through photochemical reactions driven by UV radiation. “We still have a lot to learn,” Johnson notes.
Assuming the gap between the rise of photosynthesis and the global oxidation event is real, some researchers propose that early oxygen was consumed by reactions with reducing substances such as dissolved iron in the oceans. This consumed oxygen contributed to the formation of iron‑manganese nodules with dark red bands of oxidized iron, found in Australia and elsewhere. Later, it was suggested that a steady influx of reducing compounds from Earth’s interior prevented oxygen from accumulating. However, Johnson points out that it is unclear how such a chemical “sponge” could have remained effective for hundreds of millions of years.
A new explanation, proposed by Fakhraie and his co‑authors in a Goldschmidt conference presentation, ties the rise of oxygen to continental shelves — shallow ocean areas that today are among the most productive zones for photosynthetic oxygen despite covering only 9 % of the ocean’s surface. When marine cyanobacteria die in these shelf regions, their carbon is rapidly buried in sediments, preventing it from being oxidized and thus from consuming atmospheric oxygen. Evidence suggests that these shelves began to expand around the same time as the formation of young continents. According to Fakhraie, modeling indicates that once shelf areas exceeded about 10 % of their current extent, atmospheric oxygen levels would have begun to rise. “This supports the idea that tectonic activity could have driven large‑scale environmental changes,” he says.
Still, these changes were not stable. Earlier this decade, a team of geochemists reported that signs of low oxygen levels in sulfides from South African marine sediments recurred over a 200‑million‑year span following the onset of the GOE — including periods believed to correspond to three separate “Snowball Earth” episodes. “We realized that [global glaciation] was more of an episodic process,” says geochemist Andrei Becker of the University of California, Riverside, co‑author of a 2021 Nature paper.
At the Goldschmidt conference, Colin Goldblatt, a climate scientist at the University of Victoria, argued that Snowball Earth episodes may have been responsible for the unstable start to Earth’s oxygenation. Geological data largely support the idea that the first Snowball Earth occurred before the Permian mass extinction, possibly triggered by a drop in carbon dioxide (CO₂) due to increased uptake by rocks or reduced volcanic output.
Whatever the cause, such a global freeze would have severely limited the oceans’ ability to absorb CO₂, allowing it to build up in the atmosphere. This would eventually trigger a rebound warming, melting the ice within about 10,000 years, while CO₂ remained in the atmosphere for roughly 100,000 years. In the resulting hot climate, large amounts of phosphorus would have been washed from the continents into the oceans, fueling a bloom of marine microorganisms and boosting oxygen production. This, in turn, would have helped form an ozone layer, which would then limit the breakdown of oxygen by UV radiation.
When the next Snowball Earth episode occurred and microbial productivity in the oceans declined, oxygen levels would have fallen again. “Oxygen levels varied with Earth’s climate until the planet escaped the cycle of low‑latitude glaciations,” Goldblatt explains.
According to Goldblatt, early Earth researchers are only just beginning to piece together a coherent picture of global glaciation and oxygenation — one that likely involves factors beyond continental shelves. “It’s likely that everyone in this room has their own idea,” he remarks.
About the Author
